pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
27303412 | PMC4880568 | pmc | 3,400 | {
"abstract": "The isolation or engineering of algal cells synthesizing high levels of medium-chain fatty acids (MCFAs) is attractive to mitigate the high clouding point of longer chain fatty acids in algal based biodiesel. To develop a more informed understanding of MCFA synthesis in photosynthetic microorganisms, we isolated several algae from Great Salt Lake and screened this collection for MCFA accumulation to identify strains naturally accumulating high levels of MCFA. A diatom, Chaetoceros sp. GSL56, accumulated particularly high levels of C14 (up to 40%), with the majority of C14 fatty acids allocated in triacylglycerols. Using whole cell transcriptome sequencing and de novo assembly, putative genes encoding fatty acid synthesis enzymes were identified. Enzymes from this Chaetoceros sp. were expressed in the cyanobacterium Synechococcus sp. PCC 7002 to validate gene function and to determine whether eukaryotic enzymes putatively lacking bacterial evolutionary control mechanisms could be used to improve MCFA production in this promising production strain. Replacement of the Synechococcus 7002 native FabH with a Chaetoceros ketoacyl-ACP synthase III increased MCFA synthesis up to fivefold. The level of increase is dependent on promoter strength and culturing conditions.",
"introduction": "Introduction Biologically derived diesel from water-oxidizing, photosynthetic microorganisms (PSMs) is considered an efficient and promising next-generation technology for the production of renewable fuels ( Radakovits et al., 2010 ; Work et al., 2013 ). These photosynthetic organisms are capable of high photon conversion efficiencies and could be deployed so that their cultivation does not directly compete with the contemporary global food supply ( Schenk et al., 2008 ; Mata et al., 2010 ). To improve commercial feasibility, research efforts have focused on areas ranging from optimizing photosynthetic yields to performing scalability assessments ( Larkum et al., 2012 ; Ho et al., 2014 ; Quinn and Davis, 2015 ). These efforts can be combined with advances in next-generation DNA sequencing and metagenomic analysis to characterize the diversity of phototrophic life at the enzyme/molecular level, which facilitates the discovery of genetic “parts” that can be transformed into “chassis” organisms to genetically improve biotechnological phenotypes in production organisms ( Kodzius and Gojobori, 2015 ). While most contemporary research efforts are focused on improving biomass accumulation, relatively few efforts are targeting improved oil quality. However, lipid oil quality is a significant issue in diesel fuel utilization, as biological products typically have longer chain lengths and higher degrees of unsaturation relative to conventional petrodiesel ( Durrett et al., 2008 ). These features lead to poor cold-flow temperature properties (longer chain lengths) and oxidative instability (higher unsaturation) ( Knothe, 2005 , 2008 ). The genetic manipulation of fatty acid chain length in plants has succeeded in changing oil crops to synthesize more saturated, medium-chain fatty acids (MCFAs, C 8-14 ), which are preferable for biodiesel ( Voelker et al., 1992 ; Thelen and Ohlrogge, 2002 ). Expressing acyl-ACP thioesterases (acyl-ACP TEs) alone, or with ketoacyl-ACP synthase (KAS) from plants that produce MCFA, leads to MCFA accumulation in transgenic hosts ( Jones et al., 1995 ; Leonard et al., 1998 ; Slabaugh et al., 1998 ). However, there are relatively few successful reports regarding MCFA accumulation in PSMs ( Radakovits et al., 2011 ), and only limited information is available regarding mechanisms to increase MCFA synthesis in PSMs to produce a higher quality biofuel. Studies of TEs, the enzymes most commonly used to control fatty acid chain length, indicate differences between algal and plant TEs in terms of substrate recognition and phylogeny ( Jing et al., 2011 ; Blatti et al., 2012 ; Beld et al., 2014 ). In this study, we used bioprospecting, cell sorting, and fatty acid methyl ester (FAME) profiling to identify a diatom from Great Salt Lake (GSL), Utah ( Chaetoceros GSL56 hereafter) that naturally accumulated high levels of C 14:0 fatty acid (>16%), and initiated studies probing fatty acid synthesis (FAS) in this organism. We chose GSL as a sampling site because the growth of halophilic algae in high-salt production systems has the potential to yield strains that are productive in seawater raceways where evaporation leads to increased salt concentrations; and because unique enzymology of potential biotechnological relevance can be found in extremophiles isolated from hypersaline ecosystems ( Meuser et al., 2013 ). Chaetoceros GSL56 accumulates among the highest levels of C 14:0 in a PSM that we have observed to date. As an initial characterization of the MCFA accumulation phenotype, we probed the physiological parameters influencing C 14:0 accumulation in this alga, and attained and assembled a whole-cell transcriptome to begin probing FAS enzymes. To validate gene annotations and to search for superior FAS enzymes, a targeted set of genes encoding eukaryotic FAS enzymes from this diatom were transformed into a cyanobacterium [ Synechococcus sp. PCC 7002 (hereafter, Synechococcus 7002)] engineered to secrete fatty acids. We establish that a eukaryotic type III ketoacyl-ACP synthase (KASIII) enzyme can functionally replace the endogenous FAS enzyme FabH in Synechococcus 7002, and that expression of this non-native, eukaryotic enzyme improves MCFA yields under the culturing conditions used.",
"discussion": "Discussion Medium chain fatty acids are desirable in both the fuel and chemical industries ( Battey et al., 1989 ). However, large quantities of MCFAs are only available in some specific oilseeds, such as coconut and Cuphea . Expression of MCFA specific enzymes can lead to the accumulation of shorter chain fatty acids in crop plants, providing a way to increase MCFA productivity ( Voelker et al., 1992 ). In this study, we explored FAS in a halophilic alga ( Chaetoceros GSL56) that is enriched for MCFA, and used whole cell transcriptome sequencing to identify genes encoding putative FAS enzymes. Significantly, we identified a gene encoding a KASIII enzyme that when expressed in Synechococcus 7002 is able to functionally replace the native cyanobacterial FabH enzyme, and when co-expressed in a C 12:0 MCFA secreting strain results in increased levels of MCFA synthesis. Chaetoceros GSL56 primarily synthesizes C 14:0 , C 16:0 , C 18:0 , C 18:N , and C 20:5 fatty acids that are differentially incorporated into all lipid classes ( Table 1 ; Figures 1 and 2 ). Depending on growth conditions, saturated C 14:0 fatty acids range from ∼15–40% of all fatty acids in Chaetoceros GSL56 which is amongst the highest native C 14:0 levels that we have observed in an alga ( Figure 1A ). Additional marine algae have also been documented to contain MCFA, primarily within the Bacillariophyta and Haptophyta phyla ( Chen et al., 2007 ; Guihéneuf et al., 2010 ). Analysis of individual lipid classes in Chaetoceros GSL56 revealed that MCFAs are preferentially incorporated into TAG ( Figure 2 ; Supplementary Table S1), which is consistent with the MCFA profiles in plant seeds and the other diatom species ( Battey and Ohlrogge, 1989 ; Alonso et al., 2000 ; Chen et al., 2007 ). In contrast to neutral oil droplets, the phospholipids and galactolipids, which are used as membrane constituents, are enriched in long chain fatty acids (C 16-20 ). Although Chaetoceros GSL56 has a promising MCFA phenotype, this alga has poor growth metrics relative to biotechnologically relevant strains. We therefore identified enzymes in the MCFA biosynthetic pathway that could be used as “parts” in more biotechnologically promising strains. Synechococcus 7002 is a cyanobacterium with among the fastest growth rates of any PSM, and has become the recent focus of several biotechnology efforts ( McNeely et al., 2010 ; Mendez-Perez et al., 2011 ; Davies et al., 2014 ; Work et al., 2015 ). We therefore transferred selected components of the Chaetoceros GSL56 FAS machinery into Synechococcus 7002 to (i) verify functional annotations and (ii) determine whether enzymes from Chaetoceros GSL56 could be expressed in a cyanobacterium to increase MCFA synthesis. Based on precedence in plant studies, ( Voelker et al., 1992 ; Dehesh, 2001 ), we targeted substrate specific enzymes, such as acyl-ACP TEs and β-ketoacyl ACP synthase (KAS), to influence MCFA synthesis ( Voelker et al., 1992 ; Leonard et al., 1998 ). Initially, we probed whether five genes encoding putative thioester hydrolysis enzymes contained acyl-ACP TE activity that could be used to produce MCFAs in Synechococcus 7002. Expression of each of these genes individually while concurrently knocking out the gene encoding the fatty acid recycling enzyme FadD ( Kaczmarzyk and Fulda, 2010 ; Liu et al., 2011 ; Ruffing, 2014 ; Work et al., 2015 ), did not result in the production of any MCFA in Synechococcus 7002. This may be because the enzymes tested were not genuine acyl-ACP TEs, or because they did not fold into functional enzymes in Synechococcus 7002. We then targeted the KAS enzyme family. The β-ketoacyl ACP synthase III (KASIII) functionally replaced the FabH enzyme in Synechococcus 7002, confirming the annotation of KASIII as a genuine β-ketoacyl ACP synthase III. Furthermore, expression of KASIII in a Synechococcus 7002 strain that coexpresses a plant-sourced medium chain specific thioesterase ( UcfatB ) produced up to 40% of total fatty acid as lauric acid (SAK03), which is approximately four times more than corresponding strains expressing the native FabH enzyme. Enhanced lauric acid production in transgenic strains may be due to, (i) an increased acyl-ACP pool size since fabH is the rate-limiting enzyme of fatty acid synthase in Synechococcus 7002 ( Kuo and Khosla, 2014 ); (ii) modified overall FAS rate that better matches thioesterase activity ( Dehesh et al., 2001 ); and/or, (iii) concentrated short to medium chain acyl-ACPs pools caused by KASIII ( Abbadi et al., 2000 ). The exact mechanism of KASIII enhancement, and testing whether KASIII participates in MCFA synthesis in Chaetoceros GSL56, is the subject of further investigation. The euryhaline cyanobacterium Synechococcus 7002, is a particularly promising host for FFA production, because it shows a unique high tolerance to FFAs ( Ruffing, 2014 ). However, the total metabolic flux to FAS represents a small portion (∼5–10%) of cell dry weight, indicating evolutionary limitations to fatty acid productivities in this host ( Work et al., 2015 ). Replacing FabH, which is the kinetically rate-limiting enzyme in FAS, with KASIII did not improve fatty acid production when transformed into the wild-type host (data not shown), suggesting that downstream elements that traffic fatty acids to membrane lipids may now be rate limiting. However, the expression of acyl-ACP thioesterase allows cleavage of MCFA from acyl-ACP, and has been reported to increase malonyl-ACP turnover rates and reduce the inhibition of acyl-ACP in other systems ( Magnuson et al., 1993 ). Transgenic strains expressing KASIII and thioesterase (SAK01 and SAK03) accumulated higher amounts of C 12:0 and C 14:0 fatty acids, relative to their parental strain SA01, and these increases are consistent with UcfatB hydrolytic activity ( Voelker and Davies, 1994 ). Levels of MCFA are influenced by promoter strength, indicating a correlation between KASIII expression levels and MCFA production. It is still undetermined how Chaetoceros GSL56 regulates C 14 fatty acid production as we were unable to identify an acyl-ACP TE with C 14:0 specificity in this alga, which is consistent with other studies in red algae and diatoms ( Beld et al., 2014 ). Future studies of the role of KASIII on medium chain FAS, as well as the substrate specificities of other enzymes, such as acyltransferase, will be important in understanding fatty acid chain length regulation in this and other diatoms. In future research, we also intend to explore whether, the expression of other eukaryotic FAS enzymes in Synechococcus 7002, which typically has less than 10% of its biomass in fatty acyl lipids, results in further yield improvements. It is possible that these eukaryotic enzymes, which did not evolve under the regulatory mechanisms used by cyanobacteria, remain active outside of the metabolic context that evolved to limit FAS in cyanobacteria such as Synechococcus 7002."
} | 3,156 |
26974339 | null | s2 | 3,402 | {
"abstract": "Polyhydroxyalkanoates (PHAs) are carbon and energy storage polymers produced by a variety of microbial organisms under nutrient-limited conditions. They have been considered as an environmentally friendly alternative to oil-based plastics due to their renewability, versatility, and biodegradability. PHA synthase (PhaC) plays a central role in PHA biosynthesis, in which its activity and substrate specificity are major factors in determining the productivity and properties of the produced polymers. However, the effects of modifying the substrate side chain are not well understood because of the difficulty to accessing the desired analogues. In this report, a series of 3-(R)-hydroxyacyl coenzyme A (HACoA) analogues were synthesized and tested with class I synthases from Chromobacterium sp. USM2 (PhaCCs and A479S-PhaCCs) and Caulobacter crescentus (PhaCCc) as well as class III synthase from Allochromatium vinosum (PhaECAv). It was found that, while different PHA synthases displayed distinct preference with regard to the length of the alkyl side chains, they could withstand moderate side chain modifications such as terminal unsaturated bonds and the azide group. Specifically, the specific activity of PhaCCs toward propynyl analogue (HHxyCoA) was only 5-fold less than that toward the classical substrate HBCoA. The catalytic efficiency (kcat/Km) of PhaECAv toward azide analogue (HABCoA) was determined to be 2.86 × 10(5) M(-1) s(-1), which was 6.2% of the value of HBCoA (4.62 × 10(6) M(-1) s(-1)) measured in the presence of bovine serum albumin (BSA). These side chain modifications may be employed to introduce new material functions to PHAs as well as to study PHA biogenesis via click-chemistry, in which the latter remains unknown and is important for metabolic engineering to produce PHAs economically."
} | 456 |
35634798 | null | s2 | 3,404 | {
"abstract": "Genetically engineered silk-elastin-like-proteins (SELPs) synthesized with the combination of silk and elastin domains are bioengineered to also contain a graphene oxide (GO) binding domain. The conductivity and mechanical stability of graphene, combined with SELP-specific graphene interfaces are pursued as dynamic hybrid materials, toward biomaterial-based electronic switches. The resulting bioengineered proteins with added GO demonstrate cytocompatibility and conductivity that could be modulated by changing hydrogel size in response to temperature due to the SELP chemistry. Upon increased temperature, the gels coalesce and contract, providing sufficient condensed spacing to facilitate conductivity via the graphene domains, a feature that is lost at lower temperatures with the more expanded hydrogels. This thermally induced contraction-expansion is reversible and cyclable, providing an \"on-off\" conductive switch driven by temperature-driven hydrogel shape-change."
} | 244 |
39509456 | PMC11575764 | pmc | 3,406 | {
"abstract": "Previous work shows that a host’s resident microbial community can provide resistance against an invading pathogen. However, this community is continuously changing over time due to adaptive mutations, and how these changes affect the invasion resistance of these communities remains poorly understood. To address this knowledge gap, we used an experimental evolution approach in synthetic communities of Escherichia coli and Salmonella Typhimurium to investigate how the invasion resistance of this community against a bacterium expressing a virulent phenotype, i.e., colicin secretion, changes over time. We show that evolved communities accumulate mutations in genes involved in carbon metabolism and motility, while simultaneously becoming less resistant to invasion. By investigating two-species competitions and generating a three-species competition model, we show that this outcome is dependent on the strength of interspecies interactions. Our study demonstrates how adaptive changes in microbial communities can make them more prone to the detrimental effects of an invading species.",
"introduction": "Introduction An important factor determining the outcome of host-pathogen interactions is the interaction between the host’s residential microbial community and the invading pathogen. Several studies investigating these interactions in different host species have shown them to either protect against [ 1 – 5 ] or facilitate a pathogen’s invasion [ 6 , 7 ]. Mechanisms that protect the host include resource competition, colonization resistance, and modulation of the host’s immune response, while mechanisms that facilitate invasion include the secretion of metabolites that can allow the pathogen to initiate infections in the host. Building on this knowledge, research is also being conducted to identify and characterize synthetic microbial communities that could protect a host against specific pathogens [ 8 – 10 ], which potentially could be utilized in precision medicine and therapeutics. Interestingly, the outcome of these interactions is not only dictated by the identity of the species present in these communities but also by other factors such as the secretion of specific molecules or species abundances [ 11 , 12 ]. Given the importance of these interactions to a host’s health, it is thus important to determine the underlying mechanisms, both physiological and ecological, that can affect the outcome of the competition between the host’s residential microbial community and a pathogen. To overcome the invasion resistance posed by the host’s residential microbial community, several pathogens have been shown to secrete various toxins. For example, using bacterial genome-wide association studies and high-throughput phenotyping, it was suggested that the enteric pathogen Shigella sonnei outcompetes the gut commensal Escherichia coli by secreting colicins, which are proteinaceous toxins that specifically target E . coli strains [ 13 ]. The secretion of such toxins has been hypothesized to result in the epidemiological success of this pathogen. Similarly, an analysis of 1,181 E . coli strains of fecal origin showed a strong positive correlation between bacteriocin-producing genes and virulence factors, suggesting the role of these toxins during the pathogen invasion [ 14 ]. This correlation was specifically observed for ExPEc (extraintestinal) pathogenic E . coli strains. In another study [ 15 ], Salmonella Dublin, which is a cattle-adapted pathogen, was shown to carry 2 different Type VI secretion systems that were involved in interbacterial competition. These studies have given us valuable insights into how pathogens can dominate in competition against the host’s residential microbial community by secreting specific toxins. On the other hand, the host’s resident microbial community is also continuously changing with several studies demonstrating adaptive evolution in these communities [ 16 , 17 ]. These studies have demonstrated how these evolutionary changes can be observed over different time scales [ 18 ] and can affect important resource utilization pathways [ 19 ]. Importantly, these adaptive changes can in turn affect the community’s resistance against an invading microbial species. Thus, increased productivity of a microbial community, either through faster growth rates or increased cell density of its community members, results in a lower likelihood of invasion by a microbial species due to lower nutrient availability for the invader [ 20 ]. Another study showed that the joint evolution of microbial communities growing in a toxic medium increased invasion resistance [ 21 ]. Although the mechanism underlying this is unclear, the authors hypothesized that this observation could either be due to resource specialization through niche-partitioning or due to increased productivity of the evolved community members, both of which would result in fewer nutrients for the invading microbial species. Although these studies provide an important framework to investigate the relationship between an evolving community and invasion by a microbial species, whether or not these rules are applicable during an invasion by a toxin-secreting pathogen remains an outstanding question. To answer this question systematically, we employed an experimental evolution approach in synthetic microbial communities and determined how adaptive changes in these communities affect the invasion by a toxin-secreting bacteria. Specifically, we determined the rate of invasion of a colicin-producing natural isolate of E . coli (ECOR 11, henceforth designated as colicin-producing E . coli ) in ancestral and evolved synthetic two-species communities of Escherichia coli (henceforth mentioned as E . coli ) and Salmonella enterica serovar Typhimurium LT2 (henceforth designated as S . Typhimurium). Colicins are a commonly occurring class of proteinaceous polymorphic toxins that specifically target E . coli [ 22 ]. Importantly, colicin secretion by the pathogen has been shown to contribute to its success during an invasion [ 13 , 14 , 23 ]. This setup allowed us to mechanistically link the success of an invasion due to a virulent phenotype to adaptive changes in a microbial community. Our results demonstrate strong eco-evolutionary dynamics influencing the success of the colicin-producing E . coli strain’s invasion, where we observed a higher rate of invasion in evolved communities as compared to the ancestral communities. Two-species competition experiments and whole-genome sequencing analysis of evolved clones demonstrate that this outcome was observed due to increased resource utilization and ecological specialization in the evolved communities.",
"discussion": "Discussion Several studies have demonstrated the importance of the host’s residential microbial community against pathogen invasion [ 26 , 27 ], but how evolutionary changes within a microbial community can affect its interaction with invading pathogens is lacking. Our study addresses this by determining how invasion by a colicin-producing E . coli is affected by adaptive changes in synthetic microbial communities. We used a colicin-producing E . coli in our experiments because colicin production is a widespread phenotype that can contribute to the invasion by several pathogens. Our results show that communities that have coexisted for longer periods and have accumulated adaptive changes are more prone to invasion by a colicin-producing bacterium ( Fig 5 ). However, this is opposite to what is expected from existing ecological theory [ 20 , 28 , 29 ], which states that evolved communities demonstrate increased resistance to invasion. Our experimental data and mathematical simulations show that these observations are an outcome of a complex combination of evolutionary and ecological changes, based on increased resource consumption abilities and ecological specialization in these communities ( Fig 5 ). Additionally, previous studies have not investigated the invasion due to a pathogen-specific phenotype (i.e., toxin production) in a microbial community but have focused on replacement through resource competition, which may explain the different results between our and previous studies. 10.1371/journal.pbio.3002889.g005 Fig 5 An increase in resource competition within a community results in a decrease in invasion resistance. A schematic representation of the interplay between resource competition and invasion resistance in microbial communities. As the strength of resource competition increases within the community, the resistance to a toxin-producing invader reduces. Furthermore, these invasion dynamics are not predictable based on two-way competition outcomes between the community members and the invader. Importantly, our study highlights the interplay between interspecies resource competition and invasion dynamics by a toxin-secreting bacterium. Although both of these occurrences are common in microbial communities across different ecosystems [ 30 – 32 ], the interplay between them is rarely investigated. Our study shows that the interaction between these features results in unintuitive dynamics. Thus, invasion resistance is lower when the community members that are not affected by the toxin evolve to have differential resource utilization abilities. In our experiments, this was observed when S . Typhimurium which was isolated from the evolved communities had increased in fitness compared to the ancestral S . Typhimurium, had a higher fitness against the cognate E . coli strains (as compared to the fitness of the ancestral S . Typhimurium against the ancestral E . coli ), and demonstrated reduced competition against the invading colicin-producing E . coli (as compared to the ancestral S . Typhimurium). These observations could either be due to more efficient usage of resources or due to the use of different resources, both of which are likely outcomes in the complex growth media used in our study. We further corroborated the dependence of invasion resistance on the strength of interspecies competition using a Lotka–Volterra interspecies competition model. In this model, we observed that invasion resistance to the colicin-producing E . coli decreases as the strength of interspecific competition between S . Typhimurium and E . coli increases. Furthermore, this effect becomes more pronounced with increasing ecological specialization of S . Typhimurium in these communities, which was achieved in our simulations by reducing the effect of S . Typhimurium on the growth of the invading colicin-producing E . coli (as was observed in our experiments). Interestingly, in our synthetic communities, interspecific competition (measured by the increased resource overlap and fitness measurements of S . Typhimurium with respect to susceptible E . coli ) and ecologically specialization (measured by fitness measurements of colicin-producing E . coli with respect to S . Typhimurium) both increased over time. A possible explanation for these observations is that in these synthetic communities, the joint evolution of E . coli and S . Typhimurium results in both species becoming better adapted to utilizing common resources thus increasing interspecific competition. At the same time, the preferred usage of these common resources would result in other resources being available for any incoming invading species, thus resulting in the resident species being more ecologically specialized in relation to any incoming species. Although our results and model predictions are suggestive of these conclusions, more work is needed to understand the mechanisms and nature of ecological specialization of E . coli and S . Typhimurium in these communities. Whole-genome sequencing revealed that all the mutations in the evolved S . Typhimurium clones were in genes that contributed to motility, i.e., motA , cheR , fimH , and fljB . The evolved clones harboring these mutations were also found to be more resource-specialized, as indicated by the reduced competition between the invading E . coli and S . Typhimurium strains ( Fig 3B ). Furthermore, evolved clones of E . coli , had mutations in genes that were involved in carbon metabolism and in cell-adhesion phenotypes. Although these results suggest the relevance of these phenotypes in contributing to the community dynamics we observe in our experiments, future experiments involving genetic reconstruction of these mutations in clean ancestral backgrounds are needed to conclusively establish a causal link between these mutations and resource utilization. Interestingly, we also observed an opposite trend between invasion dynamics in two-species (toxin-susceptible and toxin-producing bacterium) versus three-species system (toxin-susceptible, toxin-producing bacterium, and toxin-resistant competitor). Thus, the ancestral E . coli was outcompeted by colicin-producing E . coli at a faster rate compared to the evolved E . coli . However, in a community consisting of ancestral E . coli and S . Typhimurium invasion by a colicin-producing E . coli was slower than in a community consisting of jointly evolved E . coli and S. Typhimurium clones. The differences between the two-species and three-species dynamics highlight the importance of understanding the mechanisms driving these interactions to make better predictions about invasion dynamics in microbial communities. Connecting ecological theory with evolutionary outcomes for medically relevant phenotypes is an important step in making predictions about the evolution of these phenotypes. This study takes a step in this direction by using synthetic microbial communities, where invasion dynamics become predictable based on phenotypes of resource competition and toxin production. Disentangling the interactions between different members of the community further demonstrated why our observations are opposite to what one expects from the ecological theory of invasion, highlighting the need to understand the mechanisms that dictate these dynamics."
} | 3,531 |
39624713 | PMC11609156 | pmc | 3,407 | {
"abstract": "Soil amendments, including various types of fertilizers, are often used to control the uptake of heavy metals such as cadmium in cropping fields. The influence of these amendments on other members of the agroecosystem, such as arbuscular mycorrhizal fungi (AMF), remains less well investigated. Here, we established an experiment with the application of woody peat organic fertilizer and phosphate rock powder to examine its effects on AMF communities in two cadmium-contaminated vegetable crop fields (cucumber and pepper). We found that the application of phosphate rock powder enhanced soil phosphorus content, while the application of woody peat organic fertilizer enhanced soil nitrogen content, but neither influenced AMF abundance. We also found little influence of either amendment on measures of AMF diversity, except in one case where the Shannon index of diversity was lower in pepper fields amended with phosphate rock powder. We did, however, find significant shifts in the community composition and relative abundances of AMF taxa in the two vegetable fields, primarily as a result of shifts in the soil pH and nitrogen content.",
"conclusion": "5 Conclusion We examined the response of AMF communities to the application of soil amendments intended to allow safer utilization of crops in Cd contaminated vegetable fields. We found that in two types of Cd-contaminated vegetable fields (peppers and cucumbers), the AMF community composition and structure changed with the addition of soil amendments, but no change in the overall abundance. Interestingly, this response was primarily associated with changes in soil nitrogen and the nitrogen-to-phosphorus ratio, rather than phosphorus itself, reflecting the important role of the absolute levels of nutrients in regulating AMF communities. Additionally, we preliminarily identified certain AMF taxa that were more responsive to soil amendments. These findings provide valuable insights into exploring and identifying more efficient AMF strains for Cd remediation. However, further research is needed to elucidate the mechanisms underlying these responses and determine whether the selected taxa play specific roles in regulating plant heavy metal uptake. Long-term field trials across a wider range of crops and soil conditions are crucial to ensure the scalability and reliability of these strategies for broader agricultural applications.",
"introduction": "1 Introduction As a result of ongoing anthropogenic activities, agricultural soils are being continually degraded ( Durdu et al., 2023 ). Notably, heavy metals are accumulating in agricultural soils globally because of increased uses of pesticides and fertilizers, as well as other industrial actions ( Rashid et al., 2023 ). This heavy metal contamination can negatively influence crop production through a number of pathways, and can also influence human health. For example, cadmium (Cd) is a non-essential element for plant growth, but owing to its high chemical activity in soil, can be easily absorbed by plants and accumulate within plant tissues, ultimately posing risks to human health through consumption ( McLaughlin et al., 2021 ). Levels of Cd have increased in agricultural soils in recent decades ( Six and Smolders, 2014 ; Shi et al., 2019 ), and now reducing the Cd content and activity in soils, as well as decreasing its absorption and accumulation by plants, has become a priority for safe production of edible agricultural products ( Mubeen et al., 2023 ). For lightly and moderately polluted soils, a number of soil amendments, including organic material and phosphate, have been commonly applied to reduce the bioavailability of heavy metals in the soil through the mechanisms of adsorption, complexation and precipitation ( Gao J. et al., 2023 ; Wang F. et al., 2022 ; Cui et al., 2023 ). While there has been work to understand the impact of these activities on farmland soil, such as the positive effects of reducing plant heavy metal content and crop yield ( Feng and Cheng, 2023 ), less attention has been paid to the influence of these amendments on environmental and ecological risks, such as the impact of these measures on soil microbial community structure and their functions ( Holland et al., 2018 ). For example, how these amendments might influence crop interactions with soil microbes, such as arbuscular mycorrhizal fungi (AMF) that form symbiotic relationships with over 70% of terrestrial plants ( Brundrett and Tedersoo, 2018 ), obtaining carbon from plants in exchange for helping plants to absorb mineral nutrients and water ( Smith and Smith, 2011 ). In agriculture, AMF can promote crop yields while reducing the application of chemical fertilizers, controlling disease occurrence, and enhancing the host plant’s adaptability to drought, salinity and heavy metal pollution ( Hao et al., 2022 ). AMF can also reduce the uptake and transport of Cd in crops through various mechanisms, including regulating plant and root growth, binding heavy metals with root and hyphal exudates, increasing antioxidant enzyme activities, interacting with soil beneficial microorganisms, and immobilizing heavy metals via extraradical hyphae as well as their improvement on soil aggregation ( Wang, 2017 ; Wang H. et al., 2022 ; Moreno Jiménez et al., 2024 ). Studies on the application of AMF inoculant, combined with other remediation strategies, to mitigate Cd contamination in soils has shown promising results for various crops ( Cakmak et al., 2023 ; Zulfiqar et al., 2023 ; Liu et al., 2024 ). Given that AMF are widespread soil fungi in agricultural ecosystems, native AMF communities may also significantly influence the Cd remediation efficacy of soil amendments. Therefore, evaluating the interactions between soil amendments and native AMF communities is essential when developing strategies to mitigate Cd contamination. Findings from studies examining the effects of soil environmental changes on AMF under conditions that do not involve or explicitly consider Cd contamination remain inconclusive and highly context-dependent ( van der Heyde et al., 2017 ; Mohamed et al., 2023 ). For example, some studies report that AMF abundance are enhanced by organic amendments or reduced by excessive phosphate, while others find more nuanced or contrasting outcomes ( Treseder, 2004 ; Park et al., 2024 ). The diversity of outcomes highlights the complexity of AMF interactions with soil environments and the necessity for more targeted research that considers the specific conditions under which AMF are studied. This includes exploring how soil amendments can influence the structure and function of AMF communities in the context of Cd contamination. Here, we used an experimental approach to explicitly investigate the effects of soil amendments on the AMF community in Cd contaminated vegetable fields. We examined changes in both soil chemistry and the AMF community, and their associations, to understand the primary factors influencing AMF community. We used two types of vegetable crops, pepper ( Capsicum annuum ) and cucumber ( Cucumis sativus ), and two types of soil amendments (phosphate rock powder and woody peat organic fertilizer). In all, our research will provide insights into the responses, protection and potential application of AMF communities in farmland soils.",
"discussion": "4 Discussion Overall, it was not surprising that the application of organic fertilizer and phosphate rock powder as soil amendments led to significant increases in soil nitrogen and phosphorus levels ( Table 1 ). The increase in soil nutrients could allow plants to rely more on their own root system for nutrient acquisition, reducing their dependence on AMF ( Treseder, 2004 ; Singh et al., 2022 ) because the carbon costs of AMF associations may outweigh the nutritional benefits they provide ( Johnson, 2010 ). However, despite the significant negative correlation between soil nitrogen and AMF abundance, we found no influence of the application of soil amendments on mycorrhizal colonization rates and hyphal length density ( Figure 1 ). This is likely because the ambient levels of nitrogen and phosphorus are high in the soil, suggesting that nutrient acquisition may not be a primary role of AMF in symbiosis with plants. Indeed, the positive impacts of AMF on soil aggregation, water uptake and transport, and disease resistance can also play an important role in the plant-AMF symbiosis, but be less influenced by nutrient availability ( Delavaux et al., 2017 ). Moreover, the relatively high ambient total phosphorus and organic carbon in the soil may have mitigated their responses to the common amounts of amendment addition (7% of total phosphorus and 2% of total organic carbon). This could partly explain the lack of significant changes in AMF abundance. Despite the lack of response of AMF abundance to soil amendments, we did find lower Shannon diversity (but not richness) of AMF communities after applying phosphate rock powder. The adverse effects of nutrient additions on AMF diversity are consistent with results in many other ecosystems, including grasslands ( Santos et al., 2006 ), alpine meadows ( Liu et al., 2012 ), tropical rainforests ( Camenzind et al., 2014 ), and other agroecosystems ( Lin et al., 2012 ; Gosling et al., 2013 ). This is also consistent with the NMDS ordination results which indicated that applying phosphate rock powder led to a phylogenetically more dispersed AMF community ( Figure 5 ). This community is more likely to be structured by competition, permitting taxa with various traits to respond to nutrient variations in different ways ( Webb et al., 2002 ). However, the lack of response of AMF diversity to the nitrogen inputs from organic fertilizer may have been because the differential responses of AMF taxa to nitrogen and phosphorus addition ( Camenzind et al., 2014 ). Even when there was no influence of amendments on abundance or richness of AMF, we did consistently find shifts in the AMF community structure among treatments ( Figure 5 ). Correspondingly, at both the family and OTU levels, there were notable changes in the relative abundance of dominant groups after the application of soil amendments, leading to increases of some taxa and decreases of others ( Figures 3 , 4 ). Different AMF taxa vary in their allocation to the intraradical structures that obtain carbon from plant roots and extra-radical hyphae that are responsible for obtaining nutrients from the soil ( Maherali and Klironomos, 2012 ). As a result, taxa with a higher proportion of intraradical allocation (such as the Glomeraceae) tend to be more competitive in carbon resource acquisition ( Chagnon et al., 2013 ), and thus tend to be more dominant. However, while we found that Glomeraceae dominated in the control treatment, their dominance diminished with soil amendments, while the abundance of Gigasporaceae, which has a greater advantage in acquiring soil nutrient resources, increased ( Figure 3 ). This suggests that processes other than the availability of soil nutrients more likely regulates AMF community structure. Furthermore, it is possible that the effects of soil amendments on certain AMF groups may become more pronounced over an extended period. Due to the constraints imposed by the vegetable cropping system, the duration of our study was limited to 4 months, which represents one of the key limitations of this research. Although we cannot full disentangle mechanisms leading the observed shifts in AMF community composition with the application of organic fertilizer, the declines in soil pH, observed specifically in pepper fields, may play an important role. Soil pH regulates both nutrient availability ( Hartemink and Barrow, 2023 ) and AMF growth and reproduction ( Wang et al., 1993 ; Van Aarle et al., 2002 ). As a result, AMF taxa with variation in nutrient and growth traits ( Hart and Reader, 2002 ; Koch et al., 2017 ) can shift in their relative abundances under different levels of soil pH ( Davison et al., 2021 ; Bodenhausen et al., 2023 ), such that soil pH may be a key driver of changes in AMF community composition ( Figures 5 , 6 ). A lack of pH change under organic fertilizer addition in cucumber fields suggests that crop-specific factors such as root exudates, nutrient uptake patterns, or differential microbial activity might mediate the soil’s response to amendments ( Vives-Peris et al., 2020 ; Baggs et al., 2023 ; Shu et al., 2023 ). This discrepancy highlights the complexity of plant–soil-microbe interactions and underscores the importance of considering crop-specific responses when evaluating the effects of soil amendments. At our study, not only was the ambient nitrogen and phosphorus content in the soil high, but so was the content of Cd (average of 0.7 mg/kg), leading to high risk for Cd pollution ( Chinese National Standardization Administration, 2018 ). Although the response of AMF to variation in Cd content is mixed ( Weissenhorn et al., 1993 ; Krishnamoorthy et al., 2015 ), we found no evidence for an influence of Cd on AMF abundance. However, we did find an influence of Cd on the composition and structure of AMF communities ( Figures 5 , 6 ). While we cannot explicitly disentangle the mechanisms underlying this effect, it is likely related to the differential responses of different AMF taxa to Cd ( Gai et al., 2018 ; Chen et al., 2023 ). For example, the growth of Gigasporaceae family was markedly inhibited by Cd exposure ( Figure 6 ). Existing evidence suggests that AMF communities in low P or high nitrogen-to-phosphorus ratio environments tend to favor Gigasporaceae species likely due to their competitive life history traits ( Johnson et al., 2003 ; Chagnon et al., 2013 ). This is consistent with the significant increase in their abundance following the addition of organic matter in this study, whereas their response to phosphate addition was negligible ( Figure 3 ). These findings suggest that the response of Gigasporaceae to varying resource inputs may be largely unaffected by the inhibitory effects of Cd, indicating a potential resilience of this family to heavy metal stress in specific environmental contexts. While the two soil amendments did not significantly influence the total Cd concentration in the soil, they have the potential to mitigate plant Cd uptake by decreasing Cd bioavailability ( Wang F. et al., 2022 ). However, the absence of data on variations in available Cd content constrains our ability to comprehensively assess the extent to which these amendments may have impacted AMF abundance and community composition through the modulation of Cd bioavailability. The compositional shifts of AMF communities we observed indicate that the use of soil amendments can have certain ecological impacts and/or risks ( van der Heijden et al., 1998 ; Rillig and Mummey, 2006 ; Hawkins et al., 2023 ). Specifically, the increased abundance of Gigasporaceae in response to organic matter addition may enhance nutrient use efficiency in crops, improve soil aggregation, and boost carbon sequestration due to their significant investment in extraradical hyphae ( Hart and Reader, 2002 ; Chagnon et al., 2013 ). In contrast, a decline in Glomeraceae could increase the risk of excessive Cd accumulation in agricultural products, as they generally demonstrate higher remediation efficiency than Gigasporaceae ( Ferrol et al., 2009 ). Given the significant gaps in our understanding of the functional diversity of AMF communities, it remains challenging to predict the specific consequences of these changes ( Wang and Rengel, 2024 ; Martin and van der Heijden, 2024 ) and how these changes in AMF communities will influence production in heavy metal-contaminated farmland. Nevertheless, the application of AMF inoculants on Cd contaminated farmland soils is considered an effective strategy to reduce plant Cd uptake and enhance crop yields ( Liu et al., 2018 ; Gao Y. et al., 2023 ). Therefore, the practical application of our study lies in our ability to select representative taxa with differential responses to soil amendments, then explore their roles in reducing plant Cd uptake, increasing plant yields and improving soil aggregation, and to ultimately identify fungal strains with potential for widespread application. However, we did find that AMF responses to soil amendments differed between the two different kinds of crops in our study, underscoring the need for considering crop-dependent strategies for AMF inoculation in agricultural soils, where targeted AMF inoculation could be tested for its efficacy in reducing Cd uptake and ensuring food safety."
} | 4,163 |
25028968 | PMC4152358 | pmc | 3,409 | {
"abstract": "Amyloid curli fibers and cellulose are extracellular matrix components produced in the stationary phase top layer of E. coli macrocolonies, which confer physical protection, strong cohesion, elasticity, and wrinkled morphology to these biofilms. Curli and cellulose synthesis is controlled by a three-level transcription factor (TF) cascade with the RpoS sigma subunit of RNA polymerase at the top, the MerR-like TF MlrA, and the biofilm regulator CsgD, with two c-di-GMP control modules acting as key switching devices. Additional signal input and fine-tuning is provided by an entire series of small RNAs—ArcZ, DsrA, RprA, McaS, OmrA/OmrB, GcvB, and RydC—that differentially control all three TF modules by direct mRNA interaction. This review not only summarizes the mechanisms of action of these sRNAs, but also addresses the question of how these sRNAs and the regulators they target contribute to building the intriguing three-dimensional microarchitecture and macromorphology of these biofilms.",
"conclusion": "Conclusions and Perspectives In E. coli macrocolony biofilms, amyloid curli fibers and cellulose are the matrix components that confer physical protection to cells and generate strong cohesion, elasticity, and the striking macroscopic morphology of these biofilms ( Fig. 1 ). Their production in the stationary phase top layer of macrocolonies is controlled by a cascade of three transcription factors (TF; Fig. 3 ), i.e., RpoS (I), YdaM/MlrA (II), and CsgD (III). This cascade is complemented by two c-di-GMP switch modules (A and B in Fig. 3 ), which trigger signal propagation between TF modules I and II and differential control of curli and cellulose production at the very end of the cascade, respectively ( Fig. 3 ). As detailed in this review, at least eight sRNAs have a crucial input into all three TF modules, and thereby, connect these modules to a wide range of environmental signals. A question remaining for future research is whether even more sRNAs may have an impact in this intriguing network. Even though most of what we currently know about the regulation of and by sRNAs is based on research with liquid or planktonic cultures, this knowledge has implications for biofilm growth and morphogenesis that deserve further studies, in particular, in situ in macrocolonies. ArcZ and McaS expression responds to signals that in fact reflect the long range gradients of oxygen, energy, and carbon sources in a macrocolony and—by inversely regulating FlhDC and RpoS/CsgD—are highly likely to contribute to setting up the two physiologically distinct layers with flagella-producing growing cells in the bottom layer and outer colony edges and matrix-producing stationary phase cells in the top layer. 14 Other sRNAs integrate stress signals such as cell envelope imbalances, increased osmolarity, low temperature, acidification, or oxidative stress that further shape macrocolony biofilms, but also contribute to sudden acute stress reactions which may be more relevant for “free-living” planktonic cells. Thus, RprA and OmrA/OmrB could be key regulators in making the decision whether to get immobilized in a biofilm matrix or staying potentially motile, because in the curli/cellulose control cascade, they can still downregulate CsgD-driven matrix production under stationary phase conditions, where RpoS is already fully active and the YciR-controlled c-di-GMP switch has been thrown to “ON” (see Fig. 3 ). By interfering with both flagella and curli production, OmrA/OmrB and OxyS may actually act as stress-induced “emergency” sRNAs, which in general inhibit production of large cellular appendices, and thereby, allow cells to focus their resources on coping with acute stress effects (for a more detailed discussion of these complex regulatory patterns and their physiological implications, see ref. 46 ). The set of sRNAs in the curli/cellulose control cascade also provides an ideal opportunity to study the sometimes-surprising behavior of RNA-based networks, in which an sRNA can bind to multiple mRNAs, which in turn, bind additional sRNAs ( Fig. 3 ). In contrast to classical hierarchical and amplificatory transcription factor networks, such a sRNA/mRNA network operates in a non-hierarchical and stoichiometric mode in which sRNAs “target” mRNAs as well as vice versa. The output of this network crucially depends on competition and titration reactions that are determined by the rates of synthesis and cellular levels of sRNAs, mRNAs, and Hfq, and the affinities between all these players, which generates threshold-linear responses and priorities in cross-talk within the sRNA/mRNA network. 46 , 131 In the curli/cellulose control cascade, this means that csgD mRNA is not only “targeted” by several sRNAs (which bind with different affinities to overlapping sites on csgD mRNA), but that strong accumulation of csgD mRNA—as e.g., during entry into stationary phase—can also scavenge sRNAs, and thereby, indirectly affect the function of their other target mRNAs. 93 Thus, effects of dynamic changes in sRNAs as well as distinct “hub” mRNAs can “horizontally” propagate through the network, which can be a challenge for the interpretation of genetic analyses, which in principle rely on hierarchical “one-way” causal relationships. Another point of interest for future analysis is the heterogenous expression of CsgD in the transition zone between the two physiological layers of a macrocolony biofilm, which features matrix-surrounded CsgD ON cells right next to “naked” CsgD OFF cells. 8 As these cells are exposed to the same environmental conditions, this heterogeneity may reflect bistability in the control of CsgD. The underlying control network ( Fig. 3 ) indeed features several double-negative feedback loops, i.e., a potentially bistability-generating motif. One is the above-mentioned non-hierarchical mutual inhibition of csgD mRNA and sRNAs. The other double negative feedback loops surround the trigger enzyme and PDE YciR, i.e., the key switch protein in the c-di-GMP module ( Fig. 3 ). It will be a challenge for future work to sort out, how these motifs contribute to the apparent bistable expression of CsgD, and therefore, to heterogenous matrix production in distinct zones of the macrocolony biofilm. Finally, the intricate interplay between c-di-GMP-mediated control and the sRNA network in curli/cellulose regulation suggests that searching for additional regulatory connections between nucleotide second messenger signaling and sRNAs may yield interesting new insights.",
"introduction": "Introduction Bacterial biofilms are multicellular communities in which cells are surrounded by a self-produced extracellular matrix of polymeric substances that can include exopolysaccharides, proteins, amyloid fibers, and exo-DNA. Matrix components allow cells to attach to each other and mediate adherence to biotic or abiotic surfaces. Bacteria inhabiting a biofilm are protected from physical stress, antimicrobials, and the host immune system, and thereby cause severe medical, environmental and technical problems. 1 - 3 Biofilm formation and architecture is controlled by complex regulatory networks that integrate multiple environmental signals at the transcriptional as well as post-transcriptional levels by using alternative sigma factors, two-component systems, small regulatory RNAs (sRNA), and second messengers (in particular c-di-GMP). In this review, we highlight the multiple roles of several sRNAs in controlling the expression of the stationary phase sigma factor RpoS (σ S ), the c-di-GMP generating diguanylate cyclase YdaM and the transcription factor CsgD, which are key factors in the regulatory network that controls biofilm matrix production in the model bacterium Escherichia coli . By focusing on highly structured macrocolony biofilms, we also address the question how these sRNAs and the regulators they target contribute to building the intriguing three-dimensional architecture and macromorphology of these biofilms ( Fig. 1 ). Figure 1. Morphology and physiological stratification of E. coli macrocolony biofilms. Macrocolonies of the curli- and cellulose-producing E. coli K-12 strain AR3110 ( A ) and the curli-producing, but cellulose-deficient original K-12 strain W3110 ( C ) were grown for 5 d at 28 °C on salt-free LB plates containing the amyloid dye Congo red. AR3110 is a direct derivative of W3110, in which a “domesticating” point mutation that generated a stop codon in the cellulose synthase operon was '”repaired.” 12 Cryosections through macrocolonies of AR3110 ( B ) and W3110 ( D ) grown for 5 d at 28 °C on salt-free LB supplemented with thioflavin S (TS) are shown, with brightfield and fluorescence microscopic images merged. TS is an amyloid-staining dye that binds to both curli fibers and cellulose, which are produced under the control of RpoS in the upper layer of starving cells in the macrocolonies. Fluorescence images were false-colored green for TS. The figure illustrates previously described results, 12 but only ( D ) is a section of an image actually published in the previous study and is shown here with permission."
} | 2,284 |
29207531 | PMC5750922 | pmc | 3,410 | {
"abstract": "Environmental pollution from hazardous waste materials, organic pollutants and heavy metals, has adversely affected the natural ecosystem to the detriment of man. These pollutants arise from anthropogenic sources as well as natural disasters such as hurricanes and volcanic eruptions. Toxic metals could accumulate in agricultural soils and get into the food chain, thereby becoming a major threat to food security. Conventional and physical methods are expensive and not effective in areas with low metal toxicity. Bioremediation is therefore an eco-friendly and efficient method of reclaiming environments contaminated with heavy metals by making use of the inherent biological mechanisms of microorganisms and plants to eradicate hazardous contaminants. This review discusses the toxic effects of heavy metal pollution and the mechanisms used by microbes and plants for environmental remediation. It also emphasized the importance of modern biotechnological techniques and approaches in improving the ability of microbial enzymes to effectively degrade heavy metals at a faster rate, highlighting recent advances in microbial bioremediation and phytoremediation for the removal of heavy metals from the environment as well as future prospects and limitations. However, strict adherence to biosafety regulations must be followed in the use of biotechnological methods to ensure safety of the environment.",
"conclusion": "10. Conclusions This review highlighted the effects of heavy metal contamination caused by some human activities on the environment, the possible health hazards, as well as the various mechanisms and enzymatic reactions used by plants and microbes to effectively remediate polluted environments. It revealed the usefulness of bioremediation as a better substitute for the removal of heavy metals from contaminated sites compared to the physico–chemical methods which are less efficient and expensive due to the amount of energy expended. Microorganisms and plants possess inherent biological mechanisms that enable them to survive under heavy metal stress and remove the metals from the environment. These microbes use various processes such as precipitaton, biosorption, enzymatic transformation of metals, complexation and phytoremedation techniques of which phytoextraction and phytostabilization have been very effective. However, the environmental conditions need to be adequate for effective bioremediation. The use of hyperaccumulator plants to remediate contaminated sites depends on the quantity of metal at that site and the type of soil. Environmental factors play a major role in the success of bioremediation as the microbes used will be hampered if appropriate environmental conditions are not available. More rapidly growing plants with high phytoextraction ability should be identified for the remediation of pollutants from soil. Moreover, assessment of metal stress on beneficial rhizospheric microorganisms and crops should be carried out and effective ways of enhancing the bioremediation process predicted. Transgenic microbes and plants could effectively remediate contaminated sites of heavy metal and organic pollutants but its use should be subject to stringent biosafety procedures to ensure that there is no health or environmental hazards. More efficient ways of using transgenic plants and microbes should be identified that will enable effective remediation of polluted environments without horzontal transfer of recombinant plasmids or pollens to indigenous organisms, which is currently a major challenge. Synthetic biology is an emerging technology that will be useful to synthesize microbial consortia with the ability to degrade and remove heavy metals from agricultural soils using the metabolic properties of such consortia of organisms. This technology should be promoted for more effective remediation of the environment from pollutants. Metagenomic approaches and metabolic analysis should also be used to study the functional composition of microbial communities within the polluted sites for metal resistance genes that could be used to improve specific heavy metal degradation strains of microbes. Public perception of the use of gene technology for bioremediation will also need to change for its effective utilization; this will require cooperation between researchers and environmentalist.",
"introduction": "1. Introduction Pollution of the environment keeps on increasing at an alarming rate due to the activities of man such as urbanization, technological advancement, unsafe agricultural practices and rapid industrialization which degrades the environment. Heavy metals released into the environment are persistent due to their toxicity which poses a severe threat to organisms exposed to high levels of such pollutants. Metals are essential to the biological functions of plants and animals but at elevated levels, they interfere with metabolic reactions in systems of organisms. Toxic heavy metals such as lead (Pb), cadmium (Cd), mercury (Hg), chromium (Cr), zinc (Zn), uranium (Ur), selenium (Se), silver (Ag), gold (Au), nickel (Ni) and arsenic (As) which are not useful to plants, are capable of reducing plant growth due to reduced photosynthetic activities, plant mineral nutrition, and reduced activity of essential enzymes [ 1 , 2 ]. Heavy metals are cytotoxic at low concentrations and could lead to cancer in humans [ 3 ]. These toxic metals could accumulate in the body when consumed in contaminated food through the food chain and become health risks to living organisms [ 4 ]. This causes oxidative stress, an unevenness involving the production of free radicals and the capacity of cells to eradicate them or repair the damage [ 5 , 6 ]. This leads to base damage through formation of reactive oxygen species (ROS) which includes oxygen radicals (superoxide and hydroxyl) [ 7 ] and non-radical derivatives of molecular oxygen (O 2 ) such as hydrogen peroxide (H 2 O 2 ), as well as breakage of the DNA molecule [ 5 , 6 ]. Heavy metal toxicity increases the production of ROS thereby decreasing the antioxidant systems (glutathione, superoxide dismutase, etc.) which protect cells. If this condition continues, the normal functioning of the organism is affected and may invariably lead to cell death. Bioremediation is gradually being accepted as the standard practice for the restoration of heavy-metal-contaminated soils since it is more eco-friendly and cost effective compared to the conventional chemical and physical methods, which are often very expensive and ineffective when metal concentrations are low, in addition to producing significant amounts of toxic sludge [ 8 , 9 ]. The cost effectiveness of bioremediation was reported by Blaylock et al. [ 10 ] who were able to save 50–65% of cost, when bioremediation was used for the treatment of one acre of Pb-polluted soil compared with the use of conventional methods such as excavation and landfill [ 7 ]. The ability of microorganisms to degrade pollutants depends on the suitability of environmental conditions for their growth and metabolism which include suitable temperature, pH, and moisture [ 11 ]. This review discusses the effects of heavy metals on the environment and how they can be effectively remediated using plants and microorganism. It further discusses the various mechanisms utilized by these organisms for remediating heavy metal contamination. The possible prospects and limitation of genetically modified organisms for bioremediation are also discussed."
} | 1,871 |
31882839 | PMC6934488 | pmc | 3,411 | {
"abstract": "Transgenic switchgrass overexpressing Lolium perenne L. delta1-pyrroline 5-carboxylate synthase ( LpP5CS ) in group I (TG4 and TG6 line) and group II (TG1 and TG2 line) had significant P5CS and ProDH enzyme activities, with group I plants (TG4 and TG6) having higher P5CS and lower ProDH enzyme activity, while group II plants had higher ProDH and lower P5CS enzyme activity. We found group II transgenic plants showed stunted growth, and the changed proline content in overexpressing transgenic plants may influence the growth and development in switchgrass. RNA-seq analysis showed that KEGG enrichment included phenylpropanoid biosynthesis pathway among group I, group II and WT plants, and the expression levels of genes related to lignin biosynthesis were significantly up-regulated in group II. We also found that lignin content in group II transgenic plants was higher than that in group I and WT plants, suggesting that increased lignin content may suppress switchgrass growth and development. This study uncover that proline can appropriately reduce lignin biosynthesis to improve switchgrass growth and development. Therefore, appropriate reduction in lignin content and increase in biomass are important for bioenergy crop to lower processing costs for biomass fermentation-derived fuels.",
"introduction": "Introduction Proline is known as a compatible solute (osmolyte) and a scavenger of reactive oxygen species (ROS) providing protection against oxidative damage 1 . In addition to its well established role in coping with environmental stress, proline also plays an increasingly significant role in plant development. It was reported that proline might have certain regulatory functions during protein synthesis and may act as a signaling molecule during plant development 1 , and proline plays an important role in plant growth and life cycle, such as regulating cyclin genes and affecting general protein synthesis 2 . Additionally, proline may also play critical roles in cellular metabolism both as a component of proteins and as a free amino acid 3 . Many genes are involved in the proline synthesis and degradation pathways. In higher plants, Δ1-pyrroline-5-carboxylate synthetase (P5CS) and proline dehydrogenase (ProDH) are rate-limiting enzymes during the synthesis and degradation of proline respectively. Glutamate is reduced to pyrroline-5-carboxylate (P5C) by P5CS, which is then converted into proline by Δ1-pyrroline-5-carboxylate reductase (P5CR) 1 . On the other hand, proline degradation takes place in the mitochondria, where it is catalyzed into glutamate by ProDH and P5C dehydrogenase (P5CDH) 4 . Proline metabolism occupies a central place in plant metabolism and is connected to other pathways through both ornithine and glutamate. It is connected to the pentose phosphate pathway and the TCA cycle as a way of moving the reductants and buffering the redox status of the chloroplast 4 , 5 . In the proline biosynthesis pathway, the consumption of the reductants (NADPH) buffers the redox status of the chloroplast, which is linked with the pentose phosphate pathway. In the pentose phosphate pathway, glucose-6-phosphate (G-6-P) is reduced to ribulose-5-phosphate (Ru-5-P) by the rate-limiting enzymes glucose 6-phosphate dehydrogenase (G6PDH) and 6-phosphogluconate dehydrogenase (6PGDH), and concurrently consumes NADP + to generate NADPH in this process 6 . In plants, transketolase (TK) plays a role in Calvin cycle while in non-photosynthetic organisms it connects the phosphate pentose pathway and glycolysis for generating NADPH 7 . In addition, proline degradation contributes carbon for TCA cycle 4 . The pentose phosphate pathway and Calvin cycle provide erythrose-4-phosphate (E4P) which together with phosphoenolpyruvate acts as precursor for phenylalanine biosynthesis through the shikimic acid pathway 5 . Phenylalanine, as the precursor amino acid for lignin biosynthesis, is essential for secondary cell wall biosynthesis 8 . In most plants, lignin is mainly composed of hydroxyphenyl (H), guaiacyl (G) and syringyl (S) monolignol subunits that are derived from p -coumaryl, coniferyl and sinapyl monolignols, respectively. Several enzymes are required for monolignol biosynthesis, including phenylalanine ammonia-lyase (PAL), caffeic acid 3-Omethyltransferase (COMT), caffeoyl-CoA O-methyltransferase (CCoAOMT), hydroxycinnamoyl-CoA reductase (CCR), cinnamyl alcohol dehydrogenase (CAD) 9 . A number of reports have showed that lignin content is reduced by down-regulating the expression of monolignol biosynthesis related genes in plants 10 . Many studies had showed that lignin content had the negative relationship with the growth and development of plants. Transgenic aspen with suppressed Pt4CL1 expression exhibited up to a 45% reduction of lignin and a 15% increase in cellulose, but leaf, root, and stem growth were substantially enhanced 11 . Additionally, compared with WT plants, AtLOV1 transgenic switchgrass plants had higher total lignin content, delayed flowering time and less aboveground biomass 12 . Besides, a number of studies suggested that the most severe deficiency of lignin showed stunted growth in plants 13 . However, many studies also showed that lignin can be reduced without reducing yield or fitness 14 , 15 . Switchgrass ( Panicum virgatum L.) is a perennial C4 grass native to North America, considered as a potential dedicated bioenergy crop due to its high biomass production and tolerance on marginal land 9 . We found that overexpression plants showed two different phenotypes, the two groups transgenic plants had different P5CS and ProDH enzyme activities, as well as proline content. To shed light on potential changes in growth and development, we analyzed the RNA-seq data. The results showed that phenylpropanoid biosynthesis pathway was significantly up-regulated in the group II plants compared with the group I and WT plants. In particular, lignin content in group II transgenic plants was higher than that in group I and WT plants, suggesting that proline affects switchgrass growth and development by coordination with lignin biosynthesis.",
"discussion": "Discussion P5CS activity was lower in group II plants, which could led to the decreased proline accumulation. In the present study, KEGG enrichment included pentose phosphate pathway (PPP), fatty acid synthesis and phenylalanine biosynthesis pathway, and at the same time, NADP + and NADPH content showed a significant difference between the two groups of transgenic and WT plants. NADP + and NADPH are generated during the synthesis and degradation of proline, respectively 18 . The cycling of proline substrate is coupled to maintaining the NADP + /NADPH ratio via the pentose phosphate pathway 19 , and PPP provides NADPH for fatty acid synthesis in plastids 6 . So it seems that proline degradation was accelerated in the group II transgenic plants, which led to an increase in the NADPH content. It was reported that antisense AtP5CS transgenic Arabidopsis led to proline depletion and abnormal leaf formation 20 . Down-regulation of proline biosynthesis genes, on the other hand, resulted in growth defects 21 , indicating the importance of proline for plant development. In our study, the stems were significantly thicker in group I transgenic plants and thinner in group II plants, when compared with the WT plants (Fig. 3D ). Furthermore, as the results of the RNA-seq analysis indicated, the enriched KEGG pathways also include the photosynthesis and circadian rhythm-plant pathways (Supplementary Fig. S2 ). This leads us to conclude that proline may play important roles in plant growth and development. Proline metabolism is connected to the pentose phosphate pathway (PPP) and the TCA cycle, and the PPP and Calvin cycle provide erythrose-4-phosphate (E4P) which together with phosphoenolpyruvate acts as a precursor for phenylalanine biosynthesis through the shikimic acid pathway 5 . In our result of RNA-seq, the enriched KEGG pathways were including arginine and proline metabolism, flavonoid biosynthesis, phenylpropanoid biosynthesis and pentose phosphate pathway (Supplementary Fig. S3 ). To precisely understand the relationship between proline metabolism and lignin biosynthesis in switchgrass, we summarized the metabolic network that exists between the phenylpropanoid biosynthesis pathway, pentose phosphate pathway and proline metabolic pathway in Fig. 6 . In these metabolism pathways, PvG6PDH and Pv6PGDH are genes encoding key enzymes involved in the pentose phosphate pathway; PvPAL , PvCOMT , PvCCoAOMT , PvCCR and PvCAD are the genes encoding for important enzymes in lignin biosynthesis, and transketolase (TK) participates in the Calvin cycle. The proline biosynthesis pathway is linked with the pentose phosphate pathway through consuming the reductants (NADPH), while the pentose phosphate pathway and the Calvin cycle provide erythrose-4-phosphate (E4P) for phenylalanine biosynthesis. Thus, proline metabolism pathway is coordinated with phenylpropanoid biosynthesis pathways in switchgrass. In particular, our results showed that the lignin content (S + G + H monomer content) was lower in group I plants and higher in group II plants compared with the WT. So, reducing proline biosynthesis may induce the increase of lignin content in group II transgenic plants. Many studies showed that lignin content affected plant growth and development. Suppressing Pt4CL1 expression can reduce the lignin content and enhance the growth of leaf, root, and stem in aspen 11 , and RNA interference of Pv4CL1 lead to a reduction in lignin content with uncompromised biomass yields 15 . Additionally, overexpressing AtLOV1 gene increased total lignin content, delayed flowering time and reduce the aboveground biomass in switchgrass 12 . Besides, extreme lignin deficiency can also contribute to stunted growth in plants 13 . In our result, group II transgenic plants which had lower proline and higher lignin content exhibited more stunted growth. Thus, we speculate that proline affects switchgrass growth and development by coordination with lignin biosynthesis. Figure 6 Metabolic network diagram between phenylpropanoid biosynthesis pathway, pentose phosphate pathway and proline metabolic pathway. In conclusion, unlike the previous studies that proline plays an important role in regulating cyclin genes and affecting general protein synthesis to improve plant growth and development. Our study showed that lacking proline could result in an increased lignin content, and leading to the stunted growth in switchgrass. However, the regulatory mechanism how proline plays roles in plant growth and development is still unclear, and understanding the molecular mechanism will help us to prime candidates in crop genetic engineering for improving the biomass of plants. Moreover, appropriate reduction in lignin content and increase in biomass is one important strategy of bioenergy crop to lower processing costs for biomass fermentation-derived fuels."
} | 2,759 |
24183010 | null | s2 | 3,413 | {
"abstract": "Historical milestones in neuroscience have come in diverse forms, ranging from the resolution of specific biological mysteries via creative experimentation to broad technological advances allowing neuroscientists to ask new kinds of questions. The continuous development of tools is driven with a special necessity by the complexity, fragility, and inaccessibility of intact nervous systems, such that inventive technique development and application drawing upon engineering and the applied sciences has long been essential to neuroscience. Here we highlight recent technological directions in neuroscience spurred by progress in optical, electrical, mechanical, chemical, and biological engineering. These research areas are poised for rapid growth and will likely be central to the practice of neuroscience well into the future."
} | 207 |
37851813 | PMC10584339 | pmc | 3,415 | {
"abstract": "An iontronic-based artificial tactile nerve is a promising technology for emulating the tactile recognition and learning of human skin with low power consumption. However, its weak tactile memory and complex integration structure remain challenging. We present an ion trap and release dynamics (iTRD)–driven, neuro-inspired monolithic artificial tactile neuron (NeuroMAT) that can achieve tactile perception and memory consolidation in a single device. Through the tactile-driven release of ions initially trapped within iTRD-iongel, NeuroMAT only generates nonintrusive synaptic memory signals when mechanical stress is applied under voltage stimulation. The induced tactile memory is augmented by auxiliary voltage pulses independent of tactile sensing signals. We integrate NeuroMAT with an anthropomorphic robotic hand system to imitate memory-based human motion; the robust tactile memory of NeuroMAT enables the hand to consistently perform reliable gripping motion.",
"introduction": "INTRODUCTION Tactile memory is a crucial sensory modality that enables human skin to perceive and interpret its surrounding environment—vital tasks in human daily activities ( 1 – 3 ). The tactile encoding enables humans to identify objects via touch and reliably and repeatedly handle (e.g., grip and lift) them nondestructively through automatic access to previously acquired tactile information (i.e., motor skills) ( 3 , 4 ). This ability originates from tactile-based implicit memory [unconscious long-term memory (LTM)], enabling humans to rapidly and efficiently perform specific motion tasks without consciously retrieving object information ( 1 , 5 ). Such biological tactile recognition and LTM-based motion processing, realized by the tactile neural network of the somatosensory system, are essential functions that a neuro-inspired robotic system performing human motion must emulate. These functions can facilitate a human-like (anthropomorphic) robot to rapidly interact with a dynamic environment and consistently manipulate target objects without reduction of motion accuracy and requirement of iterative control command ( Fig. 1A ) ( 3 , 4 , 6 , 7 ). Fig. 1. Concepts of tactile memory–based human motion and ion trap and release dynamics (iTRD)-driven neuro-inspired monolithic artificial tactile neuron (NeuroMAT). ( A ) Schematics of tactile perception and learning procedure in biological tactile neuron and tactile memory–based reliable human motion. ( B ) Comparison of conventional ion behavior and iTRD mechanism under external E-field and tactile stimuli. ( C ) Tactile perception and learning characteristics of iTRD-driven NeuroMAT. Several artificial tactile neural systems, e.g., triboelectric- or ferroelectric-driven single-synaptic devices, have been proposed for mimicking biological tactile perceptual memory performance ( 8 – 10 ). The systems showed plausible LTM toward tactile stimuli but have the complex tactile perception (magnitude, frequency, and duration) or high-voltage operation, which are undesirable characteristics in artificial tactile neuromorphic devices ( 11 ). On the other hand, neural processing of sensory information in biological synapses, driven by charges (particularly ions), has inspired artificial synapse disciplines to exploit extremely low power consumption and high data processing efficiency ( 11 – 14 ). Approaches based on this concept primarily involve the integration of tactile sensors (triboelectric, piezocapacitive, or piezoresistive), signal converters (ring oscillator or ionic cable), and iongel-gated synaptic memory transistors ( 15 – 18 ) or fabrication of a suspended iongel-gated synaptic transistor with an additional control gate electrode ( 19 ). However, they suffer from complicated electrical interconnections, low circuit compatibility, data latency, and additional power requirements for operating each device component ( 19 , 20 ). The major drawback is a lack of tactility-induced LTM (typically <200 s), which is a critical feature for incorporating learning behavior in an artificial tactile neural system (table S1). This results from the intrinsic behavior of ions to reach an electrochemical equilibrium state within conventional iongel dielectrics in synaptic transistors. Ions accumulated on or diffused into the semiconductor layer can easily dissipate at zero voltage, which causes rapid erasure of the conductance encoded by the synaptic device (i.e., volatile tactile memory). In the case of the integrated tactile neural system, voltage inputs transmitted from tactile sensors to synaptic transistors are tactile event–driven, thereby causing difficulty in the independent application of auxiliary voltage stimuli to suppress the ion equilibrium behavior for preserving the induced tactile memory. Even if external voltage stress is intentionally applied, the histories of the tactile event–driven synaptic signal and memory are easily distorted because the internal ion flux (i.e., ion migration and diffusion) in the iongel dielectrics is dominantly manipulated by voltage signals not generated by the tactile signals of the system. From this perspective, the ion behavior of conventional ionic materials poses a challenge to the augmentation of tactile memory and the design of a simple artificial tactile neural system capable of perceiving and learning tactile stimuli in a monolithic device. Here, we report a neuro-inspired, monolithic (capable of sensing, memorizing, and even learning, all in a single device) artificial tactile neuron, termed NeuroMAT, with nonintrusive and augmented memory based on ion trap and release dynamics (iTRD), in which the free ion density and ion flux can be independently modulated using tactile stimuli (pressure) and an external electric field (E-field), respectively. We use this system to demonstrate a reliable human motion–mimicking an anthropomorphic robotic hand. In the ion trap state ( n free ions ≈ 0), the ion flux of the iTRD-iongel is not modulated by the external E-field; hence, the potential drop is uniformly distributed over the iongel film ( Fig. 1B ). This behavior is the opposite of that in a conventional iongel, wherein the potential drop is concentrated at the iongel–semiconductor layer interface because of E-field–induced formation of an electric double layer (EDL). Tactile stimulus (pressure) application can release the trapped ions ( n free ions ≠ 0), at which instant only the released ions can contribute to the total ion flux of the iTRD-iongel; this consequently causes a considerable potential drop in the interface regions through released ion-induced EDL formation. This implies that the ions released under tactile stimuli can exclusively generate synaptic memory signals (reflecting tactile information), and the induced memory level can be augmented by applying an auxiliary external E-field (suppression of equilibrium behavior of the released ions). We used this unique ion behavior to develop the iTRD-driven NeuroMAT with nonintrusive and augmented memory, capable of simultaneously sensing, memorizing, and learning tactile stimulation ( Fig. 1C ). The NeuroMAT can perform highly sensitive tactile recognition under a series of voltage pulses; concurrently, retention of the encoded tactile memory is effectively improved by the application of voltage signals independent of the tactile signals. This characteristic cannot be explored in conventional iongel-based synaptic transistors because the internal ion flux is already dominated by the E-field so that ion-induced synaptic signal cannot be manipulated by tactile stimuli.",
"discussion": "DISCUSSION In summary, we developed an iTRD-driven, neuro-inspired monolithic artificial tactile neuron (NeuroMAT) with nonintrusive and augmented memory. The ion behavior of the iTRD is completely distinguished from conventional ion dynamics in that the free ion density and ion flux can be independently modulated by tactile and voltage stimulations, respectively. The mechano-driven iTRD mechanism enabled the proposed NeuroMAT to simultaneously perceive, memorize, and learn tactile information in a single device, without requiring any additional sensors and memory devices. We used the augmented tactile memory of NeuroMAT to develop a NeuroMATICS-based robotic hand system, which consistently and reliably emulated tactile memory–driven human motion. Our theoretical study and practical demonstration of the iTRD behavior can provide inspiration for artificial nerve disciplines and can be extended to the field of neuromorphic electronic skin and machine learning."
} | 2,144 |
24672515 | PMC3949148 | pmc | 3,416 | {
"abstract": "Energy conservation via the pathway of dissimilatory sulfate reduction is present in a diverse group of prokaryotes, but is most comprehensively studied in Deltaproteobacteria . In this study, whole-genome microarray analyses were used to provide a model of the energy metabolism of the sulfate-reducing archaeon Archaeoglobus fulgidus , based on comparative analysis of litoautotrophic growth with H 2 /CO 2 and thiosulfate, and heterotrophic growth on lactate with sulfate or thiosulfate. Only 72 genes were expressed differentially between the cultures utilizing sulfate or thiosulfate, whereas 269 genes were affected by a shift in energy source. We identified co-located gene cluster encoding putative lactate dehydrogenases (LDHs; lldD , dld , lldEFG ), also present in sulfate-reducing bacteria. These enzymes may take part in energy conservation in A. fulgidus by specifically linking lactate oxidation with APS reduction via the Qmo complex. High transcriptional levels of Fqo confirm an important role of F 420 H 2 , as well as a menaquinone-mediated electron transport chain, during heterotrophic growth. A putative periplasmic thiosulfate reductase was identified by specific up-regulation. Also, putative genes for transport of sulfate and sulfite are discussed. We present a model for hydrogen metabolism, based on the probable bifurcation reaction of the Mvh:Hdl hydrogenase, which may inhibit the utilization of Fd red for energy conservation. Energy conservation is probably facilitated via menaquinone to multiple membrane-bound heterodisulfide reductase (Hdr) complexes and the DsrC protein—linking periplasmic hydrogenase (Vht) to the cytoplasmic reduction of sulfite. The ambiguous roles of genes corresponding to fatty acid metabolism induced during growth with H 2 are discussed. Putative co-assimilation of organic acids is favored over a homologous secondary carbon fixation pathway, although both mechanisms may contribute to conserve the amount of Fd red needed during autotrophic growth with H 2 .",
"introduction": "Introduction The sulfate-reducing prokaryotes (SRP) have played a central role in cycling of carbon and sulfur in anoxic environments throughout long periods of Earth's geological history. Despite early characterization of the cytoplasmic pathway of dissimilatory sulfate reduction (Peck, 1962 ) it is only in recent years that the mechanisms facilitating energy conservation in SRP have been more comprehensively characterized (Pereira et al., 2011 ; Grein et al., 2013 ). The genus Archaeoglobus comprises of archaeal, (hyper)thermophilic, dissimilatory sulfate reducers (Stetter et al., 1987 ; Stetter, 1988 ) and is phylogenetically associated with the lineages of Methanosarcinales , Methanomicrobiales , and uncultured ANME-1 (Brochier-Armanet et al., 2008 ; Guy and Ettema, 2011 ). The type species A. fulgidus VC16 is a chemolithoautotroph that utilizes H 2 or formate as electron donors for autotrophic growth. In addition, A. fulgidus grows carboxydotrophically on CO/CO 2 and as a chemoorganoheterotroph utilizing a wide range of substrates including fatty acids, alkenes, complex peptides, and specific amino acids (Stetter et al., 1987 ; Stetter, 1988 ; Hartzell and Reed, 2006 ; Henstra et al., 2007 ; Khelifi et al., 2010 ; Parthasarathy et al., 2013 ). For the complete oxidation of organic substrates to CO 2 , A. fulgidus uses a modified acetyl-CoA pathway with similar enzymes and cofactors as in the methanogens (Möller-Zinkhan et al., 1989 ; Möller-Zinkhan and Thauer, 1990 ; Vorholt et al., 1995 ; Estelmann et al., 2011 ). Reduction of sulfate (SO 2− 4 ) to sulfide (S 2− ) in A. fulgidus proceeds via the highly conserved dissimilatory sulfate reduction pathway of the SRP (Peck, 1962 ; Klenk et al., 1997 ; Pereira et al., 2011 ). This was probably acquired by Archaeoglobales via multiple lateral gene transfer events from an early ancestor of clostridial SRP (Klein et al., 2001 ; Zverlov et al., 2005 ; Meyer and Kuever, 2007 ). The energy conservation mechanisms in A. fulgidus are incompletely understood. During growth on lactate, the reduced coenzyme F 420 (F 420 H 2 ) is generated from the oxidative acetyl-CoA pathway. The presence of both menaquinone and a homolog of the respiratory NAD(P)H:quinone oxidoreductase complex, the F 420 H 2 :quinone oxidoreductase complex (Fqo), suggest that electrons from F 420 H 2 are transferred to the membrane-bound respiratory chain by the Fqo complex. Fqo probably couples the reduction of menaquinone and proton translocation. (Tindall et al., 1989 ; Kunow et al., 1993 ; Baumer et al., 2000 ; Brüggemann et al., 2000 ) A d -lactate dehydrogenase is confirmed to be present (Reed and Hartzell, 1999 ), but it is unclear how this membrane associated enzyme facilitates energy conservation, as it is shown to interact with a NADH oxidase (Pagala et al., 2002 ). Also, the cofactor NAD(P)H plays a negligible role in energy conservation (Noll and Barber, 1988 ; Kunow et al., 1993 ; Warkentin et al., 2001 ). There is also a possible alternative energy conservation pathway in A. fulgidus . In D. vulgaris , cytochrome c mediated “hydrogen cycling” is suggested as an energy conservation mechanism during growth with lactate (Odom and Peck, 1981 ; Keller and Wall, 2011 ). In this reaction, formation of hydrogen is a result of cytoplasmic oxidation of lactate. The subsequent diffusion and periplasmic oxidation of hydrogen contributes to the formation of a proton gradient. In Methanosarcina barkeri , the Vht/Vhx dehydrogenase also facilitates a hydrogen cycling mechanism under heterotrophic growth conditions, and sustains growth when Fpo (Fqo) is absent in deletion mutants (Kulkarni et al., 2009 ). The presence of a cytoplasmic as well as a periplasmic hydrogenase in A. fulgidus (Mander et al., 2004 ) potentially fulfills requirements for a “hydrogen-cycling” mechanism. Two co-located heterodisulfide reductase (Hdr)-associated hydrogenases are present in the genome of A. fulgidus , which are homologous to those involved in energy conservation in the methanogens (Mander et al., 2004 ). These are the soluble [NiFe]hydrogenase/heterodisulfide-like (MvhABC/HdlABC) complex and the membrane-bound uptake hydrogenase, “F 420 -non-reducing hydrogenase” (Vho/Vht). Reduced ferredoxin (Fd red ) is essential for fixation of CO 2 through the acetyl-CoA pathway. In methanogens, the Mvh:Hdl complex homolog, Mvh/Hdr, couples the exergonic reduction of the heterodisulfide, CoM-S-S-CoB, with endergonic reduction of ferredoxin with H 2 , by a flavine-based bifurcation mechanism (Kaster et al., 2011 ). The periplasmic Vht hydrogenase reduces the quinone-like cofactor methanophenazine coupled to the membrane-bound HdrDE, facilitating energy conservation during growth on H 2 (Ide et al., 1999 ; Thauer et al., 2010 ). Despite the absence of genes and cofactors for terminal methanogenesis (Stetter et al., 1987 ; Klenk et al., 1997 ), several factors suggest that thiol/disulfide conversions catalyzed by Hdr are involved in electron transfer and energy conservation in A. fulgidus , as has been proposed for methanogens and more recently for SRP (Mander et al., 2002 , 2004 ; Pereira et al., 2011 ; Grein et al., 2013 ). All known SRP, including A. fulgidus , encode HdrA and HdrDE related genes, which almost ubiquitously form membrane-bound redox complexes (Pereira et al., 2011 ; Grein et al., 2013 ). These complexes may facilitate energy conservation during different steps of sequential dissimilatory sulfate reduction. The quinone-interacting membrane-bound oxidoreductase (QmoABC) complex probably links the electron transfer chain to the first reductive step of sulfate reduction catalyzed by adenosine-5′-phosphosulfate (APS) reductase (AprAB) (Pires et al., 2003 ; Zane et al., 2010 ; Grein et al., 2013 ). In Desulfovibrio it has recently been proposed that the Qmo subunit homologous to the bifurcating HdrA, QmoB, may facilitate a “confurcation” mechanism (Ramos et al., 2012 ). The “confurcating” Qmo complex may catalyze energy conservation by proton translocation via an endergonic periplasmic menaquinol oxidation, driven by an exergonic cytoplasmic oxidation reaction coupled to terminal reduction of APS. The second complex, DsrMK, is a homolog of HdrDE, and is ubiquitous amongst SRP (Pereira et al., 2011 ). This complex probably facilitates energy conservation and is linked by electron transfer via disulfide/thiol redox reactions, to the terminal step of sulfite reduction by bisulfite reductase/sulfite reductase (DsrAB) (Mander et al., 2002 ; Pires et al., 2006 ). Similarly to the HdrDE of methanogens, the DsrMK complex probably couples periplasmic oxidation of reduced menaqinone (instead of reduced methanopenazine) to cytoplasmic cysteine disulfide (Cys-S-S-Cys) reduction, in the enzyme DsrC (in stead of a CoM-S-S-CoB) (Mander et al., 2005 ). Unusually, dsrMK is encoded by multiple homologs in A. fulgidus , corresponding to multiple DsrMK and a DsrMK(JOP) complex, which differ in domain composition and among lineages of SRP (Klenk et al., 1997 ; Pereira et al., 2011 ). The dsrC gene is ubiquitously present in SRP, and DsrC is the probable link between heterodisulfide reductase (DsrK) and DsrAB (Oliveira et al., 2008 ; Pereira et al., 2011 ; Grein et al., 2013 ). However, it should be noted that although it is likely that the DsrMK(JOP) complexes may facilitate proton translocation by MQH 2 oxidase:DsrC reductase, it is questioned whether this reaction is thermodynamically favorable (Thauer et al., 2007 ; Grein et al., 2013 ). The role of reduced ferredoxin (Fd red ) in energy conservation in SRP remains unclear, as it has been proposed as an electron donor for both APS and sulfite reduction (Oliveira et al., 2008 , 2011 ; Ramos et al., 2012 ). In A. fulgidus , this offers a potential coupling between ferredoxin and electron transport phosphorylation, but also represents a significant bioenergetic challenge, as fixation of CO 2 through the acetyl-CoA pathway requires Fd red . Interestingly, while chemoorganotrophic and carboxydotrophic growth are coupled to sulfate reduction in A. fulgidus , only thiosulfate or sulfite are utilized with H 2 as energy source (Stetter et al., 1987 ; Steinsbu et al., 2010 ). This may potentially be coupled to the role of Fd red in energy and carbon metabolism. To provide a deeper insight into electron transport and energy conservation mechanisms in A. fulgidus , we used whole genome microarrays to identify redox complexes expressed under different growth conditions. Previously, only the heat shock response in A. fulgidus has been characterized by global transcriptional profiling (Rohlin et al., 2005 ). We examined heterotrophic growth with lactate and litoautotrophic growth with H 2 , as well as the differential use of the electron acceptors thiosulfate and sulfate. The results form an overall energy conservation model where the Fqo and membrane-bound electron transport, facilitated by menaquinone, Qmo and multiple DsrMK, are central to energy conservation during growth with lactate. During growth with hydrogen, our model suggests that Fd red , generated by Mvh:Hdl, is utilized primarily for carbon assimilation and probably does not contribute to energy conservation. From the data and comparative genomics it seems likely that the inability of A. fulgidus to grow with sulfate when hydrogen is an energy source is caused by transcriptional regulation of the gene for pyrophosphatase, resulting in the blocking of APS formation. Overall, the results point to a key role in energy conservation for electron transfer from hydrogen to thiosulfate, facilitated by thiol/disulfide conversions catalyzed by membrane-bound DsrMK in A. fulgidus .",
"discussion": "Discussion In the present work, a model of the energy metabolism in A. fulgidus for the utilization of lactate and hydrogen with thiosulfate or sulfate as terminal electron acceptors is presented based on transcriptome profiling. Lactate metabolism Lactate is the “classical” substrate of sulfate reducers, and its link to energy conservation in Desulfovibrio has been the subject of intense study (Keller and Wall, 2011 ). Several transcriptional shifts were observed in A. fulgidus , involving expression of LDH and putative LDH genes (Table 1E , Figure 3 ). Our results indicate that during growth with T-L, activity of multiple LDH isozymes (Figures 3 , 5 ) may occur in A. fulgidus , as suggested in D. vulgaris (Keller and Wall, 2011 ). When sulfate is used as an electron acceptor, oligomeric LdlEFG may operate together with monomeric lldD and dld in the oxidation of lactate (Figure 5 ). The conserved “modular” domain composition of the proteins encoded by the genes dld, the ORF “AF0808” and AF0809, may facilitate a multimeric complex that functions as monomeric homologs encoded in other species (Dvu3071, Figures 3 , 5 ). The presence of a gene cluster with identical arrangement in the lactate utilizing A. sulfaticallidus and A. fulgidus , supports a potential role of the LdlEFG in linking lactate oxidation with sulfate reduction in A. fulgidus . Acquiring the lldEFG gene cluster may have been essential for Archaeoglobales in order to perform dissimilatory sulfate reduction with lactate as an energy source, potentially via the QmoABC complex to APS reductase. Oligomeric lldEFG is widely distributed in Bacteria, including sulfate-reducing Deltaproteobacteria (Pinchuk et al., 2009 ; Pereira et al., 2011 ), but has previously not been identified in Archaea. Various functions have, however, been suggested for LdlEFG in Bacteria. In S. oneidensis MR-1, the LdlEFG is found to stimulate the activity of Dld-II (Figure 3 ), indicating a functional relationship (Pinchuk et al., 2009 ). Interestingly, in D. alaskensis the LdlEFG is required in syntrophic growth with Methanococcus (Meyer et al., 2013 ). In this model an LdhAB-1 (GplCD) catalyses the primary oxidation of lactate, and transfers electrons, possibly through thiol/disulfide, to an LdlEFG homologous complex. The LdlEFG may transfer electrons to the QmoABC complex, which facilitates menaquinol reduction (Meyer et al., 2013 ). However, the LdlEFG is also present in species without a QmoABC complex and functions independently as a membrane associated l -LDH capable of reducing quinone (Chai et al., 2009 ; Pinchuk et al., 2009 ; Thomas et al., 2011 ). In order to verify the specific role of the LdlEFG homologs in A. fulgidus , biochemical studies are required (enzyme activity and protein-protein interaction) to understand its relation to Qmo and energy conservation. Perhaps, prior to construction of deletion mutants as a genetic system is not yet available for this species. With the exception of the cdhAB-1 (see next section), genes encoding the acetyl-CoA pathway were constitutively expressed at high levels (Table S2 ). This was also true for the F 420 H 2 : quinone oxidoreductase (Fqo) complex, which probably catalyzes proton translocation utilizing F 420 H 2 generated by the oxidative acetyl-CoA pathway (Brüggemann et al., 2000 ). The hydrogenases in A. fulgidus were specifically induced during growth with hydrogen, and low transcriptional expression of hydrogenases was observed during growth on lactate (Table 1 ). Therefore, it may be questioned whether “hydrogen cycling” (Odom and Peck, 1981 ; Kulkarni et al., 2009 ) is used as a mechanism for energy conservation with lactate as the energy source. This would emphasize the role of the respiratory Fqo complex and a menaquinone-based respiratory system (Figure 5 ) in energy conservation in A. fulgidus during growth with lactate. Several distinct putative menaquinol oxidase:Hdr complexes are present in the genome of A. fulgidus . The DsrMKJOP (AF0499-AF503) complex and the DsrMK(K) (AF0543-AF0544) were constitutively highly expressed (Table 1C ), whereas a second DsrMK (AF0543-AF0544) was expressed at average expression levels (Table 1C ). Multiple membrane-bound DsrMK complexes may therefore oxidize the menaquinol (MQH 2 ) generated by the Fqo complex (Figure 5 ). The DsrK components may transfer electrons to DsrC by breaking the disulfide bonds between the two C—terminal cysteines of the enzyme (Mander et al., 2005 ; Oliveira et al., 2008 ; Grein et al., 2010 ). The dsrC gene (AF2228) is, however, expressed at average transcriptional abundance vs. high transcriptional abundance for dsrAB (Table 1C , Figure S1 ). This is lower than previously estimated in D. vulgaris , where the gene of dsrC is expressed at high levels (Wall et al., 2008 ). Although these values are more rigorously estimated in the previous study, our results point toward a lower expression ratio between dsrAB and dsrC in A. fulgidus . This may indicate involvement of additional electron transport components from Hdr to DsrAB. However, other electron carrying proteins, such as ferredoxin, are expressed at similar levels as dsrC ( fdx , <1.4 average expression, Table S3 ). The true significance of the role of electron flow via DsrC requires further evaluation on translational level. Reduction of thiosulfate The mechanism of thiosulfate reduction and the import of sulfate for cytoplasmatic reduction is uncharacterized in A. fulgidus . The specific growth rate of A. fulgidus cultivated with lactate was increased by the utilization of thiosulfate, vs. sulfate, as terminal electron acceptor (Figure 1A ). The reduction of thiosulfate is thermodynamically favorable (Badziong and Thauer, 1978 ). However, utilization of thiosulfate vs. sulfate is reported as inhibiting for growth rate in D. vulgaris Hildenborough, and has been attributed to the toxicity of increasing intracellular concentrations of sulfite (Badziong and Thauer, 1978 ; Pereira et al., 2008 ). The genes corresponding to thiosulfate reductase in A. fulgidus are identified by specific up-regulation of a molybdopterin oxidoreductase (AF2384-AF2386, Table 1A ). This reductase is active during both lactate and H 2 -oxidation, and is probably a membrane-integrated complex with a periplasmic facing active site (Figure 5 ). The presence of a Tat signal peptide (Figure 4 ) indicates that export is facilitated by the twin arginine translocation pathway (Coulthurst et al., 2012 ). A periplasmic reduction of thiosulfate may exclude the build-up of toxic intracellular levels of sulfite and may partly explain the high growth rate observed for A. fulgidus during cultivation with lactate and thiosulfate. It is unlikely that the reduction of thiosulfate to sulfite (E°′ −402 mV) contributes to energy conservation (Badziong and Thauer, 1978 ; Stoffels et al., 2012 ). Rather, final intracellular reduction of SO 2− 3 to S 2− (E°′ −116 mV) has a redox potential sufficient for energy conservation (Thauer et al., 2007 ). In most Desulfovibrio spp., an indistinguishable ion gradient symport has been found for thiosulfate and sulfate (Cypionka, 1987 ; Stahlmann et al., 1991 ). However, such a mechanism has not been identified in A. fulgidus (Rabus et al., 2006 ). The genes previously annotated “sulfate ABC transporter permease” (AF0092-AF0094) are probably a molybdate-specific transporter (Klenk et al., 1997 ; Hollenstein et al., 2007 ). The induction of genes corresponding to a ABC-type transport system (AF1136-AF1138) during growth with sulfate (S-L) may ambiguously be assigned as a putative sulfate transporter, as these genes are also induced during growth with lactate (Table 1B , Figure 5 ). The gene tauE is proposed to encode a sulfite exporter in Cupriavidus necator ( Ralstonia eutropha ) during sulfoacetaldehyde degradation (Weinitschke et al., 2007 ). The constitutive highly expressed tauE homolog (AF1562) may be assigned a putative function for sulfite import in A. fulgidus (Table 1B , Figure 5 ). The utilization of thiosulfate is a common property of all Archaeoglobus spp. and Ferroglobus placidus , however, homologous of the putative periplasmic AF2384-AF2386 gene cluster can only be found in the species A. fulgidus and F. placidus (BLASTp, Absynte, and Syntax-tools Oberto, 2013 ). The DsrAB of A. fulgidus displays a high level of in vitro thiosulfate reductase activity (Parey et al., 2010 ) and may play a role as a parallel process of cytoplasmatic thiosulfate reductase. Although a common trait, different Archaeoglobus spp. seem to have diverging enzyme systems for thiosulfate reduction. Unexpectedly, when thiosulfate was substituted for sulfate as electron-acceptor, a second copy of the cdhAB-1 subunits in the ACS/CODH complex was induced (AF1100-AF1101, Table 1F , Figure 5 ). The specific regulation corresponding to terminal electron acceptor (thiosulfate) may indicate a preferential utilization of different electron carriers between dissimilatory sulfate and thiosulfate reduction. A similar shift in genes of cobalamin/vitamin B 12 biosynthesis may also affect the function of the ACS subunit (Banerjee and Ragsdale, 2003 ). Previous studies have shown that regulation of CODH/ACS complexes in M. acetivorans are related to carbon source (Matschiavelli et al., 2012 ). Hydrogen metabolism Archaeoglobus fulgidus possesses only two hydrogenases; the periplasmic Vht hydrogenase and the soluble Mvh:Hdl (Mander et al., 2004 ). The latter may take part both in energy conservation and in generation of Fd red for CO 2 —fixation through the acetyl-CoA pathway. The reductive acetyl-CoA pathway requires at least 3 mol Fd red for the generation of one mole pyruvate from CO 2 (Fuchs, 2011 ). Similar to the methanogens, a bifurcation reaction is obligate in A. fulgidus for the generation of Fd red while growing autotrophically with hydrogen. In addition, Fd red can be hypothesized as an electron donor to APS reductase, through the QmoABC (Ramos et al., 2012 ) and the DsrAB (Oliveira et al., 2008 , 2011 ) offering a potential coupling between ferredoxin and electron transport phosphorylation. The genes of periplasmic Vht hydrogenase represented the highest transcriptional shift of any genes in relation to growth on T-H 2 /CO 2 and was expressed at a high level relative to average signal abundance (Table 1D , Figures 4 , 5 ). The resulting two protons from a periplasmic hydrogenase reaction catalyzed by Vht may contribute directly to the formation of pmf during growth. In Methanomicrobiales , the Vht hydrogenase homolog donates electrons via methanophenazine (MP) to a cytoplasmic-facing, membrane-bound Hdr (HdrDE, Figure 6A ) (Deppenmeier et al., 1992 ; Ide et al., 1999 ; Thauer et al., 2010 ). Similarly, the VhtABC complex in A. fulgidus may reduce menaquinone (MQ). A subsequent menaquinol (MQH 2 ) oxidation, facing the periplasm, by a membrane-bound Hdr may translocate two protons (Figures 5 , 6A ). The observed co-induction of vht hydrogenase genes and a fused hdrDE homolog encoding dual [CCG] domains (AF0755, Figure 4 ) suggests a close physical interaction between the two encoded complexes that may form a distinct path of electron flow to DsrAB. However, the genes of DsrMKJOP or DsrMK(K) were constitutively expressed, and therefore, electron flow is also possible via these complexes (Table 1 , Figures 4 , 5 ). The reaction probably represents the major pathway of energy conservation during growth with H 2 . Figure 6 (A) A schematic comparison of hydrogenotrophic and methanogenic metabolism with that of A. fulgidus . The Wolfe cycle of autotrophic hydrogenotrophic methanogens without cytochromes is shaded in red. The Mvh:Hdr is the only known ferredoxin-reducing complex present in A. fulgidus . A pathway analogs to the Wolfe cycle is indicated in orange—if Fd red is required for either APS or SO 2− 3 reduction, prior to the reduction of di-thiol (X-SH HS-X), an anaplerotic ferredoxin-reducing hydrogenase is required. (B) An overview of alternative complexes for Fd reduction (dotted borders), which are absent in A. fulgidus . Most methanogens and SRP maintain several of these complexes. The Wolfe cycle requires an anaplerotic ferredoxin (Fd)-reducing hydrogenase (Ech/Eha), in order to fixate carbon in anabolic processes. (C) Putative mechanisms for generation of F 420 H 2 . Complexes absent in A. fulgidus outlined in (B) : the electron bifurcating [Fe-Fe] hydrogenase (Huang et al., 2012 ), the Rnf complex (Biegel et al., 2011 ; Tremblay et al., 2013 ), the energy converting hydrogenase (Ech), (C) : the F 420 -reducing hydrogenase (Frh) (Thauer et al., 2010 ). In contrast to most methanogens and SRP, A. fulgidus possesses only one potential mechanism for ferredoxin generation from hydrogen; namely the Mvh:Hdl(Hdr) catalyzed reaction (Klenk et al., 1997 ; Thauer et al., 2010 ; Pereira et al., 2011 ) (Figure 6A ). The Mvh:Hdr hydrogenase is one of, so far, 4 perceived reaction mechanisms for the reduction of ferredoxin from H 2 during autotrophic growth (Fuchs, 2011 ) (Figures 6A,B ). In methanogens the Mvh:Hdr is the key enzyme of the recently named Wolfe cycle (Thauer, 2012 ), which catalyses the crucial bifurcation reaction that couples the first (Fd red is required for the fixation of CO 2 ) and last step of methanogenesis (reduction of heterodisulfide, CoM-S-S-CoB). No net Fd red is generated from this reaction, as generation of Fd red and heterodisulfide reduction are interdependent (Figure 6A ). In order to assimilate carbon—an anaplerotic hydrogenase—the energy-conserving membrane-associated hydrogenase (Ech) is required in these methanogens for additional generation of Fd red for anabolic processes (Figure 6B ) (Lie et al., 2012 ; Thauer, 2012 ). The presence of ferredoxin-binding sites ([4Fe-4S] clusters) in the structures of DsrAB indicate that soluble Fd red or a ferredoxin reductase complex may facilitate the steps of two-electron transfer to SO 2− 3 ; from redox state +IV, to +II and 0 (Schiffer et al., 2008 ; Oliveira et al., 2011 ). However, if the reduction of sulfite required 2 mol Fd red prior to reduction by 1 mol reduced DsrC, the pool of available oxidized DsrC would soon be depleted (Figure 6A ). Therefore, disulfide (X-S-S-X) would not be available for recycling Fd ox to Fd red by Mvh:Hdl mediated bifurcation.If the reduction of sulfite was dependent on only 1 Fd red , an anaplerotic hydrogenase would still be required for the generation of Fd red for subsequent CO 2 fixation; analogous to the Wolfe cycle (Thauer, 2012 ). During growth with hydrogen, the absence of an anaplerotic ferredoxin reductase in A. fulgidus requires multiple two-electron transfers for the reduction of sulfite by other mechanisms, either by an unknown electron donor or repetitive association, oxidation and dissociation of DsrC. According to this model Fd red is not a viable electron donor for reduction of sulfite in A. fulgidus during growth with H 2 . The electrons for reduction of sulfite must therefore be provided by the Vht hydrogenase (Figure 6A ). Hence, Fd red is probably utilized in biosynthesis rather than energy conservation during growth with T-H 2 , and may be a plausible explanation to the low transcriptional levels of mvh:hdl (Table 1D , Figure 6A ). Requirement of a Fd red -driven “confurcation” mechanism via the QmoABC complex for APS reduction in A. fulgidus would according to our model inhibit fixation of CO 2 (Figure 6 ). Accordingly, no growth with sulfate and hydrogen (S-H 2 /CO 2 ) has been observed for A. fulgidus (Stetter et al., 1987 ; Steinsbu et al., 2010 ). While growth on sulfate with hydrogen (S-H 2 ) does not occur; A. fulgidus is capable of utilizing sulfate as terminal electron acceptor with CO or formate as electron donors (Stetter et al., 1987 ; Henstra et al., 2007 ). The redox potential of CO/CO 2 indicates the capacity to reduce ferredoxin directly without the need for a bifurcation reaction (Thauer et al., 2007 ). The redox potential of formate/CO 2 is similar to that of hydrogen (Thauer et al., 2007 ), and requires bifurcation for the generation of Fd red . In addition, A. sulfaticallidus grows on S-H 2 /CO 2 (Steinsbu et al., 2010 ) and genome analysis of A. sulfaticallidus (Stokke et al., 2013 ), did not provide an alternative mechanism of Fd red generation. An unknown cytoplasmic formate dehydrogenase (AF1203-AF1202, Figure 4 ; Henstra et al., 2007 ) probably associates with the HdrA subunit of the Mvh/Hdl complex in order to catalyze the reduction of ferredoxin (Lie et al., 2012 ). In order to escape the proposed physiological impasse of Fd red as an intermediate of APS reduction, a formate dehydrogenase would also be needed to associate with QmoB (a HdrA homolog, Figure 4 ) and drive energy conservation by a confurcation mechanism. An unknown mechanism may also be facilitated by the gene product of AF1238 (Figure 4 ). Similarly, in A. sulfaticallidus , the Mvh hydrogenase subunit may, tentatively, form two complexes; one with HdrA and one with a homologous QmoB. Considering growth with formate and sulfate, and the similar genomic composition of A. fulgidus and A. sulfaticallidus ; the most plausible explanation for the inability of A. fulgidus to grow on S-H 2 is a regulatory link between hydrogen and observed down-regulation of a pyrophosphatase gene ( ppx , AF0756). Additionally, we observed minor down-regulation of genes corresponding to the Sat-ORF2-AprAB operon (Table 1A , Figure 5 ). A reduced expression of Ppx would inhibit or limit the formation of APS, as the pyrophosphatase reaction drives the total reaction to completion (Peck, 1962 ). Uniquely for A. fulgidus , the ppx gene is close to the inversely induced membrane-bound Hdr gene ( hdrDE ; AF0755) located on the opposite strand (Tables 1A,C , Figure 4 ), which may suggest a regulatory link. Therefore, the inability of A. fulgidus to grow with sulfate and H 2 may relate to transcriptional regulation rather than a physiological limitation. Clearly, further biochemical characterization is needed to verify the proposed regulatory mechanism. There is also a need to characterize a mechanism for growth with sulfate and formate for A. fulgidus , and S-H 2 for A. sulfaticallidus (Figure 6A ). Generation of F 420 H 2 in the absence of a dedicated hydrogenase (Frh) Archaeoglobus fulgidus lacks the cytoplasmic F 420 -reducing hydrogenase (FrhABG) that catalyzes the reduction of F 420 in most methanogens (Alex et al., 1990 ; Thauer et al., 2010 ) (Figure 6C ) Therefore, mechanism for generating the reduced F 420 H 2 required for carbon fixation through the reductive acetyl-CoA is unknown (Figure 5 ). A negligible role of NAD(P)H is supported by low expression levels of F 420 H 2 :NADP + oxidoreductase genes in our study (AF0892, AF1209; Table S3 ). Independently of the Fpo complex and Frh hydrogenase; the FpoF subunit is shown to reduce F 420 coupled with oxidation of Fd red in M. mazei (Welte and Deppenmeier, 2011 ). It is therefore possible that FqoF in A. fulgidus catalyzes a similar mechanism for the generation of reduced F 420 (Figure 6C , fqoF : AF1833, Table S2 ). The required Fd red must be provided by the bifurcation reaction facilitated by the Mvh:Hdl hydrogenase (Figures 6A,B ). As discussed in the previous section, the main route of energy conservation is probably provided by the periplasmic hydrogenase. The low transcriptional levels of mvh:hdl may be a reflection of translational levels of Mvh:Hdl hydrogenase, if an alternative pathway of F 420 reduction is present independently of Fd red . Vorholt et al. ( 1995 ) suggested the possibility that reduced F 420 H 2 may be generated by reverse electron flow through menaquinol oxidation. The Fqo complex, including the FqoF subunit, is also a potential MQH 2 oxidase (Figures 5 , 6C ). The constitutive expression of the entire Fqo complex indicates that the complex may function in reverse as a pmf (μΔH + )-dependent menaquinol—F 420 oxidoreductase, where the menaquinol (E°′ −75 mV) may donate electrons for the reduction of F 420 (E°′ −360 mV). The resulting positive redox potential (E°′ +285 mV) may be overcome in a process assisted by the consumption of proton gradient. Further support for such a mechanism can be found in the common ancestry of the respiratory complex I and Ech hydrogenase (Hedderich, 2004 ; Moparthi and Hägerhäll, 2011 ). In mitochondria and iron-oxidizing Thiobacillus ferrooxidans the NAD(P)H-oxidoreductase (Complex I) has been shown to catalyze this reaction at the expense of ATP hydrolysis, which is perceived to be coupled to generation of a pmf by reversal of ATPase (Chance and Hollunger, 1960 ; Vinogradov, 1998 ; Elbehti et al., 2000 ). The Eha/Ech hydrogenase activity is linked to the reduction of ferredoxin (Figure 6B ) (Meuer et al., 1999 ), as the reverse electron flow of the Ech dehydrogenase catalyzes the formation of CO from CO 2 , and H 2 by consumption of pmf (Bott et al., 1986 ; Bott and Thauer, 1987 ; Lie et al., 2012 ). Therefore, the reduction of Fd red (E°′ −500 mV) with H 2 (E°′ −414 mV; −300 mV at 10 Pa H 2 ) results in a positive redox potential (E°′) of at least +86 mV (or +200 mV at 10 Pa H 2 ) and is considered possible with the utilization of a proton gradient (Figure 6B ) (Thauer et al., 2007 ). It remains to be shown in A. fulgidus whether it is possible to drive the reduction of F 420 by MQH 2 oxidation (E°′ +280 mV) and a proton gradient by e.g., constructing deletion mutants of fqo genes encoding MQ interacting components, or biochemical characterization by inverted vesicles (Baumer et al., 2000 ) coupled to ATP hydrolysis. Co-assimilation of organic substrates The ambiguous roles of genes corresponding to fatty acid metabolism (COG; I, Figures 2 , 7 ) during growth with H 2 is discussed. Putative co-assimilation of organic acids is considered more likely than a homologous secondary carbon fixation pathway, although, both mechanisms may contribute to conserve Fd red during autotrophic growth with H 2 . Figure 7 Genes corresponding to fatty acid and propionate metabolism, compared with putative steps in the 4-hydroxybutyryl (4HB) and 3-hydroxyproprionyl (3HP) pathway of autotrophic Creanarchaea . Homologous genes correspond to several steps in the 4HCD and 3HP pathway (intermediate names in gray), and may not be easily-distinguished by homology alone. Genes of the acetyl-CoA pathway are expressed at a higher level than the key enzymes of putative 4-hydroxybutyryl-CoA dehydratase (4HCD), which are not uniformly induced. Genes that are not induced by T-H 2 /CO 2 are underlined, for up-regulated genes—the number of up-regulated vs. total number of homologs is indicated (genes are annotated in Table S1 ). The genes of methylmalonyl metabolism (AF2215-AF2219, Table S1b , Figure 7 ) were continuously highly expressed. These enzymes may serve as a pathway of propionate degradation (Takaki et al., 2010 ; Moon et al., 2012 ). During growth with T-H 2 /CO 2 , several genes related to fatty acid biosynthesis and metabolism were induced (COG; I, Figure 2 , Table S1a ). Expression of the genes may be affected by trace amounts of fatty acids present in yeast extract (0.03% weight) and points to a potential for fatty acid scavenging/co-assimilation of organic substrates during autotrophic growth (Klenk et al., 1997 ; Zarzycki and Fuchs, 2011 ). Recently, Parthasarathy et al. ( 2013 ) demonstrated induced activity of phenylalanine degradation in the presence of yeast extract. Despite amino acids being a major component of yeast extract, none of the putative genes reported in the previous study (Table 2 in Parthasarathy et al., 2013 ) were induced during T-H 2 /CO 2 growth in this study. Thus, there is no uniform induction of putative scavenging mechanisms for organic carbon during autotrophic growth. The genes related to fatty acid biosynthesis also encode enzymes in the 3-hydroxypropionate/4-hydroxybutyrate (3HP/4HB) cycle identified in Metallosphaera sedula , and could potentially represent a secondary carbon fixation pathway in A. fulgidus (Berg et al., 2007 ; Estelmann et al., 2011 ). The genome of A. fulgidus includes 3 homologs of the 4-hydroxybutyryl-CoA dehydratase ( 4hcd ), (AF0333, AF0885 and AF1027, Table S1c ), which is a key enzyme of the 4-hydroxybutyrate carbon dioxide assimilation pathway (Figure 7 ) (Berg et al., 2007 ). During growth with T-H 2 /CO 2 , one of the 4hcd homologs (AF0885) was induced (1.5 fold). The induced gene displayed average signal intensity (1.3), and was expressed at a similar level as the two other unregulated homologs (AF0333 and AF1027). The differential transcriptional expression of 4hcd (Msed1321) in M. sedula was related to autotrophic vs. heterotrophic growth and resulted in a more than 7 fold up-regulation (Auernik and Kelly, 2010 ). The presence of a 3HP/4HB cycle was refuted by Estelmann et al. ( 2011 ) by a subsequent study on the obligate autotroph “ A. litotrophicus ,” where enzyme activity of key processes could not be detected. The presence of 5 homologs of A. fulgidus 4hcd in the genome of Desulfatibacillum alkenivorans suggest that this enzyme is involved in alkene degradation in these species (Estelmann et al., 2011 ). Analogously to A. fulgidus , the facultative autotroph and fatty-acid and alkene degrading D. alkenivorans utilizes the bacterial acetyl-CoA/Wood-Ljungdal pathway (So and Young, 1999 ; Callaghan et al., 2012 ), indicating physiological similarities between the distantly related species. Alternative mechanisms for up-regulation of homologous of propionate and beta-oxidation may be co-assimilation of organic substrates that may supplement the reductive acetyl-CoA pathway. In the light of the considerations of the role of Fd red during growth with T-H 2 /CO 2 , this may provide a significant advantage by supplementing reduction of CO 2 with the uptake of reduced organic acids. The constitutively highly expressed EtfAB may also provide a source of Fd red by an unknown bifurcation reaction from ambient fatty or amino acids (Buckel and Thauer, 2013 ; Parthasarathy et al., 2013 ). In summary, our data may add support to the theory that the 3HB/4HP cycle may have originated from a heterotrophic pathway; or as a co-assimilatory pathway in Archaea (Fuchs, 2011 ; Zarzycki and Fuchs, 2011 ). Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest."
} | 9,728 |
23035866 | null | s2 | 3,419 | {
"abstract": "Biofilm formation by Vibrio fischeri is a complex process involving multiple regulators, including the sensor kinase (SK) RscS and the response regulator (RR) SypG, which control the symbiosis polysaccharide (syp) locus. To identify other regulators of biofilm formation in V. fischeri, we screened a transposon library for mutants defective in wrinkled colony formation. We identified LuxQ as a positive regulator of syp-dependent biofilm formation. LuxQ is a member of the Lux phosphorelay and is predicted to control bioluminescence in concert with the SK AinR, the phosphotransferase LuxU and the RR LuxO. Of these, LuxU was the only other regulator that exerted a substantial impact on biofilm formation. We propose a model in which the Lux pathway branches at LuxU to control both bioluminescence and biofilm formation. Furthermore, our evidence suggests that LuxU functions to regulate syp transcription, likely by controlling SypG activity. Finally, we found that, in contrast to its predicted function, the SK AinR has little impact on bioluminescence under our conditions. Thus, this study reveals a novel connection between the Lux and Syp pathways in V. fischeri, and furthers our understanding of how the Lux pathway regulates bioluminescence in this organism."
} | 318 |
24492620 | PMC4438999 | pmc | 3,420 | {
"abstract": "Polymers bearing dynamic covalent bonds may exhibit dynamic properties, such as self-healing, shape memory and environmental adaptation. However, most dynamic covalent chemistries developed so far require either catalyst or change of environmental conditions to facilitate bond reversion and dynamic property change in bulk materials. Here we report the rational design of hindered urea bonds (urea with bulky substituent attached to its nitrogen) and the use of them to make polyureas and poly(urethane-ureas) capable of catalyst-free dynamic property change and autonomous repairing at low temperature. Given the simplicity of the hindered urea bond chemistry (reaction of a bulky amine with an isocyanate), incorporation of the catalyst-free dynamic covalent urea bonds to conventional polyurea or urea-containing polymers that typically have stable bulk properties may further broaden the scope of applications of these widely used materials.",
"discussion": "Discussion Integrating dynamic moieties to polymers, especially those capable of dynamic covalent bonding in widely-used conventional polymers are of tremendous interest. In one particular application, for example, materials bearing such dynamic moieties may theoretically self-heal for unlimited times if the healing is directed by the reversible exchange of dynamic covalent bonds. However, bringing dynamic properties to conventional polymers often involves special dynamic moieties 13 – 17 that may require tedious synthesis and/or use of external stimuli such as catalyst 35 , 36 or high temperature 37 , 38 for dynamic property activation and control. In this study we report the design of dynamic polyureas and poly(urethane-urea)s by attaching bulky substituents to the urea nitrogen to create the so-called hindered urea bonds (HUBs, Fig. 1 ). As HUBs are synthesized through the reaction of isocyanate and amine, conventional polyureas and poly(urea-urethane)s with stable bulky properties can thus be readily made dynamic by replacing regular amine with bulky amines (amines containing bulky substituents). By screening five HUBs with different substituent bulkiness, we identified N -tertbutyl- N -ethylurea (TBEU) as a promising HUB and successfully demonstrated its application in making dynamic polymers ( Fig. 2e ) and self-healing materials ( Fig. 4d ). Polymer chain reshuffling was observed for a linear polyurea based on TBEU ( Fig. 2e ). We incorporated TBEU moieties into a cross-linked poly(urethane-urea) to obtain catalyst-free self-healing materials under mild condition with good mechanical strength, dimensional rigidity and chemical stability. TBEU has a large K eq for retaining strong bonding and a reasonably large k −1 for efficient dynamic bond exchange under mild conditions ( Table 1 ). Very bulky N -substituent in HUBs may result in faster dynamic bond exchange, as shown in the case of TMPCA, but its weak bond strength (smaller K eq , Table 1 ) makes it less favoured for self-healing applications ( Supplementary Fig. 18 ). HUBs present a number of desirable properties for the synthesis of dynamic and self-healing polymers. They can be easily synthesized through the reaction of an isocyanate and a hindered amine, both of which are widely available and inexpensive. The dynamic properties of HUBs can be well controlled by adjusting bulkiness of the substituents. HUBs possess a hydrogen-bonding motif via its urea bond that can increase the mechanical strength of polymers, a property that most of other dynamic covalent chemistries lack. We anticipate that HUB structure and chemistry can be readily integrated in the design of a wide range of materials, bringing modular and tuneable dynamic properties to conventional polyureas and urea-containing polymers."
} | 940 |
26610025 | PMC4878690 | pmc | 3,421 | {
"abstract": "Summary Nitrification is a two-step process where ammonia is considered to first be oxidized to nitrite by ammonia-oxidizing bacteria (AOB) and/or archaea (AOA), and subsequently to nitrate by nitrite-oxidizing bacteria (NOB). Described by Winogradsky already in 1890 1 , this division of labour between the two functional groups is a generally accepted characteristic of the biogeochemical nitrogen cycle 2 . Complete oxidation of ammonia to nitrate in one organism ( co mplete amm onia ox idation; comammox) is energetically feasible and it was postulated that this process could occur under conditions selecting for species with lower growth-rates but higher growth-yields than canonical ammonia-oxidizing microorganisms 3 . Still, organisms catalysing this process have not yet been discovered. Here, we report the enrichment and initial characterization of two Nitrospira species that encode all enzymes necessary for ammonia oxidation via nitrite to nitrate in their genomes, and indeed completely oxidize ammonium to nitrate to conserve energy. Their ammonia monooxygenase (AMO) enzymes are phylogenetically distinct from currently identified AMOs, rendering recent acquisition by horizontal gene transfer from known ammonia-oxidizing microorganisms unlikely. We also found highly similar amoA sequences (encoding the AMO subunit A) in public sequence databases, which were apparently misclassified as methane monooxygenases. This recognition of a novel amoA sequence group will lead to an improved understanding on the environmental abundance and distribution of ammonia-oxidizing microorganisms. Furthermore, the discovery of the long-sought-after comammox process will change our perception of the nitrogen cycle."
} | 432 |
28635267 | PMC5520101 | pmc | 3,422 | {
"abstract": "The\nhigh quantum efficiency of photosynthetic reaction centers (RCs) makes\nthem attractive for bioelectronic and biophotovoltaic applications.\nHowever, much of the native RC efficiency is lost in communication\nbetween surface-bound RCs and electrode materials. The state-of-the-art\nbiophotoelectrodes utilizing cytochrome c (cyt c )\nas a biological wiring agent have at best approached 32% retained\nRC quantum efficiency. However, bottlenecks in cyt c -mediated electron transfer have not yet been fully elucidated. In\nthis work, protein film voltammetry in conjunction with photoelectrochemistry\nis used to show that cyt c acts as an electron-funneling\nantennae that shuttle electrons from a functionalized rough silver\nelectrode to surface-immobilized RCs. The arrangement of the two proteins\non the electrode surface is characterized, revealing that RCs attached\ndirectly to the electrode via hydrophobic interactions and that a\nfilm of six cyt c per RC electrostatically bound\nto the electrode. We show that the additional electrical connectivity\nwithin a film of cyt c improves the high turnover\ndemands of surface-bound RCs. This results in larger photocurrent\nonset potentials, positively shifted half-wave reduction potentials,\nand higher photocurrent densities reaching 100 μA cm –2 . These findings are fundamental for the optimization of bioelectronics\nthat utilize the ubiquitous cyt c redox proteins\nas biological wires to exploit electrode-bound enzymes.",
"conclusion": "Conclusions This work shows that\ncyt c electrostatically binds and saturates a Ag R |mSAM surface, forming a full electroactive monolayer. In\ncontrast, the RC adsorbs on the electrode via hydrophobic regions\nthat are normally embedded in the membrane interior. The finding that\nhigh cyt c loading yields larger photocurrents and\nmore positive onset potentials is important for improving the short-circuit\ncurrents and open-circuit voltages of cyt c -based\nbiophotovoltaic systems. We suggest this is due to an interconnected\ncyt c layer, which acts as an electron-funneling\nantenna and electron-storing capacitor for delivery of electrons to\nthe RC. Our data show that cyt c mobility is crucial\nfor ET from the cyt c to the surface-bound RCs, likely\ndue to reorientation of the heme group from the electrode surface\nfor lateral cyt c– cyt c exchange,\nor toward the photo-oxidized RC primary electron donor cofactors.\nFinally, we have obtained definitive evidence on the central role\nof cyt c -mediated ET to RCs immobilized on electrodes.\nWe show the cyt c indeed gates ET to the RC, via\nthe shift in the half-wave potential that is consistent with the positive\nshift in cyt c formal potential. Collectively,\nthese findings suggest that further improving both the coverage and\nthe mobility of surface-adsorbed cyt c will improve\nthe electrical communication between the electrode and the RC, thus\nenhancing the photocurrent and onset potential. The maximum RC turnover\nrates achieved in the present study (19 e – s –1 ) are still orders of magnitude lower than those determined\nfor solubilized RCs (1000 e – s –1 ), showing that cyt c -mediated ET, even when optimized\nto the extent described above, is still insufficient to meet the demands\nof RC turnover. Thus, it is necessary to further engineer the cyt c or substitute it with a more efficient ET mediator that\nis better suited to interface the RC to the electrode.",
"introduction": "Introduction Photosynthetic reaction centers (RCs)\ndrive a photochemical charge separation that forms the energetic basis\nof most life on our planet. They do so with a near-unity quantum yield,\ntransforming almost every absorbed photon into a displaced high-energy\nelectron that drives biochemical processes within the organism. In\nthe bacterial species Rhodobacter (Rba.) sphaeroides , the key photovoltaic machinery is located within a 6 nm diameter\ntransmembrane protein that is essentially a tractable nanoscale solar\nbattery that can be removed from the bacterial cell and interfaced\nwith man-made electrode materials. This has been achieved using a\nvariety of strategies on metal electrodes, 1 − 10 resulting in composite biohybrid photoelectrodes that support current\ndensities of up to ∼400 μA cm –2 . These\nhalf-cells form the platform for biophotovolatics, which seek to exploit\nphotosynthetic pigment proteins to develop alternative cheap and sustainable\nmaterials for solar energy conversion. 11 , 12 In-depth summaries\nof progress in this field of RC-based biophotolectrodes, including\nparallel work using photosystem I (PSI) and photosystem II (PSII)\nRCs, have been published in recent years. 13 − 18 A key aspect in the construction of biophotoelectrodes is\nthe interface between the photovoltaic protein and the adjacent working\nelectrode. Cytochrome c (cyt c )\nhas been utilized as an electron-transfer (ET) element between electrodes\nand surface-bound enzymes, such as cyt c peroxidase, 19 , 20 bacterial RCs, 1 , 2 , 4 , 7 PSI, 21 , 22 and PSII RCs, 23 supporting turnover rates approaching 71 s –1 and peak photocurrents of up to 416 μA cm –2 . 1 , 21 , 22 , 24 − 26 1 , 21 , 22 , 24 − 26 In comparison, net photocurrents of 322 μA cm –2 have been obtained for PSI connected to planar glassy carbon electrodes\nwith engineered redox hydrogel films, but at much larger RC turnover\nrates of up to 335 s –1 that surpass those measured\nin vivo. 16 This suggests that there is\nan untapped capacity for photocurrent output from biophotoelectrodes\nutilizing cyt c to aid ET to RCs. The benefits\nof horse heart cyt c -mediated ET to bacterial RCs\nhave been characterized on functionalized gold electrodes, resulting\nin photocurrent densities of a few μA cm –2 and RC turnovers on the order of tens per second. 1 , 4 PSI\nRCs have also been interfaced with working electrodes using cyt c , either to form multilayer systems that produce photocurrent\ndensities on the order of a few μA cm –2 on\nflat substrates, or up to 150 μA cm –2 on mesoporous\nsubstrates. 21 , 22 These reports suggested that\nslow heterogeneous ET from the electrode to the cyt c limits RC turnover and thus photocurrents. Furthermore, both works\nsuggest that the natural binding affinity between cyt c and the RC donor side allows cyt c to effectively\nact as a docking site for the oriented attachment of the RC onto immobilized\nmolecules of cyt c . 1 , 21 Cytochromes\nare a class of heme-based proteins, some of which perform ET functions\nin both respiratory and photosynthetic electron transport chains. 27 Class I cytochromes of type c (cyt c ) are defined by a covalently bound porphyrin,\nwhich is partially exposed to the solution through a crevice, and\nhave an amino acid sequence that is heavily preserved such that the\nstructure, electrochemical formal potential and binding properties\nare strongly conserved. 27 , 28 This allows the substitution\nof cyt c between species as divergent as photosynthetic\nbacteria and mammals with negligible effects on enzyme turnover capacity. 21 , 27 , 29 In purple photosynthetic bacteria,\nthis protein (termed cyt c 2 ) is typically\nlocated in the periplasmic space or loosely attached to the periplasmic\nside of the membrane 28 , 30 and shuttles electrons from the\nubiquinol-cyt c oxidoreductase (cyt bc 1 complex) to the RC. 31 There,\nit reduces the photo-oxidized primary electron donor bacteriochlorophyll\npair (P + ) produced by photochemical charge separation within\nthe RC. This charge separation also produces a reduced, mobile ubiquinone\nspecies at the so-called Q B site, which may be substituted\nby a water-soluble ubiquinone analogue to optimize photocurrents in\nan RC-based photocathode. 32 Cyt c has long been at the heart of bio-electrochemistry, beginning\nwith studies of mediated ET using soluble mediators such as 4,4 bipyridyl\nto direct ET on functionalized electrodes. 33 − 37 The introduction of self-assembled monolayers (SAMs)\non bare metal electrodes has enabled diffusionless voltammetry for\nmore accurate and physiologically relevant characterization of cyt c ET phenomena by promoting strong binding of the electroactive\ncytochrome. 38 , 39 The use of mixed SAMs (mSAMs)\nwith various alkyl chain lengths and terminal functional groups has\nfurther increased the possibility of fine tuning the relevant parameters\nsuch as the magnitude of the electric field at the SAM–cyt c interface and the degree of protonation of the SAM, which\nhas a strong impact on the ET properties of the bound protein. 40 This control may be advantageous\nin building bioelectrodes for designing faster ET or in improving\nprotein binding. In previous work, we have employed bare metal\nelectrodes to bind both cyt c and RCs in a manner\nthat results in significant photocurrents on the order of hundreds\nof μA cm –2 . 2 , 22 However, cytochromes\nare known to bind to bare metals in a manner that hampers ET, not\nonly limiting subsequent ET to the RC, but also making it difficult\nto electrochemically characterize the surface-adsorbed proteins and\noptimize ET bottlenecks. In the present work, we resolve these issues\nby functionalizing the bare metal surface using a mSAM that promotes\neffective SAM–cyt c binding, resulting in\ndistinct faradaic currents ( Figure S1 ),\nfast ET at the SAM–cyt c interface, as well\nas moderate RC loading. 24 This facilitated\nprotein film voltammetry (PFV) of the electroactive cyt c , which yielded key parameters such as the reversibility and rate\nof ET as well as the quantity of electroactive cytochrome. In conjunction\nwith the recently coined technique “protein film photoelectrochemistry”, 41 this has allowed us to simultaneously characterize\nthe nature of cyt c -mediated ET and activity of the\nlight-dependent enzyme. This is important for the design and full\nexploitation of a high-turnover enzyme/electrode system that uses\ncyt c as a biological wire. We find that optimal\nphotocurrents of ∼100 μA cm –2 were\nobtained in low ionic strength buffers and high cyt c loading, resulting in a higher RC turnover. Furthermore, we deduced\nthat cyt c mobility plays a crucial role in mediating\nET between cyt c and the RC.",
"discussion": "Discussion Physiological Relevance The mechanism of cytochrome\nET between transmembrane proteins in biological ET chains has been\nproposed to involve one or a combination of three possible processes:\n(1) complete three-dimensional diffusion whereby the cytochrome detaches\nfrom the membrane/protein surface into bulk solution to diffuse to\nthe next membrane-embedded redox partner, (2) two-dimensional diffusion\non the membrane surface via rotational mobility and/or lateral diffusion,\nand (3) an immobilized film configuration in which the cytochromes\nact as a network of relay centers between transmembrane redox proteins. 51 Self-exchange rates ( k ex ) for horse heart cyt c of 10 3 s –1 at a low ionic strength (0.1 M NaCl) and 10 4 s –1 at a high ionic strength (1 M NaCl)\nhave been determined. 52 However, whether\nlateral ET occurs in a monolayer of cyt c on a membrane\nsurface is not known, but has been shown for three-dimensional multilayers. 46 When adsorbed to a negatively charged\nSAM in man-made systems, cyt c undergoes a redox\nshift of −45 mV relative to the solution value, which is comparable\nto that reported for cyt c bound to biological membranes\ndue to stabilization of the cyt c 3+ form. 35 Bowden et al. have suggested that surfaces with\na higher defect density, resulting in irregularities in the chemical\nand topographical texture of the mSAM on flat gold, more closely resemble\nthe natural membrane-protein environment and binds ET proteins with\noptimal electronic coupling. 40 If we also\nconsider that the rough silver topography promotes better protein\nadsorption versus a smooth topology, 53 it\nmay reveal why the Ag R |mSAM is such a good candidate for\nbinding cyt c and transmembrane proteins with fast\nET, despite the added tunneling barrier inferred by the 11-carbon\nalkyl chain. In the absence of precise information on the structure\nof the RC–cyt c complex on the electrode surface,\nit is not possible to decisively discriminate between the different\nmodels for ET on our electrodes. However, our data support the model\nof a cyt c film on the membrane surface, with rotational\nand perhaps lateral mobility playing a role in ET, as illustrated\nin the TOC."
} | 3,098 |
39660009 | PMC11631310 | pmc | 3,423 | {
"abstract": "Abstract Ferruginous conditions prevailed through Earth’s early oceans history, yet our understanding of biogeochemical cycles in anoxic iron-rich, sulfate-poor sediments remains elusive in terms of redox processes and organic matter remineralization. Using comprehensive geochemistry, cell counts, and metagenomic data, we investigated the taxonomic and functional distribution of the microbial subsurface biosphere in Lake Towuti, a stratified ferruginous analogue. Below the zone in which pore water becomes depleted in electron acceptors, cell densities exponentially decreased while microbial assemblages shifted from iron- and sulfate-reducing bacterial populations to fermentative anaerobes and methanogens, mostly selecting Bathyarchaeia below the sulfate reduction zone. Bathyarchaeia encode metabolic machinery to cycle and assimilate polysulfides via sulfhydrogenase, sulfide dehydrogenase, and heterodisulfide reductase, using dissimilatory sulfite reductase subunit E and rubredoxin as carriers. Their metagenome-assembled genomes showed that carbon fixation could proceed through the complete methyl-branch Wood-Ljungdahl pathway, conducting (homo)acetogenesis in the absence of methyl coenzyme M reductase. Further, their partial carbonyl-branch, assumed to act in tetrahydrofolate interconversions of C1 and C2 compounds, could support close interactions with methylotrophic methanogens in the fermentation zone. Thus, Bathyarchaeia appeared capable of coupling sulfur-redox reactions with fermentative processes, using electron bifurcation in a redox-conserving (homo)acetogenic Wood-Ljungdahl pathway, and revealing geochemical ferruginous conditions at the transition between the sulfate reduction and fermentation zone as their preferential niche.",
"introduction": "Introduction Ferruginous (iron-rich, sulfate-poor) conditions dominated much of Earth’s early oceans history [ 1 ] before sulfate became available and abundant [ 2 ]. While the geochemistry of past marine environments has been studied extensively, metabolic activities of a microbial biosphere slowly buried in the subsurface remain elusive in sedimentary records [ 3 , 4 ]. Recognizing the major biogeochemical transitions, such as oceanic bottom water oxygenation [ 5 ] and the rise of seawater sulfate [ 6 ] during the Proterozoic Eon, requires empirical constraints on microbial processes in terms of redox changes and organic matter (OM) mineralization of sediment substrates during early diagenesis [ 7 ]. Although ferruginous geochemical conditions are common in sediment pore water below the depth of sulfate depletion [ 8 ], modern sedimentary environments in which to study microbial communities and metabolic processes analogous to those that operated in Proterozoic oceans enriched in iron oxides and devoid of sulfate, are scarce [ 9 , 10 ]. Moreover, modern sediments contain substantially more primary OM and their iron sources are rather of detrital than hydrothermal origin, usually orienting the system towards methanogenesis as the main remineralization process [ 11 ]. Ferruginous sediments are deposited in the Malili Lakes, a chain of five interconnected tectonic lakes hosted in ultramafic rocks on Sulawesi Island, Indonesia ( Fig. 1A ). Weathering of lateritic soils in the catchment ( Fig. 1B ) supplies considerable amounts of iron (oxyhydr)oxides [ 12 ], but very little sulfate to the lakes [ 13 ]. These iron oxides scavenge phosphorus in the water column and limit its productivity [ 14 ]. Lake Towuti (2.5° S, 121.3° E), the largest of the Malili Lakes, is currently stratified, displaying oligotrophic conditions with persistent ferruginous anoxia in its bottom waters ( Fig. 1C ). A substantial part of the sinking OM is remineralized through microbial reduction of ferrihydrite [ 15 ], and potentially methanogenesis [ 11 ], suggesting analogies to depositional processes in early ferruginous ecosystems [ 9 ]. Previous geomicrobiological investigations showed that, despite extremely low sulfate concentrations (<20 μM), biogeochemical processes related to sulfur, iron and methane co-occur in sediments [ 10 , 16 ], which thereby may provide clues to early diagenesis by selective benthic microbial communities. Figure 1 \n Site description of Lake Towuti. ( A ) World map displaying the location of Sulawesi Island (square), with close-up on the Indonesia archipelago and location of the Malili Lake system (square). ( B ) Map of Sulawesi illustrating the geological context of the Malili Lake system (square). ( C ) Bathymetric map of Lake Towuti with position of the coring site (star) corresponding to 156 m water depth; oxygen, iron, pH profiles and sulfate concentrations of Lake Towuti’s water column [ 15 ]. The figure is modified from [ 17 ]. Using metagenomic data, we investigate Lake Towuti’s subsurface biosphere in terms of environmental selection, and identify metabolic traits that reflect specific adaptation to anoxic ferruginous conditions. We report bulk sediment, methane and pore water geochemical data along a sediment core spanning ~3000 years of sediment history [ 18 ]. Using sulfate reduction rate (SRR), cell count, 16S rRNA gene and metagenomic sequencing data, we characterize the abundance, taxonomic distribution, and metabolic functions of microbial communities, and trace their assembly during burial in ferruginous deposits. By adding taxonomy of functional marker genes and analysis of metagenomes-assembled genomes (MAGs), we determine which metabolic features among the main microbial clades relate to processes of metal and sulfate reduction, OM remineralization, and methanogenesis. Finally, we detail the metabolic machinery that selects Bathyarchaeia in sediments and consider its relevance to a primeval biosphere in Earth’s ferruginous oceans.",
"discussion": "Discussion In Lake Towuti’s anoxic ferruginous sediments, FeRB and SRB exhibited a certain degree of metabolic versatility, allowing them to switch from respiration of reactive ferric minerals and SO 4 2− to fermentation ( Figs. 3 - 5 ). After pore water depletion in electron acceptors, we expected that FeRB and SRB would be gradually replaced by primary (i.e. acidogens) and secondary (i.e. acetogens) fermenters [ 64 ]. Still, the Acidobacteriota (class Aminicenantia), which are predicted to act as primary degraders of buried OM during acidogenesis, fermenting proteinaceous substrates and different sugars producing hydrogen and acetate as by-products [ 65 ], were identified as potential FeRB ( Fig. 5D ). Redox cofactors ( trx , rbx , grx , fdx ) are not ferric iron-specific enzymes, but also allow for the reduction of other organic or inorganic compounds [ 54 ], which implies a certain degree of metabolic versatility among FeRB and SRB. Thus, we suggest that microbial reduction of mineral ferric iron could proceed at slow rates [ 17 ] via redox cofactors, enhancing hydrogenase ( frh , mvh , hdr ) activity [ 66 ] alongside electron bifurcation ( Rnf , Nuo ) and fermentation ( Fig. 5 ). Similarly, the Nitrospirota (class Thermodesulfovibriona) and the Desulfobacterota, while being SRB ( dsrAB , aprAB ), were metabolically capable of energy conservation via a partial hydrogenotrophic and acetoclastic ( pta , akn ) oxidative W-L pathway. Several redox cofactors ( trx , rbx , grx ), which are hypothesized to mitigate oxidative stress during sulfate reduction and, by extension, under sulfate-poor ferruginous conditions [ 67–69 ], could also function with hydrogenases ( mvh , hdr ) and electron bifurcation complexes ( Ech , Rnf , Nuo ) to provide redox-conserving energy. As SRB are known to be fast degraders of labile OM [ 43 , 70 ], the aforementioned metabolic traits suggest that alternative electron donors may be used below the depth of sulfate depletion, allowing them to persist as fermenters outside their geochemical niche ( Fig. 5D ). During fermentation stages, C1 and C2 compounds, like formate, methanol, CO − , and acetate, are remineralized into methanogenic substrates or taken up by non-syntrophic anaerobes [ 64 , 71 ]. Among those, the Hadarchaeota became increasingly abundant below 20 cmblf ( Fig. 2B ). This phylum exhibits metabolic traits centered on H 2 and CO − oxidation combined with a partial W-L pathway [ 72 ]. Similarly, the Chloroflexota (class Dehaloccoidia) have a fermentative lifestyle with a reductive W-L pathway to facilitate cell carbon synthesis from acetate ( acs , acss2 ). The fact that these clades remained relatively abundant down to 50 cmblf suggested that (homo)acetogens could outperform methanogens in the uptake of C1 and C2 metabolites [ 73 ] in either an oxidative or reductive W-L pathway [ 61 , 62 ]. The methanogens identified ( Supplementary Fig. S1 ) appeared to be mostly methylotrophic (Methanomassiliicoccales, Methanofastidiosales, Methanomethyliales), with fewer taxa related to hydrogenotrophic methanogenesis (e.g. Methanomicrobiales) and methanotrophs (ANME-1). We detected mcrA genes in three specific MAGs ( Fig. 5D ; Supplementary Fig. S12 ) that were concomitant either with mta and mtb genes (i.e. methylotrophic) in Candidatus Methylarchaeales [ 50 ], or with mtr genes (i.e. hydrogenotrophic) in Methanomicrobiales [ 63 , 74 ]. Despite the relatively low abundance of methanogens, methane accumulated to 550 μM at 50 cmblf, with upward diffusion apparently feeding ANME-1 populations [ 10 , 11 ] in the upper 20 cmblf ( Figs. 2A- 2B ). We further suspected that C1 compounds and molecular hydrogen sustaining methylotrophic/hydrogenotrophic methanogenesis might be produced during OM fermentation by Bathyarchaeia [ 57 , 75 ], which constitute the majority of the microbial assemblage below 20 cmblf ( Fig. 2B ). Bathyarchaeia can utilize a wide range of labile and refractory organic compounds, e.g. proteins, carbohydrates, volatile fatty acids, short-chain alkanes, methylated compounds, cellulose, lignin [ 57 , 76 ]. At the same time, they also use multiple C1 compounds [ 49 ] and drive CO 2 fixation with molecular hydrogen to convert it to acetyl-CoA and acetate via the W-L pathway. The question arose as to whether Bathyarchaeia play an active role in syntrophic interactions in lake sediments, as already proposed for paddy fields [ 77 , 78 ], peatlands [ 58 , 75 ], and estuaries [ 79 , 80 ], or whether they are linked to the plethora of reactive iron minerals (>10 weight %) and scarcity of sulfate in Lake Towuti, or simply to OM types derived from tropical vegetation in the catchment [ 3 , 11 ]. In the maar Potrok Aike [ 47 ], Bathyarchaeia were mainly found in deep layers composed of organic-poor detrital mafic material while Atribacteriota, Chloroflexota, Aminicenentia formed the predominant subsurface biosphere in methane-rich sediments. In sulfate-rich (~2 mM) Lake Cadagno [ 81 ], Bathyarchaeia were abundant at shallow depths around the sulfate–methane transition and prevailed below the depth of sulfate depletion. Overall, the geographical distribution of Bathyarchaeia has been proposed to result from salinity gradients and OM types, enabling to trace their orders in terms of sediment provenance [ 49 , 57 , 75 , 82 ], i.e. soils (40CM-2-53-6), groundwater (RBG-16-48-13), hot springs (B24), estuaries (TCS64), cold marine sediments (B26-1), and hydrothermal vents (EX4484–135, B25). Apart from hydrothermal vents, Lake Towuti’s ferruginous sediments harbored groups of all origins ( Supplementary Figs. S9 and S10 ), suggesting that ferruginous geochemical conditions constitute a preferential niche for Bathyarchaeia. Bathyarchaeia thrive below the sulfate reduction zone Low but sustained SRR values (≤ 2 nmol × cm −3 × day −1 ) coincided with the minor detection of protein-encoding genes ( asr, phs/psr ) involved in the reduction of sulfur intermediates (SO 3 2− , S 2 O 3 2− , S N 2− ) in the presence of few ANMEs and methanogens ( Figs. 2 - 3 ), suggesting that cryptic sulfur cycling and AOM processes occurred across the sulfate–methane transition [ 43 , 83 ]. Below the active sulfate reduction zone, Bathyarchaeia prevailed in the fermentative zone ( Fig. 2B ). Several of their MAGs revealed metabolic capabilities either in sulfide dehydrogenase ( sud ) activity [ 84 ] with polysulfide ( hyd ) as electron acceptor [ 85 , 86 ], or reverse in fermentation coupling heterodisulfide reductase ( hdr ) with electron bifurcation ( Ech , Nuo ), or both [ 70 ]. By metabolizing sulfide into sulfur, and sulfur into polysulfide via rubredoxin [ 84 , 85 ], and assimilating persulfide with the dsrE protein as acceptor [ 56 , 87 ], Bathyarchaeia feature a redox complex offering an alternative metabolic pathway that makes elemental sulfur disproportionation energetically more favorable. Soluble polysulfides can effectively scavenge hydrogen sulfide (H 2 S) to form elemental sulfur (S 0 ) chains of up to eight atoms [ 86 ], which allows continuous sulfide removal without the need for sulfur disproportionation to sulfate. Thus, microbial cycling of sulfur intermediates via polysulfides could account for residual SRR measured below 5 cmblf. Although polysulfide is more soluble than sulfide in the pore water [ 86 ], it reacts abiotically with OM to produce poorly reactive sulfurized compounds [ 43 , 88 ]. In the absence of pore water polysulfide, we hypothesize that this function could be reversed towards fermentation. In the fermentative zone, Bathyarchaeia could rely on their heterotrophic capabilities to perform fermentation of polysaccharides and proteins [ 57 , 89 ], while able to couple the use of molecular hydrogen ( frh , mvh , hdr ) with electron bifurcation ( Ech , Nuo ) in the complete methyl-branch of the W-L pathway [ 90 , 91 ] to convert CO 2 into acetyl-CoA under low-energy conditions [ 92 ]. In the absence of mcrA genes ( Supplementary Fig. S12 ), none of the bathyarchaeal groups identified (B24, B26-1, 40CM-2-53-6, TCS64, RBG-16-48-13) corresponded to those performing methanogenesis (BA1, BA2). Instead, their methanopterin-based W-L pathway appeared to be partially methylotrophic [ 49 , 93 ], using methanol ( fae , mta ), methylamines ( mtb ), formate ( fdh ) and CO 2 ( codh , mtr ) as diverse C1 compounds, to produce energy and assimilate acetate into biomass ( acs , acss2 ). Yet, the groups B24 and B26-1 [ 90 ] qualified as full homoacetogens as they exhibited metabolic capability to catalyze the conversion of acetyl-CoA into acetate ( pta , akn ). While acting as an electron sink in heterotrophic acetogenesis ( Fig. 5D ), their partial carbonyl-branch (i.e. bacterial type) combined with SLP could participate in tetrahydrofolate interconversions of C1 and C2 compounds, mostly from CO 2 to formate and acetate [ 49 , 93–95 ]. Similar to termite guts [ 96 ], such W-L pathway combined with complex carbohydrate degradation ( Supplementary Fig. S15 ) could support syntrophic associations with methylotrophic methanogens. Potential metabolic features inherited from ancient ferruginous ecosystems Bathyarchaeia are postulated to have arisen at hydrothermal vents [ 93 , 94 ] and diversified in ferruginous and euxinic environments during the Proterozoic [ 49 , 57 ], perpetuating several primitive sulfur-based and energy-conserving metabolic features of methylotrophic archaea [ 59 ]. Among these, cryptic sulfur cycling coupled with CO 2 fixation via the W-L pathway [ 97 ] are presumedly prominent features of primeval microbial life in ancient ferruginous systems [ 93 ]. Similarly, dissimilatory iron reduction is thought to be one of the oldest energy-generating processes [ 54 ] that may have evolved on Earth from catalytic minerals into electron-transporting metalloenzymes [ 66 ]. In Earth’s early oceans, the metabolic capability to access limited resources in bioavailable sulfur, metals and OM appear to have played a key role in driving enzymatic reactions relevant to the W-L pathway at the origin of life [ 36 , 66 ]. By analogy, in modern ferruginous sediments, Bathyarchaeia prevailed below the sulfate reduction zone upon depletion of reactive iron minerals and labile organic substrates. Their MAGs included metabolic machinery for respiration of soluble polysulfides [ 87 ] and assimilation via the hdr -complex, dsrE and rubredoxin, inferring an ability to harness reactive sulfur species [ 98 ] via polysulfide cycling. The presence of an archaeal type RuBisCO III pathway with genetic potential to recycle AMP, methanol and formaldehyde ( deoA , fae ) further suggested that phosphoriboses (i.e. necromass) could originally be fermented under organic-lean, high CO 2 conditions [ 41 , 99 ]. By taking advantage of both a partial catabolic (carbonyl-branch) and complete anabolic (methyl-branch) W-L pathway, Bathyarchaeia developed a mixotrophic lifestyle as methylotrophic/acetogenic fermenters capable of fixing CO 2 [ 61 , 95 ]. Their metabolic potential to degrade complex carbohydrates and proteins [ 82 ] grafted onto such redox-conserving homoacetogenic W-L pathway may have led to excess production of C1-C2 compounds and molecular hydrogen during OM fermentation, thereby developing close (syntrophic) interactions with methylotrophic methanogens [ 50 , 52 ] and non-methanogens [ 51 ] in modern ecosystems."
} | 4,326 |
22899899 | null | s2 | 3,424 | {
"abstract": "We present the integration of a natural protein into electronic and optoelectronic devices by using silk fibroin as a thin film dielectric in an organic thin film field-effect transistor (OFET) ad an organic light emitting transistor device (OLET) structures. Both n- (perylene) and p-type (thiophene) silk-based OFETs are demonstrated. The measured electrical characteristics are in agreement with high-efficiency standard organic transistors, namely charge mobility of the order of 10(-2) cm(2)/Vs and on/off ratio of 10(4). The silk-based optolectronic element is an advanced unipolar n-type OLET that yields a light emission of 100nW."
} | 159 |
23774788 | PMC3684811 | pmc | 3,427 | {
"abstract": "Direct current (DC) piezoelectric power generator is promising for the miniaturization of a power package and self-powering of nanorobots and body-implanted devices. Hence, we report the first use of two-dimensional (2D) zinc oxide (ZnO) nanostructure and an anionic nanoclay layer to generate piezoelectric DC output power. The device, made from 2D nanosheets and an anionic nanoclay layer heterojunction, has potential to be the smallest size power package, and could be used to charge wireless nano/micro scale systems without the use of rectifier circuits to convert alternating current into DC to store the generated power. The combined effect of buckling behaviour of the ZnO nanosheets, a self-formed anionic nanoclay layer, and coupled semiconducting and piezoelectric properties of ZnO nanosheets contributes to efficient DC power generation. The networked ZnO nanosheets proved to be structurally stable under huge external mechanical loads.",
"discussion": "Discussion We also investigated the performance of the ZnO nanosheet-based nanogenerator using several top electrodes of various work functions such as Au (Schottky contact with ZnO nanosheets), graphene (weak Schottky contact with ZnO nanosheets), indium tin oxide (ITO) (very weak Schottky contact with ZnO nanosheets), and Al (ohmic contact with ZnO nanosheets) (see Supplementary Figs. S5 and S6 ). The output performances of voltage and current pulses from the graphene, ITO, and Al top electrode-based nanogenerators are poor compared to that of the Au top electrode-based nanogenerator. It is due to lack and less preservation of piezoelectric potential in case of weak Schottky contact, very weak Schottky contact, and ohmic contact compared to preservation of piezoelectric potential in the case of good Schottky contact with the Au electrode which helps to store larger charges in the LDH, resulting in stronger voltage and current pulses in response to applied force. To clearly understand the mechanical stability of the 2D ZnO nanosheet-based nanogenerator under an applied force, compression testing of ZnO nanosheets and ZnO nanorods grown on copper foil was performed with a nanoindentation system equipped with a 2 μm diameter flat-punch type diamond tip. Compression testing up to 1 μm from the top-end of the nanorods and nanosheets by nanoindentation gives insight into the ZnO nanorods and nanosheets which respond differently under uniaxial compression ( Fig. 4a ). Such response of nanorods and nanosheets under compression provides an intricate means of comparing mechanical properties of these two distinctively different types of nanostructures. Simple comparison of mechanical properties between two nanostructures by buckling forces or elastic modulus is not a proper method because force distribution on the two distinctive nanostructures is completely different. Furthermore, nanorods and nanosheets do not have completely normal to the surface as well as indentation direction, so compression force could not be uniformly distributed over all the structures under the flat-punch indenter tip. Hence, we adopted strain-energy density concept to understand the elastic stability of the two nanostructures. Since total energy introduced to the structures by compression will be absorbed by the structures in unit area, strain-energy density, which represents energy absorption ability until the structure is permanently deformed (in this case, abruptly changes in normal load) could be a representative property to explain how the materials respond to the external forces. As indicated in Fig. 4a , abrupt changes in load-displacement curves of nanorods and nanosheets are presented by arrows, and named critical load (P crt ). We selected P crt of nanorods at the largest increase in the displacement. Several ‘pop-in' events (abrupt increases in the displacement) are observed before the critical load, P crt , which could be due to the buckling of the nanorods. It could not be the representative compressed force on nanorods by the indentation tip. We calculated the area under the curve before P crt to determine the strain-energy density of the nanorods. In the case of nanosheets, special care to select P crt is needed because it does not show a distinct change of load or displacement as shown in Fig. 4a . We differentiated the load-displacement curve, and we then selected an inflection point at the constriction of the structure. This force before constriction of the network structure of nanosheets could be used as the representative force for nanosheets under the indentation tip. Since the strain energy U of the nanostructures is distributed uniformly throughout its volume (under the flat-punch), it can be expressed as, where δ is displacement. We can determine the strain-energy density by dividing the total strain energy with the volume of the structures. The estimated strain-energy density of the nanorods is about 7.88 × 10 6 J/m 3 whereas that of the nanosheets is 8.20 × 10 7 J/m 3 ( Table 1 ). The strain-energy density of the ZnO nanosheets is about ten times larger than that of the the ZnO nanorods. The absorption of higher strain-energy density in the nanosheets is due to their networked structure. In summary, we have demonstrated the first use of 2D ZnO nanostructure and an anionic layer to generate large DC electrical output. The device made from such 2D piezoelectric nanosheets/anionic layer heterojunction has potential for use as a mechanically durable and smallest size power package, which can be used to charge wireless nano/micro-scale systems with no need for a rectifier circuit. The combined effect of buckling the ZnO nanosheets, anionic layer, and coupled semiconducting and piezoelectric properties of ZnO nanosheets are the main causes of the efficient DC power generation. Due to morphologically networked ZnO nanosheets, the ZnO nanosheets proved to be more structurally stable under huge external mechanical loads compared to ZnO nanorods. Furthermore, it was found that the ZnO nanosheet-based piezoelectric nanogenerators reveal much higher and more stable power-generating performance compared to previous piezoelectric DC output nanogenerators."
} | 1,548 |
26433202 | PMC4682429 | pmc | 3,429 | {
"abstract": "Highlight The first evaluation of lignification in Brachypodium distachyon grain is reported. Moderately down-regulated BdCOMT6 alters grain and stem lignification, which improves stem saccharification without major detrimental effects on grain development and composition.",
"conclusion": "Conclusion In this study focused on the Bd5139 Brachypodium mutant, it was established that modifying the lignin-related BdCOMT6 gene induced alterations in grain lignins which nicely mirrored those observed in stem lignins. The accumulation of 5-OH G units, which is the most specific signature for COMT deficiency, was reported here for the first time in the grain of a COMT grass mutant. The single mutation in the BdCOMT6 protein did not completely annihilate its enzyme activity. It induced substantial alterations in lignin structure but only a moderately reduced lignin level. The moderate lignin reduction did not compromise the vegetative and reproductive development of the Brachypodium plant model, but facilitated the straw saccharification, opening up the possibility of a sustainable cereal grain production with improved straw end-use potential.",
"introduction": "Introduction Grass lignocellulosics are important potential feedstocks for the production of cellulosic ethanol. Cereal by-products, such as wheat straw or corn stover, and dedicated-energy grass crops, such as miscanthus or switchgrass, are mainly composed of cellulose, hemicellulose, and lignins. However, their enzymatic saccharification into fermentable sugars is detrimentally affected by lignins and their cross-linking to wall carbohydrates. Indeed, lignins limit the accessibility of enzymes to polysaccharides and make necessary the use of costly pre-treatments aimed at improving this accessibility ( Yang and Wyman, 2008 ). Lignins are currently the target of biotechnology with the objective to design plant cell walls more amenable to the saccharification process. Lignin genetic engineering is still more challenging for grass lignins due to the specificities of lignified grass cell walls. While grass lignins are made of guaiacyl (G) and syringyl (S) units together with a lesser amount of p -hydroxyphenyl (H) units, like any other angiosperm lignins, their major peculiarity is that p -coumaric acid (CA) and ferulic acid (FA) participate substantially in their assembly in the cell wall ( Ralph, 2010 ). FA acylates the arabinose substituents of arabinoxylans, and grass lignins are oxidatively cross-linked to these ferulate esters ( Jacquet et al. , 1995 ). CA acylates mainly S lignin units ( Ralph, 2010 ) and to a lower extent arabinose units ( Mueller-Harvey et al. , 1986 ). Designing grass cell walls more amenable to bioethanol production requires the use of appropriate model plants provided with the unique specificities of grass lignins. Brachypodium distachyon is a plant species whose genomic sequence is released for the Bd21 accession ( International Brachypodium Initiative, 2010 ) and which has recently been championed as a model grass to identify genes important for cereals and energy grasses ( Draper et al. , 2001 ; Vogel et al. , 2006 ; Opanowicz et al. , 2008 ; Bevan et al. , 2010 ; Brkljacic et al. , 2011 ; Mur et al. , 2011 ; Rancour et al. , 2012 ). In the challenging context of designing grass cell walls with improved end-use properties, a B. distachyon ( Brachypodium ) mutant collection associated with a TILLING (targeting induced local lesion in genome) platform has been recently developed ( Dalmais et al. , 2013 ). Using these tools, several Brachypodium mutants affected in a lignin-related caffeic acid O -methyltransferase (COMT) activity were identified ( Dalmais et al. , 2013 ). This study focused on the Bd5139 line displaying a single point mutation in the BdCOMT6 gene to demonstrate that this lignin-related gene (or its homologues in other grasses) is a promising target for lignin genetic engineering. The prospect of a sustainable use of cereal crops will rely on breeding programmes to improve the potential of by-products for bioethanol production while preserving grain quality. The impact of the Bd5139 mutation on the cell wall phenolics occurring not only in Brachypodium lignified stems, the most conventionally studied organ in lignin-deficient mutants or transgenics, but also in Brachypodium grains, is reported. This grain study was performed due to the fact that lignins, CA, and FA are thought to have important roles in the conductive and protective tissues of grass grains ( Beaugrand et al. , 2004 ; Barron et al. , 2007 ). It was therefore important to evaluate the impact of the lignin-targeted mutation not only on stems, but also on seeds. It is demonstrated that this single mutation substantially affects the cell wall phenolics of both stems and grains, and improves the saccharification of Brachypodium mature stems without impairing grain development and composition.",
"discussion": "Results and Discussion The lignin-related BdCOMT6 gene is expressed both in stem and in grain In a recent study ( Dalmais et al. , 2013 ), Brachypodium mutants for the BdCOMT6 gene were identified in a sodium azide-induced mutant collection established in the Bd21-3 genetic background. Several mutant lines displayed a lower lignin content in mature stems together with a reduced frequency of S lignin units, and Bd5139 was the most affected line ( Dalmais et al. , 2013 ). These results established that the BdCOMT6 gene is involved in lignification of Brachypodium stems. To characterize the BdCOMT6 gene further, its expression pattern was obtained using the online expression platform PlaNet/Brachypodium ( http://aranet.mpimp-golm.mpg.de/; \n Mutwil et al. , 2011 ). This investigation revealed a wide expression pattern of the BdCOMT6 gene. It was found to be expressed not only in vegetative lignified organs, such as nodes and internodes, but also in grains ( Supplementary Fig. S1 available at JXB online). The BdCOMT6 expression level was found to be similar in the WT and the BdCOMT6 -deficient Bd5139 lines ( Supplementary Fig. S1 ). The Bd5139 mutant has a normal growth phenotype, but an altered lignification in stem and grain As compared with the WT line, the Bd5139 mutant did not show any visible growth or developmental phenotype when grown in a growth chamber or in a greenhouse even though stem biomass was found to be slightly reduced ( Supplementary Fig. S2 at JXB online). In 40-day-old plants, Maüle staining revealed the occurrence of S lignin units ( Nakano and Meshitsuka, 1992 ) in the cell walls of several stem tissues (epidermis, sclerenchyma, and vascular bundles) without any noticeable difference between the WT and the Bd5139 lines ( Fig. 1 ). This result suggests that the Bd5139 mutation might not affect the frequency of S lignin units to a large extent. Fig. 1. Maüle staining of stem cross-sections of Brachypodium WT (Bd21-3) and Bd5139 lines. Brachypodium Bd21-3 (WT; A) and Bd5139 (B) stem cross-sections from 40-day-old plants were stained with the Maüle reagent that stains S lignin units. e, epiderm; if, interfascicular fibres, s, sclerenchyma; vb, vascular bundle. (This figure is available in colour at JXB online.) It is now well established that lignin-related COMT enzymes of angiosperm species catalyse the 5-OH group methylation on the pathway towards sinapyl alcohol, the precursor of S lignin units ( Bonawitz and Chapple, 2010 ; Vanholme et al. , 2010 ). According to several studies, their most likely in vivo substrates would be 5-hydroxyconiferaldehyde and 5-hydroxyconiferyl alcohol ( Zubieta et al. , 2002 ; Louie et al. , 2010 ; Green et al. , 2014 ). In contrast, caffeic acid, which has been shown to be an in vitro COMT substrate converted into FA ( Higuchi, 2006 ), would be a poor in vivo candidate ( Parvathi et al. , 2001 ; Green et al. , 2014 ). Manipulation of lignin in plants has repeatedly been carried out by targeting the lignin-related COMT genes, which most often resulted in a deficit of S lignin units and the appearance of unusual levels of 5-OH G units (reviewed in Rastogi and Dwivedi, 2008 ). In agreement with a preliminary study on COMT-deficient Brachypodium lines ( Dalmais et al. , 2013 ), both lignin content and lignin structure were found to be affected by the single Gly256Asp mutation occurring in the Bd5139 line ( Table 1 ). Relative to the WT level, the lignin content of Bd5139 stem CWR was moderately reduced by ~10%. Lignin structure was more markedly affected by the BdCOMT6 mutation, as revealed by thioacidolysis ( Table 1 ). This method provides specific monomers from lignin units that are only involved in labile β-O-4 bonds. When expressed relative to the KL content, the total yield of thioacidolysis monomers released from Bd5139 lignins was found to be lower than the WT value. This reduced yield is diagnostic for an increased frequency of resistant interunit bonds in lignins. In addition, the relative frequency of S thioacidolysis monomers was reduced by the mutation whereas 5-OH G monomers were recovered in an unusually high amount. As previously described in the maize bm3 mutant ( Lapierre et al. , 1988 ; Barrière et al. , 2004 ), sorghum bmr12 mutant ( Palmer et al. , 2008 ), Arabidopsis Atomt1 mutant ( Goujon et al. , 2003 ), COMT1-silenced transgenic tobacco ( Pinçon et al. , 2001 ), or COMT-deficient transgenic poplars ( Lapierre et al. , 1999 ; Jouanin et al. , 2000 ), these alterations in lignification are the hallmarks of lignin-related COMT deficiency in angiosperms, the most diagnostic signature being the increased frequency of lignin 5-OH G units. While these 5-OH G units were found to build up substantially in the lignins of Bd5139 stems (up to 5% of the total amount of thioacidolysis monomers), the frequency of S thioacidolysis monomers was found to be only moderately reduced (from 66% in the WT to 57% in Bd5139 ; Table 1 ). This result is consistent with the observation that the Maüle stain, specific for S lignin units, was not affected to a large extent in Bd5139 mature stem cross-sections ( Fig. 1 ) in contrast to the observations made for severe comt mutants ( Jouanin et al. , 2000 ; Pinçon et al. , 2001 ). Taken together, these data reveal that the biosynthesis of sinapyl alcohol is affected to a noticeable but limited extent in the Bd5139 stems. Table 1. \n Lignin content and structure of cell wall residues prepared from wild-type (WT, accession Bd21-3) and Bd5139 3-month-old mature stems Lignin content is evaluated as the Klason lignin (KL) level and lignin structure is evaluated by thioacidolysis. Line % KL Main H, G, S, and 5-OH G thioacidolysis monomers (total yield and relative mol%) \n \n Yield (μmol g –1 KL) %H %G %S %5-OH G WT 19.57±0.30 1258±30 3.1±0.2 29.3±1.2 66.9±1.1 0.8±0.0 \n Bd5139 \n 17.66±0.93* 888±31** 2.9±0.2 36.4±1.2** 55.6±1.5** 5.1±0.4** Values are means ±SD from three different plants. The KL level is expressed as weight percentage of the stem cell wall residue. Asterisks indicate significant differences ( t -test) compared with the WT value at * P <0.05 or ** P <0.01 As BdCOMT6 was found to be expressed in grains and as it is now well established that lignins occur in the outer layers of grass grains ( Desvignes et al. , 2006 ; Grefeuille et al. , 2006 ), the lignins from Brachypodium grains were studied. The occurrence and distribution of lignins in grain tissues were first studied on grain cross-sections using the Wiesner and the Maüle tests, two commonly used stains specific for lignified tissues. The Wiesner test stains the p -hydroxycinnamaldehyde end-groups occurring in native lignins (mainly as coniferaldehyde end-groups) reddish-purple while the Maüle test stains S lignin units deep-purple ( Nakano and Meshitsuka, 1992 ). Brachypodium grains are hulled grains consisting of the caryopsis surrounded by two husks, a large lemma, and a small palea ( Evers and Millar 2002 ; Guillon et al. , 2011 ). The lemma can be easily removed while the palea adheres to the pericarp in the crease. In the palea, both the Wiesner and the Maüle tests positively stained the epidermis and vascular bundles ( Fig. 2E , G , H ), which establishes the occurrence of lignin coniferaldehyde end-groups and of S lignin units in this protective husk. In addition, the outer layer of the seed testa (t2 in Fig. 2C , F ) was positively stained by the Maüle reagent whereas no positive Maüle staining could be detected in the brownish-pigmented t1 layer. The Maüle-stained t2 layer did not positively react to the Wiesner test, a phenomenon probably accounted for by the distinct detection sensitivity level of these lignin staining methods. Consistently with the positive Wiesner and Maüle stainings of the seed palea and t2 testa, immunolabelling experiments carried out with an antibody targeting a β-5 dimer of coniferyl alcohol ( Kiyoto et al. , 2013 ) confirmed the occurrence of lignins in the grain palea and testa ( Fig. 2J ). Fig. 2. Cytological observations of Brachypodium grain. Brachypodium Bd21-3 (WT) mature grain cross-sections. (A) Unstained sections of a whole grain with labelled frames indicating the corresponding areas in the subsequent parts of the figure. (B) Unstained section and (C) section stained with toluidine blue focusing on the testa area to visualize the two layers of the testa t1 and t2. (D) Unstained section focusing on one vascular bundle of the grain palea. (E) Section stained with phloroglucinol-HCl revealing positive staining in the palea epidermis and vascular bundle. (F–H) Section stained with Maüle reagent revealing positive staining in the testa outer layer t2, in the palea epidermis, and in the vascular bundle. (J) Section labelled with KM1, an antibody targeting a lignin β-5 structure and showing a positive signal in the testa and in the palea epidermis and vascular bundle, to be compared with (I) the corresponding control without primary antibody. al, aleurone; e, epiderm; ne, nucellar epidermis; p, pericarp; pal, palea; sc, silica cells; se, storage endosperm; t. testa; t1, testa inner layer pigmented; t2, testa outer layer not pigmented but lignified; vb, vascular bundle. (This figure is available in colour at JXB online.) To evaluate the impact of BdCOMT6 deficiency on the grain lignin level, the WT and Bd5139 grain CWRs were subjected to ABL determination. The ABL content was found to be similar in the Bd5139 and WT grain samples ( Table 2 ), in contrast to the mature stem samples ( Table 1 ). CWR samples prepared from whole grain were then subjected to thioacidolysis with the objective of finding a more diagnostic lignin signature. In agreement with previous studies performed on wheat grains ( Desvignes et al. , 2006 ; Grefeuille et al. , 2006 ), thioacidolysis unambiguously revealed the occurrence of H, G, and S lignin units in Brachypodium whole grains from the detection of H, G, and S thioacidolysis monomers ( Table 2 ). The relative frequency of H thioacidolysis monomers was found to be higher from grain CWR compared with the levels observed from stem CWR. In addition, the relative frequency of S thioacidolysis monomers was found to be substantial (60% of thioacidolysis monomers for the WT sample; Table 2 ), in agreement with the Maüle positive staining of the palea and of the outer t2 testa layer. More importantly and relative to the WT sample, lignins of Bd5139 grain displayed severe structural alterations, namely a reduced frequency of S monomers together with an increased frequency of 5-OH G monomers (up to 10% of the thioacidolysis monomers; Table 2 ; Supplementary Fig. S3 at JXB online). When expressed relative to the ABL content, thioacidolysis yield was found to be reduced by the Bd5139 mutation, which reveals that Bd5139 grain lignins are richer in resistant interunit bonds than those of the WT. Taken together, the present results establish that the alterations induced by the Bd5139 mutation in the lignins of mature grains mirror those observed in mature stems. In both samples and relative to the WT, marked lignin structural alterations can be observed, namely a lower frequency of S units together with an increased frequency of 5-OH G units and of resistant interunit bonds. In both Bd5139 stem and grain samples and surprisingly enough, the frequency S units was found to be reduced, but only to a moderate extent. This persistence of substantial levels of S units in Bd5139 lignins suggests that some enzyme activity survives in the mutated BdCOMT6 protein. Table 2. \n Lignin content and structure of cell wall residues prepared from wild-type (WT, accession Bd21-3) and Bd5139 whole grain samples Lignin content is evaluated as the acetyl bromide lignin (ABL) level and lignin structure is evaluated by thioacidolysis. Line % ABL Main H, G, S, and 5-OH G thioacidolysis monomers (total yield and relative mol%) \n \n Yield (μmol g –1 ABL) %H %G %S %5-OH G WT 3.66±0.50 208±27 5.9±0.1 33.8±2.7 60.2±2.7 Trace \n Bd5139 \n 3.62±0.15 115±31** 6.2±0.6 39.5±1.8** 44.1±2.4** 10.1±2.6** Whole grain samples correspond to the whole caryopsis with the adhering palea. Four biological replicates were prepared for each genotype, with 70–100 grains collected per replicate. Values are means ±SD. The ABL level is expressed as weight percentage of the sample cell wall residue. Asterisks indicate significant differences ( t -test) compared with the WT value at ** P <0.01. Evaluation of p -coumaric and ferulic acid linked to Brachypodium stem and grain cell walls It is now well established that CA and FA are linked to the polymers of grass cell walls (reviewed in Ralph, 2010 ). For many grass species, such as maize, sorghum, miscanthus, or wheat, CA and FA units are quite distinctly distributed between the wall polymers of lignified stems. While FA predominantly acylates the arabinose substituents of arabinoxylans, CA is primarily ester-linked to S lignin units. However, such a distinct distribution is not as clearly observed in Brachypodium cell walls, which contain a substantial proportion of CA acylating arabinoxylans ( Christensen et al. , 2010 ; Petrik et al. , 2014 ). The effect of the Bd5139 mutation on the amount of CA or FA ester-linked to the cell wall polymers was monitored by mild alkaline hydrolysis ( Table 3 ). Mature stem cell walls were found to contain higher amounts of CA and FA esters than mature grain cell walls. This compositional trait reflects the higher abundance of lignins and arabinoxylans in stems than in grains, as lignins and arabinoxylans, respectively, are the major p -coumaroylated and feruloylated wall polymers. Table 3. \n Determination of p -coumaric acid (CA) and of ferulic acid (FA) released by mild alkaline hydrolysis (NaOH 1M, room temperature, overnight) of wild-type (WT, accession Bd21-3) and Bd5139 cell wall residues (CWRs) prepared from mature stem or whole grain samples \n Line Mature stem CWR Mature whole grain CWR \n CA (mg g –1 ) FA (mg g –1 ) CA (mg g –1 ) FA (mg g –1 ) WT 8.88±0.41 5.25±0.48 0.98±0.1 1.83±0.47 \n Bd5139 \n 6.30±0.66* 5.33±0.17 0.59±0.04* 1.55±0.11* Whole grain samples correspond to the whole caryopsis with the adhering palea. Four biological replicates were prepared for each genotype, with 70–100 grains collected per replicate. Values are means ±SD ( n =3). Asterisks indicate significant differences ( t -test) compared with the WT value at * P <0.05. Relative to the WT level, the amount of FA released by mild alkaline hydrolysis was not affected in the Bd5139 stem CWR and slightly decreased in the mutant grain CWR. This result suggests that BdCOMT6 deficiency has no or little effect on the level of FA released by mild alkaline hydrolysis of Brachypodium cell walls. In contrast, mature internodes of COMT-deficient bm3 maize ( Provan et al. , 1997 ; Marita et al. , 2003 ; Barrière et al. , 2004 ) or bmr12 sorghum ( Palmer et al. , 2008 ) lines have been shown to release more FA when subjected to mild alkaline hydrolysis, a phenomenon related to their markedly lower lignin level as measurable FA is reduced by lignification ( Grabber et al. , 2004 ). These results confirm that in planta the lignin-related COMT activity is not involved in the synthesis of FA that subsequently acylates arabinoxylans. Relative to WT values, the content of CA esters was reduced by 30% and 40% in Bd5139 stem and grain samples, respectively ( Table 3 ). A similar reduction of CA units ester-linked to stem cell walls has been reported for maize bm3 and sorghum bmr12 mutant lines ( Barrière et al. , 2004 ; Palmer et al. , 2008 ). On the rationale that CA mainly acylates maize or sorghum S lignin units, the CA reduction observed in these COMT-deficient lines could be directly assigned to the COMT deficiency-induced reduction of S lignin units. In contrast to maize or sorghum cell walls, a noticeable amount of CA units acylates the arabinoxylans of Brachypodium cell walls. It was therefore necessary to clarify whether the lower level of CA esters in Bd5139 samples affected only lignins or both lignins and arabinoxylans. To address the issue of CA acylation targets, a recently developed mild acidolysis method which efficiently releases CA-Ara and FA-Ara from grass arabinoxylans was employed ( Petrik et al. , 2014 ). The CA-Ara quantities released from Bd5139 stem CWR were found to be very close to the WT values ( Supplementary Table S1 at JXB online). On this basis, it could be ascertained that the BdCOMT6 mutation specifically reduces lignin p -coumaroylation in Bd5139 CWR whereas the p -coumaroylation of arabinoxylans is not affected. Similar to what was reported in COMT-deficient bm3 maize or bmr12 sorghum lines, the BdCOMT6 mutation affects the biosynthesis of sinapyl alcohol and thereby of sinapyl p -coumarate, the precursor of p -coumaroylated S lignin units. Complementation experiments reveal that the mutated BdCOMT6 protein is still functional in the Bd5139 mutant The mutation in Bd5139 changed a relatively conserved Gly256 residue into an aspartic acid residue ( Supplementary Fig. S4 at JXB online). The overall impact of this single mutation in the BdCOMT6 protein was found to mimic the impact of the bm3 or bmr12 mutations on the lignification of maize or sorghum mutant lines, respectively, but to a less severe extent. Indeed, S thioacidolysis monomers accounted for ~56% of the total amount of thioacidolysis monomers recovered from stem Bd5139 lignins (versus 67% for the WT; Table 1 ). In contrast, their frequency was reduced to a value close to zero in the lignins of a bmr12 sorghum mutant ( Palmer et al. , 2008 ) and to a twice lower value relative to the WT level in bm3 lignins ( Barrière et al. , 2004 ). To evaluate the functionality of the mutated BdCOMT6 protein (BdCOMT6 5139 ), complementation experiments were performed in an Arabidopsis comt-1 mutant which is almost completely depleted in S lignin units ( Vanholme et al. , 2012 ; Van Acker et al. , 2013 ). Such a complementation experiment was performed in order to determine to what extent the BdCOMT6 \n 5139 allele was able to rescue the biosynthesis of syringyl lignin units in the Arabidopsis comt-1 mutant. Not unexpectedly, the Maüle reagent specific for S lignin units negatively stained Arabidopsis stem cross-sections from the comt-1 line ( Fig. 3 ). A positive Maüle staining was restored in Arabidopsis comt-1 samples complemented with the BdCOMT6 \n 5139 allele ( Fig. 3 ). This result revealed that the BdCOMT6 5139 protein has conserved sufficient activity to catalyse the methylation step involved in the pathway to sinapyl alcohol. To evaluate the functionality of the mutated protein more precisely, the impact of the complementation on the frequency of lignin-derived H, G, S, and 5-OH G thioacidolysis monomers specifically released from the lignins of mature Arabidopsis stems was examined ( Table 4 ). As expected and relative to the WT levels, the relative frequency of the 5-OH G monomers was dramatically increased while that of S monomers was reduced to <3% in the comt-1 Arabidopsis mutant. The transformation of this mutant with BdCOMT6 \n 5139 rescued the S lignin units to a frequency that surprisingly exceeded that of the WT sample and for three different complemented lines ( Table 4 ). In addition, the frequency of 5-OH G thioacidolysis monomers was decreased to a value close to that of the WT. These thioacidolysis results unambiguously confirmed that the mutated BdCOMT6 has conserved sufficient enzyme activity so as to rescue the COMT-deficient Arabidopsis mutant efficiently. The surprisingly higher frequency of S units in the complemented lines and relative to the control value might be accounted for by the promoter employed, the strong maize ubiquitin promoter, which would lead to a non-specific overexpression of the mutated BdCOMT6 \n 5139 gene as compared with the regulated WT AtCOMT1 transcript. Fig. 3. \n Arabidopsis comt-1 complementation assays using mutated BdCOMT6. Stem cross-sections of Arabidopsis thaliana stained using the Maüle reagent to reveal S lignin units. (A) WT (Col-0) with intense positive staining in lignified interfascicular fibres, (B) comt-1 mutant depleted in S lignin units (negative Maüle staining), (C) comt-1 line complemented with the mutated BdCOMT6 gene under the control of the maize ubiquitin promoter (restoration of positive Maüle staining). Scale bar=100 μm (A–C). (This figure is available in colour at JXB online.) Table 4. \n Relative molar frequency of the p -hydroxyphenyl (H), guaiacyl (G), syringyl (S), and 5-hydroxyguaiacyl (5-OH G) monomers released by thioacidolysis of the cell wall residues from Arabidopsis mature stems \n The examined genotypes (in the Col-0 background) are the wild type (WT), the comt-1 mutant, and three lines (Cp-line) obtained by complementation of the comt-1 mutant with the mutated BdCOMT6 gene. Genotype %H %G %S % 5-OH G WT 0.83±0.02a 69.3±0.2 a 29.3±0.4a 0.51±0.02 a \n comt-1 \n 0.84±0.04 a 90.4±0.6 b 2.40±0.69 b 6.34±0.09 b Cp line 4–7 1.63±0.10 b 59.6±0.2 c 38.2±0.3 c 0.59±0.01 a Cp line 7-5 1.74±0.19 b 59.7±0.3 c 38.0±0.2 c 0.60±0.02 a Cp line 15–6 1.57±0.12 b 58.7±0.5 c 39.1±0.4 c 0.61±0.03 a Values are means ±SD (with three biological replicates). Within each row, different letters indicate significant differences (one-way ANOVA) at P <0.01. In previous studies, it has been shown that an Arabidopsis T-DNA mutant knockout for the lignin-specific comt1 gene was not only affected in stem lignification, but also in the pool of soluble phenolics ( Goujon et al. , 2003 ; Do et al. , 2007 ). This mutant contained a severely reduced level of sinapoyl malate (SIM) and abnormal amounts of 5-hydroxyferuloyl malate (5-OH FM). In addition and at the plantlet stage, AtCOMT1 deficiency induced the complete disappearance of isorhamnetin glycosides. These results established that the AtCOMT1 enzyme is involved not only in the lignin pathway, but also in the synthesis of SIM and of methylated flavonol derivatives (i.e. isorhamnetin derivatives). To establish further the functionality of the mutated BdCOMT6 protein, the impact of the complementation experiment on the soluble phenolics of 5-day-old comt-1 Arabidopsis plantlets was investigated. In agreement with previous results ( Do et al. , 2007 ) and relative to the WT values, the comt-1 plantlets contained a severely reduced SIM level, this reduction being partially compensated for by the appearance of 5-OH FM ( Supplementary Table S2 at JXB online). In addition, isorhamnetin derivatives that could be observed as minor soluble phenolics of WT plantlets were entirely absent from comt-1 plantlets. The complementation with the mutated BdCOMT6 protein efficiently restored the SIM and the isorhamnetin levels to the WT value and induced the total disappearance of the 5-OH FM. Taken together, these results ascertained that the mutated BdCOMT6 gene introduced in the AtCOMT1 -deficient Arabidopsis mutant was able to rescue not only its altered stem lignification, but also its reduced pool of methylated soluble phenolics. A similar rescue of syringyl units and of sinapate esters has been reported for the fah Arabidopsis mutant complemented by the Eucalyptus globulus coniferaldehyde-5 hydroxylase ( Garcia et al. , 2014 ). In the present study, such an efficient complementation was done with the mutated BdCOMT6 gene, which definitely established that the mutated BdCOMT6 protein is still functional. The Bd5139 mutant displays an improved saccharification of mature stems without major alteration in grain quality A promising breeding strategy of cereal crops would consist of making their straw more amenable to saccharification, through appropriate changes in lignin content and/or structure, without introducing deleterious effects on biomass production and on grain quality. Indeed, as lignins are essential to plant health and development, substantial lignin reduction would inevitably reduce the agricultural fitness of grass crops ( Pedersen et al. , 2005 ). Many studies of COMT-deficient transgenic or mutant angiosperms have established that their reduced lignin level was associated with improved digestibility ( Cherney et al. , 1991 ; Bernard-Vailhé et al. , 1996 ; Sewalt et al. , 1997 ; Goujon et al. , 2003 ; Barrière et al. , 2004 ; Chen et al. , 2004 ; Trabucco et al. , 2013 ) or saccharification ( Chen and Dixon, 2007 ; Dien et al. , 2011 ; Fu et al. , 2011 ; Jung et al. , 2012 ; Van Acker et al. , 2013 ; Baxter et al. , 2014 ). The currently studied Bd5139 mutant displayed moderate lignin reduction, as the BdCOMT6 mutated protein, provided with a single point mutation, has retained enough enzyme activity to ensure a lignin level reduced only by 10–15%. On this basis, the impact of the mutation on the saccharification of mature stems was investigated. A saccharification assay was conducted on small amounts (30mg) of stem CWR without any pre-treatment and using a commercially available cellulase preparation (cellulase Onozuka from T. viride , provided with cellulase, hemicellulase, and β-glucosidase activities). The efficiency of cell wall enzymatic hydrolysis was evaluated both as the weight loss induced by the enzyme treatment and as the glucose amount released from the cell walls. Both evaluation methods consistently revealed that the saccharification efficiency of the stem CWR was improved (by ~20%) by the mutation and relative to the WT values ( Table 5 ). Such a result is consistent with the lower lignin level of Bd5139 stems as lignins detrimentally affect the enzymatic degradation of lignocellulosic biomass. Table 5. \n Saccharification assays of wild-type (WT, accession Bd21-3) and Bd5139 cell wall residues prepared from mature stem samples \n Saccharification efficiency is measured as the weight loss induced by the enzymatic treatment (as a percentage of the initial weight) and by the glucose released from the cell walls. Line Weight loss % Glucose mg g –1 a \n WT 23.0±1.5 80.9±3.1 \n Bd5139 \n 27.7±1.0** 97.7±2.5** Values are means ±SD ( n =6 with three biological replicates, each one analysed as analytical duplicates). Asterisks indicate significant differences ( t -test) compared with the WT value at ** P < 0.01. \n a Expressed as anhydroglucose equivalent. Lignins are important components of the stem vascular tissues necessary to convey water and nutrients to the developing grain. In addition, lignins occur in several grain outer layers that fulfil protective and nutritive functions towards the developing seed. Lignin mutants with seed defects have been reported. Maize field trials where bm3 lines were assessed revealed a reduced grain yield that was explained by a lower number of ears per plant and by a lower number of kernels per ear (reviewed in Pedersen et al. , 2005 ). Several Arabidopsis mutants altered in lignin polymerization exhibit seed phenotypes such as defects in seed pigmentation, permeability, and germination ( Liang et al. , 2006 ) or increased numbers of siliques and enlarged seeds ( Wang et al ., 2014 ). The Bd5139 line was assessed to evaluate the effect of moderate changes in lignin content and composition on cereal grains. Grain size and morphology were not affected in the Bd5139 line ( Supplementary Fig. S5 at JXB online). No major effect was noticed on grain histology or development ( Supplementary Fig. S5 ). No effect of the Bd5139 mutation could be evidenced on grain polysaccharide storage compounds ( Supplementary Table S3 ). Therefore, the moderate decrease in grain lignins and CA esters does not seem to affect grain development and polysaccharide composition. Conclusion In this study focused on the Bd5139 Brachypodium mutant, it was established that modifying the lignin-related BdCOMT6 gene induced alterations in grain lignins which nicely mirrored those observed in stem lignins. The accumulation of 5-OH G units, which is the most specific signature for COMT deficiency, was reported here for the first time in the grain of a COMT grass mutant. The single mutation in the BdCOMT6 protein did not completely annihilate its enzyme activity. It induced substantial alterations in lignin structure but only a moderately reduced lignin level. The moderate lignin reduction did not compromise the vegetative and reproductive development of the Brachypodium plant model, but facilitated the straw saccharification, opening up the possibility of a sustainable cereal grain production with improved straw end-use potential."
} | 8,467 |
37840744 | PMC10571051 | pmc | 3,430 | {
"abstract": "Soil microorganisms play a crucial role in remediating contaminated soils in modern ecosystems. However, the potential of combining microorganisms with legumes to enhance the remediation of heavy metal-contaminated soils remains unexplored. To investigate this, we isolated and purified a highly efficient cadmium and lead-tolerant strain. Through soil-cultivated pot experiments with two leguminous plants ( Robinia pseudoacacia L. and Sophora xanthantha ), we studied the effects of applying this microbial agent on plant nutrient uptake of soil nutrients, heavy metal accumulation, and the dynamics of heavy metal content. Additionally, we examined the response characteristics of inter-root microbial and bacterial communities. The results demonstrated that microorganisms screened from heavy metal-contaminated soil environments exhibited strong survival and adaptability in heavy metal solutions. The use of the Serratia marcescens WZ14 strain-phytoremediation significantly increased the soil’s ammonium nitrogen (AN) and organic carbon (OC) contents compared to monoculture. In addition, the lead (Pb) and cadmium (Cd) contents of the soil significantly decreased after combined remediation than those of the soil before potting. However, the remediation effects on Pb- and Cd-contaminated soils differed between the two legumes following the Serratia marcescens WZ14 inoculation. The combined restoration altered the composition of the plant inter-rhizosphere bacterial community, with the increase in the relative abundance of both Proteobacteria and Firmicutes. Overall, the combined remediation using the tolerant strain WZ14 with legumes proved advantageous. It effectively reduced the heavy metal content of the soil, minimized the risk of heavy metal migration, and enhanced heavy metal uptake, accumulation, and translocation in the legumes of S. xanthantha and R. pseudoacacia . Additionally, it improved the adaptability and resistance of both legumes, leading to an overall improvement in the soil’s environmental quality. These studies can offer primary data and technical support for remediating and treating Cd and Pb in soils, as well as rehabilitating mining sites.",
"conclusion": "5. Conclusion Microorganisms residing in heavy metal environments for extended periods have developed a notable tolerance and resistance to heavy metal ions. In order to address heavy metal contamination in soil, one of the effective approaches can be exploring heavy metal-resistant bacteria for their capacity to absorb and accumulate heavy metals ( Zheng et al., 2022b ). In this study, we successfully isolated and purified an efficient Cd- and Pb-tolerant strain from the contaminated soil of lead-zinc-silver mine in Zhijiadi, Shanxi Province. This strain exhibited robust acclimation in high Pb 2+ concentration cultures (2,000 mg/L and 1,500 mg/L), with peak values ranging from 2.92 cfu·mL −1 to 3.14 cfu·mL −1 . Additionally, it demonstrated strong viability and adaptability in Cd 2+ cultures at varying concentration gradients, showing growth even at concentrations exceeding 100 mg/L, with growth ranging from 1.45 cfu·mL −1 to 2.77 cfu·mL −1 . Both R. pseudoacacia and S. xanthantha are suitable for cultivation in heavy metal-contaminated soils, as they can absorb and translocate heavy metals, such as Pb and Cd, resulting in an improved soil microenvironment. Combined with crop-applied WZ14 inoculants, the remediation efficacy was enhanced, resulting in increased soil AN and OC levels and decreased total Pb and active heavy metals compared to monoculture. Furthermore, the application of WZ14 significantly facilitated the uptake of Cd by S. xanthantha , particularly in the roots, where the Cd uptake reached 147.44 mg/kg after microbial inoculant. This represented a substantial increase of 130.79% compared to monoculture, signifying the root’s significant role in the remediation of soil Cd contamination by S. xanthantha . The community structure of soil microorganisms in heavy metal-contaminated areas exhibited significant differences between planting and combined remediation treatments, resulting in notable changes in microbial abundance and diversity. The combination of R. pseudoacacia monoculture with WZ14- R. pseudoacacia proved to be particularly effective in enhancing microbial diversity, as the WZ14- R. pseudoacacia soils displayed the highest abundance and diversity of microorganisms compared to other treatments. In heavy metal-contaminated soils, the dominant bacterial phyla were identified as Proteobacteria , Bacteroidetes , Patescibacteria , Chloroflexi , and Acidobacteria . The inoculation with WZ14 significantly increased the relative abundance of Proteobacteria and Firmicutes . Soil microorganisms have been proven crucial in remediating heavy metal-contaminated soils, with combined mycorrhizal agents and plant remediation proving highly promising for reducing heavy metal content and enhancing the soil environment.",
"introduction": "1. Introduction The issue of heavy metal pollution in soil, driven by unnatural factors from rapid industrialization, is increasingly severe and requires urgent solutions. Soil, as the most abundant and diverse ecosystem on Earth, plays a crucial role in reflecting soil health and function through the dynamics of soil quality, surface vegetation, and microbial communities in complex environments ( Jiang et al., 2016 ). Mine remediation using microorganisms has been vastly studied over recent decade for remediation and ecological systems restoration at various mine sites ( Xiao et al., 2023 ). The unnatural uptake of heavy metals during mining is the primary cause of soil heavy metal contamination ( Rachelle et al., 2018 ; Wang et al., 2022 ). Heavy metals (HMs) contamination has led to severe environmental issues, including soil nutrient losses, sharp reductions in soil microbial diversity, and hindered plant growths ( Liu et al., 2019 ; Gonalves et al., 2020 ; Shuaib et al., 2021 ). Unlike organic pollutants, heavy metal pollution is characterized by difficult degradation, hidden, long-term, and high toxicity, and it is difficult to achieve the intended effect of complete removal in the short term in the remediation of soil heavy metal pollution ( Luo et al., 2019 ). As traditional remediation techniques like physical and chemical methods are increasingly limited in addressing soil heavy metal pollution, phytoremediation has gained significant attention as an alternative due to its ecological, economic, and sustainable advantages ( Fatima et al., 2016 ; Saxena et al., 2019 ; Xiao L. et al., 2021 ), emerging as one of the most promising remediation approaches. In addition, plants play a vital role in improving soil quality and optimizing the soil microbial community ( Schloter et al., 2018 ; Beiyuan et al., 2021 ). Various microorganisms living in the rhizosphere, can have beneficial effects on plant growth, and health and increase plant biomass production ( Evlat et al., 2023 ). Therefore, investigating changes in soil nutrients and microbial community structure during phytoremediation is crucial for successful ecological restoration. Most studies on phytoremediation for soil heavy metal remediation, especially on phytoextraction, have primarily focused on utilizing super-enriched plants (HMH) to extract soil heavy metals ( Duan et al., 2020 ; Atikur-Rahman et al., 2022 ). Although HMH has demonstrated favorable remediation results, certain studies have presented its limitations, such as slow growth and shallow root systems, which hinder its ability to reach deeper soil layers and extract heavy metals to a treatable level ( Słomka et al., 2012 ). In addition, Wood et al. (2016) observed that non-HMH species often extracted more heavy metals than HMH when measuring the net number of metals extracted per plant, with the number of extracted heavy metals closely correlated to plant biomass. Although HMH remains valuable in practical phytoremediation, non-HMH species with high biomass may represent a more suitable option for developing efficient phytoextractors in the future. Leguminosae, with 172 genera, 1,485 species, and 153 varieties in China ( Hei et al., 2019 ), are widely distributed throughout the country and hold significance in soil improvement and ecological restoration ( Dary et al., 2010 ; Cai et al., 2015 ). Legumes have exhibited excellent tolerance and effectiveness in heavy metal remediation, with some species exhibiting remediation capabilities comparable to HMH due to their robust root biomass ( Shi et al., 2012 ; Guo and Chi, 2017 ; Zeng, 2017 ). The advantages of legumes in this regard include: (i) their strong root biomass can produce abundant secretions that confer resistance to heavy metal pollution stress ( Pereira et al., 2006 ), with the dissolution of insoluble heavy metals in the soil by the organic acids released from the roots, which can enhance plant uptake of soil heavy metals. (ii) Legume roots contain abundant rhizobia, effectively improving soil quality ( Maynaud et al., 2013 ). These traits enable legumes to enhance, maintain, and develop stable soil systems. Therefore, employing legumes in phytoremediation holds significant potential for soil heavy metal remediation, contributing to the future refinement of plant species screening and phytoextraction in ecological restoration. In order to improve phytoextraction efficiency, the inoculation of characteristic microorganisms into plant roots is a common strategy ( Abhilash et al., 2012 ; Sessitsch et al., 2013 ). The identification and development of new, effective PGPR strains would be very efficient, providing several beneficial activities such as improved nutrient uptake, improved stress tolerance, enhanced plant growth, and resistance to fungal or bacterial pathogens ( Xiao C. Q. et al., 2021 ). The phytoremediation coupled with Pb-resistant phosphate-solubilizing bacteria effectively improved the efficiency of Pb bioremediation. For example, The inoculation of soil with strain LA greatly promoted the growth of ryegrass and sonchus, increased the concentration of bioavailable P and Pb in plants, and decreased the bioavailability of Pb in the soil ( Guo et al., 2021 ). In the “biotrophic bacteria-plant” mechanism, on the one hand, “bacteria” generally refers to inter-root biotrophic bacteria that can promote plant growth or improve soil quality, and the microorganisms promote plant growth through the production of iron carriers, phytohormones, organic acids and functional enzymes to enhance the remediation of heavy metals in the soil; On the other hand, some microorganisms themselves have the ability to dissolve and activate heavy metals, which reduces the content of heavy metals in the soil, increases the base of heavy metals that can be absorbed by the soil, and improves the possibility of plant uptake of heavy metals in the soil, so that the total amount of heavy metals in the soil is reduced to a harmless level ( Ni et al., 2019 ). Several studies have demonstrated the microorganisms can convert Cr (VI) in soil into Cr (III) or directly adsorb it in their bodies by means of bioreduction and biosorption, while plants uptake and accumulate Cr in tissues, thus reducing the total Cr content in the soil ( Wu et al., 2014 ). Li (2021) inoculated Variovorax paradoxus DE5 into Celosia argentea plants and found that DE5 was able to significantly promote the growth of Celosia argentea plants and enhance the uptake of soil cadmium by Celosia argentea . The results of the pot test showed that the application of Lactobacillus casei at 105 cfu·mL −1 reduced the pH of the soil, increased the soil enzyme activity, promoted the growth and development of cabbage mustard, and facilitated the remediation efficiency of cabbage mustard on Cd and Zn composite contaminated soil ( Liu, 2022 ). It has been shown that the application of organic acid-secreting endophytic bacteria effectively increased the conversion of insoluble Pb into the effective state of Pb, increased the content of soil heavy metal Pb in the effective state, and significantly improved the enrichment efficiency of oilseed rape for soil heavy metal Pb ( Ma Y. et al., 2022 ). Typically, the candidate microorganisms inoculated into plants originate from internal plant tissues or root secretions and function as “probiotics,” directly or indirectly influencing plant growth ( Li et al., 2017 ). However, such microorganisms often exhibit weak resistance against external pollution stress ( Zhang et al., 2011 ). Conversely, microorganisms exposed to and surviving in heavy metal-contaminated soil over extended periods may possess higher tolerance and resistance ( Ibiang et al., 2020 ). Therefore, understanding the source of microorganisms, their impact on heavy metal morphology, and their probiotic effects on plants prove beneficial for soil heavy metal remediation efforts. Our study addressed these limitations through a comprehensive approach. We first screened lead (Pb)- and cadmium (Cd)-tolerant microorganisms (resistant bacteria) for inoculum in experiments, using long-term heavy metal-contaminated soil as substrate. In addition, the heavy metal environment was simulated through indoor resistance growth experiments. Concurrently, outdoor pot experiments were conducted to investigate the impact of legumes ( Sophora xanthantha and Robinia pseudoacacia L.) and resistant bacteria-legume combinations on the remediation of heavy metal-contaminated soil. The objective of this study was (a) to determine whether tolerant bacterial inoculum would enhance the effectiveness of legume-based remediation, (b) to examine whether tolerant bacterial inoculum would lead to plant improvement in soil physicochemical properties and reduction in heavy metal content, and (c) to investigate key species of contaminated soil bacterial communities using high-throughput sequencing techniques. These studies have provided essential data and technical support for the utilization of legumes in the remediation of Cd and Pb in soils and the rehabilitation of mining sites.",
"discussion": "4. Discussion 4.1. Effect of soil physicochemical properties under combined remediation Soil physicochemical properties are fundamental indicators used to characterize soil nutrients, which in turn determine the overall fertility level of the soil. The presence of volatile nutrients, susceptible to environmental influences, directly reflects soil quality ( Glick, 2010 ). In this context, microorganisms act as essential “regulators” of geobiochemical processes, playing a vital role in maintaining soil vitality and ecological functions ( Liu et al., 2023 ). The total soil nutrient content, including total N and total P, and fast-acting nutrients demonstrated a decrease but without significant difference, consistent with the findings of Hussain et al. (2018) . Soil nutrient loss is directly related to heavy metal contamination, with higher contamination levels leading to more significant loss of soil nitrogen, phosphorus, potassium, and fast-acting nutrients. Moreover, in this study, the observed changes in soil phosphorus content after applying mycorrhizal fungi can be attributed to the enhancement of plant nutrient uptake by these fungi. This, in turn, promoted plant growth and optimized the plant’s ability to remediate heavy metals through phosphorus solubilization ( Table 1 ; Guo et al., 2013 ; Hu et al., 2019 ). Soil pH changes can directly affect the effective levels of soil nutrients, microbial activities, and the toxic effects of heavy metal ions. Under Cd and Pb contamination, both monoculture and combined remediation improved the soil organic carbon levels to some extent, with the combined system of inoculated bacteria and legumes showing the best results. However, propagating plants under heavy metal contamination reduced the alkaline-dissolved nitrogen levels of the soil. The inoculation of bacteria in plant monocultures inhibited the declining impact of plants on soil alkaline digestion of nitrogen. It also mitigated the decline in soil alkaline-dissolved nitrogen, consistent with the findings of Cui (2019) and Hussain et al. (2018) . This may be attributed to the ability of inoculation treatment to enhance the accumulation of soil carbon and nitrogen nutrient to varying degrees, promote plant root growth, and improve the antagonistic ability of plants. 4.2. Effect of soil Pb and Cd uptake under combined remediation Heavy metals in soil exist in various forms. The exchangeable form of heavy metals in soil is highly mobile and easily absorbed by plants, unlike the organic, carbonate, and ferro-manganese oxidation forms, which are less readily absorbed. Determination of the effective state content of heavy metals is an effective way to determine the extent of heavy metal contamination and to predict the impact of heavy metals on ecosystems ( Zhou et al., 2017 ). Therefore, the primary objective of remediation is to reduce the amount of active heavy metals in the soil, which in turn affects their crop uptake ( Wang et al., 2020 ). Numerous studies have demonstrated that the interaction between plant and microorganisms can enhance plant biomass and improve heavy metal tolerance, facilitating the absorption, fixation, and reduction of heavy metal concentrations in the soil. This process reduces the toxic effects of heavy metals ( Khan et al., 2017 ; Fu et al., 2022 ; Zheng et al., 2022a ). In this study, the combined treatment with WZ14 was more effective than individual plant treatments. The roots of R. pseudoacacia and S. xanthantha showed significantly higher levels of Pb and Cd than stems and leaves, which were the primary sites of soil Pb and Cd uptake. This difference was due to the limited translocation capacity of non-enriched plants, leading to the accumulation of Pb and Cd in their roots. These findings aligned with previous research ( Sun and Mao, 2015 ). Microorganisms in the soil play a crucial role in both the formation of soil humus and the mineralization of organic matter, the uptake and accumulation of soil Pb and Cd by R. pseudoacacia and S. xanthantha were strongly influenced by soil properties and fertility. In this study, the addition of WZ14 bacterial agent increased the OC content of soils occupied by R. pseudoacacia and S. xanthantha . Additionally, it enhanced the metabolic processes of inter-root microorganisms and their byproducts. These changes influenced the migration and release mechanisms of Pb and Cd, leading to reduced toxicity of heavy metals in the soil. 4.3. Correlation analysis of soil environmental factors and soil bacterial communities Strong correlations exist between plant and soil microbial communities, with plant root secretions and residues influencing the function and structure of soil microbial communities ( Bian et al., 2018 ). In this study, the addition of WZ14 bacterial agent had contrasting effects on soil bacterial communities of R. pseudoacacia and S. xanthantha . Although it enriched the richness and diversity of soil bacterial communities in R. pseudoacacia , it reduced the richness and diversity in S. xanthantha . These changes could be due to short-term responses of bacterial communities to environmental and pollution changes, reflecting their microenvironmental conditions ( Yao et al., 2019 ). Although the exogenous soil microorganism WZ14 partially promoted the development of heavy metal-tolerant microorganisms, it also intensified competition among soil microorganisms. Consequently, less adaptable microorganisms struggled to cope with environmental changes and experienced a decline in abundance and diversity ( Zhang et al., 2020 ). The majority of bacterial communities in this study showed a significant negative correlation with soil physicochemical factors, possibly due to the ability to decompose and utilize soil nutrients for energy, either for themselves or plants ( Yao et al., 2022 ). The results of this study revealed that Proteobacteria had the highest relative abundance in both monoculture and grafted-plant combination models, indicating the prevalence of Proteobacteria in heavy metal-contaminated soils and their tolerance to high levels of Cd and Pb contamination ( Gupta et al., 2017 ). Proteobacteria emerged as the dominant phylum in the heavy metal-contaminated soil of the mining area and positively contributed to improving soil contamination status. However, its relative abundance was negatively correlated with the overall soil quality ( Yu et al., 2020 ; Ma J. D. et al., 2022 ). In contrast, our study discovered the highest abundance of Proteobacteria in the soil of R. pseudoacacia and S. xanthantha when the heavy metal contamination level decreased after the application of the microbial inoculant. This discrepancy may arise from the specific focus of our experiment on a combined Pb and Cd contamination pattern, resulting in different responses of bacterial communities to Pb and Cd under heavy metal-contaminated conditions."
} | 5,279 |
37419944 | PMC10329038 | pmc | 3,431 | {
"abstract": "Social entrainment is important for functioning of beehive organization. By analyzing a dataset of approximately 1000 honeybees ( Apis mellifera ) tracked in 5 trials, we discovered that honeybees exhibit synchronized activity (bursting behavior) in their locomotion. These bursts occurred spontaneously, potentially as a result of intrinsic bee interactions. The empirical data and simulations demonstrate that physical contact is one of the mechanisms for these bursts. We found that a subset of honeybees within a hive which become active before the peak of each burst, and we refer to these bees as \"pioneer bees.\" Pioneer bees are not selected randomly, but rather, are linked to foraging behavior and waggle dancing, which may help spread external information in the hive. By using transfer entropy, we found that information flows from pioneer bees to non-pioneer bees, which suggest that the bursting behavior is caused by foraging behavior and spreading the information through the hive and promoting integrated group behavior among individuals.",
"introduction": "Introduction It is well known that populations of agents, whether animate or inanimate, sometimes exhibit complex self-organization and emergence phenomena. An example of such self-organization in social organisms is known as \"social entrainment/synchronization\" 1 . The synchronization of group behavior is essential for a group's daily survival, as it enables the achievement of common goals, such as breeding, defense against predators, collective hunting, and energy conservation through gathering. Moreover, for a group to function as a superorganism beyond a mere collection of individuals, synchronization phenomena are likely indispensable. Groups of starlings and sardines are well-known examples of groups that function as superorganisms 2 , 3 . Grouping reduces the susceptibility to predator attacks. Drosophila increases the probability of mating by synchronizing its circadian rhythm with that of the group 4 – 6 . Also, it is known that honey bees use waggle dance to notify the location of feeding area and to synchronize circadian clock within a hive is needed to transfer the information 7 . In honeybees, numerous studies have been published regarding temporal synchronization among groups within the hive, particularly focusing on circadian rhythms. Southwick et al. (1987) suggested that bees use direct contact or vibrations, rather than volatile substances, to synchronize activity levels among individuals 8 . Korst et al. suggested that trophallaxis is a major form of communication mechanism among honeybees 9 . A recent study suggests that direct contact, for example via contact pheromones or tactile communication, is not necessary for the synchronization of circadian rhythm in the hive 10 . Furthermore, in some species of ants, which are also eusocial insects like honeybees, synchronization not only occurs on long timescales such as circadian rhythms but also in the form of self-organizing synchronization of periodic activities that occur in approximately 30-min cycles. The phenomenon was discovered in the 1990s, yet it is believed that these activity cycles are not functional but rather the inevitable outcome of interactions within social groups 11 , 12 . We can find parallels to the complex processes occurring within the brains and there are studies that discuss animal groups in comparison to cognitive science research 13 , 14 . The brain is well-known for generating a large number of periodic/non-periodic synchronization activities, which are derived not only from responses to external stimuli but also from so-called \"spontaneous activities\" that regularly carry out firing activities without external inputs 15 , and this phenomenon is believed to play a role in stabilizing the performance of cortical circuits 16 . In the past decade, tracking technologies have made significant progress in tracking individuals within biological populations, allowing for more detailed analysis of individual behavior within the group 17 – 20 . This consequently enables us to examine which individual (micro) characteristics contribute to self-organization at the population (macro) level. Gernat et al . (2018) developed a high-throughput automatic monitoring system of artificial honeybee hive of a single layer in a transparent cage. By attaching a \"bCode\" device (a custom matrix barcode) to the thorax of every individual bee in the hive, they succeeded in tracking each individual bee’s positions, speeds, and orientations using the recorded digital images. Gernat et al . used the tracking system to study bees’trophallaxis (mouth to mouth interaction to transfer food or chemicals) networks and calculate how often they communicated. They found that bees communication occurred in a temporally intermittent manner, which they refer to as bursts, much like human communication networks 17 . In the present study, using the same large dataset as their research, we report on the global synchronous activity of locomotion, here we called ‘burst’, found in a population of European honeybees ( Apis mellifera) and its relationship with individual activity. This collective burst is an extremely robust phenomenon observed under five different conditions in an artificial hive of approximately 1000 bees (Fig. S1 ). The burst occurs on a timescale smaller than the diurnal rhythm, such as circadian rhythm, and appears to be different from the oscillatory activity observed in the collective behavior of ants 11 , 12 in that the frequency of their bursts is Poissonian rather than periodic. Furthermore, this burst does not simply occur in response to external stimuli. The paper investigates the mechanism behind these bursts and analyzes whether the bees that initiate each burst called “pioneer bee (P)” (bees with increased activity prior to the burst) are randomly selected or if there are biological features that determine their selection. Moreover, based on the tracking data, we identified well-known honeybee biological behaviors, such as foraging, waggle dance, and dance following, and then examined to what extent these behaviors were performed by P. If the behavior of a small group of pioneer bees (micro), the bursting behavior, influences the behavior of the hive itself (macro), then a detailed study of pioneer bees of global bursts may provide an indicator for understanding the state of the hive.",
"discussion": "Discussion In this study, we investigated the bursting behavior of honeybees, a form of social synchronization in locomotion activities, by analyzing the tracking data of approximately 1,000 honeybees across five trials. These bursts were predominantly spontaneous, and that these bursts in honeybees occurred irrespective of the hive entrance being open or closed. We aimed to understand the underlying mechanisms that drive these bursts, their relationship with the roles of bees within the hive, and the potential implications for the overall functioning of the hive. Initially, to examine global bursts, we calculated information flows between the time series of bursts (i.e., global kinetic energy, \\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}$${K}_{G}(t)$$\\end{document} K G ( t ) ), dance events and the number of bees outside the hive (O). Our findings indicate that events transpire in the following sequence after a dance event is confirmed: dance, burst, and departure from the hive. Furthermore, we demonstrated that physical contact could be a contributing factor to spontaneous bursts, both from experimental analysis and agent-based model simulations (see Supplementary Information for details). These results suggest that bees becoming active prior to a spontaneous burst may trigger the burst. We designated these bees as \"pioneer bees\"(P) and employed non-negative factorization to identify them and investigated the characteristics of P. We examined each burst from the perspective of the members of P and the condition (for example, the entrance was opened or not) of the hive. During the phase before the hive entrance opened, bees had no direct interaction with the external environment, yet spontaneous burst events still occurred. At this stage, forager (F), waggle dancer (W), and dance follower (DF) did not exist. We speculate that these bursts may resemble the spontaneous periodic activity cycle observed by Cole in isolated ant populations in 1991 12 . The 1–2 days after the entrance opened, approximately 30% of 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}$${\\mathrm{P}}_{\\mathrm{b}}$$\\end{document} P b performed foraging before the burst. However, based on the multidimensional scaling (MDS) analysis results, it seemed that the selection of the members of P at this stage was slightly organized compared to random selection, but no significant difference was observed. When we examined the spatial distribution of foraged P (FP bees, 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}$${\\mathrm{F}}_{\\mathrm{b}}\\wedge {\\mathrm{P}}_{\\mathrm{b}}$$\\end{document} F b ∧ P b ) and non-foraged P (NFP) inside the hive, both were aggregated near the entrance. From these findings, we hypothesize that the burst at this stage was not so much ordered as it was a “spontaneous commotion” resulting from the hive being opened to the outside. The 3–4 days after the entrance opened, we observe dancing and following behaviors, thus P members included not only F but also W and DF. Multidimensional scaling (MDS) analysis results revealed that P members were somewhat organized, unlike the bursts in a closed case. Moreover, when examining the spatial distribution inside the hive, FP was located near the entrance, as before, while NFP was situated toward the back of the hive. When assessing the information flow between FP, NFP, and non-pioneer (NP) using a method called transfer entropy, we discovered a significant flow of information from FP to NFP (Note that there was no significant information flow between FP and NFP during the 1–2 days after the entrance opened) (Fig. S11 ). The tracking data used for this analysis only covers the inside of the hive, so it cannot demonstrate whether bees which performed foraging or dance actually found the feeding site and brought this information back to the hive. However, considering that the waggle dance is known to convey information about external feeding sites to bees inside the hive, and taking into account the results thus far including the temporal correlation between the dance event and the burst, it is plausible that bursts are utilized to efficiently convey such information. Furthermore, considering the information flow between FP and NFP, it is suggested that the commotion caused by returning bees near the entrance of the hive spreads the burst to the interior of the hive. In other words, we may argue that these bursts were induced by “external information ” rather than purely self-excited. Note that not all bursts following the appearance of W are necessarily information-induced. Figure 3 B showed that there were bursts which P did not contain W even after their appearance inside the hive. Even after bursts with W occur, there may still be spontaneous bursts that are not apparently “information induced”. From the perspective of regulating the division of labor, addressing the question of how workers acquire information on colony requirements, Robinson (1992) proposed a hypothesis that sampling behavior (i.e., exploration behavior of states in the hive) with social interaction may be facilitated by worker activity synchrony observed in ant colonies 34 . Furthermore, “mature hive” may indicate that the activities of individual bees are constrained by the entire hive (downward causation) 35 , and as proposed by E.O. Wilson et al . it would imply role differentiation 36 . Burst phenomena in bee colonies may also promote or regulate division of labor; however, to prove this, it is necessary to prepare a hive in which bursts are intentionally suppressed and compare it with a normal hive. Additionally, since this study only has data for the first seven days after bees were born, analyzing longer tracking data can provide more precise insights into the hive's differentiation. Lastly, the question of which indicators can be used as measures of the state of populations (hives) is an attractive challenge in the research of collective behavior. The pioneer bees' behavior and composition may serve as indicators of the state of the hive, providing valuable insights into the dynamics of collective behavior in honeybee colonies."
} | 3,311 |
22716969 | null | s2 | 3,433 | {
"abstract": "Magnetotactic bacteria (MTB) use magnetosomes, membrane-bound crystals of magnetite or greigite, for navigation along geomagnetic fields. In Magnetospirillum magneticum sp. AMB-1, and other MTB, a magnetosome gene island (MAI) is essential for every step of magnetosome formation. An 8-gene region of the MAI encodes several factors implicated in control of crystal size and morphology in previous genetic and proteomic studies. We show that these factors play a minor role in magnetite biomineralization in vivo. In contrast, MmsF, a previously uncharacterized magnetosome membrane protein encoded within the same region plays a dominant role in defining crystal size and morphology and is sufficient for restoring magnetite synthesis in the absence of the other major biomineralization candidates. In addition, we show that the 18 genes of the mamAB gene cluster of the MAI are sufficient for the formation of an immature magnetosome organelle. Addition of MmsF to these 18 genes leads to a significant enhancement of magnetite biomineralization and an increase in the cellular magnetic response. These results define a new biomineralization protein and lay down the foundation for the design of autonomous gene cassettes for the transfer of the magnetic phenotype in other bacteria."
} | 321 |
34796477 | PMC9299204 | pmc | 3,436 | {
"abstract": "Abstract Lignin valorization may offer a sustainable approach to achieve a chemical industry that is not completely dependent on fossil resources for the production of aromatics. However, lignin is a recalcitrant, heterogeneous, and complex polymeric compound for which only very few catalysts can act in a predictable and reproducible manner. Laccase is one of those catalysts and has often been referred to as an ideal “green” catalyst, as it is able to oxidize various linkages within lignin to release aromatic products, with the use of molecular oxygen and formation of water as the only side product. The extent and rate of laccase‐catalyzed lignin conversion were measured using the label‐free analytical technique isothermal titration calorimetry (ITC). IITC provides the molar enthalpy of the reaction, which reflects the extent of conversion and the time‐dependent power trace, which reflects the rate of the reaction. Calorimetric assessment of the lignin conversion brought about by various fungal and bacterial laccases in the absence of mediators showed marked differences in the extent and rate of conversion for the different enzymes. Kraft lignin conversion by Trametes versicolor laccase followed Michaelis–Menten kinetics and was characterized by the following thermodynamic and kinetic parameters Δ H \n ITC = −(2.06 ± 0.06)·10 3 kJ mol −1 , K M \n = 6.6 ± 1.2 μM and V max \n = 0.30 ± 0.02 U/mg at 25°C and pH 6.5. We envision calorimetric techniques as important tools for the development of enzymatic lignin valorization strategies.",
"introduction": "1 INTRODUCTION Biorefineries create lignin waste streams, the amount of which is expected to exceed 200 megatons by 2022 (Bruijnincx et al., 2016 ). These waste streams are mainly burnt to produce energy and utilized commercially only in a limited manner (2%) although it could serve as a source of valuable aromatics (S. Xie et al., 2016 ). The chemical industry is currently completely dependent on fossil resources to satisfy the demand for aromatics and the valorization/depolymerization of lignin could offer a sustainable alternative for the future (Roberts & Caserio, 2021 ). The move to using lignin waste streams could render the biorefineries more economically viable as lignin can gain up to 10 times in value upon valorization (Nanayakkara et al., 2014 ). Lignin's structure is complex as it is a heterogeneous polymer made of aromatic units (Figure 1 ); p ‐coumaryl, coniferyl, and sinapyl alcohol monomers are coupled together with mostly ether and C‐C linkages (de Gonzalo et al., 2016 ; Den et al., 2018 ; Venkatesagowda & Dekker, 2020 ). The resulting lignin polymer constitutes different proportions of p‐hydroxyphenyl (H), guaiacyl (G), and syringyl (S) units. The inherent heterogeneity, arising due to the multiple types of bonds that may form among the monomeric groups, leads to the recalcitrant nature of the lignin (S. Xie et al., 2016 ). Figure 1 The basic monomeric units of lignin. These units are linked to one another to make lignin. The amount of each monomer and the type of linkages among them, depend upon the source of lignin, its treatment, and handling (Bugg & Rahmanpour, 2015 ) Only a few biocatalysts act upon lignin in a predictable and reproducible manner (Bruijnincx et al., 2016 ; Bugg & Rahmanpour, 2015 ; Cajnko et al., 2021 ; Christopher et al., 2014 ; de Gonzalo et al., 2016 ; Pollegioni et al., 2015 ; Venkatesagowda & Dekker, 2020 ). Among the enzymes which act upon lignin, laccases are important in nature to catalyze lignin conversion by plants, fungi, and bacteria (de Gonzalo et al., 2016 ; Mate & Alcalde, 2017 ; Pollegioni et al., 2015 ; Rodgers et al., 2010 ; Thurston, 1994 ). Other oxidative enzymes, such as lignin peroxidases (LiPs), manganese peroxidases (MnPs), and versatile peroxidases (VPs), are extracellular and use H 2 O 2 as co‐substrate whereas laccases use molecular oxygen. LiP generally possesses a very high redox potential of the active site heme cofactor (1.2 V vs. NHE at pH 3.0) compared to the other enzymes (Kersten et al., 1990 ). This property enables it to catalyze the oxidation of non‐phenolic aromatic compounds directly, without the need for a mediator. Laccases from plants are generally involved in polymerization (i.e., production of lignin), and those from fungi and bacteria are generally involved in depolymerization (i.e., degradation of lignin), which reflects their physiological role (Abdel‐Hamid et al., 2013 ; Zakzeski et al., 2010 ). For this study, the focus was on lignin conversion brought about by fungal and bacterial laccases to add to the body of knowledge regarding lignin depolymerization/valorization in a more economical and sustainable manner. Oxidation by laccases is brought about with molecular oxygen and low‐molecular weight products, aromatic in nature, are formed with water as side product. Laccase possesses moderately high redox potentials of the active site T1 Cu center (0.42–0.79 V) and can oxidize, in principle, only phenolic residues of lignin directly. Thus, in most studies of laccase‐catalyzed lignin conversions, mediators have been used to mitigate the challenges associated with the large size of lignin and the high redox potential of the non‐phenolic reactions (Cañas & Camarero, 2010 ; Christopher et al., 2014 ; Hilgers et al., 2018 ; Hilgers, Kabel, et al., 2020 ; Hilgers, van Erven, et al., 2020 ; Rich et al., 2016 ; Zhu et al., 2020 ). Laccase catalyzes the one‐electron oxidation of the mediator to cation radicals which subsequently react with the high redox potential groups within bulky lignin. ABTS (2,2ʹ‐azino‐bis(3‐ethybenzothiazoline‐6‐sulphonic acid)) has been frequently used as a mediator to enhance the substrate scope of the enzyme (Hilgers et al., 2018 ; Könst et al., 2013 ). Mediators also prevent re‐polymerization of the phenoxy radical products of laccase (Bourbonnais et al., 1995 ). Most mediators, however, show side‐reactions with substrates and products and even inactivate enzymes in certain cases (Christopher et al., 2014 ; Hilgers et al., 2018 ). Thus, in this report, lignin conversion by laccases was studied without the addition of mediators. For measuring the rates of enzymatic reactions with difficult, that is, complex and/or non‐ (or partially) soluble substrates, such as lignin, many of the conventional assay techniques are not suitable. Generally, the change in color or fluorescence of the substrates or the products are followed in time—the absence of such can be tackled by using labeling, coupled reactions, or chemical analysis. However, labeling or using coupled enzymatic or chemical reactions is not always possible and may require extensive optimization for each reaction. Over the past two decades, several non‐spectrophotometric label‐free methods to measure enzyme activity have been developed (Das et al., 2002 ; He et al., 2014 ; Hennig et al., 2007 ; Leung et al., 2013 ; Lu et al., 2014 ; Orosco et al., 2009 ; Syahir et al., 2015 ) and among them, isothermal titration calorimetry (ITC) allows reactions to be followed based upon a universal thermodynamic parameter, the molar enthalpy of the catalyzed reaction, ΔH (Bianconi, 2007 ; Todd & Gomez, 2001 ). In an enzyme‐catalyzed reaction, the total heat generated (or consumed) is dependent on many factors, but, if the temperature and pH are well controlled, it is proportional to the total conversion that takes place under the influence of the enzyme and the molar enthalpy of the reaction itself. The ITC set‐up contains a sample cell and a reference cell and the temperature difference among them is monitored continually using a thermoelectric device. Any observed difference is corrected by a PID (proportional integral derivative) controller which changes the power supplied to the sample cell to maintain a constant temperature difference, for example, an exothermic reaction in the sample cell is countered by a lowering of the power to compensate for the heat generated. The area under the curve of the power supply to maintain isothermal conditions for the duration of the experiment, gives information about the total heat released/consumed during the experiment. The molar enthalpy is calculated based on the principles of Wiseman from this area under the curve (Wiseman et al., 1989 ). Furthermore, with proper calibrations, the power readings during calorimetric experiments are reflections of the heat flows which, in turn, are directly proportional to the rates of the reaction (Hobbs et al., 2013 ; Honarmand Ebrahimi et al., 2015 ; Watt, 1990 ). However, in practice, the measured heat‐flow is not a real‐time reflection of the rate of the enzyme. The instrumental response time has to be taken into account, which has not always been properly done in literature (Todd & Gomez, 2001 ). An empirical solution to this problem is the IrCal method (Honarmand Ebrahimi et al., 2015 ) which is based on the fact that the initial ITC power data can be used to obtain an apparent initial rate of a reaction. An empirical calibration constant can be obtained for the ITC setup using a standard enzyme‐catalyzed reaction carried out both in the ITC device and a conventional assay setup, for example, a spectrophotometer, under the same reaction conditions. This empirical calibration constant can subsequently be used for reporting the rates of other reactions catalyzed by their respective enzymes within the same ITC setup. Since the rate of an enzyme‐catalyzed reaction increases linearly with the enzyme concentration, the ITC power changes that occur upon adding a specific concentration of substrate to different enzyme concentrations also correlate linearly. Under conditions of the same temperature, pH, and enzyme concentration, changes in the rates of the enzyme‐catalyzed reactions can be brought about by changing the concentration of the substrate and these would be reflected in the ITC power measurements. As reaction rates brought about by changes in substrate concentration follow appropriate rate laws describing the enzyme kinetics, the power measurements may thus be a means, in principle, of determining the appropriate rate laws (Honarmand Ebrahimi et al., 2015 ). To summarize, enzyme calorimetry using ITC and IrCal gives us two parameters: (1) the area under the curve—reflecting the total conversion (and equilibrium) and the molar enthalpy of the reaction and (2) the power trace over time which reflects the rate of the reaction and which can be further utilized to determine appropriate kinetic parameters of the enzyme‐catalyzed reaction. Here we present the study, with ITC and IrCal method of analysis, of various bacterial and fungal laccases acting upon lignin, without the use of mediators. ABTS oxidation was used as a reference substrate for laccase as the formation of the blue‐green colored ABTS •+ radical can be followed by both UV‐visible spectrophotometry and ITC (Honarmand Ebrahimi et al., 2015 ).",
"discussion": "4 DISCUSSION The extent of the recalcitrant nature of lignin varies according to the source (e.g., softwood vs. hardwood, age of wood), treatment, and handling (Bugg & Rahmanpour, 2015 ; Christopher et al., 2014 ). It is anticipated that the full potential of lignin can be realized when efficient and viable depolymerization processes are in place to obtain the low molecular weight aromatic components which are valuable and/or which could serve as building blocks after further downstream processing (Bruijnincx et al., 2016 ; Bugg & Rahmanpour, 2015 ; Lora, 2008 ; S. Xie et al., 2016 ). Although various routes of valorization are available (Cao et al., 2018 ; Wendisch et al., 2018 ; S. Xie et al., 2016 ), enzymatic valorization is desirable since chemical and thermal treatments are generally harsh in comparison to conditions of enzymatic reactions and they lack specificity and efficiency. Biocalorimetry has been used previously to measure the enzymatic action of laccases, and also to measure the biological degradation of wood (Wadsö et al., 2017 ). The action of the plant laccase from Rhus vernicifera on different small phenolic substrates has been measured using microcalorimetry (Wang et al., 2000 ; Wong & Yu, 1999 ; X. Y. Xie et al., 2002 ). Molar enthalpy values in the range of −40 to −550 kJ mol −1 were found, depending on the number of phenolic groups. The oxidation of the lignin model compound syringic acid by Galerina sp HC1 and T. versicolor laccase was measured using ITC previously (Volkova et al., 2012 ). A molar enthalpy of −230 kJ mol −1 was reported. However, this value is based on an unusually slow conversion of 8 h, after an initial sharp peak of less than 30 min. Since the authors showed independently that the reaction was complete within 30 min we attribute the very slow reaction to subsequent non‐catalyzed radical polymerization reactions. According to the authors, the enthalpy of the initial 30 min reaction is 15% of the total, resulting in an apparent molar enthalpy of −35 kJ mol −1 . The conversion of lignosulfonic acid by fungal Lentinus sp. nLcc4 laccase was measured using the ITC multi‐injection approach (Maestre‐Reyna et al., 2015 ). An apparent molar enthalpy of −105 kJ mol −1 was reported. The kinetic parameters for lignosulfonic acid conversion were k \n \n cat \n = 0.234 s −1 and K \n \n M \n = 56.7 µM. Although lignosulfonic acid is a heterogeneous lignin‐derived mixture of soluble compounds, the authors have performed docking studies with their lignosulfonic acid substrate, which suggests it has a very defined structure containing one sinapyl and one coniferyl unit and has one phenol group, two methoxy groups, and two sulfonate groups (CAS 8062‐15‐5). Quantitative measurements of the reaction kinetics of fungal laccases T. versicolor and Ganoderma lucidum with monomeric lignin units and a dimeric lignin model compound OH‐dilignol was performed using HPLC‐MS (Perna et al., 2018 ). The conversion of OH‐dilignol with T. versicolor laccase showed Michaelis–Menten kinetics with apparent kinetic parameters K \n \n M \n = 12.89 µM and V \n max = 0.39 µM min −1 . Using a different MS approach the kinetics of the conversion of a labeled lignin model substrates by T. versicolor laccase was measured. This approach involves nanostructure‐initiator mass spectrometry (NIMS), which involves surface immobilization of the labeled substrates and products and laser desorption/ionization MS analysis (Deng et al., 2018 ). Using this approach it was possible to determine the time‐dependent formation of specific dimerization and degradation products. Electron paramagnetic resonance (EPR) spectroscopy has been shown to be a label‐free method to directly measure the kinetics of radical formation in lignin by laccases (Munk et al., 2017 ). The kinetics of lignin (organosolv) conversion by fungal laccases T. versicolor , M. thermophila , and Ganoderma lucidum has been measured using this technique. The apparent kinetic parameters based on the extent of radical formation were determined to be V \n \n max \n = 0.30 ± 0.04 µM min −1 and K \n \n M \n \n = 73.0 ± 5.9 mM lignin for T versicolor laccase (Perna et al., 2019 ). Different apparent kinetic parameters were obtained for the different laccases when compared after normalization for their syringaldazine oxidation activities. These relative kinetic parameters showed that T. versicolor laccase had a five to tenfold higher specific activity than M. thermophila and G. lucidum laccase. We find that the specific activity for lignin conversion by T. versicolor laccase is circa threefold higher than for M. thermophila laccase after normalization for ABTS oxidation activity (Table 3 ). Without normalization M. thermophila laccase has a circa 80‐fold higher specific activity than T. versicolor laccase, although this represents a much lower extent of lignin conversion. The modeling of the enzyme kinetics assuming (global) Michaelis–Menten is a simplification and more advanced kinetic modeling will be necessary to fully understand the process (Cajnko et al., 2021 ). Lignin degradation products have been shown to act as a competitive inhibitor for laccase catalyzed ABTS oxidation, and therefore are likely to influence the lignin oxidation kinetics as well (Pamidipati & Ahmed, 2020 ). Here we present the conversion of kraft lignin (with low sulfonate content) by three different laccases, which resulted in an apparent molar enthalpy of −263 to −2063 kJ mol −1 depending on the type of enzyme. On the basis of the molar enthalpy values that have been previously reported for laccase oxidation of phenolic substrates (−50 to −100 kJ mol −1 per phenolic group), this may correspond to the conversion of 3–5 to 21–41 phenolic groups per 10 kDa unit of lignin. The large difference in the extent of conversion of the lignin substrate by the different laccases can be partially explained by the differences in their T1 Cu reduction potentials. Another explanation could be the accessibility of the substrate or substrate recognition. The performance of the enzymes at their optimal pH and temperature may be different, but here we chose to compare the enzymes under identical conditions. The slower uncatalyzed polymerization or depolymerization reactions brought about by the phenoxy radicals are more difficult to analyze using ITC. In principle, one can distinguish the two processes, enzyme‐catalyzed oxidation, and uncatalyzed radical reactions, by performing either relative short measurements (minutes) and very long measurements (hours/days). This has been observed for the oxidation of syringic acid by laccase using ITC (Volkova et al., 2012 ). The effect of the uncatalyzed radical reactions on the enthalpy is difficult to predict. As the polymerization and depolymerization reactions involve different reaction types, each of which can be either exothermic or endothermic (also depending on conditions), the observed thermogram will be the net result of all these contributions. Most studies describing lignin degradation using laccase involve so‐called laccase‐mediator systems (Christopher et al., 2014 ). Mediators are small molecules that undergo one‐electron oxidation to form cation radicals with some stability. The radical forms of the mediators subsequently chemically react with the different linkages in lignin. Also, small molecule lignin degradation products can act as a natural mediator. It has been suggested that the phenolic units of lignin are initially oxidized, followed by the more sluggish transformation of the non‐phenolic part. We may assume that, in the absence of added mediators, laccase will only be able to oxidize phenolic lignin units within the timeframe of the measurements. Degradation of Kraft lignin using different isoforms of T. versicolor laccase with ABTS as mediator has been reported (Bourbonnais et al., 1995 ). Lignin conversion with laccase without added mediator increased the average molecular weight of the product, while it decreased after the addition of ABTS. These reactions were performed up to 6 days, so they do reflect a different timescale than the reactions reported here. Although this study focused on laccase as a lignin converting enzyme, the outcomes have implications for both other lignin converting enzymes and enzymatic conversions of other complex polymeric substrates. ITC is a powerful tool to gain independent information on the extent and rate of conversion of complex substrates such as lignin. In principle, this method can be adapted for reporting on enzymes that catalyzes similar reactions, for example, similar enzymes from different sources or genetically engineered enzyme variants. A number of lignin‐converting enzymes have been identified, including lignin peroxidase, versatile peroxidase, dye decolorizing peroxidase, and Mn peroxidase, β‐etherase, C‐C hydrolase, O‐demethylase, and other oxidative enzymes in lignin degradation pathways (Pollegioni et al., 2015 ). We envision that by combining the conversion and rate information obtained by ITC with advanced analytical approaches, such as the aforementioned MS and EPR approaches as well as NMR and HPLC to establish the precise nature of the observed transformation, a very powerful approach for the biotransformation of lignin compounds and materials can be established."
} | 5,113 |
34344892 | PMC8333059 | pmc | 3,437 | {
"abstract": "The rapid development of Internet of Things and artificial intelligence brings increasing attention on the harvesting of distributed energy by using triboelectric nanogenerator (TENG), especially the direct current TENG (DC-TENG). It is essential to select appropriate triboelectric materials for obtaining a high performance TENG. In this work, we provide a set of rules for selecting the triboelectric materials for DC-TENG based on several basic parameters, including surface charge density, friction coefficient, polarization, utilization rate of charges, and stability. On the basis of the selection rules, polyvinyl chloride, used widely in industry rather than in TENG, is selected as the triboelectric layer. Its effective charge density can reach up to ~8.80 mC m −2 in a microstructure-designed DC-TENG, which is a new record for all kinds of TENGs. This work can offer a basic guideline for the triboelectric materials selection and promote the practical applications of DC-TENG.",
"introduction": "Introduction With the development of the Internet of Things (IoTs) and artificial intelligence (AI), our daily life is embracing unimaginable and complex distributed arrays of electronics and sensors, which drive urgent demand for distributed energy harvesters 1 – 3 . Triboelectric nanogenerator (TENG), as the energy supply units and the self-powered sensor units, represents one of the core groups of potential technologies with great potential applications to support the fast development of IoTs and AI 4 – 7 . In addition, its high efficiency at low frequency, light-weight, and low cost make it a prime candidate for integration with multifunctional electronics 8 – 13 . According to the working mechanism and output signal type, TENG can be divided into alternative current TENG (AC-TENG) based on contact electrification and electrostatic induction effects, and direct-current TENG (DC-TENG) based on contact electrification and electrostatic breakdown effects. Among them, DC-TENG possesses more advantages, such as driving electronics directly without a rectifier unit, anti-electromagnetic interference of output signal, and no limitation of dielectric breakdown 14 – 17 . The excellent output performance of TENG, namely high effective charge density, is the essential prerequisite as an energy harvester or a self-powered sensor 18 – 21 . Very recently, the limiting factor of charge density of DC-TENG ( σ DC-TENG ) has been described as 13 : 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}$${\\sigma }_{{{{{{\\rm{DC}}}}}}-{{{{{\\rm{TENG}}}}}}}=k\\,\\times \\,{{{{{\\rm{min}}}}}}({\\sigma}_{{{{{{\\rm{triboelectrification}}}}}}},{\\sigma}_{{{{{{\\rm{c}}}}}},\\;{{{{{\\rm{electrostatic}}}}}}\\;{{{{{\\rm{breakdown}}}}}}})$$\\end{document} σ DC − TENG = k × min ( σ triboelectrification , σ c , electrostatic breakdown ) where k is the electrode structure factor, σ triboelectrification is the triboelectrification charge density, and σ c, electrostatic breakdown is the collected effective charge density by the electrostatic breakdown. The k and σ c, electrostatic breakdown have been optimized to significantly enhance the output of DC-TENG by electrode structure design or external environment controlling (e.g., temperature or atmosphere), and the effective charge density has reached a new milestone of 5.4 mC m −2 12 , 15 , 22 . The σ triboelectrification is the basic parameter for contact electrification effect, which largely determines whether the electrostatic breakdown (generally is air breakdown for DC-TENG in the air) occurs or not 15 , and has a close relationship with the species of triboelectric materials. Thus, the triboelectric materials series has been provided as a guideline for the materials selection of TENG 23 , however, it is only applicable for the AC-TENG, especially for contact-separate AC-TENG. As for the DC-TENG, a kind of sliding TENG derived from the electrostatic breakdown, except for the σ triboelectrification , more complex parameters (friction properties, dielectric properties, etc.) should be considered in order to select suitable triboelectric materials. Thus, providing a guideline for selecting appropriate triboelectric materials is the key to designing a high-performance DC-TENG and promoting the practical applications of DC-TENG. In this work, we propose a set of selection rules to determine whether a kind of triboelectric material is suitable for DC-TENG. Taking the surface charge density, friction coefficient, strength of polarization, the utilization rate of triboelectric charges, and stability as basic parameters, we find that the triboelectric material of polyvinyl chloride (PVC) film, which is widely utilized in the industry but not in TENG field, showed an excellent DC output performance. The effective charge density for a microstructure-designed DC-TENG with PVC can reach ~8.80 mC m −2 , setting a new TENG’s record. The proposed comprehensive model for selecting triboelectric materials not only can be used for DC-TENG to achieve high-performance DC-TENG, but also can provide a guideline for the applications of DC-TENG in practice.",
"discussion": "Results and discussion Judgement rules of triboelectric materials in DC-TENG The working mechanism of DC-TENG is shown in Fig. 1a , which couples the contact electrification and electrostatic breakdown effects (detailed explanation is shown in Supplementary Note 1 ). The movement of DC-TENG is a kind of sliding process, and therefore the low friction coefficient ( μ ) is essential to obtain a high-efficiency DC-TENG. Here, taking generally used copper as the friction electrode (FE), the μ of various commercial polymers under different loads are tested (Fig. 1b ). The test diagram is shown in Supplementary Fig. 1 and the calculation of μ is shown in Supplementary Note 2 . The average μ of most dielectric films is <0.4 (Supplementary Fig. 2 ), and the polytetrafluoroethylene (PTFE) film and polydimethylsiloxane (PDMS) film presents the smallest μ of ~0.17 and the largest μ of ~1.35, respectively. The surface roughness Ra of the triboelectric materials has no obvious influence on their friction coefficient (Supplementary Fig. 3 ). The Ra value of PTFE is ~0.15 μm, which is larger than that of most of the triboelectric materials used in this work, but the μ of PTFE is the smallest. Large μ between the copper electrode and polymer film will increase the wear process during the sliding movement. Thus, the films with μ lower than 0.4 are utilized as the triboelectric materials to confirm their surface charge density ( σ SCD ) during the triboelectrification process. Consequently, the PDMS, Nitrile rubber, and poly(styrene) (PS) films are left out, although PDMS has been used as good triboelectric material in contact-separation type AC-TENG. Fig. 1 Friction and triboelectric performance of TENG with different triboelectric materials. a Schematic diagram of DC-TENG. b The friction coefficient ( μ ) between Cu with various triboelectric layers and the polarization intensity of triboelectric layers at 10 kV cm −1 (error bars represent standard deviation). c The surface charge density of different triboelectric layers. d The electric field distribution curves in the gap between CCE and triboelectric layers’ surface (simulated by COMSOL software) and e corresponding summary of average electric field (the inset is the simulated result of PVC film). The sliding AC-TENG is utilized to confirm the σ SCD between Cu and different triboelectric layers, as shown in Supplementary Fig. 4 , and the detailed working mechanism is determined in Supplementary Note 3 . The σ SCD and short circuit current ( I sc ) is presented in Fig. 1c and Supplementary Fig. 5 (the slider of AC-TENG is 10 mm × 20 mm). The polyethylene terephthalate (PET), polypropylene (PP), polycarbonate (PC), and polyethylene (PE) show quite inferior σ SCD , which is <0.05 mC m −2 . A moderate σ SCD (0.05 mC m −2 < σ SCD <0.15 mC m −2 ) is obtained when using polyetherimide (PEI), polyphenylenesulphide (PPS), polyimide (PI), or polyetheretherketone (PEEK) as the triboelectric layer. The rest films, e.g., PTFE, fluorinated ethylene propylene (FEP), poly(vinylidene fluoride) (PVDF), and PVC, provide large σ SCD , exceeding 0.15 mC m −2 . Especially, the σ SCD of 0.30 mC m −2 can be obtained when using PVC as the triboelectric material. For the sliding AC-TENG with nylon (polyamide, PA) film as triboelectric material, the flow direction of output charges is opposite to the other films, which indicates that PA film loses electrons during the contact electrification process with Cu electrode. According to the working mechanism of DC-TENG, the electrostatic breakdown occurs in the gap between triboelectric film and charge collecting electrode (CCE) due to the electrostatic field generated by the charges on the surface of triboelectric layers (Supplementary Note 1 ). The electric field distribution in the gap is simulated by COMSOL software and provided to depict the intensity difference of various triboelectric materials (from a to b point in Supplementary Fig. 6 ), where the σ SCD is set according to the results in Fig. 1c . As shown in Fig. 1d, e and Supplementary Fig. 7 , the electric field strength in the gap shows the same trend as their σ SCD . The whole/partial simulated electric curves of PA, PET, PP, PE, and PC films are in the non-breakdown area (less than the air breakdown strength of 3 MV m −1 ). Thus, they only have a weak electrostatic field for air breakdown and generate a weak DC output for the DC-TENG or sometimes they even cannot cause air breakdown (e.g., PET and PP). It is worth noting that the electric field in the gap between PA and CCE is opposite to that of the other triboelectric films, resulting in the opposite DC signal in the external circuit. With the σ SCD increasing, the electric field in the gap becomes stronger, which is beneficial for the air breakdown process, thereby increasing the number of charges collected by the CCE and improving the DC output performance 13 . Moreover, on the basic of triboelectrification effect and the mechanism of DC-TENG, the triboelectric charges not only generate the electrostatic field in the gap to achieve the air breakdown but also generate an electric field in the polymer near the surface (Supplementary Fig. 6 ), which might induce the internal polarization effect near the surface along the electric field and constrain part of surface charges, making them hard to participate in the air breakdown process (Supplementary Fig. 8 ). The polarization intensity ( P ) vs. electric field loops of different triboelectric materials except for the PS, Nitrile, and PDMS films were carried out because of their overlarge friction coefficient, as shown in Fig. 1b and Supplementary Figs. 9 , 10 . Compared with other polymer films, PVDF and PEI exhibit a stronger polarization effect at the same electric field, indicating that they show a stronger polarization effect under the same electric field. The large polarization means more dipolar under the electric field (Supplementary Fig. 8 ), meaning that more charges will be bound and the air breakdown process might be restrained in the PVDF and PEI film. Primary material selection rules of DC-TENG According to the working mechanism of DC-TENG, based on the intrinsic parameters of μ , P , and σ SCD , as shown in Fig. 2 , we can provide a model to formulate the primary selection rules, which can determine whether a material is suitable for use as the triboelectric material in DC-TENG, and also determine the possible DC output performance. As shown in Fig. 2 , the primary selection rules can be divided into three steps. As shown in Fig. 2 , first, the DC-TENG is a kind of sliding-mode TENG, which is based on the sliding friction process and triboelectrification between the electrode and film. Thus, an appropriate friction coefficient of the triboelectric film is needed to reduce the wear and heat during the friction process between the electrode and triboelectric film. Therefore, the triboelectric layer with overlarge μ should be excluded, and the threshold value μ th is set as (~0.40), which is just a preliminary estimate for this primary rule based on the various triboelectric materials in this work. The μ th in this work is the middle value between the average μ of Cu and PI (~0.38, Supplementary Fig. 2 ) and the average μ of Cu and PS (~0.42, Supplementary Fig. 2 ). Second, on the basis of the mechanism of DC-TENG, the strong polarization of the triboelectric layer has an adverse effect on the air breakdown process and decreases the DC output of DC-TENG. Thus, the triboelectric layer with overlarge P should be ruled out. Here, except for the PVDF and PEI, the polarization intensity of the rest triboelectric materials is always higher than 0.05 μC m −2 (for PE) but lower than 0.5 μC m −2 (for PI) at 10 kV m −1 , as shown in Fig. 1b and Supplementary Figs. 9 , 10 . Therefore, the upper limit value of P ( P th ) is set as a preliminary estimate for this primary rule in this work, which is the middle value (~0.6 μC m −2 at 10 kV m −1 ) between 0.5 μC m −2 (polarization intensity of PI at 10 kV m −1 ) and 0.7 μC m −2 (polarization intensity of PEI at 10 kV m −1 ). It should be noted that the thresholds ( μ th and P th ) are preliminary estimates based on the triboelectric materials used in this experiment and will be further refined as more materials are studied in future works. Third, the possible DC output of the remaining triboelectric materials can be estimated based on their respective σ SCD value, and only the triboelectric materials with high σ SCD can form a high DC output owing to the high electric field in the gap (Fig. 1d, e ). These primary selection rules will provide a guideline to select appropriate triboelectric materials, and reduce the trial-and-error cost for DC-TENG’s research. Fig. 2 Primary selection rules of triboelectric materials for DC-TENG. The input parameters μ is friction coefficient, P is polarization, and σ SCD is surface charge density. Validation of material selection rules of DC-TENG The DC output performance of these 13 kinds of triboelectric materials is determined by the microstructure-designed DC-TENG device (Supplementary Fig. 11a ), whose size is similar to that of an AC-TENG device: 10 mm × 20 mm and DC unit = 20. The detailed structure is shown in Supplementary Fig. 11b . The interlaced FEs and CCEs are arranged on the acrylic substrate orderly, and a tiny air breakdown gap (30–40 μm) is formed between CCE and triboelectric layer (Supplementary Fig. 11c ). The friction coefficients with a micro-structured copper electrode of various triboelectric materials are shown in Supplementary Fig. 12 . The friction coefficients with a micro-structured copper electrode show a similar tendency with the friction coefficients with flat copper electrodes for the different triboelectric materials (Fig. 1b ). The working mechanism is shown in Supplementary Fig. 13 and Note 4 . As shown in Fig. 3a, b and Supplementary Fig. 14 , the PET and PP do not have the DC output owing to their low σ SCD value (Fig. 1d ), and the PE, PEI, and PC present low DC effective charge density ( σ DC ) of <0.5 mC m −2 ( I sc <0.15 μA), whereas the PI, PEEK, PPS and PVDF show moderate σ DC in the range from 0.5 to 2.0 mC m −2 and corresponding I sc of ~0.5 μA. The PVC, FEP, and PTFE provide high σ DC and I sc , which is >2.0 mC m −2 and 1.5 μA, respectively. It should be noticed that, when the PA film is utilized as triboelectric material, the DC output is also opposite to the other friction films (Fig. 3b ), because its surface charges are positive when in contact with Cu electrode. The working mechanism is shown in Supplementary Fig. 15 and Supplementary Note 5 . Fig. 3 DC output performance and corresponding relationship with surface charge density. a Effective charge density and b short circuit current of microstructure-designed DC-TENG (DC unit: 20) with different triboelectric layers. c The relationship between surface charge density and effective charge density of DC-TENG of different triboelectric layers. The relationship between σ SCD and σ DC of different triboelectric films is shown in Fig. 3c . It can be clearly seen that the PVDF and PEI films show relatively high surface charge density but low DC output, and moderate surface charge density but low DC output, respectively, indicating that they are not suitable as the triboelectric layers for DC-TENG due to their relatively strong polarization effect. Except for them, the DC output of the other triboelectric materials is increased with the increase of σ SCD . Meanwhile, the relationships between σ SCD and σ DC can be divided into four types: a. the films show strong surface charge density with strong DC output (e.g., PVC, FEP, PTFE); b. the films present moderate surface charge density and moderate DC output (e.g., PI, PEEK, PPS); c. the films show low surface charge density and low DC output (e.g., PC, PE, PP, PET); d. the film shows opposite DC output (PA). On one hand, the larger σ SCD of triboelectric films will lead to a stronger electric field in the gap which is beneficial to the air breakdown effect. On the other hand, when the triboelectric film has a larger σ SCD , more surface charges on the film will participate in the air breakdown process and can be collected by DC-TENG. In other words, the utilization rate of contact electrification charges in single DC unit ( η , the ratio of collected charges by one DC unit and the σ SCD ) will increase. For example, the PVC, FEP, and PTFE films show high η of >60%, but the PE just has a low η , <10% (Supplementary Fig. 16 ). Therefore, this relationship between σ SCD and σ DC for different triboelectric films confirms the effectiveness of the selection rules in Fig. 2 . Comprehensive material selection model of DC-TENG In practical applications, one or more key indexes should be decided to satisfy the application requirements of DC-TENG usage. For example, for energy harvester, the triboelectric materials should possess a high DC output and high η . For sensors, the stability of triboelectric materials with an appropriate voltage/current output is necessary to maintain repeatability and accuracy. Therefore, in the radar chart shown in Fig. 4 , a comprehensive model of the performance selection of triboelectric materials for DC-TENG is given. Five parameters are shown in Supplementary Table 1 , σ SCD *, σ DC *, η *, 1/ μ *, and ρ *, is the normalization of σ SCD , σ DC , η , the reciprocal of the friction coefficient (1/ μ ), and the stability ( ρ , shown in Supplementary Fig. 17 and Note 6 ), respectively. They are utilized as the performance evaluation indexes of triboelectric materials in DC-TENG (taking PVC, FEP, PTFE, PEEK, PVDF, and PI as the representative materials), as shown in Fig. 4 . The PVC, FEP, and PTFE films present high σ SCD values as well as high DC output. At the same time, they also provide a high utilization rate of contact electrification charges in the air breakdown process, with η reaching ~60% (Supplementary Table 1 ). However, FEP film shows inferior stability for DC-TENG because its surface layer is easily stripped off during the stability test (Supplementary Fig. 18 ), which will jam the gap and make the air breakdown hard to occur (Supplementary Fig. 19 ). Both of the PVC and PTFE films have outstanding comprehensive properties, and the PVC shows a larger effective charge density and PTFE shows a better friction property. Compared with the PVC, FEP, and PTFE, the PEEK and PI film show relatively moderate comprehensive properties. However, the PVDF film has inferior properties with quite low η and poor stability, so it is not suitable for use as a triboelectric layer in DC-TENG. Fig. 4 Comprehensive selection rules of triboelectric materials for DC-TENG. The radar chart of normalized indexes of PVC, FEP, PTFE, PEEK, PI, and PVDF, where the σ SCD *, σ DC *, η *, 1/ μ *, and ρ * is the normalization of σ SCD : surface charge density, σ DC : DC charge density, η : (σ DC /20)/ σ SCD , 1/μ : the reciprocal of friction coefficient, ρ : stability, respectively. Application of selected PVC film for DC-TENG It can be seen that the PVC film shows good comprehensive properties as triboelectric material for DC-TENG owing to its largest stacked normalized indexes, as shown in Fig. 5a . Utilizing PVC as the triboelectric layer, DC-TENG can provide an increased effective charge density and current as the number of DC units increases: 0.82 mC m −2 and 0.56 μA for DC-TENG with 5 DC units, 8.80 mC m −2 and 5.88 μA for the DC-TENG with 50 DC units (sliding distance: 5 cm, sliding width: 1 cm, Fig. 5b and Supplementary Fig. 20 ). This highly effective charge density (8.80 mC m −2 ) is >20 times that of the previous DC-TENG 12 , and breaks the existing records of output charge density for various-type TENG (Fig. 5c ) 15 , 24 – 28 . Fig. 5 Application of selected PVC film as friction for DC-TENG. a The stacked bar plot of normalized properties of PVC, PTFE, FEP, PEEK, PI, and PVDF. b The effective charge density of microstructure-designed DC-TENG with different DC units. c Comparison of the charge density of TENGs 15 , 24 – 28 . d Photograph of rotary DC-TENG ((i) stator (scale bar: 1.0 cm), (ii) partial enlarged detail of stator (scale bar: 1.5 mm), (iii) SEM image of stator (scale bar: 250 μm) and (iv) rotor (scale bar: 1.0 cm)). e Output current of rotary DC-TENG with PVC as triboelectric material under various rotating speeds. f Charging curves of capacitors (10, 47, 110 µF) charged by rotary DC-TENG with PVC as triboelectric material (600 r min −1 ). g Photograph of driving commercial LED bulbs and thermo-hygrometer directly by rotary DC-TENG with PVC as triboelectric material (600 r min −1 , scale bar: 2.0 cm). h Monitored voltage of the 660 µF commercial capacitors with driving thermo-hygrometer simultaneously by rotating DC-TENG with PVC as triboelectric material (600 r min −1 ). The microstructure-designed DC-TENG was fabricated into rotary DC-TENG to realize the continuous DC output, whose device structure and microscopic photo are shown in Fig. 5d(i–iii) . As the same structure with the sliding DC-TENG in Supplementary Fig. 13 , the interlaced FEs and CCEs are orderly arranged on the surface of the acrylic substrate (Fig. 5d(i, ii) ), and the distance between FE and adjacent CCE is ~750 μm (Fig. 5d(iii) ). The PVC film is attached to the sponge, which is utilized as the rotor (Fig. 5d(iv) ). With the rotating speed gradually increasing, the continuous DC output of rotary DC-TENG also increases from 3.3 μA to 75 μA (Fig. 5e ). Utilizing the 22 μF, 47 μF, and 110 μF capacitors as the energy storage device, it just takes 0.8 s, 2.2 s, and 4.5 s to charge the capacitors to 3 V under the rotation of 600 r min −1 , respectively, as shown in Fig. 5f . The rotary DC-TENG with PVC film as the triboelectric layer, the high DC output can directly drive the electronic devices, e.g., a commercial thermo-hygrometer or four commercial LED bulbs, as shown in Fig. 5g , Supplementary Movie 1 , 2 . The self-powered system consisting of the rotary DC-TENG (the energy supply unit), capacitor (660 μF, the energy storage unit), and commercial thermo-hygrometer (the energy consumption unit) is built, and the corresponding circuit is shown in Supplementary Fig. 21 . The monitored voltage of the capacitor at different rotary DC-TENG working conditions is shown in Fig. 5h . When the energy supply by DC-TENG is shut down, the charges in the capacitor are gradually consumed by the thermo-hygrometer, resulting in the decrease of the voltage of the capacitor. With the extra energy supplied by the DC-TENG, the voltage starts to increase because the supplied energy is larger than that consumed by thermo-hygrometer. However, as the energy supply is shut down again, the voltage of the capacitor falls again due to the consumption of thermo-hygrometer (Fig. 5h ).\n\nDiscussion In summary, we propose a set of selection rules of triboelectric materials in direct current (DC-TENG) based on surface charge density, friction coefficient, polarization, the utilization rate of charges, and stability to screen whether a triboelectric material is suitable to be used in DC-TENG. Taking advantage of this model, the PVC film is found to be utilized as the triboelectric layer for DC-TENG, whose effective charge density can reach 8.80 mC m −2 , which is >20 times that of the previous DC-TENG, and breaks the existing record of various-type TENGs. Furthermore, the proposed comprehensive model can be used to provide a guideline for the triboelectric material selection of DC-TENG to satisfy its practical application requirements. With the widespread applications of TENG, the database of triboelectric materials used in TENG becomes more abundant. The selection rule is not limited to the matching of copper and organic materials, but can also be used in other suitable combinations of triboelectric materials, such as other metals and organic materials or organic materials and organic materials, to obtain higher DC output. Moreover, on the basis of the selection model, extensive data search and prediction can be carried out with the help of machine learning or big data analysis, so as to realize the intelligent design of triboelectric materials for high-performance DC-TENG."
} | 6,484 |
22131292 | null | s2 | 3,439 | {
"abstract": "Nucleic acid nanotechnology exploits the programmable molecular recognition properties of natural and synthetic nucleic acids to assemble structures with nanometer-scale precision. In 2006, DNA origami transformed the field by providing a versatile platform for self-assembly of arbitrary shapes from one long DNA strand held in place by hundreds of short, site-specific (spatially addressable) DNA 'staples'. This revolutionary approach has led to the creation of a multitude of two-dimensional and three-dimensional scaffolds that form the basis for functional nanodevices. Not limited to nucleic acids, these nanodevices can incorporate other structural and functional materials, such as proteins and nanoparticles, making them broadly useful for current and future applications in emerging fields such as nanomedicine, nanoelectronics, and alternative energy."
} | 215 |
25202723 | PMC4150482 | pmc | 3,440 | {
"abstract": "In this paper, we present an implementation scheme of memristor-based multilayer feedforward small-world neural network (MFSNN) inspirited by the lack of the hardware realization of the MFSNN on account of the need of a large number of electronic neurons and synapses. More specially, a mathematical closed-form charge-governed memristor model is presented with derivation procedures and the corresponding Simulink model is presented, which is an essential block for realizing the memristive synapse and the activation function in electronic neurons. Furthermore, we investigate a more intelligent memristive PID controller by incorporating the proposed MFSNN into intelligent PID control based on the advantages of the memristive MFSNN on computation speed and accuracy. Finally, numerical simulations have demonstrated the effectiveness of the proposed scheme.",
"conclusion": "6. Conclusions A mathematical closed-form charge-governed memristor model is recalled firstly and the corresponding Simulink model is presented. Using the change rule of memconductance, a memristive realization scheme for synaptic weight is proposed. Moreover, the activation functions in electric neurons are also implemented based on the single-input and double-output package of the memristor. Combining the proposed memristive synapse and activation functions, a memristor-based MFSNN is addressed. It exhibits advantages in computation speed and accuracy over the traditional multilayer neural networks by considering the small-world effect. Meanwhile, it has potential of hardware realization of the neural network because of the nanoscale size of the memristive synapse. These superior properties can further improve the application of the neural networks, such as in the intelligent controller design. Motivated by this, we apply the memristor-based MFSNN to classical PID control, and the proposed memristive PID controller may possess the following superiorities. (i) Its nanoscale physical implementation could promote the development of the microcontroller. (ii) Because of the participation of the memristive neural network, the proposed PID controller can realize the parameters self-adjustment. (iii) The control speed and accuracy are improved. Eventually, extensive numerical simulations justify the effectiveness and efficiency of the memristive PID controller over the regular neural network PID controller. This work may provide a theoretical reference to physically realize the small-world neural networks and further promote the development of modern intelligent control technology.",
"introduction": "1. Introduction In 1971, Professor Chua theoretically formulated and defined the memristor and described that the memristance (short for resistor of a memristor) is characterized by the relationship between the electrical charge q and flux φ passing through a device [ 1 ]. However, it was only after the first physical realization of the memristor in nanoscale at Hewlett-Packard (HP) Lab in 2008 that it immediately garnered extensive interests among numerous researchers [ 2 – 4 ]. The reported experiments confirmed that the memristor possesses switching characteristic, memory capacity, and continuous input and output property. Due to these unique properties, memristors are being explored for many potential applications in the areas of nonvolatile memory [ 5 , 6 ], very-large-scale integrated (VLSI) circuit [ 7 ], artificial neural networks [ 8 – 10 ], digital image processing [ 11 – 13 ], and signal processing and pattern recognition [ 14 ]. At present, a considerable number of models of different complexity have been proposed in the literatures, such as Pickett's model [ 15 ], spintronic memristor model [ 16 ], nonlinear ionic drift model [ 17 ], boundary condition-based model [ 18 ], and threshold adaptive memristor model [ 19 ]. These published models exhibit desired nonlinearity of nanoscale structures. This paper still applies the TiO 2 memristor model on account of its simplified expressions and the same ideal physical behaviors. Brain neural network emerges from the interactions of dozens, perhaps hundreds, of brain regions, each containing millions of neurons [ 20 ]. They are highly evolved nervous systems capable of high-speed information processing, real-time integration of information across segregated sensory channels, and brain regions [ 20 , 21 ]. In order to obtain the similar intelligence of human brain, artificial neural network is designed to imitate the human brain not merely on architecture but also on work patterns. The connection structure of artificial neural networks is generally divided into feedforward, feedback, single-layer, multilayer, and so forth. Most of these connection architectures are approximately regular. However, the bioneurological researches show that brain neural network has random features to a certain degree and exhibits “small-world” effectiveness, that is, high levels of clustering and short average path length [ 22 ]. Therefore, it becomes a hot issue to design bionics neural network with randomness in architecture based on the background of neurobiology. Notably, Watts and Strogatz revealed a significant effect that is in common among complex networks. They pointed out that the real architecture of network is nearly a middle model between regular connection and random connection and defined it as small-world network (WS model) in 1998 [ 23 ]. Over the past several years, a large number of investigations on complex networks have provided new insight into biological neural networks. Bassett concluded that human brain functional networks have small-world network topology derived from a series of magneto encephalography experiments [ 22 ]. Douw et al. found that the cognition is related to the resting-state small-world network topology [ 24 ]. In literature [ 25 ], the authors applied small-world properties into prefrontal cortex that correlate with predictors of psychopathology risk, which holds promise as a potential neurodiagnostic for young children. Taylor has studied the protein structures and binding based on small-world network strategies and has made great progress [ 26 ]. Simard built up a small-world neural network through rewiring the regular connections and found that the small-world neural network has faster learning speed and smaller error than that of the regular network and random network with the same size [ 27 ]. In this paper, we incorporate the memristor into the multilayer feedforward small-world neural network to build up a new type of memristive neural network that is easy of VLSI implementation and closer to biological networks. Furthermore, based on the proposed memristive neural network, a novel memristive intelligent PID controller is put forward. The nanoscale memristor is beneficial for easily adjusting the PID control parameters and the hardware realization of modern intelligent microcontrol system. This paper is organized as follows. In Section 2 , we derive the mathematical model of a nonlinear memristor which takes into account the nonlinear dopant drift effect nearby the terminals and the boundary conditions and give its Simulink model correspondingly. Following that, the concepts and design algorithm of the memristive small-world neural network are described in detail in Section 3 . Section 4 designs a memristive PID controller by combining the proposed neural network with the standard PID control theory. In order to guarantee the feasibility and effectiveness of the proposed scheme, the computer simulations are performed in Section 5 . Finally, we give the conclusions in Section 6 ."
} | 1,908 |
34056333 | PMC8154022 | pmc | 3,442 | {
"abstract": "Aging infrastructure,\nincreasing environmental regulations, and\nreceiving water environment issues stem the need for advanced wastewater\ntreatment processes across the world. Advanced wastewater treatment\nsystems treat wastewater beyond organic carbon removal and aim to\nremove nutrients and recover valuable products. While the removal\nof major nutrients (carbon, nitrogen, and phosphorus) is essential\nfor environmental protection, this can only be achieved through energy-,\nchemical-, and cost-intensive processes in the industry today, which\nis an unsustainable trend, considering the global population growth\nand rapid urbanization. Two major routes for developing more sustainable\nand circular-economy-based wastewater treatment systems would be to\n(a) innovate and integrate energy- and resource-efficient anaerobic\nwastewater treatment systems and (b) enhance carbon capture to be\ndiverted to energy recovery schemes. This Mini-Review provides a critical\nevaluation and perspective of two potential process routes that enable\nthis transition. These process routes include a bioelectrochemical\nenergy recovery scheme and codigestion of organic sludge for biogas\ngeneration in anaerobic digesters. From the analysis, it is imperative\nthat integrating both concepts may even result in more energy- and\nresource-efficient wastewater treatment systems.",
"conclusion": "Concluding Remarks The need for developing resource-efficient wastewater treatment\nsystems can never be overstressed considering the critical influence\nand role of treated effluents on the receiving environment. An evaluation\nand comparison of five different alternatives with a conventional\ntreatment scheme, all including an anaerobic digestion step for biogas\ngeneration, shows that wastewater treatment plants can achieve energy\nself-sufficiency by enhancing carbon capture, by accepting external\norganic feedstock for energy recovery, and by implementing anaerobic\nbioelectrochemical treatment for electricity generation. All these\nsteps taken together can help wastewater treatment plants march toward\ncircular economy and energy sustainability, pending significant scientific\nand process related challenges to be overcome in near future.",
"introduction": "Introduction To protect the quality of receiving water\nbodies and the environment,\ncarbonaceous and nutrient (nitrogen and phosphorus – N and\nP) compounds must be efficiently removed from wastewater sources.\nMost wastewater treatment plants remove organic matter only through\nenergy-inefficient configurations. Conventional wastewater treatment\nsystems for nitrogen removal rely on excess chemicals and energy to\ncreate aerobic conditions for biological nitrification and use organic\ncarbon to remove nitrate by biological denitrification. Instead, the\ndesign of sustainable wastewater treatment systems must focus on environmental\nprotection and minimize energy and resource consumption. In this context,\nnovel wastewater treatment systems that produce net energy while removing\nnutrients are actively sought. Anaerobic wastewater treatment is a\nproven and beneficial route from energy and environmental perspectives.\nSimilarly, an alternative approach for baseline nitrification-denitrification\nprocess is to use anaerobic ammonium-oxidizing (anammox) bacteria,\nwhich reduces energy and chemical costs. Under anaerobic conditions,\nbioelectrochemical systems provide efficient wastewater treatment\nwhile generating clean electricity directly from organic substrates.\nIntegrating these technologies is essential to further enhance the\npotential for energy recovery and for developing sustainable wastewater\ntreatment systems.",
"discussion": "Discussion There are many possible solutions for accomplishing energy self-sufficiency\nor energy-positive status in wastewater treatment plants. Here, we\ncompared six different process configurations and identified two key\nprocesses that may enable this transition more efficiently. While\nthe proposed scenarios (S5 and S6) are quite attractive from circular\neconomy and energy sustainability perspectives, many barriers and\nchallenges need addressing for their practical feasibility and wide\napplication. Regarding S5, codigestion method has shown promising\npotential\nbecause the addition of FOG or other organic wastes for codigestion\nhas a positive effect on the digestion process with higher methane\nyields and stable operations. Biogas production due to FOG codigestion\ncould also increase from 15 to 30%, which is a significant contribution\nto electricity and heat recovery. 16 C/N\nratio, organic loading rate, pH, temperature, HRT, and pretreatment\nof organic substrates are all critical factors for the success of\nthe codigestion process. 55 , 56 While there are several\nbenefits of integrating codigestion of mixed waste for energy production,\nthere are a few challenges that are worth mentioning. One of the biggest\nchallenges is the generation of sludge caused by digesting additional\nwaste such as FOG with inconsistent characteristics and their land\napplication. Significantly higher nutrient concentrations present\nin supplemental feedstock also pose permit-specific issues. Material\nhandling infrastructure may face corrosion because of acidic streams.\nFurther, AD effluents containing nutrient residues encourage proliferation\nof unwanted microorganisms during the CAS process. Overall, the economic,\nenergy, and environmental benefits outweight the aforementioned issues.\nFOG addition increases biogas production by 2-fold. Upfront or upstream\nremoval of FOG prior to the acitivated sludge unit has a significant\nimpact on the efficiency conservation and efficiency of the downstream\nprocess. Moreover, codigestion of FOG in anaerobic digesters would\nbe more cost-effective and environmentally friendly. 57 Food and agricultural wastes also can be valorized using\nanaerobic digester technology. 58 An estimated\n1.6 billion tons of food waste is available worldwide, 56 which can be converted to valuable energy products\n(biofuels such as biodiesel, bioethanol, and biooil or biocrude) through\nvarious thermochemical and biochemical processes. However, AD is a\npreferred and most feasible method for waste valorization especially\nwhen combined with CHP units. 59 Regarding\nS6, the long-term performance of the anammox process\nboth in laboratory and pilot/field scale studies has shown evidence\nof slow nitrogen removal rates. The reasons behind the poor performance\nare still not well understood, although it is sometimes attributed\nto the accumulation of dissolved oxygen content over time. The long-term\nperformance and process feasibility for ABCs are still unknown. There\nare many process parameters that require deeper investigation and\noptimization. Understanding the evolution and maintenance of anammox\nbiofilms or granules under bioelectrochemical system configuration\nis important for improving the stability of this process. In the ABCs’\nprocess, the aerobic ammonium-oxidizing bacteria (AOB) play a very\nimportant role in the reduction of ammonium. The main role of AOB\nin the ABCs’ process is to convert approximately half of the\nammonium to nitrite under partial oxygen condition, and in turn, anammox\nbacteria convert nitrite, together with the remaining ammonium, into\nnitrogen gas. Therefore, the development of the strategy for maintaining\nthe balance between AOB and AnAOB (anaerobic ammonia-oxidizing bacteria)\nin the biofilm of single-stage ABCs and simultaneously inhibiting\nthe nitrite-oxidizing bacteria (NOB) would lead the ABCs toward higher\nnitrogen removal efficiency. Development and implementation of three-dimensional\nconducting electrodes in ABCs can assist anammox granules/biofilm\nin easily attaching in the conducting medium, reducing electrical\nresistance between anammox bacteria and electrode increasing energy\ngeneration. The partial aeration step used in the ABC process\nis still an energy-consuming\nprocess. A complete replacement of mechanical aeration system with\nbiological oxygen-producing microorganisms would make ABCs more energy-positive.\nIn this case, it would be interesting to consider the integration\nof anammox and microalgae polishing configurations. The oxygen produced\nby microalgae during the day time would provide enough dissolved oxygen\nfor the partial conversion of ammonium to nitrite (partial nitritation\nprocess), and during the night time, microalgae will consume the dissolved\noxygen to create an anoxic condition, which will be more favorable\nfor anammox bacteria to remove the remaining portion of ammonium and\ncombine with nitrite to form nitrogen gas."
} | 2,140 |
26858288 | PMC4790474 | pmc | 3,443 | {
"abstract": "Lignin poses a major challenge in the processing of plant biomass for agro-industrial applications. For bioengineering purposes, there is a pressing interest in identifying and characterizing the enzymes responsible for the biosynthesis of lignin. Hydroxycinnamoyl-CoA:shikimate hydroxycinnamoyl transferase (HCT; EC 2.3.1.133) is a key metabolic entry point for the synthesis of the most important lignin monomers: coniferyl and sinapyl alcohols. In this study, we investigated the substrate promiscuity of HCT from a bryophyte ( Physcomitrella ) and from five representatives of vascular plants (Arabidopsis, poplar, switchgrass, pine and Selaginella ) using a yeast expression system. We demonstrate for these HCTs a conserved capacity to acylate with p -coumaroyl-CoA several phenolic compounds in addition to the canonical acceptor shikimate normally used during lignin biosynthesis. Using either recombinant HCT from switchgrass (PvHCT2a) or an Arabidopsis stem protein extract, we show evidence of the inhibitory effect of these phenolics on the synthesis of p -coumaroyl shikimate in vitro, which presumably occurs via a mechanism of competitive inhibition. A structural study of PvHCT2a confirmed the binding of a non-canonical acceptor in a similar manner to shikimate in the active site of the enzyme. Finally, we exploited in Arabidopsis the substrate flexibility of HCT to reduce lignin content and improve biomass saccharification by engineering transgenic lines that overproduce one of the HCT non-canonical acceptors. Our results demonstrate conservation of HCT substrate promiscuity and provide support for a new strategy for lignin reduction in the effort to improve the quality of plant biomass for forage and cellulosic biofuels.",
"introduction": "Introduction Plant biomass is routinely utilized in diverse agro-industrial sectors such as pulp and paper, forage, bio-manufacturing and bioenergy. Among several sustainable energy strategies, the conversion of plant cell walls into fermentable sugars offers the possibility for microbial production of bio-based products including materials (e.g. biopolymers), commodity chemicals, fine and specialty chemicals, and liquid fuels ( Vickers et al. 2012 , George et al. 2015 ). Lignin is a hydrophobic polymer that permeates plant cell walls and impedes the biochemical processes used for plant biomass conversion into fermentable sugars ( Zeng et al. 2014 ). Lignin from ferns and gymnosperms contains p -hydroxyphenyl (H) and guaiacyl (G) units derived from p -coumaroyl alcohol and coniferyl alcohol, respectively; and lignin from certain lycophytes and angiosperms additionally contains syringyl (S) units ( Boerjan et al. 2003 , Weng et al. 2010 ). These lignin monomers (or monolignols) are synthesized from phenylalanine in the cytosol and exported to the cell wall for oxidative polymerization ( Boerjan et al. 2003 ). In the lignin pathway, the metabolic entry point responsible for the production of both coniferyl and sinapyl alcohols involves hydroxycinnamoyl-CoA:shikimate hydroxycinnamoyl transferase (HCT), which catalyzes the coupling of p -coumaroyl-CoA with shikimate to produce p -coumaroyl shikimate ( Fig. 1 ) ( Hoffman et al. 2003 , Hoffman et al. 2004 ). Following this step, the hydroxycinnamoyl moiety of p -coumaroyl shikimate is hydroxylated by p -coumaroyl shikimate 3′-hydroxylase (C3′H) to produce caffeoyl shikimate which is then transesterified by HCT to release caffeoyl-CoA, or, as recently described in Arabidopsis, cleaved by caffeoyl shikimate esterase (CSE) to release caffeate ( Fig. 1 ) ( Franke et al. 2002 , Vanholme et al. 2013a ). Reduction of HCT activity limits the synthesis of coniferyl and sinapyl alcohols and causes a backlog of upstream metabolites, which leads to a redirection of part of the metabolic flux of p -coumaroyl-CoA into p -coumaryl alcohol. Therefore, the lignin from plants exhibiting strong HCT down-regulation is enriched in H units and its content is heavily reduced due to the poor capacity of p -coumaroyl alcohol to incorporate the polymer ( Hoffman et al. 2004 , Vanholme et al. 2013b ). Considering its central role in lignin biosynthesis, HCT has been a target in genetic engineering strategies for improving the commercial utility of plant biomass. For example, down-regulation of HCT expression reduces lignin content and improves forage digestibility and saccharification efficiency in alfalfa ( Chen and Dixon 2007 , Shadle et al. 2007 ).\n Fig. 1 Simplified representation of the lignin biosynthetic pathway. Abbreviations: HCT, hydroxycinnamoyl-CoA:shikimate hydroxycinnamoyl transferase; C3′H, p -coumaroyl shikimate 3′-hydroxylase; CSE, caffeoyl shikimate esterase. Dashed arrows represent the CSE-mediated esterification of caffeoyl shikimate as described in Arabidopsis ( Vanholme et al. 2013a ). HCT belongs to the BAHD family of acyl-CoA-dependent transferases, which includes several enzymes that use hydroxycinnamoyl-CoAs as a donor for the transfer reaction ( D’Auria 2006 ). These transferases are capable of acylating a wide variety of acceptors, and some of them exhibit broad substrate flexibility ( Landmann et al. 2011 , Molina and Kosma 2014 ). For example, it is well known that some HCTs can use quinate as a substrate in addition to shikimate ( Hoffman et al. 2003 , Lallemand et al. 2012 , Walker et al. 2013 , Escamilla-Treviño et al. 2014 ). Moreover, HCT from globe artichoke (CcHCT) was shown to accept 3-hydroxyanthranilate as a substrate ( Moglia et al. 2010 ), and similar substrate promiscuity was described for HCT from Coleus blumei , which can also use 3-amino- and 3-hydroxybenzoates as acceptor substrates, forming both p -coumarate ester and amide conjugates ( Sander and Petersen 2011 ). Although HCT is a well-conserved enzyme among land plants ( Xu et al. 2009 , Tohge et al. 2013 ), characterization of its substrate flexibility remains poorly described. Furthermore, until now, it had yet to be tested whether the substrate promiscuity of HCT could be exploited for lignin engineering and improving plant biomass. In this study, using a yeast expression system, we investigated the substrate promiscuity of HCT from a bryophyte ( Physcomitrella patens ), a lycophyte ( Selaginella moellendorffii ), a coniferous gymnosperm ( Pinus radiata ), a monocot angiosperm ( Panicum virgatum ) and two dicot angiosperms ( Populus trichocarpa and Arabidopsis thaliana ) toward a series of benzene and benzoate derivatives. Recombinant yeast strains that individually co-express these phylogenetically related HCTs with 4-coumarate-CoA ligase were capable of producing p -coumaroyl conjugates when fed with p -coumarate and shikimate, 3-hydroxybenzoate, 2,5-dihydroxybenzoate (gentisate), 2,3-dihydroxybenzoate, 3,4-dihydroxybenzoate (protocatechuate), 3-aminobenzoate, 3-hydroxyanthranilate, 5-hydroxyanthranilate, ortho- hydroxyphenol (catechol) and para- hydroxyphenol (hydroquinone). Purified recombinant HCT from switchgrass showed activity towards these new ‘non-canonical’ substrates, which confirmed their capacity to bind the active site of the enzyme. Moreover, we demonstrate that in vitro formation of p -coumaroyl shikimate by either a switchgrass recombinant HCT (PvHCT2a) or an Arabidopsis stem protein extract is partially inhibited in the presence of the non-canonical acceptors, presumably via a mechanism of competitive inhibition. Finally, we extended this characterization to in vivo inhibition assays by engineering Arabidopsis lines to overproduce one of these HCT-competitive inhibitors (protocatechuate) in lignifying tissues, and showed an approximately 30% reduction of lignin content, providing a significant improvement of biomass saccharification efficiency. Our results demonstrate the potential of inhibiting HCT in various bioenergy crops by overproducing unusual HCT acceptors in planta. The substrate promiscuity of HCT was observed in all the plant taxa we tested, and it appears to have been conserved since plants first appeared on land about 450 Ma ago. This novel approach could be particularly valuable to reduce lignin in crops for which genomic information is still unavailable.",
"discussion": "Discussion The substrate promiscuity of HCT was demonstrated by its capacity to form p -coumaroyl conjugates using various acceptors other than shikimate. This substrate flexibility appears conserved between HCTs from different representatives of land plants, occurring at least in one bryophyte and one lycophyte, as well as in euphyllophytes including one gymnosperm, one monocot angiosperm and two dicot angiosperms. Although P. patens contains orthologs of all the genes required for the synthesis of p -coumaryl and coniferyl alcohols ( Xu et al. 2009 ), the presence of lignified tissues has never been demonstrated in this plant. We show here that P. patens possesses a bona fide HCT, but its physiological role remains unclear. Although PpHCT1 falls into the clade of known HCTs in the phylogenetic tree of hydroxycinnamoyl-CoA-dependent BAHD transferases ( Supplementary Fig. S1) , it may instead be involved in the synthesis of alternative phenylpropanoid conjugates such as chlorogenic acid (3- O -caffeoylquinic acid) which has been identified in P. patens ( Erxleben et al. 2012 ). Moreover, this is, to our knowledge, the first demonstration of HCT activity for an enzyme from Selaginella. However, its involvement in lignification remains questionable due to the presence in this plant of a dual meta -hydroxylase (SmF5H), which can by-pass the enzymatic steps catalyzed by C3′H and HCT ( Weng et al. 2010 ). More generally, BAHD acyltransferases from Selaginella could have a role in the synthesis of phenolic esters other than p -coumaroyl shikimate ( Weng and Noel 2013 ). Although the formation of p -coumaroyl shikimate occurs via an ester linkage between p -coumaroyl-CoA and shikimate, these HCTs can also catalyze the formation of p -coumaroyl conjugates via an amide bond, as exemplified with the use of 3-aminobenzoate as acceptor. Both ester and amide linkages are possible for the formation of p -coumaroyl 3-hydroxyanthranilate and p -coumaroyl 5-hydroxyanthranilate, but only one product was observed in our assays when these acceptors were tested. For these two cases, an amide bond can be anticipated, as previously determined for the p -coumarate conjugate formed by CcHCT using 3-hydroxyanthranilate ( Moglia et al. 2010 ). In this work, the apparent K m value observed for PvHCT2a towards shikimate (∼700 µM) is comparable with the values observed for tobacco HCT (750 µM), which has been genetically implicated in lignin biosynthesis ( Hoffmann et al. 2003 ), and for poplar HCT (895 µM; Kim et al. 2011 ), but higher than the values reported for HCT from coffee (75 µM; Lallemand et al. 2012 ), sorghum (153 µM; Walker et al. 2013 ) and Coleus blumei (332 µM; Sander and Petersen 2011 ). The slightly different structure of the new HCT acceptors described in this study could influence their binding to the active site of the enzyme and explain the higher apparent K m values of HCT for these substrates compared with shikimate. The catalytic mechanism of HCT has been elegantly described previously by Walker et al. (2013) . In their mutagenesis experiments using shikimate as a substrate, the mutant T384A-SbHCT (equivalent to Thr382 in PvHCT2a) showed only 9% of the enzyme activity compared with wild-type SbHCT ( Walker et al. 2013 ). We therefore propose that the loss of the C5 hydroxyl interaction with Thr382 in the PvHCT2a– p -coumaroyl-CoA–protocatechuate complex affected the effective binding of protocatechuate, thus engendering a similar effect to the T384A mutation in SbHCT on the turnover rate of the substrate. This is most probably the reason why it was possible to capture crystallographically the substrate PvHCT2a– p -coumaryl-CoA–protocatechuate ternary complex, compared with the product state when shikimate was used in soaking experiments. Our goal was to use the substrate promiscuity of HCT to reduce lignin. In this attempt, we generated Arabidopsis lines that overproduce protocatechuate, a non-canonical acceptor and competitive inhibitor of HCT in vitro. The engineered plants accumulate approximatey 25-fold more protocatechuate than wild-type plants, and showed significant lignin reductions (31–37%) in biomass. Interestingly, transgenic lines expressing the pobA gene alone accumulated approximately 1.8-fold more protocatechuate than wild-type plants presumably from the conversion of the endogenous pool of p- hydroxybenzoate but showed no reduction of lignin content, suggesting that a threshold amount of protocatechuate is necessary to inhibit significantly HCT activity in vivo. Since lignin is reduced in the protocatechuate-overproducing plants, either the amount of p -coumaroyl protocatechuate produced by HCT is too low to sustain lignin biosynthesis, or p -coumaroyl protocatechuate is not metabolized further in the lignin biosynthetic pathway. In particular, it is not known whether C3′H can hydroxylate p -coumaroyl protocatechuate, and whether CSE would cleave the putative resulting p -caffeoyl protocatechuate. Neither p -coumaroyl protocatechuate nor putative p -caffeoyl protocatechuate were detected in the metabolite extracts from ubiC-pobA transgenic plants, possibly because of their conjugation to other metabolites such as sugars, as was recently shown for other lignin oligomers ( Dima et al. 2015 ). Our dual transgene strategy consists of a two-step conversion of chorismate into protocatechuate. Steady-state levels of phenylalanine and salicylate, two metabolites derived from chorismate, are not significantly changed in the ubiC , pobA and ubiC-pobA transgenics compared with the wild type, suggesting that chorismate is not limiting in these lines, as previously hypothesized in the case of tobacco plants that express ubiC and show normal growth characteristics ( Siebert et al. 1996 , Viitanen et al. 2004 ). Therefore, the reduction of lignin observed only in the ubiC-pobA lines (and not in the ubiC single lines) is unlikely to be the consequence of phenylalanine depletion, although our data cannot rule out a possible slowing down of the metabolic flux. In addition, the height reduction observed in the ubiC-pobA lines at the mature stage is probably a consequence of lignin reduction rather than drastic changes in levels of salicylic acid, a metabolite whose overaccumulation in some transgenic plants often results in dwarf phenotypes ( Rivas-San Vicente and Plasencia 2011 ). In contrast to previous studies of plants in which HCT activity was down-regulated, the lignin composition of our transgenic lines does not show any enrichment in H units, possibly due to a lower degree of inhibition of activity and the consumption of p -coumaroyl-CoA by HCT for the formation of both p -coumaroyl protocatechuate and p -coumaroyl shikimate. Nevertheless, the increase of the S:G ratio in our transgenics, which results from the decrease of G units in favor of an increase in S units, is a lignin feature predicted by metabolic flux analysis in the case of HCT down-regulation, and was previously observed in plants affected in HCT activity ( Chen and Dixon 2007 , Ziebell et al. 2010 , Vanholme et al, 2013b , Wang et al. 2014 , Eudes et al. 2015 ). In addition, analysis of HCT transcript levels in the ubiC-pobA transgenic lines showed no major difference compared with the wild type ( Supplementary Fig. S10 ), which is consistent with the hypothesis that protocatechuate affects the activity of HCT rather than its expression in the ubiC-pobA lines. Engineering strategies for the overproduction of HCT non-canonical acceptors other than protocatechuate could be developed as alternative approaches to reduce lignin. For example, plant lines could be engineered to express anthranilate 3-hydroxylase, 2,3-dihydroxybenzoate synthase/dehydrogenase, salicylate 5-hydroxylases, 3-aminobenzoate synthase or 3-hydroxybenzoate synthase for the accumulation of 3-hydroxyanthranilate, 2,3-dihydroxybenzoate, 2,5-dihydroxybenzoate, 3-aminobenzoate or 3-hydroxybenzoate respectively, using the metabolites anthranilate, isochorismate, salicylate, 3-dehydroshikimate and chorismate as substrates ( Gehring et al. 1997 , Liu et al. 2010 , Andexer et al. 2011 , Hirayama et al. 2013 , Zhang et al. 2013 , Fang and Zhou 2014 ). It is interesting to note that very low amounts of 2,3- and 2,5-dihydroxybenzoates could be detected in our metabolite extracts from wild-type plants (∼10 nmol g –1 FW), whereas hydroxyanthranilates, 3-hydroxybenzoate and 3-aminobenzoate were below the detection limit, and, to our knowledge, have never been measured in plant extracts. With the exception of the occurrence of the phytoalexin p -coumaroyl 5-hydroxyanthranilate in oats ( Collins 1989 ), p -coumaroyl conjugates of such phenolics have never been reported. Lastly, although catechol and hydroquinone are found in plants ( Mageroy et al. 2012 , Rychlińska and Nowak 2012 ), no corresponding p -coumaroyl esters have been described, maybe as a consequence of poor HCT activity towards these diphenols ( Supplementary Fig. S6 ). HCT seems to have appeared at the same time that plants began colonizing land. It is unclear why such promiscuity has been retained, and this question will require further study. However, a strategy of exploiting this substrate promiscuity of HCT offers a novel way to reduce lignin in plants. Because this enzymatic activity appears to be conserved among land plants, this strategy can be valuable to affect HCT activity in bioenergy crops for which no HCT gene has been cloned or for which multiple, functionally redundant copies are present. In this respect, particular attention should be paid to the spatio-temporal activity of the transgene promoters employed for the overproduction of HCT alternative substrates to limit possible growth defects associated with lignin reduction as observed in this study using Arabidopsis. Lastly, as we show here in the case of protocatechuate overproduction, such an approach has the potential to add value to plant biomass by increasing the amount of valuable metabolites utilized in the bio-based chemical industry ( Linger et al. 2014 )."
} | 4,598 |
31762134 | null | s2 | 3,444 | {
"abstract": "Soft matter systems and materials are moving toward adaptive and interactive behavior, which holds outstanding promise to make the next generation of intelligent soft materials systems inspired from the dynamics and behavior of living systems. But what is an adaptive material? What is an interactive material? How should classical responsiveness or smart materials be delineated? At present, the literature lacks a comprehensive discussion on these topics, which is however of profound importance in order to identify landmark advances, keep a correct and noninflating terminology, and most importantly educate young scientists going into this direction. By comparing different levels of complex behavior in biological systems, this Viewpoint strives to give some definition of the various different materials systems characteristics. In particular, the importance of thinking in the direction of training and learning materials, and metabolic or behavioral materials is highlighted, as well as communication and information-processing systems. This Viewpoint aims to also serve as a switchboard to further connect the important fields of systems chemistry, synthetic biology, supramolecular chemistry and nano- and microfabrication/3D printing with advanced soft materials research. A convergence of these disciplines will be at the heart of empowering future adaptive and interactive materials systems with increasingly complex and emergent life-like behavior."
} | 365 |
25670779 | PMC4337570 | pmc | 3,447 | {
"abstract": "ABSTRACT Reef-building corals form essential, mutualistic endosymbiotic associations with photosynthetic Symbiodinium dinoflagellates, providing their animal host partner with photosynthetically derived nutrients that allow the coral to thrive in oligotrophic waters. However, little is known about the dynamics of these nutritional interactions at the (sub)cellular level. Here, we visualize with submicrometer spatial resolution the carbon and nitrogen fluxes in the intact coral-dinoflagellate association from the reef coral Pocillopora damicornis by combining nanoscale secondary ion mass spectrometry (NanoSIMS) and transmission electron microscopy with pulse-chase isotopic labeling using [ 13 C]bicarbonate and [ 15 N]nitrate. This allows us to observe that (i) through light-driven photosynthesis, dinoflagellates rapidly assimilate inorganic bicarbonate and nitrate, temporarily storing carbon within lipid droplets and starch granules for remobilization in nighttime, along with carbon and nitrogen incorporation into other subcellular compartments for dinoflagellate growth and maintenance, (ii) carbon-containing photosynthates are translocated to all four coral tissue layers, where they accumulate after only 15 min in coral lipid droplets from the oral gastroderm and within 6 h in glycogen granules from the oral epiderm, and (iii) the translocation of nitrogen-containing photosynthates is delayed by 3 h.",
"conclusion": "Conclusion. By combining pulse-chase double stable-isotopic labeling ( 13 C and 15 N) with TEM ultrastructural and NanoSIMS isotopic imaging, we have visualized and quantified at subcellular levels the incorporation and turnover of C and N in the symbiotic reef-building coral P. damicornis . These results provide a qualitative baseline of subcellular allocation and turnover of C- and N-containing photosynthates in the coral-dinoflagellate symbiosis. In the future, more precise quantitative C and N budgets for symbiotic reef corals could be constructed, using additional respiration and photosynthesis measurements. Moreover, alterations in the pattern of C and N utilization by symbiotic corals might be more precisely characterized in response to heterotrophic feeding and to global environmental changes.",
"introduction": "INTRODUCTION Photosynthesis plays a central role in many aquatic animals symbiotically associated with microalgae or cyanobacteria ( 1 ). Shallow-water reef-building scleractinian corals hosting photosynthetic dinoflagellates of the genus Symbiodinium (“zooxanthellae”) represent an emblematic example of such a stable mutualistic endosymbiotic relationship, which is critical for the development and health of coastal coral reef ecosystems in (sub)tropical oceans. The dinoflagellate endosymbionts, located within the coral gastrodermal cells (see Fig. S1 in the supplemental material), significantly contribute to the nutrition of their animal host partner by transferring a large fraction (up to 90%) of their photosynthetically assimilated carbon (C) and nitrogen (N) to support growth, respiration, reproduction, and biocalcification of the coral in nutrient-poor marine environments ( 2 , 3 ). These photosynthates are produced by dinoflagellates through the fixation of dissolved inorganic carbon (DIC) via the Calvin-Benson “C3” photosynthetic pathway ( 4 ) and through the photosynthesis-dependent acquisition of dissolved inorganic nitrogen (DIN), ultimately via the glutamine synthetase-glutamate synthase (GS-GOGAT) enzymatic cycle ( 5 , 6 ). The nature of translocated photosynthates (“mobile compounds”) ranges from soluble low-molecular-weight compounds, such as glycerol, glucose, amino acids, and organic acids ( 7 – 9 ), to more complex molecules, such as free fatty acids ( 10 ) or glycoconjugates ( 11 ). However, the detailed pathway of this nutritional autotrophic flux from the dinoflagellate endosymbionts to the different cellular layers composing the coral host tissue, as well as the precise fate and turnover of photosynthates in the symbiotic system, remain poorly documented at the (sub)cellular level. Symbiotic reef-building corals are regarded as “fat organisms” because they contain 9 to 47% (dry weight) lipids in their tissue, mostly in the form of neutral lipids (triglycerides, wax esters, and sterols) packed into lipid droplets (LDs), which are hypothesized to be a main sink for C-rich photosynthates translocated by dinoflagellates to the coral tissue ( 12 – 15 ). In support of this view, most previous bulk-level studies using radioactive ( 14 C) or stable ( 13 C) isotope labeling found preferential incorporation of translocated photosynthates into a chemically extracted lipid fraction, as well as structural polymeric compounds such as proteins ( 16 – 21 ). Additionally, recent observations indicate morphological and compositional changes of coral LDs upon coral bleaching (i.e., loss of dinoflagellates or their pigmentation) and a positive correlation between abundance of coral LDs and dinoflagellate density or light intensity ( 22 , 23 ). Nevertheless, despite their supposed key role in the trophic interactions within the coral-dinoflagellate endosymbiosis, a direct demonstration that coral LD biosynthesis is linked with the release of photosynthates by dinoflagellates is still lacking. Glycogen is another potentially important C reserve pool in the endosymbiosis, previously detected in stony corals both biochemically and ultrastructurally ( 24 , 25 ). Gene expression for glycogen synthase and glycogen phosphorylase enzymes, which regulate the production and mobilization of glycogen stores, was detected in the reef coral Acropora aspera transcriptome ( 26 ). However, the possible incorporation of photosynthates such as glucose ( 9 ) into coral glycogen has not been investigated. Furthermore, little attention has been paid to the allocation and turnover of photosynthates within the dinoflagellate subcellular compartments, especially in their C storage structures, which are LDs and starch granules ( 27 , 28 ). Nanoscale secondary ion mass spectrometry (NanoSIMS) ion microprobe imaging is a powerful tool to simultaneously image and quantify the distribution and turnover of stable isotopic tracers (e.g., 13 C and 15 N) inside cells, especially when correlated with ultrastructural transmission electron microscopy (TEM) imaging ( 29 – 33 ). Here, we used this methodological approach on microcolonies (nubbins) of the common Indo-Pacific symbiotic reef-building coral Pocillopora damicornis , which were pulse-labeled in an aquarium for 6 h simultaneously with [ 13 C]bicarbonate and [ 15 N]nitrate ( 15 NO 3 − ), followed by an extended chase of 186 h under either normal light/dark cycling (12 h/12 h) or prolonged darkness. We used [ 15 N]nitrate to unambiguously track the flow of N in the endosymbiotic system because nitrate is assimilated by the dinoflagellates only, in contrast to ammonium, the preferred source of DIN for most reef corals, which is simultaneously assimilated by both dinoflagellate and coral cells ( 32 , 33 ). The two main objectives of this study were to visualize and measure in situ , with subcellular resolution, the photosynthesis-dependent incorporation, fate, and turnover of inorganic C in the dinoflagellate endosymbionts and to track the translocation of dinoflagellate photosynthates toward the coral host tissue layers (see Fig. S1 in the supplemental material), especially their incorporation and turnover in coral LDs and glycogen granules.",
"discussion": "DISCUSSION Subcellular imaging of photosynthetic C and N assimilation and utilization by the dinoflagellate endosymbionts. This study demonstrates that in the tropical reef-building coral P. damicornis , a substantial fraction of photosynthetically assimilated inorganic C and N was retained in the dinoflagellate cells during the pulse-chase experiments. C-containing photosynthates rapidly (within 15 min) accumulated into dinoflagellate lipid droplets (LDs) and starch granules (both primary and secondary), with subsequent rapid turnover. Most accumulated 13 C was depleted from dinoflagellate C reserves within the first 18 h of the chase under light/dark cycling, over a period that includes the first 12-h dark phase. This result strongly suggests a diurnal rhythmicity in the formation (under light) and utilization (under dark) of LDs and starches by dinoflagellates. The 15 N labeling observed in the 13 C-enriched dinoflagellate LDs and starch granules most likely reflects the incorporation of 15 N-labeled proteins, possibly enzymes involved in the synthesis and further catabolism of neutral lipids or carbohydrates, onto the surface or into the internal matrix of these compartments ( 35 , 36 ). In addition, assimilated 13 C and 15 N were allocated to the various other dinoflagellate compartments (including, e.g., the nucleus and plastid), albeit for C with a lower efficiency and a slower turnover than for the C reserves, most likely reflecting the utilization of C and N for dinoflagellate maintenance, growth, and division. These NanoSIMS results obtained in situ (i.e., in the intact coral-dinoflagellate association) are in agreement with data from previous bulk-level isotopic incubation analyses with 14 C- or 13 C-labeled bicarbonate, which report a rapid loss of 14 C or 13 C enrichment in the dinoflagellate fraction within the first hours of the chase, especially during the first night ( 17 , 21 , 37 – 39 ). In particular, by labeling the reef coral Acropora pulchra simultaneously with [ 13 C]bicarbonate and [ 15 N]nitrate, Tanaka et al. ( 21 ) found a dramatic nighttime decrease in the dinoflagellate fraction of the C:N ratio of light-produced compounds, suggesting rapid consumption of photosynthates with high C content (i.e., lipids and carbohydrates) by dinoflagellate respiration. Conversely, in Stylophora pistillata colonies, the rapid decrease in 14 C or 13 C labeling of the dinoflagellate fraction, observed over a chase period of 24 to 48 h, was mainly ascribed to delayed translocation of 14 C- or 13 C-labeled photosynthates to the coral host partner ( 17 , 39 ). Here, we observed that the rapid 13 C decrease in dinoflagellate endosymbionts of P. damicornis is occurring via isotopic depletion of their intracellular LDs and starch granules and not through depletion of their other cell compartments. This result supports the hypothesis that lipids and carbohydrates stored by dinoflagellates during daytime are quickly remobilized, especially during nighttime, via mitochondrial respiration (releasing 13 CO 2 ) to sustain dinoflagellate metabolism. Nevertheless, the following additional mechanisms for such rapid 13 C depletion in dinoflagellate C reserves cannot be excluded: (i) the effect of an isotopic dilution due to the additional storage of newly produced [ 12 C]photosynthates during the light periods of the chase, (ii) the translocation of 13 C-labeled compounds toward the coral host tissue (during light and dark periods of the chase), and/or (iii) the remobilization of stored 13 C as building blocks for dinoflagellate maintenance, growth, and division. Interestingly, by maintaining P. damicornis microcolonies under constant darkness, a treatment known to trigger coral bleaching after about 4 days ( 40 , 41 ), the ultrastructural disappearance of dinoflagellate LDs and starch granules was systematically accompanied by features of cell vacuolization and damage of organelles, indicative of in situ degradation of the endosymbionts. These observations are additional evidence that photosynthates stored in dinoflagellates under light are essential to further sustain their respiration and metabolism, especially during nighttime. In marine microalgae, photosynthates produced and stored under light might provide C skeletons and energy to support DIN (ammonium and nitrate) uptake and assimilation in the dark ( 42 ). The existence of an internal C reservoir in dinoflagellate endosymbionts, metabolized during nighttime to sustain N acquisition, is suggested by the extended time of darkness (at least 15 h) needed to efficiently inhibit ammonium incorporation in symbiotic reef corals ( 5 , 33 ). Similarly, we demonstrate that prolonged dark pretreatment (24 h) fully repressed nitrate assimilation in dinoflagellates of the reef coral P. damicornis . Interestingly, in coastal marine environments, migrating free-living dinoflagellates reach the illuminated sea surface, poor in DIN, during the day, to accumulate excess photosynthates not channeled toward protein synthesis, whereas they descend during the night to a depth enriched in nitrate, which is then efficiently assimilated through the remobilization of light-produced C reserves ( 43 ). Hence, the storage of lipids and carbohydrates by dinoflagellates during daytime might constitute a C reserve helping reef corals to efficiently sustain DIN acquisition during nighttime, representing an adaption to nutrient-poor environments. The fate of C and N translocated to the coral host. The present study reveals the subcellular pathways of photosynthetic C translocation from the dinoflagellate cells toward all four epithelia composing the coral host tissue, extending our previous NanoSIMS observations of transepithelial movements of nitrogenous compounds translocated by the endosymbionts ( 32 , 33 ). In early autoradiographic investigations, transepithelial metabolite fluxes have been reported in marine cnidarian tissue, albeit at light microscopy level in thick histological sections ( 44 , 45 ). Moreover, using bulk-level isotopic measurements of separated fractions of dinoflagellate-containing oral gastroderm and dinoflagellate-free oral epiderm prepared from tentacles of a sea anemone, Trench ( 18 ) previously found that 18 to 31% of total net photosynthates moved toward the oral epiderm within 10 h of labeling in light with [ 14 C]bicarbonate. Here, we observed that coral LDs from the oral gastroderm epithelium constitute a major accumulation site for translocated photosynthetic 13 C, providing direct validation of previous hypotheses ( 12 , 13 , 15 , 22 , 23 ). The rapid C translocation toward gastrodermal coral LDs, which we visualized already at 15 min from the onset of the pulse, is consistent with results from bulk-level isotopic investigations of symbiotic sea anemones, which reported rapid translocation of photosynthates to the host fraction, within a few minutes following their production ( 9 , 46 , 47 ). Interestingly, we frequently observed 13 C-enriched LD-like structures located in the symbiosomal space between the dinoflagellate endosymbiont and the coral gastrodermal host cell (white arrows in Fig. S3 in the supplemental material). These structures have previously been interpreted as “extra-algal” LDs produced by dinoflagellates, in the process of exocytosis toward the host gastrodermal cells ( 48 , 49 ). However, ultrastructural evidence for such a potential exocytotic process is still lacking. Moreover, the occurrence of LDs within coral cells of dinoflagellate-free epithelia (oral epiderm and calicoderm) implies the existence of other, still unknown, mechanisms of coral LD formation. This study also provides evidence that coral glycogen granules in the coral oral tissue constitute another major sink for 13 C-photosynthates translocated by dinoflagellates. In bulk isotopic analyses of sea anemones, glucose was found to be a major metabolite translocated within minutes from the dinoflagellate endosymbionts to the host fraction ( 7 – 9 ). Moreover, transcriptome analyses have revealed that scleractinian ( Acropora genus) corals have the enzymatic machinery for synthesis and remobilization of glycogen from and to glucose ( 26 ). Consistent with these reports, combined TEM and NanoSIMS observations show incorporation (within 6 h) of external [ 13 C 6 ]glucose (30 µM) into glycogen granules of the coral oral epidermal cells (see Fig. S9 in the supplemental material). Thus, our results provide direct evidence of the functional mechanisms of storage of photosynthetic C, translocated by the dinoflagellate endosymbionts (probably in the form of glucose) to coral glycogen in the oral tissue. Similar to dinoflagellate C reserves, the 13 C depletion observed for coral LDs in the oral gastroderm and for the glycogen granules in the oral tissue most likely reflects the breakdown of neutral lipids and carbohydrates to sustain coral cell respiration. Nevertheless, we cannot exclude the potential contribution of (i) an isotopic dilution effect resulting from the translocation of photosynthates with normal C isotopic composition to coral LDs and glycogen and (ii) the reallocation of stored C toward coral cell maintenance, growth, and division. Translocation of N-containing compounds by dinoflagellates was not recorded immediately in the coral host partner but was observed with a delay of 3 h following the onset of the pulse-chase experiment under light/dark cycling, confirming previous observations ( 33 ). These results suggest a temporal separation between the translocation by dinoflagellates of C- and N-containing photosynthetic assimilates. Compounds bearing C are released within a few minutes after their production, compared to a time scale of hours for N-bearing compounds (summarized in Table 1 ). It is also possible that early translocation of N to the coral tissue has been partly masked by a potentially high rate of N recycling by the dinoflagellates ( 5 ) or by extraction of low-molecular-weight soluble nitrogenous compounds (e.g., free amino acids) during sample preparation. Alternative NanoSIMS sample preparation methods (e.g., cryofixation and cryosubstitution) might improve tracking of low-molecular-weight soluble photosynthates. Conclusion. By combining pulse-chase double stable-isotopic labeling ( 13 C and 15 N) with TEM ultrastructural and NanoSIMS isotopic imaging, we have visualized and quantified at subcellular levels the incorporation and turnover of C and N in the symbiotic reef-building coral P. damicornis . These results provide a qualitative baseline of subcellular allocation and turnover of C- and N-containing photosynthates in the coral-dinoflagellate symbiosis. In the future, more precise quantitative C and N budgets for symbiotic reef corals could be constructed, using additional respiration and photosynthesis measurements. Moreover, alterations in the pattern of C and N utilization by symbiotic corals might be more precisely characterized in response to heterotrophic feeding and to global environmental changes."
} | 4,663 |
24367357 | PMC3853793 | pmc | 3,449 | {
"abstract": "The Antrim Shale in the Michigan Basin is one of the most productive shale gas formations in the U.S., but optimal resource recovery strategies must rely on a thorough understanding of the complex biogeochemical, microbial, and physical interdependencies in this and similar systems. We used Illumina MiSeq 16S rDNA sequencing to analyze the diversity and relative abundance of prokaryotic communities present in Antrim shale formation water of three closely spaced recently fractured gas-producing wells. In addition, the well waters were incubated with a suite of fermentative and methanogenic substrates in an effort to stimulate microbial methane generation. The three wells exhibited substantial differences in their community structure that may arise from their different drilling and fracturing histories. Bacterial sequences greatly outnumbered those of archaea and shared highest similarity to previously described cultures of mesophiles and moderate halophiles within the Firmicutes , Bacteroidetes , and δ- and ε- Proteobacteria. The majority of archaeal sequences shared highest sequence similarity to uncultured euryarchaeotal environmental clones. Some sequences closely related to cultured methylotrophic and hydrogenotrophic methanogens were also present in the initial well water. Incubation with methanol and trimethylamine stimulated methylotrophic methanogens and resulted in the largest increase in methane production in the formation waters, while fermentation triggered by the addition of yeast extract and formate indirectly stimulated hydrogenotrophic methanogens. The addition of sterile powdered shale as a complex natural substrate stimulated the rate of methane production without affecting total methane yields. Depletion of methane indicative of anaerobic methane oxidation (AMO) was observed over the course of incubation with some substrates. This process could constitute a substantial loss of methane in the shale formation.",
"conclusion": "Concluding remarks Microbial community composition in closely spaced production wells can differ substantially due to local variation in both reservoir quality (e.g., lithology, matrix and natural fracture microstructure, fluid chemistry and saturation) and completion quality (e.g., hydraulic fracture design and fluid composition, well cleanup workflow, and subsequent production strategy). The microbial fermenting community in our study was capable of rapidly utilizing substrates, such as glucose and yeast extract, suggesting that they are well-adapted to decompose relatively labile OM including organic additives used in drilling and fracturing fluids. Hydrogenotrophic methanogens were detected, but the production waters from recently fractured wells appear to be dominated by methylotrophic methanogens, capable of utilizing high concentrations of methanol likely stemming from the fracturing fluids. Furthermore, we found that increased surface area stimulated methane production. However, a loss of methane over the course of incubation suggests that AMO also occurs in formation waters, which might constitute a substantial loss of methane in the shale formation. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.",
"introduction": "Introduction Microbial gas formation through decomposition of sedimentary organic matter (OM) comprises roughly 20% of the world's natural gas resources (Rice, 1992 ), and it is estimated that even more microbial gas is retained in the corresponding source rocks of unconventional biogenic gas shales (Milkov, 2011 ). In the U.S. unconventional gas resources account for nearly 10% of the total natural gas generation (Martini et al., 2003 ). The Michigan Basin, centered on the lower peninsula of Michigan in the U.S., is home of the Devonian (~380 Ma-old) Antrim Shale formation, one of the most productive gas shale formations in the U.S. The finely laminated Antrim Shale contains thermally immature OM with a total organic carbon (TOC) content of 0.5–24% (Shurr and Ridgley, 2002 ). In the central and eastern Michigan basin the Antrim shale contains thermogenic gas, created by pressure and thermal cracking of OM. Here, gas production rates in drilled wells are low and economically not successful (Martini et al., 2008 ). In contrast, shale gas production along the northern margin of the Michigan Basin is high and became a target area of rapid development starting with 100 wells in 1985 to over 12,000 gas-and water producing wells installed today (Martini et al., 2004 ; DEQ, 2012 ). Previous geochemical studies provided evidence for biologically mediated methane generation in the recent geological past along the northern margin of the Michigan Basin (Martini et al., 1996 ). Dilution of deep basin brines with meteoric waters and deep groundwater recharge during Pleistocene glaciations resulted in a steep salinity gradient in Antrim Shale pore waters with extremely diluted water along the margins to greater than 5 M NaCl at the center of the basin (McIntosh et al., 2002 ). The Antrim shale has been hydrologically isolated from surface water for at least 7000 years (Martini et al., 1998 ; McIntosh et al., 2002 ) and the formation water lacks abundant inorganic electron acceptors other than carbon dioxide, including sulfate and iron oxyhydroxides (Waldron et al., 2007 ). Such conditions are favorable for methanogens which often thrive in environments where carbon dioxide is the sole available electron acceptor (Whitman et al., 2006 ). Although the high salt concentrations in the Antrim formation water inhibit many microorganisms, halophilic organisms can thrive under a wide range of NaCl concentrations (Oren, 1999 ). A recent study on enrichment cultures from formation water of the methane-generating zone of the Antrim shale provided evidence of halophilic methanogenic communities growing at up to 2.5 M NaCl concentrations (Waldron et al., 2007 ), indicating that active methanogenesis is an ongoing process in the northern margins of the Antrim gas shale. Although little is known of microbial communities in gas shale formations, similarly hydrocarbon-rich, anaerobic environments such as petroleum reservoirs and subsurface coal beds have been relatively well-studied (Magot et al., 2000 ; Strąpoć et al., 2008 , 2011 ). Similar to the Antrim Shale, carbon dioxide is the predominantly bioavailable electron-acceptor in coal bed formation water (Strąpoć et al., 2011 ). Several studies have demonstrated the presence of active methanogenic archaea in coal beds (Shumkov et al., 1999 ; Green et al., 2008 ; Harris et al., 2008 ; Orem et al., 2010 ) and some have reported enhanced methane production with increases in surface area (Green et al., 2008 ), addition of inorganic nutrients (Harris et al., 2008 ) and trace elements (Ünal et al., 2012 ). The conversion of refractory OM to methane involves primarily the fermentation of polymers and monomers to fatty acids, organic acids, alcohols, hydrogen, and carbon dioxide. Subsequently degradation follows via secondary fermenting bacteria, homoacetogenic bacteria and acetoclastic, methylotrophic and hydrogenotrophic methanogens (Schink, 2006 ). Similarly to coal bed formations, gas shale formations have shown enhanced methane production with increased surface area (Curtis, 2002 ). Horizontal drilling and hydraulic fracturing have now become two key technologies in shale gas exploration (Arthur et al., 2008 ; Kerr, 2010 ). During the drilling operation large volumes of drilling mud are pumped into the formation to cool and lubricate the drilling bit. Drilling mud contains cellulose, barite, and lignosulfonates (Caenn et al., 2011 , www.fracfocus.org ), which could serve as carbon and sulfate sources for microorganisms. Hydraulic fracturing is a widely used technique to fracture gas shale by pumping fluids and sand into production wells at high pressure (Arthur et al., 2008 ), resulting in enhanced methane extraction. Biocides are added to the fracturing water to control bacterial growth (Arthur et al., 2008 ), although recent studies have shown that bacteria can survive this treatment (e.g., Struchtemeyer and Elshahed, 2012 ) and additives such as polyacrylamide and sugar-based polymers might even be utilized microbially (Arthur et al., 2008 ). Despite the widespread utilization of the drilling and hydraulic fracturing procedures not much is known about the impact on the microbial communities present in formation water. Here, we phylogenetically characterized the prokaryotic community in formation waters of three recently fractured gas-producing wells (denoted A3-11, B1-12, and C1-12) from the western margin (Manistee county) of the Antrim Shale to identify possible key microbial players in methanogenic shale gas production. Incubation experiments with formation water were performed to stimulate methanogenic communities using a variety of substrates. Direct methanogenic substrates tested include small organic acids, methanol, and trimethylamine (TMA). Sterile powdered shale, yeast extract, propionate, and glucose were used to test the ability of bacteria to convert these compounds into methanogenic substrates. Phylogenetic surveys of the incubation experiments were performed to identify the microbial communities that were stimulated by the various substrate additives. Methanol, TMA, and yeast extract substrates were monitored for consumption.",
"discussion": "Discussion Production water characteristics and implication on microbial community The methane-producing zone in the Antrim Shale formation contain formation waters whose temperature and salinity range is suitable for mesophilic moderately halophilic microorganisms (Martini et al., 1996 ; Waldron et al., 2007 ). Nonetheless, as mentioned earlier, the three wells were all recently fractured and their production water profiles suggest that water samples collected for this study might have entrained volumetrically significant amounts of hydraulic fracture flowback fluid, with an especially large proportional contribution implied for the A3-11 well. Thus, the well water microbial community identified in our survey most likely represents a mixture of indigenous shale communities and allochthonous species, which were introduced to these reservoirs during drilling and fracturing procedures (Struchtemeyer et al., 2011 ; Struchtemeyer and Elshahed, 2012 ). The residence time of the introduced drilling mud and fracturing fluid into the shale formation impacts the formation water chemistry and biology. During the drilling procedure large volumes of drilling mud are often lost in the shale formation (Gray et al., 1980 ; Grace, 2007 ) and only 30–70% of fracturing fluids injected into wells are recovered in the flowback waters (Veil, 2010 ). A recent study on hydraulic fractured thermogenic wells in the Barnett Shale (USA) showed that the concentration of salt, iron and total dissolved solids were higher in flowback water of a well that had been in contact with fracturing fluids for 2 months compared to a well with a much shorter contact time of 24 h (Struchtemeyer and Elshahed, 2012 ). The authors suggested that the differences in the bacterial community in the flowback water of the two wells might be influenced by the different time intervals (2 month vs. 24 h) between fracturing and flowback at the two sites (Struchtemeyer and Elshahed, 2012 ). Our data support these findings. Namely, the prokaryotic beta diversity in the two wells (A3-11 and C1-12) with a longer contact time of fracturing fluid was more similar than in B1-12 with a shorter contact time of fracturing fluid as indicated by the differences in methanol concentration. In addition to methanol, cellulose, lignosulfonates, and sugar-based polymers are components of fracturing fluids and drilling muds (Caenn et al., 2011 , www.fracfocus.org ). For example, guar gum or hydroxyethyl cellulose is frequently used in fracturing fluids to thicken the water in order to suspend the sand. These additives might explain the presence of an active fermenting microbial community in the production water as observed in the incubation experiments by the quick utilization of fermentable substrates such as yeast extract and glucose. The prominent heterotrophs in the production water and the incubation experiments were members of the genus Haloanaerobium ( Firmicutes ) and the genera Marinilabilia and Cytophaga ( Bacteroidetes ). Haloanaerobium species usually ferment saccharides and produce H 2 , CO 2 , and C 2 compounds (mainly acetate and sometimes ethanol, as was detected in the glucose treated incubations) (Ollivier and Cayol, 2005 ). One prominent OTU was closely related to Marinilabilia salmonicolor (Nakagawa and Yamasato, 1996 ), a bacterium capable of utilizing cellulose and complex carbohydrate. Additional sequences were related to Cytophaga, Dechloromonas , and Pseudomonas and members of these genera have been reported to be capable of growing on crude oil, benzene, xylene, and toluene (Prince, 2005 ), and might play an important role in the degradation of refractory OC in the Antrim shale. Also, the unclassified bacteria constitute a substantial fraction of the total bacterial reads, especially in the B1-12 well. These bacteria might also be capable of utilizing complex organic shale matter. Furthermore, a number of archaeal OTUs unaffiliated with known methanogenic groups were detected in the initial production water samples, which were not subsequently detected during the bottle incubations. These euryarchaeota may possess non-methanogenic metabolisms and also be involved in bitumen degradation in situ . Methanogenic capabilities in incubation experiments Methanol was quickly utilized in the incubation experiments and high concentrations of added methanol did not inhibit methylotrophic methanogenesis suggesting that the methanogenic community is well-adapted to high methanol concentrations most likely derived from the fracturing procedure since methanol is a standard ingredient in fracturing fluids. Halophilic methanogens generally utilize substrates such as methanol, methylamines, or dimethyl sulfide, which yields more free energy compared to acetate or hydrogen (Oren, 2001 ). This extra energy might be used for their osmoregulatory system to balance the energetic demands of the saline environment (Oren, 2001 ). TMA was below detection limit in our formation waters suggesting that this compound plays a minor role in comparison to methanol as a substrate for methylotrophic methanogens in this environment. TMA could potentially be formed in formation water through fermentation of betaine, which is an osmoregulatory compound synthesized by some halophilic bacteria (Oren, 1990 ). The bacterial fermentation of betaine can yield acetate and TMA. Also, similar to coal matrices (Strąpoć et al., 2011 ) the organic-rich Antrim Shale might be a source of methylamines. Acetate was not utilized by methanogens in any of the three analyzed well waters and we did not find any phylogenetic evidence for acetoclastic methanogens. This is in agreement with the finding that acetoclastic and hydrogenotrophic methanogens generally thrive in freshwater or lower salinity environments (Oren, 1999 , 2001 ). Instead, hydrogenotrophic methanogens were detected in the initial well waters and after incubation with yeast extract. Hydrogenotrophic methanogens were also identified when formate was used as a substrate, but only in B1-12. Here, hydrogen was generated by fermentation and consumed concomitant with methane production. B1-12 incubated with formate yielded an OTU closely related to Methanocalculus halotolerans , the most halotolerant hydrogenotrophic methanogen known to date and capable of withstanding concentrations of up to 12% NaCl (Ollivier et al., 1998 ). One OTU closely related to Methanoplanus limicola (Wildgruber et al., 1982 ) was detected in B1-12 and C1-12 well waters incubated with yeast extract. This methanogen tolerates salt concentrations between 0.4 and 5.4% (Wildgruber et al., 1982 ), which is substantially lower than the salt concentration of 8–10% in our formation waters. Most likely this OTU represents a novel hydrogenotrophic methanogen with a higher salinity tolerance. Both hydrogenotrophic methanogens might possess different hydrogen affinities since the utilization of formate resulted in a 10-fold higher hydrogen concentration then the amount of hydrogen resulting from the fermentation of yeast extract. The OTU related to M. limicola , might be better adapted to grow at low hydrogen concentrations, whereas the OTU related to M. halotolerans , is capable of thriving at elevated hydrogen concentrations. This could be similar to methanogens from rice paddies, which grow slowly if at all at high hydrogen concentrations (Sakai et al., 2007 ). The low methane generation in the glucose-treated incubations could be explained by an inhibition of the methanogenic community caused by the rapid depletion of nutrients and/or the decrease in pH from the accumulation of organic acid intermediates (Strąpoć et al., 2011 ) as a result of bacterial fermentation. The accumulation of ethanol caused by fermentation might also have had an inhibitory effect on the methanogens present in the glucose treatment. The increase of mineral surface area greatly stimulated methane production rates as observed from formation waters that were incubated with sterile powdered shale. However, the addition of shale did not result in consistent increases in methane yield in the C1-12 well water, and the overall methane yield in the other two well waters was identical to the no substrate background. This suggests that over the course of incubation the bitumen in the powdered shale was not utilized by fermentative prokaryotes and did not yield substrates for the methanogenic communities. In contrast, a previous enrichment experiment showed that fermentative bacteria derived from Antrim well water could be enriched by using only water-soluble shale OM and subsequent methane accumulation was detected, demonstrating that methanogenic communities can be supported by using shale derived DOM as the only source of energy and carbon (Huang, 2008 ). Most likely our fermenting microbial community was adapted to the presence of easy degradable carbon compounds such as those stemming from the drilling and fracturing procedure, and was not adapted to degrade bitumen from the shale. Methanol and methane consumption The discrepancy between the theoretically expected methane yields based on methanol consumption and measured methane yields requires non-methanogenic methanol-utilizing metabolisms, or the ability of anaerobic oxidation of methane to be present. This could be explained by OTUs related to methanol-utilizing bacteria such as Geoalkalibacter subterraneus (Greene et al., 2009 ) present in the incubation experiments. Furthermore, some sulfate reducers are known to utilize methanol (Qatibi et al., 1991 ; Tarasov et al., 2011 ), and might be involved in methanol consumption. Also, the detected archaea without close cultured relatives, which were detected in the methanol incubation experiment, might possess the capability to utilize methanol. The depletion of methane suggests anaerobic methane oxidation (AMO) at least in C1-12 well water incubated with powdered shale. Methane can be oxidized anaerobically by a consortium of methane-oxidizing archaea (ANME clusters), and sulfate-reducing bacteria (SRB) (Knittel and Boetius, 2009 ). We did find sequences of SRB in our incubation experiments, however no ANME-related sequences were recovered despite a good coverage of the universal primers used to generate template DNA for Illumina sequencing. Although the archaeal partner involved in AMO was not found, the depletion of methane in some of the incubations and possible involvement of AMO is in agreement with previous geochemical evidence for anaerobic hydrocarbon oxidation in formation water of the western margin of the Antrim shale (Martini et al., 2003 ). The western margin formation water contains some of the highest sulfate concentrations relative to salinity in the Antrim shale (Walter et al., 1997 ), and C2 and C3 gasses are enriched in δ 13 C values, suggesting that anaerobe hydrocarbon oxidation occurred in the western margin (Martini et al., 2003 ). The sulfate present in the formation water is most likely derived from the shale since a recent incubation-based study showed that water-soluble organic shale material sustains SRB (Huang, 2008 ). This is supported by our incubation experiments where a shift toward δ- and ε- Proteobacteria , which comprise many iron-, sulfur-, sulfate-, and nitrate reducers, was observed with the addition of powdered shale. Sulfate can be quickly utilized by SRB, which compete for hydrogen and other low-molecular-weight substrates with methanogens. Recently, a study on formation water from wells along the northern production trend of the Antrim shale showed an increase in sulfate concentrations and SRB over the last two decades (Kirk et al., 2012 ). The authors explained these changes by ongoing processes driven by commercial gas production such as ground water inflow from the sulfate-rich Travers limestone, which is in contact with the Antrim shale (Kirk et al., 2012 ). It was furthermore suggested that this development could have negative implications for commercial gas production by creating conditions that favor growth of SRB (Kirk et al., 2012 ). These bacteria can compete with methanogens for substrate or, as indicated previously (Martini et al., 2003 ) and suggested in our study, SRB could be involved in AMO and responsible for loss of methane in the shale formation."
} | 5,486 |
19961607 | PMC2800837 | pmc | 3,452 | {
"abstract": "Background The temporal dynamics and formation of plant-pollinator networks are difficult to study as it requires detailed observations of how the networks change over time. Understanding the temporal dynamics might provide insight into sustainability and robustness of the networks and how they react to environmental changes, such as global warming. Here we study an Arctic plant-pollinator network in two consecutive years using a simple mathematical model and describe the temporal dynamics (daily assembly and disassembly of links) by random mechanisms. Results We develop a mathematical model with parameters governed by the probabilities for entering, leaving and making connections in the network and demonstrate that A. The dynamics is described by very similar parameters in both years despite a strong turnover in the composition of the pollinator community and different climate conditions, B. There is a drastic change in the temporal behaviour a few days before the end of the season in both years. This change leads to the collapse of the network and does not correlate with weather parameters, C. We estimate that the number of available pollinator species is about 80 species of which 75-80% are observed in each year, D. The network does not reach an equilibrium state (as defined by our model) before the collapse set in and the season is over. Conclusion We have shown that the temporal dynamics of an Arctic plant-pollinator network can be described by a simple mathematical model and that the model allows us to draw biologically interesting conclusions. Our model makes it possible to investigate how the network topology changes with changes in parameter values and might provide means to study the effect of climate on plant-pollinator networks.",
"conclusion": "Conclusion General remarks Our study highlights various interesting points. Based on our model, we have demonstrated that the plant-pollinator network shows strong dynamic stability over the two years; i.e. the dynamic features of the network are highly conserved from one year to the next. The length of the season, temperature and other weather parameters differ (Table 1 ) and also the visiting pollinators are not the same. Despite this we find that the development of the network could be described by very similar parameters in 1996 and 1997. We described the number of links attached to a new insect by a modified geometric distribution. This distribution does not have the characteristic power-law shape that has been reported for other types of network and is in concordance with a previous analysis of the same data [ 6 ]. In addition, our model has distinctive random features; e.g. the number of new species per day is independent of the number already present and the number of insects being removed from the network daily is binomial such that each insect has the same probability of being removed. Studies about the impact of habitat loss on pollination networks highlight the existence of a habitat destruction threshold at which both plants and pollinators disappear suddenly, leading to the collapse of the network [ 21 ]. The transition from maintenance to destruction of the network is very sharp leading to the conclusion that the network's fate might change by a slight modification of the parameter controlling the transition - almost like a phase transition. We observe a similar collapse of the network with parameters changing many folds within a few days. In our case, it could be that a sudden short rise in temperature causes plants to stop flowering [ 17 ]. Alternatively, it could be that night frost towards the end of the season creates north-facing snow patches that persist during day time, thereby reducing the resources available for the insects and making it more difficult for them to recover and regain activity. However, to test this hypothesis we need data on frost (and snow coverage) in the study site, which are not available. Finally, we demonstrated by simulation that the arrival and departure of insects in network has not reached equilibrium when the collapse of the network appears at the end of the season and we estimated that the number of available species in the area is about 80, of which 75%-80% are observed. Climate change might have an impact on plant-pollinator networks [ 7 , 8 , 22 - 25 ]. For example an increase in the average temperature is likely to increase the length of the season and change the conditions for the existence and maintenance of the network. The effect of temperature rise is not well-understood, though some evidence is available. In [ 26 ], it is argued that experimental warming does not alter the length of the flowering season, whereas [ 27 ] and [ 28 ] (a study on butterflies) provide evidence that the adult life cycle of insects is unchanged with increasing temperature. However, the availability of pollinator species or plant species might change as the length of the season changes, as well as when the species are present [ 7 ]. To study consequences of climate changes, observations over several years would help us to relate the parameters to climatic variables; however we might still need to impose more assumptions on the model, e.g. for how long are plants flowering, to what extent do pollinators overlap with their plants etc. Remarks on and limitations of the model We consider our model a first step in modelling the temporal dynamics of plant-pollinator networks, a complex process of network formation. As discussed in the sections above, species-specific information about plant and pollinators is not included in the model. This information includes species identity, species abundance, known species-specific interactions and weather-related parameters, such as temperature thresholds for when species are active and when they rest. In our case, species identity (Additional files 2 and 3 ) and the specific plant-pollinator interactions [ 6 ] are available. All this information potentially influence the dynamics of the network, e.g. when and which connections are made in the network [ 10 , 16 , 29 ]. However, it is not straightforward how to include such elaborate information in the model and the analysis we have proposed. Our model is based on network data gathered over two years and we have demonstrated that many aspects of the dynamics can be accounted for using simple mathematical tools. With detailed observations over several years we might be able to provide models that can account for further aspects of the temporal dynamics and include more detailed describtions of plants, pollinators and their characteristics.",
"discussion": "Results and Discussion The start and end of the season In the Arctic the start of the flowering season correlates strongly with snow melt and a corresponding rapid large change in net and outgoing radiation [ 9 , 10 , 15 ]. This is also the case for our data sets, see Figure 1 . In contrast, the end of the season is not as strongly correlated with changes in net and outgoing radiation, or other climatic parameters [ 9 ]. However, the wind direction does show some correlation with both the start and the end of the season, see Figure 1 . Further, in our case, we note that weather parameters generally differ between years (Table 1 ). Figure 1 The start and end of the season . The figure shows the amount of outgoing radiation at 12 noon from June to August in 1996 and 1997 and the wind direction at 12 noon from April to September, also in 1996 and 1997. The start of the season correlates strongly with reduced level of outgoing radiation, whereas both the start and the end of the season correlate mildly with wind direction (0° -360°). The season is marked with two vertical blue lines; good days (days of observation) are blue and bad days (days where observations were not done due to bad weather conditions) are red. In the case of outgoing radiation (top two figures) days are counted from June 1st (i.e. June 1st = Day 0) and in the case of wind direction days are counted from April 1st (i.e. April 1st = Day 0). Note that 0° and 360° represent the same direction. Table 1 Summary of weather parameters Year Month Temp (C) Net (W/m 2 ) Wind (m/s) 1996 June 1.90 (2.76) 106.58 (145.84) 1.39 (0.93) July 5.84 (3.58) 137.10 (149.48) 2.28 (1.38) August 4.40 (3.20) 68.78 (106.68) 2.50 (1.89) 1997 June 2.23 (2.88) 80.48 (105.61) 2.06 (1.65) July 3.72 (2.52) 123.35 (126.24) 2.40 (1.86) August 5.05 (3.69) 70.74 (111.34) 2.45 (1.74) Shown are mean values with standard deviations in parenthesis of air temperature (Temp), net radiation (Net) and wind velocity (Wind) for the three summer months in 1996 and 1997. The collapse of the network Figure 2 shows that the number of pollinators decreases drastically a few days before the end of the season. This is the case for both years and indicates the beginning of the end of the season. Though the end of the season is not characterized by marked changes in weather conditions (see section above), the collapse of the network could still be related to (minor) changes or fluctuations in weather conditions that cause pollinators to disappear and/or reduce activity or plants to end flowering; see e.g. [ 10 , 16 ] where activity levels of Alpine insects are discussed in relation to varying weather conditions. To test this hypothesis we performed a linear regression of the number of pollinators leaving the network on the ZMS climatic variables from the same day (or the day before; we did both analyses); the analyses showed that the hypothesis is not supported. Generally we overestimate the number of leaving pollinators in the beginning of the season and underestimate the number at the end of the season (Additional file 1 ). It indicates that other factors than the available climatic variables affect the end of the season. Figure 2 Number of pollinators . a) Evolution of the number of pollinators in the network during the season. For both years, the number of pollinators increases until a few days before the end of the season where it collapses suddenly. At the same time, the number of pollinators leaving the network peaks. b) Shown is the estimated sigmoid shape of the parameter p ( x ) (see Methods, Analysis and modelling) for the removal of insects in 1996 and 1997. 'Day' is the good days. In contrast, we found that the disappearance of pollinators from the network is highly correlated with the disappearance of plants from the network (linear regression: r 2 = 0.8 for 1996, and r 2 = 0.7 for 1997). Since we only know the time span a plant is visited and not for how long it is flowering, we cannot say whether links disappear A) because plants stop flowering or B) because insects disappear for other reasons. Some of the plants that stay in the network longest also enter the network very early in the season (see Additional files 2 and 3 ). Based on the available data, it is therefore difficult to distinguish between A and B. However, we note that the collapse of the network follows immediately after the temperature has peaked: In 1996, the temperature rises above 15°C for several days and in 1997, the temperature rises above 20°C in a single day (Additional file 4 ). It is plausible that flowering of plants adapted to cold climate could be affected by high temperatures [ 17 ]. When interpreting the results it should be kept in mind that weather conditions to some extent could be local and heterogeneous; e.g. variations in the orientation of the ground level towards the sun could impose restrictions on the hours of sun, wind and the snow coverage of plants compared to ZMS. Heterogeneous snow distribution has previsouly been reported to influence e.g. alpine ecosystems [ 10 , 16 ]. Simple distributions describe the temporal development Our main focus is to describe the temporal dynamics of the network with simple probability distributions and compare the results between the two years. The composition of the plant community is the same for the two years; in total 31 flowering plants comprising the same species each year; see [ 6 ] and Additional file 2 . Therefore we consider the plants as fixed and study the network from the view of the pollinators. In particular, we are interested in the following features: • How many pollinators enter the network daily? • How many pollinators leave the network daily? • How many links do pollinators get when they first enter the network? • How many links do pollinators gain daily while they are in the network? • How many links do pollinators loose daily while they are in the network? We use the answers to the above questions as a scaffold to develop a model of network assembly and disassembly. The daily arrival of pollinators can be described by a Poisson distribution (Table 2 ). The empirical graph of pollinators leaving the network shows a drastic change a few days before the end of both seasons (Figure 2 , as discussed above). To accommodate this we fitted a binomial distribution to the number of leaving pollinators assuming a gradual sigmoid change in the parameter (see Table 2 and Methods, Analysis and modelling). Since the climatic parameters did not explain species disappearance we did not use these parameters for modelling. Table 2 Model summary Parameter 1996 1997 1996&1997 LTR Arrival of insects Poisson λ 2.417 (0.317) 2.423 (0.305) 2.420 (0.220) Yes p 0.094 0.15 0.28 0.99 Links for new insects Modified geometric r 1 0.414 (0.065) 0.429 (0.062) 0.421 (0.045) Yes r 0.351 (0.049) 0.356 (0.048) 0.354 (0.034) p 0.65 0.37 0.55 0.98 Death of insects Sigmoid binomial α 0.058 (0.012) 0.037 (0.008) 0.046 (0.007) Yes β 0.436 (0.071) 0.469 (0.143) 0.446 (0.063) H 4.901 (4.942) 2.797 (3.688) 3.847 (3.984) T 19.66 (0.362) 22.31 (0.747) p 0.41 0.51 0.75 0.57 Addition of links Geometric q 0.752 (0.019) 0.836 (0.015) 0.799 (0.012) No p 0.13 0.078 0.031 < 10 -3 Removal of links Sigmoid binomial γ 0.006 (0.002) 0.006 (0.002) 0.008 (0.002) No δ 0.187 (0.021) 0.401 (0.036) 0.271 (0.020) h 18.28 (9.497) 40.12 (-)* 48.68 (-)* t 19.06 (0.031) 22.55 (-)* p 0.15 0.73 < 10 -3 < 10 -3 Shown are the fitted parameters for the two years; estimated for each year separately and jointly (except T and t ) with standard deviations in parenthesis. The p-values in the 3rd, 4th and 5th columns are goodness-of-fit probabilities based on the chi-square test; the p-values in the 6th column are test probabilities for the LRT (Yes/No indicates whether the LRT is accepted or not). Even when the LRT is not accepted the parameters are of similar magnitude in the two years. The distributions are explained in Methods, Analysis and modelling. Standard deviations marked with * could not be estimated reliably. When pollinators first enter the network, many have few links while fewer have many links; 57% (resp. 63%) of the pollinators enter the network with one or two links and 12% (resp. 13%) have six or more links in 1996 (resp. 1997). We used a modified geometric distribution that allows the ratio of insects with one link to insects with two or more links to be higher compared to a pure geometric distribution (see details in Methods, Analysis and modelling). Once in the network, pollinators can keep the same number of links - which is the most frequent situation; 65% (resp. 75%) of the cases in 1996 (resp. 1997) - get one or more additional links or loose one or more links each day. For the addition of links we fit a geometric distribution and for the loss of links a binomial distribution. The loss of links is more pronounced at the end of the season and we allowed a sigmoid form of the parameter again. All the computed parameters, and results of the tests for the chosen distributions are gathered in Table 2 , Figure 3 (year 1996) and Additional file 5 (year 1997). Figure 3 Empirical and fitted distributions, 1996 . Dynamic features of the 1996 network and the associated distributions. a) Number of pollinators entering the network each day fitted to a Poisson distribution, b) Number of pollinators leaving the network each day fitted to a binomial distribution with sigmoid-shaped parameter, c) Number of links assigned to pollinators when they enter the network. Here fitted to a modified geometric distribution, d) Number of links added or removed each day from pollinators in the network. The model is a geometric distribution for the added links and a binomial distribution with a sigmoid-shaped parameter for the removed ones. See Additional file 5 for 1997. For the arrival of new pollinators and new links the fit is not improved significantly if a seasonal change in the parameters are allowed (results not shown) and there is not a visual indication of seasonal change. For the death of insects and removal of links the change in parameter happens around the same time (the parameters T and t are very similar, Table 2 ) and a model assuming T = t fits the data (results not shown). We also note that the change in parameter values over the season is very sharp and appears to happen within a day or two (Figure 2 ). Stability in the parameters over the two years An important observation is that the estimated model parameters are similar for the two years, which indicates stability in the network dynamics. This hypothesis was validated for the arrival and leave of pollinators and for the number of links when they enter the network (Table 2 ). However the evolution of the number of links (addition or removal of links for pollinators while in the network) cannot be described by joined parameters. Despite of this, the parameters are still of the same order of magnitude. (In the remaining of the paper, the 'joined model' refers to a model which uses joined parameters whenever possible.) While the plant community is the same for the two years, one fifth of the pollinator species and two thirds of all links are only observed in one of the two years; see [ 6 ] and Additional file 3 . It is therefore interesting to see that the fitted distributions for the evolution of the number of pollinators and links through the season are identical for the two years with similar parameters. For other networks it has already been reported that overall network properties such as connectance or nestedness are conserved over the years despite the turnover in pollinator species and links [ 3 , 18 ]. Our findings reinforce this observation at the more detailed level, the level of the dynamical assembly and disassembly. Simulations Based on the above results we implemented a computer program reproducing the arrival/leaving of pollinators and the evolution of their number of links with plants during their stay in the network (see Methods, Simulation). Addition of links is assumed to be independent of the pollinator's present linkage level. Also each link is removed with a probability independent of the linkage level. When pollinators are removed from the network they are chosen with probability inversely proportional to their linkage level. This provides a better fit to the observed networks than removing pollinators randomly and might reflect sampling properties; i.e. a pollinator with many links might have higher chance of being observed over consequetive days than a pollinator with few links. We simulated 50 networks using the estimated parameters and compared the results to the observed data. Table 3 shows the simulated results and the comparisons; the observed data fits nicely to the model in that the observed data are within the 95%-confidence limits (the mean ± two times the standard deviation) obtained by simulation (Table 3 ). Table 3 Validation of the model 1996 1997 Maximum number of links 163 (29.5) 200 164 (35.8) 190 Total number of insects 57.7 (6.80) 61 64.6 (9.36) 64 Total number of interactions 277 (36.0) 286 266 (48.4) 268 Maximum number of insects 30.8 (5.29) 30 34.0 (6.00) 39 Connectance 0.155 (0.012) 0.15 0.133 (0.014) 0.14 Pollinator average linkage level 4.80 (0.381) 4.7 4.11 (0.421) 4.2 Distribution of Number of links when at maximum 35 48 Phenophase 50 38 Plants per pollinator 44 44 For each year and summary statistic, the mean and standard deviation (in parenthesis) are shown, based on 50 simulations. For each summary statistic, the second row shows the observed value. We used the parameters of the joined model. Connectance is defined as the number of observed links divided by the total number of potential links between all species of plants and animals. Bottom three rows: For each simulation it is tested whether the simulated distribution is similar to the observed distribution using Kolmogorov-Smirnov's test. Shown is the number of times (out of 50) the test gives a p-value larger than 0.05. Consequences of the model The model assumes a Poisson number of arriving insects per day. This is compatible with a scenario where there are N available species and each insect has the same probability to visit and be observed at the study site; λ = Np i / G i where p i is the probability in year i and G i the number of good days ( λ is assumed to be the same in both years). Under this scenario we expect N 1 N 2 / N = N 12 insects to be observed both years. Here N i is the number of insects in year i and N 12 the number of insects observed in both years. We find that N = N 1 N 2 / N 12 = 61·64/49 = 79.7 pollinators (see Methods, Data sets), hence p 1 = N 1 / N = 76.6% of the available pollinator species are observed in 1996, while p 2 = 80.3% are observed in 1997. Solving λ = Np i / G i , gives p 1 = 72.9% and p 2 = 79.0% which are close to 76.6% and 80.3%, respectively, but derived through the Poisson model. The per day probability is independent of the year (since λ is the same), p i / G i = λ / N = 0.03. If we assume that the parameters are fixed at the values they attain in the beginning of the season and that the season in principle goes on for ever, the network will eventually reach an equilibrium where the average pollinator phenophase approaches a constant level. Using 1996 parameters, the equilibrium phenophase is approximately 17.2 days, because a pollinator stays a geometric (with probability α = 0.058, Table 2 ) number of days. However the equilibrium level is not reached for season lengths observed in the Arctic (Table 4 ). Also the average number of pollinators in the network eventually reaches a constant level which is balanced by the arrival of new insects and the departure of insects already in the network. Our simulation shows that the real network is far from equilibrium and in the process of being built up when the collapse of the network appears (Table 4 ). Table 4 Prolonging length of the season Length of the season Maximum number of links Total number of insects Average phenophase Total number of links Maximum number of insects Pollinator average linkage level 24 200 61 7.89 286 30 4.7 30 230 (42.3) 74.1 (8.49) 8.72 (7.85) 385 (55.9) 37.0 (6.25) 5.19 (0.40) 40 312 (47.0) 95.3 (9.09) 10.7 (10.5) 552 (60.6) 41.7 (6.18) 5.79 (0.36) 60 475 (64.7) 147 (11.9) 13.6 (15) 992 (101) 49.1 (6.85) 6.73 (0.37) 80 625 (77.7) 193 (13.7) 15.9 (18.8) 1448 (138) 54.5 (6.12) 7.50 (0.40) Shown are mean values with standard deviations in parenthesis based on 50 simulated networks. We used 1996 joined parameters (see Table 2). The length of the season is the number of good days only and the first row shows the observed data for the 1996 network. The 6th column shows the maximum number of insects present in a single day. The model stipulates randomness in the development of the network. The linkage level of a pollinator is described by a sum of geometric (new links) and binomial (removal of links) variables. As shown in [ 6 ] the linkage level is far from a power-law and closer to an exponential (or geometric) distribution. Also a truncated power-law distribution (a power law restricted to a certain range defined by a cut-off) is fitted to the data in [ 6 ], but the cut-off is here difficult to reconcile with biological interpretation (see [ 19 ] for a discussion of power-laws in biology and [ 20 ] for a discussion and review of cut-offs/characteristic scales in ecology)."
} | 6,051 |
32970846 | PMC7816256 | pmc | 3,456 | {
"abstract": "Abstract The mechanisms causing invasive species impact are rarely empirically tested, limiting our ability to understand and predict subsequent changes in invaded plant communities. Invader disruption of native mutualistic interactions is a mechanism expected to have negative effects on native plant species. Specifically, disruption of native plant‐fungal mutualisms may provide non‐mycorrhizal plant invaders an advantage over mycorrhizal native plants. Invasive Alliaria petiolata (garlic mustard) produces secondary chemicals toxic to soil microorganisms including mycorrhizal fungi, and is known to induce physiological stress and reduce population growth rates of native forest understory plant species. Here, we report on a 11‐yr manipulative field experiment in replicated forest plots testing if the effects of removal of garlic mustard on the plant community support the mutualism disruption hypothesis within the entire understory herbaceous community. We compare community responses for two functional groups: the mycorrhizal vs. the non‐mycorrhizal plant communities. Our results show that garlic mustard weeding alters the community composition, decreases community evenness, and increases the abundance of understory herbs that associate with mycorrhizal fungi. Conversely, garlic mustard has no significant effects on the non‐mycorrhizal plant community. Consistent with the mutualism disruption hypothesis, our results demonstrate that allelochemical producing invaders modify the plant community by disproportionately impacting mycorrhizal plant species. We also demonstrate the importance of incorporating causal mechanisms of biological invasion to elucidate patterns and predict community‐level responses.",
"introduction": "Introduction Biological invasions threaten biodiversity, with broad impacts on the economy, environment, health, and culture. Global meta‐analyses demonstrate that invasive species can have a range of negative impacts that include lower abundance and fitness of native species and altered microbial activity and nutrient availability in invaded sites (Vilà et al. 2011 , Pyšek et al. 2012 ). Despite these observed negative consequences, we are still largely unable to predict the potential outcomes of an introduced species prior to its establishment and spread (Ricciardi et al. 2013 , Sofaer et al. 2018 ). Since the impacts of invasion depend on the interaction of traits of individual invasive species (Gurevitch and Padilla 2004 ) and traits of the invaded community (Hejda et al. 2009 ), it is unlikely that a single mechanism will fully explain the impacts of invasion. However, understanding causal mechanisms by which certain types of invasive species exert impacts on species in the resident community can lead to more accurate predictions of where and when negative effects will occur in the invaded range (Levine et al. 2003 ). Many studies have analyzed spatiotemporal patterns of community change to assess the impacts of an invasive species, but knowledge of the mechanism of impact is indispensable for generalizing individual or species‐level effects to those expected at the community level. Studies of community‐level patterns show that invasions can lead to biotic homogenization across locations (Sax et al. 2002 ) but rarely reduce local species richness within one location (Powell et al. 2011 ). In addition, invasive species alter community evenness by disproportionately displacing dominant native species (Powell et al. 2013 , Pearse et al. 2019 ). Yet such patterns do not generally predict which species within the resident community will be most vulnerable to the invader, and thereby reliably link invader impacts at the individual, population, and community levels. In contrast, by understanding mechanisms of impact, we may be able to predict responses across levels of biological organization. For example, by understanding how herbivorous insects identify their host plants, we can anticipate the species‐specific impacts of invasive plants or insects on other trophic levels in the native food web (Pearse and Altermatt 2013 , Desurmont and Pearse 2014 ). Moreover, evolutionary divergence time between native range host tree and invaded range host trees is a strong predictor of high impacts of introduced insects on forests (Mech et al. 2019 ). For trophic interactions (such as herbivory, above), the mechanism of impact is a function of diet, and can be relatively straightforward and consistent across space and time. The indirect or competitive interactions that are typical of plant invaders are particularly difficult to detect and predict (Suding et al. 2008 ), and understanding mechanisms of invasion involving these types of interactions typically requires manipulative studies. Many processes can underlie invader impact (Ricciardi et al. 2013 ), but the mutualism disruption hypothesis (Hale and Kalisz 2012 ) is particularly well suited to evaluating invader impact from individual species to the community because we can make predictions about which native species within a community will be affected based on their known mutualism dependence. Under this hypothesis, an invasive species can gain a competitive advantage over mutualism‐dependent native species by reducing or eliminating the benefits of these mutualisms (Hale and Kalisz 2012 , Traveset and Richardson 2014 ). The effects of mutualism disruption have been observed in a variety of invaded systems. Compelling examples include the disruption of ant seed dispersal in the presence of the Argentine ant ( Linepithema humile ) that resulted in a shift of the resident plant community based on seed and elaiosome size (Christian 2001 , Rowles and Dowd 2009 ), the predation of seed dispersing birds to near extinction by the brown tree snake ( Boiga irregularis ) that severely reduced native tree recruitment on Guam (Rogers et al. 2017 ), and the usurpation of native pollination services by the showy flowered purple loosestrife ( Lythrum salicaria ) that led to a steep decline in pollinator visitation to native flowers (Brown et al. 2002 ). Mutualism disruption is widespread, and can negatively affect dispersal, reproductive, or nutritive mutualisms. The global threat of mutualism disruption is particularly substantial for plant–mycorrhizal mutualisms because of their ubiquity (Soudzilovskaia et al. 2020 ) and their importance for plant growth, function, and community dynamics (Van Der Heijden et al. 2008 ). Invasive plants can impact plant–mycorrhizal mutualisms in various ways (Grove et al. 2017 ): by directly competing with native plants, which weakens the native plant’s contribution to their mycorrhizal partnership (Grove et al. 2017 ), by changing soil properties to alter the balance of resource exchange between plant and fungal partners (Allison et al. 2006 , Johnson 2010 ), or by directly inhibiting the mutualism via invader‐produced toxic secondary chemistry (allelochemicals) (Hale and Kalisz 2012 ). The specific mechanism of mycorrhizal mutualism disruption may vary among invasive species but toxic allelochemistry is common. Using published lists of invasive plant families (Pyšek et al. 2012 ) and plant families with known allelopathic properties (Hale and Kalisz 2012 ), we estimate that 41% of invasive plant families are allelopathic ( unpublished data ). Of these allelopathic plant invaders, Alliaria petiolata (Brassicaceae, Bieb. Cavara & Grande, hereafter garlic mustard) is emerging as a model for understanding allelopathic mycorrhizal mutualism disruption (Cipollini and Cipollini 2016 ). Experimental evidence in both controlled and field settings has established that garlic mustard reduces native plant performance and population growth through the disruption of their mycorrhizal associations (e.g., Stinson et al. 2006 , Callaway et al. 2008 , Wolfe et al. 2008 , Barto et al. 2011 , Hale et al. 2011 , 2016 , Brouwer et al. 2015 , Bialic‐Murphy et al. 2019 ). Furthermore, garlic mustard is a poor direct competitor (Meekins and McCarthy 1999 , Bossdorf et al. 2004 ), and has no known direct phytotoxic effects (Brouwer et al. 2015 , Hale et al. 2016 ). In addition, there is evidence that an increase in garlic mustard abundance is associated with lower native plant diversity (Stinson et al. 2007 ), and that community‐level responses to garlic mustard vary by species and by abiotic conditions (Haines et al. 2018 ). However, it is unknown whether the physiological consequences of mutualism disruption for individual plant species, mediated by declines in function of plant–mycorrhizal fungal partnerships, are linked to the shifts in plant community diversity and composition following garlic mustard invasion (Stinson et al. 2007 , Callaway et al. 2008 , Cipollini and Cipollini 2016 , McCary et al. 2019 ). Here, we exploit the power of this well‐established mechanism of invasion impact to test for a link between mutualism disruption of mycorrhizal plant individuals and community‐level response of the forest understory to garlic mustard invasion. A large portion of the diversity in most Eastern deciduous forests exists in the understory, and these species are also diverse in their mycorrhizal associations (Gilliam 2007 , Soudzilovskaia et al. 2020 ), making the forest understory community a highly relevant study system. We present analyses of data collected from a 11‐yr garlic mustard weeding experiment in replicated plots to assess whether garlic mustard differentially impacts native mycorrhizal vs. non‐mycorrhizal herbaceous understory plant species. We ask two questions: (1) Does the release and recovery from garlic mustard invasion (i.e., garlic mustard weeded treatment) differentially affect the community composition and diversity of mycorrhizal and non‐mycorrhizal herbaceous understory plant communities? (2) Are changes in community composition and diversity explained by shifts in abundance of mycorrhizal and non‐mycorrhizal plant species? Based on the mutualism disruption hypothesis, we expect a greater change in the community composition, diversity, and abundance of mycorrhizal plants (i.e., the functional groups most sensitive to garlic mustard’s allelochemicals) than non‐mycorrhizal plants, culminating in a distinct signature of mutualism disruption at the community level.",
"discussion": "Discussion Numerous studies demonstrate that invasive plants have negative impacts on native species (Vilà et al. 2011 , Pyšek et al. 2012 ). However, to anticipate how resident plant communities shift and which native species are most likely to be impacted by an invader, we must explore the mechanisms by which invasive plants displace natives (Levine et al. 2003 , Sofaer et al. 2018 ). Using detailed data from a 11‐yr manipulative field experiment, we found that the removal of garlic mustard (i.e., weeded treatment) altered the composition, decreased the diversity, and increased the abundance of the mycorrhizal plant community in comparison to ambient levels of garlic mustard invasion. In contrast, there were no effects of the garlic mustard treatments on the non‐mycorrhizal plant community, which is not susceptible to garlic mustard’s disruption of mycorrhizal mutualisms. Our study illustrates that the consequences of garlic mustard invasion for the plant community may be anticipated by incorporating an understanding of mutualism disruption into future invasion predictions, because garlic mustard weeding favored mycorrhizal, but not non‐mycorrhizal plant species. Our results are consistent with previous work that found variation in response to garlic mustard by mycorrhizal dependency of forest tree seedlings (Stinson et al. 2006 ). Together, our findings and previous work demonstrate that garlic mustard invasion results in a functional shift in plant community membership that can be reversed over time with removal of the invader. The shifts in the composition, diversity, and abundance of the mycorrhizal, but not the non‐mycorrhizal, functional group in response to garlic mustard weeding are indicative of mutualism disruption. Other mechanisms of impact, such as direct competition or direct phytotoxicity, would not disproportionately affect species that form mycorrhizal associations. Garlic mustard is a poor direct competitor in its invasive range (Meekins and McCarthy 1999 , Bossdorf et al. 2004 ), and disrupts mycorrhizal mutualisms to indirectly compete against other plants (Brouwer et al. 2015 , Hale et al. 2016 ). Removal of a competitive dominant plant species from a community is expected to result in an increase in abundance of other community members (MacDougall and Turkington 2005 ), but unlike mutualism disruption, exploitative competition would not explicitly predict different responses for mycorrhizal and non‐mycorrhizal plants to the removal of a competitor. Since mycorrhizal and non‐mycorrhizal plants differ in their methods of resource acquisition (via roots alone or via fungal partners), they may respond differently to limiting resources under exploitative competition. However, soil resources do not differ between garlic mustard ambient and weeded treatments at our site (Bialic‐Murphy, L. et al. unpublished manuscript ) and are, therefore, unlikely to cause the trends we observed here. Observed patterns of the plant community response to garlic mustard in our field site are consistent with those predicted by the mechanistic process of mutualism disruption. Previous work has quantified responses of root fungal communities to garlic mustard and provides context for interpreting our results. At our study site, root fungal communities within some species’ roots changed in response to our garlic mustard treatments (Burke 2008 ), particularly in the presence of deer (Burke et al. 2019 ). Ongoing work shows that species in the ambient garlic mustard treatments have reduced hyphal colonization in their roots, and higher evenness in the soil AM fungal community (Bialic‐Murphy, L. et al. unpublished manuscript ). In other locations, studies have documented no recovery of mycorrhizal fungi after three years (Anthony et al. 2019 ) or only partial recovery after six years (Lankau et al. 2014 ). Since we did not observe a response in the aboveground community until 4–6 yr of garlic mustard weeding, it is possible that the aboveground response is mirroring the recovery of the belowground mycorrhizal fungal community, or that the mycorrhizal fungal community has not yet responded to garlic mustard removal. In addition, it is possible that changes in the abiotic soil conditions that are associated with garlic mustard invasion attribute to the observed response of mycorrhizal plant species (Rodgers et al. 2008 , Anthony et al. 2019 ). However, no differences in abiotic soil conditions between ambient and weeded garlic mustard treatments after 14 yr were found by Bialic‐Murphy, L. et al. ( unpublished manuscript ). Thus, findings from our experiment provide a link between mechanistic evidence of mutualism disruption from careful fine‐scale experiments (e.g., Stinson et al. 2006 , Hale et al. 2011 , 2016 , Brouwer et al. 2015 ) to observed shifts in mycorrhizal plant communities. Our findings support the utility of our coarse classification into mycorrhizal and non‐mycorrhizal functional groups for anticipating responses to invasions but also highlight limitations in the degree of predictability within ecological communities. In particular, within the mycorrhizal group there was considerable variation in response to garlic mustard (RII) (Fig. 5 ), such that a simple binary assignment of mycorrhizal status would be insufficient to reliably predict changes in species‐level abundance. Plant species’ dependency on mycorrhizal partners (i.e., biomass response to mycorrhizal colonization) generally falls on a spectrum from highly positive plant biomass responses to mycorrhizal colonization to somewhat negative (Tawaraya 2003 ), and variation in dependence of mycorrhizal plants on their fungal partners is known to influence the response of a particular species to mutualism disruption (Stinson et al. 2006 , McCary et al. 2019 ). We do not know where our naturally occurring plant species fall along the spectrum of dependency on their mycorrhizal partners, and it is possible that differences among trajectories of the experimental plots in response to garlic mustard weeding (Fig. 2a ) arise from variation in the dependency of the mycorrhizal species naturally occurring in those plots. Exploring the factors that may have driven the different trajectories of the unique plant communities in each plot is an important next step in understanding the intricacies of community responses to garlic mustard invasion. Our results revealed that the mycorrhizal plant community had decreased evenness and diversity in the garlic mustard weeded treatments. The decreases in Shannon and inverse Simpson diversity of the mycorrhizal plant community are likely driven by the lower evenness in the garlic mustard weeded treatments, since there was no change in species richness for this group. Evenness of the mycorrhizal plant community decreased in the garlic mustard weeded plots because of the dramatic increase in the abundance of a common annual species, Impatiens spp., after 2006 (Appendix S1 : Fig. S2). This created a less balanced community with low evenness where one species ( Impatiens spp.) has a very high abundance whereas the rest of the species remain at a low abundance. Since Impatiens spp . have an annual life history and often exhibit ruderal growth patterns in comparison to many of the long‐lived perennials at our site, we expect that the long‐lived perennial species will rebound as well with more time. Over the course of this study, we found that the main effect of time on the abundance of both plant communities was significant, but the interactions between the garlic mustard treatment and year were not. The non‐mycorrhizal plant abundance decreased over time in both garlic mustard treatments. Therefore, this decline in non‐mycorrhizal plant abundance cannot be attributable only to the increase in mycorrhizal plant abundance in the garlic mustard weeded treatment. We see similar declines in non‐mycorrhizal plant abundance in other plots at our site that are not protected from deer herbivory by fences ( unpublished data ), indicating that exclusion of deer is also not driving the declines in non‐mycorrhizal plant abundance. These results suggest that there are other sources of environmental variability (e.g., annual fluctuations in precipitation patterns) that likely influenced the abundance of both plant community types over time. Despite this, there is a strong signature of the effects of garlic mustard weeding on mycorrhizal plant abundance. Garlic mustard consistently has negative effects on native mycorrhizal plant species (e.g., Stinson et al. 2006 , Cipollini et al. 2008 , Wolfe et al. 2008 ). These negative effects on mycorrhizal plant species are consistent with garlic mustard’s disruption of native mycorrhizal mutualisms, which leads to water stress and carbon stress (Brouwer et al. 2015 , Hale et al. 2016 ) and decreases in vital rates (Brouwer et al. 2015 ) and population growth rates (Bialic‐Murphy et al. 2019 ). In addition, garlic mustard alters plant communities (Stinson et al. 2007 ). Our results build on these findings to show that the outcomes symptomatic of mutualism disruption in individual plants follow a similar pattern for plant communities in the field. Our 11‐yr long‐term field experiment shows that, even at the modest garlic mustard abundances at Trillium Trail, mutualism disruption alters the community composition, changes diversity, and reduces the abundance of only the mycorrhizal plants within the entire plant community. We stress that the length of our experiment provides strong support for the long‐term effects of garlic mustard on the mycorrhizal plant community. Most studies that evaluate the impacts of invasive species last only 1 yr (Stricker et al. 2015 ) even though the effects of invaders are known to lag after the invasion establishes and have legacy effects after they are removed. Land managers attempting to restore native understories through garlic mustard removal may expect to see an increase in abundance of mycorrhizal plant species, but likely no changes in species richness. However, responses may differ by species’ sensitivity to garlic mustard (see Fig. 5 and Appendix S1 : Fig. S3), and may require multiple years (in our experimental study 4–6 yr) of annual garlic mustard removal to occur. Many invasive plant species are thought to impact native species by allelopathy and/or mutualism disruption (e.g., Alliaria petiolata, Tamarisk spp., various Fallopia species; Stinson et al. 2006 , Wolfe et al. 2008 , Bainard et al. 2009 , Murrell et al. 2011 , Hale and Kalisz 2012 , Meinhardt and Gehring 2012 ), so our findings likely apply to the impacts of invasive plants beyond garlic mustard. In addition, since approximately 80% of understory herbs globally (and two‐thirds of the study species in our experimental plots; Appendix S1 : Table S1) form mycorrhizal associations (Soudzilovskaia et al. 2020 ), there is potential for garlic mustard to drastically shift understory plant communities across its invasive range. We demonstrate that mycorrhizal mutualism disruption by an invasive plant leads to a shift in plant communities but found that responses were variable among species within functional groups. This suggests that the ability to predict invader impacts across space and time may require refinement of the focal functional trait, in this case a shift from a binary to a continuous view of mycorrhizal dependency. Our study highlights the utility of exploring the mechanisms by which invasive species cause impact because we find a shift in functional groups of the resident plant community that would only be anticipated by knowing that mechanism."
} | 5,527 |
19775247 | null | s2 | 3,458 | {
"abstract": "Cannibalism is a mechanism to delay sporulation in Bacillus subtilis. Cannibal cells express the skf and sdp toxin systems to lyse a fraction of their sensitive siblings. The lysed cells release nutrients that serve to feed the community, effectively delaying spore formation. Here we provide evidence that the subpopulation of cells that differentiates into cannibals is the same subpopulation that produces the extracellular matrix that holds cells together in biofilms. Cannibalism and matrix formation are both triggered in response to the signalling molecule surfactin. Nutrients released by the cannibalized cells are preferentially used by matrix-producing cells, as they are the only cells expressing resistance to the Skf and Sdp toxins. As a result this subpopulation increases in number and matrix production is enhanced when cannibalism toxins are produced. The cannibal/matrix-producing subpopulation is also generated in response to antimicrobials produced by other microorganisms and may thus constitute a defense mechanism to protect B. subtilis from the action of antibiotics in natural settings."
} | 278 |
29230370 | PMC5721908 | pmc | 3,459 | {
"abstract": "Sugarcane bagasse is an abundant source of lignocellulosic material for bioethanol production. Utilisation of bagasse for biofuel production would be environmentally and economically beneficial, but the recalcitrance of lignin continues to provide a challenge. Further understanding of lignin production in specific cultivars will provide a basis for modification of genomes for the production of phenotypes with improved processing characteristics. Here we evaluated the expression profile of lignin biosynthetic genes and the cell wall composition along a developmental gradient in KQ228 sugarcane. The expression levels of nine lignin biosynthesis genes were quantified in five stem sections of increasing maturity and in root tissue. Two distinct expression patterns were seen. The first saw highest gene expression in the youngest tissue, with expression decreasing as tissue matured. The second pattern saw little to no change in transcription levels across the developmental gradient. Cell wall compositional analysis of the stem sections showed total lignin content to be significantly higher in more mature tissue than in the youngest section assessed. There were no changes in structural carbohydrates across developmental sections. These gene expression and cell wall compositional patterns can be used, along with other work in grasses, to inform biotechnological approaches to crop improvement for lignocellulosic biofuel production.",
"introduction": "Introduction Sugarcane is a C4 perennial grass of high economic importance in many parts of the world ( Suprasanna et al., 2011 ). In addition to the production of high levels of sucrose in the stem, it produces large amounts of lignocellulosic biomass that has the potential to be used for the production of bioethanol ( Canilha et al., 2012 ). Sugarcane is a particularly attractive source of biomass for lignocellulosic biofuels production, as it is already transported to a central location for the production of sugar. As a result, the bagasse remaining following sugar production is does not require transport costs that otherwise represent a significant cost in bioethanol production. However, as with all potential lignocellulosic feedstocks, the recalcitrance of the biomass presents challenges that need to be addressed. The deposition of the secondary cell wall is an important step in terrestrial plant development ( Weng & Chapple, 2010 ), involving the ordered deposition of cellulose and hemicellulose followed by the impregnation of lignin polymers into this polysaccharide matrix ( Vogel, 2008 ). Lignin polymers are comprised of guaiacyl (G), syringyl (S) and p -hydroxyl-phenyl (H) units, through oxidative polymerization of coniferyl, sinapyl and p -coumaryl alcohols respectively, that are produced through the lignin biosynthesis pathway ( Boerjan, Ralph & Baucher, 2003 ; Liu, 2012 ). Due to the importance of lignin in structural stability and water transportation, the role and function of each gene within the lignin biosynthesis pathway is well established ( Boerjan, Ralph & Baucher, 2003 ; Bonawitz & Chapple, 2010 ). The relationship between lignin and efficiency of lignocellulosic bioethanol production has led to increased focus into lignin biosynthesis and manipulation, and advances the possibility of cost-competitive bioethanol being produced from lignin-altered sugarcane bagasse. Given the influence lignin has on cell wall digestibility, further understanding of control and timing of lignin deposition will be applicable for the genetic modification of plants to specifically alter lignin characteristics. While a number of studies have looked at cell wall formation in sugarcane previously ( Bottcher et al., 2013 ; De Souza et al., 2013 ; Lingle & Thomson, 2012 ), here we aim to assess the expression profile of lignin biosynthetic genes and cell wall composition of a commercially relevant Australian sugarcane cultivar (KQ228). In particular, we look at a developmental gradient to further understand the relationship between gene expression and cell wall formation and composition, with the goal of providing critical information for the biotechnological development of improved varieties of sugarcane for second generation biofuel production.",
"discussion": "Discussion Sugarcane bagasse has great potential as a lignocellulosic biofuels source, in part, due to its already being moved to a centralized location for sugar production. In order to effectively produce fermentable sugars from bagasse, the challenge of cell wall recalcitrance needs to be overcome. Improved understanding of lignin biosynthesis and deposition in sugarcane will be of great value when deciding the most appropriate approaches to facilitate the development of commercial lines with increased saccharification potential. The work herein uses an economically important smut-resistant Australian sugarcane cultivar, KQ228, and assesses lignin biosynthetic gene expression and cell wall composition along a developmental gradient in an attempt to further characterize the timing and location of lignin deposition to guide attempts to improve bagasse for lignocellulosic biofuels production. It has been shown that disease resistance and lignin content are often related ( Cass et al., 2015 ; Yang et al., 2017 ), making this an attractive cultivar for this work. Whereas in our study we look only at one specific homologue for each lignin biosynthetic gene, we have provided a comparison to previous work ( Bottcher et al., 2013 ) wherein multiple homologues are assessed ( Table 1 ). Our work was based on available genome data at the time of the study, and despite being somewhat limited relative to Bottcher et al. (2013) , provides confirmation and comparison with another economically important cultivar. In addition to this, we also examined lignin gene expression in buttress roots as this can be a key storage sink for carbon. The trends in the stem expression data dichotomize the lignin biosynthesis genes. The two ‘expression pattern groups’ are genes for which expression decreases with tissue age ( PAL , CCR , 4CL , COMT and CAD ) or genes for which expression remains constant during maturation ( C3H , F5H , C4H and CCoAOMT ). As PAL catalyzes the entry of metabolites into the lignin biosynthesis pathway ( Liu, 2012 ; Weng & Chapple, 2010 ), its high level of expression in younger tissue found in this study may represent an initial metabolic flux to provide a burst of metabolites for the various phenylpropanoid pathways including lignin biosynthesis. CCR functions in the final stages of lignin biosynthesis and is considered a committed step, key in the production of the individual lignin monomers ( Vogt, 2010 ; Weng & Chapple, 2010 ). Given the position of CCR in the lignin biosynthesis pathway, it may act as a regulating control point for directing the metabolic flux into lignin monomer production ( Lacombe et al., 1997 ). As high expression of PAL in young tissue may act to stimulate metabolic flux into phenylpropanoid production, high expression of CCR in young tissue may ensure a high level of metabolite commitment into lignin biosynthesis, which is fundamentally important for healthy plant development ( Weng & Chapple, 2010 ). The expression profiles of 4CL and COMT are similar to that of PAL and CCR , but they retain slightly higher expression levels in more mature tissue. 4CL represents an important branch where metabolites are directed either into lignin biosynthesis or to alternative phenylpropanoid biosynthesis pathways ( Vogt, 2010 ; Weng & Chapple, 2010 ). Its position also allows for direct metabolite contribution into H monomer biosynthesis or redirection of metabolites for G or S monomer biosynthesis. The high level of 4CL expression in young tissue may reflect its response to the metabolic flux into the phenylpropanoid pathway initiated by PAL . COMT is the last of two enzymes entirely responsible for the production of the S lignin monomer within the lignin biosynthesis pathway ( Bonawitz & Chapple, 2010 ). The increased expression of COMT in young tissue in this research may be to ensure S monomer production during the availability of the initial metabolic flux. Previous work has shown that the RNAi suppression of COMT in sugarcane resulted in decreased lignin and altered S:G ratio ( Bewg et al., 2016 ). The final gene showing a reduction in expression as stem tissue matures was CAD , though the trend was not as strong as the previously discussed genes. CAD represents the final enzyme in the lignin biosynthesis pathway catalyzing the production of precursor monolignols and committing them to lignin monomer synthesis ( Ferrer et al., 2008 ). The initial high expression of CAD in young tissue may relate to the increased metabolic flux through the lignin biosynthesis pathway. Whereas overall trends between our work and previous research are the same for these genes with decreasing expression for increasing maturity, there are differences, particularly with CAD and COMT . The discrepancies between the current and published research may be a result of various experimental differences between the current research and published findings, but it is more likely that the differences arise from the differences in cultivars. Three genes were identified with relatively consistent expression across the maturity gradient: C4H , F5H , and CCoAOMT . Results for C4H were consistent with other results ( Papini-Terzi et al., 2009 ). For F5H , expression in the Brazilian low and high lignin cultivars was highest in intermediate aged internodes with the exception of the high lignin pith samples wherein it was highest in the mature tissue ( Bottcher et al., 2013 ). In the 30 cultivars with varying Brix levels, F5H expression levels were higher in maturing stem tissue than in young tissue ( Papini-Terzi et al., 2009 ). Our results for CCoAOMT closely mirrored the results of Bottcher et al. (2013) . We have previously published work describing the downregulation of both the F5H and CCoAOMT in sugarcane, with the result being increased glucose release by enzymatic hydrolysis but with no decrease in lignin. In the F5H lines this was attributed to a change in the lignin monomer ratio ( Bewg et al., 2016 ) The final gene assessed in our study, C3H , had the highest expression level in Section B, immediately below the most juvenile Section A. C3H catalyzes the second aromatic hydroxylation reaction in the lignin biosynthesis pathway and is an important hub in controlling metabolic flux into G and S lignin monomer synthesis ( Barriere et al., 2004 ; Weng & Chapple, 2010 ). CCoAOMT , along with C3H , is hypothesized to be important control points for cell wall lignification by acting as part of the ferulate production pathway ( Barriere et al., 2004 ). CCoAOMT is responsible for the 3′ methylation of caffeoyl-CoA to produce feruloyl-CoA, a key step in the production of G and S lignin monomers ( Hisano, Nandakumar & Wang, 2009 ; Raes et al., 2003 ). The feruloyl residues aid in cross-linking within the cell wall and may increase the resistance of the cell wall to hydrolysis by adding to its structural stability ( Barriere et al., 2004 ; Bonawitz & Chapple, 2010 ; Grabber, 2005 ). The relatively steady expression of CCoAOMT and C3H within the maturing sugarcane stem may reflect their continued requirement for feruloyl residue production for ongoing cell wall lignification and not just their role in lignin monomer biosynthesis. Hydroxycinnamoyl transferase ( HCT ) was not included in qRT-PCR analysis as at the time of this study a specific sequence could not be confidently identified. At that time, only one published accession for sugarcane HCT was found ( CA210265 ) ( Casu et al., 2007 ). When analyzed by BLAST it showed very close alignment with Zea mays anthranilate N-benzoyltransferase ( NM_001153992 ) ( Soderlund et al., 2009 ). Further BLAST searching in the sugarcane nucleotide and EST databases of NCBI with alternative HCT sequences from maize ( AY109546 , DR807341 ) ( Barrière et al., 2007 ) and from MAIZEWALL (2478084.2.1_REV, 2619423.2.1) ( Guillaumie et al., 2007 ), Medicago sativa L. ( AJ507825 ) ( Shadle et al., 2007 ), Nicotiana benthamiana ( AJ555865 ) ( Hoffmann et al., 2004 ), Coffea arabica ( AM116757 ) ( Salmona et al., 2008 ) and Triticum aestivum L. ( CK193498 , CK199765 ) ( Bi et al., 2011 ) did not highlight any potential sugarcane HCT sequences, nor any conserved regions of sufficient length to design primers (standard or degenerate) for potential use in sugarcane. To our knowledge, this is the first paper that has looked at lignin biosynthetic gene expression in sugarcane buttress roots. There were no significant differences in expression levels between root tissue and the five stem sections (A–E) for C3H , CCoAOMT , F5H and CAD . Of these, C3H , CCoAOMT and F5H are all in the group with plateaued gene expression during development and may highlight the promoters of these three genes as potential biotechnological tools to drive continuous and even expression of transgenes in sugarcane stem and root tissue. The only gene with an unexpected level of expression was C4H that had approximately 9-fold higher expression in roots than in any stem section. This suggests that the C4H promoter may be useful for preferential expression of transgenes in sugarcane root tissue, however further analysis, including the functionality of this promoter in additional tissue types, such as leaves, would need to be assessed. In addition to the assessment of lignin biosynthetic gene expression, we also examined the cell wall composition along the same developmental gradient. It is well known that the composition of the cell wall material changes as a plant matures due to secondary cell wall deposition. Following cell elongation, the secondary cell wall is formed through the deposition of cellulose and hemicellulose, followed by lignification ( Vogel, 2008 ; Weng & Chapple, 2010 ). Within sugarcane, rapid elongation of young internode cells precedes cell wall thickening, including lignification ( Casu et al., 2007 ). No significant differences were seen in levels of structural carbohydrates including glucose, xylose or galactose indicating that the deposition of structural polysaccharides into the secondary cell wall had also occurred before harvesting of samples (in more juvenile tissue). This is in contrast to published findings in sugarcane ( Lingle & Thomson, 2012 ) and maize ( Jung & Casler, 2006 ). In sugarcane, cellulose peaked and then declined below internode 5, whereas hemicellulose was highest in young tissue before reducing to a steady state ( Lingle & Thomson, 2012 ). In maize, glucose content increased as tissue matured before plateauing, and hemicellulose (xylose and arabinose) decreased as tissue matured before also reaching a steady state ( Jung & Casler, 2006 ). The decrease in xylose and arabinose coincided with an increase in ferulates, and the authors suggest ferulates may be replacing the xylose and arabinose within the cell wall, hence their decrease during tissue maturation ( Jung & Casler, 2006 ). Results suggest the lignin deposition was complete by Section B as lignin content plateaued and no differences were detected between Sections B through E. Other studies have also found that overall lignin content increased with tissue maturity in wheat and maize ( Jung & Casler, 2006 ; Ma, 2007 ). In maize stem, lignin content decreased initially before increasing to a plateau ( Jung & Casler, 2006 ). In sugarcane, marked internodes harvested over a period of twelve weeks had increased lignin content over time ( Lingle & Thomson, 2012 ). In a second experiment, odd numbered internodes harvested at a single time point, showed lignin content increased with maturity, with the exception of a significant decrease in internode 3 ( Lingle & Thomson, 2012 ). The results of the second experiment are similar to the maize results of Jung & Casler (2006) , who suggest that young maize tissue is comprised of a higher percentage of lignified protoxylem vessels than more mature tissue, that initially results in a high lignin content in very young tissue ( Jung & Casler, 2006 ). It is likely that the Section A tissue (from internodes 2 and 3) was in what is the second zone identified in the studies by Jung & Casler (2006) and Lingle & Thomson (2012) . This is supported by the results of Bottcher et al. (2013) who showed lower lignin levels in internode 2–4 of two sugarcane cultivars before reaching a relatively steady state lignin level for internodes 5 to 18. Only one paper has examined the lignin content of root tissue, and the authors found only small changes in lignin content in the first 5 cm of root development, with lignin levels ranging from 5–9% ( Leite et al., 2017 ). Of note was the inverse relationship seen between lignin and arabinose content. This is likely the result of the pattern of cell wall deposition, in which there is a natural progression from highly substitute arabinoxylans to less branched xylan as cells fully expand. This natural progression would also correspond with increasing lignin deposition ( Carpita, 1996 ; De O Buanafina, 2009 ). This is consistent with previous results comparing cell wall properties across three Miscanthus genotypes ( De Souza et al., 2015 ). The work presented herein provides a profile of lignin biosynthetic gene expression and cell wall composition for an economically important Australian sugarcane cultivar. The results support findings of previous groups and add additional information on gene expression in sugarcane buttress roots. As a key potential biofuels crop, detailed information from multiple cultivars will help to improve the understanding of lignin and cell wall formation in this species and to inform biotechnological approaches to crop improvement."
} | 4,504 |
38403652 | PMC10894876 | pmc | 3,460 | {
"abstract": "Scalable, high-capacity, and low-power computing architecture is the primary assurance for increasingly manifold and large-scale machine learning tasks. Traditional electronic artificial agents by conventional power-hungry processors have faced the issues of energy and scaling walls, hindering them from the sustainable performance improvement and iterative multi-task learning. Referring to another modality of light, photonic computing has been progressively applied in high-efficient neuromorphic systems. Here, we innovate a reconfigurable lifelong-learning optical neural network (L 2 ONN), for highly-integrated tens-of-task machine intelligence with elaborated algorithm-hardware co-design. Benefiting from the inherent sparsity and parallelism in massive photonic connections, L 2 ONN learns each single task by adaptively activating sparse photonic neuron connections in the coherent light field, while incrementally acquiring expertise on various tasks by gradually enlarging the activation. The multi-task optical features are parallelly processed by multi-spectrum representations allocated with different wavelengths. Extensive evaluations on free-space and on-chip architectures confirm that for the first time, L 2 ONN avoided the catastrophic forgetting issue of photonic computing, owning versatile skills on challenging tens-of-tasks (vision classification, voice recognition, medical diagnosis, etc.) with a single model. Particularly, L 2 ONN achieves more than an order of magnitude higher efficiency than the representative electronic artificial neural networks, and 14× larger capacity than existing optical neural networks while maintaining competitive performance on each individual task. The proposed photonic neuromorphic architecture points out a new form of lifelong learning scheme, permitting terminal/edge AI systems with light-speed efficiency and unprecedented scalability.",
"introduction": "Introduction Artificial intelligence (AI) tasks become increasingly abundant and complex fueled by large-scale datasets 1 – 4 . One open question in the field of machine learning is how artificial agents could propagate in a smarter manner with exceptional learning scalability and realize versatile advanced AI tasks 5 – 8 . With the plateau of Moore’s law and end of Dennard scaling, energy consumption becomes a major barrier to more widespread applications of today’s heavy electronic deep neural models 9 – 12 , especially in terminal/edge systems 13 , 14 . The community is imminently looking for next-generation computing modalities to break through the physical constraints of electronics-based implementations of artificial neural networks (ANNs). Photonic computing has been promised to overcome the inherent limitations of electronics and improve energy efficiency, processing speed and computational throughput by orders of magnitude 15 – 17 . Such extraordinary properties have been exploited to construct application-specific optical architectures 18 – 22 for solving fundamental mathematical and signal processing problems with performances far beyond those of existing electronic processors. Optical neural networks (ONNs) are constructed to validate simple visual processing tasks 23 – 26 such as hand-written digit recognition 27 – 29 and saliency detection 30 , 31 , using wave-optics simulations or small-scale photonic computing systems. Meanwhile, some works combine the photonic computing units with a variety of electronic ANNs to enhance the scale and flexibility of optical architectures, e.g., deep optics 32 – 34 , amplitude-only Fourier ONNs 31 , and hybrid optical-electronic CNN 35 . However, existing optics-based implementations are limited to a small range of applications and cannot continually acquire versatile expertise on multiple tasks to adapt to new environments. The main reason is that they inherit the widespread problem of conventional computing systems, which are prone to train new models interfering with formerly learned knowledges, rapidly forget the expertise on previously learned tasks when trained on something new, i.e., ‘catastrophic forgetting’ 36 – 40 . Such an approach fails to fully exploit the intrinsic properties in sparsity and parallelism of wave optics for photonic computing, which ultimately results in poor network capacity and scalability for multi-task learning. In contrast, humans possess the unique ability to incrementally absorb, learn and memorize knowledge. In particular, neurons and synapses perform work only when there are tasks to deal with, in which two important mechanisms participate: sparse neuron connectivity 41 – 43 and parallelly task-driven neurocognition 44 – 47 , together contribute to a lifelong memory consolidation and retrieval. Accordingly, in ONNs, these characteristic features can be naturally promoted from biological neurons to photonic neurons based on the intrinsic sparsity and parallelism properties of optical operators 31 , 48 – 51 . An optical architecture imitating the structure and function of human brains demonstrates its potential to alleviate the aforementioned issues, which shows more advantages than electronic approaches in constructing a viable lifelong learning computing system. Herein, we propose L 2 ONN: a reconfigurable photonic computing architecture for lifelong learning (Fig. 1 ). Neuromorphically inspired, L 2 ONN can incrementally learn tens-of-tasks in one model with light-speed efficient computation. We show that the unique characteristics of light, spatial sparsity and multi-spectrum parallelism that for the first time developed in photonic computing architecture, endow ONNs with lifelong learning capability. Specifically, considering the physical propagation of free-space coherent light field (Fig. 2 ): Phase change materials (PCM)-based sparse optical filters are employed to modulate photonic neuron connections of each single task; And a multi-spectrum light diffraction-based optical computing module is constructed to extract the multi-task features allocated with different wavelengths. Throughout the architecture, photonic neurons are selectively activated according to the input signals. Unlike existing ONNs trying to imitate ANN structures, the photonic lifelong learning of L 2 ONN is initially designed following the physical nature of light-matter interaction, to fully explore the functional and performance potentials of wave optics in photonic computing. Fig. 1 Principle of the photonic neuromorphic architecture. a Illustration of human lifelong learning. The brain can incrementally absorb, learn and memorize knowledge throughout its lifespan. Neurons and synapses are adaptively connected by task-driven neurocognition. b Diagram of the neuromorphic photonic lifelong learning. The photonic connections in each optical layer are gradually activated with different tasks. Photonic neurons only lighten when activated by corresponding signals, in which the active connections are relatively sparse and the information is parallelly transmitted in spectrum. c Workflow of the L 2 ONN multi-task inference. Input information of multiple tasks is encoded into coherent light with different wavelengths, and processed with the sparse photonic computing module to obtain the final results Fig. 2 Free-space implementation of photonic lifelong learning (L 2 ONN). a Overall structure of L 2 ONN. Inputs of multi-tasks are projected into coherent light field with the multi-spectrum representations \\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}$${U}_{k}^{{\\lambda }^{i}}$$\\end{document} U k λ i . Beam splitter (BS), mirrors (M), lens (L) and optical filters are employed to guide and modulate the light. The cascaded sparse optical layers are realized by configuring the light-controlled optical filters at the Fourier plane of a 4 \\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}$$f$$\\end{document} f optical system. With propagation of optical feature embeddings O at the output plane, the final results can be obtained through an electronic read-out layer. b Detailed construction of the reconfigurable optical layers. Each layer receives sparse features as the inputs. PCM-based filters are all-optically switched, which sparsely conducts spatial and spectrum-wise photonic neuron activations. The activated photonic neurons are then connected in the subsequent diffractive computing module. c Training strategy of photonic lifelong learning on an 8 × 8 optical filter. Training of each task initially learns a dense activation map, which is further pruned to a sparse one. The activation map of each task is retained and stay fixed in the following evolution of learning. The final filter shares optical weights learned from all seen tasks The free-space L 2 ONN can adaptively allocate computational resources with unprecedented scalability and versatility, permitting ONNs to increment capabilities and memorize knowledges with enhanced performance. In the experiments, for the first time, we evaluate that L 2 ONN can progressively learn challenging tens-of-tasks, e.g., from hand-written digit classification to complex scene recognition (Fig. 3 ). The network achieves up to 14× larger learning capacity than the vanilla ONN 52 while maintaining competitive accuracy on each individual task, and more than an order of magnitude higher efficiency than the representative electronic based neural networks, e.g., LeNet 53 . It is worth noting that the learning sequence on complexity of tasks affects much on overall network performance (Fig. 4 ). The smarter way is to start from an easy task and slowly transition to more difficult ones, which corresponds with the progressive learning styles of human. Fig. 3 Evaluation on the photonic lifelong learning capability. a 5 representative vision classification tasks used for training L 2 ONN. Evolution of the activation map in layer 1 of b L 2 ONN and c vanilla ONN. With task learning, the photonic neuron connections in L 2 ONN are initially sparse and constantly enlarged, colored with red, yellow, green, blue and purple, respectively, while in vanilla ONN are quite dense from the first task. d Convergence comparison between L 2 ONN and vanilla ONN. Each task is trained for 5 epochs, L 2 ONN can increment its capabilities and memorize all seen tasks, while vanilla ONN rapidly forgets what was learned before and falls into the catastrophic forgetting area (below 20% accuracy) Fig. 4 Numerical performance of photonic lifelong learning. a Accuracy comparison among different benchmarks of vanilla ONN, L 2 ONN and LeNet. The electronic approach LeNet is installed with similar pruning rate and incrementally learns tasks using the same training strategy. b Analysis on network sparsity with the individual FashionMNIST task. All approaches are configured with fixed pruning settings under the same sparsity. c Evolution of the proposed optical filters with various training sequences. 5 tasks are divided into 3 task difficulty grades according to the photonic neuron activation density of each individual training (row 1). Tasks 1 and 2, and tasks 3 and 4 share the same grade due to they have the similar densities. Based on such criteria, d training sequences of easy to hard and hard to easy (rows 2 and 3) are evaluated, and e shifts of interior sequences on grades 1 and 2 (rows 4 and 5) are further reported An on-chip L 2 ONN is designed and fabricated for further validation, which experimentally verifies its lifelong learning performance on representative classification tasks in an all-optical and scalable manner (Fig. 5 ). The chip can realize a low-cost mass manufacturing based on standard CMOS technology, it is promising to implement L 2 ONN as a photonic accelerator onto the highly-integrated terminal/edge AI systems. We expect that our study will provide a light-speed and low-power solution to practically tackle real-world manifold tasks, meanwhile breaking through the energy and scaling walls towards more extensive applications of transformative AI techniques. Fig. 5 Photonic lifelong learning on chip. a Schematic of an on-chip L 2 ONN architecture. Each color represents a different task, multi-task inputs are encoded into optical signals and transmitted by multi-spectrum wave sources. The optical features are propagated with several sparse diffractive layers composed of silicon slots, each slot represents a single photonic neuron. b Evolution of a single layer during the multi-task learning process. c Micrograph of a real fabricated all-optical chip including 2 hidden layers, 16 input waveguides and 4 detectors. d Confusion matrices of the on-chip lifelong learning on 2 representative classification datasets. e The optical field propagation of L 2 ONN with FDTD that performs task 2 inference after learning both tasks",
"discussion": "Discussion This paper innovates a reconfigurable photonic neuromorphic architecture for scalable tens-of-task lifelong learning (L 2 ONN). It learns each single task by adaptively activating sparse photonic neuron connections, while continually acquiring expertise on various tasks by gradually enlarging the photonic activation, multi-task optical features are parallelly processed by multi-spectrum representations allocated with different wavelengths. An on-chip L 2 ONN is fabricated and experimentally verified its lifelong learning performance by incrementally implementing tasks on a single chip. Mechanism of the photonic lifelong learning is inspired by the fact of brain functions of protecting memories and accommodating new knowledges by leveraging sparse neuron connections and parallel task-driven neurocognition. Optics own more inherent advantages in sparsity and parallelism than electronic computing systems due to the massive optical information. Unlike the existing artificial intelligence methods are prone to train new models interfering with formerly learned knowledges, the proposed photonic neuromorphic architecture increments capabilities on multiple tasks and avoids the catastrophic forgetting issue. With the speed of light, L 2 ONN gains high capacity to continually acquire versatile expertise when confronted with continuous streams of new data. In summary, we have demonstrated the photonic lifelong learning provides a turnkey solution for large-scale real-life AI applications with unprecedented scalability and versatility. L 2 ONN shows its extraordinary learning capability on challenging tens-of-tasks, such as vision classification, voice recognition and medical diagnosis, supporting various new environments. We anticipate that the proposed neuromorphic architecture will accelerate the development of more powerful photonic computing as critical support for modern advanced machine intelligence and towards beginning a new era of AI."
} | 3,816 |
27683573 | PMC5021923 | pmc | 3,461 | {
"abstract": "While a functional quorum sensing system has been identified in the acidophilic chemolithoautotrophic Acidithiobacillus ferrooxidans ATCC 23270 T and shown to modulate cell adhesion to solid substrates, nothing is known about the genes it regulates. To address the question of how quorum sensing controls biofilm formation in A . ferrooxidans T , the transcriptome of this organism in conditions in which quorum sensing response is stimulated by a synthetic superagonist AHL (N-acyl homoserine lactones) analog has been studied. First, the effect on biofilm formation of a synthetic AHL tetrazolic analog, tetrazole 9c , known for its agonistic QS activity, was assessed by fluorescence and electron microscopy. A fast adherence of A. ferrooxidans T cells on sulfur coupons was observed. Then, tetrazole 9c was used in DNA microarray experiments that allowed the identification of genes regulated by quorum sensing signaling, and more particularly, those involved in early biofilm formation. Interestingly, afeI gene, encoding the AHL synthase, but not the A. ferrooxidans quorum sensing transcriptional regulator AfeR encoding gene, was shown to be regulated by quorum sensing. Data indicated that quorum sensing network represents at least 4.5% (141 genes) of the ATCC 23270 T genome of which 42.5% (60 genes) are related to biofilm formation. Finally, AfeR was shown to bind specifically to the regulatory region of the afeI gene at the level of the palindromic sequence predicted to be the AfeR binding site. Our results give new insights on the response of A. ferrooxidans to quorum sensing and on biofilm biogenesis.",
"conclusion": "Conclusion The exogenous use of tetrazole superagonist AHL analog 9c allowed the first overview of the QS regulon of A. ferrooxidans , an acidophilic bacterial species involved in bioleaching processes. This study gave some insights into the molecular chain reactions involved in the first steps of mineral adherence and colonization of this bacterium. As expected, tetrazole 9c activates the positive feedback previously reported ( Rivas et al., 2005 ) by inducing the transcription of afeI gene, likely through its binding to the transcriptional regulator AfeR, and therefore its activation, as early as the third day of growth. The data obtained from planktonic cells revealed that tetrazole 9c triggers the QS system to drive gene expression toward sessile state by reprogramming some cellular processes. These mainly include: (i) induction of the genes encoding the F0-ATPase subunit leading to the PMF allowing AHL efflux and influx, (ii) repression of several genes involved in carbohydrate metabolism to orientate carbon flow to maltodextrin and EPS building block precursor synthesis for adhesion and biofilm formation, respectively; (iii) induction of phosphate and ammonium transporters to anticipate inorganic ion gradient within and around the biofilm structure. Whereas QS and c-di-GMP pathway have been linked in different bacterial species ( Waters et al., 2008 ; Zhang, 2010 ; Kozlova et al., 2011 ; Suppiger et al., 2016 ), it is noteworthy that no change in the transcriptional profiling of the seven genes related to the c-di-GMP pathway in A. ferrooxidans ( Ruiz et al., 2012 ; Castro et al., 2015 ) has been observed in the presence of tetrazole 9c. This result indicates that QS does not modulate c-di-GMP signaling in this Gram-negative species. Finally, the high transcription level of afeI gene in sessile cells observed after 3 days of growth lead not only to A. ferrooxidans biofilm stabilization but also to the synthesis of a large spectrum of AHL molecules ( Farah et al., 2005 ; Valenzuela et al., 2017 ), some of which are sensed by secondary colonizers such as A. thiooxidans to form a mixed biofilm ( Bellenberg et al., 2014 ) through a not yet identified non-canonical AHL-binding protein.",
"introduction": "Introduction Due to its low operating cost, biomining is a very successful geobiotechnology that actually produces approximately 15 per cent of the world’s extracted copper ( Johnson, 2014 ). Withstanding low pH and high heavy metal concentrations, Acidithiobacillus species are acidophilic key players in biomining industry recovering valuable metals from sulfidic ores such as copper or gold ( Jerez, 2009 ). However, these bacteria are also involved in Acid Mine/Rock Drainage (AM/RD), which represents a worldwide problem of water pollution, from natural and anthropogenic environments ( Johnson, 2009 , 2012 ). Indeed, several studies recently indicated that Acidithiobacillus species play a pivotal and structural role in acidophilic communities ranging from 6°C to 90°C ( Chen et al., 2015 ; Liljeqvist et al., 2015 ; Menzel et al., 2015 ). Nevertheless, due to an insufficient understanding of the microbiological processes, most biohydrometallurgical plants operate far from maximum efficiency and natural AM/RD are to a large extent uncontrolled. Acidithiobacillia has been recently defined as a new class of Proteobacteria in which the genus Acidithiobacillus is the main one characterized ( Williams and Kelly, 2013 ). Actually, the genus Acidithiobacillus encompasses seven closely related Gram-negative, chemolithoautotrophic bioleaching species: (i) Acidithiobacillus thiooxidans, A. caldus , and A. albertensis , which oxidize only reduced inorganic sulfur compounds (RISC) and (ii) A ferrooxidans, A. ferrivorans, A. ferridurans , and A. ferriphilus that oxidize both ferrous iron and RISC ( Amouric et al., 2011 ; Hedrich and Johnson, 2013 ; Williams and Kelly, 2013 ; Falagan and Johnson, 2015 ). It has been well established that all Acidithiobacillus species are able to form biofilms on the surface of ores. This bacterial attachment on the mineral has been reported to increase metal leaching due to the formation of a close and enlarged “reaction space” between the metal sulfide surface and the cell ( Pogliani and Donati, 1999 ; Harneit et al., 2006 ; Rohwerder and Sand, 2007 ). Therefore, deciphering molecular mechanisms underlying biofilm formation in acidophilic leaching bacteria has been early pointed out as an important field of investigation. Quorum sensing (QS) and the secondary messenger c-di-GMP signaling pathway [for recent reviews see ( Hengge, 2009 ; Decho et al., 2011 ; Kalia et al., 2013 ; Romling et al., 2013 ; Hengge et al., 2015 )] are the most studied mechanisms controlling biofilm development in bacteria. Both pathways have been shown to be linked in several bacterial species ( Ryan et al., 2006 ; Waters et al., 2008 ; Ueda and Wood, 2009 ; Zhang, 2010 ; Kozlova et al., 2011 ) and to control more particularly polysaccharide production and biofilm formation ( Ueda and Wood, 2009 ). QS is an important mechanism for the timing of collective behaviors through the regulation of population density-dependent cellular processes, such as the production of virulence factors, motility, exopolysaccharide production and biofilm formation ( Parsek and Greenberg, 2005 ; Waters and Bassler, 2005 ; Ng and Bassler, 2009 ). In Gram-negative bacteria, the main characterized QS system involves three key molecular elements ( Venturi and Subramoni, 2009 ): (i) N-acyl homoserine lactones (AHLs), which act as autoinducers (AIs); (ii) the AHLs synthase encoded by a luxI -like gene; (iii) a transcriptional regulator, which is encoded by a luxR -like gene and which binds AHL molecules and modulates the expression of different target genes that constitute the QS regulon. Depending on the bacterial species and also on the experimental strategies (transcriptomic or proteomic), the size of the QS regulons oscillates between 3 and 8% of the identified ORFs ( Vasil, 2003 ; Wagner et al., 2003 ; Cantero et al., 2006 ; Qin et al., 2007 ; Stevens et al., 2011 ; Majerczyk et al., 2014 ). Even if several reports related to biofilm formation regulation by acidophilic bacteria belonging to Acidithiobacillus genus have been released recently ( Farah et al., 2005 ; Bellenberg et al., 2012 , 2014 ; Ruiz et al., 2012 ; Diaz et al., 2013 ; Montgomery et al., 2013 ; Vera et al., 2013 ; Castro et al., 2015 ), the molecular cascade involved in exopolysaccharide production and biofilm formation by Acidithiobacillus species is still undeciphered. While c-di-GMP pathway has been identified in all Acidithiobacillus spp. ( Ruiz et al., 2012 ; Diaz et al., 2013 ; Castro et al., 2015 ), the species that oxidize only RISC do not possess the genes related to canonical QS systems ( Valdés et al., 2008 ). Indeed, a functional QS system has been reported only in the iron/RISC-oxidizing species A. ferrooxidans ( Farah et al., 2005 ; Rivas et al., 2005 ; Valenzuela et al., 2007 ). In addition, it has been recently reported that the RISC-oxidizing species A. thiooxidans cannot adhere to pyrite if this mineral is not previously colonized by an iron-oxidizing species ( Bellenberg et al., 2014 ) pointing out A. ferrooxidans as a key player for mineral colonization. Acidithiobacillus ferrooxidans ATCC 23270 T QS system involves two divergent genes afeI and afeR coding for the AHL synthase and the transcriptional regulator, respectively ( Farah et al., 2005 ). AfeR has the conserved amino acid residues located in the active site of LuxR-protein family and possesses the canonical AHL and DNA binding domains based on a 3D-structural model ( Soulere et al., 2008 ). In A. ferrooxidans ATCC 23270 T , nine different AHL molecules are synthesized with medium or large acyl side chains ( Valenzuela et al., 2007 ). In this strain, transcription of afeI is increased under the physiological conditions that promote biofilm formation, such as growth in the presence of sulfur (solid energetic substrate) or in low phosphate medium ( Farah et al., 2005 ), suggesting a role of QS system in the attachment of A. ferrooxidans to ores (e.g., pyrite). In agreement with this hypothesis, addition of synthetic AHL that are AIs naturally synthesized by A. ferrooxidans such as C14-AHL and 3-hydroxy-C14-AHL has been shown to enhance A. ferrooxidans ATCC 23270 T cell adhesion, exopolysaccharide production and biofilm development on elemental sulfur and pyrite ( Ruiz et al., 2008 ; Gonzalez et al., 2013 ). However, to date this phenotypic result is still uncoupled with genotypic data that will allow the understanding of the molecular chain reaction going from the AHL-sensing by AfeR to ore colonization. A bioinformatics analysis has recently allowed the identification of a putative QS regulon in A. ferrooxidans ATCC 23270 T that encompasses 75 possible AfeR target-genes, including genes likely involved in polysaccharide biosynthesis ( Banderas and Guiliani, 2013 ). However, biological data are required to fully identify the A. ferrooxidans genes whose expression is modulated by AHL signaling. Here, we report the first biological study focused on deciphering the QS regulon of A. ferrooxidans ATCC 23270 T . The effects of AI 3-hydroxy-C14-AHL and of tetrazolic AHL-analog 9c , on A. ferrooxidans adhesion to sulfur were first compared by fluorescence and scanning electronic microscopy. Then, DNA microarray experiments were performed to compare total RNA of A. ferrooxidans ATCC 23270 T cells induced or not by tetrazole 9c . These allowed the identification of 141 genes from which at least 48 can be linked with QS pathway, exopolysaccharide production and biofilm development. If we include the genes encoding hypothetical proteins that colocalized and are coregulated with these 48 genes, this number would increase to 60 and represents 1.9% of the ATCC 23270 T genome.",
"discussion": "Results and Discussion To develop biological strategies for improving biomining activities and preventing environmental damages caused by AM/RD, it is well documented that mineral colonization by acidophilic bacteria such as Acidithiobacillus species is a key step to decipher ( Rohwerder and Sand, 2007 ). If synthesis of specific exopolysaccharides rich in α-mannopyranosyl and α-glucopyranosyl sugar residues has been revealed by fluorescently labeled lectin Concanavalin A within 1 day for EPS (extracellular polymeric substances)-deficient ferrous-iron grown cells after transfer to cultures with pyrite as sole nutrient ( Bellenberg et al., 2014 ), a clear understanding of the molecular cascade involved in exopolysaccharide production and biofilm formation by Acidithiobacillus species is actually missing. However, as a molecular relationship between QS and cell adhesion has been clearly established in A. ferrooxidans ( Gonzalez et al., 2013 ) and it has to be pointed out that the canonical QS systems are missing in Acidithiobacillus species that can oxidize only RISC ( Valdés et al., 2008 ), the iron-oxidizing species such as A. ferrooxidans as primary colonizers are now considered fundamental players for mineral colonization by the bioleaching community. Therefore, to address the question of how A. ferrooxidans regulates the physiological processes involved in cell adhesion, EPS production and biofilm formation, we focused on the deciphering of the QS molecular network by using a synthetic QS-activator molecule and DNA array technology. The Tetrazolic AHL Analog 9c Accelerates Cellular Adhesion of Acidithiobacillus ferrooxidans on Sulfur Coupons To further investigate the molecular mechanisms underlying this pathway, we first challenged the identification of synthetic AHL analogs capable to induce better A. ferrooxidans cell adhesion than natural AIs previously tested ( Gonzalez et al., 2013 ). Thus, a tetrazolic derivative that displays a much higher affinity to the LuxR protein than the natural AI and acts as a superagonist of AHL signaling molecules ( Sabbah et al., 2012 ) was tested. Its effect on biofilm formation by A. ferrooxidans was compared to the natural AI 3-hydroxy-C14-AHL ( Figure 1 ). Growth curves revealed that both tetrazolic AHL analog and 3-hydroxy-C14-AHL have no effect on A. ferrooxidans growth compared to the control in the absence of exogenous AHL ( Figure 1A ). Fluorescence ( Figure 1B ) and SEM ( Figure 1C ) clearly indicated that tetrazole 9c also promoted cell adhesion. Moreover, confirming in the A. ferrooxidans model the superagonistic behavior of tetrazole 9c previously found in V. fisheri ( Sabbah et al., 2012 ), the results obtained on day 3 strongly suggest that tetrazole 9c is biologically more efficient than the natural AI 3-hydroxy-C14-AHL in promoting biofilm formation ( Figure 1C ). FIGURE 1 Effect of tetrazole 9c on biofilm formation. (A) Growth curves in the absence or the presence of 5 μM 3-hydroxy-C14-AHL or tetrazole 9c . Arrows indicate when aliquots were sampled for microscopy analysis. (B) Fluorescence microscopy of sulfur coupons after 1, 3, or 5 days of Acidithiobacillus ferrooxidans ATCC 23270 T cells grown in the absence (-) or the presence of 5 μM 3-hydroxy-C14-AHL or tetrazole 9c . Cells were stained with fluochrome Syto9. (C) Electron microscopy of sulfur coupons treated as above. White bars represent 20 μm. The QS System is Triggered after 3 days in the Presence of the Tetrazolic AHL Analog 9c in Planktonic Cells The results presented in Figure 1 suggest that QS was triggered by 5 μM tetrazole 9c between 2 and 3 days of growth versus 4–5 days in the absence of this AHL analog. To assess whether the tetrazole 9c indeed switched on QS system by inducing the transcription of the genes known to be involved in QS response ( Farah et al., 2005 ), i.e. afeI (AFE_1999) and afeR (AFE_1997), the transcription of these genes was analyzed by quantitative real-time PCR after 2, 3, and 4 days of growth in the presence or the absence of 5μM tetrazole 9c . The results indicated that afeR expression was constitutively expressed under the conditions analyzed, while afeI was induced by tetrazole 9c from the third day of growth in planktonic cells ( Table 1 ). Table 1 Quantitative real-time PCR expression data for afeI, afeR, zwf , AFE_0233 (glycosyl transferase), and AFE_1339 (putative polysaccharide export protein) genes from Acidithiobacillus ferrooxidans ATCC 23270 T planktonic cells grown with sulfur prills in the presence or the absence of 5 μM tetrazole 9c after 2, 3, and 4 days of growth. Gene or locus name Growth condition Day of growth Gene mRNA/ rrs ± SD a afeI (AFE_1999) DMSO 2 1 ± 0 Tetrazole 9c 2 1.69 ± 0.15 DMSO 3 4.49 ± 0.58 Tetrazole 9c 3 12.75 ± 1.66 DMSO 4 5.19 ± 6.95 Tetrazole 9c 4 49.35 ± 5.72 afeR (AFE_1997) DMSO 2 1 ± 0 Tetrazole 9c 2 1.58 ± 0.19 DMSO 3 1.59 ± 0.19 Tetrazole 9c 3 1.21 ± 0.08 DMSO 4 1.64 ± 0.11 Tetrazole 9c 4 1.78 ± 0.21 zwf (AFE_2025) DMSO 2 1 ± 0 Tetrazole 9c 2 1.43 ± 0.12 DMSO 3 2.70 ± 1.11 Tetrazole 9c 3 1.66 ± 0.16 DMSO 4 2.82 ± 0.32 Tetrazole 9c 4 3.90 ± 0.58 AFE_0233 DMSO 2 1 ± 0 Tetrazole 9c 2 1.04 ± 0.01 DMSO 3 1.13 ± 0.13 Tetrazole 9c 3 1.00 ± 0.13 DMSO 4 1.14 ± 0.17 Tetrazole 9c 4 0.91 ± 0.02 AFE_1339 DMSO 2 1 ± 0 Tetrazole 9c 2 1.72 ± 0.02 DMSO 3 1.58 ± 0.32 Tetrazole 9c 3 1.62 ± 0.32 DMSO 4 1.56 ± 0.11 Tetrazole 9c 4 1.69 ± 0.12 a Values were related to those obtained after 2 days of growth in the absence of tetrazole 9c . Biofilm formation after 3 days was strongly enhanced in cells treated with 5 μM tetrazole 9c compared to cells from control experiments without agonist ( Figure 1 ). Therefore, expression of some genes predicted to be linked to EPS biosynthesis [ zwf (AFE_2025), AFE_0233, and AFE_1339] was also monitored in planktonic ( Table 1 ) and sessile (Supplementary Table S2 ) cells after 3 days of growth with 5 μM tetrazole 9c . The gene zwf encodes glucose-6-phosphate 1-dehydrogenase that is involved in the intracellular levels of glucose-6P, a precursor of the EPS. AFE_0233 encodes a glycosyl transferase and is located in a gene cluster predicted to encode cell wall constituents (polysaccharides, and lipopolysaccharides). AFE_1339 encodes the putative polysaccharide export protein Wza and is located close to the gal operon proposed to be involved in the formation of EPS in iron-grown cells ( Barreto et al., 2005 ). Besides, AfeR-AHL binding sites were predicted in the regulatory region of zwf , AFE_0233, and AFE_1339 ( Banderas and Guiliani, 2013 ). Surprisingly, tetrazole 9c had no effect on AFE_0233, AFE_1339 and zwf transcription and only the expression of the afeI gene was induced by tetrazole 9c ( Table 1 ; Supplementary Table S2 ). These data indicate that afeI , and not afeR , is regulated by QS and suggest either that zwf , AFE_0233, and AFE_1339 genes were not regulated by AfeR or that their expression was induced later during biofilm biogenesis. QS Regulon in Acidithiobacillus ferrooxidans Cells Quorum sensing response and biofilm formation were obvious within 3 days of growth in the presence of the tetrazolic AHL analog 9c ( Figure 1 ; Table 1 ). Consequently, total RNAs from planktonic and sessile cells of A. ferrooxidans ATCC 23270 T were isolated from 3-days cultures in the presence or the absence of the superagonist AHL analog 9c . They were used to probe gene expression using microarrays displaying two specific oligonucleotides for each gene of this bacterium (3147 predicted genes). Only the genes filling the conditions described in the Materials and Methods section were analyzed. It has to be pointed out that the microarray and quantitative real-time PCR data agreed with the constitutive expression of afeR, zwf , AFE_0233, and AFE_1339 genes under the conditions tested ( Table 1 ; Supplementary Tables S2 – S4 ). In planktonic cells, a total of 133 genes were differentially expressed, 34 induced and 99 repressed by tetrazole 9c (Supplementary Table S3 ). In sessile cells under the same conditions, only eight genes presented significant differences in expression, four induced and four repressed by tetrazole 9c (Supplementary Table S4 ). Therefore, 141 genes were QS regulated, which represent 4.5% of the total number of A. ferrooxidans gene analyzed in this study (see Materials and Methods). These genes were grouped according to their COG classification. Their percentage relative to all the A. ferrooxidans ATCC 23270 T genes present in the same COG class is given in Table 2 . In planktonic cells, mainly the genes involved in inorganic ion transport and metabolism (4.86%), and nucleotide transport and metabolism (3.39%) were induced in the presence of tetrazole 9c . Mainly those involved in carbohydrate transport and metabolism (11.11%), posttranslational modification, protein turnover, chaperones (8.27%), energy production and conversion (5.76%), cell motility (3.70%), and transcription (2.92%) as well as poorly characterized proteins (11%) were repressed by this AHL analog. In sessile cells, mainly induction by tetrazole 9c of secondary metabolites biosynthesis, transport and catabolism (1.61%), and signal transduction mechanisms (1.15%) was observed while repression was detected for energy production and conversion genes (1.05%). Only the genes differentially expressed in cells that were cultivated with or without the tetrazolic AHL analog and which have known or reliable predicted function are presented in Table 3 for the planktonic cells and in Table 4 for the sessile cells and are discussed below. Table 2 COG classification of the genes differentially expressed in planktonic and sessile cells grown with (+) and without (-) tetrazole 9c . Process COG functional categories COG class Planktonic cells a,b Sessile cells a,b + - + - Cellular processes and signaling Cell cycle control, cell division, chromosome partitioning D 0.00% 0.00% 0.00% 0.00% Cell wall/membrane/envelope biogenesis M 0.93% 0.93% 0.00% 0.00% Cell motility N 0.00% 3.70% 0.00% 0.00% Posttranslational modification, protein turnover, chaperones O 0.00% 8.27% 0.75% 0.00% Signal transduction mechanisms T 0.00% 1.15% 1.15% 0.00% Intracellular trafficking, secretion, and vesicular transport U 1.89% 0.94% 0.00% 0.00% Defense mechanisms V 0.00% 0.00% 0.00% 0.00% Extracellular structures W 0.00% 0.00% 0.00% 0.00% Information storage and processing RNA processing and modification A 0.00% 0.00% 0.00% 0.00% Chromatin structure and dynamics B 0.00% 0.00% 0.00% 0.00% Translation, ribosomal structure, and biogenesis J 1.89% 0.00% 0.00% 0.00% Transcription K 0.00% 2.92% 0.00% 0.00% Replication, recombination, and repair L 0.00% 1.35% 0.00% 0.00% Metabolism Energy production and conversion C 2.09% 5.76% 0.00% 1.05% Amino acid transport and metabolism E 1.01% 1.52% 0.00% 0.00% Nucleotide transport and metabolism F 3.39% 0.00% 0.00% 0.00% Carbohydrate transport and metabolism G 1.59% 11.11% 0.00% 0.00% Coenzyme transport and metabolism H 0.00% 1.74% 0.00% 0.00% Lipid transport and metabolism I 0.00% 1.41% 0.00% 0.00% Inorganic ion transport and metabolism P 4.86% 1.62% 0.54% 0.54% Secondary metabolites biosynthesis, transport, and catabolism Q 0.00% 1.61% 1.61% 0.00% Poorly characterized General function prediction only R 0.31% 4.97% 0.62% 0.31% Function unknown S 0.00% 6.03% 0.00% 0.00% a The numbers represent the percentage relative to all the A. ferrooxidans ATCC 23270 T genes present in this COG class. b Bold numbers are discussed in the text. Table 3 Microarray expression data for genes with known or predicted function differentially expressed in planktonic cells in the presence of tetrazole 9c. Table 4 Microarray expression data for genes with known or predicted function differentially expressed in sessile cells in the presence of tetrazole 9c. Genes Differentially Expressed in the Presence of Tetrazole 9c in Planktonic Cells In planktonic cells, tetrazole 9c modified the expression of a number of genes related to biofilm formation, few being induced and several repressed. Among the induced genes, those involved in inorganic ion transport and energy conversion were mainly found. Not surprisingly, genes involved in the transport of phosphate [ pstS (AFE_1939) and pstC (AFE_1940)] and ammonium [ glnK (AFE_2915) and amt (AFE_2916)] were upregulated. The phosphate specific transport (Pst) system is known to be important in biofilm formation in a number of bacteria [see ( O’May et al., 2009 ; Heindl et al., 2014 ) and references therein] including Leptospirillum ferrooxidans ( Moreno-Paz et al., 2010 ) and A. ferrooxidans ( Vera et al., 2013 ), in which phosphate metabolism was early linked to QS regulatory pathway ( Farah et al., 2005 ). Deep cDNA sequencing experiments also revealed that several genes related to ammonium metabolism ( amt-1, amt-2 , and glnK-1 ) were upregulated in A. ferrooxidans planktonic cells induced by hydroxyl-C14-AHL compared to not induced (unpublished data). Biofilm formation occurs also in response to the availability of nutrients supplied by the ammonium transporter (AFE_2916) which expression is regulated by GlnK (AFE_2915), as shown recently in Streptococcus mutans ( Ardin et al., 2014 ). This might anticipate gradient of inorganic ions within and around microbial biofilm. The other gene class that was induced by tetrazole 9c in planktonic cells is involved in energy production and conversion, in particular the genes atpBEF (AFE_3207–3209) encoding the membrane-embedded proton channel F0 of the ATPase. This upregulation could allow more protons to pass through the ATP synthase complex generating a proton motive force (PMF) rather than ATP. PMF is required not only for early biofilm formation ( Saville et al., 2011 ), but also in influx and efflux involved in QS since PMF inhibition enhances the intracellular accumulation of AHL leading to decrease in biofilm formation ( Ikonomidis et al., 2008 ; Varga et al., 2012 ). Along the same lines, genes encoding a putative MolA/TolQ/ExbB proton channel family protein (AFE_2273) and TonB family protein (AFE_2275) were upregulated in the presence of the tetrazole 9c and could contribute to PMF-dependent import through the outer membrane of substrates necessary for QS and/or early EPS synthesis. Another interesting gene that was more expressed in the presence of the tetrazolic AHL analog in planktonic cells is ndk (AFE_1929) encoding a nucleoside diphosphate kinase. A ndk knockout mutant of Pseudomonas aeruginosa was shown to be deficient in polysaccharide synthesis ( Kapatral et al., 2000 ), because it was unable to provide GTP necessary for the incorporation of mannuronate in alginate. It is therefore possible that nucleotide triphosphates are required in an early step of A. ferrooxidans EPS biosynthesis. The genes that were repressed in the presence of the tetrazolic AHL analog in planktonic cells were mainly involved in energy production and conversion, carbohydrate transport and metabolism, posttranslational modification, protein turnover, chaperones, and transcription. Most of the energy production and conversion class genes encoded two out of the four hydrogenases described in A. ferrooxidans . One is a group one membrane-bound respiratory enzyme enabling the cell to use H 2 as an energy source [ hynS (AFE_3283) and hynL (AFE_3286)]. The genes encoding this hydrogenase physiological partners [ isp1 (AFE_3284) and isp2 (AFE_3285)] and biogenesis machinery [ hynD (AFE_3281), hynH (AFE_3282), hynL (AFE_3286), hypA (AFE_3287), hypB (AFE_3288), hypC (AFE_3289), and hypD (AFE_3290)] were also repressed under this condition. The second hydrogenase is a sulfhydrogenase, a group 3b cytoplasmic hydrogenase [ hoxH (AFE_0937) and hoxF (AFE_0940)], with both hydrogenase and sulfur reductase activities, likely serving as an electron sink under highly reducing conditions by recycling redox cofactors using either protons or polysulfides as the electron acceptor. It is worth mentioning that, in different bacteria, some hydrogenases were shown to be upregulated in sessile cells, others in planktonic cells ( Caffrey et al., 2008 ; Clark et al., 2012 ; Kassem et al., 2012 ). Our data suggest that the groups 1 and 3 hydrogenases of A. ferrooxidans are specific to the non-attached cells. The number of genes belonging to the carbohydrate transport and metabolism class that were differentially expressed with/without tetrazole 9c agrees with an alteration in the carbon flow when planktonic cells switched to sessile state. It has to be pointed out that all these genes were downregulated in the presence of the tetrazole 9c . Three pathways seemed to be affected: the glycolysis [ pyk (AFE_1801), AFE_1802, gpmL (AFE_1815)], the pentose phosphate pathway [ tal (AFE_0419), AFE_1857, and AFE_2024] and the glycogen biosynthesis/degradation pathway [AFE_1799, AFE_2082, AFE_2083, and glgB (AFE_2836)]. In the case of glycolysis, this could mean that the pathway was directed toward β- D -fructose-1,6-bisphosphate, β- D - fructose-6-phosphate, α- D -glucose-6-phosphate, and α- D -glucose-1-phosphate production ( Figure 2A ). Similarly, in the pentose phosphate pathway, the repression would lead toward β- D -glucose, β- D -glucose-6-phosphate and β- D -fructose-6-phosphate direction and therefore to α- D -glucose-6-phosphate and α- D -glucose-1-phosphate accumulation ( Figure 2A ). Noteworthy, α- D -glucose-6-phosphate and α- D -glucose-1-phosphate are the precursors of UDP glucose, UDP-galactose, dTDP-rhamnose and GDP-mannose, which are the building blocks in EPS biosynthesis ( Quatrini et al., 2007 ). Another interesting results was the repression of three genes predicted to be involved in trehalose synthesis [ treT (AFE_2083), treZ (AFE_2082) and treY (AFE_2081)] by tetrazole 9c . In the first case, α- D -glucose-1-phosphate consumption will be prevented, in agreement with the data presented above, and, in addition, maltodextrin synthesis will be favored. In the second case, maltodextrin consumption will be avoided ( Figure 2B ). Notably, maltodextrin has been shown to increase E. coli adhesion ( Nickerson and McDonald, 2012 ). Along the same lines, genes involved in maltodextrin consumption [AFE_1799 and glgB (AFE_2836)] were repressed in the presence of tetrazole 9c ( Figure 2B ). Therefore, in planktonic cells, it appears that tetrazole 9c directed the carbon flow toward adhesion (maltodextrin), EPS precursor biosynthesis (α- D -glucose-6-phosphate, α- D -glucose-1-phosphate) and therefore biofilm formation. FIGURE 2 Acidithiobacillus ferrooxidans pathways showing gene repression by tetrazole 9c . ( A ) Genes involved in carbon flow that were downregulated in planktonic cells (for clarity, only the connections of the pathways discussed in the text have been shown). (B) Genes involved in formaldehyde oxidation that were downregulated in sessile cells. The precursors of the EPS (extracellular polymeric substances) building blocks are surrounded by a black rectangle. The genes that were repressed by tetrazole 9c are indicated in a white background. In a number of microorganisms, including L. ferrooxidans , heat shock chaperones ( Moreno-Paz et al., 2010 ; Singh et al., 2012 ; Becherelli et al., 2013 ; Grudniak et al., 2013 ) and proteases ( Doern et al., 2009 ; Moreno-Paz et al., 2010 ; Singh et al., 2012 ; Yepes et al., 2012 ), in particular O-sialoglycoprotein endopeptidase ( Wickstrom et al., 2013 ), have been shown to be required in sessile cells for biofilm development. Furthermore, the uspA gene, encoding an universal stress protein, is necessary for optimal biofilm formation in Porphyromonas gingivalis ( Chen et al., 2006 ). In A. ferrooxidans , tetrazole 9c repressed the genes encoding the heat shock response RNA polymerase σ32 factor [ rpoH (AFE2750)], Hsp20 family heat shock proteins (AFE_0871 and 2086), a putative universal stress protein (AFE_0751), as well as protease [ lon (AFE_0872)] and O-sialoglycoprotein endopeptidase [ gcp (AFE_0123)] in planktonic cells, indicating that these proteins are not required at the early step of biofilm biogenesis. Interestingly, bioD (AFE_1675) was repressed in the presence of the tetrazolic analog. This gene encodes dethiobiotin synthetase involved in biotin synthesis from 7-keto-8-aminopelargonate. This pathway consumes S -adenosyl- L -methionine ( Streit and Entcheva, 2003 ). The down-regulation in the presence of tetrazole 9c of this gene could save this substrate that is required for AHL biosynthesis. Another important data is the repression of proB (AFE_2464) in the presence of tetrazole 9c . The proB gene encodes glutamate-5-kinase and its repression could lead to glutamate accumulation. Glutamate metabolism has been reported to be essential for biofilm formation. Amino acid levels in general increased in biofilm cells and are used as precursors for energy production with gluconeogenesis ( Yeom et al., 2013 ). In harsh environments, such as acidic conditions, a high demand for amino acids as substrates for energy production may indeed exist in biofilms. Very recently, it has been proposed that amino acids, including glutamate, may also have another role as a signal for biofilm maturation and eventual disassembly ( Wong et al., 2015 ). Finally, two genes encoding transcriptional regulators (AFE_2209 and AFE_2641) were repressed in planktonic cells in the presence of tetrazolic AHL analog. Therefore, we cannot exclude the possibility that genes differentially expressed in the presence of this superagonist AHL analog were indirectly regulated by one of these regulators rather than directly by the AfeR/AfeI QS system. It is noteworthy that members of the TetR-protein family, as is the case for AFE_2209, have been directly involved in the regulation of cellular processes and in particular the QS in different Gram-negative species ( Cuthbertson and Nodwell, 2013 ; Longo et al., 2013 ). To summarize, in planktonic cells, tetrazole 9c led to the induction of genes encoding (i) proton channel proteins to allow PMF energized transport system of AHL and substrates required for EPS synthesis, (ii) an enzyme required in an early step of polysaccharide synthesis, and (iii) transport system to anticipate phosphate and ammonium gradients within the biofilm. On the other hand, it repressed genes involved in (i) biofilm maturation (heat-shock proteins and chaperone encoding genes), (ii) biotin synthesis to prevent the consumption of S -adenosyl- L -methionine required for AHL biosynthesis, (iii) glutamate conversion to proline to use it as an energy source and/or as a signal for biofilm maturation, and (iv) carbohydrate metabolism to redirect the carbon flow toward proteins necessary for adhesion and EPS precursor biosynthesis. It seems therefore reasonable to conclude that tetrazole 9c reprograms planktonic cells toward early biofilm formation. Genes Differentially Expressed in the Presence of Tetrazole 9c in Sessile Cells In sessile cells, only four genes, encoding proteins with known or predicted functions, presented significant differences in expression. Not surprisingly, the gene with the highest fold difference was afeI (AFE_1999) encoding the AHL synthase, with at least an eight-fold expression increase in the presence of tetrazole 9c indicating that indeed the QS was triggered. The three other genes fdhD (AFE_0690), adhI (AFE_0697), and fghA (AFE_0698) encoding a putative formate dehydrogenase family accessory protein FdhD, a S -(hydroxymethyl) glutathione dehydrogenase, and a S -formylglutathione hydrolase, respectively, are involved in formaldehyde oxidation to formate ( Figure 2B ). Their repression could lead to the accumulation of formaldehyde, shown to lead to higher biofilm density in a biofilm reactor ( Ong et al., 2006 ). Another not exclusive possibility is that this system is to prevent formate formation that could acidify A. ferroxidans cytoplasm and lead to cell death. Surprisingly, only three genes differentially expressed in the presence of tetrazole 9c (Supplementary Tables S3 and S4 ) have the AfeR binding site inferred from bioinformatic prediction ( Farah et al., 2005 ; Banderas and Guiliani, 2013 ): AFE_0582 and AFE_1998 encoding hypothetical proteins as well as afeI (AFE_1999). This could be due to an indirect regulation through a regulator whose expression is controlled by QS. However, the two genes encoding a transcription regulator whose expression was downregulated in the presence of tetrazole 9c (AFE_2209 and AFE_2641) do not exhibit this predicted AfeR binding site. On the other hand, three genes [ zwf (AFE_2025), AFE_0233, and AFE_1339] in which this site was predicted, are constitutively expressed in the conditions analyzed. Therefore, another possibility is that a different transcriptional regulator than AfeR binds to the proposed AfeR binding site. All in all, the QS regulon of A. ferrooxidans seems to involve a complex regulatory cascade. AfeR Binds Specifically to the afeI Regulatory Region To check that the afeI induction in the presence of tetrazole 9c observed by transcriptomic data was mediated by the QS regulator AfeR, we have produced AfeR in E. coli and analyzed its binding to the regulatory region of the afeI gene. AfeR with a hexa-histidine tag fused to its C terminus (AfeR-Histag) was mainly found in the inclusion bodies, even when the 3-hydroxy-C14-AHL ( Gonzalez et al., 2013 ) was added at the induction time. The recombinant AfeR-Histag produced in the presence of 3-hydroxy-C14-AHL was purified on an affinity cobalt column. As shown in Figure 3A , a major band of the expected mass (theoretical molecular mass: 27,876 Da including one molecule of 3-hydroxy-C14-AHL) was visualized on Coomassie blue-stained SDS-polyacrylamide gels. This same protein was recognized by anti-hexahistidine tag antibodies ( Figure 3A ) strongly suggesting that it was AfeR-Histag. The analysis by MALDI-TOF mass spectrometry of this protein digested with Trypsin after reduction by DTT and alkylation by iodoacetamide confirmed that it was AfeR-Histag (54% sequence coverage). FIGURE 3 Production of the recombinant AfeR-Histag in Escherichia coli and its binding on the afeI regulatory region. ( A ) Coomassie brilliant blue stained SDS-PAGE (left) and Western immunoblot with antisera raised against the hexahistidine tag (right). The size of the unstained protein molecular weight marker standards (Euromedex) is indicated. (B) Schematic representation of the afeI locus with the DNA fragments analysed. (C) Gel mobility shift assays with an internal DNA fragment of the rrs gene of Thiomonas arsenitoxydans (left upper part) and the DNA fragments depicted in (B) . Binding of AfeR-Histag to the regulatory region of afeI was analyzed by EMSA in the presence of 3-hydroxy-C14-AHL. A retarded band was detected with 1.3 μM AfeR-Histag and higher concentrations ( Figure 3C ) with DNA fragments encompassing the palindromic sequence predicted to be the AfeR binding site ( Farah et al., 2005 ; Banderas and Guiliani, 2013 ) in the afeI regulatory region ( Figure 3B ). This binding was specific to this region since no binding was observed on an internal fragment of the rrs gene of Thiomonas arsenitoxydans ( Figure 3C ). These results indicate that AfeR-Histag binds to the regulatory region of afeI in the presence of 3-hydroxy-C14-AHL, in agreement with the induction of this gene in the presence of tetrazole superagonist AHL analog 9c . Since AfeR was constitutively expressed under the conditions analyzed (i.e., with or without tetrazole 9c ), these results suggest that the binding of 3-hydroxy-C14-AHL to AfeR induces a conformational change allowing its specific binding to the target DNA, as it has been proposed for several members of LuxR-like protein family ( Choi and Greenberg, 1991 )."
} | 9,982 |
27658446 | PMC5034286 | pmc | 3,462 | {
"abstract": "The beautiful structural colors in bird feathers are some of the brightest colors in nature, and some of these colors are created by arrays of melanin granules that act as both structural colors and scattering absorbers. Inspired by the color of bird feathers, high-visibility structural colors have been created by altering four variables: size, blackness, refractive index, and arrangement of the nano-elements. To control these four variables, we developed a facile method for the preparation of biomimetic core-shell particles with melanin-like polydopamine (PDA) shell layers. The size of the core-shell particles was controlled by adjusting the core polystyrene (PSt) particles’ diameter and the PDA shell thicknesses. The blackness and refractive index of the colloidal particles could be adjusted by controlling the thickness of the PDA shell. The arrangement of the particles was controlled by adjusting the surface roughness of the core-shell particles. This method enabled the production of both iridescent and non-iridescent structural colors from only one component. This simple and novel process of using core-shell particles containing PDA shell layers can be used in basic research on structural colors in nature and their practical applications.",
"discussion": "Results and Discussion Synthesis of PSt@PDA core-shell particles Monodisperse PSt particles with different diameters, which were prepared by emulsifier-free emulsion polymerization using hydrophilic comonomers, were used as the core material 29 . According to the method in our previous report, PSt@PDA core-shell particles were prepared in the DA polymerizations performed at different monomer concentrations in the presence of PSt core particles 30 31 . The coating of the PDA shell layer onto the PSt core particles was confirmed using FT-IR spectroscopy (see Figure S1 ). While the surface ζ potential of the PSt core particles was found to be approximately +60 mV, that of the PSt@PDA(Y) core-shell particles was measured to be negative (approximately −60 mV), which also indicated the presence of PDA shell layers (see Figure S2 ). Due to the high ζ potentials, the dispersions obtained were very stable in water, and no significant change was observed after 3 months of storage. The upper row of Fig. 2a shows photographs of the dispersion of PSt(237)@PDA(Y) core-shell particles (0.5 wt% in water). The brown coloration increased as the DA monomer concentration increased from 0.3 to 2.0 mg/mL. The PDA shell thickness, which was measured by SEM, gradually increased from approximately 2.5 nm to approximately 22 nm as the DA concentration increased, without affecting the core size ( Fig. 2b ). The blackness of the particles is an important factor related to obtaining high-visibility structural colors. Thus, we measured the UV-vis spectra of 0.005 wt% water dispersions of PSt(237)@PDA(Y) core-shell particles (light path = 10 mm) to investigate their behaviors. As shown in Fig. 2c , the transmittance at 600 nm decreased as the shell thickness increased, indicating that the variation of the particles’ blackness was effectively controlled by the PDA coating method. Fabrication of high-visibility structural color pellets from core-shell particles The lower row of Fig. 2a shows the structural color pellets prepared by PSt(237)@PDA(Y) core-shell particles together with the pellets from bare PSt particles. When conventional PSt particles were used as components, color pellets that were milky-white due to light scattering were obtained. In contrast, PSt@PDA core-shell particles produced bright structural color pellets. These pellets were created using only PSt@PDA core-shell particles as components. When the pellets were deliberately broken, brown powders were obtained (see Figure S3 ). A PDA shell layer of only 2.5 nm dramatically increased the visibility of structural colors to the naked eye. As the thickness of the PDA shell layer increased, the colors of the pellets changed from blue to green to red. Figure 2d shows the reflection spectra of the pellets. The reflection peaks underwent a red shift as the PDA layer thickness increased, in accordance with a previously reported result that structural colors were controlled by altering the size of particles 16 21 . A change in the PDA layer thickness also influenced the reflectance percentage. When PSt@PDA core-shell particles with thin shell layers (0–4.5 nm) were used as components, the highest reflectance of the pellets was approximately 40–80%, while the highest reflectance of the pellets made from particles with a thick PDA layer (8–22 nm) dramatically decreased to approximately 5%, at which point saturated structural colors were observed by the naked eye; this result is in good agreement with that reported previously 16 21 . Under practical conditions, Bragg’s law can be approximately expressed as Equation (1) by considering the effective refractive index of the system: where m is the order of diffraction, λ is the wavelength of light, n is the refractive index of colloidal particles, d is the center-to-center distances between the nearest particles, and θ is the angle between the incident light and the diffraction crystal planes 32 . The refractive index of PSt(237)@PDA(Y) particles can be calculated using Equation (1) assuming that Bragg’s law holds. As shown in Fig. 2e , the refractive index of PSt@PDA particles with thin shell layers (0–4.5 nm) was approximately 1.59, which is relatively the same as the refractive index of PSt 24 . The refractive index of the PSt@PDA particles will be depending on the volume averaged value of PSt and PDA. When PDA shell thickness were thin, the samples would be almost similar to PSt. In contrast, the refractive index of PSt@PDA particles increased to ca. 1.7 as the PDA shell thickness increased. The refractive index value of the barbules’ melanin, measured by the Jamin-Lebedeff interference microscopy method, was approximately 1.7-1.8 25 . While matching the refractive index of PSt@PDA with that of melanin has no physical meaning, that of the PSt@PDA particles with a thick PDA layer (>17 nm) was relatively the same as that of native melanin. These data clearly indicate the usefulness of the PDA coating method for controlling the particles’ refractive index as well as their blackness. The process of creating these pellets was investigated. A total of 100 μL of a water dispersion of PSt(237)@PDA(4.5) particles was dropped onto a silicone rubber plate using a micropipette, and the suspensions were allowed to dry. Structural color pellets were generally obtained at room temperature for 12 h. However, it takes a lot of time to measure the reflection spectra. Thus, in the present experiments, pellets were prepared at 50 °C for about half an hour. The reflection spectra of these dispersions were measured with a microscopic spectrophotometer as a function of the evaporation time. As shown in Fig. 3a , the λ max of the reflection peaks decreased as time increased, clearly indicating the blue-shift of the reflection spectra. Figure 3b shows a photograph of the dispersions. Before the water evaporated, dispersions of PSt(237)@PDA(4.5) particles (5 wt% in water) were light brown in color due to absorption. During the evaporation of the water, the colors changed, and finally, a blue-green structural color pellet was obtained. The center-to-center distances between the nearest particles ( d ), calculated using Equation (1) (n = 1.59), gradually decreased from approximately 347 nm as the evaporation time increased, because of the increase of particle concentrations. Finally, d reached a plateau at approximately 249 nm, which is consistent with the diameter of PSt(237)@PDA(4.5) particles (246 nm), indicating the formation of close-packed structures. By selecting the core size and feed concentration of DA monomers, we obtained pellets in a variety of colors ( Fig. 4a ). The reflectance spectra of each of the structural color pellets were measured by the microscopic spectrophotometer. The λ max values of the reflection peaks of the pellets were plotted as a function of the diameter of the core-shell particles ( Fig. 4b ). The λ max of the reflection peaks increased as the diameters of the core-shell particles increased. As the PDA shell thicknesses are controlled by the feed concentration of the DA monomer ( Fig. 2b ), the refractive indices of particles prepared under the same monomer conditions will be relatively similar. Thus, the theoretical lines in Fig. 4b were calculated from Equation (1) using the refractive indices for PSt(237)@PDA(Y) particles ( Fig. 2e ). The model values correspond reasonably well with the experimental data, suggesting that structural coloration was controlled by the refractive indices and the diameters of the particles. In Fig. 4c , the colors of each pellet are plotted on the International Commission on Illumination (CIE) 1931 chromaticity diagram labeled with the limits of the sRGB color gamut. By controlling the diameters of the core particles and PDA shell thicknesses, we were able to obtain nearly the full range of colors. Figure 4d shows the CIE chromaticity chart for the pellets prepared by PSt(237) core particles with different PDA shell thicknesses. The colors obtained indicated a gradual transition from the blue to the green to the red portions of the visible spectrum. Next, we investigated the effect of the PDA coating on the colorfulness of the structural colors. Figure 4e shows plots of the colors prepared by PSt core particles (gray plots) and PSt(X)@PDA(Y) core-shell particles (black plots, DA concentration = 0.5 mg/mL). After PDA coating, the colors moved to the outside of the plot, indicating the increase of colorfulness. These results strongly indicate the ability of the present method to create high-visibility structural color materials with a full color range. Figure 4a shows that two types of colors were obtained: iridescent and non-iridescent structural colors. When the PDA shell thickness was thin, the obtained pellets exhibited iridescent structural colors ( Fig. 5 ). In contrast, the core-shell particles with thick shell layers produced non-iridescent structural color pellets ( Fig. 5 ). This result is likely due to the arrangement of the particles. We performed SEM measurements of the pellet surface to investigate the arrangement of the particles ( Fig. 6a ). Iridescent structural colors were obtained from colloidal crystal structures, and non-iridescent colors were obtained from amorphous structures. Fourier transforms (FFT) of the SEM images of the pellets were investigated 14 16 21 . Sharp hexagonal peaks were observed for colloidal crystals, indicating the formation of close-packed structures. In contrast, circular patterns were produced by roughly packed arrays, forming amorphous structures. In the amorphous state, diffraction will be suppressed and the reflection spectra related to the size of the particles will be selectively enhanced, producing non-iridescent structural colors 14 . These results clearly indicated that two type of colors were observed for the different particle arrangements. To investigate the effects of thick PDA shell layers on the arrangement of particles, high resolution field emission-SEM (FE-SEM) images of PSt(285)@PDA(15) particles were taken ( Fig. 6b ). While the coefficient of variation (CV) of the particles remained approximately 3%, the PSt(285)@PDA(15) core-shell particles had rough surfaces. DA polymerization occurs in both the surface of the particles and in the solution phase. When the monomer concentration was high, DA polymerization primarily occurred in the water phase. The resulting DA oligomers would aggregate onto the particles, producing a rough surface. The surface roughness of these particles would prevent the production of colloidal crystal structures. Figure 6c,d shows the cross-section of the pellets. While close-packed structures were observed in PSt(285)@PDA(2.5) pellets ( Fig. 6c ), amorphous structures were obtained in PSt(285)@PDA(15) pellets ( Fig. 6d ), indicating the uniformity of the pellets. These results indicated that both iridescent and non-iridescent high-visibility structural colors were easily prepared using the same process. Therefore, this method enables the separate formation of colloidal crystal structures and amorphous colloidal structures by core-shell particles whose only difference is their shell thicknesses. The simplicity of our method facilitates the practical use of these particles because this method did not require any additives. Finally, we examined the effects of particle components on structural coloration. To investigate the effects of the ratio of the PSt region and the PDA region, four samples were prepared: PSt(237), PSt(237)@PDA(2.5), PSt(221)@PDA(9.5), and PDA(242) particles. While the four particles have relatively similar diameters, the colors of the pellets were different ( Fig. 7a ). As shown in Fig. 7b , the λ max of the reflection peaks, as measured by the microscopic spectrophotometer, increased as the PDA regions increased. The variation in the λ max of the reflection peaks is due to variations in the refractive index of the particles. Figure 7b also shows the refractive index of the particles as calculated using Equation (1) . The refractive index of the particles gradually increased as the PDA ratio increased. While it is difficult to accurately determine the refractive index for PDA, which is a strongly adsorbing polymer, the refractive index of PDA particles was calculated to be approximately 1.76, which is in agreement with a previous report 33 . For the refractive index, a similar trend is observed with λ max , strongly indicating that the structural coloration produced by colloidal particles depends on their refractive index. In conclusion, we developed both iridescent and non-iridescent high-visibility structural colors from biomimetic colloidal particles. This process was inspired by the structural color found in bird feathers. To create high-visibility structural colors, we designed and synthesized PSt@PDA core-shell particles that met the following four conditions: size, blackness, refractive index, and arrangement. By controlling the feed concentration of the DA monomers, core-shell particles with any size, blackness, and refractive index were easily prepared under mild conditions. Additionally, the surface roughness of the particles, which is an important factor for controlling the arrangement of the particles, was adjusted by varying the PDA thickness. While PSt@PDA particles with smooth surfaces produced colloidal crystal structures, PSt@PDA particles with rough surfaces produced amorphous structures. As a result, both iridescent and non-iridescent structural colors were separately prepared from only one component. Because PSt@PDA core-shell particles have properties similar to those of melanin, this method will be useful for understanding biological systems in nature. Furthermore, this method enables the production of high-visibility structural colors with a simple process. Additional studies in progress include the development of structural color inks based on this methodology. The results of these investigations will be described in future reports."
} | 3,840 |
33998118 | null | s2 | 3,465 | {
"abstract": "Vibrio campbellii BB120 (previously classified as Vibrio harveyi) is a fundamental model strain for studying quorum sensing in vibrios. A phylogenetic evaluation of sequenced Vibrio strains in Genbank revealed that BB120 is closely related to the environmental isolate V. campbellii DS40M4. We exploited DS40M4's competence for exogenous DNA uptake to rapidly generate greater than 30 isogenic strains with deletions of genes encoding BB120 quorum-sensing system homologues. Our results show that the quorum-sensing circuit of DS40M4 is distinct from BB120 in three ways: (i) DS40M4 does not produce an acyl homoserine lactone autoinducer but encodes an active orphan LuxN receptor, (ii) the quorum regulatory small RNAs (Qrrs) are not solely regulated by autoinducer signalling through the response regulator LuxO and (iii) the DS40M4 quorum-sensing regulon is much smaller than BB120 (~100 genes vs. ~400 genes, respectively). Using comparative genomics to expand our understanding of quorum-sensing circuit diversity, we observe that conservation of LuxM/LuxN proteins differs widely both between and within Vibrio species. These strains are also phenotypically distinct: DS40M4 exhibits stronger interbacterial cell killing, whereas BB120 forms more robust biofilms and is bioluminescent. These results underscore the need to examine wild isolates for a broader view of bacterial diversity in the marine ecosystem."
} | 354 |
21971585 | PMC3277704 | pmc | 3,468 | {
"abstract": "The spatial scale of disturbance is a factor potentially influencing the relationship between disturbance and diversity. There has been discussion on whether disturbances that affect local communities and create a mosaic of patches in different successional stages have the same effect on diversity as regional disturbances that affect the whole landscape. In a microcosm experiment with metacommunities of aquatic protists, we compared the effect of local and regional disturbances on the disturbance–diversity relationship. Local disturbances destroyed entire local communities of the metacommunity and required reimmigration from neighboring communities, while regional disturbances affected the whole metacommunity but left part of each local community intact. Both disturbance types led to a negative relationship between disturbance intensity and Shannon diversity. With strong local disturbance, this decrease in diversity was due to species loss, while strong regional disturbance had no effect on species richness but reduced the evenness of the community. Growth rate appeared to be the most important trait for survival after strong local disturbance and dominance after strong regional disturbance. The pattern of the disturbance–diversity relationship was similar for both local and regional diversity. Although local disturbances at least temporally increased beta diversity by creating a mosaic of differently disturbed patches, this high dissimilarity did not result in regional diversity being increased relative to local diversity. The disturbance–diversity relationship was negative for both scales of diversity. The flat competitive hierarchy and absence of a trade-off between competition and colonization ability are a likely explanation for this pattern. Electronic supplementary material The online version of this article (doi:10.1007/s00442-011-2140-8) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction Disturbance has long been proposed as a mechanism enabling the coexistence of species, with a unimodal relationship between diversity and the level of disturbance predicted by the intermediate disturbance hypothesis (IDH; Connell 1978 ). Diversity is expected to be low in both weakly and strongly disturbed systems through competitive exclusion and extinction of disturbance-intolerant species, respectively. Diversity should be maximized by intermediate disturbances that disrupt competitive displacement before elimination of weak competitors but do not drive disturbance-intolerant species to extinction. However, only a minority of experimental and observational studies on the disturbance–diversity relationship reports the expected unimodal pattern (reviewed in Mackey and Currie 2001 ). One factor possibly influencing the disturbance–diversity relationship is the spatial scale of the disturbance. There has been discussion on whether the IDH is a between-patch mechanism relying on the presence of a mosaic of patches in different successional stages (Wilson 1994 ), or works both within a single patch and between patches (Collins and Glenn 1997 ). In a modeling approach, Roxburgh et al. ( 2004 ) compared between-patch and within-patch disturbances. Between-patch disturbances completely destroyed patches within a landscape and required recolonization by reimmigration from outside the disturbed patch, while within-patch disturbances affected the whole landscape but left residuals behind. This study demonstrated that coexistence via disturbance of intermediate frequency is possible for both disturbance types, as long as the species show a trade-off between the ability to recolonize the disturbed patch and competitive ability. However, the two disturbance types require different traits for recolonization. Spatially patchy disturbances enable species coexistence when the inferior competitor is better at dispersing to and colonizing a recently disturbed patch. Theoretically, such a trade-off between competition and colonization ability can enable coexistence of an unlimited number of species, assuming that inferior competitors have sufficiently higher colonization abilities than superior competitors (Tilman 1994 ). Global within-patch disturbances require other traits for recolonization. Such disturbances favor species that either have higher maximal growth rates enabling them to profit from the resource-rich conditions after the disturbance (Pacala and Rees 1998 ; Rees et al. 2001 ), or that leave more residuals behind due to production of resting eggs or cysts, vegetative resprouting, higher survival in the seed bank, protective structures or less susceptible growth forms (Sousa 1984b ; Pake and Venable 1996 ; Cáceres 1997 ; Turner et al. 1998 ). Both types of disturbance are common in nature. Examples for spatially patchy disturbances are wave disturbances in intertidal boulder fields or mussel beds (Sousa 1979 , 1984a ; Paine and Levin 1981 ), mound building by burrowing animals in grasslands (Platt 1975 ; Guo 1996 ; Questad and Foster 2007 ), and small floods creating a mosaic of differently disturbed patches for river invertebrates (Matthaei et al. 1999 , 2000 ). Examples for within-patch disturbances are droughts in marshes and large chaparral fires with subsequent colonization from the seed bank or by vegetative resprouting (Keeley et al. 1981 ; Bonis et al. 1995 ). However, differences between these two disturbance types are not always unequivocal and different groups of organisms may experience the same disturbance as patchy and as global, respectively. For example, some species survive the drying of temporary ponds as resting stages, while others have to reimmigrate from permanent ponds after refilling (Wellborn et al. 1996 ; Chase 2007 ). It is therefore a within-patch disturbance for the former, but a spatially patchy disturbance for the latter species. Microcosm experiments with aquatic protists have often been used to study the effect of disturbance on diversity. However, studies have examined either disturbances within single patches (e.g., Scholes et al. 2005 ; Haddad et al. 2008 ; Jiang and Patel 2008 ) or disturbances affecting local communities within a metacommunity, thereby creating a mosaic of patches in different successional stages (Warren 1996 ; Östman et al. 2006 ; Cadotte 2007 ). To our knowledge, the present experiment is the first that directly compares the effect of these two spatial scales of disturbance on diversity. In a microcosm experiment, we assembled metacommunities consisting of four local communities connected by active dispersal of organisms. Disturbances that reduced abundances of all species in a non-selective way and introduced fresh resources were applied at three intensities and at two spatial scales. Disturbances at the local scale resulted in a complete destruction of selected local communities and required reimmigration from neighboring communities for recolonization. Regional disturbances, however, affected the whole metacommunity but left part of each local community intact. We hypothesized that regional disturbances would have a less severe effect on diversity, as they allowed recolonization not only through reimmigration but also through growth within a patch. Moreover, we predicted that local disturbances would have a positive effect on beta diversity by creating a mosaic of patches in different successional stages and would thus increase regional diversity relative to local diversity. We further predicted that species with high growth rates would be favored by intense regional disturbances, while both high growth and high dispersal rates would be important for survival under strong local disturbances.",
"discussion": "Discussion High intensity disturbances reduced diversity, irrespective of the spatial scale of disturbance. However, the mechanism behind this decrease in diversity was clearly dependent on the scale of disturbance. Strong local disturbance decreased diversity by reducing species richness, while strong regional disturbance decreased diversity by reducing evenness (Fig. 1 ). Apart from different effects on the two components of diversity, spatial scale of disturbance also affected the speed of diversity decline. Species extinctions due to strong local disturbances occurred even early in the experiment, while the decrease in evenness with high regional disturbance occurred only at the end of the experiment. We hypothesized that local disturbances would have a stronger effect on diversity than regional disturbances. The confirmation of this hypothesis depends on whether one gives more weight to species richness, evenness, or the combination of both. Species extinctions in the regional species pool may indeed have a more severe effect on the community since there is no chance for recovery. Conversely, with reduced evenness, all species are still present in the species pool and recovery to a community of higher diversity is possible. By reducing species richness, local disturbances can thus be regarded as having a stronger effect on the community. Disturbance scale and species richness The reasons for species loss with strong local disturbance can be elucidated when looking at the species’ traits (Table 2 ). While a regression analysis between species’ traits and persistence in disturbed treatments was not possible due to the low number of data points, some qualitative comparisons between species with different responses to disturbance can be made. With strong local disturbance, only the two species with the highest growth rates survived ( Stylonychia , Tachysoma ), while the slow-growing species rapidly went extinct. A microcosm experiment analyzing the importance of several species’ traits in predicting the response to within-patch disturbances also found growth rate to be the best predictor for persistence under disturbance (Haddad et al. 2008 ). Interestingly, a high growth rate was also the most important trait for survival in our between-patch disturbance treatments, while dispersal rate was unimportant. Paramecium , one of the species that went extinct, had by far the highest dispersal rate in our model community, while one of the survivors ( Tachysoma ) had the lowest dispersal rate. Within our experimental set-up, the ability to rapidly build up large populations seems to have been more important than fast dispersal to undisturbed patches. However, the importance of dispersal rate for survival with between-patch disturbances will probably depend on the distance between disturbed patches. With greater distance, a high dispersal rate will become increasingly important. With strong regional disturbance, all four species survived, demonstrating that this type of disturbance is a less strong environmental filter than local disturbance. Presence or absence of surviving organisms and propagules has also been found to strongly influence post-disturbance succession in many natural systems (Turner et al. 1998 ; Turner 2010 ). Succession in forests and other plant communities is more predictable and re-establishment of the pre-disturbance community is faster when residuals are present, as more species are able to recolonize a patch where they still have propagules or individuals present than when they have to recolonize a distant patch by reimmigration. Despite the simplicity of our model system, the results of our experiment are thus comparable to observations in natural communities. Disturbance scale and evenness While regional disturbances had no effect on species richness, they led to a decrease in evenness due to strong dominance of the fastest-growing species. It has been suggested that inferior competitors with high maximal growth rates are able to temporally dominate after disturbances of small spatial extent, as they are best at exploiting the resource-rich conditions after the disturbance (Pacala and Rees 1998 ; Rees et al. 2001 ). Accordingly, Tachysoma , the species with the highest growth rate in our community, reached high abundances with strong regional disturbance (Figs. 2 c and 3 d). In our experiment, disturbances increased mortality but introduced fresh resources, thereby creating conditions that were best exploited by species with high growth rates. In contrast, when disturbances reduce nutrients, it is particularly the fast-growing species that are negatively affected since their growth rates are more severely reduced than those of slower-growing species (Haddad et al. 2008 ). The effect of disturbance on the dominance structure of the community will thus depend on whether disturbances increase or decrease productivity. We only tested disturbances increasing resource concentrations, a type of disturbance common in nature. Examples are the input of nutrient-rich river water into floodplain lakes during flood events (Van den Brink et al. 1993 ), increased light availability within treefall gaps (Denslow 1987 ), and higher palatability of early-successional plant species that establish after disturbances (Cates and Orians 1975 ). While Tachysoma reached high absolute abundance with strong regional disturbance early in the experiment, its relative abundance increased and thus evenness decreased only during the final 2 weeks, concurrent with a sharp decline in Stylonychia abundance (Fig. 2 d). Stylonychia gained high abundances with strong local disturbance due to high growth and dispersal rates, its low abundance with strong regional disturbance in the end of the experiment is thus surprising. Even more, since it was also the strongest competitor in the community and dominated in the undisturbed control. However, the competitive hierarchy was rather flat (Table 2 ) and an earlier experiment has shown temporal variability of competitive effects. We found Tachysoma to have negative effects on Stylonychia during the early phase of succession (Limberger and Wickham 2011 ). With strong regional disturbance, succession was regularly set back to early-successional and resource-rich conditions, and competitive interactions might thus be the reason for the reduction of Stylonychia abundance. Also, the decline of Stylonychia abundance in the weakly and undisturbed treatments during the intermediate period of the experiment probably is a result of competition. The mid-successional species Frontonia was found to have strong negative effects on Stylonychia (Limberger and Wickham 2011 ). Our results show that regional disturbances that increase productivity of the system lead to high dominance of the fastest-growing species. Despite a similar input of resources with strong local disturbance, the double requirement of dispersal and growth with high local disturbance prevented the slow disperser Tachysoma from reaching similar dominance as with high regional disturbance. The two survivors attained nearly equal abundances, leading to high evenness. Thus, while the requirement of dispersal seemed to be unimportant for survival with strong local disturbance, it did affect the abundances of the surviving species. Results on species’ responses to disturbance treatments could have been influenced by trait evolution. There is increasing empirical evidence that rapid evolution affects ecological dynamics (Hairston et al. 2005 ; Fussmann et al. 2007 ). Accordingly, treatments with high disturbance intensity could have selected for individuals with high growth rates. With our data, however, we cannot assess whether evolutionary processes influenced species’ responses to disturbance. Measuring species’ growth rates at the beginning and the end of disturbance experiments would be an interesting avenue for future studies. Scale of disturbance not only influenced species richness, evenness, and species’ abundances, it also affected resource biomass (Fig. 4 ). Although strong local and regional disturbances both introduced the same amount of resources, algal biomass was higher with local disturbance. The probable reason was a negative effect of strong local disturbance on ciliate abundance, and since only the two smallest species survived, the effect on ciliate biomass was even larger (Fig. 2 b). This negative effect on grazer biomass translated to the lower trophic level by leading to a less efficient resource use. Local, regional, and beta diversity Disturbance scale had a strong influence on beta diversity (Fig. 1 d). Regional disturbances affected each local community to the same extent and thus resulted in high similarity of local communities. While we also expected high similarity of local communities in the undisturbed control since each local community was treated the same (i.e. not at all), similarity between undisturbed local communities decreased during the second half of the experiment. Similarly, Chase ( 2007 ) found drought disturbance to result in high similarity of pond communities, while species composition in undisturbed ponds was more variable. He suggested stochastic effects to be more important in undisturbed communities, leading to higher dissimilarity of communities, while disturbances provide a harsh environmental filter and increase the importance of niche-selective forces. In contrast to regional disturbances, local disturbances led to an increase in beta diversity by creating a mosaic of differently disturbed patches. With strong local disturbance, this increase was only temporal and beta diversity quickly decreased as soon as the disturbance-intolerant species had gone extinct. Weak local disturbances were applied only every other week and resulted in pronounced fluctuations of beta diversity, with dissimilarity depending on the time since last disturbance. The amplitude of the fluctuations decreased with increasing time of the experiment, when slow colonizers had either died out or built up populations large enough to rapidly colonize disturbed patches. Since regional diversity depends on both local and beta diversity, the high dissimilarity of local communities with local disturbance let us expect an increase of regional relative to local diversity. However, there were only marginal differences between local and regional diversity: with strong local disturbance, the initial decline in species richness was faster for local than for regional richness. By comparing absolute species’ abundances of the four local communities, Bray–Curtis dissimilarity probably detected fine-scale differences that were not appearent when comparing local and regional richness or diversity. Other studies comparing the effects of local disturbances on local, regional, and beta diversity also found a negative effect of disturbance on local diversity (Warren 1996 ; Östman et al. 2006 ; Cadotte 2007 ). However, results for dissimilarity and regional diversity are inconsistent. Similar to our experiment, Warren ( 1996 ) found a negative effect of disturbance on regional diversity despite a positive effect on beta diversity, at least under low dispersal rates. The reason was that disturbance reduced local richness to a stronger extent than regional richness. Therefore, strong disturbances resulted in a comparatively large difference between local and regional diversity, and thus high beta diversity. In contrast, an increase of both beta diversity and regional diversity with at least intermediate disturbance has also been found (Östman et al. 2006 ; Cadotte 2007 ). Disturbance–diversity relationship We found a negative relationship between disturbance intensity and diversity, irrespective of the spatial scale of disturbance and of whether local or regional diversity was regarded. One basic requirement for a unimodal relationship is the presence of a strong competitive hierarchy. When comparing communities with strong versus weak competitive hierarchy, Svensson et al. ( 2009 ) found a unimodal disturbance–diversity relationship in the former, but negative or nonsignificant patterns in the latter communities. In our model community, the competitive hierarchy was rather flat, with large variability of competitive ranks (Table 2 ). No competitive exclusions occurred in the undisturbed control, yet, without competitive displacement at undisturbed conditions, an increase in species richness with intermediate disturbances is not possible. Apart from a strong competitive hierarchy, the presence of a trade-off between colonization and competitive ability is a prerequisite for the IDH to be operating (Haddad et al. 2008 ). Even the presence of such a trade-off does not guarantee a unimodal disturbance–diversity relationship when the number of good colonizers and good competitors is not balanced (Cadotte 2007 ). In our community, however, the species that were best at colonizing new habitats due to their high growth rates were also the species that finally dominated in the undisturbed control. Thus, succession was rather circular instead of serial, with the same species dominating at the beginning and at the end of succession. Given the flat competitive hierarchy, lack of a trade-off between competition and colonization ability and absence of a serial succession, the negative disturbance–diversity relationship is unsurprising. A model community with more species might have been a more robust test of the IDH, but this was not the main aim of our study. Rather than describing the relationship between disturbance intensity and diversity, we wanted to test whether the pattern of this relationship differed between local and regional disturbances. Our results suggest that local disturbances provide a harsher environmental filter that fewer species are able to pass than regional disturbances."
} | 5,428 |
34502980 | PMC8434175 | pmc | 3,470 | {
"abstract": "Controlling the residence time of drops on the solid surface is related to a wide spectrum of engineering applications, such as self-cleaning and anti-icing. The symmetry-breaking dynamics induced by the initial drop shape can promote drop bouncing. Here, we study the bouncing features of spherical and ellipsoidal drops on elliptical surfaces that continuously change curvatures inspired by natural succulent leaves. The bounce characteristics highly depend on the geometric relations between the ellipsoidal drops and curved surfaces. Numerical results show that ellipsoidal shapes of the drops amplify asymmetries of the mass and momentum in synergy with an influence of the surface curvature during the impact, which is verified by experiments. Effects of the surface anisotropy and drops’ ellipticity on the residence time are investigated under various surface morphologies and Weber numbers. The residence time is closely associated with an initial surface curvature at the apex. The underlying principle of modifying the residence time via the drops’ ellipticity and initial surface curvature is elucidated based on momentum asymmetry. The understanding of the bouncing features on curved surfaces will offer practical implications for enhanced heat transfer performances and controlled water repellency, etc.",
"conclusion": "4. Conclusions We investigated the bouncing features and t c of ellipsoidal drops on the E -surfaces, compared with spherical drops. The numerical results revealed that the dynamics highly depended on the geometric configuration between the elliptical curves and ellipsoidal drops. The t c obtained from the different geometric relations could be determined by the hydrodynamic interplay between the influence of the initial drop shape and surface curvature. To better understand the roles of the two factors in drop dynamics, we investigated the evolutions of shapes and momentum asymmetries of the drops on various E -surfaces, which revealed that the rapid bouncing appeared at high κ 0 and high e . t c of e + drops on E (2.0, 4.0) surfaces could diminish by approximately 46% below the spherical cases on flat surfaces at We = 24, which presented the enhancement of asymmetries of the mass and momentum. t c of e − drops generally increased with |e| on surfaces with increasing κ , whereas those were independent of e on surfaces with decreasing κ , in general. In addition, we found that the residence time could be closely related to the maximal asymmetry of the momenta, which showed that τ c decreased roughly linearly against ( p t – p z ) m for the spherical and e + drops. The drops’ ellipticity could have the ability to adjust the repellency from surfaces by covering a broad range of the residence time beyond the spherical cases. We believe that reshaping of the ellipsoidal drops will provide new insight into the strategies for further lower retention on bioinspired surfaces, such as an array of cylinders or corrugated surfaces. The fundamental understanding of the drop dynamics will be able to assist practical applications, such as dropwise condensation, anti-corrosion, and anti-icing.",
"introduction": "1. Introduction Bouncing dynamics of drops on solid surfaces have gained substantial attention over the last two decades for industrial applications, such as self-cleaning [ 1 ], anti-icing [ 2 ], low friction [ 3 ], and dropwise condensation [ 4 ]. The bounce characteristics are highly dependent on the surface roughness, temperature, wettability, and ambient conditions [ 5 , 6 ]. The residence time is considered essential because it determines the extent to which mass, momentum, and energy are exchanged between surfaces and drops. Drops impacting on superhydrophobic surfaces can lift off quickly because of the low wetting hysteresis at the contact line, inspired by effects of the lotus leaves and pitcher plant [ 7 , 8 ]. The residence time can be shortened to the inertio-capillary time scale of ( ρD 3 / σ ) 1/2 with circular symmetry, where ρ , D , and σ are the density of liquid, diameter of drop, and interfacial tension, respectively [ 9 , 10 , 11 ]. The emphases of recent works focus mainly on compelling drops to depart from the surfaces as fast as possible. Many efforts have been devoted to further reducing the residence time by utilizing surface textures and chemical compositions. The surface morphologies were relevant to the sub-millimeter textures decorated on flat substrates, including center-assist bouncing along macro-ridge structures [ 12 , 13 ] and complicated cross-shaped macro-textures [ 14 ], counter-intuitive bouncing on lattice patterns of posts with nanostructures [ 15 ], and water-ring bouncing on point-like defects [ 16 ]. The research progress in this field developed some fresh regimes to enhance drop mobility based on the surface modifications. Along with the development, recent works have demonstrated asymmetric bouncing dynamics on surfaces with the radius of curvature comparable to drop size, including asymmetric bouncing on tubular surfaces and natural succulent leaves [ 17 ], ribbed-curved surfaces [ 18 ], and cylindrical ridges in millimetric size [ 19 , 20 ]. The shape and size of the cylindrical ridges altered the drops’ behavior significantly, and hence the residence time could reduce further, compared with the dynamics on the flat surfaces. Particularly, there are two distinct regimes where the variation of the ridge in size has reverse effects on the residence time [ 19 ]. For cylindrical ridges greater than the drop size [ 17 ], the mass and momentum redistribute in the manner of a hydrodynamic interplay between the momenta in the tangential direction along the curve side and axial direction. The lift-off time further reduced at higher curvatures because the bouncing dynamics were closely related to the anisotropic curvature of the surface. On the contrary, when the ridge is much smaller than the drop size [ 12 , 13 ], a role of the flat substrate on which the ridge laid would be played in an interaction with the drops. After the impact, the drop split into two parts that retracted and bounced separately. Further demonstrations on Y-shaped or cross-shaped macro-texture surfaces reported that drops could be configured into several subunits and take off at the reduced residence time [ 13 , 14 ]. The lift-off time decreased as the ridge size increased in this regime [ 19 ]. Another methodology for the control of the residence time with non-spherical shaping has been suggested by our group [ 21 ]. The symmetry-breaking dynamics induced by the initial drop shape could potentially open up opportunities for modifying the impact dynamics. Ellipsoidal shapes allowed the peculiar spreading and retraction behaviors, thereby forming liquid alignment with the principal axis during retraction [ 21 ]. When the ellipsoidal drops were colliding on the heated [ 22 ] and superhydrophobic surfaces [ 23 ], the drops produced the preferential flow to the minor axis of the initial elliptical footprint, which led to the decrease in the residence time and bounce height. The distinguishable features of shape-dependent dynamics could significantly change the outcome of impact without target surface modification or additional chemical composition of the liquid. However, most of the previous works focused their attention solely on the scenarios of drop impact on flat substrates, and the effect of the initial drop shapes on hydrodynamics on anisotropic surfaces has yet to be explored [ 21 , 22 , 23 ]. Furthermore, there was a lack of knowledge on how shape distortions of impinging drops had an influence on the bouncing characteristics on the surfaces in practical spraying systems. Initial drop shapes would potentially amplify asymmetries of the mass and momentum in synergy with an influence of the surface curvature, which can alter the residence time. The recent study of our group investigated the bouncing features of the ellipsoidal drops on superhydrophobic cylinders and reported a further decrease in the residence time, compared with spherical drops impacting on the surfaces [ 24 ]. However, the latter study showed the shape-dependent dynamics on curved surfaces that were limited to circular cylinders. The knowledge may still be insufficient for practical applications, such as self-cleaning and biomimetic strategies. The current work was motivated by the symmetry-breaking bouncing on the Echeveria succulent leaves [ 17 ], which can be represented as the bouncing on elliptically curved surfaces in a more realistic situation, as captured in Figure 1 a. The other leaves we found also exhibited the curved surfaces that change curvature along the surfaces, as shown in Figure 1 b. To understand how the initial drop shape and surface curvature affect the dynamics, we study the bouncing features of the drops on elliptically curved surfaces ( E -surfaces). The current work focuses on the above-mentioned hydrodynamic regime where the drops never contact the flat surface on which the ridge laid, and the ridge width is equal to or more than two-fold of the drop size. The influences of the drops’ ellipticity ( e ) and surface curvature ( κ ) on the residence time are scrutinized under various surface curvatures and impact velocities. In the momentum analysis, we discuss how the momentum asymmetry is relevant to the decrease in the residence time.",
"discussion": "3. Results and Discussion We established the model of E ( a , b ) surfaces that represent the semi-elliptic surfaces with the semiaxes of a and b , as shown in Figure 1 c. The bouncing dynamics are highly affected by the geometric configurations between curved surfaces and drops. The ellipsoidal drops have the ellipticities of e + ( e > 0) and e − ( e < 0) when arranged to be the major axes of the z and x , respectively. Figure 1 d indicates dimensionless surface curvatures that are derived from κ ( x , y ) = ( ab ) 4 /( b 4 x 2 + a 4 y 2 ) 3/2 , and the value at the apex is equal to an initial surface curvature κ 0 = b / a 2 . Distinct from circular cylindrical surfaces with the constant κ ( a = b ), the curvatures of the E -surfaces are increasing ( a > b ) or decreasing ( b > a ) along the x axis. The color bar shown in Figure 1 d means the magnitude of curvature, which will be used as the color of the E -surface later. To understand the effects of the drops’ ellipticity and surface curvature on the bouncing behavior, we studied the temporal evolutions of the drops under b = 1.2 ( b 1.2) and 2.8 ( b 2.8), as shown in Figure 2 a,b. e + drops spread wider along the x axis and leave the surface earlier than the spherical and e − drops on the surfaces. In addition, all the drops on the b 2.8 surfaces exhibit further extensions along the x axis and bounce off the surface faster, compared with the drops on the b 1.2 surfaces, as shown in Figure 2 a,b at 6 ms. The bouncing characteristics are confirmed by the temporal variations of the x and z widths and the detachment times pointed out by using the single-circle symbols for each solid line, as shown in Figure 2 c. On the flat surfaces, as the limiting case ( κ 0 = 0), the e + ( e − ) drops have the hydrodynamic features of the spreading and retraction behaviors and the subsequent liquid alignment on the x axis ( z axis) before bouncing, as shown in Figure 2 d. The initially ellipsoidal shapes induce the predominant outward flow to the direction of the minor axis of the elliptical footprint during the impact. Thus, the e + drops stretch wider to the x axis than the z axis. In addition, the maximal extension of the x axis ( w xm ) is found later and greater than that of the z axis ( w zm ). Figure 2 e represents the maximal extensions ( w m ) as a function of e , which shows that the drops have significant variations in w xm , but only slight variations in w zm in the simulation and experiment. The maximum relative error of the maximal extensions between numerical and experimental results is within 6%. We investigated the effects of the surface anisotropy on the residence time of the drops. The t c on the flat and a 2.0 surfaces (i.e., E (2.0, b ) surface) are plotted in Figure 3 a as a function of e , which reveals that rapid bouncing is found at high b/a and e . The e + drops provide an efficient pathway to reduce t c on E -surfaces, which presents a striking contrast to e − drops. On the b 1.2 surfaces, for example, t c of the e + drops could reduce by approximately 19% and 40% below the spherical cases on the b 1.2 surfaces and flat surfaces, respectively. On the b 4.0 surfaces, t c of the e + drops could reduce by approximately 16% and 46% below the spherical cases on the b 4.0 surfaces and flat surfaces, respectively. By contrast, t c of the e − drops generally increases with |e| on surfaces with increasing κ ( a > b ), whereas those are generally independent of e on surfaces with decreasing κ ( b > a ). For instance, t c of the e − drops increases up to 7.7 and 6.0 ms on the b 1.2 and b 2.8 surfaces, which correspond to increases of approximately 22% and 3% above the spherical cases on the same surfaces, respectively. Distinctively, t c of the e − drops on the b 0.4 surfaces has a peak value at low |e|. This result is because the roles of the initial drop shape and surface curvature in symmetry-breaking in drop bouncing might be comparable to each other at the peak. Residence times on the curve surfaces result from a hydrodynamic interplay between the flow induced by effects of the drops’ initial shape and surface curvature. Assuming that e + drops are impacting on flat surfaces, the e + shapes induce a pronounced flow in the x axis, which could be intensified on the curved surfaces by the positive influence of the tangential momentum ( p t ) due to anisotropy of the surface. Conversely, assuming that e − drops are colliding on flat surfaces, the e − shapes induce the pronounced flow in the z axis, which could be suppressed on the curved surfaces by the negative influence of the p t that is orthogonal to the axial momentum ( p z ). Figure 3 b shows t c of the e + and e − drops as a function of b under several We . As We increases, t c on the b 1.2 surfaces increases up to 8.2 ms for e − drops and decreases down to 4.8 and 6.1 ms for e + and spherical drops, respectively. In addition, t c of the e − drops has peak values on the b 1.2 surfaces at the given Weber numbers. As the height, b, increases, the role of the surface curvature in the asymmetric momentum transfer becomes dominant, thereby causing a monotonous decrease in t c of all the drops. To interpret the mechanism of reducing the residence time, the drop dynamics were analyzed in terms of the axial momenta. Figure 4 shows the dimensionless momenta in the normal to the surface ( p n ), tangential along the ridgeline ( p t ), and axial directions ( p z ) of the elliptic cylindrical coordinates, and y -directions ( p y ). The signs of p were imposed in the spreading (positive) and the retraction processes (negative) based on its definition. The insets represent snapshots at the distinct times. After touching the surfaces, the spherical drops enhance p t and p z and reach the maximum values, p tm and p zm , in the spreading process at nearly 1 ms, as shown in Figure 4 a. Thereafter, the p t and p z become negative values at 3.5 and 2.6 ms, the onset times for the retraction process along their own directions, respectively. The gap between p t and p z could be closely relevant to the asymmetric mass and momentum transfer because a significant gap leads to the massive transfer in one direction and the subsequent liquid alignment with the same direction, as depicted in snapshots of Figure 4 a at 6.0 ms. After this time, the drop leaves the surfaces at 6.3 ms, as pointed out by the single-circle symbol for each line. e + drops expand the discrepancy between p t and p z and complete the liquid alignment, as shown in Figure 4 b, at 4.0 ms. The drops enhance the asymmetries of the mass and momentum, thereby leading to fast bouncing at 5.1 ms, which is in marked contrast to e − drops. This is because the e − drops display a minor difference between p t and p z , even until nearly 7.0 ms, and are thereby detached from the surfaces at 7.7 ms, as presented in Figure 4 c. The switching times of the p t and p z required to change from the spreading to retraction process are comparable to each other for the e − drops. The snapshots of the inset at 4–8 ms indicate that the retraction dynamics are approximately axisymmetric, comparable to the features of the spherical drops. Figure 4 d shows the temporal evolutions of p n and momentum asymmetry ( p t – p z , in the inset) on the flat and E -surfaces, which reveals that the e + drops show the rapid rising of p n and ( p t – p z ) at the earliest time and detach from the surfaces at the lowest t c among the several drops. The momentum asymmetry of the drops shown in the inset of Figure 4 d is also closely connected to the bounce speed. The enhanced asymmetry of the momenta is found at high values of | p t – p z |. Asymmetry of the spherical and e + drops reach the maximum values of ( p t – p z ) m , as indicated by the arrow of the inset. Meanwhile, the maximal asymmetry of the e − drops on the b 1.2 (increasing κ ) surfaces occurs during the spreading because of the dominant role of the drops’ initial shape in the hydrodynamics at the low surface curvatures. We suggest that the momentum asymmetry can be convincing evidence for the decrease in the residence time. Details in the velocity fields for ellipsoidal drops are described in Figure S1 of the Supplementary Materials . To better elucidate the effects of the height, b, on the residence times, we examined the evolutions of the shapes on b 0.8 and b 4.0 surfaces. Figure 5 a shows that e + drops on the b 4.0 surfaces exhibit severe elongation to the x axis, compared with those on the b 0.8 surfaces, as shown at 6 ms. On the contrary, e − drops on the b 0.8 surfaces behave as roughly axisymmetric by evolving into the vertical liquid column before bouncing, whereas those on the b 4.0 surfaces massively redistribute to the x axis and then depart from the surfaces early. Figure 5 b,c indicate the temporal variation of the axial momenta of the ellipsoidal drops on b 0.8 (solid line) and b 4.0 surfaces (dashed line). Figure 5 d shows the temporal variations of p n and ( p t – p z ) for several drops. The solid and dashed lines correspond to the momenta of the e + and e − drops, respectively. The discrepancy between p t and p z of the e + drops on the b 4.0 surface is more intensified than that on the b 0.8 surfaces. Meanwhile, the discrepancy of the e − drops on the b 0.8 surfaces has negative signs because the role of the drops’ initial shape is dominant in the hydrodynamics at the low surface curvatures, as discussed earlier. The e + drops on the b 4.0 surfaces show the fastest growths of p n and ( p t – p z ) among several drops. As the height, b, increases, the initial surface curvature and the momentum asymmetry increase. Figure 6 a shows t c of the ellipsoidal drops with e = ±0.45 and spherical ones (insets) under three different a . The t c of the e − drops has peaks around b = 1.2, 2.7, and 4.0 at a = 2.0, 3.0, and 4.0, respectively. When t c is plotted by the initial surface curvature, κ 0 , the t c is roughly gathered around one line for each drop, as shown in Figure 6 b. Thus, residence times could be characterized in terms of κ 0 . The figure reveals that the drops have the only slight deviation of t c between the surfaces with the different widths at high κ 0 , in spite of the relatively high deviation of t c at low κ 0 . Figure 6 c shows shape evolutions of the drops on E (2.0, 0.4) and E (4.0, 1.6) surfaces, commonly with κ 0 = 0.1. The figure indicates that the ellipsoidal drop dynamics on the two surfaces are similar to each other. Accordingly, the conditions of the high κ 0 and high e play an important role in the symmetry-breaking in the mass and momentum distribution. Temporal variations of the momentum symmetries on the several E -surfaces with κ 0 = 0.1, including E (10, 10) surfaces, were comparable to each other, which is described in Figure S2 of the Supplementary Materials . The residence time can be closely related to the maximal asymmetry of the momenta. We plotted the normalized residence time, τ c , with P sp , where P sp = ( p tm – p zm ), the difference between the maximum momenta within the spreading process, as shown in Figure 7 a. The solid lines and symbols represent the drops’ several ellipticities ranging from −0.53 to +0.53. The inset represents the plot of P sp as a function of κ 0 , which reveals that P sp has a significant value at high e and high κ 0 . Additionally, the τ c is generally in inverse proportion to P sp , partly satisfying the slope of −0.45 at very high P sp . However, when the τ c is plotted by P wh , it is well-fitted by the single lines with slopes of −0.55 and 1.1 at the positive and negative P wh respectively, as shown in Figure 7 b, where P wh = ( p t – p z ) m , the maximum difference between the momenta within the whole spreading and retraction processes. The spherical and e + drops hold the slope of −0.55. Meanwhile, the e − drops hold the slope of 1.1, retaining the negative P wh at low κ 0 , as shown in the inset of Figure 7 b. Therefore, the maximal asymmetry, P wh , can serve to determine the reduction in τ c by emphasizing the richness of the physics of how surface curvatures and initial drop shape affect the bouncing features. We also concluded that the drops’ ellipticity is considered capable of controlling the drop mobility by offering a broader range of the τ c (approximately from 0.5 to 0.9), compared with the spherical case ( τ c ≈ 0.6–0.8). As an additional note, we found the shape evolution of the oblate drops on curved surfaces to compare with those of the prolate spheroidal shapes, which showed that the dynamics of the prolate and oblate drops were similar. In addition, we also predicted the bouncing dynamics of ellipsoidal drops that were rotated 45° on the y axis, and indicated that the symmetry-breaking bouncing on E -surfaces was also significantly affected by anisotropic curvatures. Details of the impact dynamics were described in Figures S3 and S4 of the Supplementary Materials ."
} | 5,696 |
24774563 | null | s2 | 3,477 | {
"abstract": "Once seen as anomalous, facilitative interactions among plants and their importance for community structure and functioning are now widely recognized. The growing body of modelling, descriptive and experimental studies on facilitation covers a wide variety of terrestrial and aquatic systems throughout the globe. However, the lack of a general body of theory linking facilitation among different types of organisms and biomes and their responses to environmental changes prevents further advances in our knowledge regarding the evolutionary and ecological implications of facilitation in plant communities. Moreover, insights gathered from alternative lines of inquiry may substantially improve our understanding of facilitation, but these have been largely neglected thus far. Despite over 15 years of research and debate on this topic, there is no consensus on the degree to which plant-plant interactions change predictably along environmental gradients (i.e. the stress-gradient hypothesis), and this hinders our ability to predict how plant-plant interactions may affect the response of plant communities to ongoing global environmental change. The existing controversies regarding the response of plant-plant interactions across environmental gradients can be reconciled when clearly considering and determining the species-specificity of the response, the functional or individual stress type, and the scale of interest (pairwise interactions or community-level response). Here, we introduce a theoretical framework to do this, supported by multiple lines of empirical evidence. We also discuss current gaps in our knowledge regarding how plant-plant interactions change along environmental gradients. These include the existence of thresholds in the amount of species-specific stress that a benefactor can alleviate, the linearity or non-linearity of the response of pairwise interactions across distance from the ecological optimum of the beneficiary, and the need to explore further how frequent interactions among multiple species are and how they change across different environments. We review the latest advances in these topics and provide new approaches to fill current gaps in our knowledge. We also apply our theoretical framework to advance our knowledge on the evolutionary aspects of plant facilitation, and the relative importance of facilitation, in comparison with other ecological processes, for maintaining ecosystem structure, functioning and dynamics. We build links between these topics and related fields, such as ecological restoration, woody encroachment, invasion ecology, ecological modelling and biodiversity-ecosystem-functioning relationships. By identifying commonalities and insights from alternative lines of research, we further advance our understanding of facilitation and provide testable hypotheses regarding the role of (positive) biotic interactions in the maintenance of biodiversity and the response of ecological communities to ongoing environmental changes."
} | 752 |
37399977 | PMC10406623 | pmc | 3,480 | {
"abstract": "Bacteria and fungi catabolize plant-derived aromatic compounds by funneling into one of seven dihydroxylated aromatic intermediates, which then undergo ring fission and conversion to TCA cycle intermediates. Two of these intermediates, protocatechuic acid and catechol, converge on β-ketoadipate which is further cleaved to succinyl-CoA and acetyl-CoA. These β-ketoadipate pathways have been well characterized in bacteria. The corresponding knowledge of these pathways in fungi is incomplete. Characterization of these pathways in fungi would expand our knowledge and improve the valorization of lignin-derived compounds. Here, we used homology to characterize bacterial or fungal genes to predict the genes involved in the β-ketoadipate pathway for protocatechuate utilization in the filamentous fungus Aspergillus niger. We further used the following approaches to refine the assignment of the pathway genes: whole transcriptome sequencing to reveal genes upregulated in the presence of protocatechuic acid; deletion of candidate genes to observe their ability to grow on protocatechuic acid; determination by mass spectrometry of metabolites accumulated by deletion mutants; and enzyme assays of the recombinant proteins encoded by candidate genes. Based on the aggregate experimental evidence, we assigned the genes for the five pathway enzymes as follows: NRRL3_01405 ( prcA ) encodes protocatechuate 3,4-dioxygenase; NRRL3_02586 ( cmcA ) encodes 3-carboxy- cis,cis -muconate cyclase; NRRL3_01409 ( chdA ) encodes 3-carboxymuconolactone hydrolase/decarboxylase; NRRL3_01886 ( kstA ) encodes β-ketoadipate:succinyl-CoA transferase; and NRRL3_01526 ( kctA ) encodes β-ketoadipyl-CoA thiolase. Strain carrying Δ NRRL3_00837 could not grow on protocatechuic acid, suggesting that it is essential for protocatechuate catabolism. Its function is unknown as recombinant NRRL3_00837 did not affect the in vitro conversion of protocatechuic acid to β-ketoadipate.",
"discussion": "Discussion While the biochemistry and molecular biology of the β-ketoadipate pathways are well-characterized in bacteria, in fungi the information we have is partial and often not linked to genes. Conversely, the genomes of fungi are full of sequence annotations that link genes to function in these pathways, with very little supporting evidence. In this study, we provide evidence for the specific gene complement that is involved in the β-ketoadipate pathway for 3,4-dihydroxybenzoate (3,4-DHB) using a combination of homology, transcriptome analysis, mutational analysis, and biochemical characterization. Figure 6 summarizes the genes involved and the available evidence supporting the functional assignments. The supporting evidence is as follows: (1) EXP, inferred from experiment, where mutation in the gene results in the accumulation of metabolite that corresponds to the product of the reaction catalyzed by the enzyme of the preceding step of the pathway; (2) IDA, inferred from direct assay, where the recombinant protein encoded by the gene catalyzes the stated enzyme activity by biochemical assay; (3) IEP, inferred from expression pattern, where the gene is differentially upregulated when cultured on 3,4-DHB; (4) IMP, inferred from mutant phenotype, where strain carrying the mutated gene displays severe growth deficiency on 3,4-DHB; and (5) ISO, inferred from sequence orthology, where orthologues of the gene in related organisms have been characterized to have the stated function. Figure 6 Summary of the 3,4-dihydroxybenzoic acid catabolic pathway in A. niger . Genes assigned for enzymes in the pathway based on the evidence from various approaches: EXP, inferred from experiment; IDA, inferred from direct assay; IEP, inferred from expression pattern; IMP, inferred from mutant phenotype; ISO, inferred from sequence orthology. Sequence similarity to orthologues of characterized enzymes suggested multiple A. niger genes are potentially involved in four of the five steps of the 3,4-DHB catabolic pathway ( Table 1 ). Comparison of the transcriptomes of growth on fructose and 3,4-DHB, both those obtained by transfer culture ( Table S3 ) and those obtained by batch fermentation in the bioreactor ( Table S4 ), provided useful clues to determine which of the candidate genes for the other four steps were likely involved in the catabolism of 3,4-DHB. Although both methods of cultivation led to similar results concerning the candidate genes, the transcriptomes had several notable differences. Perhaps due to the reduced exposure time to 3,4-DHB, the transfer culture samples had higher expression levels for each of the genes predicted to be in the pathway. However, compared to the bioreactor samples, many additional genes were also more highly expressed. Filtering the transcriptomes to include only genes with an average TPM on 3,4-DHB of 50 or greater and a fold change of 5 or greater resulted in a list of 430 genes for transfer cultures but only 223 for the bioreactor cultures. Among these genes, 84 appeared on both lists, including one of the candidate genes for each enzyme except for ChdA, 3-carboxymuconolactone hydrolase/decarboxylase, which had three candidates on both lists: NRRL3_00837 , NRRL3_01409 , and NRRL3_08340 . Thus, for each step except that catalyzed by ChdA, the most highly induced candidate was subsequently confirmed by deletion analysis to be required for growth on 3,4-DHB as described below. Many of the remaining genes which appeared on both lists were transporters and transcription factors, though there were no clear candidates for the 3,4-DHB pathway, as well as hypothetical proteins and dehydrogenases. Genes appearing solely in the transfer culture list included several genes annotated as being involved in other aromatic catabolic pathways including homogentisate and 2,3-DHB. The 3,4-DHB transcriptomes in the bioreactor cultures included a higher expression level for biosynthetic gene clusters involved in secondary metabolism; for example, the TAN-1612/BMS-192548 cluster (NRRL3_09545-NRRL3_09550) and the yanuthone D cluster (NRRL3_06287-NRRL3_06296) ( 47 ), perhaps a sign of stress as shown by slow growth on 3,4-DHB. Previously, Martins et al . ( 30 ) assigned four genes to the 3,4-DHB catabolic pathway in A. nidulans . Table 3 lists the A. niger proteins involved in 3,4-DHB catabolism, revealed in this study, and their A. nidulans orthologues. Our results agreed with three of the gene assignments made by Martins et al . ( 30 ): AN8566 as protocatechuate 3,4-dioxygenase, AN1151 as 3-carboxy- cis,cis -muconate cyclase, and AN10495 as β-ketoadipate:succinyl-CoA transferase. The A. nidulans protein AN5232 assigned as 3-carboxymuconolactone hydrolase/decarboxylase is the orthologue of A. niger NRRL3_00837. We showed here that NRRL3_00837 is an essential component of the 3,4-DHB catabolic pathway, but its role is currently unresolved. We identified NRRL3_01409 as 3-carboxymuconolactone hydrolase/decarboxylase and NRRL3_01526 as β-ketoadipyl-CoA thiolase, their A. nidulans orthologues AN10520 and AN5698 had not been characterized previously. Table 3 Aspergillus niger proteins implicated in the 3,4-DHB catabolic pathway by this study compared to their orthologues in A. nidulans 3,4-DHB catabolic pathway proteins A. niger A. nidulans orthologue Sequence identity Protocatechuate 3,4-dioxygenase NRRL3_01405 AN8566 a 94% 3-carboxy- cis,cis -muconate cyclase NRRL3_02586 AN1151 a 91% 3-carboxymuconolactone hydrolase/decarboxylase NRRL3_01409 AN10520 72% β-ketoadipate:succinyl-CoA transferase NRRL3_01886 AN10495 a 86% β-ketoadipyl-CoA thiolase NRRL3_01526 AN5698 91% Essential protein of unknown function NRRL3_00837 AN5232 a b 72% a Proteins previously assigned by Martins et al . ( 30 ). b Protein previously assigned by Martins et al . ( 30 ) as 3-carboxymuconolactone hydrolase/decarboxylase. Previous studies used transcriptome analysis, 3,4-DHB degradation by strains overexpressing NRRL3_01405 , purified enzyme activity, and mutant growth phenotype to conclude that this gene encodes protocatechuate 3,4-dioxygenase in A. niger ( 31 , 32 ). Furthermore, our study showed the purified enzyme produced 3-carboxymuconic acid from 3,4-DHB, and the Δ prcA strain accumulated 3,4-DHB on a medium containing quinic acid as a carbon source. Residual growth in this strain may be due to the ability of one of the other catechol oxygenases encoded in the genome to catalyze this step, although in a previous study we were unable to detect activity of purified forms of these enzymes, 2 hydroxyquinol dioxygenases (NRRL3_02644 and NRRL3_05330) and a catechol dioxygenase (NRRL3_04787), on 3,4-DHB ( 32 ). However, Lubbers et al . ( 31 ) showed that a strain with deletions of both prcA and the hydroxyquinol dioxygenase, encoded by NRRL3_02644 , converted 3,4-DHB slowly to hydroxyquinol, suggesting the possible existence of an enzyme that catalyzes the oxidative decarboxylation of 3,4-DHB to hydroxyquinol, as in the fungus Trichospon cutaneum ( 48 ). This enzyme has been suggested to be encoded by NRRL_04986 ( 37 ). Only one gene, NRRL3_02586 ( cmcA ), displays strong homology to a known N. crassa orthologue with 3-carboxy- cis,cis -muconate cyclase activity ( 40 ). The corresponding A. nidulans orthologue, AN1151 ( Table 3 ), was upregulated in both the transcriptome and proteome of benzoate-grown cells and a knockout of this gene eliminated growth on benzoate and resulted in the accumulation of an intermediate putatively identified as 3-carboxymuconate ( 30 ). Transcriptome analysis showed that cmcA was upregulated 60-fold when A. niger was cultured with 3,4-DHB, compared with fructose, as the sole carbon source ( Table 1 ). In Δ cmcA , the complete lack of growth ( Fig. 2 ) establishes that it is required for growth on 3,4-DHB. Our biochemical analysis establishes that CmcA catalyzes the conversion of 3-carboxy- cis,cis -muconic acid to the 3-carboxymuconolactone that has been reported in other fungal systems, as opposed to the bacterial 4-carboxymuconolactone ( 19 ). As expected, the Δ cmcA strain accumulated a compound with the mass of 3-carboxy- cis,cis-muconic acid in the medium ( Fig. 3 D ). Although there were several other compounds with a similar score, the Agilent Molecular Structure Correlator software predicted 3-carboxy- cis,cis -muconic acid as one of the most likely compounds based on the MS/MS fragment spectra. Four candidates for the catabolism of 3-carboxymuconolactone were identified, of which NRRL3_08340 was at least 3.5 times more highly induced than the other candidates ( Table 1 ). The Δ NRRL3_08340 strain, however, grew unimpeded on 3,4-DHB indicating that it is not essential for the pathway. This should perhaps not be surprising since sequence comparisons suggest that it is an orthologue of a bacterial enzyme involved in the decarboxylation of 4-carboxymuconolactone, rather than the 3-carboxymuconolactone produced by fungi, including A. niger as we have shown here. Deletion of each of the remaining two 3,4-DHB-induced candidates resulted in no growth on 3,4-DHB in Δ NRRL3_00837 strain and severely reduced growth in the Δ NRRL3_01409 strain, possibly due to activity of an alternative enzyme. No pathway-related metabolites accumulated in either deletion strain. However, biochemical analysis showed unambiguously that purified NRRL3_01409 catalyzed the conversion of 3-carboxymuconolactone to β-ketoadipate, whereas purified NRRL3_00837 did not. These results show the importance of using multiple different approaches when making gene assignments. Since deletion of the A. nidulans orthologue ( AN5232 ) of NRRL3_00837 resulted in a lack of growth and the appearance of a small amount of unconfirmed intermediate from benzoate ( 30 ), we tested the possibility that it had some kind of accessory role in the conversion. However, the addition of NRRL3_00837 has no effect on either the rate or nature of products from NRRL3_01409 ( Fig. 5 ). The protein encoded by NRRL3_01409 is consistent with the size of an A. niger enzyme that was purified previously and shown to possess both decarboxylase and hydrolase activities ( 33 ). A BLASTP comparison shows that the sequence over the C-terminal half of the protein is 28% identical with a 3-oxoadipate enol-lactonase from the bacterial protocatechuate branch of the β-ketoadipate pathway in A. baylyii ( 40 ). Furthermore, an α/β hydrolase domain (pfam00561) was detected in the C-terminal. The N-terminal half contains an uncharacterized protein domain (IPR003497, BRO N-terminal domain). Together with the biochemical characterization, these observations suggest two domains in this bifunctional enzyme: a hydrolase domain at the C-terminal and a decarboxylase domain at the N-terminal. The observation that the Δ NRRL3_01409 strain still shows weak growth on 3,4-DHB suggests the possibility that paralogous enzymes, potentially those involved in the catechol branch of the β-ketoadipate pathway may be partly functional in the delactonization step. Motif analysis of NRRL3_00837 based on InterProScan shows the presence of the AhpD domain (InterPro entry IPR029032). However, a BLASTP comparison (using default parameters) between the sequence of NRRL3_00837 and AhpD, a characterized alkylperoxide reductase from M. tuberculosis (AHPD_MYCTU) upon which the domain is predicted, revealed no significant sequence identity. To gain additional insight into the function of NRRL3_01409 and NRRL3_00837, we used AlphaFold ( 49 ) to predict their structures and to search for proteins that are structurally similar to them. Using the AlphaFold model of NRRL3_01409 as a query, the best hit was to crystal structure 2XUA from Paraburkholderia xenovorans in the Protein Data Bank ( https://www.rcsb.org ). Crystal structure 2XUA is described as a β-ketoadipate enol-lactonase because it displays 38% identity to the biochemically characterized A. baylyi β-ketoadipate enol-lactonase. Crystal structure 2XUA matches closely the AlphaFold-predicted structure of A. baylyi β-ketoadipate enol-lactonase ( Fig. S12 ) and to the C-terminal half of NRRL3_01409 ( Fig. S13 ). These results support our conclusion that the C-terminal half of NRRL3_01409 possesses hydrolase activity similar to that of β-ketoadipate enol-lactonase and that the N-terminal half of NRRL3_01409 is a previously uncharacterized structure with decarboxylase activity. The AlphaFold model of NRRL3_00837 showed no similarity to the AlphaFold or crystal structures of fungal or bacterial enzymes involved in 3,4-DHB catabolism except to the A. nidulans ortholog AN5232. A structural similarity search of the Protein Data Bank using the AlphaFold predicted model of NRRL3_00837 returned a transcription regulator, FapR, from Staphylococcus aureus (PDB ID 4A0Z). This may suggest NRRL3_00837 plays a role in the regulation of the β-ketoadipate pathway, though no significant similarity was found in the sequences of these two proteins. For the last two steps of the pathway, conversion of β-ketoadipate to acetyl-CoA and succinate via β-ketoadipyl-CoA, three candidates each were identified based on homology, of which NRRL3_01886 (encoding β-ketoadipate:succinyl-CoA transferase, KstA) and NRRL3_01526 (encoding β-ketoadipyl-CoA thiolase, KctA) were each 70-fold upregulated by 3,4-DHB. Deletion of each gene in turn resulted in near-complete or complete loss of growth on 3,4-DHB, as well as on 2,3-DHB, which is likely degraded via β-ketoadipate. But as catabolism of 2,3-DHB does not use the enzymes between 3,4-DHB and β-ketoadipate, deletion of their encoding genes had no effect on the growth of 2,3-DHB ( Fig. 2 ). As would be expected from its predicted function as β-ketoadipate CoA transferase, Δ NRRL3_01886 accumulated β-ketoadipate from quinic acid. Unfortunately, no metabolites were detected in culture supernatants for Δ NRRL3_01526 , possibly since its expected product, β-ketoadipyl-CoA, would not be expected to pass the cell membrane. However, attempted analysis of intracellular metabolites did not reveal accumulation either. Although we attempted to heterologously express NRRL3_01886 , only inclusion bodies were produced, so we were unable to confirm this functional assignment by direct biochemical analysis. Interestingly, it has been shown in bacteria such as Pseudomonas putida that β-ketoadipate CoA transferase is a heteromeric enzyme encoded by two genes ( 27 ). These authors noted the similarity of each gene to the two halves of homodimeric pig heart succinyl-CoA:3-ketoacid-CoA transferase and concluded that a gene fusion event occurred during evolution of the eukaryotic enzyme. As noted in the NRRL3 annotation for this gene, the closest orthologue (56% identity) in the Uniprot database to KstA is human succinyl-CoA:3-ketoacid coenzyme A transferase (accession number P55809), with an amino-terminal corresponding to one subunit (InterPro domain IPR012792) and the C-terminal encompassing the other subunit (InterPro domain IPR012791)."
} | 4,280 |
35626550 | PMC9141356 | pmc | 3,482 | {
"abstract": "When computers started to become a dominant part of technology around the 1950s, fundamental questions about reliable designs and robustness were of great relevance. Their development gave rise to the exploration of new questions, such as what made brains reliable (since neurons can die) and how computers could get inspiration from neural systems. In parallel, the first artificial neural networks came to life. Since then, the comparative view between brains and computers has been developed in new, sometimes unexpected directions. With the rise of deep learning and the development of connectomics, an evolutionary look at how both hardware and neural complexity have evolved or designed is required. In this paper, we argue that important similarities have resulted both from convergent evolution (the inevitable outcome of architectural constraints) and inspiration of hardware and software principles guided by toy pictures of neurobiology. Moreover, dissimilarities and gaps originate from the lack of major innovations that have paved the way to biological computing (including brains) that are completely absent within the artificial domain. As it occurs within synthetic biocomputation, we can also ask whether alternative minds can emerge from A.I. designs. Here, we take an evolutionary view of the problem and discuss the remarkable convergences between living and artificial designs and what are the pre-conditions to achieve artificial intelligence.",
"introduction": "1. Introduction With the evolution of life came cognition [ 1 ]. As soon as cells were able to evolve into autonomous agents, the combination of receptors gathering signals and mechanisms of response to those signals rapidly transformed into rich molecular networks. Those networks provided the basis for the smaller scale of computation: survival requires exploiting resources in a reliable way that allows reproduction. Since this is a combination of growing and being robust against fluctuations over a minimal time window, computation was tied to predictive power [ 2 , 3 , 4 , 5 ]. It is this power what actually might foster the evolution towards brains [ 6 ], large and small: in order to reduce the uncertainty of external fluctuations, prediction is a convenient faculty. If we follow the steps towards cognitive complexity that predate the emergence of brains, several key ingredients seem necessary. Looking at their evolutionary emergence is relevant for our discussion concerning the space of possible cognitive networks. One of them was the invention of neurons: specialized cell types with a marked elongated, branched shape capable of establishing connections. In most cases, these are polar, unidirectional structures, with response functions that involve nonlinear thresholds. The power of neurons became a reality as soon as groups of them became interconnected, leading to the first neural networks. Among the key innovations associated to these early assemblies, interneurons must have been a crucial step towards information processing beyond the sensor–actuator chain. With the Cambrian explosion of life, the rise of animals favored the development of sensory organs, learning, and movement [ 7 ]. All these factors came together within a novel developmental design: brains emerged within bilateral animals, and those newcomers actively explored their worlds, moving around. A compelling proposal concerning the origins of brains is, in fact, the so-called moving hypothesis : it posits that the active exploration of the external world fostered the evolutionary path that led to brains [ 6 ]. In a novel biosphere dominated by predator–prey arms races, brains were an optimal solution to deal with information. If we fast-forward in time, several important changes took place paving the way towards complex brains. This is particularly dramatic for human brains: a rapid expansion during evolution facilitated the addition of microcircuit modules to a multilayered neocortex [ 8 ]. Turning our attention to machines, we can see how inventors and scholars have repeatedly drawn inspiration from nature’s cognitive systems. Sometimes through outright imitation, as in the case of mechanical automata ( Figure 1 a). Others by focusing efforts on human-specific cognitive problems (e.g., chess, Figure 1 b) [ 9 ]. In yet other cases, through converging metaphors—e.g., from Cajal’s flows within neurons and across neural circuits to technological networks that enable the flow of information ( Figure 1 c). The exchange of ideas between computational designs and theoretical neuroscience has been constant. Prediction too has been a major force in the development of a large part of technology, particularly after the rise of Information Technology from the 1950s [ 10 ]. In parallel with the development of the theory of computation, the first steps towards a theory of neural networks came to life, starting from early comparisons between brains and computers. The first computers were plagued with problems associated to faulty units: vacuum tubes were prone to failure. Far from the reliable nature of brains, where neurons can die with no major consequences for the system-level performance, single-element failures could cause large disruptions. The comparative analysis between brains and computers (or computational analogies of brains) has been a recurrent topic since von Neumann’s book The computer and the brain [ 11 ]. In the original formulation, the main problem was how to design reliable computers made of unreliable parts. Such approximation was largely forgotten within computer science with the rise of integrated circuits, although the problem became a central topic in the domain of neural networks. With the potential of simulating neural systems some of the early metaphors of memory involved strong analogies with magnetic materials. Such analogies would be developed in depth with the use of the statistical physics of spin glasses as computational substrates [ 12 , 13 ]. These similarities eventually provided the basis for the attractor picture that is now widespread. Over the last decade, a new wave of excitement has emerged with the rise of Deep Learning networks [ 14 ]. These descendants of the early multilayer neural networks developed in the 1990s have been accompanied with a considerable set of expectations (and hype). Because of their remarkable power to deal with specific problems far beyond the capacities of humans, claims have been repeatedly made suggesting that larger systems will eventually achieve cognitive skills similar (if not greater) than human brains, including consciousness or awareness. (See the recent stir caused by OpenAI’s chief scientist, Ilya Sutskever, claiming that “it may be that today’s large neural networks are slightly conscious”. https://lastweekin.ai/p/conscious-ai?s=r , accessed on 2 May 2022). However, as it has happened before many times (the winter–spring cycles of A.I.), artificially intelligent systems are still rather far from our general intelligence [ 15 ], and, indeed, they often appear brittle when taken even slightly out of their well-controlled closed worlds [ 16 ]. All this mimicry, inspiration, and convergences bear some pressing questions: do natural cognitive designs exhaust the space of the possible? Will every artificial cognitive solution ever found correspond to an earlier invention in nature? If so, then understanding natural cognition should be sufficient to learn everything that there is to know about intelligence and cognition. Might comprehending nature be necessary as well—i.e., does every cognitive solution respond to some natural challenge or feature that we need to understand in order to ultimately grasp cognition? If so, which is a minimal set of such challenges and features that can generate the range of cognitive designs? It is also possible that nature is not so restrictive and all-encompassing. This would leave a large elbow room for artificial cognition, human invention, and open-ended progress. Yet, more wondrous questions are also put forward: what solutions might have been missed in the natural history of cognition? Are there artificial cognitive designs that cannot be reached by extant evolutionary forces alone? In this paper, we argue that very relevant lessons can (and must) be obtained from a comparative analysis between evolved and designed cognitive systems. On the one hand, there are several non-trivial observations that suggest a limited repertoire of design principles that pervade and constrain the space of the possible: evolved and artificial architectures often converge. Secondly, there is a question regarding certain dynamical patterns exhibited by living systems that seldom appear in artificial neural networks. Brains seem to operate close to critical states: is this a relevant trait to be considered when building artificial counterparts? Third, we will consider a list of attributes of human brains that define a gap between our species and any other living organism and we will see why A.I. systems might require to include evolutionary dynamics to get there.",
"discussion": "5. Discussion Can machines ever achieve true intelligence? In a recent paper entitled “Building machines that learn and think like people” [ 198 ], it has been argued that, for ANN to rapidly acquire generalization capacities through learning-to-learn, some important components are missing. One is to generate context and improve learning by building internal models of intuitive physics. Secondly, intuitive psychology is also proposed as a natural feature present since early childhood (children naturally distinguish living from inanimate objects) which could be obtained by introducing a number of Bayesian approximations. Finally, compositionality is added as a way to avoid combinatorial explosions. In their review, Lake et al. discussed these improvements within the context of deep networks and problem-solving for video games, and thus considered the programming of primitives that enrich the internal degrees of freedom of the ANN. These components would expand the flexibility of deep nets towards comprehending causality. (See also the life-long work of Jürgen Schmidhuber for important developments over the last decades in the meta-learning or learning-to-learn paradigms: https://people.idsia.ch/~juergen/metalearning.html , accessed on 2 May 2022). Lake et al. also pointed at several crucial elements that need to be incorporated, being language a prominent one. So far, despite groundbreaking advances in language processing, the computational counterparts of human language are very far from true language abilities. These improvements will without doubt create better imitations of thinking, but they are outside an embodied world where—we believe—true complex minds can emerge by evolution. Are there alien minds? Yes and no. An affirmative answer emerges from the obvious: artificial systems do not need to follow biological constraints or Darwinian evolutionary paths. Being designed by humans or evolved within computers using ad hoc optimization procedures, the final outcome can depart from biology in multiple ways. A deep network can outperform humans in a very specific task using a training algorithm based on feed-forward convolutional nets that, although inspired by experiments, lack the re-entrant loops that might be crucial to achieve true intelligence and awareness. Robotic agents can have behavioral patterns of response to complex environments, but the cognitive skills are externalized: algorithms are being executed in a rather powerful computer that resides somewhere else outside the body. However, these are systems where agency plays a minor role. Perhaps, the really relevant question is this: are there autonomous alien minds? If convergent designs are an indication that there is a limited repertoire of possible neural architectures and cognitive autonomous agents, the future of A.I. is in the evolutionary arena. That means that the roads not taken necessarily cross the land of embodiment: as it occurs with naturally evolved systems, moving in an uncertain world was one of the engines of brain evolution. Moreover, another crucial component of evolutionary innovations is the emergence of new forms of cooperation. Cognitive agents mean evolving communication and dealing with information [ 193 ]. What kind of interesting phenomena can be observed using these two ingredients? Evolved robotic systems illustrate fairly well the ways in which evolutionary dynamics simultaneously link some of these components of cognition. As an example, robotic agents moving on a landscape where both positive and negative inputs (sources of charge and discharge, respectively) are located on given spots develop communication along with cooperative strategies that improve group fitness [ 199 , 200 ]. Each robot is equipped with a set of sensors and lights and start foraging with a random configuration. A feedforward ANN allows evolving the interactions between sensors and lights and to generate communication among robots that allows for cooperation and altruism. Finding and avoiding positive and negative scenarios create the conditions for increasing group fitness. However, crowding also triggers cheating and deception (a familiar trait of evolution): robots can also evolve into lying to each other. Despite the simple nature of the players, a combination of some key evolvable features can lead to unexpected insights. As pointed out in the Introduction, the paths that lead to brains seem to exploit common, perhaps universal properties of a handful of design principles and are deeply limited by architectural and dynamical constraints. Is it possible to create artificial minds using completely different design principles, without threshold units, multilayer architectures or sensory systems such as those that we know? Since millions of years of evolution have led, through independent trajectories, to diverse brain architectures and yet not really different minds, we need to ask if the convergent designs are just accidents or perhaps the result of our constrained potential for engineering designs. Within the context of developmental constraints, the evolutionary biologist Pere Alberch wrote a landmark essay that can further illustrate our point [ 201 ]. It was entitled “The Logic of Monsters” and presented compelling evidence that, even within the domain of teratologies, it is possible to perceive an underlying organization: far from a completely arbitrary universe of possibilities (Since failed embryos are not the subject of selection pressures, it can be argued that all kinds of arbitrary morphological “solutions” could be observed), there is a deep order that allows to define a taxonomy of “anomalies”. Within our context, it would mean that the universe of alien minds might be also deeply limited."
} | 3,725 |
28286360 | PMC5324606 | pmc | 3,483 | {
"abstract": "Abstract Aim Coral reefs rely on the symbiosis between scleractinian corals and intracellular, photosynthetic dinoflagellates of the genus Symbiodinium making the assessment of symbiont diversity critical to our understanding of ecological resilience of these ecosystems. This study characterizes Symbiodinium diversity around the Arabian Peninsula, which contains some of the most thermally diverse and understudied reefs on Earth. Location Shallow water coral reefs throughout the Red Sea ( RS ), Sea of Oman ( SO ), and Persian/Arabian Gulf ( PAG ). Methods Next‐generation sequencing of the ITS 2 marker gene was used to assess Symbiodinium community composition and diversity comprising 892 samples from 46 hard and soft coral genera. Results Corals were associated with a large diversity of Symbiodinium , which usually consisted of one or two prevalent symbiont types and many types at low abundance. Symbiodinium communities were strongly structured according to geographical region and to a lesser extent by coral host identity. Overall symbiont communities were composed primarily of species from clade A and C in the RS , clade A, C, and D in the SO , and clade C and D in the PAG , representing a gradual shift from C‐ to D‐dominated coral hosts. The analysis of symbiont diversity in an Operational Taxonomic Unit (OTU)‐based framework allowed the identification of differences in symbiont taxon richness over geographical regions and host genera. Main conclusions Our study represents a comprehensive overview over biogeography and molecular diversity of Symbiodinium in the Arabian Seas, where coral reefs thrive in one of the most extreme environmental settings on the planet. As such our data will serve as a baseline for further exploration into the effects of environmental change on host–symbiont pairings and the identification and ecological significance of Symbiodinium types from regions already experiencing ‘Future Ocean’ conditions.",
"conclusion": "Conclusions This study utilized next‐generation sequencing of the ITS2 marker gene to analyze Symbiodinium composition associated with 46 coral genera and 892 specimens around the Arabian Peninsula. As such our study provides a comprehensive catalog and comparative assessment of symbiont diversity in this comparatively understudied region, where coral reefs thrive in one of the most extreme environmental settings on the planet. Our data show that corals around the Arabian Peninsula are associated with a large diversity of Symbiodinium types that are strongly structured by geographical location and to a lesser extent by coral host identity. The application of high‐resolution symbiont typing enabled the analyses of symbiont diversity in an OTU‐based framework that highlights differences in OTU richness associated with geographical region and host genus, the significance of which remain to be determined. In addition, our analysis highlights a set of potential thermotolerant Symbiodinium types outside clade D that warrant further investigation and emphasize that thermal tolerance is a species‐ or type‐specific, rather than clade‐specific trait.",
"introduction": "Introduction Reef‐building corals are the foundation of reef ecosystems and provide habitats to a diverse set of marine species, many of which are economically and ecologically important (Roberts et al ., 2002 ). The ability of scleractinian corals to build reef structures critically relies on their ability to form symbioses with photosynthetic dinoflagellates of the genus Symbiodinium Freudenthal, 1962 (Muscatine & Porter, 1977 ). These intracellular algae provide up to 95% of the energy needs of the coral host (Falkowski et al ., 1984 ). Symbiodinium species are ecologically diverse, exhibiting discrete associations with different coral hosts that can differ over large geographical scales, depth, season, and exposure to stressors (LaJeunesse et al ., 2004 , 2010 ; Finney et al ., 2010 ; Ziegler et al ., 2015 ). Furthermore, Symbiodinium species vary in their nutritional benefits to the hosts (Stat et al ., 2008a ; Cantin et al ., 2009 ; Baker et al ., 2013 ) and in their response to thermal stress and varying light intensity (LaJeunesse, 2001 ; Iglesias‐Prieto et al ., 2004 ; Ziegler et al ., 2014 ; Pettay et al ., 2015 ). Hence, detailed knowledge of Symbiodinium coral pairings is arguably critical to our understanding of ecological resilience of coral reefs. Of the currently nine clades of Symbiodinium (Pochon & Gates, 2010 ), the clades A, B, C, and D are most commonly associated with corals (Pochon et al ., 2014 ). These clades can be further subdivided into subclades and types likely comprising hundreds of species. However, delineation of Symbiodinium diversity is not straightforward. Due to the deep phylogenetic divergence in the genus Symbiodinium , differences between clades can match differences observed at the level of order in other dinoflagellates (Rowan & Powers, 1992 ). Hence, no single universal marker gene exists to tease apart all ecologically discrete units (LaJeunesse, 2001 ; Pochon et al ., 2014 ). Recent studies have applied specific multigene phylogenies to a single clade under study that successfully characterized distinct species and lineages (LaJeunesse & Thornhill, 2011 ; Lajeunesse et al ., 2012 ). Despite the limitations of single gene phylogenies, the Internal Transcribed Spacer 2 (ITS2) region remains the most commonly used marker for Symbiodinium diversity typing. Because of the tandem repeat arrangement of rRNA genes, the ITS2 gene is a multicopy marker, which makes discerning inter‐ from intragenomic variation critical (Thornhill et al ., 2007 ; Sampayo et al ., 2009 ). Denaturing gradient gel electrophoresis (DGGE) can be used to address this issue by identifying the numerically dominant ITS2 variant(s) that in many cases represents a reproducible ITS2 ‘type’ that can be associated with the underlying dominant Symbiodinium species (LaJeunesse, 2001 ; Sampayo et al ., 2009 ; Arif et al ., 2014 ). However, DGGE lacks sensitivity to identify background symbiont types, especially if their abundance is < 5–10% (Thornhill et al ., 2006 ; LaJeunesse et al ., 2008 ). Bacterial cloning, in comparison, tends to overestimate ITS2 diversity by potentially amplifying genomically rare variants that may further complicate discrimination of intra‐ and intergenomic variation of ITS2‐based Symbiodinium diversity (Thornhill et al ., 2007 ). Recently, next‐generation sequencing (NGS) has been utilized for typing Symbiodinium ITS2 diversity, enabling the identification of distinct ITS2 types below 1% in abundance (Arif et al ., 2014 ; Quigley et al ., 2014 ; Thomas et al ., 2014 ). High‐throughput sequencing of the ITS2 gene locus creates an opportunity to derive Operational Taxonomic Unit (OTU) cut‐offs via the assessment of average intragenomic diversities retrieved from the sequencing of isoclonal cultures that represent different species, as demonstrated by Arif et al . ( 2014 ). In the study by Arif et al . ( 2014 ), a range of isoclonal cultures representing different clades was sequenced and intragenomic ITS2 variance was successfully collapsed into distinct OTUs at a 97% similarity cut‐off. This approach was subsequently applied to field‐collected specimens where Arif et al . ( 2014 ) showed that this approach efficiently reduced the complexity of ITS2 NGS data and allowed for economic and efficient comparative analysis of a large number of coral species in a common and reproducible framework. Yet, sequence‐based analysis of NGS data faces some of the same challenges as bacterial cloning in discerning intra‐ from intergenomic ITS2 diversity. As a consequence, OTU‐derived Symbiodinium species diversity estimates must be considered provisional and can be overconservative for some specimens (but see Arif et al . ( 2014 )). Thus, concomitant analyses of ITS2 sequence data in combination with an OTU‐based approach provide a means to interrogate NGS data in a manner that allows elucidating symbiont diversity of known Symbiodinium types and an assessment of taxon richness that addresses the challenges associated with a multi‐copy marker such as ITS2. The seas surrounding the Arabian Peninsula, including the Red Sea (RS), the Sea of Oman (SO), and the Persian/Arabian Gulf (PAG), represent an understudied marine region despite hosting a large diversity of coral reef ecosystems (Riegl et al ., 2012 ; Bauman et al ., 2013 ; Coles et al ., 2015 ). The RS is an oligotrophic system with high temperature variation and high salinity due to low influx of freshwater, high evaporation and limited exchange with the Indian Ocean (Sheppard et al ., 1992 ). The environmental conditions in the PAG are arguably the most extreme in the world under which corals exist. Corals in the PAG are exposed to extreme fluctuations of temperatures (from 11 to 36 ° C) and high salinity (often > 44 PSU) (Coles & Riegl, 2013 ). The conditions experienced by corals in the RS and the PAG are generally beyond the limits of what corals experience and survive elsewhere, which for the PAG has been shown to be partially attributable to a recently identified symbiont species, Symbiodinium thermophilum (Hume et al ., 2015 ), that is prevalent in the PAG due to its preference to high salinity (D'Angelo et al ., 2015 ). On the contrary, the temperature (22–32 ° C) and salinity (> 37 PSU) conditions in the SO are less extreme than in the PAG, due to greater mixing with the wider Indian Ocean (Piontkovski & Al‐Jufaili, 2013 ). To date, Symbiodinium diversity has been primarily studied in coral species from various locations in the Caribbean, the Central Pacific, and the Great Barrier Reef (LaJeunesse et al ., 2004 ; Finney et al ., 2010 ; Tonk et al ., 2013 ), but data from the Arabian region are limited. Studying symbiont diversity in reefs of the RS and PAG might provide critical insight to our understanding of coral resilience and the underlying adaptations that allow corals to survive under future climate change scenarios. In this context, the more moderate conditions in the SO, comparable to other major coral reef habitats around the world, such as the Great Barrier Reef (Tonk et al ., 2013 ), can serve an important baseline to disentangle geographically‐ from environmentally prompted patterns of Symbiodinium diversity and abundance. To provide a comprehensive assessment of coral‐associated Symbiodinium diversity in the seas around the Arabian Peninsula, we conducted next‐generation sequencing‐based ITS2 typing of 892 coral colonies representing 46 coral genera from the RS, SO, and PAG. Furthermore, we investigated Symbiodinium diversity and community structure within and between coral colonies and genera across regions.",
"discussion": "Discussion This study represents a comprehensive survey on coral– Symbiodinium association using next‐generation sequencing techniques in one of the most extreme, but understudied geographical regions in the world. We investigated Symbiodinium diversity associated with almost 900 coral specimens encompassing 46 genera from the Red Sea, the Sea of Oman, and the Persian/Arabian Gulf. Given the large number of specimens collected, Symbiodinium diversity typing using next‐generation sequencing methodology was not only more convenient, but also allowed for the study of host–symbiont patterns including both sequence‐based ITS2 analysis and OTU‐based ITS2 diversity and richness analysis. A sequence‐based analysis permits direct assessment of coral‐associated symbionts (typically focusing on the more/most dominant members). Its reliance on previously recorded Symbiodinium types allows the comparison between studies and sites, but makes the assignment of provisionally new entities difficult. A benefit of the application of an OTU‐based framework may lie in its ability to estimate Symbiodinium diversity without the need for a priori and formal description of ITS2 symbiont types. Consequently, we used both approaches in this study in order to provide a comprehensive description of symbionts associated with corals of the Arabian Seas. Biogeography of Symbiodinium types around the Arabian Peninsula Our analyses of host– Symbiodinium association in 46 coral genera across the Arabian Peninsula showed that coral–symbiont association was strongly defined by geographical location. Beyond the prevalence of regionally specific Symbiodinium types that distinguished the three regions, symbiont communities shifted from clade C and clade A dominance in the RS, over decreasing proportions of clade A and increasing proportions of clade D in the SO, to a Symbiodinium community dominated by clade D and to a lesser extent by clade C in the PAG. Not all coral genera could be sampled in all regions, yet sampling the most abundant taxa at each location enabled us to record the common Symbiodinium types in each region. In contrast to many other studies (LaJeunesse, 2002 ; LaJeunesse et al ., 2004 ; Tonk et al ., 2013 ), we found that coral host genus played a relatively small role in the identity of the dominant Symbiodinium type, but largely followed biogeographical patterns within most coral host genera around the Arabian Peninsula. Possible explanations for this observation could lie in the difference of environmental conditions between the three regions (see introduction), the comprehensive sampling approach over many diverging host genera, differences in sampling times between locations, or the use of highly resolving molecular techniques to characterize Symbiodinium associations. A common feature of Symbiodinium communities in the Arabian Seas with those in the Indo‐Pacific and Atlantic‐Caribbean is the overproportionally high diversity in Symbiodinium clade C (LaJeunesse, 2005 ). This diversity can be attributed to a series of adaptive radiation events based on few ancestral Symbiodinium types, C1 and C3 (LaJeunesse, 2005 ), which were also present in our data set. Possible examples of regional diversification may be represented by Symbiodinium C41, which is separated from Symbiodinium C1 by a single base pair difference in the ITS2 sequence. Symbiodinium C41 was endemic, but regionally prevalent in the RS; and in the SO Symbiodinium C39 may represent another example of regional diversification. Moreover, fine‐scale genetic divergence within these types may not be fully resolved by the ITS2 marker, masking other potentially divergent lineages (Thornhill et al ., 2014 ), as exemplified by the cryptic, but regionally prevalent C3–type species Symbiodinium thermophilum in the PAG (Hume et al ., 2015 , 2016 ), which likely also dominated the clade C assemblage observed in the southern PAG in this study (Hume et al ., 2016 ). Insight into thermotolerant Symbiodinium types Studying Symbiodinium diversity in one of the hottest regions where corals persist offers the opportunity to search for environmentally tolerant Symbiodinium types. Probably the most obvious such case lies in S. thermophilum in the PAG (Hume et al ., 2015 , 2016 ). Porites harbouring this Symbiodinium species were more resilient to heat stress than their Pacific counterparts harboring C15 (Hume et al ., 2013 ). However, further investigation into this symbiosis has revealed a concordant adaptation to the high salinity of the PAG, deeming it unlikely that its thermal tolerance can be extended beyond its current range (D'Angelo et al ., 2015 ). Furthermore, in the RS the main Symbiodinium type in Porites changed from C15 at cooler offshore locations to Symbiodinium D1a (and not C3) at warmer nearshore locations (Ziegler et al ., 2015 ). Within the PAG, the northern reefs were distinct from the southern locations, mainly due to the higher prevalence of Symbiodinium from clade D in the north, an intriguing pattern that has been described before (Baker et al ., 2004 ). While Symbiodinium from clade D may offer higher thermal tolerance, it may actually perform inferior compared to other types in supporting essential functions such as coral growth under non‐stressful conditions (Pettay et al ., 2015 ). Beyond known heat‐resistant symbionts, we identified other Symbiodinium types prevalent around the Arabian Peninsula, whose tolerance to high temperature is presently unknown. For example, Symbiodinium type C41 and C39 were prevalent and restricted to the RS and the SO providing potential candidate endosymbiont types that display increased thermotolerance. Another source of heat‐resilient candidate endosymbionts may be derived from the identification of Symbiodinium types associated with coral genera that were stable between regions. For instance, Stylophora was commonly and dominantly associated with Symbiodinium A1 throughout the Arabian Peninsula, an association that is congruent with previous reports from the northern RS (LaJeunesse, 2001 ). By comparison, in other geographical locations Stylophora spp. are more commonly associated with Symbiodinium from clade C, such as in the Western Indian Ocean (LaJeunesse et al ., 2010 ) and in the Great Barrier Reef (Sampayo et al ., 2007 ; Stat et al ., 2008b ). This symbiont association is quite surprising, because outside the Caribbean, Symbiodinium from clade A are more commonly associated with non‐scleractinian taxa, such as zooanthids (Reimer et al ., 2006 ), giant clams (Baillie et al ., 2000 ), and jellyfish (LaJeunesse, 2001 ). \n Symbiodinium community composition and richness align with geographical regions and environmental settings In addition to the biogeographical pattern of the most abundant Symbiodinium types between the three regions around the Arabian Peninsula, the composition of the Symbiodinium OTU community was more similar in the SO and the PAG compared to the RS. Commonly, Symbiodinium communities are structured along environmental gradients, for example, from nearshore to offshore locations across the shelf (Tonk et al ., 2013 ) or over latitudinal temperature gradients (Loh et al ., 2001 ; Macdonald et al ., 2008 ). In light of the diverging environmental conditions between the SO and PAG, our observations suggest that the geographical proximity and thus the higher connectivity of populations between the two regions may be responsible for the higher similarity compared to the more distantly located RS. Besides the similarities and differences in OTU community composition across geographical regions, we found distinct OTU richness patterns within corals depending on site and genus. For instance, coral colonies belonging to the genus Porites on average harbored twice as many Symbiodinium OTUs than corals from any other host genus; they also contained the largest number of OTUs encountered in a single colony by far. Coral colonies of Porites were found in association with a large diversity of other Symbiodinium types in addition to the two main types (C3 and C15), and the genus Porites roughly contained half of all recorded Symbiodinium OTUs in each of the three regions, but this may potentially be confounded by sampling of different (cryptic) species in this genus. In this regard, our observations contradict the notion of Porites as a symbiont specialist genus (Silverstein et al ., 2012 ) and support the latest observations suggesting a large symbiont flexibility in Porites (Ziegler et al ., 2015 ). Overall, corals from the PAG hosted the least diverse Symbiodinium communities compared to the RS and the SO, reflecting patterns observed in species diversity of coral hosts, fishes, and other reef‐associated fauna (Sheppard et al ., 1992 ; Burt et al ., 2011 ; Bauman et al ., 2013 ). This might be due to the comparably young age of the PAG or the extreme environmental conditions with respect to high temperature variation as well as elevated salinity in the PAG constituting a selective bottleneck in which only highly specific host–symbiont pairings prevail (D'Angelo et al ., 2015 ; Hume et al ., 2015 ). Local environmental pressure may thus also be a limiting factor for diversity and distribution of both corals and Symbiodinium in this region (D'Angelo et al ., 2015 ). At the same time, the biological significance of differences in symbiont richness over geographical regions and host genera is entirely unclear. For instance, it would be desirable to understand whether the increased OTU richness in Porites colonies constitutes an untapped resource of symbiont plasticity or provides a possible explanation for the environmental flexibility of this coral genus."
} | 5,199 |
29862335 | PMC5968083 | pmc | 3,486 | {
"abstract": "Here, we demonstrate a very efficient simultaneous approach of bioenergy generation from wastewater and added-value compounds production by using a photosynthetic microalgae microbial fuel cells (PMFC), based on polybenzimidazole (PBI) composite membrane as separator. The use of PBI was proved to be very promising, even more convenient than Nafion™ in terms of energy performances as well as cost and sustainability. This polymer is also easily autoclavable, so allowing a re-use of the separator with a consequent beneficial cost effect. Two PMFCs were investigated: 1) Pt electrocatalysed and 2) Pt-free. They were operated as microbial carbon capture (MCC) device under continuous illumination, by using a domestic wastewater as anolyte and Scenedesmus acutus strain in the catholyte. The Pt-based cell allowed to generate higher volumetric power density (∼400 mW m −3 ) after more than 100 operating days. This resulted in an improved wastewater treatment efficiency, determined in terms of normalised energy recovery ( NER > 0.19 kWh kg COD −1 in case of Pt). The CO 2 fixation of the PMFC-grown microalgae leaded to a high accumulation of added-value products, namely pigments and fatty acids. A significant quantity of lutein was observed as well as a relevant amount of other valuable carotenoids, as violaxanthin, astaxanthin and cantaxanthin. The lipids were even excellently accumulated (49% dw ). Their profile was mainly composed by fatty acids in the range C 16-18 , which are particularly indicated for the biofuel production. These results demonstrate the feasibility and the implemented sustainability of such PMFCs as a great potential technology for the wastewater treatment and the simultaneous production of valuable products.",
"conclusion": "4 Conclusions Here, we described the performances of two PBI-based photosynthetic algae MFCs, one with Pt-electrocatalysed biocathode, the other one Pt-free, as devices for a synergistic approach in the simultaneous wastewater treatment and biorefinery. The PMFCs were operated under continuous illumination, using a microbial community at the anode, inoculated by a domestic wastewater, and Scenedesmus acutus microalgae at the biocathode compartment. They worked as microbial carbon capture cells, because the microalgae in the catholyte use CO 2 coming from the wastewater degradation at the anode as a carbon sources for photosynthesis and for the consequent oxygen production. The presence of an electrocatalyst greatly affects the electrochemical performances of the fuel cell, leading to higher power density, OCV and to a more efficient wastewater treatment. In contrast, no differences are observed for what concerns the microalgae properties, which are similar in both the reactors. The efficiency of such integrated bioelectrochemical system resulted particularly encouraging. The algal photocatalytic activity was enough to guarantee high cell vitality, quite fast growth kinetics and a rich production of added-value compounds, as pigments and lipids. Such products were obtained with significant yields, due to the stressing conditions of the MFC, which are favorable to a secondary metabolism and to the accumulation of reserve products. Indeed, neutral lipids amount of 49% with respect the biomass dry weight were quantified, whose majority part is composed by fatty acids in the range C 16-18 . For this reason, the PBI-based PMFCs are potential electrochemical devices for a more sustainable simultaneous bioenergy generation, biofuel production and accumulation of added-value compounds.",
"introduction": "1 Introduction Bioenergy is renewable energy made available from materials derived from biological resources (e.g. forest, agricultural and algae-derived biomass, biogenic fraction of municipal and industrial waste), which can be converted into biofuels, as well as directly into bioelectricity [1] . Nowadays, biofuels production seems to offer new opportunities to reduce greenhouse gases (GHG) emissions and to replace the conventional fossil fuels for uses in conventional engines [2] . Biofuel is typically produced by food, plants or animal oil, and/or by agricultural residues. Recently, a great deal of attention has been devoted to the production of biodiesel from microalgae [ 3 , 4 , 5 ], which seem to be an important source of bioenergy, sustainable both in terms of economy, because they use free CO 2 and free sunlight, and of environment, because they consume flue gas or CO 2 from the atmosphere and from industry. Microalgae possess an excellent photosynthetic efficiency and require less water than terrestrial crops to be cultivated. Moreover, some algal strain can have a very high lipid content (up to 50% on the dried system) [4] ; ii) and in general, according to the kind of strain and to the culture conditions in proper photobioreactors, microalgae can accumulate several important secondary metabolites with health and nutritional implications. In addition to lipids and fatty acids, in fact, other co-products with relevant commercial value may be obtained as proteins, sugars, or carotenoids, animal feed, super-food, advanced plastics for smart packaging. Exploiting of algae, therefore, has a dual benefit: it serves as biomass for the biofuels production and it has potential for the concurrent extraction of valuable co-products with a wide range of applications [ 4 , 6 ]. Currently, there are numerous technologies for microalgae biomass production, which can be based on photoautotrophic or heterotrophic growth [3] . Among them, microalgae production based on the integration with microbial fuel cells (MFCs) seems to be a very promising synergistic strategy, since phototropic organisms act as in-situ O 2 generators, which promote the bioelectrochemical reactions in MFCs. Basically, in a microbial fuel cell, CO 2 is produced at the anode as the oxidised product of an organic substrate (for instance wastewater), whereas O 2 is required at the cathode to accept protons coming from the anodic chamber and free electrons from the external circuits. In presence of microalgae and of light, photosynthesis occurs at the cathode, where CO 2 is metabolised by such microorganisms with the help of light to produce biomass. The general flow chain reaction may be stated from the following steps: 1) half reaction @anode (oxidation): Organics → CO 2 + H + + e − ; 2) half reaction @cathode (reduction): 2O 2 + 4H + + 4e − → 2H 2 O; 3) overall electrochemical reaction: i) Organics + O 2 → CO 2 + H 2 O + external power; ii) CO 2 + H 2 O + light → biomass + O 2 [ 7 , 8 ]. Electricity may be produced and harvested at the expenses of organic substrate decomposition, with algae growth in the cathodic chamber, and subsequent oxygen generation by the photosynthetic process. For this reason, these systems are commonly known as photosynthetic MFC (PMFCs) [9] . Several geometries were discussed in literature, as for instance tubular, coupled PMFCs, single-chamber PMFCs, sediment PMFCs, dual-chamber PMFCs, depending on the presence of algae at anode or cathode, the need of a chemical or biological mediator to allow the electron shuttling, the use of an ion exchange membrane (IEM), the coupling with a photo-bioreactor [8] . Among them, dual-chamber PMFCs with algae bio-cathode are particularly interesting, because they offer several advantages: i) the treatment and purification of urban or industrial wastewaters, ii) the growth of functional microalgae, iii) the harvesting of bioelectricity. The presence of microalgae at the bio-cathodes also makes this technology more sustainable in terms of costs, because it helps to replace the mechanical aeration methods. In addition, it reduces also the CO 2 generated from bacterial metabolisms and respiration [8] . PMFCs properties depend on several parameters: the dark/light regime, the geometry of electrode (e.g. brush or planar), the electrode distance, and the family of the microalgae ( Scenedesmus, Chlorella vulgaris, Chlamydomonas reinharditi , etc.) [ 7 , 10 , 11 , 12 , 13 , 14 ]. Oxygen production is normally sufficient for the optimum operation of PMFCs, even if the electricity generation may be not constant due to fluctuations of diluted O 2 concentration, which is dependent on the illumination of the chamber. Furthermore, the use of electrode without precious metals leads to good performances and, in particular, carbon brush cathodes allow reaching power density 6.5 times higher than plain electrodes (30 mW m −2 vs 4.6 mW m −2 ). Key-factors as light, nutrients, and pH are also important parameters contributing to the biomass growth and composition. Under stressing limitation conditions, in fact, microalgae increase the production of lipids, carbohydrates and pigments [ 11 , 15 ]. Despite these promising results, some bottlenecks are still present, concerning the understanding of biologic aspects related to algae growth mechanisms, but also PMFCs performances, which gradually worsen due to issues related to the biocathode, as the cell start-up time, the sensibility of pH-splitting, biofouling, substance crossover across the membrane, catalysts poisoning and adjustment of CO 2 , light and nutrients supply to microalgae, in addition to the MFC costs [ 7 , 8 ]. All these aspects may be improved by optimising the cell design and materials, namely ion exchange membrane (IEM) and electrodes, in order to increase the power output and to reduce the cost of the whole device. Regarding the IEM, which plays a fundamental role to avoid contamination between bacteria and algae [16] , the research should find materials cheaper and electrochemically more performing as MFCs separators [17] . Recently, we described a microbial fuel cell with a pyridine-PBI composite as proton exchange membrane (PEM) for the treatment of municipal wastewater. Such device resulted extraordinarily promising in terms of membrane cost/power generation ratio [18] . This work aims at evaluating the potential of a PBI-based photosynthetic MFCs with algae-assisted biocathode for the simultaneous wastewater bioremediation and biomass growth, necessary for the production and subsequent extraction of products with relevant commercial value, such as lipids, fatty acids and carotenoids. A pure culture of Scenedesmus acutus microalga as the biocathode, and a returned sludge from a domestic wastewater treatment plant as the anode were used. The power generation of the cell was investigated under continuous illumination. The content of valuable products as oils, lutein and other carotenoids produced by microalgae through such bioelectrochemical route was evaluated. The aim was to verify how the stress conditions occurring in the MFC reactor affect the algae metabolism and in particular the accumulation of important secondary metabolites with nutritional and health properties.",
"discussion": "3 Results and discussion 3.1 Performance of the photosynthetic microbial fuel cells 3.1.1 PMFC operation As already described before, the aim of this work is the investigation of PMFCs with microalgae-assisted biocathode, which can produce O 2 , namely the electron acceptor for electricity generation, by means of photosynthesis, without any need of aeration. In particular, these devices behave like microbial carbon capture cells (MCCs), because under light illumination the microalgae in the cathode chamber can use CO 2 coming from the wastewater degradation at the anode as a carbon source for photosynthesis and, consequently, for the oxygen production. Two PMFC reactors were simultaneously studied, both of them consisting of similar bioanode, inoculated by domestic wastewater, and biocathode, inoculated by the same algae culture ( Scenedesmus acutus PVUW12 ). The two compartments were externally connected to pipe CO 2 . In one case, the cathodic gas diffusion electrode (GDE) was Pt-free, in the other case, a Pt-coated electrode was used (Pt loading = 0.5 mg cm −2 ). The proton exchange membrane was a PBI-based composite, which is significantly promising as separator for MFCs for what concerns functional performances as well as cost. In our recent study, in fact, we showed that such a system allows a strong increase of power density, durability and wastewater efficiency with respect to Nafion™ [18] . Moreover, this kind of membrane does resist against sterilization processes in autoclave at 120 °C under water, without any mechanical/structural degradation, so allowing a re-use of the separator. This is certainly a beneficial aspect in terms of reduced cost and implemented sustainability of the MFC technology. The functional performances of the PBI-based PMFCs were investigated by means of long-term operation experiments over about 100 days under continuous illumination. Both the cells were kept under the same conditions of substrate, fed-batch cycles and algal inoculum regeneration, necessary when the culture stopped to grow (after ∼3 weeks, as better described in the following sections). The electrochemical characterization was performed by collecting polarization curves every 3 days, after that the microbial acclimation was complete (typically when the open circuit voltage, OCV , did not change more than 5–7% during one day). The polarization was carried out at low voltage scan rate (0.1 mV s −1 ) in order to avoid power overestimation [26] . Fig. 3 a–b shows, as an example, the voltage and power density ( PD v ) plots vs. the current density for the Pt-based and Pt-free PMFCs, after 25 and 62 operation days, respectively, which correspond to the obtained maximum values during time operation (see below). In order to rule out variations of electrode configuration and size (e.g. surface area), volumetric power density (mW m −3 ) was used as the parameter to describe the MFC performances [24] . Both the devices did not evidence any curve distortion, and this is an indication of a stable biofilm deposited onto the anode during the anode-enrichment period and the fed-batch cycles. However, the figures show different MFC performances. As expected, the presence of an electrocatalyst (Pt) at the biocathode was beneficial for the cathodic oxygen reduction reaction (ORR) and, consequently, in terms of generated power density. Table 1 compares the MFCs properties observed in the cells, showing better performances when Pt was used as a catalyst. Indeed, the maximum power density generated by the Pt-based devices exceeded 1200 mW m −3 ( PD MAX = 1217 ± 58 mW m −3 ), which is almost 10 times higher than that one observed in case of a Pt-free GDE, where PD MAX of 170 ± 16 mW m −3 was obtained. These values are quite in line with what reported in literature for PMFCs with microalgae at the cathode [ 8 and refs. therein cited], even if a thorough comparison is challenging because the functional performances of such devices may strongly depend on several parameters, including MFC architecture, light intensity, illumination period (continuous or dark cycles), nature of wastewater and microalgae strain. However, the obtained values of power density confirmed the capability of the algae to produce enough oxygen for optimal MFC operation. Fig. 3 Polarization and power density curves of the investigated Scenedesmus acutus PMFCs. (a): Pt-free biocathode; (b): Pt-biocathode. Fig. 3 Table 1 Operation parameters ( OCV ; power density normalized vs. the cathode surface, PD c/a and vs. the cell volume, PD v ), and treatment efficiency ( COD removal rate, COD r.r. ; normalized energy recovery, NER; CO 2 production) of the two investigated PMFCs. As stated in the experimental details, the electrochemical measurements were carried out in duplicate. Error on PD data lower than 10%. Table 1 PMFC OCV (V) PD c/a (mW m −2 ) PDv (mW m −3 ) COD r.r. (g m −3 h −1 ) NER (kWhkg COD −1 ) CO 2 production (g m −3 h −1 ) Pt-free 0.50 1.3 98 3.3 0.196 3.3 Pt 0.64 5.3 400 3.4 0.014 3.3 In order to check the durability of the two PMFCs (Pt-free and Pt-coated cells), long-term experiments were carried out, by monitoring functional parameters as the generated maximum power density, open circuit voltage ( OCV ) and the amount of dissolved oxygen ( DO ), which is produced at the cathode compartment by algae photosynthesis. Fig. 4 a–c shows the behavior of volumetric PD MAX , OCV and DO vs. the operating time, respectively. In the Pt-free PMFC reactor the power density initially increased up to a maximum of about 170 mW m −3 after ∼60 days, then it gradually decreased to a value of 108 ± 10 mW m −3 that remained constant until the end of experiment ( Fig. 4 a). The Pt-based cell showed a comparable trend but, in this case, a remarkably higher power density was generated throughout the entire experiment. Indeed, after at least 100 days PD v of 400 ± 31 mW m −3 was obtained, which is 4 times higher than that observed in the Pt-free device. Fig. 4 Long-term performances of the investigated PMFCs. (a) volumetric power density; (b) OCV; (c) dissolved oxygen. In part (b) and (c) the error is inside the symbol dimensions. Fig. 4 Similarly, the Pt-MFC open circuit voltage was significantly higher than that one measured in the Pt-free cell. Fig. 4 b shows the OCV behavior vs. the operating time for both the systems. Average values ranging between 0.8 V and 0.6 V were found with the cathode containing Pt, whereas values gradually increasing up from 0.35 V to 0.5 V were detected in case of Pt-free cell, in agreement with the voltages generally reported in the literature, which are typically below 0.6 V [26] . The behavior of the actual cell voltage against the operating time is a quite complex phenomenon. Indeed, it depends on the contribution of several factors, as for instance the charge transfer process, mass transport and in particular the bacteria metabolism [26] . Basically, the open circuit voltages remained stable for several days. When the voltage generation decreased below 0.3 V, the anolyte containing the wastewater substrate was regenerated [14] . The better performances of the Pt cell seemed to be exclusively due to the Pt catalytic effect on the ORR reaction at the biocathode. In fact, with respect to other critical aspects, such as membrane biofouling and oxygen production, the two cells showed a totally overlapping response. Regarding the separator biofouling, for instance, the PBI-composite did not show any evidence of microbe adhesion on the membrane surface, as we already reported in our recent paper [18] . Furthermore, similar reproducible cycles of O 2 production at the algae-assisted cathode compartment were generated by both the system. A maximum oxygen amount of about 25 mg L −1 was always produced by MFC-grown Scenedesmus ac. that remained stable within the overall algae growth cycle, namely 3 weeks before the addition of a fresh inoculum ( Fig. 4 c). Despite the higher power density, Pt-based MFC suffered higher performance loss during the operating time with respect to the Pt-free reactor. Such a result may be likely attributed to a significant presence of crystals or agglomerates, produced during the electrochemical process, which were deposited onto the biocathode, so covering the electroactive sites of its surface. This is well evident by comparing the results of scanning electron microscopy carried out on the two devices. Fig. 5 a–d reports the SEM images of the Pt-free GDE (a), Pt-GDE (b) of the tested PMFC reactors, and those corresponding to the pristine biocathode @t = 0 (before MFC functional tests), (c) Pt-free GDE and (d) Pt-based GDE, respectively. The images show several large particles (>10 μm) that are copiously distributed along the surface only in case of the electro-catalyzed cathode. Fig. 5 SEM analysis of the biocathode surface (membrane side). (a): Pt-free PMFC; (b) Pt-PMFC and of the pristine biocathode (before MFC functional tests). (c) Pt-free GDE; (d) Pt-based GDE. Fig. 5 3.1.2 Efficiency of the wastewater treatment The PMFCs efficiency in the treatment of the domestic wastewater was investigated by measuring the chemical oxygen demand ( COD ) during the operation time, and the normalized energy recovery ( NER ), defined as the energy production in terms of kilowatt-hour per kilogram of removed COD \n [24] . Fig. 6 shows the COD removal as a function of time after the cell stabilization, calculated as the ratio between the removed and influent chemical oxygen demand values. This parameter reveals the amount of the organic substrate, converted into bioelectricity by the MFC reactor. In both cases, the substrate degradation is efficient. Indeed, a COD removal exceeding 87% was obtained during the first 19 days of the long-term experiment. This is in very nice agreement with what already found in our recent work on PBI-based open-to-air-cathode MFC for wastewater treatment [18] . After this time the anolyte was re-feed in order to continue the investigation of the cell durability. Fig. 6 Efficiency of the Chemical Oxygen Demand, COD , removal for the investigated PMFCs. The error is inside the symbol dimensions. Fig. 6 As introduced before, the COD removal is used to determine the normalized energy recovery, NER , which is a convenient way to describe MFC energy production in kWh m −3 , taking into account the wastewater treatment capacity. Recently, it was discussed that NER allows a better understanding of the actual organic substrate conversion into energy with respect to the net power density or the coulombic efficiency, ε c . It also makes possible a more appropriate cross-comparison of the several MFCs proposed in the literature, which is still a critical issue when only power density is used in the evaluation of the wastewater treatment efficiency [24] . NER parameters were calculated both in case of Pt-free and Pt-PMFCs by using Eq. (2) , as reported in the Experimental Section. As reported in Table 1 , NER = 0.196 kW h kg COD −1 and 0.014 kW h kg COD −1 was the energy production determined in the two reactors, respectively. This result suggests a remarkably more efficient treatment in presence of Pt at the biocathode (more than a factor of 10), which could likely mitigate the cost issue due to the use of a precious metal as the electrocatalyst. 3.2 Microalgae biomass production: efficiency of the synergistic strategy and biorefinery Investigation on some important aspects, such as i) cell growth rate; ii) chlorophyll content; iii) composition of some carotenoids, and iv) total triglycerides content and their identification, was carried out on the Scenedesmus acutus culture grown in the MFC reactors. These data were necessary to assess the efficiency of such an integrated system to produce energy, treat wastewater and, in the same time, to accumulate several important secondary metabolites. Regarding these properties, it was observed that the presence of the Pt electrocatalyst at the biocathodes does not affect in any way the growth kinetics and the production of valuable products. Similar photosynthetic activity and pigments/lipids composition were, in fact, obtained in both the tested PMFCs. 3.2.1 Microalgae growth kinetics Fig. 7 shows an optical microscopy image (500X) of the Scenedesmus acutus population cultivated in the PMFC reactor. As already described before, the cell density was measured by settling the microalgae in a Burker chamber and counting them via microscopic route. Fig. 7 Optical microscopy image (500X) of the Scenedesmus acutus strain, grown in the PMFC. Fig. 7 Fig. 8 describes the growth of Scenedesmus in the PMFC determined as described by Damiani et al. [21] . The initial cell concentration and biomass content were 10 6 cells mL −1 and 0.02 g L −1 , respectively. Basically, the growth process follows a quasi-exponential behavior. The algae needed few days (about 4 days) of adaptation before enhancing their biomass. The concentration increased of one order of magnitude, up 1.4 ± 0.1 ×10 7 cells mL −1 , then a growth recovery could be observed after 22 days. Subsequently, the algal cell number began to decrease, as expected, because of the environmental stressing conditions typical of the PMFC, namely exhaustion of nutrients in cultural medium and oxidative stress. The cell growth rate, μ , determined only in the exponential phase, is 0.17 cell mL −1 day −1 with a doubling time, t D , of ∼6 days. The biomass collected from the culture grown in MFC reactor was 0.29 ± 0.2 g L −1 after 22 days of cultivation. The good growth rate accompanied by the great biomass production, registered in algal population cultivated in the MFC reactor, was essentially due to enough CO 2 produced by bacteria at the anode of the MFC reactor during the electrochemical wastewater degradation of the substrate. The quantity of produced carbon dioxide was ∼3 g m −3 h −1 in both the PEMFC reactors (Pt and Pt-free). It is well known in literature the positive effect of CO 2 on the algal biomass concentration and cell growth, which also depends on factors as light intensity, algae family, gas flow rate etc. [27] . This result is an index that the MFC electrochemical processes properly run in terms of bioenergy production as well as harvested CO 2 , whose concentration seems to be enough to stimulate the cell growth and to fix gas into added value products. Fig. 8 Growth rate of the Scenedesmus acutus strain, μ , grown at the PMFC biocathode. Fig. 8 3.2.2 Migroalgae photosynthetic activity The chlorophyll concentration of Scenedesmus acutus was measured by means of spectrophotometric methods, as described in the experimental section. The value obtained in case of MFC-grown culture was 106 ± 3 μg mL −1 . This value is satisfying and suggests that the environmental stress of the MFC reactor was not so high to inhibit the photosynthetic activity of algae. In order to further check the cell vitality, oxygen production measurements were carried out on the Scenedesmus acutus , cultivated in the PMFC reactor. Fig. 9 shows the amount of oxygen produced by microalgae, arranged under illumination after 16 hours dark phase (see Fig. 7 ). The time-dependent behavior of the O 2 production was similar for both types of PMFCs (with- and without Pt at the biocathode). The oxygen dissolved in the catholyte increases exponentially up to ∼25 mg L −1 within 7 hours, before reaching a plateau. This content, whose value is in fair agree with what recently reported in literature [28] , seems to be enough for the electricity generation of MFC and the microbial growth inoculated in the anodic chamber. Fig. 9 Photosynthetic activity of the PMFC-grown Scenedesmus acutus culture, measured in terms of O 2 production in the reactor. The error is inside the symbols dimensions. Fig. 9 3.2.3 Production of added-value products: carotenoids and fatty acids Scenedesmus acutus is considered an efficient strain for CO 2 mitigation and for pigments and lipids accumulation [ 6 , 23 ]. In order to evaluate the efficiency of the algae-assisted PMFC device to produce carotenoids and fatty acids, qualitative/quantitative analyses were carried out on the culture grown in the MFC during a long-term functional experiment. The results were compared with the pigments and total lipid amount found in the culture grown in a flask at the same conditions of light, temperature and pH of the MFC reactor. The carotenoids composition was investigated by HPLC analysis, which revealed a similar total content of pigments for both types of tested fuel cells (Pt-free and Pt), namely 3.7 ± 0.5 and 3.9 ± 0.7 mg g dw −1 , respectively. In terms of composition, data showed a very relevant level of lutein and a significant level of other carotenoids, as astaxanthin, violaxanthin and cantaxanthin. Fig. 10 reports the content of these pigments, expressed as mg per gram of dried biomass, in Scenedesmus culture cultivated in the PMFC reactor. Fig. 10 Pigments extracted from the microalgae-grown PMFC. Data from HPLC analysis. Fig. 10 The lutein content in Scenedesmus culture in PMFC was more than 15 times higher than the average level of other carotenoids, and almost the double with respect to that one found in flask (1.85 mg g dw −1 ), whereas the lowest level was obtained for the cantaxanthin (0.16 mg g dw −1 ), namely 1.4 times lower than the average value of astaxanthin and violaxanthin. In contrast, no other carotenoids, besides lutein, were observed in the flask culture. This obtained result demonstrated that the PMFCs operative conditions stimulate the biosynthesis of pigments. Therefore, the high amount of lutein and the presence of the other ketocarotenoids with a very important antioxidant effect in the algal cell grown in the MFC reactors are probably due to the environmental stress conditions able to induce in algal cells an accumulation of these xanthophylls that play a fundamental role in the protection of the photosystems against the oxidative stress. The production of other carotenoids, besides lutein, in algae grown in a PMFC was previously described in literature, for instance in presence of Chlorella vulgaris and interpreted in terms of faster carotenogenesis favored by the fuel cell environment [29] . The lipid composition of the Scenedesmus acutus was preliminary investigated by TLC. This qualitative analysis method allowed to separate different kinds of lipids. Fig. 11 shows TLC runs for the lipids extracted from microalgae culture grown in MFC. Comparing the resulting spots with those reported in the literature at the same TLC conditions [30] , it was possible to identify monoacyglicerols (MAG), diacylglicerols (DAG), free fatty acids (FFA) and triglycerides (TAG) spots. The different intensity of the spots showed a highest amount of monoacylglicerols and a lowest presence of diacylglicerols. In particular, the spot intensity of free fatty acids and triglycerides, which are the most interesting lipids for nutritional aims, resulted very well defined. Fig. 11 TLC profiles of oil content in Scenedesmus acutus grown in PMFC. Fig. 11 The effects of the electrochemical environment on the total lipids accumulation in Scenedesmus acutus were quantitatively evaluated by means of GC-MS analysis. In particular, fatty acids were identified and quantified after methyl esterification to form the corresponding fatty acid methyl esters (FAMEs). The total content of neutral lipids accumulated in the MFC-grown microalgae was remarkably higher than that found for the culture grown in flask, namely 49% and 6% of the biomass dry weight, respectively. Such a high FAs concentration is particularly encouraging because the content of neutral lipids accumulated in PMFC-grown microalgae is typically lower than 40%. Recently, lipid productions ranging around 30–37% were observed in different microalgae strain (for instance Golenkinia , Chlorella vulgaris , Scenedesmus SDEC-8 and SDEC-13 ), used as biocathode in a PMFC for kitchen waste effluent treatment [28] . In case of the Scenedesmus species, a total lipid content of about 30 wt% was found, despite of a faster growth process. The fatty acids concentration in algal culture grown in MFC reactor described in this work is even higher that that obtained for the same strain ( Scenedesmus acutus PVUW12 ) previously cultivated in a tubular photobioreactor placed outdoor, where microalgae were grown in nitrogen-free medium (N-stress). In this case, about 32% of total lipids were accumulated with a predominance of monounsaturated fatty acid (oleic acid) with respect to the saturated ones, namely palmitic and stearic acids [23] . Obtained results confirm how the stress conditions in the MFC lead to an accumulation of reserve substances that may ensure survival and recovery of the microalgae on long times. Nutrient depletion and oxidative stress might direct the algal metabolism towards the synthesis of reserve compounds and secondary metabolites such as lipids and protective pigments, rather than primary metabolites and cell division. The FAMEs profile, obtained from the Total Ion Current (TIC) Chromatograms shown in Fig. 12 , is reported in Table 2 . Saturated lipids are mainly accumulated, even if some amounts of unsaturated systems are also present. The reason of such a result may be related to the low content of CO 2 coming from the anodic chamber to aerate the microalgal biocathode, which favors the accumulation of saturated fatty acid and short-chain fatty acids. It was described in literature, in fact, that a decrease of the feeding gas amount could lead to a O 2 increase with a negative effect on the enzymatic desaturation and a consequent enhancement of the saturated FAs [ 26 and refs. therein indicated]. Fig. 12 Total Ion Current (TIC) chromatogram of the alga extract (up) and corresponding zoomed spectrum in the range 16–60 minutes. Fig. 12 Table 2 Identification and composition of the FAMEs, extracted from the Scenedesmus acutus microalgae grown in a PMFC reactor. Data from GC-MS analysis. ( R t = Retention time). Table 2 R t (min) FAMEs MW (g mol −1 ) FAME composition (mg g −1 ) 16.97 Methyl myristate 242 0.40 19.21 Methyl pentadecanoate 256 0.25 21.04 Methyl hexadecatetraenoate 262 0.71 21.54 Methyl palmitoleate or isomer 268 0.61 21.69 Methyl palmitoleate or isomer 268 <0.12 22.50 Methyl palmitate 270 12.3 27.43 Methyl margarate (heptadecanoate) 284 <0.12 31.21 Methyl stearidonate 290 0.31 31.96 Methyl linoleate 294 0.51 32.48 Methyl oleate 296 3.07 32.96 Methyl vaccenate 296 0.41 35.00 Methyl stearate 298 28.8 49.94 Methyl arachisate 326 0.29 57.91 Methyl erucate 352 0.44 59.61 Methyl behenate 354 <0.12 The lipid composition in Scenedesmus acutus , grown in the MFC, is mainly in the C 16 -C 18 range, even if some amounts of C 14 and C 20 -C 22 are present as well, similarly to what typically observed in plant oil [27] . This result is in fair agreement with what reported in literature for Scenedesmus sp ., whose fatty acids profile is mainly represented by 14:0, 16:0, 16:4n1, and 18:3n3 along with minor levels of monounsaturated (16:1n7, 18:1n9) and polyunsaturated fatty acids (16: 3, 18:2 and 18:3n3) [ 31 , 32 ]. In our strain grown in the fuel cell, a relevant production of methyl palmitate (palmitic acid, C16:0), methyl oleate (oleic acid, C18:1) and above all methyl stereate (stearic acid, C18:0) was observed, in the fraction of 25%, 6% and 59% of the total lipid content, respectively. The predominance of C 16 -C 18 lipids is an index of the potential use of Scenedesmus acutus cultivated in PMFC reactors in biodiesel feedstock production."
} | 8,677 |
27152931 | PMC4859558 | pmc | 3,487 | {
"abstract": "Diatoms are eukaryotic microalgae that contain genes from various sources, including bacteria and the secondary endosymbiotic host. Due to this unique combination of genes, diatoms are taxonomically and functionally distinct from other algae and vascular plants and confer novel metabolic capabilities. Based on the genome annotation, we performed a genome-scale metabolic network reconstruction for the marine diatom Phaeodactylum tricornutum . Due to their endosymbiotic origin, diatoms possess a complex chloroplast structure which complicates the prediction of subcellular protein localization. Based on previous work we implemented a pipeline that exploits a series of bioinformatics tools to predict protein localization. The manually curated reconstructed metabolic network i LB1027_lipid accounts for 1,027 genes associated with 4,456 reactions and 2,172 metabolites distributed across six compartments. To constrain the genome-scale model, we determined the organism specific biomass composition in terms of lipids, carbohydrates, and proteins using Fourier transform infrared spectrometry. Our simulations indicate the presence of a yet unknown glutamine-ornithine shunt that could be used to transfer reducing equivalents generated by photosynthesis to the mitochondria. The model reflects the known biochemical composition of P . tricornutum in defined culture conditions and enables metabolic engineering strategies to improve the use of P . tricornutum for biotechnological applications.",
"conclusion": "Conclusion Our assembled reconstruction represents the current, comprehensive biochemical, genetic, and genomic knowledge about P . tricornutum and contains information such as reaction stoichiometry and associations between genes and reactions. We especially focused on lipid metabolism since diatoms are attractive candidates for industrial-scale lipid production [ 67 , 84 ]. The reconstruction is anticipated to facilitate model-driven exploration of the organism’s complex metabolism and hypothesis generation. Furthermore, the manually curated metabolic network facilitates visualization and analysis of different data types including metabolomics, fluxomics or common genomic data such as RNA-Seq. We have demonstrated that the model reflects the known biochemical composition of these algae in defined culture conditions ( Fig 4 ) and that it enables the study of light-dependent carbon partitioning ( Fig 6 ). Diatoms thrive in highly dynamic environments and this model will provide a template for future studies that aim to understand how diatoms balance photosynthesis and heterotrophic metabolism over light-dark cycles or the stochastic supply of nutrients. This model will also enable metabolic engineering strategies to improve the use of P . tricornutum for biotechnological applications.",
"introduction": "Introduction Diatoms are unicellular photosynthetic eukaryotes ubiquitous in marine and freshwater habitats and are responsible for about 20% of the photosynthetic carbon fixation on Earth [ 1 ]. Diatoms are evolutionary evolved from secondary endosymbiosis and harbor many genes of bacterial origin [ 2 ] which is predicted to give these microalgae a wide range of metabolic functions that are distinct from plants, green algae, and red algae [ 3 ]. Some of these distinct functions include the formation of silica nanostructures [ 4 ], the incorporation of an assimilatory urea cycle [ 5 ], and the breakdown of fatty acids in mitochondria and peroxisomes [ 6 ]. Diatoms also produce high intracellular concentrations of ω-3 fatty acids and other valuable compounds of biotechnological interest [ 7 ]. The marine diatom Phaeodactylum tricornutum is an emerging model diatom because of its relatively small genome (27.4 megabases) [ 2 ], ease of cultivation, and amenability to genetic engineering. Indeed, genetic systems in P . tricornutum may be the most advanced in microalgae, with the recently developed ability to assemble whole chromosomes in yeast [ 8 ], knock-out genes using TALEN technology [ 9 , 10 ], and introduce stable nucleus-localized episomes the size of small chromosomes via conjugation [ 11 ]. Previously developed technologies include transgenic gene overexpression [ 12 ] and gene expression knockdown using RNA interference or antisense transcript interference [ 13 ]. The development of these genetic engineering systems means that computationally directed experimental manipulations of the diatom genome are not only possible, but necessary. One promising strategy that investigates the yet unexplored metabolic capabilities of distinct organisms such as P . tricornutum is the metabolic network reconstruction, which enables computational analysis of systems-level responses. Genome-scale metabolic network reconstructions are derived from the annotated genome and contain information about all known metabolic reactions in an organism including the stoichiometry, subcellular localization, and the gene products by which they are catalyzed. The reconstruction process itself is laborious and iteratively and described, for example, in detail in [ 14 ]. The reconstructed network can be transformed into a genome-scale model of metabolism that can be used to predict metabolic phenotypes which are represented by flux distributions and have proven to be useful tools, for example, in the analysis of biological network properties, model-driven discovery, metabolic engineering and strain design [ 15 , 16 ]. Here, we report the reconstruction of a detailed and compartmentalized genome-scale metabolic model for P . tricornutum which provides a comprehensive insight into yet unexplored metabolic capabilities. We constrained the model with organism-specific biomass equations generated by Fourier transform infrared spectroscopy. The model predicts the presence of a surprising chloroplast glutamine-ornithine shunt that transfers reducing equivalents generated by photosynthesis to the mitochondria. Our findings demonstrate the utility of whole genome metabolic reconstructions to uncover unexpected biochemistries and to provide an important in silico template for directing future metabolic engineering efforts.",
"discussion": "Results and Discussion Metabolic network reconstruction Genome-scale network reconstructions are biochemically, genetically and genomically structured knowledge-bases which provide a framework to analyze and predict genotype-phenotype relationships. The reconstruction process is divided into four main steps [ 14 ] and summarized in Fig 1 . 10.1371/journal.pone.0155038.g001 Fig 1 Metabolic network reconstruction workflow. In step one we obtained a draft reconstruction based on P . tricornutum ’s genome annotation and reference reconstructions. This draft reconstruction was manually curated using several resources such as an improved genome annotation, subcellular localization predictions and external databases. All reactions were elementally and charge balanced, QC/QA was performed and a biomass objective function was defined before transforming the reconstruction into a computational model. In an iterative process, the in silico predictions are compared with experimental observations to validate and improve the metabolic model. First, we generated a draft reconstruction based on the P . tricornutum genome annotation and protein homology to template organisms having reconstructions [ 39 – 41 ]. Diatoms are taxonomically and functionally distinct from other algae and vascular plants; in fact, many nuclear genomic contents are more closely related to metazoans, demonstrating the diversity of diatom metabolism [ 2 ]. Although the diversity complicated the generation of a homology-based draft reconstruction, it also makes diatoms, such as the model organism P . tricornutum , attractive candidates for the analysis of cellular processes at a systems level, as they add to the biochemical diversity of microbes in a biotechnology setting, thereby increasing available production systems. Second, the draft reconstruction was manually curated and refined using additional resources such as the genome annotation, subcellular localization predictions and external databases (see Materials and Methods ). Once the manual curation was completed, the reconstruction was converted into a mathematical model in the third step. We added the biomass objective function and defined system boundaries (i.e., carbon and nitrogen uptake) according to experimental results (see Materials and Methods ). Qualitative tests were performed during the manual curation and the final step of model refinement and analysis. We verified that all biomass components and vitamins for which P . tricornutum is autotrophic could be produced under realistic growth conditions. Blocked pathways could be resolved with the addition of one or two reactions; in most cases transport reactions between intracellular compartments were missing. Furthermore, we ensured that ATP could not be produced without inputs. We also performed several in silico tests to assess the consistency of our model and verify that known physiological behaviors can be computationally reproduced. Diatoms are able to utilize a variety of nitrogen sources, both inorganic (such as nitrate and ammonium [ 54 ]) and organic (e.g. amino acids or urea [ 55 ]). Therefore, we examined the ability of the model to simulate biomass production on different nitrogen sources. Biomass was not produced in the presence of histidine, tryptophan, cysteine, or methionine as sole nitrogen sources in our initial in silico model, which contradicted literature results [ 55 ]. Histidine catabolism is not well understood in diatoms or plants and was not incorporated in the model at first. Since we could not identify genes that are involved in histidine catabolism in P . tricornutum , we added histidine catabolism as one lumped, low confidence reaction degrading histidine and water into ammonium, formamide and glutamate. Formamide is split into formate and ammonium with formate accumulating during histidine catabolism in silico ; a demand reaction was added to allow the accumulated formate to leave the system. Biomass production for growth on methionine or cysteine as sole nitrogen sources was achieved by adding a demand reaction for dimethylsulphoniopropionate (DMSP). DMSP levels are known to increase with light intensity or nitrogen starvation but its metabolism is not well understood in diatoms and while the biosynthetic pathway is currently unknown [ 56 ], a sensible starting point would be an amino acid with an already reduced sulfur atom. Indole accumulation prohibited growth on tryptophan as nitrogen source. To account for the unknown indole degradation, a demand reaction was added. With these changes, the model could simulate biomass production using the different nitrogen sources tested. Leveraging a genome-scale model in the exploration and contextualization of lipid metabolism requires an accurate representation of the metabolic pathways and intermediate metabolites. To this end, a lipid module was developed ( i LB1027_lipid, see S3 File ) that encompasses the full range of lipid metabolites and metabolic reactions. This module allows incorporation of experimental fatty acid and lipid class characterization to be reflected in the biomass composition. Incorporation of experimental FAME data was possible via a linear optimization based data fitting algorithm (see Materials and Methods ). After fitting the model to the data, the deviation from the experimental values to the model was 350 times lower in the lipid module compared to the core model. This result demonstrates the utility of the lipid module when investigating fatty acid and lipid metabolism in P . tricornutum . The curated genome-scale metabolic network for P . tricornutum including the lipid module, i LB1027_lipid, accounts for 1,027 genes associated with 4,456 reactions and 2,172 metabolites distributed across six compartments (Tables M-O in S2 and S3 Files). Compared to the draft reconstruction, the number of genes (446 genes) was more than doubled during the manual curation phase. All reactions are associated with at least one of 90 subsystems which can be categorized into ten groups, e.g., carbon or lipid metabolism ( Fig 2 ). Additionally, a core model with substantially reduced lipid metabolism ( i LB1025) was constructed. The reduced lipid metabolism subsystem accounts for 1,029 reactions compared to 3,325 reactions involved in lipid metabolism in i LB1027_lipid. The core model yields comparable flux distributions and is suitable, for example, if detailed data on the lipid composition under the simulated condition are missing. 10.1371/journal.pone.0155038.g002 Fig 2 Reconstruction characteristics i LB1027_lipid. (A) Reactions per subsystem. Most reactions are involved in lipid metabolism. Our FTIR measurements underline the fact that the lipids make up the highest fraction of biomass. Due to the presence of multiple compartments and the fact that many pathways are split among compartments, many reactions are attributed to intracellular transport. The modeling subsystem contains ATP maintenance, biomass, demand, sink, and exchange reactions. (B) Percent reactions and metabolites per compartment. Most reactions and metabolites are present in the cytosol, followed by chloroplast and mitochondria in the case of reactions and mitochondria and chloroplast for metabolites. Peroxisome, extracellular space, and thylakoid contain less than 5% and 8% of all reactions and metabolites in the reconstruction, respectively. Prediction of enzyme subcellular localization One challenging aspect of eukaryotic reconstructions is the subcellular localization prediction of proteins. Due to their endosymbiotic origin, photosynthetic heterokonts including diatoms possess chloroplasts that are surrounded by four membranes. This complex structure concurs with distinct plastid targeting signals in diatoms that restrict the use of available subcellular prediction tools for other eukaryotes. We enhanced a previously developed pipeline which combined different bioinformatics programs to predict the subcellular localization of proteins in diatoms [ 57 ] (see Fig 3 , Materials and Methods , and Section A in S1 File ). 10.1371/journal.pone.0155038.g003 Fig 3 Subcellular localization prediction pipeline. Schematic representation of the implemented subcellular localization prediction pipeline for Phaeodactylum tricornutum adapted from previous work [ 57 ]. Subcellular compartments are given in ellipses and bioinformatics programs are displayed in rectangles. Our added steps are highlighted in gray. The ER retention signal is (K/D)-(D/E)-E-L in the protein C-terminal region. A protein is categorized as peroxisomal if the signal (S/A/C)-(K/R/H)-(L/M) or S-S-L is found in the C-terminal region. To evaluate the accuracy of the improved pipeline, we compared our predictions to Sunaga et al .’s results and experimentally validated subcellular protein localizations taken from [ 5 , 12 , 58 – 64 ]. By using the refined pipeline, 15 out of 19 subcellular localization predictions coincided with experimental data as summarized in Table 1 . 10.1371/journal.pone.0155038.t001 Table 1 Validation of the in silico subcellular localization prediction pipeline. Protein ID Status Phatr3 Experimental Prediction original pipeline Prediction improved pipeline TargetP MitoProt Note Phatr2 Phatr3 localization Phatr2 Phatr3 Phatr2 Phatr3 Fructose-bisphosphate aldolase Phatr_825_bd 304098 Kept Chloroplast [ 58 ] Chloroplast Chloroplast Chloroplast Chloroplast ER 0.2507 Glyceraldehyde-3-phosphate dehydrogenase Phatr_25308 301292 Kept Mitochondrion [ 59 ] Mitochondrion Mitochondrion Mitochondrion Mitochondrion Mitochondrion 0.9647 Glyceraldehyde-3-phosphate dehydrogenase Phatr_22122 308678 Kept Chloroplast [ 59 ] Chloroplast Chloroplast Chloroplast Chloroplast Chloroplast 0.946 Transaldolase Phatr_20779 304241 Kept Chloroplast [ 58 ] Chloroplast Chloroplast Chloroplast Chloroplast Mitochondrion 0.9717 Glutamine synthetase Phatr_22357 310769 Kept Mitochondrion [ 12 ] Mitochondrion Mitochondrion Mitochondrion Mitochondrion Mitochondrion 0.9972 Glutamine synthetase Phatr_51092 306624 Kept Chloroplast [ 12 ] Chloroplast Chloroplast Chloroplast Chloroplast ER 0.4399 Carbamoyl phosphate synthase Phatr_24195 309585 Modified Mitochondrion [ 5 ] Cytoplasm Mitochondrion Cytoplasm Mitochondrion Mitochondrion 0 . 9588 As mentioned by Sunaga et al ., the Phatr2 gene model is incomplete . Usage of Phatr3 gene model yields correct prediction . Fructose-1,6-bisphosphatase Phatr_54279 311663 Kept Chloroplast [ 58 ] Chloroplast Chloroplast Chloroplast Chloroplast ER 0.1645 Phatr2 ID 2793 shorter version of 54279. Δ12 desaturase Phatr_48423 300552 Kept Chloroplast [ 60 ] ER ER Chloroplast Chloroplast Chloroplast 0 . 813 Usage of improved pipeline yields correct prediction . Note that we could not reproduce Sunaga et al . 's result using the original pipeline . ATPase δ subunit Phatr_20657 301964 Modified Chloroplast [ 61 ] Chloroplast Chloroplast Chloroplast Chloroplast Chloroplast No prediction PtCA1 Phatr_51305 306874 Kept Chloroplast [ 62 ] Chloroplast Chloroplast Chloroplast Chloroplast ER 0.5189 Triosephosphate translocator Phatr_24610 308968 Kept Chloroplast [ 63 ] Chloroplast Chloroplast Chloroplast Chloroplast ER 0.6211 CA-I Phatr_35370 303871 Kept Chloroplast [ 64 ] ER ER Chloroplast Chloroplast Mitochondrion 0 . 4648 Usage of improved pipeline yields correct prediction . CA-II Phatr_44526 311660 Modified Chloroplast [ 64 ] ER ER ER ER ER No prediction Wrong prediction. CA-III Phatr_55029 300968 Modified Chloroplast [ 64 ] ER No prediction ER No prediction - No prediction For Phatr_55029, the pipeline predicts endoplasmic reticulum instead of chloroplast. The improved gene model 300968 was extended at the N-terminus and does not start with methionine anymore. Mitoprot does not predict localization if the protein does not start with methionine and, therefore, the pipeline does not predict localization for the Phatr3 gene model. PtCA2 Phatr_45443 311919 Kept Chloroplast [ 62 ] Chloroplast Chloroplast Chloroplast Chloroplast ER 0.4762 CA-VI Phatr_54251 303635 Kept Chloroplast [ 64 ] ER ER ER ER ER 0.3987 Both pipeline versions predict endoplasmic reticulum instead of chloroplast. CA-VII Phatr_42574 304857 Modified Chloroplast [ 64 ] ER ER ER ER ER No prediction Both pipeline versions predict endoplasmic reticulum instead of chloroplast. CA-VIII Phatr_20030 311877 Kept Mitochondrion [ 64 ] Mitochondrion Mitochondrion Mitochondrion Mitochondrion Mitochondrion 0.932 Phatr2 ID 35304 shorter version of 20030. The table compares predictions of protein localizations to experimental data. For all considered proteins, Phatr2 and Phatr3 IDs and the status of the gene model in Phatr3 are given. If the gene models were modified, the pipeline predictions for both gene models are given. We distinguish between two versions of the in silico pipeline; original refers to the version as published by Sunaga et al . [ 57 ] and the improved version is the one presented in this study. Entries for which the improved pipeline or usage of Phatr3 gene models improved the prediction are formatted italic. Discrepancies between prediction and experimental localization are shown in bold. ER: Endoplasmic reticulum. Determination and modeling of biomass composition In order to mathematically solve the genome-scale model using FBA, the observed cellular phenotype is manifested as a biological objective function [ 51 ]. This objective function is a metabolic reaction in the model that is maximized or minimized in order to achieve a desired phenotypic state. In order to simulate cellular growth, the macromolecular constituents of the cell are defined as the objective function (see Table L in S2 File ). This biomass objective function accounts for all known cellular components and their fractional contributions to the overall cellular biomass, defines the anabolic requirements for cell division, and provides mass balance. The biomass composition used in heterotrophic genome-scale models is typically fixed based on experimentally derived values at a given culture condition [ 65 ]. However, phototrophic organisms have a dynamic biomass composition that changes not only across the diel cycle, but also along the duration of the culture. In P . tricornutum , biomass changes in the light period is dominated by the generation of carbon storage compounds, while the dark period is dominated by the anabolic processes necessary for cell division [ 66 ]. There is also dramatic remodeling of the cellular biomass composition that accompanies nutrient limitation in diatoms [ 67 ]. High confidence intracellular flux predictions are dependent on the biomass composition being accurately reflected during the simulation. To this end, we determined P . tricornutum ’s biomass composition over a growth curve that resulted in nitrogen deprivation after the high accumulation of biomass ( Fig 4 ). Selected samples of this growth curve were examined using time consuming biochemical methods for determining lipid, carbohydrate, and protein content of the cells. Parallel samples were used to develop linear models relating FTIR peaks to biomass composition (Fig A in S1 File ). These calibrated models were then used to determine the biomass composition for all time points. The linear models are most robust when a large gradient for biomass composition values (i.e., percent lipid, protein, and carbohydrate) are achieved, thus our experiment was designed to maximize the changes in content. Nitrogen starvation, low CO 2 , and low light all can contribute to high lipid content and all three scenarios were achieved in our engineered culture experiment, resulting in very high lipid values at the end of the experiment ( Fig 4C ). The lipid values are elevated relative to previous experiments that examined more realistic bioproduction conditions, but this was planned and resulted in the expected fashion. We were able to achieve large changes in the cellular contents for all of these cellular components in smooth gradients. 10.1371/journal.pone.0155038.g004 Fig 4 FTIR spectrum and culture data. A typical FTIR spectrum for Phaeodactylum tricornutum is shown in (A). Peaks corresponding to lipids, proteins and carbohydrates are highlighted (see Table A in S1 File for specific wavelengths). Panel (B) shows the growth curve and photosynthetic efficiency of the culture used for model calibrations and the biomass objective function. The decline in F v /F m indicates the onset of nitrogen starvation (n = 1). Percent dry weight of the cells in terms of carbohydrates, lipids, and proteins according to FTIR spectra and the calibrated linear model (n = 5, error bars represent five independent FTIR scans) is displayed in (C). Additionally, FAME data at each sample point was incorporated into the biomass composition via a linear optimization based fitting algorithm to ensure changes in fatty acid biosynthesis were taken into consideration during simulations (see Materials and Methods and Section B in S1 File ). Interestingly, diatoms store large amounts of nitrogen in the cell in the form of inorganic compounds [ 30 ], probably in the vacuole [ 68 ]. A demand reaction for NO 3 was added to account for cellular nitrate that has not yet been assimilated into other biomass components such as proteins but is included in the dry weight measurements. By defining the cellular composition at each sampling point, differences in the metabolic network usage could be analyzed along the duration of the culture. Commonly, maximizing the biomass equation is selected as an appropriate objective function for the growth phenotype. Since cell division in P . tricornutum is relegated to the dark period when cells are grown in a light-dark regimen, the common biological objective function of maximizing growth is not applicable to simulations during the light period. Thus, maximizing carbon uptake was selected as the biological objective function that best represents the cellular phenotype during the light period. Mass balance was achieved by allowing fixed carbon to accumulate as either carbohydrates or neutral lipids in accordance with previous observations of P . tricornutum [ 66 ]. Comparison to other models Several metabolic models for P . tricornutum have been constructed to date ( Table 2 ). Kroth and coworkers investigated the localization of enzymes and pathways involved in carbohydrate metabolism [ 69 ]. This model served as foundation for the first genome-scale model for P . tricornutum which was presented in form of a detailed pathway/genome database named DiatomCyc [ 45 ]. DiatomCyc comprises a high number of pathways and offers different software tools, e.g. for network analysis, but it lacks subcellular compartments which are important to account for distinct environments required for different metabolic processes. A smaller and compartmentalized version of the DiatomCyc metabolic network was used to compute elementary flux modes and investigate light-dependent changes in P . tricornutum ’s metabolism [ 70 , 71 ]. Here, little information about the reconstruction process is given and reactions and metabolites are poorly annotated. Kim et al . developed the most recent genome-scale metabolic network for P . tricornutum and explored flux distributions for autotrophic, mixotrophic and heterotrophic growth conditions [ 72 ]. For all three modes, the same biomass objective function was exploited. The prediction of protein localization was based on MitoProt [ 19 ] and TargetP [ 22 ]. Reactions are annotated using EC numbers which might be ambiguous and hamper clear identification of reaction mechanism or model comparison based on reaction content. Gene reaction associations are not formulated as Boolean rules making it impossible to distinguish between isozymes, enzyme complexes, or subunits. No information about the performance of quality control or mass and charge balancing is given. 10.1371/journal.pone.0155038.t002 Table 2 Characteristics of available models for Phaeodactylum tricornutum . Property Kroth et al . [ 69 ] DiatomCyc [ 45 ] Hunt et al . [ 71 ] Kim et al . [ 72 ] i LB1027_lipid (this study) Reactions 88 1719 metabolic reactions 67 transport reactions 318 849 (not including biomass equation) 4456 ( i LB1025: 2156) Metabolites Not available 1173 335 587 2172 ( i LB1025: 1704) Genes 151 1613 680 607 1027 ( i LB1025: 1025) Compartments Cytoplasm, Mitochondria, Chloroplast, Endoplasmic reticulum, Peroxisome Cytoplasm Cytoplasm, Mitochondria, Chloroplast, Peroxisome Cytoplasm, Mitochondria, Chloroplast (stroma and lumen), Peroxisome Cytoplasm, Mitochondria, Chloroplast (stroma and thylakoid), Peroxisome Reconstruction software Not available Pathway Tools CellNetAnalyzer MOST COBRA Toolbox, COBRApy Availability No mathematical model available Online access SBML SBML SBML, MAT Notes Carbohydrate metabolism Genome-wide model, not compartmentalized Simplified and compartmentalized version of DiatomCyc; see [ 70 ] for simulations Genome-wide model, GPRs not in Boolean format Genome-wide model, detailed lipid metabolism Metabolic model characteristics are compared between four available models for P . tricornutum and the one presented in this study. Here, we based our reconstruction effort on the updated and improved genome annotation which yields more precise localization predictions due to refined gene models. Compared to predictions of each bioinformatics tool, the sophisticated protein localization pipeline more often coincides with experimental findings ( Table 1 ). Since diatom metabolism and consequently biomass components strongly vary with growth conditions ( Fig 4 ), we determined P . tricornutum ’s biomass composition over a growth curve that resulted in nitrogen deprivation after the high accumulation of biomass. In order to assess i LB1027_lipid’s overall model coverage, we compared the ratio of genes accounted for in the reconstruction to genes predicted in the genome against the genome size for different eukaryotic organisms, namely Arabidopsis thaliana , Brassica napus , Chlamydomonas reinhardtii , Zea mays , Saccharomyces cerevisiae , Homo sapiens and Mus musculus ( Fig 5 ). The considered reconstructions span a large range in genome size. The i LB1027_lipid model includes a higher ratio of genes in reconstruction per genes in genome (10%) than the median of all models (6%). B . napus (bna572+) has a comparable ratio of genes in the reconstruction (996) to predicted genes in the genome (9873) but contains far fewer reactions (671). The only model with a higher ratio belongs to the well-studied model organism S . cerevisiae , though this model i TO977 also contains fewer total reactions. 10.1371/journal.pone.0155038.g005 Fig 5 Genes in reconstruction over predicted genes in genome against genome size for selected eukaryotic metabolic reconstructions. The three reconstructions with the highest ratio of genes in reconstruction per genes in genome are highlighted. bna572+ has a comparable ratio as i LB1025 and i LB1027_lipid, i TO977 has a higher ratio. Compared to i TO977 and bna572+, i LB1025 and i LB1027_lipid contain more reactions. The number of reactions in the respective reconstructions is used to scale the circle diameters. Note the discontinuous x-axis. Abbreviations: AraGEM: Arabidopsis thaliana [ 73 ]; bna572+: Brassica napus [ 74 ]; AlgaGEM: Chlamydomonas reinhardtii [ 75 ]; i RC1080: Chlamydomonas reinhardtii [ 39 ]; i RS1563: Zea mays [ 76 ]; i LB1025 and i LB1027_lipid: Phaeodactylum tricornutum , this study; i TO977: Saccharomyces cerevisiae Sc288 [ 77 ]; Recon2: Homo sapiens [ 78 ]; i MM1415: Mus musculus [ 79 ]. Carbon partitioning Recently, there has been a focus on using diatoms for biotechnological applications such as biofuel production, because of their high rate of neutral lipid accumulation [ 80 , 81 ]. Maximization of lipid biomass is a prerequisite for optimizing biofuel production in diatoms. Typical strategies for neutral lipid accumulation in P . tricornutum involve environmental stress, such as nitrogen or phosphorous limitation [ 37 ]. However, nutrient stress induced TAG accumulation also initiates growth arrest. TAGs store not only fixed carbon but also photosynthetically derived reducing equivalents. Storage of photosynthetically derived electrons into biomass also serves as photoprotection in diatoms [ 82 ]. Using the genome-scale model, we investigated the light-dependent partitioning of fixed carbon between storage carbohydrates and storage lipids, as shown in Fig 6A . Carbon fixation increased linearly with photon flux until saturation at the upper bound of CO 2 uptake (experimentally determined, see Materials and Methods ). Demand reactions added to the model allowed dynamic allocation of carbon and redox power into storage compounds and ensured mass balance with nutrient uptake. Resources could be fixed into biomass via nitrate reduction into ammonia, sulfate reduction into DMSP, carbohydrates or a representative TAG (see Materials and Methods ). Prior to saturation at a photon uptake of 16 mM, all of the fixed carbon was stored as carbohydrates (see Fig 6A ). Upon saturation, excess redox potential was stored as lipid and then as ammonia when all fixed carbon has been stored as TAG. No accumulation of DMSP was predicted. 10.1371/journal.pone.0155038.g006 Fig 6 Light-dependent carbon partitioning. (A) Simulations indicated as photon uptake exceeds carbon uptake, excess redox potential is stored in triacylglycerol. The saturation of carbon uptake is shown in black. (B) Percent of carbon fixed in TAG against percent of metabolite flow through NADHOR (vNADHOR; EC 1.6.5.3,1.6.99.3) over metabolite flow through PSI (vPSI; EC 1.97.1.2) at a super-saturating photon uptake of 22 mM. According to our simulations TAG accumulation is inversely proportional to energetic coupling. TAG accumulation is prohibited when at least 35% of photosynthetically fixed electrons are redirected to the mitochondria. Energetic coupling between mitochondria and plastid A recent, in depth characterization of photosynthetic electron flux in P . tricornutum enabled high quality constraints to be applied to the photosystem ( Table 3 ). Results in Bailleul et al . indicated cyclic electron flow (CEF_h) accounted for approximately 30% of total electron flow at low irradiances and as low as 5% at high irradiances [ 83 ]. Fixing the CEF reaction boundaries to 0.3 mM approximated these ratios. Water-water reactions (plastid terminal oxidase (PTOX, EC 1.10.3.11), and Mehler reaction) constituted approximately 10% of the total electron flow. To allow the electron flow into these reactions to scale with photon uptake in silico , 5% of electron flow through the cytochrome b6f complex (CBFC_u) was routed to elemental oxygen mimicking the electron drain to PTOX while 5% of the electron flow through photosystem I (PSI_u) was committed to a Mehler-like reaction. Combined, these accounted for the 10% of electron flow to water-water reactions. Independent PTOX and Mehler reactions in the model are blocked by default but the boundaries can be adjusted to fit experimental results that deviate from the 10% value. In accordance with Bailleul et al .’s findings, the model predicts the use of mitochondrial oxidative phosphorylation to balance ATP and NADPH ratios. 10.1371/journal.pone.0155038.t003 Table 3 Photosynthetic electron flow constraints as determined by Bailleul et al . [ 83 ]. Abbreviation Description Constraint CEF_h Cyclic electron flow around PSI LB = UB = 0.3 mM FNOR_h Ferredoxin:NADP + Oxidoreductase 5% electron flow to Mehler reaction CBFC_u Cytochrome b6f complex 5% of electron flow to PTOX PTOX_h Plastid terminal oxidase Default bounds set to 0 flux MEHLER_h Mehler reaction Default bounds set to 0 flux Constraint e-flow Energetic coupling of mitochondria and plastid νNADHOR—C ∙ νPSI ≥ 0 The model abbreviations refer to in silico reaction or metabolite identifiers. Abbreviations: LB, lower bound of reaction flux; UB, upper bound of reaction flux; νNADHOR, metabolite flow through the mitochondrial NADH:ubiquinone oxidoreductase; νPSI, metabolite flow through photosystem I, a proxy for total electron flow; C, a scalar value representing the percent of photosynthetically derived electrons coupled to mitochondrial respiration. The model did not initially predict the use of the alternative oxidase (AOX, EC 1.10.3.11) to vent excess reducing equivalents. Our results predicted that flow of reductant from the plastid to the mitochondria was dependent on the ATP needs of the cell; however the results of Bailleul et al . suggest that this ratio is fixed over a range of low to moderate light intensities. To simulate the observed energetic coupling between the mitochondria and plastid, an inequality constraint was added to the model. This constraint forced a minimum amount of the photosystem flux to be routed to the mitochondrial electron transport chain. Upon adding energetic coupling, the model predicted AOX was a primary electron sink at high irradiances. Additionally, the energetic coupling affected accumulation of neutral lipid biomass. Storage of lipid biomass was inversely proportional to energetic coupling with TAG accumulation being abolished when at least 35% of photosynthetically fixed electrons were redirected to the mitochondria at super-saturating photon uptake ( Fig 6B ). Since lipid biosynthesis is dependent on plastid localized reducing power, it is possible that energetic coupling of the mitochondria and plastid is an inherent limit on the accumulation of neutral lipids, as predicted by the model. These results indicate that disrupting the energetic coupling of the plastid to the mitochondria while upregulating plastid lipid biogenesis and taking advantage of increased NADPH pools in AOX knockdown lines may result in increased TAG accumulation during exponential phase while alleviating the observed growth defect [ 83 ]. This would allow for the decoupling of growth process (e.g. nutrient limitation) from TAG production and increase overall yields of biofuel precursors. The mechanism by which reducing equivalents are shuttled to the mitochondria during energetic coupling is still unknown. In addition to the malate shuttle as proposed by Bailleul et al ., our reconstruction uncovered a previously undescribed plastid ornithine biosynthetic pathway ( Fig 7 ) that may represent an important metabolic connection between plastid and mitochondria. The compartmentalization pipeline indicated plastid targeting of acetylglutamate kinase (AGK_h, EC 2.7.2.8), N-acetyl-γ-glutamyl-phosphate reductase (AGPR_h, EC 1.2.1.38), acetylornithine transaminase (ACOAT_h, EC 2.6.1.11), and ornithine acetyltransferase (GACT_h, EC 2.3.1.35). Biomass yield simulations suggested that in silico the ornithine-glutamine shuttle is used to transfer reducing equivalents generated by photosynthesis to the mitochondria. Four photosynthetically derived electrons are used; two by the oxidation of ferredoxin molecules by plastid glutamate synthase (GLTS_h, EC 1.4.7.1) and two via oxidation of NADPH by AGPR_h. Ornithine is then proposed to be shuttled from the plastid to the mitochondria. The activity of 1-pyrroline-5-carboxylate dehydrogenase (P5CDH_m, EC 1.2.1.88) and glutamine dehydrogenase (GLUDH2_m, EC 1.4.1.2) produce NADH further suggesting that this novel ornithine-glutamate pathway coupling these two organelles is possible. 10.1371/journal.pone.0155038.g007 Fig 7 Chloroplastic ornithine cycle as revealed by the model. Metabolic network usage of a chloroplastic ornithine cycle is shown under a saturating photon constraint of 16 mM allowing maximum carbon uptake. Minor reactants and products are omitted for visual clarity (i.e., water, protons and phosphate). Metabolite and reaction abbreviation suffixes indicate cellular compartment; c, cytosol; h, chloroplast; m, mitochondria. Reversible reactions are indicated by arrowheads at both ends. The filled arrowhead indicates the direction in which the reaction is running, i.e. from substrate (open arrowhead) to product (filled arrowhead). Abbreviations used: ACOAT, acetylornithine transaminase; AGK, acetylglutamate kinase; AGPR, N-acetyl-δ-glutamyl-phosphate reductase; GACT, glutamate N-acetyltransferase; GLNA, glutamine synthase; GLTS, glutamate synthase (ferredoxin dependent); GLUDH2, glutamine dehydrogenase (NAD dependent); GLUSA, glutamate semialdehyde degradation (spontaneous); OAT, ornithine aminotransferase; P5CDH, 1-pyrroline-5-carboxylate dehydrogenase; acorn, N-acetylornithine; acglu, N-acetyl-L-glutamate; acg5p, N-acetyl-L-glutamate 5-phosphate; acg5sa, N-Acetyl-L-glutamate 5-semialdehyde; adp, ADP; akg, α-ketoglutarate; atp, ATP; fdxox, ferredoxin (oxidized); fdxrd, ferredoxin (reduced); gln__L, L-glutamine; glu__L, L-glutamate; glu5sa, L-glutamate 5-semialdehyde; nad, NAD + ; nadh, NADH; nadp, NADP + ; nadph, NADPH; nh4, ammonium ion; orn, ornithine; 1pyr5c, (S)-1-Pyrroline-5-carboxylate. Storage of metabolites such as glutamine and ornithine could serve a photoprotective role by sequestering reducing equivalents as well as assimilated nitrogen. Indeed when intermediates of this ornithine shuttle were allowed to accumulate during simulations, the model predicted they were preferred over TAG biosynthesis. Ornithine concentrations were previously investigated in the context of the diatom ornithine-urea cycle (OUC) [ 5 ]. Although one of the most abundant metabolites in the cell, ornithine levels were not correlated with OUC intermediates, which indicated a possible alternative function [ 5 ]. We hypothesize storage of reducing power and electron transport into the mitochondria, potentially coupled to OUC consumption, is this alternative function."
} | 10,042 |
26136581 | null | s2 | 3,488 | {
"abstract": "All plants are inhabited internally by diverse microbial communities comprising bacterial, archaeal, fungal, and protistic taxa. These microorganisms showing endophytic lifestyles play crucial roles in plant development, growth, fitness, and diversification. The increasing awareness of and information on endophytes provide insight into the complexity of the plant microbiome. The nature of plant-endophyte interactions ranges from mutualism to pathogenicity. This depends on a set of abiotic and biotic factors, including the genotypes of plants and microbes, environmental conditions, and the dynamic network of interactions within the plant biome. In this review, we address the concept of endophytism, considering the latest insights into evolution, plant ecosystem functioning, and multipartite interactions."
} | 203 |
30909637 | PMC6470648 | pmc | 3,489 | {
"abstract": "Improvement of energy harvesting performance from flexible thin film-based energy harvesters is essential to accomplish future self-powered electronics and sensor systems. In particular, the integration of harvesting signals should be established as a single device configuration without complicated device connections or expensive methodologies. In this research, we study the dual-film structures of the flexible PZT film energy harvester experimentally and theoretically to propose an effective principle for integrating energy harvesting signals. Laser lift-off (LLO) processes are used for fabrication because this is known as the most efficient technology for flexible high-performance energy harvesters. We develop two different device structures using the multistep LLO: a stacked structure and a double-faced (bimorph) structure. Although both structures are well demonstrated without serious material degradation, the stacked structure is not efficient for energy harvesting due to the ineffectively applied strain to the piezoelectric film in bending. This phenomenon stems from differences in position of mechanical neutral planes, which is investigated by finite element analysis and calculation. Finally, effectively integrated performance is achieved by a bimorph dual-film-structured flexible energy harvester. Our study will foster the development of various structures in flexible energy harvesters towards self-powered sensor applications with high efficiency.",
"conclusion": "4. Conclusions In summary, we have demonstrated two different dual-film-structured flexible thin film energy harvesters for effectively integrated energy harvesting performance: a stacked structure and a bimorph structure. The LLO process-based fabrication has been selected for the basic energy harvesting device because it was confirmed to realize flexible high-performance harvesting unit devices. However, there have still been only a few ways to enhance the performance and/or integrate the energy level of the flexible film energy harvesters, which is highly significant for the future progress of commercialization, sensor applications, and self-powered electronics based on piezoelectric device principles. Multiple device connection is not useful due to the complicated features. Material modification approach was too expensive to achieve the improvement of the flexible energy harvesting unit. Therefore, we have suggested integrated structures of PZT film devices in a single area, like the approaches of MLCC. Both stacked and bimorph structures of dual PZT thin films fabricated by multistep LLO processes are well established on the flexible PET sheet substrate. However, the stacked dual-film-structured flexible energy harvester cannot generate effective energy harvesting signals, while the bimorph structure shows the general performance of PZT thin film energy harvesters. Based on the theoretical simulations and calculations, it is well discussed that the phenomena is due to the ineffective position of mechanical neutral plane in the stacked dual-films structure. Therefore, we propose the bimorph structure to effectively integrate and improve the performance of flexible PZT thin-film energy harvester as a single device configuration. As we expected, the energy harvesting performance of the double-faced device are the voltage up to ~280 V and the current up to ~2.2 μA, which indicated the good performance enhancement and integration in the bending deformation. Because the piezoelectric energy device and technology can be directly associated with various sensor systems [ 41 , 42 , 43 , 44 ], our experimental and theoretical demonstration for performance improvement of flexible thin film energy harvesters using the intuitive and simple approach will promote more rapidly the practical developments of flexible piezoelectric energy harvesting devices for future self-powered and self-sufficient sensor systems.",
"introduction": "1. Introduction In recent years, energy harvesting technologies have drawn attention from many researchers hoping to establish self-powered sensors and Internet of Things (IoT) systems for future applications [ 1 , 2 ]. Among the various energy sources in our surroundings, mechanical energy sources are highly promising for individual energy harvesting devices because mechanical energy is pervasive (e.g., machinery vibration, body activity, biomechanical movement, natural stimulation, etc.) but often wasted unwittingly [ 3 , 4 ]. In terms of this aspect of mechanical energy sources, the energy harvesting technology largely means the field of mechanical energy harvesting [ 1 , 2 ]. Moreover, mechanical energy harvesting devices can be also used as mechanical sensors without external power sources [ 5 , 6 ]. Therefore, energy harvesting materials and devices are important concepts for a new era of sensor applications. Although some principles have been developed to convert mechanical energy to electrical energy, piezoelectric materials and devices have still been considered as prospective energy harvesting technology owing to the simple device structures and environmental robustness regardless of wear, humidity, and serious heaviness [ 4 , 7 ]. In addition, piezoelectric energy harvesters can be fabricated as flexible electronics for next-generation electronic systems using piezoelectric ceramics as well as polymers on flexible plastic substrates [ 8 , 9 , 10 ]. For example, piezoelectric energy harvesters can be applied to tiny vibration-based energy harvesting and sensor applications polyvinylidene fluoride or ceramics [ 11 , 12 , 13 , 14 ]. In fact, triboelectric energy harvesting, which is based on the coupling between tribelectrification and electrostatic induction, is also a promising technology for self-powered sensor devices. Triboelectric energy harvesters can generate very high voltage signals [ 15 , 16 , 17 ]. In contrast, they are weak during high humidity and very long duration owing to the mechanism of contact electrification [ 18 , 19 ]. Moreover, piezoelectric devices can be more easily fabricated in a flexible configuration [ 19 ]. Therefore, triboelectric and piezoelectric energy harvesters should be utilized according to proper and complementary applications. Various technologies have been reported to fabricate piezoelectric ceramics-based flexible energy harvesters, such as the stamping transfer method, polymer-hybrid composites, single-crystalline ceramic lamination, and so on [ 20 , 21 , 22 , 23 , 24 , 25 ]. However, these approaches suffer from instable fabrication, non-reliable performance, insufficient output, and/or very high cost. In contrast, the recently developed technology for flexible ceramic film energy harvesters using laser lift-off (LLO) processes is a highly plausible approach due to low-cost materials, excellent scalability, and high output density [ 26 , 27 , 28 , 29 ]. Furthermore, the LLO process is already commercialized in the field of optoelectronics and display devices, which guarantees the processing reliability and the compatibility with conventional fabrications of electronics [ 27 , 30 , 31 ]. Nonetheless, the LLO fabrication-based energy harvester should be developed for enhancing the current or charge density performance because the current level of the previously reported device is much weaker than that of single-crystalline laminated flexible energy harvesters [ 25 , 32 ]. Although the energy harvesting signals (indicating the generated output voltage and current) can increase by connecting multiple devices to each other, it is a very complicated configuration in terms of practical energy or sensor applications. Additionally, for sensors based on energy harvesters and self-powered applications, the generated power output is very important. In the field of piezoceramic thin/thick film-based flexible energy harvesters in bending mode, state-of-the-art power output in instantaneous bending motions has been reported from 160 μW to 200 μW [ 29 , 33 ]. Even though the performance enhancement of the LLO-based flexible thin film energy harvesters has been studied using the quantitative simulations and the textured polycrystalline piezoelectric films resulting from modifying basal mother wafers [ 33 ], the crystallographic approach requires a special, expensive wafer (e.g., MgO, LaAlO 3 , etc.), while the processing yield becomes low. Therefore, the original LLO process for the flexible thin film energy harvesters should be revisited to improve the device performance and stability. Similar to the technology of multilayer ceramic capacitor (MLCC), flexible thin film energy harvesters can also be enhanced by using layered piezoelectric ceramic films on a single area of plastic substrate. However, there has not yet been a systematic investigation into layered flexible piezoelectric thin film energy harvesters for effectively enhanced and integrated energy harvesting performance. In this study, we demonstrate two different types of flexible thin film energy harvesters as prototypes of multilayered (dual-film-structured) piezoelectric generators fabricated by the reported stable LLO process. One type is a stacked dual-film structure, while the other is a double-faced (bimorph) dual-film structure. Because both devices are fabricated by the guaranteed LLO process and piezoelectric ceramics, the intrinsic material properties are very reasonable and reliable. Nevertheless, the performance of the two structures is highly different due to the differences in mechanical strain and generated piezopotential of piezoceramic films. The stacked dual-film energy harvester cannot generate effective energy harvesting performance, while the bimorph energy harvester presents good energy harvesting performance and reasonably integrated energy harvesting output for performance enhancement. The important discrepancy originates from the location of mechanical neutral plane for applied effective strain in piezoelectric active layers. Thus, it is noted that the device performance cannot simply be improved by increasing the number of piezoelectric active layers in a single flexible film energy harvester. Our investigation provides an appropriate design and perspective for multilayered piezoelectric film-based flexible energy harvesters towards performance-enhanced self-powered mechanical sensor systems.",
"discussion": "3. Results and Discussion 3.1. Analyses of PZT Films As-Deposited and Transferred by Multiple Laser Lift-Off Processes Figure 2 a,b shows photographs of stacked dual-structured PZT films and bimorph PZT films, which are multiple-transferred onto the single flexible PET sheet substrate. Both structures were well established by the optimum processes. The first LLO of PZT film has already been well defined by the previously reported strict procedure [ 26 , 27 , 33 ]. However, the second LLO step of next PZT films onto the as-fabricated device can cause material failure due to undesirable effects such as abnormal thermal gradients, mechanical deteriorations, interfacial roughness, and so forth. For the stacked device, failure (e.g., crack) easily occurs when the second PZT film is detached from the sapphire wafer after LLO because it should be aided by mechanical force applied by tweezers, which can induce the fracture of the underlying first PZT film. This problem can be solved by a method in which the second PZT film is transferred onto the first layer device so as not to overlap completely, as shown in Figure 2 a. The bimorph device is also generally hard to fabricate using LLO, because the backsideof the PET sheet is not flat due to the already-established first device part, which may hinder the laser focusing for backside LLO. Adopting a semi-cured polydimethylsiloxane (PDMS) coated glass substrate as a handle wafer during the second LLO can alleviate this problem because the first device part could be conformally buried into the PDMS buffer on the glass substrate. Thus, the backside of the sheet can be relatively flat for the laser focusing. Throughout our optimized multistep LLO processes for dual-structured PZT films, there are no serious visual problems, as presented by the optical microscopy images (insets of Figure 2 a,b). Note that the square patterns on the transferred PZT result from the shape of the pulsed laser beam [ 26 , 27 , 33 ]. To investigate the material characteristics of PZT films in detail, we compared the X-ray diffraction (XRD) patterns and Raman spectra of the as-deposited PZT film on the sapphire wafer, the first transferred PZT film, and the second transferred PZT films in the two different dual-film structures. Figure 2 c shows the XRD patterns of the PZT thin film on the sapphire wafer deposited by the sol-gel processing and the first transferred PZT film on the flexible PET substrate by the initial LLO. As expected, the crystallized PZT film of MPB composition was defined by the sol-gel deposition, and well maintained on the PET sheet after the LLO process [ 27 ]. In addition, the secondly transferred PZT films by subsequent LLO processes also present same perovskite MPB crystal structures, as shown in Figure 2 d. In both stacked and bimorph structures, the second PZT films are stably transferred without degradation. It means that there is no crystallographic and compositional modulation in the transferred PZT films after multistep LLO processes because the duration time of pulsed excimer laser is very short (~30 ns), localizing the laser heating only within the interface [ 27 ]. Raman spectroscopy also shows the well-retained MPB phase structure of PZT films during the multistep LLO processes, as presented by Figure 2 e,f. Raman spectroscopy was performed to analyze the phase of the PZT films using a 514.5 nm Ar + laser source as an excitation source at ambient temperature. The bands of the Raman shifts around 205, 275, 330, and 591 cm −1 correspond to the typical property of perovskite MPB phase of PZT [ 26 , 33 ]. The above bands correspond to E(2TO), B1+E, A1(2TO) and A1(3TO) modes, respectively [ 37 ]. Note that there may be also more complicated Raman mode bands of the perovskite phase, but they do not appear clearly in the spectrum of thin film materials due to weak signals and substrate effects [ 26 ]. The Raman characteristics of the secondly transferred PZT films are also the same as those of the as-deposited and the first transferred PZT films. 3.2. Energy Harvesting Performance of Two Different Dual-Film-Structured Flexible Energy Harvesters A stacked structure is the first way to enhance and integrate the energy harvesting performance of PZT thin-film energy harvesters in a single device because various integrated devices exist in vertically stacked structures like the MLCC. Therefore, our stacked dual-film-structured flexible film energy harvester was investigated first. One might expect that each PZT layer can generate high-performance energy harvesting signals and the performance can be merged for higher voltage and current signals. As shown in Figure 3 a,b, however, the first PZT layer produces excessively low energy harvesting signals, i.e., ~9 V of voltage and ~4 nA of current, in the bending stimulation. This energy harvesting performance is useless compared to the previous general flexible PZT thin film energy harvester made by LLO processes, which generated ~140 V of voltage and ~1 μA. Although the signals from the second PZT layer (~25 V and ~50 nA) are better than those of the first layer ( Figure 3 c,d), they are still much poorer than from the general flexible PZT film generator. If there is no consideration of the mechanics of flexible devices, this phenomenon looks weird because the basic structure, fabrication, and measurement conditions (bending radius of ~1.6 cm) are identical to our previous reports for flexible single-layered PZT film energy harvesters. In contrast, the bimorph dual-film-structured energy harvester generates the expected and high-performance signals in the same bending condition [ 26 , 27 ]. As presented by Figure 4 , both the first (top-sided) and the second (bottom-sided) PZT layers converted the convex and concave bending strain into the electrical peaks, respectively. The voltage level is up to 140 V and the current peak is up to 1.3 μA. Note that the peak amplitude of the bottom-side PZT layer is slightly asymmetric presumably because it is governed by compressive stress and strain, different from the general measurements of other flexible energy harvesters as well as the top-side PZT layer under tensile stress and strain. 3.3. Theoretical Simulations and Final Selection for Integrated Performance To investigate the difference between the performance of the stacked dual-film-structured flexible energy harvester and the bimorph dual-film-structured flexible device, we performed the finite element analysis (FEA) simulations using COMSOL Multiphysics computation program ( Figure 5 and Figure 6 ). All material parameters were based on the confirmed values and previously reported simulations. As shown in Figure 5 , interestingly, the simulated piezoelectric potential of the first PZT layer of the stacked structure is relatively very low (~20 V), while that of the second layer is relatively decent (~70 V). This serious discrepancy should be analyzed by the mechanical neutrality in the flexible device configuration. The position of the mechanical neutral plane in a layered device structure can be described as below Equation (1): (1) h neutral = ∑ i = 1 N { Y i ¯ t i ( ∑ j = 1 i t j − t i 2 ) } ∑ i − 1 N Y i ¯ t i , \nwhere N is the total number of layers, t i is the thickness of the i th layer (from the top), and is Y i ¯ = Y i / ( 1 − υ i 2 ) is defined as the effective Young’s modulus ( Y i ≡ absolute Young’s modulus and υ i ≡ Poisson’s ratio of the i th layer) [ 38 ]. For the stacked dual-film-structured flexible PZT thin film energy harvester, the layer structure is the sequence of SU-8 encapsulation, second PZT thin film, SU-8 encapsulation, first PZT thin film, and PET sheet substrate (from the top). The mechanical parameters and thicknesses are (1) Y SU-8 = 4.02 GPa, υ SU-8 = 0.22, and t SU-8 = 5 μm; (2) Y PZT = 62.5 GPa, υ PZT = 0.35, and t PZT = 2 μm; and (3) Y PET = 3 GPa, υ PET = 0.4, and t PET = 125 μm. The mechanical neutral plane is ~48 μm below the top surface. Therefore, the distance between the mechanical neutral plane and the middle of first PZT film is about 35 μm. The applied bending strain in the PZT layer is calculated by Equation (2): (2) ε = δ r , \nwhere r is the bending radius and δ is the distance from the mechanical neutral plane [ 38 ]. The calculated strain in the first PZT thin film at bending radius of 1.6 cm is about 0.217%. On the other hand, the distance between the mechanical neutral plane and the middle of second PZT layer is about 42 μm, indicating the strain applied to the second layer is about 0.260% in the same mechanical stimulation. The much lower energy harvesting performance in both PZT layers of the stacked dual-film-structured device is because these much smaller applied strain in bending stimulations than that of the general one-layer PZT thin film energy harvester. Thus, this phenomenon results from the mechanics of layered structures in flexible device configuration. In contrast, the generated piezopotential of each PZT layer of the bimorph structure is generally high (~160 V), as shown in Figure 6 . This is in accordance with the trend of the experimental measurement. Note that the polarity of piezopotential between the first and second PZT layers is opposite because they are under tensile and compressive stress, respectively. It can be merged and integrated by the proper poling process and wire connection. The computational result can also be analyzed by Equations (1) and (2) for the mechanical neutral plane and the actually applied strain. According to Equation (1), the location of the mechanical neutral plane in a layered device structure is ~69.5 μm below the top surface. Hence, the distance between the mechanical neutral plane and the middle of each PZT film is about 63.5 μm because the bimorph dual-film-structured device has the symmetric cross-sectional configuration. This means that the applied bending strain to the PZT layers of bimorph device is about 0.394% in the same bending stimulation, slightly higher than that of the general one-layer PZT thin film energy harvester [ 26 , 33 ]. As declared previously, the performance enhancement of energy harvesting by signal integration in a single device is very important without any complicated multiple device arrangements or complex material modifications, in order to facilitate more rapid technological growth for commercialization of the advantages of LLO technology-based flexible thin film energy harvesters with high performance. Therefore, we integrated efficiently the generated signals from both PZT layers of the double-faced dual-film-structured flexible energy harvester for effectively enhanced performance. As presented by Figure 7 , the energy harvesting signals are well correspondingly merged and integrated, showing the voltage up to ~280 V in series and the current up to ~2.2 μA in parallel. Consequently, it is reasonable that the bimorph structure is adoptable and rational for the effectively integrated performance of a flexible PZT thin film energy harvester in a single device. It is easily guaranteed by the performance of single-faced (uni-morph) structured device with the same condition ( Figure S2 ). The uni-morph energy harvester generates ~120 V and 0.8 μA, which is smaller even than the single part of bimorph device. This is presumably due to the applied strain in the uni-morph device is smaller [ 26 , 33 ]. All generated energy harvesting signals are also briefly summarized in Table S1 to show our results at a glance. Note that the integrated output signals are not the exact sum of each single-faced device, which can be considered a minor non-linear effect. This phenomenon also commonly occurs in the widely used linear superposition test for energy harvesters. It is presumably due to the parasitic capacitance in circuits of multiple-connected electrodes [ 39 , 40 ]. Moreover, this is more visible in the current output because the current signals are more irregular than voltage output generally in the piezoelectric energy harvesters with bending motions. Using the bimorph integrated flexible energy harvester, we also measured the instantaneous power output according to the external circuit load resistance, as shown in Figure S3 . The maximum power, ~180 μW, was presented at ~100 MΩ, which is a similar matching resistance in IDE-based flexible energy harvesters [ 26 ]. According to the area of electrode pairs, the maximum instantaneous power density is about 250 μW·cm −2 ."
} | 5,728 |
30894842 | PMC6415583 | pmc | 3,490 | {
"abstract": "Shear stress is an important factor that affects the formation and structure of anode biofilms, which are strongly related to the extracellular electron transfer phenomena and bioelectric performance of bioanodes. Here, we show that using nitrogen sparging to induce shear stress during anode biofilm formation increases the linear sweep voltammetry peak current density of the mature anode biofilm from 2.37 ± 0.15 to 4.05 ± 0.25 A/m 2 . Electrochemical impedance spectroscopy results revealed that the shear-stress-enriched anode biofilm had a low charge transfer resistance of 46.34 Ω compared to that of the unperturbed enriched anode biofilm (72.2 Ω). Confocal laser scanning microscopy observations showed that the shear-stress-enriched biofilms were entirely viable, whereas the unperturbed enriched anode biofilm consisted of a live outer layer covering a dead inner-core layer. Based on biomass and community analyses, the shear-stress-enriched biofilm had four times the biofilm density (136.0 vs. 27.50 μg DNA/cm 3 ) and twice the relative abundance of Geobacteraceae (over 80 vs. 40%) in comparison with those of the unperturbed enriched anode biofilm. These results show that applying high shear stress during anode biofilm enrichment can result in an entirely viable and dense biofilm with a high relative abundance of exoelectrogens and, consequently, better performance.",
"conclusion": "Conclusion Anodes were started and operated under nitrogen sparging rates from 0 to 80 mL/min. Increasing the nitrogen sparging rate from 0 to 10, 40, and 80 mL/min improved the anode performance by 47, 69, and 69%, respectively (2.37 ± 0.15 vs. 3.48 ± 0.12, 4.0 ± 0.25 and 4.05 ± 0.25 A/m 2 ). The improved anode performance was attributed to the viability, physical and microbial community structures of the anode biofilms. For the viability structure, the unperturbed enriched anode biofilm showed a two-layer structure with a live outer layer on top of a dead inner-core layer. The perturbed enriched biofilms exhibited only a single viable layer. Regarding the physical structure, the biomass and biofilm density increased with the increasing nitrogen sparging rate. For the microbial community structure, compared to that of the unperturbed enriched anode biofilm, the low nitrogen sparging rate (10 mL/min)-enriched anode biofilm had a higher abundance of exoelectrogens, such as Rhodocyclaceae , Rhodobacteraceae , Desulfovibrionaceae , and Comamonadaceae , and the high nitrogen sparging rate (40 and 80 mL/min)-enriched anode biofilms had a higher abundance of Geobacteraceae (over 80 vs. 40%).",
"introduction": "Introduction Anodic microorganisms in microbial electrochemical systems (MESs) are biocatalysts that oxidize organic matter to transfer electrons to an electrode ( Xiao et al., 2015 ; Hodgson et al., 2016 ). Normally, anodic microorganisms perform this process in the form of a biofilm. The anode biofilm is a complex aggregation of microbial communities and substances developed from planktonic microorganisms attached to the anode surface. It has been found that several factors can affect the formation, structure, and performance of anode biofilms, including the substrate concentration ( Hari et al., 2017 ), electron acceptor ( Ucar et al., 2017 ), anode solution ( Liu et al., 2017 ), electrode potential ( Bosire and Rosenbaum, 2017 ), and electric field intensity ( Du et al., 2018 ). The shear stress that arises from solution disturbance is an important factor affecting the formation, structure and performance of anode biofilms because of the physical force exerted on the anode biofilm and enhanced substance diffusion (substrate and metabolic end products) in the anode biofilm. Physical force can affect the attachment and detachment of microorganisms and the biofilm. For example, applying a high potential to the anode ( Bosire and Rosenbaum, 2017 ), modifying the anode with positively charged compounds ( Cheng and Logan, 2007 ) or improving the surface hydrophilicity ( Du et al., 2017 ) can accelerate the attachment of bacteria due to the increased electrostatic attraction between anodic microorganisms and the electrode surface. Additionally, enhanced substance diffusion [either the substrate or metabolic end products (H + )] can increase the biomass, viability and performance of the anode biofilm. It has been shown that using a highly concentrated phosphate-buffered saline (PBS) solution as the anode electrolyte or increasing the pH of the anode electrolyte from a weak acid (pH = 6–7) to alkalinity (pH = 7–9) to mitigate proton accumulation in the anode biofilm or adjusting the gravity settling of planktonic bacteria and bioanode can increase the biomass, viability and current generation of anode biofilms ( Liu et al., 2005 ; Cheng and Logan, 2007 ; Patil et al., 2011 ; Dhar et al., 2017 ; Li et al., 2017 ). In previous studies, we found that aerobically enriched anode biofilms with sufficient substance diffusion in the inner layer had a thicker inner layer and a higher current generation. Nonetheless, few studies have examined the effect of shear stress on the formation, structure, and performance of anode biofilms. The impacts of shear stress on non-electrochemically active biofilms have been investigated. In general, a high shear stress usually results in thin, dense, and strong biofilms with low microbial diversity ( Liu and Tay, 2002 ; Rickard et al., 2004 ; Celmer et al., 2008 ; Rochex et al., 2008 ). However, compared to these biofilms, the anode biofilm has substantially different characteristics, such as using an insoluble electron acceptor (mostly soluble molecules such as dissolved oxygen, nitrate and fermentation products are used for other biofilms) and the ability of long-distance extracellular electron transfer (EET) ( Lovley, 2012 ). Therefore, these results are not directly translatable to the impacts of shear stress on anode biofilms. In this study, we investigated the impacts of shear stress caused by nitrogen sparging (0–80 mL/min) on anode biofilm structure and performance. Solution disturbance caused by nitrogen sparging has been validated as an effective method to regulate shear stress ( Celmer et al., 2008 ; Shen et al., 2013 ). For a fixed nitrogen sparging rate, we evaluated the anode performance in terms of startup time and linear sweep voltammetry (LSV). We then revealed the anode structure and characteristics with electrochemical impedance spectroscopy (EIS), confocal laser scanning microscopy (CLSM) and high-throughput 16S rRNA gene sequencing.",
"discussion": "Discussion All MFCs were successfully started and operated under nitrogen sparging rates from 0 to 80 mL/min, indicating that exoelectrogens can form stable biofilms under shear stress. Interestingly, with an increase in the operation cycles, the stable potential of A0 gradually increased, but the stable voltages of A10, A40, and A80 remained constant ( Figure 1 ). Even when the nitrogen sparging was stopped, A10, A40, and A80 had higher LSV peak current densities ( Figure 2 ) than A0, indicating that shear stress affected the anode biofilm structure and therefore influenced the anode performance. The EIS results ( Figure 3 ) showed that the increased anode performance of the MFCs with shear-stress-enriched anode biofilms was mainly due to their lower charge transfer resistances, indicating that the change in R ct caused by the change in the anode biofilm structure is the key factor affecting anode performance. The observed anode biofilm structures that could affect R ct were considered as three different factors: the viability structure, the physical structure and the microbial community structure. With regard to the viability structure, the anode biofilms formed with nitrogen sparging (A10-80) mostly consisted of live cells in one layer ( Figure 4B–D ), though the A0 anode biofilm showed a two-layer structure with a live outer layer on top of a dead inner-core layer ( Figure 4A ). A two-layer structure generally leads to low anode performance. On the one hand, the live cells of the two-layer structure were less than that of the viable single layer with the same biomass. On the other hand, although the dead inner layer in the two-layer structure will not inhibit electron transfer from the live outer layer to the electrode, the electrochemical activity of the outer layer cells will be impaired by the dead inner layer, resulting in an increase in the charge transfer resistance ( Sun et al., 2015 , 2017 ; Dhar et al., 2017 ). The gradual decrease in the A0 stable potential was very possibly caused by the accumulation of dead cells in the inner layer of the anode biofilm. A similar decrease in current generation caused by the accumulation of dead cells in the inner layer of an anode biofilm was reported by Sun et al. (2015) . Regarding the physical structure, CLSM observations and biomass analyses showed that the biomass and biofilm density increased with the increasing nitrogen sparging rate. The EET of exoelectrogens may occur through conduction- or diffusion-based (mediated) mechanisms or a combination of mechanisms ( Lovley, 2012 ). Regardless of the specific form of EET, increasing biomass and biofilm density will lead to reduced electrical resistance either by shrinking the spacing and increasing the electron shuttle density or by increasing the density and contact of components that exhibit metal-like conduction ( Strycharz-Glaven et al., 2011 ). However, for nitrogen sparging rates over 40 mL/min, the biofilm density increased with the increasing nitrogen sparging rate, whereas the anode performance did not. This result might be because the charge transfer resistance R ct includes the electron transfer resistance between microorganisms and microorganisms and the anode ( Malvankar et al., 2012 ). Increasing the biofilm density will decrease the electron transfer resistance between microorganisms but minimally contribute to the electron transfer resistance between microorganisms and the anode. Pham et al. (2008) also reported that increasing the shear force from 10 to 120 s -1 resulted in two times more biomass and, therefore, the anode performance increased. For the microbial community structure, Geobacteraceae dominated the anode biofilms on all anodes, especially the A40 and A80 anode biofilms ( Figure 5 ). Geobacteraceae are known for their excellent electricity generation and long-range EET ( Bond and Lovley, 2003 ) and are the dominant species in the anode biofilms of bioelectrochemical systems fed with acetate ( Yates et al., 2012 ). The dominant presence of Geobacteraceae (over 80%) in the A40 and A80 compared to A0 (40%) contributed to the increased anode performance of A40 and A80. This increase in the Geobacteraceae percentage was believed to be altered by shear stress rather than oxygen. Introducing oxygen into the anode chamber during the operation of all the MFCs was strictly avoided. In the microbial community of A0, mostly of the dominating species, such as Geobacteraceae (40%) ( Bond and Lovley, 2003 ), Rikenellaceae (14%) ( Su et al., 2014 ), Porphyromonadaceae (8.8%) ( Rosenberg et al., 2014 ), Acholeplasmataceae (8.5%) ( Rosenberg et al., 2014 ) and Spirochaetaceae (2.7%) ( Freundt et al., 1984 ), all were considered anaerobic. Notably, A0 had a higher abundance of Geobacteraceae than A10 (40 vs. 22%). However, compared to those in A0, many of the dominant families in the A10 anode biofilm have previously been reported to be exoelectrogens, such as Rhodocyclaceae , Rhodobacteraceae , Desulfovibrionaceae , and Comamonadaceae . Rhodocyclaceae species were found to be dominant in acetate-fed MFCs ( Borole et al., 2009 ; Jangir et al., 2016 ), and some strains were confirmed to be exoelectrogens ( Jangir et al., 2016 ). Rhodobacteraceae consists of chemoheterotrophs and photoheterotrophs, which are typical exoelectrogens ( Kiely et al., 2011 ; Wong et al., 2016 ). Desulfovibrionaceae and Comamonadaceae have been widely found in MFCs, and their electricity generation ability has been confirmed ( Xing et al., 2010 ; Eaktasang et al., 2016 ). In contrast, of the dominant families in A0, members of Rikenellaceae are anaerobic fermentation bacteria that tend to use complicated substrates such as peptone, yeast extract, maltose and glucose, but cannot exploit some simple organic matter such as formic acid, acetate, and ethyl alcohol ( Su et al., 2014 ). Thus, as the provided substrate was acetate, the high relative abundance of Rikenellaceae might be related to the accumulation of dead cells. The next dominant family, Porphyromonadacea , was found in abundance in the anode biofilm of dual-chamber MFCs ( Sotres et al., 2015 ), but the electricity generation ability of this family has not been confirmed. Acholeplasmataceae has not been reported in MFCs. In summary, the A10 biofilm possibly had a higher abundance of exoelectrogens than A0, resulting in the higher anode performance. In this study, we showed that increasing shear stress could increase anode performance. Increasing the shear stress requires additional energy consumption, and optimization of the flow rate and power generation is needed. Moreover, we revealed, for the first time, that shear stress helps maintain a viable anode biofilm. The viability of the anode biofilm was found to be altered by the stable current density of the anode biofilm or the PBS concentration of the anode solution ( Sun et al., 2015 , 2017 ; Dhar et al., 2017 ). When the stable current density was lower than 2.3 A/m 2 or the PBS concentration was lower than 10 mM, the anode biofilm formed a two-layer structure with a live outer layer covering a dead inner layer. In contrast, when the stable current density of the anode biofilm was higher than 4.8 A/m 2 or the PBS concentration was higher than 100 mM, the anode biofilm was mostly a viable single layer. Thus, to maintain a high-performance and viable anode biofilm, an anode biofilm should be operated with a high current density or high PBS concentration. However, in application, especially for wastewater treatment, the highly variable organic matter contents and low concentrations of organic matter and ionic strengths usually result in little current production, impacting the long-term viability of the anode biofilm. The one-layer viable anode biofilm structure under nitrogen sparging provides an instructive method for maintaining a high-viability active anode biofilm at low current density (1.1 ± 0.1 A/m 2 in this work) and low PBS concentration (50 mM PBS in this work), facilitating high performance and sufficient COD removal in MESs for applications."
} | 3,676 |
34948474 | PMC8708155 | pmc | 3,491 | {
"abstract": "Belowground interactions of plants with other organisms in the rhizosphere rely on extensive small-molecule communication. Chemical signals released from host plant roots ensure the development of beneficial arbuscular mycorrhizal (AM) fungi which in turn modulate host plant growth and stress tolerance. However, parasitic plants have adopted the capacity to sense the same signaling molecules and to trigger their own seed germination in the immediate vicinity of host roots. The contribution of AM fungi and parasitic plants to the regulation of phytohormone levels in host plant roots and root exudates remains largely obscure. Here, we studied the hormonome in the model system comprising tobacco as a host plant, Phelipanche spp. as a holoparasitic plant, and the AM fungus Rhizophagus irregularis . Co-cultivation of tobacco with broomrape and AM fungi alone or in combination led to characteristic changes in the levels of endogenous and exuded abscisic acid, indole-3-acetic acid, cytokinins, salicylic acid, and orobanchol-type strigolactones. The hormonal content in exudates of broomrape-infested mycorrhizal roots resembled that in exudates of infested non-mycorrhizal roots and differed from that observed in exudates of non-infested mycorrhizal roots. Moreover, we observed a significant reduction in AM colonization of infested tobacco plants, pointing to a dominant role of the holoparasite within the tripartite system.",
"introduction": "1. Introduction Rhizosphere is a dynamic platform for complex interactions of plants with the other biotic and abiotic components of the soil ecosystem. The exchange of organic and inorganic matter varies throughout the ontogenesis, and ultimately leads to better adaptation of plants to the fluctuating environment. The chemical composition of root exudates largely shapes the plant-associated microbial communities [ 1 ]. The selective enrichment of bacterial species in the rhizosphere is based on their specific substrate preferences secured by the plant species [ 2 ]. In general, root exudates contain a variety of primary and secondary metabolites of low molecular weight, as well as macromolecules like proteins and polysaccharides. The exuded chemical components have been shown to be functionally implicated in diverse biological processes such as symbiosis, pathogenesis, allelopathy, and mineral nutrition [ 3 , 4 ]. Importantly, all other biotic factors jointly contribute to the chemical composition in the rhizosphere, and their metabolic activities are subjected to feedback regulation. The rhizosphere microbiome modulates root metabolism and exudation by driving long-distance signaling and systemic transcriptional reprogramming in plants [ 5 ]. Beneficial symbiotic interactions between arbuscular mycorrhizal (AM) fungi of the phylum Glomeromycota and land plants nicely illustrate the metabolic synchronization and mutual regulation of the processes of growth and stress adaptation. AM fungi gain from host plants photoassimilates and lipids, and in turn supply the plants with mineral nutrients via an extraradical mycelium network [ 6 ]. Mineral deficiency, such as phosphate limitation, triggers changes in the root exudate composition that positively influence the AM development for better phosphate acquisition [ 7 ]. These precisely orchestrated metabolic readjustments rely on extensive bidirectional small-molecule signaling that underlies all stages of AM symbiosis, including the presymbiotic communication [ 8 ]. In the course of land plant evolution, parasitic weeds have abused the chemical communication between host plant roots and beneficial AM fungi to ensure parasitic seed germination. Strigolactones (SLs) are the best studied class of germination stimulants for seeds of holoparasites, such as broomrapes [ 9 , 10 ]. Apart from being phytohormones, SLs are exuded by the host plant to facilitate root colonization with AM fungi via induction of hyphal branching. The core chemical structure of canonical SLs comprises a methyl butenolide ring attached to a hydrophobic tricyclic scaffold via an enol ether bridge. Besides canonical SLs, non-canonical SLs lacking the typical tricyclic scaffold [ 11 ], as well as sesquiterpene lactones [ 12 ], have also been implicated in broomrape seed germination. Based on their stereochemistry, canonical SLs are subdivided into orobanchol- and strigol-type molecules with species-specific occurrence [ 9 , 13 , 14 ]. Variations in the levels of the two SL types in the root exudate determine the differential susceptibility of host plants to infestation. For instance, sorghum genotypes with reduced levels of 5-deoxystrigol and increased exudation of orobanchol display Striga resistance [ 15 ]. In addition, different combinations of SL species in the exudate have unequal efficiency of boosting AM symbiosis and parasitic seed germination, and such uncoupling of the two processes has potential practical applications [ 16 ]. In contrast to parasitic plants, SL signaling in autotrophic plants does not seem to be directly related to induction of seed germination. SLs interact with protein receptors with α/β hydrolase activity to initiate a signaling cascade in responsive cells (reviewed in [ 10 , 17 , 18 ]). In the course of convergent evolution with photosynthetic plants, in holoparasites a subset of karrikin receptors have acquired specificity for recognition of host-derived SL signals [ 19 ]. Broadened susceptibility for perception of diverse SL classes is presumably the cause for the recently observed expansion of the plant host preferences of Orobanche cumana [ 20 ]. Unlike plant systems, the mechanisms of SL sensing by AM fungi are so far unclear. Plant host interactions with the other biotic factors are mediated by extensive phytohormonal crosstalk. Different hormone signaling pathways share common components to fine-tune the plant response to the changing environment. The latest advances in the studies on hormonal regulation in host plants in the course of root colonization with AM fungi have revealed the complex interplay of growth- and stress-related hormones (reviewed in [ 21 , 22 , 23 ]). For instance, the host DELLA proteins, negative regulators in the gibberellin (GA) signaling pathway, have been shown to play an essential role for promoting AM development [ 24 ]. GAs trigger DELLA polyubiquitination and proteasomal degradation, but other hormones such as abscisic acid (ABA) positively affect DELLA protein functions [ 25 ]. Another aspect of the hormonal crosstalk during AM colonization deals with regulation of hormone biosynthesis. The expression of SL biosynthesis-related genes is auxin-dependent, and SL exudation is reduced in conditions of low auxin content, which in turn impairs mycorrhizal colonization [ 26 ]. Furthermore, it has been demonstrated that ABA deficiency leads to an increase in the ethylene levels, thus restricting mycorrhization [ 27 ]. Hormone partitioning in shoots and roots also appears to be an important determinant for coordinated metabolic interactions between host plants and AM fungi, as has already been shown in experiments with organ-specific cytokinin (CK) depletion [ 28 ]. In contrast to mycorrhizal plant systems, much less is known about the influence of obligate parasitic plants on the hormone homeostasis in host plants. The hormone composition of root exudates is not only a result of the host plant metabolic activity. Recent findings prove that AM fungi produce phytohormones, such as auxins, cytokinins, ethylene, and GA, that might influence both fungal and host plant development [ 29 ]. In turn, plant parasites also use hormonal signals for growth regulation. Genetic approaches have elucidated the importance of auxin and ethylene signaling pathways for proper haustorium formation during parasite invasion [ 30 , 31 ]. It is still elusive as to whether parasitic plants exude metabolites with hormonal activity into the rhizosphere. Here, we used the model AM fungal species Rhizophagus irregularis (former name Glomus intraradices ) and branched broomrape to explore their impact, alone or in combination, on the phytohormone levels in roots and root exudates of a host plant (i.e., oriental tobacco). We found a dominant effect of the parasitic plant on the hormonome in the tripartite system compared to that of the AM fungus. Accordingly, we observed suppressed AM colonization in broomrape-infested tobacco plants. The antagonistic broomrape-tobacco interaction was correlated with diminished levels of exuded SL signals at the tubercle stage of development.",
"discussion": "3. Discussion Since most of the land plants are involved in symbiotic interactions with AM fungi [ 8 ], the chemical communication within pathosystems of plant hosts and holoparasites in nature is performed in the context of AM development. Based on carbon costs, host plants regulate the extent of mycorrhization to maintain the balance between beneficial AM symbiosis and AM parasitism [ 55 ]. Here, we observed broomrape-induced suppression of AM colonization, suggesting that the holoparasite might disturb the establishment of beneficial plant host-AMF interactions through competition with R. irregularis for host-derived photosynthates. Direct physical connections in the tripartite system ensure multidirectional exchange of water and nutrients, but also of small-molecule signals that concomitantly shape the metabolic activity of all partners. SLs are such key chemical signals with a central role for the autoregulation within the system. Their synthesis and exudation by the hosts are finely modulated by other biotic factors and in response to the changing environmental conditions [ 56 , 57 ]. It has already been demonstrated that plants with established AM symbiosis release less SLs in the rhizosphere [ 54 ]. Our data revealed a substantial decrease in ORB exudation by broomrape-infested tobacco plants. Like the autoregulation in conditions of excessive AM colonization, the reduced SL release by infested host roots might be an adaptive strategy to suppress further parasitic seed germination. The latter is supported by the lower germination stimulation activity of root exudates from broomrape-infested tobacco plants. Although the molecular mechanisms underlying the aforementioned feedback inhibition are yet to be explored, the crosstalk with plant defense phytohormones seems to be essential for the adjustment of SL levels. In addition to its prominent role in plant adaptation to abiotic stress conditions, ABA turns out to be intrinsically involved in plant host interactions with AM fungi and holoparasitic plants [ 21 , 37 , 58 ]. As part of a common module for defense response, JA-, SA-, and ABA-related tomato genes have been shown to be upregulated at the initial stages of interaction of tomato plants with Phelipanche ramosa [ 59 ]. Interestingly, their upregulation is accompanied by an increase in the expression of SL biosynthetic genes [ 59 ]. Moreover, SL-deficient tomato mutants are characterized by reduced levels of JA, SA, and ABA, which renders them more susceptible to fungal pathogens [ 60 ]. Like SLs, ABA is synthesized from carotenoid precursors, and the levels of the two signaling molecules are a result of interdependent regulation [ 61 ]. It should be noted that the trends of accumulation of endogenous SL hormones and those of exuded SLs might not correlate in the course of broomrape infestation. The latter reflects the dual role of SLs depending on their localization, i.e., growth regulators within plant tissues as well as small-molecule signals with an impact on other organisms in the rhizosphere. The auxin IAA has also been found to be involved in the regulation of SL exudation. Reduction in the auxin content in an IAA-deficient mutant or after stem girdling leads to a corresponding decrease in the amount of orobanchol and orobanchyl acetate in exudates of pea plants which in turn negatively influences the extent of AM symbiosis [ 26 ]. However, plants with non-impaired IAA metabolism and transport show different dynamics of IAA and SL accumulation. In tobacco mycorrhizal roots, we registered higher amounts of IAA, while the ORB levels in the root exudates were slightly lower than those in the non-colonized controls. The same trends were observed in infested versus non-infested roots and root exudates. An increase in auxin content in mycorrhizal roots has been documented in several plant species except for tobacco [ 62 , 63 ]. The absence of an effect of R. irregularis on tobacco root auxin content reported in [ 63 ] might be due to age-dependent specificities in host plant response to AM colonization, as the hormonal analyses have been done with young tobacco plants at an early stage of AM symbiosis. The AM-induced accumulation of auxin in roots appears to induce local responses essential for AM development. Such cell type-specific reprogramming triggered by auxin has been demonstrated for arbuscule-containing host root cortical cells [ 62 ]. Likewise, auxin-mediated reprogramming possibly takes place at the site of broomrape attachment to the host root in the course of infestation. Cytokinins are another class of plant hormones that act in concert with auxins to shape plant development in optimal and suboptimal conditions [ 64 ]. Analogously to IAA, we detected an increase in the content of bioactive CK bases in root tissues of mycorrhizal plants as well as in broomrape-infested samples. The rise in CK levels might confer limitation of excessive mycorrhization and might also restrict the further spread of broomrape infestation. Such a hypothesis is supported by previous findings demonstrating that CKs play a central role in coordinating the extent of AM colonization with the shoot and root growth of host plants through regulation of the exchange of carbon and phosphate between the two partners [ 28 ]. Interestingly, the increased abundance of IAA and CK bases in host roots co-cultivated with Phelipanche and/or R. irregularis did not result in respective enrichment of those hormonal species in the root exudates. Moreover, the levels of IAA and all studied CKs were consistently lower in exudates of AM colonized roots, a trend that was partially reverted when broomrape was also present in the system. The functional significance of exuded auxins and CKs has yet to be elucidated. Recent studies have demonstrated that host-derived CKs can be released into the growth medium to serve as molecular signals for induction of haustorium formation of P. ramosa [ 65 ]. Besides, R. irregularis also contributes to the CK composition of the growth medium as germinated spores have been shown to release at least one CK metabolite (i.e., isopentenyladenosine) [ 29 ]. In turn, all three partners in the system host-parasite-AM fungus produce auxin and possibly contribute to the auxin levels measured in the medium [ 29 , 31 ]. In Phelipanche , the processes of hormone production and exudation are poorly explored. Whole-genome sequencing of representatives of the Orobanchaceae family should shed light on which of the conventional pathways for phytohormone biosynthesis are functional in broomrapes. In conclusion, we managed to identify previously undescribed hormone-related metabolites present in tobacco root exudates. The comparative analysis in roots and root exudates revealed characteristic hormonal profiles associated with the impact of mycorrhization as well as with broomrape development. For most of the studied metabolites, the trends detected upon co-cultivation of tobacco concomitantly with AM spores and broomrape seeds resembled those registered in samples from broomrape-infested non-inoculated plants, an observation pointing to a dominant effect of broomrape over the AM fungus within the tripartite system. A schematic summary of the most characteristic changes in the plant hormonome identified in this study is provided in Figure 7 . In the long term, these findings might expand the possibilities for modulating the communication in the host plant-parasitic plant-AM fungi system in search for advanced strategies for improvement of host plant tolerance to Phelipanche infestation."
} | 4,060 |
34512857 | PMC8423106 | pmc | 3,492 | {
"abstract": "Abstract Spatial patterns of movement regulate many aspects of social insect behavior, because how workers move around, and how many are there, determines how often they meet and interact. Interactions are usually olfactory; for example, in ants, by means of antennal contact in which one worker assesses the cuticular hydrocarbons of another. Encounter rates may be a simple outcome of local density: a worker experiences more encounters, the more other workers there are around it. This means that encounter rate can be used as a cue for overall density even though no individual can assess global density. Encounter rate as a cue for local density regulates many aspects of social insect behavior, including collective search, task allocation, nest choice, and traffic flow. As colonies grow older and larger, encounter rates change, which leads to changes in task allocation. Nest size affects local density and movement patterns, which influences encounter rate, so that nest size and connectivity influence colony behavior. However, encounter rate is not a simple function of local density when individuals change their movement in response to encounters, thus influencing further encounter rates. Natural selection on the regulation of collective behavior can draw on variation within and among colonies in the relation of movement patterns, encounter rate, and response to encounters."
} | 348 |
23844990 | PMC4288973 | pmc | 3,493 | {
"abstract": "Arbuscular mycorrhizal (AM) fungi are the most abundant plant symbiont and a major pathway of carbon sequestration in soils. However, their basic biology, including their activity throughout a 24-h day : night cycle, remains unknown.\n \nWe employed the in situ Soil Ecosystem Observatory to quantify the rates of diurnal growth, dieback and net productivity of extra-radical AM fungi. AM fungal hyphae showed significantly different rates of growth and dieback over a period of 24 h and paralleled the circadian-driven photosynthetic oscillations observed in plants.\n \nThe greatest rates (and incidences) of growth and dieback occurred between noon and 18:00 h. Growth and dieback events often occurred simultaneously and were tightly coupled with soil temperature and moisture, suggesting a rapid acclimation of the external phase of AM fungi to the immediate environment.\n \nChanges in the environmental conditions and variability of the mycorrhizosphere may alter the diurnal patterns of productivity of AM fungi, thereby modifying soil carbon sequestration, nutrient cycling and host plant success.",
"introduction": "Introduction Of broad interest to scientists are Earth system processes that are modulated by the 24-h day : night cycle (Kayanne et al ., 1995 ; Canadell et al ., 2000 ; Mears & Wentz, 2005 ), especially diurnal patterns related to plants (McClung, 2006 ). For example, the opening of plant stomata by day and their closure at night – that is, in general, for C 3 and C 4 pathways – ultimately drives carbon (C) fluxes at ecosystem and global scales (Hetherington & Woodward, 2003 ). Arbuscular mycorrhizal (AM) fungi that exist symbiotically with the roots of plants also play a critical role in the C cycle, by increasing nutrient and water uptake, which facilitates plant growth, and by providing a C sink in the form of respired C, hyphal mass and glycoproteins, such as glomalin. Extra-radical hyphae, observed in AM fungi (Glomeromycota), grow from the root into the soil in lengths over 100 m cm −3 soil (Miller et al ., 1995 ), assimilating C received directly from the host plant in exchange for nutrients. The hyphae of AM fungi respire an estimated 3% of fixed C (Paul & Kucey, 1981 ) and turnover C during the part of the day when host plants are most photosynthetically active (Staddon et al ., 2003 ). However, we lack a basic biological understanding of the timing, magnitude and controls of productivity of AM fungi in this external phase at the daily time scale. Field-based estimates of the productivity of AM fungi have relied on the removal of soil cores for the estimation of the length, diameter and total volume of hyphae (Allen & MacMahon, 1985 ; Allen & Allen, 1986 ; Miller & Jastrow 1990 ; Miller et al ., 1995 ). This destructive method precludes the opportunity to measure changes in the hyphal length of AM fungi – that is, growth and dieback – over time. Today, destructive (e.g. soil coring, in-growth mesh bags) and other static ‘snapshot’ methods are still employed for the measurement of various dynamics of productivity of mycorrhizal fungi (i.e. net growth of AM fungal hyphae) and turnover (Miller et al ., 1995 ). Using these methods, spatial and temporal variation cannot be separated. Individual soil cores cannot be replicated in time, as the sampled material is destroyed and the change between adjacent points over distances of a few millimeters can be as large as between sample periods (Allen & MacMahon, 1985 ; Belnap et al ., 2003 ). Consequently, the study of hyphal productivity of AM fungi is hampered by logistical problems, including the difficulty in observing microscopic hyphae without concomitantly altering their function, the inability to generate replication in space from which a soil sample is removed and confounding literature from laboratory and glasshouse experiments which may not represent the conditions in natural systems (Allen et al ., 2007 ; Vargas & Allen, 2008a , b ). For example, Wang et al . ( 1989 ) found that C was allocated to AM fungal hyphae within 24 h of labeling in a growth chamber experiment, but these rates are unknown in the soil ecosystem. Treseder et al . ( 2010 ) utilized a Bartz minirhizotron zoomed in at ×100 to describe the dynamics of single large AM fungal hyphae in situ over a season, but the hand imaging and analyses are slow and tedious processes to undertake extensive analyses. Advances in soil imaging automation, environmental sensor technology and digital image processing, together, have conferred the opportunity to observe extra-radical fungal hyphae directly, at unprecedented temporal scales, and in concert with the measurements of variables that characterize the mycelium environment (Allen et al ., 2007 ; Rundel et al ., 2009 ; Iversen et al ., 2011 ). Like plants, AM fungi may express diurnal-scale patterns of growth, dieback and net productivity, in accordance with endogenous circadian systems or environmental stimuli (Heinemeyer et al ., 2006 ); however, to date, no study has tested this directly. Although not suggestive of a diurnal periodicity per se , Johnson et al . ( 2002 ) found that the in situ flux of pulse-derived 13 C peaked in its release as 13 CO 2 in AM fungi-colonized soil cores in the first 0–6 h, revealing a rapid allocation of host plant C to AM hyphae. Previous field-based studies on AM rhizomorphs – that is, distinct, large cords comprising several hyphal strands – using in situ manual minirhizotrons (Fig. 1 ), showed rapid diurnal changes in rhizomorph length, with growth up to 105.4 cm m −2 d −1 (Heinemeyer et al ., 2006 ; Vargas & Allen, 2008a , b ; Hasselquist et al ., 2010 ). Given that diel changes in rhizomorphs were dynamic over a 24-h interval, this suggests that individual fungal hyphae may respond similarly. Together, these results suggest that the productivity of mycorrhizal fungi has yet to be quantified at the time scale in which it operates and, if so, any patterns of growth and dieback over a 24-h day : night cycle remain undiscovered. Figure 1 (a) Three individual Soil Ecosystem Observatories (SEOs) without their exterior tubes (note: one single SEO was employed in this study). The Proscope camera was equipped with a 2-megapixel (1.92 million effective pixels) color sensor with a precision of 0.1 mm, accuracy of 0.3 mm and ×100 magnification. The wavelength of the light source was 322 nm. The total length of the SEO is 156.87 cm (107.95 mm in diameter). The maximum imaging area for the entire tube is 320 × 700 mm 2 with a maximum number of 32 928 images (3.01 × 2.26 mm 2 each). The SEO has an operating range of −12 to 45°C. (b) Embedded above- and belowground soil sensor network (University of California James San Jacinto Mountains Reserve, Idyllwild, CA, USA) showing the installed SEO (labeled ‘AMR Unit’ in photograph), Campbell CR1000 data logger and Campbell Li-Cor Quantum Sensor. (c) Drawing of the SEO prototype showing the Proscope camera at home position and fastened to the robotic sled. In addition, environmental stimuli can entrain or reset circadian oscillators, making even circadian rhythms responsive to short-lived external cues, such as light and temperature (Liu & Bell-Pederson, 2006 ). If the productivity of AM fungal hyphae varies at the diurnal scale, we have increasing evidence that productivity is not only a response to photosynthesis in the host plant (e.g. light stimuli), but also a direct response to changes in the mycorrhizosphere (the zone in which AM hypha–soil interactions occur). For example, Vargas & Allen ( 2008b ) found that rhizomorph growth was positively coupled with soil temperature and precipitation events. In a controlled glasshouse experiment, Heinemeyer et al . ( 2006 ) found that extra-radical AM fungi increased in length as a result of an increase in temperature. Indeed, glasshouse experiments have shown that AM fungal hyphae are directly modulated by temperature (Heinemeyer et al ., 2006 ), but experiments utilizing natural environmental variation are needed to better understand and confirm these relationships. In this study, our goal was to employ the Soil Ecosystem Observatory as a new method to quantify the diurnal growth, dieback and productivity of extra-radical AM mycorrhizal fungi – the normal condition for c . 80% of Earth's terrestrial plant species and most economic crops (Smith & Read, 2008 ). For this study, we set the temporal resolution of the Soil Ecosystem Observatory to 6-h observations of the soil to discern changes in the dynamics of AM hyphae occurring within a single day. In addition, we tested the hypothesis that AM fungal extra-radical hyphae exhibit differential rates of growth (i.e. elongation) and dieback (i.e. disappearance) throughout a day : night cycle; specifically, that certain 6-h intervals (e.g. 12:00 to 18:00 h, when plant photosynthetic activity is greatest) may show higher rates of growth and dieback than other 6-h intervals within a 24-h time period; furthermore, we examined whether growth and dieback events of AM fungal hyphae are correlated with abiotic variables in the mycorrhizosphere (i.e. soil temperature ( T s ), soil moisture (SWS)). Lastly, we were interested in determining whether the growth and dieback of AM fungal hyphae can occur simultaneously and at similar rates, or whether such events are mutually exclusive.",
"discussion": "Discussion Diurnal patterns of AM fungal hyphae in a natural environment The description of the diurnal patterns of plant physiology and their relationship with environmental factors has been of interest to scientists for over 60 yr (Larcher, 2003 ). From early on, scientists recognized the challenge and importance of taking plant ecophysiological measurements in natural settings (Tenhunen et al ., 1987 ; Mooney & Field, 1989 ; Mooney, 1991 ). Such in situ measurements of short-term dynamics provided the foundation for the elucidation of not only whole-plant function, but ecosystem level and Earth system processes (Mooney & Field, 1989 ; Zotz & Winter, 1993 ; Hetherington & Woodward, 2003 ; McClung, 2006 ). In comparison, relatively little progress was made – and still remains the case today – in understanding the diurnal patterns of the plant–fungus symbiosis, an essential relationship if we wish to truly describe plant physiology and function as a ‘whole’. Observations of AM fungi in natural settings are essential because of the inherent complexity of the plant–fungi symbioses (Treseder, 2004 ) in which fungi exhibit plant-dependent and -independent responses, and such responses both create and are impacted by numerous feedbacks. To this end, understanding the short-term dynamics of AM fungi requires an understanding of plant-dependent and -independent responses simultaneously, together with their associated feedbacks, which can only be accomplished in a natural setting. Before this study, the timing of the growth of extra-radical hyphae (and their rate of growth) during a 24-h day : night cycle was not well understood. However, several studies made prescient hypotheses alluding to their dynamism based on the rapid allocation of labeled 11 C to plant roots and mycorrhizal fungi in a phytotron (Wang et al ., 1989 ) and 14 CO 2 to soil cores colonized by AM (Johnson et al ., 2002 ), measurements of rhizomorph growth (Vargas & Allen, 2008a , b ; Hasselquist et al ., 2010 ), rates of coarse hyphal production (Treseder et al ., 2010 ) and standing crop estimates of AM hyphae interpolated from soil core measurements (Miller et al ., 1995 ). In this study, we determined that the rates of in situ AM fungal productivity in the meadow of a semiarid mixed conifer forest fluctuated throughout a 24-h day : night period. Growth rates and incidences of elongation were maximal from 12:00 to 17:59 h – when elongation often exceeded 150 μm mm −3 h −1 and soils were at peak temperature. Similar to the circadian and exogenously driven oscillation system that drives the photosynthesis observed in plants, hyphal elongation may be a sinusoidal pattern, peaking in the same 6-h interval when C 3 and C 4 plant productivity is also greatest (Larcher, 2003 ). Patterns of dieback paralleled growth, with incidences and rates of dieback of AM fungal hyphae being greatest in the interval from 12:00 to 17:59 h. Reductions in external hyphae are important to understand at the diurnal scale because, like growth, hyphal dieback may influence phosphorus inflow to the host plant (Liu et al ., 2012 ). However, unlike growth, increased hyphal dieback at diurnal timescales may be indicative of changes in resource allocation from belowground to shoots and leaves (Bloom et al ., 1985 ; Johnson et al ., 2003 , 2008 ) and/or fungivory, and also regulates directly the flow of C, either as throughput or as more stable sources, into the soil. Future studies exploring the diurnal patterns of dieback of AM fungal hyphae should evaluate how such rates and incidences of dieback differ throughout the rhizosphere (e.g. depth, distance from host plant) and across ecosystems. Our observations suggest the existence of a biotic (e.g. photosynthate availability) and abiotic (e.g. temperature) ‘optimum’ for AM fungal networks, conditions that probably operate synergistically to facilitate hyphal productivity through the soil. In that vein, the term ‘optimal’ may be somewhat misleading. This is because resource limitations (e.g. C, phosphorus, nitrogen), in the host plant or its symbiont, may also drive elongation (Orwin et al ., 2011 ). Further underscoring the complexity of the conditions influencing changes within a mycelium network, our study showed that the growth and mortality of AM fungal hyphae occurred throughout the entire 24-h day : night period, even when host plants ceased light-dependent photosynthesis and PPFD was negligible. Consequently, lags and factors external to the plant–fungi symbiosis probably further impact the timing and magnitude of the productivity of AM fungal hyphae. We anticipated a net reduction in AM fungal standing crop, as May is the end of the growing season in this Mediterranean-type ecosystem (for details, see Vargas & Allen, 2008a ). Integrating across all intervals, our study provides an estimate of daily production at 1.6 mm mm −3 and mortality at 2.5 mm mm −3 in the mycorrhizosphere in May. This equates to a loss of roughly 0.9 mm mm −3 mycelium per day throughout our campaigns. As this ratio of growth : dieback will probably change throughout the year and by ecosystem type, future studies should evaluate diurnal patterns of growth and dieback during biologically important times throughout the year and in different ecosystems to better understand the annual rates of productivity. Of importance for understanding AM fungal turnover, we observed anecdotally that runner hyphae (coarser in diameter) persisted throughout most of both campaigns, whereas hyphae strands thinnest in diameter were most likely to die back. The most dynamic hyphae are part of the absorbing network tips, which have been shown previously to respond to rapid stimuli (Friese & Allen, 1991 ). These data also confirm the observations of Treseder et al . ( 2010 ), who noted that the coarse hyphae had a residence time of > 145 d during the summer drought. Our observations also addressed the question of whether hyphal elongation and dieback are mutually exclusive within the mycorrhizosphere. Much like the simultaneous growth and senescence of different leaves on a tree, we found that elongation and dieback can occur concomitantly within AM fungal networks. Consequently, future studies should not only elucidate the factors impacting production, but also tease apart their roles in facilitating or inhibiting elongation and dieback events separately. Short-term dynamics of AM fungal hyphae, respiration and abiotic soil conditions In laboratory and glasshouse experiments, several studies have shown that AM hyphae are modulated directly by abiotic conditions in the soil (Monz et al ., 1994 ; Rillig et al ., 2002 ; Gavito et al ., 2003 ; Alberton et al ., 2005 ; Heinemeyer et al ., 2006 , 2007 ; Vicca et al ., 2009 ; Cheng et al ., 2012 ). For example, Monz et al . ( 1994 ) reported that the colonization by AM fungi in two host plants decreased when precipitation increased, and one plant showed reduced colonization in response to increasing temperature – patterns also observed in our study. Because our study conferred direct observations and measurements in a natural habitat, our findings confirm previous hypotheses on the acclimation of extra-radical AM hyphae to temperature. AM fungal hyphae also appear to respond rapidly, potentially even to hourly changes in response to host and diurnal environmental cycles. Staddon et al . ( 2003 ) found a turnover of hyphae within 5–6 d based on soil 14 C studies, showing that these fungi respond markedly at short time scales. At a nearby meadow to the current study, during May of 2006 (phase III of their study), using a Bartz conventional minirhizotron and the networked sensor instrumentation equivalent to that described here, soil respiration responded directly to soil temperature and moisture, showing no hysteresis, at a 24-h time scale (Vargas & Allen, 2008b ). Soil respiration was correlated primarily with soil temperature, vapor pressure deficit (VPD) and photosynthetically active radiation (PAR). Putting our observations into these contexts, the growth and respiration of hyphae appear to be tightly coupled to the environmental and plant growth drivers. The primary environmental variable changing within the 24-h cycle is soil temperature, peaking in the afternoon, into the evening. The plant variables are VPD, which drives transpiration, and thus CO 2 exchange through the stomata, and PAR, which drives CO 2 fixation. Thus, hyphal production and soil respiration are co-correlated with each other and with critical environmental (temperature) and plant photosynthetic (VPD, PAR) variables that cycle on a diurnal scale. Hyphal production was more dynamic in 2010 – when PPFD and soil temperatures showed higher variability – than in 2009. It is also clear that AM fungal hyphae are dynamic within a short (24 h) period of time. AM fungi clearly are modulated directly (exogenously; e.g. temperature) and indirectly (e.g. circadian oscillators, assimilation in host plant), as suggested in several studies (Monz et al ., 1994 ; Rillig et al ., 2002 ; Gavito et al ., 2003 ; Alberton et al ., 2005 ; Heinemeyer et al ., 2006 , 2007 ; Cheng et al ., 2012 ). Consequently, net growth of AM fungal hyphae may be impacted by short-term weather events (e.g. heat-waves and droughts) and also longer term changes in weather, such as climate change (Rillig, 2004 ; Govindarajulu et al ., 2005 ; Cheng et al ., 2012 ). Summary Our study's novel technological approach, the Soil Ecosystem Observatory, revealed that the productivity of AM fungal hyphae and its interaction with the soil ecosystem is more complex than previously thought, requiring a nondestructive and observation-based experimental approach. First, our study showed that AM fungal hyphae are highly dynamic across a diurnal time scale, with maximal rates and incidences of growth and mortality occurring from noon to 17:59 h. Production and mortality rates were tightly coupled with diurnal-scale changes in temperature and hydrology within the mycorrhizosphere. Second, we found that elongation and dieback can occur simultaneously within a single AM fungal network, with the magnitude of each rate determining whether there is a net input to the soil. In addition, our study suggests that variability in soil abiotic factors may increase variability in hyphal production and mortality rates, such as those observed in 2010. Lastly, our study underscores the vast potential of Soil Ecosystem Observatories to elucidate soil microbial function where direct human observation and measurements are otherwise impossible. This is a leap forward in the understanding of mycorrhizal fungal ecology, as past studies have typically measured processes on weekly or monthly time intervals, and have relatively overlooked extra-radical AM mycelium."
} | 5,084 |
40292401 | PMC12034278 | pmc | 3,495 | {
"abstract": "Abstract Helicobacter pylori ( H. pylori ) is prevalent in over 50% of the global population and is recognized as the primary etiological agent for the development of gastric cancer. With the increasing incidence of antibiotic resistance, clinical treatment of H. pylori is a significant challenge. The formation of H. pylori biofilm is an important reason for antibiotic resistance and chronic infection, and it is also one of the key obstacles to eradicating H. pylori. H. pylori biofilm acts as a physical barrier, preventing the penetration of antibiotics and increasing the expression of efflux pump genes and drug-resistant gene mutations. Therefore, the treatment of H. pylori biofilm is extremely challenging. Nanomaterials, such as inorganic nanoparticles, lipid-based nanoparticles, and polymeric nanoparticles, which have properties including disrupting bacterial cell membranes, controlling drug release, and overcoming antibiotic resistance, have attracted significant interest. Furthermore, nanomaterials have the ability to treat H. pylori biofilm owing to their unique size, structure, and physical properties, including the inhibition of biofilm formation, enhancement of biofilm permeability, and disruption of mature biofilm. Moreover, nanomaterials have targeting functions and can carry antimicrobial drugs that play a synergistic role, thus providing a prospective strategy for treating H. pylori biofilm. In this review, we summarize the formation and antibiotic-resistance mechanisms of H. pylori biofilm and outline the latest progress in nanomaterials against H. pylori biofilm with the aim of laying the foundation for the development and clinical application of nanomaterials for anti– H. pylori biofilm.",
"conclusion": "Conclusion and Outlook Antibiotic resistance and high eradication-failure rates are severe challenges currently faced by H. pylori treatment strategies. H. pylori can form biofilm not only in vitro but also in vivo, and the formation of biofilm is closely related to its antibiotic resistance. Mechanisms such as the barrier effect of biofilm, the involvement of efflux pumps, and changes in bacterial metabolic activity within biofilm lead to drug resistance and limit the clinical application of antibiotic-based eradication regimens. Moreover, the use of antibiotics not only exacerbates drug resistance but also causes a variety of adverse reactions, leading to poor patient compliance. Therefore, exploring antibiotic-independent strategies for treating H. pylori biofilm is an important direction for addressing the challenges of biofilm-forming H. pylori eradication. Nanomaterials, with their unique physicochemical properties and highly efficient antibiofilm mechanisms, have emerged as a promising new strategy. Nanomaterials can eradicate biofilm-forming H. pylori through various mechanisms, including preventing biofilm formation, improving biofilm permeability, disrupting mature biofilm, and serving as a drug-delivery system that synergizes with antimicrobials. Based on their type, nanomaterials offer certain advantages to eradicating biofilm-forming H. pylori : 1) inorganic NPs possess photosensitivity, magnetism, and thermal properties, which enable them to directly disrupt H. pylori biofilm through physical mechanisms; 2) lipid-based nanomaterials show strengths in H. pylori biofilm penetration, high biocompatibility, and drug-loading capacity; and 3) polymeric NPs offer superior functionalization, stability, and pH responsiveness. However, nanomaterials also have certain limitations. The cost of metal NPs needs to be reduced, the stability of lipid NPs needs to be improved, and the biocompatibility of polymer NPs needs to be further enhanced. Therefore, further research is needed to integrate the advantages of various types of nanomaterial for efficient H. pylori biofilm treatment, such as targeting biofilm-specific metabolites or working synergistically with antibiotics. Although existing studies have confirmed that nanomaterials have significant effects and potential in eradicating H. pylori biofilm, their application in clinical practice still faces numerous obstacles. First, the design of nanomaterials must take into account the complex physiological environment of the human body to ensure that they are safe, atoxic, and biodegradable. In addition, the processes of drug administration, metabolism, and excretion of nanomaterials need to be optimized and validated to maximize their antibiofilm efficacy and biosafety. At the same time, simplifying the synthesis steps and reducing costs to facilitate large-scale production are also crucial. In summary, with continuous optimization, nanomaterials hold promise as part of a new generation of strategies for treating biofilm-forming H. pylori and are expected to ultimately achieve the eradication of H. pylori in the near future.",
"introduction": "Introduction Helicobacter pylori ( H. pylori ) is a Gram-negative, spiral-shaped, microaerophilic bacterium that specifically colonizes the human stomach of around 50% of the global population, with a higher prevalence in developing countries. 1–4 \n H. pylori is a significant contributor to a broad range of gastrointestinal diseases, including chronic gastritis, peptic ulcer disease, autoimmune diseases, and gastric adenocarcinoma. 5 , 6 In 1994, H. pylori was classified as a class I carcinogen by the World Health Organization. 7 By the end of 2021, the United States had updated its 15th report on carcinogens, adding eight new carcinogenic substances and listing H. pylori as a definite carcinogen. 8 In addition, increasing research suggests that chronic infection with H. pylori is associated with a series of extraintestinal diseases, such as coronary heart disease, iron-deficiency anemia, and neurological disorders. 9–11 Hence, the eradication of H. pylori is necessary for the maintenance of human health. The current treatment of H. pylori mainly relies on antibiotics. However, with the widespread use of antibiotics, the problem of bacterial resistance has become increasingly prominent. 12 , 13 Among the various antibiotic-resistance mechanisms of H. pylori , the formation of bacterial biofilm is an important factor. 14 As a bacterial barrier, biofilm reduces or delays the entry or penetration of antibiotics, preventing them from contacting the bacterial cells, thereby leading to antibiotic resistance and causing persistent and chronic H. pylori infections in the host. 15 , 16 Moreover, efflux pump genes that form biofilm are highly expressed in H. pylori , promoting the development of antibiotic resistance. 17 Furthermore, the morphology of H. pylori within biofilm changes from spiral to coccoid form, leading to increased tolerance to antibiotics. 18 Therefore, when conventional antibiotics are prescribed clinically, their efficacy in eradicating biofilm-associated H. pylori infection is significantly reduced due to the aforementioned factors. Alternative therapies for H. pylori biofilm have been explored. For example, antimicrobial peptides possess broad-spectrum antibacterial activity and can penetrate the extracellular matrix of biofilm to directly target and disrupt bacterial cells. 19–21 Studies have shown that the antimicrobial peptides DJK-5 and IDR-1018 can inhibit the growth and maturation of H. pylori biofilm. 22 Some probiotics can also interfere with H. pylori biofilm. For instance, Lactobacillus fermentum UCO-979C and Lactobacillus plantarum LN66 can inhibit the formation of H. pylori biofilm. 23 , 24 Moreover, certain natural products can disrupt H. pylori biofilm. Silvia et al used Pistacia vera L. oleoresin in combination with levofloxacin to synergistically disrupt H. pylori biofilm. 25 However, these studies are still relatively immature, and the efficiency of antibiofilm activity as well as biosafety needs to be improved. Therefore, it is important to develop innovative treatment approaches that target H. pylori biofilm. With the rapid development of nanotechnology, an increasing number of nanomaterials have shown great potential in the field of biofilm treatment. 26–28 Nanomaterials consist of particles with dimensions ≤100 nm. 29 Depending on the matrix components, nanomaterials used for antibacterial purposes can be classified into various types, such as inorganic nanoparticles (NPs; including metal-based NPs or their oxides), lipid-based NPs (eg, liposomes, nanoemulsions), and polymer NPs (natural or synthetic polymers). 30 Nanomaterials have several benefits, including diminutive size, elevated specific surface area, substantial drug-carrying potential, and the ability to release drugs in response to specific stimuli. These characteristics enable them to overcome current resistance barriers and enhance the efficacy of biofilm eradication. 31–33 Some studies on these nanomaterials have already entered the clinical trial stage. For instance, amikacin liposomes have been used to treat pulmonary infections caused by M. tuberculosis in biofilm. 34 In addition, nanomaterials have been used to treat H. pylori biofilm at different stages. First, lipid–polymer NPs (LPNs) can interfere with the attachment of H. pylori to the gastric epithelium, thereby blocking biofilm formation. 35 Polymer NPs can enhance the permeability of H. pylori biofilm by adjusting their surface properties. 36 Inorganic NP-based photothermal therapy/photodynamic therapy can directly destroy mature H. pylori biofilm by interacting with the biofilm through oxidative stress and hyperthermia. 37 Moreover, an increasing number of studies are focused on drug-delivery nanocarriers that have been engineered to specifically target H. pylori biofilm and regulate the release of antimicrobial agents, aiming to reduce the usage of antibiotics and decrease adverse side effects. 38 , 39 In this review, we first provide a brief introduction to the biological traits of H. pylori biofilm and highlight its fundamental mechanisms of formation. We review current studies on the mechanisms of H. pylori biofilm–enhancing antibiotic resistance. Furthermore, we discuss advancements in research regarding the interaction of various nanomaterials with H. pylori , focusing on their biofilm-targeting capabilities and summarizing their properties and mechanisms of action against H. pylori biofilm. Finally, we assess the prospects and challenges associated with the application of nanotechnology in the treatment of H. pylori biofilm."
} | 2,644 |
25525408 | null | s2 | 3,496 | {
"abstract": "While the remarkable chemical and biological properties of DNA have been known for decades, these properties have only been imparted into materials with unprecedented function much more recently. The inimitable ability of DNA to form programmable, complex assemblies through stable, specific, and reversible molecular recognition has allowed the creation of new materials through DNA's ability to control a material's architecture and properties. In this review we discuss recent progress in how DNA has brought unmatched function to materials, focusing specifically on new advances in delivery agents, devices, and sensors."
} | 156 |
27924027 | PMC5314801 | pmc | 3,497 | {
"abstract": "Pseudomonas aeruginosa possesses at least three well-defined quorum-sensing (QS) ( las, rhl and pqs ) systems that control a variety of important functions including virulence. RsaL is a QS repressor that reduces QS signal production and ensures homeostasis by functioning in opposition to LasR. However, its regulatory role in signal homeostasis remains elusive. Here, we conducted a ChIP-seq assay and revealed that RsaL bound to two new targets, the intergenic regions of PA2228/PA2229 and pqsH/cdpR , which are required for PQS synthesis. Deletion of rsaL reduced transcription of pqsH and cdpR , thus decreasing PQS signal production. The Δ rsaL strain exhibited increased pyocyanin production and reduced biofilm formation, which are dependent on CdpR or PqsH activity. In addition, we solved the structure of the RsaL–DNA complex at a 2.4 Å resolution. Although the overall sequence similarity is quite low, RsaL folds into a HTH-like structure, which is conserved among many transcriptional regulators. Complementation results of the rsaL knockout cells with different rsaL mutants further confirmed the critical role of the DNA-binding residues (including Arg20, Gln27, Gln38, Gly35, Ser37 and Ser42) that are essential for DNA binding. Our findings reveal new targets of RsaL and provide insight into the detailed characterization of the RsaL–DNA interaction.",
"introduction": "INTRODUCTION Bacteria use small diffusible molecules, also known as autoinducers, as signals for monitoring population density and coordinating gene regulation via a process termed quorum-sensing (QS) ( 1 , 2 ). Many Gram-negative bacterial species, including several human and plant pathogens, use acylated homoserine lactones (AHLs) as QS signal molecules ( 3 , 4 ). AHLs are synthesized by LuxI-type synthases and detected by LuxR-type regulators, which serve as the signal receptors. Once AHL concentration reaches a specific threshold, the LuxR–AHL complex binds to palindromes within quorum-controlled promoters and activates the expression of QS-dependent genes ( 1 ). Pseudomonas aeruginosa is an opportunistic human pathogen that can cause both acute and chronic infections in hospitalized and immunocompromised hosts. Pseudomonas aeruginosa frequently causes life-threatening infections in cystic fibrosis patients; such infections are mediated by multiple QS-regulated virulence factors such as protease, exotoxin, pyocyanin, and biofilms ( 5 – 7 ). Two well-defined AHL QS systems, las and rhl , exist in P. aeruginosa ( 8 ). The las system consists of the transcriptional regulator LasR and the QS signal synthase LasI. The lasI gene product directs the biosynthesis of 3-oxo-C 12 -HSL, which interacts with LasR and activates target promoters. The rhl system consists of the transcriptional activator RhlR and the enzyme RhlI that is responsible for the biosynthesis of C 4 -HSL ( 5 , 9 ). In addition to 3-oxo-C 12 -HSL and C 4 -HSL, P. aeruginosa produces diverse 2-alkyl-4-quinolones (AHQs) as the third group of QS signal molecules ( 10 ). The major AHQ signals include 2-heptyl-3-hydroxy-4-quinolone (the Pseudomonas quinolone signal [PQS]) and 2-heptyl-4-quinolone (HHQ) ( 10 , 11 ). PQS synthesis is catalyzed by enzymes encoded by the pqsABCDE and phnAB operons as well as pqsH ( 11 ). DNA microarray analysis has revealed that hundreds of genes are controlled by the quorum-sensing systems in P. aeruginosa ( 12 ). The P. aeruginosa QS circuitry is complex and hierarchical. For example, the post-transcriptional regulator RsmA modulates production of virulence determinants and QS signals by binding to the lasI and rhlI promoters ( 13 , 14 ); VqsM is a global regulator of QS and virulence factors in P. aeruginosa ( 15 ). Previously, we have shown that VqsM binds directly to the promoter region of lasI , thus controlling QS-regulated phenotypes ( 16 ). A number of regulators involved in controlling the activation threshold of quorum-related genes, such as QscR ( 17 , 18 ) and QteE ( 19 ), have also been identified. QslA is an anti-activator of QS that regulates virulence factor production by interacting with LasR and preventing it from binding to its target DNA sequence ( 20 , 21 ). Recently, we identified a novel regulator, CdpR, which is required for PQS production and virulence factor expression ( 22 ). Another important regulator, RsaL, acts as a major repressor of the las system by binding to the lasI promoter, which controls the maximal level of AHLs and thus virulence factor production ( 23 ). Microarray results have shown that RsaL regulates at least 341 genes, including the most important virulence genes (i.e. lasA, rhlA and phzA1 ) ( 24 ). As a global regulator, RsaL controls gene expression through different mechanisms including repression of 3-OC 12 -HSL signal production, direct binding to target genes (such as phzA1, phzM and hcnA ), and indirect regulation of several genes via other unknown regulators ( 25 ). Moreover, it has been shown that in P. aeruginosa 3OC 12 -HSL concentration reaches a steady state long before stationary phase ( 25 ), indicating that unidentified homeostatic mechanisms contribute to limiting 3OC 12 -HSL production. In this study we searched for additional RsaL targets using a ChIP-seq assay. Our results revealed that RsaL binds to the intergenic region between pqsH and cdpR , which are involved in PQS signal synthesis. Furthermore, our experiments demonstrated that the altered phenotypes of the Δ rsaL strain are dependent on PqsH or CdpR expression. Additionally, we report the crystal structure of RsaL bound to DNA, which is reminiscent of a HTH transcription factor bound to dsDNA. In summary, the work presented here identifies RsaL targets involved in QS and provides a molecular structure of RsaL interacting with the promoter region of a QS target.",
"discussion": "DISCUSSION The QS systems of P. aeruginosa consist of complex regulatory networks and play an important role in the pathogenicity of this bacterium. However, the detailed regulatory mechanisms of a number of QS regulators remain elusive. Previous studies have shown that RsaL is a repressor of the las system ( 23 , 25 ); however, it remains uncharacterized otherwise. Here, we performed ChIP-seq and EMSA experiments that identified two new in vivo binding sites of RsaL in the P. aeruginosa genome. We also solved the crystal structure of the RsaL–DNA complex. The combination of the present ChIP-seq results and structural model of RsaL should provide a much better understanding of the regulatory mechanisms of RsaL. Of the two RsaL-bound regions, one is located in the PA2225-PA2228 operon and the other is the intergenic region between pqsH and PA2588. However, other targets identified from ChIP-seq assays are not bound by RsaL using EMSA. In addition, some previously established direct targets (such as lasI, phzA1, phzM and hcnA ) were not identified in the present work. The omission of these genes may be due to the conditions used in the present study (i.e. using the pAK1900 vector and mid-log phase cultures), thus, further optimization of this method is needed. Importantly, the ChIP-seq data and EMSA analysis showed that RsaL binds efficiently to the intergenic region between pqsH and cdpR (Figure 1 ), which explains the reduction in pqsH/cdpR transcription level and PQS production in the rsaL mutant (Figures 2 and 3 ). Previous microarray results have shown that RsaL can control several hundreds of genes, including the most important virulence genes involved in biofilm formation and pyocyanin production ( 24 , 25 ). Based on these observations, RsaL most probably enables the control of QS-regulated phenotypes via PqsH or CdpR. In agreement with this hypothesis, expression of pqsH in the rsaL mutant was sufficient to restore wild-type biofilm production levels (Figure 2E and F ). Moreover, RsaL is required for cdpR expression and the altered pyocyanin production in the Δ rsaL strain is dependent on CdpR (Figure 3 ). Therefore, RsaL can control these virulence genes via multiple pathways including direct binding to their promoters and indirectly blocking las or pqs -dependent transcription. To further elucidate the QS regulatory networks, the crystal structures of two QS-related proteins have been solved; the Bottomley et al . reported the crystal structure of the LasR ligand-binding domain bound to its autoinducer 3-oxo-C 12 -acylhomoserine lactone ( 42 ) and the Lintz et al determined the structure of the QS repressor QscR bound to N-3-oxo-dodecanoyl-homoserine lactone ( 18 ). A BLASTP search of the RsaL amino acid sequence did not reveal significant homology with functionally characterized proteins in P. aeruginosa . Therefore, we determined the crystal structure of the RsaL–DNA complex. RsaL mainly interacts with the target DNA using its α3 helix ( 37 SQSCGSRFEN 46 ), which packs against the DNA major groove and recognizes the base pairs (Figure 5 ). Furthermore, one residue (Arg20) from α1 and two residues (Gln27 and Glu28) from α2 also participate in DNA backbone recognition (Figure 5D ). RsaL has no significant sequence similarity to any known proteins in other species; however, the DALI server program ( 43 ) revealed that the overall fold of RsaL is similar to the HTH domains (Supplementary Figure S7A) conserved in many DNA transcriptional regulators, such as HNF6 ( 44 ), SATB1 ( 45 ), MqsA ( 46 ) and Oct1 ( 47 ). Although obvious differences can be observed in the α1 and α5 regions, the conformations of the three central helixes (α2–α4) of RsaL are quite similar to these HTH domain structures (Supplementary Figure S7A). Similar to RsaL, MqsA interacts with DNA as a dimer (Supplementary Figure S7B). Oct1 and HNF6 function as monomers and in addition to their primary HTH domains, their additional HTH domains (HTH2) are also involved in target DNA recognition (Supplementary Figure S7C). The relative orientations of the HTH2 domains in the Oct1 and HNF6 structures are different from one another; however, the orientations of the HTH domains and their interactions with the target DNAs are conserved among all of the structures. Structure based sequence alignment showed that most of the DNA recognition residues are also conserved in these structures (Supplementary Figure S7D); the conservation of the residues and their interactions with DNA bases suggest that other HTH domain-containing proteins may also use the same strategy in target DNA recognition. In summary, our findings extend understanding of the functions of the QS repressor RsaL. In addition to being involved in regulating the 3OC 12 -HSL, RsaL also controls the activity of pqsH and cdpR , which are both required for PQS synthesis. Therefore, cross-regulation between lasI and pqsH/cdpR in the rsaL mutant maintains a balanced level of signal production in P. aeruginosa . Importantly, the major function of RsaL in P. aeruginosa physiology is to govern the homeostasis of 3OC 12 -HSL by controlling, together with LasR, the expression of rsaL and lasI ( 25 ). Moreover, RsaL is a global regulator that is an integral part of the QS signaling network, which controls gene expression through different mechanisms, including repression of 3OC 12 -HSL signal molecule production, activation of PQS synthesis, and direct binding of target genes (Figure 7 ). Therefore, the broad range of RsaL functions further elucidates the complexity of the QS network and the detailed characterization of the RsaL–DNA complex will provide new clues for understanding other QS regulators. Figure 7. Schematic model of RsaL involved in the QS regulatory cascade and virulence factor regulation of P. aeruginosa . The potential regulatory pathways and interplays of RsaL are based on our observations and those of previous studies. It has been demonstrated that RsaL, in concert with LasR, govern the homeostasis of 3OC 12 -HSL by controlling the expression of rsaL and lasI ( 25 ). In the present study, we show that RsaL binds directly to the intergenic region of pqsH / cdpR , thus controlling PQS production and P. aeruginosa virulence phenotypes. Solid arrows indicate positive regulation and solid T-bars represent negative regulation."
} | 3,080 |
35538126 | PMC9090930 | pmc | 3,498 | {
"abstract": "Recently, brain-inspired computing models have shown great potential to outperform today’s deep learning solutions in terms of robustness and energy efficiency. Particularly, Spiking Neural Networks (SNNs) and HyperDimensional Computing (HDC) have shown promising results in enabling efficient and robust cognitive learning. Despite the success, these two brain-inspired models have different strengths. While SNN mimics the physical properties of the human brain, HDC models the brain on a more abstract and functional level. Their design philosophies demonstrate complementary patterns that motivate their combination. With the help of the classical psychological model on memory, we propose SpikeHD, the first framework that fundamentally combines Spiking neural network and hyperdimensional computing. SpikeHD generates a scalable and strong cognitive learning system that better mimics brain functionality. SpikeHD exploits spiking neural networks to extract low-level features by preserving the spatial and temporal correlation of raw event-based spike data. Then, it utilizes HDC to operate over SNN output by mapping the signal into high-dimensional space, learning the abstract information, and classifying the data. Our extensive evaluation on a set of benchmark classification problems shows that SpikeHD provides the following benefit compared to SNN architecture: (1) significantly enhance learning capability by exploiting two-stage information processing, (2) enables substantial robustness to noise and failure, and (3) reduces the network size and required parameters to learn complex information.",
"introduction": "Introduction Many applications run machine learning algorithms to assimilate the data collected in the swarm of devices on the Internet of Things (IoT). Sending all the data to the cloud for processing is not scalable, cannot guarantee a real-time response. However, the high computational complexity and memory requirement of existing DNNs hinder usability to a wide variety of real-life embedded applications where the device resources and power budget is limited 1 – 4 . Therefore, we need alternative learning methods to train on the less-powerful IoT devices while ensuring robustness and generalization. System efficiency comes from sensing and data processing. Unlike classical vision systems, neuromorphic systems try to efficiently capture a notion of seeing motion 5 – 9 . Bio-inspired learning methods, i.e., spiking neural networks (SNNs), address issues related to energy efficiency 5 , 10 – 23 . SNNs have been widely used in many areas of learning and signal processing 24 – 27 . These systems have yet to provide robustness and intelligence that matches that from embodied human cognition. For example, the existing bio-inspired method cannot integrate sensory perceptions with actions. SNN applications in machine learning have largely been limited to very shallow neural network architectures for simple problems. Using deep SNN architecture often does not improve learning accuracy and can result in a possible training divergence 9 . In addition, SNNs lack brain-like robustness and cognitive support. On the other hand, Hyperdimensional Computing (HDC) is introduced as a promising brain-inspired solution for robust and efficient learning 28 . HDC is motivated by the understanding that the human brain operates on high-dimensional representations of data originated from the large size of brain circuits 29 . It thereby models the human memory using points of a high-dimensional space, that is, with hypervectors . HDC performs a learning task after mapping data into high-dimensional space. This encoding is performed using a set of pre-generated base vectors . HDC is well suited to address several learning tasks in IoT systems as: (i) HDC is computationally efficient and amenable to hardware level optimization 30 – 32 , (ii) it supports single-pass training with no back-propagation or gradient computation, (iii) HDC offers an intuitive and human-interpretable model 33 , (iv) it is a computational paradigm that can be applied to a wide range of learning and cognitive problems 33 – 45 , and (v) it provides strong robustness to noise—a key strength for IoT systems 46 . Despite the above-listed advantages, HDC encoding schemes are not designed for handling neuromorphic data. HDC lacks the behavioral resemblance to neurons to extract features from neuromorphic data effectively. While SNN mimics the physical properties of the brain (how biological neurons are operating), HDC models the brain at a more abstract and functional level. This makes these two computational models complementary. Inspired by the classical and popular memory model, introduced by Atkinson–Shiffrin 47 , we propose a novel framework that fundamentally combines Spiking neural network and hyperdimensional computing. Our framework, called \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD , enables a scalable and strong cognitive learning system to better mimic brain functionality. \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD creates a cross-layer brain-inspired system that captures information of sensory data from different perspectives: low-level neural activity and pattern-based neural representation. Since both SNN and HDC have memorization capability, they are powerful in preserving spatial and temporal information. Therefore, \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD can ensure advanced learning capability with high accuracy. To the best of our knowledge, \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD is the first framework that fundamentally combines SNN and HDC. \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD first exploits a few layers of spiking neural network to extract low-level spatiotemporal information of raw event-based data. Then, it utilizes HDC to operate over SNN output, learn the abstract information, and classifying the data. To ensure robust, efficient, and accurate HDC learning, we present a non-linear neural encoder that transforms data into knowledge at a very low cost and with comparable accuracy to state-of-the-art methods for diverse applications. We develop an end-to-end framework that enables co-training of SNN and HDC models. Instead of using deep SNN architecture, we exploit a simple SNN architecture that updates based on gradient rule and connects it to an HDC module capable of fast and single-pass learning. Our framework trains SNN and HDC models simultaneously to ensure that the data generated by SNN is optimal for HDC learning. \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD supports online learning from the data stream. In this configuration, we keep the SNN layer static while exploiting HDC single-pass training capability to update the model in real-time. This enables \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD model to learn or update its functionality with very few samples and without paying the cost of storing large-scale train data for iterative learning. We evaluate \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD on multiple classification problems. Our evaluation 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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD provides significant benefits compared to both HDC and SNN architectures: (1) enhance learning capability by exploiting two-stage information processing, and (2) significantly reduces the network size and required parameters to learn complex information. For example, our results indicate 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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD can provide 6.1% and 3.8% higher classification accuracy on MNIST and DVS Gesture datasets.",
"discussion": "Conclusion and discussion In this paper, we propose \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD , a novel framework that combines Spiking neural network and hyperdimensional computing in order to design a scalable and strong cognitive learning system that better mimics brain functionality. \\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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD exploits spiking neural networks to extract low-level features by preserving the spatial and temporal correlation of raw event-based spike data. Then, we utilize HDC to operate over SNN output by mapping the signal into high-dimensional space, learning the abstract information, and classifying the data. Our evaluation on a wide range of classification problems 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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD provides significant benefit compared to both HDC and SNN architecture: (1) enhances learning capability by exploiting two-stage information processing, (2) significantly reduces the network size and required parameters to learn complex information. For the rest of this section, we highlight some of the open challenges that our framework has yet to overcome and encourage exploration of the question along multiple axes (Fig. 7 ). Figure 7 Overview 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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD extended. Framework \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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD incorporates spiking neural networks and vector symbolic architecture at a high level. While our work has demonstrated one efficiency-oriented instance of the model with DECOLLE and HDC, other combinations may give rise to models of different strengths. Loss transfer one direction for future work is the back-propagation of loss from VSA. Better loss propagation to the SNN layer leads to more effective SNN training. Interpretability The decoding of HDC Memory and the interpretability of SNN lead to knowledge at the sensory data level. Loss backpropagation During step III 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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD , a Moore–Penrose inverse of the HDC encoder is applied to backpropagate the loss from the HDC module to the SNN. Since HDC encoder maps vectors to hypervectors, the rank of the inverse is limited to the output dimension of the SNN at the point of injection, which may be way less than the dimension. A large amount of information may be lost from this transition. We experimented with several methods to solve this problem. One example is to continue training the original SNN, transfer the new weights 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}$${\\mathsf {SpikeHD}}$$\\end{document} SpikeHD , and then train the HDC module. This did not improve performance except in the case of the transfer learning task, where the context of the data or the task changes. One method that may be suggested is to introduce a regularization term in the loss function of the SNN layers such that it outputs an HDC-like vector as the representation of the data directly. This will avoid the explicit usage of the HDC non-linear encoder, and the loss will be optimally propagated up to the approximation introduced by the regularization term. Component choices We have selected DECOLLE as SNN and HDC as VSA for our hybrid model for the reason we’ve discussed in section 2.1. It is optimized for time and energy efficiency, and practicality, as both models are known for such traits. Readers interested in the exploration of other aspects may choose to adopt our memory framework to other components. such as Legendre Memory Units 91 and HDC, or BI-SNN and HRR 92 . Concept interpretability Our current usage of the memory framework is to directly operate on long term memory and derives decisions from its representation, which simulates what the cerebellum does. For the purpose of completing the analogy to the Atkinson–Shiffrin memory model, it has yet to be incorporated the decoding mechanisms of the HDC memory: we did not fetch the long term memory back to the hippocampus and decode it for operations in working memory. This subject is not the purpose of this paper, and the decoupling of encoding and decoding invites more possibilities, as the HDC memory may be used as heterogeneous storage such that multiple tasks may be performed in one memory model. That said, the subject of interpretability remains interesting at two levels. Each entry in the HDC memory represents the concept of the class, which can be decoded to retrieve a representation of the concept at the SNN level. The representation in SNN can then be interpreted to infer knowledge on the original sensory data."
} | 4,156 |
25359576 | PMC4215353 | pmc | 3,500 | {
"abstract": "Evolution has produced some remarkable creatures, of which silk gland is a fascinating organ that exists in a variety of insects and almost half of the 34,000 spider species. The impressive ability to secrete huge amount of pure silk protein, and to store proteins at an extremely high concentration (up to 25%) make the silk gland of Bombyx mori hold great promise to be a cost-effective platform for production of recombinant proteins. However, the extremely low production yields of the numerous reported expression systems greatly hindered the exploration and application of silk gland bioreactors. Using customized zinc finger nucleases (ZFN), we successfully performed genome editing of Bmfib-H gene, which encodes the largest and most abundant silk protein, in B. mori with efficiency higher than any previously reported. The resulted Bmfib-H knocked-out B. mori showed a smaller and empty silk gland, abnormally developed posterior silk gland cells, an extremely thin cocoon that contain only sericin proteins, and a slightly heavier pupae. We also showed that removal of endogenous Bmfib-H protein could significantly increase the expression level of exogenous protein. Furthermore, we demonstrated that the bioreactor is suitable for large scale production of protein-based materials.",
"discussion": "Discussion In summary, we generated several novel B. mori mutants by targeted editing of BmFib-H gene using customized ZFN. The genome editing efficiency was higher than any previously reported 25 26 33 . In one of the mutant lines, knock-out of BmFib-H gene resulted in a smaller and empty silk gland, abnormally developed posterior silk gland cells, an extremely thin cocoon that contain only sericin proteins, and a slightly heavier pupae. Using an EGFP fusion protein as a reporter, we showed that the genome edited silk gland could be used as a highly efficient bioreactor for recombinant protein production, whose productivity is superior to any of the known expression systems. Furthermore, by designing and successful production of an artificial B. mori silk protein, we demonstrated that the bioreactor is suitable for large scale production of protein-based materials. Genetically modified organisms have been used as bioreactors since 1980s, during which increasing the expression yield and developing strategies for large scale production has been long-lasting challenge. Given the highly protein synthesis activity 30 , the high protein content (more than 95%) 4 , and the simplicity of the silk protein components (only 11 known major proteins), the silk gland of B. mori has held promise as a low-cost, high-yield bioreactor. And this promise was proved by successful production of numerous recombinant proteins 31 . The present silk gland bioreactor generated by genome editing of BmFib-H gene further increase the content of recombinant proteins to about 50% of total cocoons, placing this bioreactor superior to any of the known bioreactors. It is reasonable to assume that other proteins such as collagen, human serum albumin and monoclonal antibodies could be expressed in this superior bioreactor with a similar yield. Furthermore, B. mori has been used for production of silk for thousand of years and are now large-scaled producing in China, Indian, Brazil and some other developing countries every year. It is very easy to achieve industrial scale production of biopharmaceutical proteins, which are in increasing demand for both analytical and clinical applications 32 . According to Tomita et al.'s calculation 3 , production of 5 kg of recombinant proteins using our bioreactor only need about 10 kg of cocoon material, which can be achieved in a facility with a floor surface of about 5 m 2 and 1 worker. Besides the amazing recombinant production ability, we also showed that the superior bioreactor could be used to produce protein-based materials, which are also in great need for biomedical applications but lack of efficient production system. As silk gland is the only organ that can store proteins at an extremely high concentration without aggregation or denaturalization, we believe that the genome edited silk gland generated here could be a unique high efficient bioreactor for the production of protein-based materials. The mechanism of silk proteins secretion and transport has long been an interesting but unresolved matter. The H-L subunit was thought to be important for the solubility of fibroin during intracellular transport, secretion, and luminal transport through the silk gland 33 . Based on this knowledge, Fib-L and P25 proteins should be absent in the cocoons of the mutant line Fib-H-I1 , and recovered in the line Fib-H-I1-TR as the small version of Fib-H (art-Fib-H) would form a disulfide bond and interact with P25. However, our results ( Fig. 4 ) showed that both Fib-L and P25 proteins existed in the cocoons of Fib-H deficient mutant line, Fib-H-I1 and other lines including WT, WT/Fib-H-I1 and Fib-H-I1-TR. This result indicated that the H-L subunit is not necessary for the secretion of Fib-L. Given the phenotypic effect of Fib-L mutation observed in the Nd-sd mutant, we suspected that the H-L subunit is only necessary for the intracellular transport, secretion, and luminal transport of large fibroin proteins such as Fib-H. The small proteins such as Fib-L, P25, and some other silk proteins are able to secrete and transport in an independent manner. However, to completely reveal the mechanism, more investigations such as knock-out or down/up regulating Fib-L and P25 are required. The fast-changing development of genome editing technologies in recent years and their rapid applications in increasing amount of organisms have greatly reshaped the field of genetic manipulation, especially in those non-model organisms such as B. mori that targeted mutagenesis was not achievable. ZFNs, TALENs CRISPR/Cas9 system emerged subsequently and served as the troika for genome editing tools. Recently, we showed that TALEN and CRISPR/Cas9 could introduce highly efficient targeted mutagenesis, large chromosomal deletion and even sophisticated genomic structure variations in B. mori 26 34 35 36 . Together with the high efficiency of ZFN for B. mori genome editing proved here, it is feasible for researchers to implement precise and sophisticated manipulation of any chosen B. mori gene. Besides the largest and most abundant fib-H protein, silk gland also secretes another 2 fibroin proteins (Fib-L and P25) and 8 main sericin proteins (Ser1 variants, Ser2 and Ser3). The next step of silk gland genetic engineering would be removing these 10 silk proteins using genome editing tools, which in principle will generate a more superior bioreactor that produce industrial scale of pure (purification-free) recombinant biopharmaceutical proteins or protein-based materials. Given that ZFNs are much more difficult to be customized to new targets, TALENs and CRISPR/Cas9 system are preferred to be used to fulfill the mission."
} | 1,754 |
37476623 | null | s2 | 3,506 | {
"abstract": "For animals to navigate an uncertain world, their brains need to estimate uncertainty at the timescales of sensations and actions. Sampling-based algorithms afford a theoretically-grounded framework for probabilistic inference in neural circuits, but it remains unknown how one can implement fast sampling algorithms in biologically-plausible spiking networks. Here, we propose to leverage the population geometry, controlled by the neural code and the neural dynamics, to implement fast samplers in spiking neural networks. We first show that two classes of spiking samplers-efficient balanced spiking networks that simulate Langevin sampling, and networks with probabilistic spike rules that implement Metropolis-Hastings sampling-can be unified within a common framework. We then show that careful choice of population geometry, corresponding to the natural space of parameters, enables rapid inference of parameters drawn from strongly-correlated high-dimensional distributions in both networks. Our results suggest design principles for algorithms for sampling-based probabilistic inference in spiking neural networks, yielding potential inspiration for neuromorphic computing and testable predictions for neurobiology."
} | 306 |
32679577 | null | s2 | 3,507 | {
"abstract": "Recent progress in artificial intelligence is largely attributed to the rapid development of machine learning, especially in the algorithm and neural network models. However, it is the performance of the hardware, in particular the energy efficiency of a computing system that sets the fundamental limit of the capability of machine learning. Data-centric computing requires a revolution in hardware systems, since traditional digital computers based on transistors and the von Neumann architecture were not purposely designed for neuromorphic computing. A hardware platform based on emerging devices and new architecture is the hope for future computing with dramatically improved throughput and energy efficiency. Building such a system, nevertheless, faces a number of challenges, ranging from materials selection, device optimization, circuit fabrication and system integration, to name a few. The aim of this Roadmap is to present a snapshot of emerging hardware technologies that are potentially beneficial for machine learning, providing the Nanotechnology readers with a perspective of challenges and opportunities in this burgeoning field."
} | 287 |
34009951 | null | s2 | 3,508 | {
"abstract": "Engineered microbial communities show promise in a wide range of applications, including environmental remediation, microbiome engineering, and synthesis of fine chemicals. Here we present methods by which bacterial aggregates can be directed into several distinct architectures by inducible surface expression of heteroassociative protein domains (SpyTag/SpyCatcher and SynZip17/18). Programmed aggregation can be used to activate a quorum-sensing circuit, and aggregate size can be tuned "
} | 122 |
23298573 | PMC3560178 | pmc | 3,509 | {
"abstract": "Background Lignin materials are abundant and among the most important potential sources for biofuel production. Development of an efficient lignin degradation process has considerable potential for the production of a variety of chemicals, including bioethanol. However, lignin degradation using current methods is inefficient. Given their immense environmental adaptability and biochemical versatility, bacterial could be used as a valuable tool for the rapid degradation of lignin. Kraft lignin (KL) is a polymer by-product of the pulp and paper industry resulting from alkaline sulfide treatment of lignocellulose, and it has been widely used for lignin-related studies. Results Beta-proteobacterium Cupriavidus basilensis B-8 isolated from erosive bamboo slips displayed substantial KL degradation capability. With initial concentrations of 0.5–6 g L -1 , at least 31.3% KL could be degraded in 7 days. The maximum degradation rate was 44.4% at the initial concentration of 2 g L -1 . The optimum pH and temperature for KL degradation were 7.0 and 30°C, respectively. Manganese peroxidase (MnP) and laccase (Lac) demonstrated their greatest level of activity, 1685.3 U L -1 and 815.6 U L -1 , at the third and fourth days, respectively. Many small molecule intermediates were formed during the process of KL degradation, as determined using GC-MS analysis. In order to perform metabolic reconstruction of lignin degradation in this bacterium, a draft genome sequence for C. basilensis B-8 was generated. Genomic analysis focused on the catabolic potential of this bacterium against several lignin-derived compounds. These analyses together with sequence comparisons predicted the existence of three major metabolic pathways: β -ketoadipate, phenol degradation, and gentisate pathways. Conclusion These results confirmed the capability of C. basilensis B-8 to promote KL degradation. Whole genomic sequencing and systematic analysis of the C. basilensis B-8 genome identified degradation steps and intermediates from this bacterial-mediated KL degradation method. Our findings provide a theoretical basis for research into the mechanisms of lignin degradation as well as a practical basis for biofuel production using lignin materials.",
"conclusion": "Conclusions This study demonstrated that C. basilensis B-8 could be a useful tool for lignin utilization. The great capability for KL degradation by this strain was confirmed. The maximum degradation rate was 44.4% at the initial concentration of 2 g L -1 after 7 days of incubation. High activity of MnP and Lac as well as the presence of many intermediates was observed during the degradation progress. Comprehensive and systematic whole genomic analysis of bacterial lignin degradation pathways was also performed via sequencing and analysis of the C. basilensis B-8 genome, and many genes related to lignin degradation were identified. Three major pathways for lignin degradation were reconstructed via genomic analysis.",
"discussion": "Result and discussion Optimization of temperature and pH on KL degradation by C. basilensis B-8 According to the actual results, the temp optimum was 30°C and the pH optimum was between 7–8.5 (Figure 1 ), as presented in previous reports, e.g. the optimum pH of Aneurinibacillus aneurinilyticus is 7.6 [ 17 ] and the corresponding values for Comamonas sp. B-9 and Bacillus strain are 7 [ 18 ] and 7.6 [ 19 ], respectively. For Streptomyces strains, the optimal pH ranges from 7.8–8.5 [ 20 ]. Figure 1 Effect of temperature and pH on KL degradation by C. basilensis B-8. ( a ) temperature and ( b ) pH. Average values of three replicates are shown with the standard error of the mean as error bars. Bacterial growth and KL degradation The rate of C. basilensis B-8 growth was evaluated under seven different initial KL concentrations ranging from 0.5 g L -1 to 6 g L -1 . C. basilensis B-8 grew well under initial concentrations from 1 g L -1 to 6 g L -1 (Figure 2a ), indicating that bacterial growth would not be inhibited under the tested concentrations. However, the optical density (OD) value of the cultured sample increased with the increase in initial KL concentration. The rates of KL degradation under different initial concentrations all surpassed 31% on day 7 (Figure 2b ), but there was no obvious correlation between the initial concentration and the KL degradation rate. The highest KL degradation rate of 44.4% was observed at an initial concentration of 2 g L -1 , and the largest KL degradation capacity of 2.1 g L -1 was observed at the initial concentration of 6 g L -1 . Figure 2 Bacterial growth and KL degradation in different initial concentration of KL. ( a ) Bacterial growth in different initial concentration, ( b ) KL degradation rate at seventh day in different initial concentration, ( c ) Bacterial growth and COD reduction in 2 g · L -1 KL. Average values of three replicates are shown with the standard error of the mean as error bars. The growth and KL degradation capacity of C. basilensis B-8 in nutrient medium with a KL concentration of 2 g L -1 were investigated in detail. The results are shown in Figure 2c . C. basilensis B-8 growth was substantially faster during the first 2 days and reached the maximum at day 4. KL degradation mainly occurred during the initial 2 days. Accordingly, the maximum KL degradation rate of 722.8 mg L -1 Day -1 was recorded during this period. From the third day, KL degradation was continuous, but the degradation rate was decreased. The chemical oxygen demand (COD) value reached 1455.4 mg L -1 from the initial 3276.4 mg L -1 at day 7. The growth and KL degradation observed in this experiment were different from those of Citrobacter strains, which must initially use glucose and peptone as carbon sources and subsequently utilize lignin as a co-metabolite [ 21 ]. Accordingly, KL could be the sole nutrition source of C. basilensis B-8, as it must be metabolized during the initial growth stage to provide carbon and energy for growth. A similar process of KL degradation was also reported for Comamonas sp. B-9 [ 18 ] and Streptomyces viridosporus [ 11 ], which also showed a great capacity for KL degradation. This predicts that bacteria that use lignin as their sole carbon source must metabolize it throughout the whole life cycle; therefore, the efficiency and total amount of lignin degradation would be relatively higher for these strains than those using lignin as co-metabolite. Analysis of enzymes and genes related to KL degradation Three major enzymes including LiP, MnP, and Lac, which use low-molecular-weight mediators to carry out lignin degradation, have been well characterized in microorganism [ 22 ]. The activity of these three enzymes from C. basilensis B-8 is shown in Figure 3 . MnP activity increased significantly during the initial 3 days, with a maximum of 1685.3 U/L at day 3, followed by a slight decrease from day 4. Lac activity was maintained at a low level on day 1. A rapid increase was then observed from day 2, with a maximum of 815.6 U/L on day 4. These results indicated that MnP played a crucial role during the entire process of KL degradation by C. basilensis B-8, whereas Lac mainly functioned during the latter stages of the reaction. A similar conclusion was also proposed in previous reports [ 23 , 24 ]. However, the mechanism of lignin micro-biodegradation is complicated; thus, a complete explanation requires further study. In addition, no obvious LiP activity was observed during the course of KL degradation, indicating that active LiP was not produced by C. basilensis B-8. Similar to this strain, some white-rot fungi and bacteria (i.e., Dichomitus squalens , Lentinula edodes [ 10 ], and Comamonas sp. B-9 [ 18 ]), which simultaneously produce MnP and Lac, were reported not to secret detectable levels of LiP. These organisms are also strong lignin degraders. Since LiP is responsible for the oxidation of non-phenolic syringyl and biphenyl model compounds (which exist in certain types of lignins, like hardwood) and subsequent ring cleavage [ 10 ], it is conceivable that the efficiency of hardwood degradation by C. basilensis B-8 and the other microorganisms mentioned above are relatively low. Figure 3 The Activity of MnP and Lac during 7 days incubation. Average values of three replicates are shown with the standard error of the mean as error bars. No LiP activity was detected in this study, and it was not surprising that no gene encoding LiP was identified via genomic analysis of C. basilensis B-8 (see supplementary files). However, a genomic search for MnP genes in the genome of C. basilensis B-8 rendered only open reading frames (ORFs) with low amino acid (aa) identities of 27.0% and 30.0% with the MnP genes of the fungi Pleurotus ostreatus and Ganoderma australe , respectively. MnP gene retrieval using the NCBI public database showed that no MnP genes from bacteria were available; furthermore, very few reports detail the presence of MnP enzymes in bacterial systems [ 25 ]. Metabolite characterization via GC-MS The low molecular weight compounds released from lignin due to KL degradation by C. basilensis B-8 were analyzed by GC-MS. The total ion chromatograph (TIC) patterns corresponding to the compounds extracted with ethyl acetate from the control (uninoculated medium sample) and degraded samples are shown in Figure 4a -c, and their peak identity is depicted in Table 1 . In the TIC pattern of the control sample (Figure 4a ), peaks at RT 8.1 and RT 10.6 were identified as acetic acid and phenol, respectively. The identification of these two important intermediate metabolites generated during the degradation of lignin by microorganisms [ 3 , 26 ] may be attributed to the chemical oxidation of lignin due to aeration and agitation. Moreover, other lignin-related compounds were also indentified, suggesting partial degradation of KL during the industrial production process [ 27 ]. Figure 4 TIC of TMS derivatives of compounds extracted with trichloromethane from kraft lignin medium incubated with Cupriavidus sp. B-8 a : 0d; b : 3d; c : 6d. Table 1 Chromatographic peak identification of metabolic products from KL degradation No RT a Present in Compound Figure 4a (0d) Figure 4b (3d) Figure 4c (7d) 1 8.1 + - - Acetic acid 2 8.5 + - - Methyl acetate 3 10.5 - + - Ethanedioic acid 4 10.6 + - - Phenol 5 10.6 - - + 2,3-dihydro-3,5-dihydroxy-6-methyl-4-pyrone 6 11.8 + - - 3,5-Dimethyl-4-hydroxybenzaldehyde 7 13.1 - + - 2-methylnaphthalene 8 15.6 - - + Ethyl gallate 9 16.5 - - + 4-hydroxy-3-methoxyacetophenone 10 16.7 + - + Cinnamic acid 11 16.7 - + - 2,3-dihydro-5-methylfuran-2-one l 12 16.8 + + + 3,5-di-tert-butyl phenol 13 18.7 - - + Gentisate 14 21.3 - - + 4-hydroxy-3,5-dimethoxy acetophenone 15 21.7 - - + 3,4-dihydroxyphenylacetic acid 16 21.8 - + - 4-Hydroxycinnamic acid 17 22.4 - + - Methyl 3-(α,α,-dimethyl)propionate 18 24.4 - + - Octadecanal 19 25.6 + - + hexadecanoic acid 20 27.2 - - + 3,5-dihydroxybenzoic acid 21 29.8 - + - hexadecamide The number of peaks in the TIC increased significantly after 3 and 6 days of incubation with C. basilensis B-8 as compared to the control. Many low molecular weight compounds such as 2, 3-dihydro-3, 5-dihydroxy-6-methyl-4-pyrone; 3, 5-dimethylbenzaldehyde; 2-methylnaphthalene; cinnamic acid; and gentisate were identified in the extract of the degraded sample (Figure 4b -c and Table 1 ), and these were not present in the extract of the control sample. The detected guaiacol-related compound and cinnamic acid could be easily related to the oxidation of the guaiacyl units from precursor coniferyl alcohol and ρ-hydroxyphenyl units generated from precursor ρ-coumary alcohol. These are considered to be basic moieties with syringyl units from precursor sinapyl alcohol that are components of the lignin structure [ 3 ]. Unfortunately, no syringyl-related compounds were identified in the sample. In addition to aromatic compounds, more acid-type compounds were identified than aldehyde and ketone-type compounds due to degradation of lignin. A similar study was also performed previously in Aneurinibacillus aneurinilyticus [ 17 ]. The low molecular weight compounds identified in the extracts of the inoculated sample favor the conclusion that KL was degraded by C. basilensis B-8. The initial degradation of lignin into low molecular weight compounds by extracellular phenoloxidases The extracellular oxidative enzymes LiP, MnP, and Lac are defined as phenoloxidases, which are responsible for the initial degradation of lignin and have been intensively studied in fungi. Although enzymology of bacterial lignin degradation has not been as thoroughly investigated as that of fungi, there are indications that bacteria use extracellular peroxidase for lignin degradation [ 11 ]. The activities of MnP and Lac from C. basilensis B-8 have been observed using colorimetric enzyme assays (Section “Analysis of enzymes related to KL degradation”). These observations indicted the existence of a novel MnP or its isozyme in C. basilensis B-8. The further study, purification, and characterization of these enzymes is currently under way. High MnP activity has also been documented in a previous report investigating Citrobacter strains [ 21 ] that need glucose as an extra carbon source to produce hydrogen peroxide, which serves as a cosubstrate for the ligninolytic activity of MnPs via glucose oxidation. However, glucose was not involved in the lignin degradation by C. basilensis B-8; in addition, the genes encoding glyoxal oxidase and aryl alcohol oxidase that are responsible for hydrogen peroxide production in fungi were not found in C. basilensis B-8, suggesting hydrogen peroxide in C. basilensis B-8 is generated via other unknown mechanisms. The genomic search of C. basilensis B-8 for a Lac gene only indicated one ORF with low aa identity (32.0%) with the LAC gene from Thermus thermophilus HB27. Given that some dye-type peroxidases, which are active against KL and lignin model compounds, have been identified in several bacteria [ 3 ], it is reasonable to predict that the detected enzymes form C. basilensis B-8 may belong to this group. Catabolism pathways for lignin components Many low molecular weight compounds were produced from initial KL degradation. The enzymes that are involved in catabolic pathways for the degradation of lignin fragments have been identified and characterized in several bacterial [ 28 - 30 ]. Here, three important degradation pathways for lignin basic derivatives, including coumaric acid, ferulic acid, cinnamic acid, phenol, salicylate, and 3-hydroxybenzoate, were predicted basing on genomic analysis. Moreover, all of these compounds were shown to support the growth of C. basilensis B-8. The β -ketoadipate central pathway for coumarate, ferulate, and cinnamate degradation The central reactions of the β -ketoadipate pathway in C. basilensis B-8 are shown in Figure 5 . The two branches of the β- ketoadipate pathway (i.e. the catechol branch encoded by cat genes and the protocatechuate branch encoded by pca genes) can convert catechol and protocatechuate into the Krebs cycle intermediates succinate and acetyl coenzyme A [ 31 ]. Biochemical studies and amino acid sequence data indicated that the enzymes of this pathway are highly conserved among phylogenetically diverse organisms that possess this pathway. Figure 5 Catabolic pathways for the catabolism of lignin and its derivatives in C. basilensis B-8. ? means that the enzyme encoding such biochemical step is still unknown. The cat and pca gene products of C. basilensis B-8 were significantly similar to proteins of known function from other bacteria, mainly Cupriavidus basilensis OR16 (Table 2 ), which has been reported earlier [ 32 ]. However, unlike many other bacteria whose cat and ben genes are usually organized in a single cluster, these genes in C. basilensis B-8 were organized in three clusters at different positions (see alignment result in supplementary files). The catA gene encoding catechol-1, 2-dioxygenase, which starts the catechol branch of the β- ketoadipate pathway, clustered together with the benABCD genes that encode the enzymes responcibe for converting benzoate into catechol. This operon was regulated by CatR, a member of the LysR family of regulatory proteins in P. pudita [ 33 , 34 ]. There are two copies of the catC gene that shared 71.0% aa identity within the genome of C. basilensis B-8; one of these genes was located upstream of the catB gene and the other was located the downstream of the catD gene. Table 2 Genes encoding the β-ketoadipate pathway and peripheral reactions Gene (orf no.) 1 Gene product Size (aa) 2 Function Organism % Identity/aa Accession no. catA (002058) CatA 304 Catechol 1,2-dioxygenase Lutiella nitroferrum 2002 80/304 ZP_03698425.1 catB (007569) CatB 142 Muconate cycloisomerase Cupriavidus basilensis OR16 91/382 ZP_09625347.1 catC1 (007567) CatC1 92 Muconolactone δ -isomerase Cupriavidus basilensis OR16 93/92 ZP_09625348.1 catC2 (006855) CatC2 91 Muconolactone δ -isomerase Alicycliphilus denitrificans BC 79/92 YP_004127266.1 catD (006858) CatD 273 3-oxoadipate enol-lactonase Acidovorax avenae subsp. avenae ATCC 19860 68/273 YP_004236298.1 catR (002057) CatR 307 Transcriptional activator (LysR family) Burkholderia cenocepacia J2315 83/307 YP_002233374.1 benA (002059) BenA 462 Benzoate dioxygenase large subunit Cupriavidus basilensis OR16 96/462 ZP_09623640.1 benB (002060) BenB 166 Benzoate dioxygenase small subunit Cupriavidus basilensis OR16 93/166 ZP_09623641.1 benC (002061) BenC 339 Benzoate dioxygenase reductase subunit Cupriavidus metallidurans CH34 77/339 YP_587015.1 benD (002063) BenD 108 2-Hydro-1,2-dihydroxybenzoate dehydrogenase Ralstonia eutropha JMP134 91/108 YP_298602.1 pobA (007971) PobA2 389 4-Hydroxybenzoate 3-Monooxygenase Cupriavidus basilensis OR16 92/389 ZP_09625369.1 pobB (001362) PobA1 394 4-Hydroxybenzoate 3-Monooxygenase Pseudomonas putida W619 92/394 YP_001748818.1 pcaB (001344) PcaB 453 3-Carboxy-cis,cis-muconate cycloisomerase Streptomyces hygroscopicus ATCC 53653 70/453 ZP_07293737.1 pcaC1 (003089) PcaC1 125 4-Carboxymuconolactone decarboxylase Pseudomonas fulva 12-X 82/125 YP_004473183.1 pcaC2 (005251) PcaC2 139 4-Carboxymuconolactone decarboxylase Cupriavidus basilensis OR16 90/139 ZP_09624697.1 pcaD1 (000893) PcaD1 280 3-Oxoadipate enol-lactonase Cupriavidus basilensis OR16 86/280 ZP_09627578.1 pcaD2 (006858) PcaD2 273 3-Oxoadipate enol-lactonase Acidovorax avenae subsp. avenae ATCC 19860 68/273 YP_004236298.1 pcaF (003109) PcaF 400 β -Ketoadipyl CoA thiolase Cupriavidus basilensis OR16 98/400 ZP_09627621.1 pcaG (002623) PcaG 192 Protocatechuate 3,4-dioxygenase, α subunit Cupriavidus basilensis OR16 88/192 ZP_09628855.1 pcaH (002624) PcaH 241 Protocatechuate 3,4-dioxygenase, β subunit Cupriavidus basilensis OR16 93/241 ZP_09628856.1 pcaI1 (003111) PcaI1 227 β -Ketoadipate succinyl-CoA transferase α subunit Ralstonia eutropha H16 93/227 YP_728364.1 pcaI2 (000701) PcaI2 252 β -Ketoadipate succinyl-CoA transferase α subunit Burkholderia xenovorans LB400 75/252 YP_554660.1 pcaI3 (003203) PcaI3 251 β -Ketoadipate succinyl-CoA transferase α subunit Bordetella petrii DSM 12804 57/251 YP_001632201.1 pcaI4 (002146) PcaI4 233 β -Ketoadipate succinyl-CoA transferase α subunit Cupriavidus basilensis OR16 95/233 ZP_09628963.1 pcaJ1 (003112) PcaJ1 215 β -Ketoadipate succinyl-CoA transferase β subunit Cupriavidus basilensis OR16 95/215 ZP_09627620.1 pcaJ2 (000700) PcaJ2 220 β -Ketoadipate succinyl-CoA transferase β subunit Bordetella petrii DSM 12804 68/220 YP_001633147.1 pcaJ3 (003202) PcaJ3 211 β -Ketoadipate succinyl-CoA transferase β subunit Cupriavidus basilensis OR16 91/211 ZP_09626665.1 pcaJ4 (002145) PcaJ4 205 β -Ketoadipate succinyl-CoA transferase β subunit Cupriavidus basilensis OR16 99/205 ZP_09628964.1 pcaK (003417) PcaK 446 4-Hydroxybenzoate transporter Cupriavidus basilensis OR16 93/446 ZP_09623076.1 pcaL (006326) PcaL 399 3-Oxoadipate enol-lactone hydrolase/4-carboxymuconolactone decarboxylase Ralstonia eutropha H16 74/399 YP_841800.1 pcaQ (002625) PcaQ 266 Transcriptional regulator (LysR family) Cupriavidus basilensis OR16 77/266 ZP_09628857.1 pcaT (001951) PcaT 436 β -Ketoadipate transporter Pseudomonas syringae pv. phaseolicola 1448A 78/436 YP_276146.1 hcaA (005714) HcaA 280 ρ -Hydroxycinnamoyl-CoA hydratase/lyase Burkholderia glumae BGR1 93/280 YP_002908324.1 hcaB (005963) HcaB 397 Vanillin dehydrogenase Burkholderia thailandensis MSMB43 86/397 ZP_02469116.1 hcaC (001154) HcaC 620 Feruloyl-CoA synthase Azotobacter vinelandii DJ 56/620 YP_002801316.1 hcaD1 (003030) HcaD1 395 Acyl-CoA dehydrogenase Cupriavidus basilensis OR16 95/395 ZP_09626604.1 hcaD2 (003645) HcaD2 383 Acyl-CoA dehydrogenase Cupriavidus basilensis OR16 95/383 ZP_09624899.1 vanA1 (007066) VanA1 351 Vanillate-O-demethylase oxygenase subunit Ralstonia solanacearum MolK2 80/351 CAQ18072.1 vanA2 (001443) VanA2 340 Vanillate-O-demethylase oxygenase subunit Cupriavidus metallidurans CH34 38/340 YP_587674.1 vanB1 (007067) VanB1 315 Vanillate-O-demethylase reductase subunit Pseudomonas syringae pv. tomato str. DC3000 56/315 NP_792703.1 vanB2 (001445) VanB2 315 Vanillate-O-demethylase reductase subunit Halomonas elongata DSM 2581 55/315 YP_003899010.1 vanR (7065) VanR 258 Transcriptional regulator (GntR family) Cupriavidus basilensis OR16 92/258 ZP_09625218.1 hcaK (002970) HcaK 371 Hydroxycinnamate transporter Ralstonia eutropha JMP134 84/371 YP_299063.1 hcaR (002968) HcaR 164 Transcriptional activator (MarR family) Ralstonia eutropha JMP134 75/164 YP_299066.1 hcaX (002969) HcaX 404 Porin transmembrane protein Ralstonia eutropha JMP134 88/404 YP_299064.1 a. Indicates the open reading frame number in the Additional file 1 . b. aa, number of amino acids. The pca genes that are responsible for the protocatechuate branch of the β- ketoadipate pathway were dispersed throughout the genome of C. basilensis B-8 (Table 2 ), and this organization has not been previously reported for other bacteria. The pcaIJF genes encoding the enzymes required for the last two steps of the β- ketoadipate pathway clustered together. In some bacteria, the expression of the pcaIJF genes is induced by β- ketoadipate, which then activates IcIR family regulatory proteins PcaR/PcaQ. However, pcaR was found in another gene cluster and pcaQ was located in the vicinity of pcaH in C. basilensis B-8. It is not uncommon that transcriptional regulators control the expression of distal genes [ 29 ], though the exact mechanism of regulation for these genes requires further study. It was surprising that four copies of the pcaI and pcaJ genes were found in C. basilensis B-8 (Table 2 ), and the reason for this is still unknown. One striking aspect of the pca genes in C. basilensis B-8 is the presence of two copies of the pcaC and pcaD genes (which encode g-carboxymuconolactone decarboxylase and β -ketoadipate enol-lactone hydrolase, respectively) as well as a unique fused gene ( pcaL ) consisting of the pcaC and pcaD ORFs (Table 2 ). Sequence analysis of the pcaL gene revealed that the predicted C-terminal third was homologous to decarboxylases, whereas the N-terminal two thirds were homologous to enol-lactone hydrolases. Furthermore, DNA sequence data has revealed a remarkable feature of catD and pacD that encode the isozyme involved in the third from last step of the C. basilensis B-8 β -ketoadipate pathway (Figure 5 ). These two genes share only a 33% aa identity. Another striking aspect of the protocatechuate branch genes in C. basilensis B-8 is the presence of two 4-hydroxybenzoate 3-monooxygenase encoding genes, pobA and pobB . These two genes share 63.4% aa identity, indicating a distant evolutionary origin. Ferulate and coumarate form a vast array of ether and ester bonds in lignin and suberin. In some bacteria, ferulate degradation follows a CoA-dependent non- β –oxidative pathway catalyzed by the feruloyl-CoA synthetase (HcaC) and enoyl-CoA hydratase/aldolase (HcaA) proteins, producing vanillin. Vanillin is further converted to protocatechuate via an aldehyde dehydrogenase (HcaB) and a demethylase (VanAB) [ 29 ]. The pathway for coumarate degradation into protocatechuate is similar to that of ferulate, which is conducted by PobA and/or PobB at the last step (Figure 5 ). Genes homologous to hcaABCDKXR have been identified in C. basilensis B-8. The hcaABCD genes are dispersed throughout the genome of C. basilensis B-8, whereas the hcaKXR genes are clustered together (Table 2 ). hcaD encoding an acyl-CoA dehydrogenase is not involved in the above biochemical pathway, but the protein could be responsible for a CoA-dependent β -oxidative pathway of ferulate degradation with HcaA and Aat ( β -ketothiolase), as has been described in other organisms [ 28 ]. hcaR , hcaK , and hcaX encode a putative regulatory protein of the MarR family, a 3-hydroxyphenylpropionic acid transporter, and a putative porin of unknown function, respectively. Two van gene clusters were also identified in the genome of C. basilensis B-8 (Table 2 ), and one of them contained a transcriptional regulator of the GntR family (vanR). The hca gene cluster was not linked to the van cluster. A similar situation is also found in P. putida KT2440 and Acinetobacter sp. ADP1, and it has been suggested that this gene organization would facilitate the appearance of spontaneous van-deficient strains in natural Acinetobacter populations, which might allow the production of vanillate from ferulate as a chemical signal between plants and bacteria [ 35 ]. Cinnamate degradation via benzoate has been described in Cupriavidus necator JMP134 [ 29 ]. However, the enzymes that are responsible for the initial steps have not been characterized. Many similar intermediates including benzoate, produced during the process of cinnamate degradation by C. basilensis B-8 were observed on basis of our GC-MS analysis (date not shown). Accordingly, we could predict that cinnamate was also degraded through benzoate (Figure 5 ). The study of the enzymes involved in the first five steps of this pathway is currently in progress. Phenol degradation pathways Bacterial catabolic pathways for phenol and its derivatives have been studied extensively in C . necator JMP134 [ 29 ]. Genomic analysis of C. basilensis B-8 showed that all orthologous genes were present except phlX , which encodes a relatively hydrophobic protein. Phenol and its derivatives metabolized via the methylcatechol ortho ring-cleavage pathway (enzymes encoded by mml genes) and the catechol meta ring-cleavage pathway (enzymes encoded by phl genes) in C. basilensis B-8 are shown in Figure 5 , and the involved enzymes are listed in Table 3 . The mml genes organized in a single cluster. The mmlJIHGFRL is maintained in the mml clusters of C. basilensis B-8 (Table 3 ) and C . necator JMP134. Similar to C . necator JMP134, no putative gene encoding an isoenzyme of β -ketoadipate enollactone hydrolase was found in the mml gene cluster of C. basilensis B-8. This point supports the idea that 4-methyl- β -ketoadipate enollactone is not further metabolized through a classical β -ketoadipate pathway. The next step in the reaction process still requires further study. Table 3 Genes encoding catabolic pathways for phenols and its derivatives degradation Gene (orf no.) 1 Gene product Size (aa) 2 Function Organism % Identity/aa Accession no. phlB (007044) PhlB 310 Catechol-2,3-dioxygenase Ralstonia sp. KN1 97/310 BAA84125.1 phlC (007045) PhlC 484 2-Hydroxymuconic semialdehyde Cupriavidus basilensis OR16 96/484 ZP_09625234.1 phlD1 (001164) PhlD1 279 2-Hydroxymuconic semialdehyde hydrolase Azotobacter vinelandii DJ 72/279 YP_002801325.1 phlD2 (002956) PhlD2 291 2-Hydroxymuconic semialdehyde hydrolase Burkholderia thailandensis MSMB43 69/291 ZP_02463428.1 phlE (007046) PhlE 260 2-Hydroxypent-2,4-dienoate hydratase Ralstonia eutropha H16 91/260 YP_728710.1 phlF (008422) PhlF 313 Acetaldehyde dehydrogenase (acylating) Pseudomonas resinovorans 82/313 NP_758577.1 phlG (008421) PhlG 335 4-Hydroxy-2-oxovalerate aldolase Dechloromonas aromatica RCB 82/335 YP_286438.1 phlH (007047) PhlH 262 4-Oxalocrotonate decarboxylase Cupriavidus basilensis OR16 95/262 ZP_09625232.1 phlI (007048) PhlI 63 4-Oxalocrotonate isomerase Cupriavidus basilensis OR16 90/63 ZP_09625231.1 phlK (007037) PhlK 72 Phenol hydroxylase subunit Ralstonia sp. KN1 93/72 BAA84118.1 phlL (007038) PhlL 331 Phenol hydroxylase subunit Ralstonia sp. KN1 95/331 BAA84119.1 phlM (007039) PhlM 94 Phenol hydroxylase subunit Cupriavidus basilensis OR16 95/94 ZP_09625240.1 phlN (007040) PhlN 504 Phenol hydroxylase subunit Ralstonia sp. KN1 97/504 BAA84121.1 phlO (007041) PhlO 119 Phenol hydroxylase subunit Ralstonia sp. KN1 90/119 BAA84122.1 phlP (007042) PhlP 355 Phenol hydroxylase subunit Cupriavidus basilensis OR16 93/355 ZP_09625237.1 phlQ (007043) PhlQ 111 Ferredoxin Ralstonia sp. KN1 89/111 BAA84124.1 phlR (007035) PhlR 571 Phenol hydroxylase regulator protein Cupriavidus basilensis OR16 95/571 ZP_09625495.1 mmlF (000701) MmlF 252 Oxoadipate-CoA Transferase α subunit Ralstonia eutropha H16 88/252 NP_943023.1 mmlG (000700) MmlG 220 Oxoadipate-CoA Transferase β subunit Cupriavidus necator N-1 91/220 YP_004687736.1 mmlH (000699) MmlH 428 Muconolactone transporter Ralstonia eutropha JMP13 86/428 YP_295715.1 mmlI (000698) MmlI 112 4-Methylmuconolactone methylisomerase Ralstonia eutropha H16 83/112 NP_943020.1 mmlJ (000697) MmlJ 87 Methylmuconolactone isomerase Cupriavidus necator N-1 83/87 YP_004687733.1 mmlL (000704) MmlL 296 Hypothetical protein (Zn-dependent hydrolases) Ralstonia eutropha H16 16/296 NP_943025.1 mmlR (000703) MmlR 304 Transcriptional activator (LysR family) Ralstonia eutropha H16 91/304 NP_943024.1 a. Indicates the open reading frame number in the Additional file 1 . b. aa, number of amino acids. Although the mml genes are clustered together, the phl genes were organized in two different clusters. phlGF encoding the 4-hydroxy-2-ketovalerate aldolase and the aldehyde dehydrogenase that catalyzes the final steps of the meta ring-cleavage pathway (Figure 5 ) are separated from the rest of the other meta ring-cleavage pathway genes. A similar arrangement is also present in C . necator JMP134, though one difference is that two copies of phlD (which encodes a 2-hydroxymuconic semialdehyde hydrolase) were not found in either of these two gene clusters (Table 3 ). The gentisate pathway for catabolism of salicylate and 3-hydroxybenzoate Salicylate is generated from benzoic acid hydroxylation or trans-cinnamic acid side chain β -oxidation in plants. The catabolism of salicylate into catechol by salicylate 1-hydroxylase, a flavoprotein monooxygenase, or by a three-component protein has been described in several bacteria [ 29 ]. Genetic analysis of the C. basilensis B-8 genome showed seven genes (data not shown) with high identity to those from Pseudomonas and Acinetobacter strains. An alternative route of salicylate degradation, via a gentisate intermediate, is initiated by a multicomponent oxygenase (salicylate-5-hydroxylase), as has been reported in Ralstonia sp. U2 [ 36 ]. The putative LysR-type transcriptional regulator encoding the hybR gene, large and small subunits of the oxygenase encoding hybB and hybC genes, as well as the ferredoxin encoding hybD gene were organized in a single gene cluster (Table 4 ), which shows a significant similarity with that from C. necator JMP134. Further study is required to develope our understanding of how salicylate is converted into catechol in C. basilensis B-8. Table 4 Genes encoding gentisate pathways and peripheral reactions Gene (orf no.) 1 Gene product Size (aa) 2 Function Organism % Identity/aa Accession no. hybR (000498) HybR 306 LysR-like regulator protein Ralstonia solanacearum UW551 80/306 ZP_00946271.1 hybA (000499) HybA 328 Putative ferredoxin oxidoreductase Ralstonia solanacearum Po82 70/328 YP_006030105.1 hybB (000500) HybB 418 Salicylate-5-hydroxylase large oxygenase component Ralstonia solanacearum UW551 83/418 ZP_00946273.1 hybC (000501) HybC 157 Salicylate-5-hydroxylase small oxygenase component Ralstonia solanacearum MolK2 72/157 CAQ37385.1 hybD (000502) HybD 103 Salicylate 5-hydroxylase ferredoxin component Achromobacter xylosoxidans A8 66/103 YP_195869.1 mhbR (003412) MhbR 319 Transcriptional regulator (LysR family) Cupriavidus necator N-1 77/316 YP_004681033.1 mhbD (003413) MhbD 348 Gentisate 1,2-dioxygenase Cupriavidus basilensis OR16 95/348 ZP_09623073.1 mhbH (003414) MhbH 232 Fumarylacetoacetate hydrolase Cupriavidus basilensis OR16 91/232 ZP_09623074.1 mhbI (003415) MhbI 214 Maleylpyruvate isomerase Ralstonia eutropha H16 76/214 YP_729032.1 mhbM (003416) MhbM 413 3-Hydroxybenzoate-6-hydroxylase Burkholderia multivorans CGD1 82/413 ZP_03585707.1 mhbT (003417) MhbT 446 Putative 3-hydroxybenzoate transporter Burkholderia sp. NCIMB 10467 59/446 ABW22835.1 a. Indicates the open reading frame number in the Additional file 1 . b. aa, number of amino acids. 3-hydroxybenzoate is degraded through gentisate by 3-hydroxybenzoate-6-hydroxylase (3H6H) or through protocatechuate by 3-hydroxybenzoate-4-hydroxylase (3H4H) in Comamonas testosteroni [ 37 ] and Bacillus sp. [ 38 ]. No homologue of the 3H4H gene was found in the genome of C. basilensis B-8. However, a gene ( mhbM ) was identified with 82.0% aa identity with that of the 3H6H gene sequence from Burkholderia multivorans CGD1 (Table 4 ). The gentisate pathway (Figure 5 ) is initiated by a gentisate-1, 2-dioxygenase (MhbD), which cleaves the aromatic ring to form maleylpyruvate. Maleylpyruvate could be further degraded by maleylpyruvate hydrolase or by maleylpyruvate isomerase, central metabolites of the Krebs cycle. Genomic analysis indicated the presence of an mhb gene cluster in the genome of C. basilensis B-8. The mhbR and mhbT genes were located upstream of the mhb gene cluster and encoded a LysR family regulator protein and 3-hydroxybenzoate transporter, respectively (Table 4 ). mhbI and mhbH are homologous to genes that encode maleylpyruvate isomerase and fumarylpyruvate hydrolase in C. necator JMP134 [ 29 ], respectively. Together, these observations suggested that gentisate is metabolized by a glutathione-dependent pathway in C. basilensis B-8, as it is in C. necator JMP134."
} | 8,737 |
31048697 | PMC6497657 | pmc | 3,511 | {
"abstract": "The ratio of syringyl (S) and guaiacyl (G) units in lignin has been regarded as a major factor in determining the maximum monomer yield from lignin depolymerization. This limit arises from the notion that G units are prone to C-C bond formation during lignin biosynthesis, resulting in less ether linkages that generate monomers. This study uses reductive catalytic fractionation (RCF) in flow-through reactors as an analytical tool to depolymerize lignin in poplar with naturally varying S/G ratios, and directly challenges the common conception that the S/G ratio predicts monomer yields. Rather, this work suggests that the plant controls C-O and C-C bond content by regulating monomer transport during lignin biosynthesis. Overall, our results indicate that additional factors beyond the monomeric composition of native lignin are important in developing a fundamental understanding of lignin biosynthesis.",
"introduction": "Introduction The success of second-generation biorefineries hinges on the effective removal and utilization of the lignin fraction of biomass (ranging from 12 to 32 wt% depending on the plant species) 1 . Lignin is a poly-aromatic polymer contained in the cell wall of the plant, which provides structural stability, aids in water transport, and assists in preventing microbial attack of plant cells 2 , 3 . Accordingly, lignin contributes to the overall recalcitrance of biomass and must be separated before carbohydrates can be successfully and selectively converted into fuels and chemicals 3 – 6 . Lignin is created by the polymerization of three monomers, sinapyl alcohol, coniferyl alcohol, and p -coumaryl alcohol, which are synthesized from phenylalanyl and tyrosine in the cytoplasm 7 , 8 . The plant transports monomers to the cell wall where they undergo free radical coupling reactions creating a variety of C–O and C–C linkages (Fig. 1 ). This polymerization is mediated by peroxidase and laccase enzymes that form radicals on the phenolic group. By resonance, the radical is shared by the 5, 1, and β position of the monomer 9 and coupling reactions at any of these positions lead to polymers linked via C–O bonds (β-O-4) and C–C bonds (β-5, 5-5, β-1, and β-β). The β-O-4 bonds are the most abundant and, due to their labile nature, are key for the depolymerization of lignin. The generation of one monomer unit in depolymerization requires cleavage of two β-O-4 bonds, one at each side of the aromatic unit. Therefore, a small number of C–C linkages in the lignin structure can reduce the maximum theoretical monomer content. For example, in the lignin polymer shown in Fig. 1 , a β-O-4 content of 69% only yields a maximum monomer yield of 36%. Monomeric units from lignin depolymerization have been shown to be highly valuable as they offer a diverse platform to synthesize chemicals 10 – 20 and functional replacements for conventional polymers 21 – 25 . Fig. 1 Lignin structure overview. The two primary monomers, sinapyl alcohol and coniferyl alcohol, shown with the different reacting carbons highlighted with the appropriate C–O or C–C bonds that can form. A hypothetical lignin structure is shown with each type of bond as well as the calculated S/G ratio, β-O-4 content, and monomer yield Influencing lignin biosynthesis to favor the production of sinapyl alcohol (S-unit) relative to coniferyl alcohol (G-unit) is hypothesized to increase the β-O-4 content in lignin. Sinapyl alcohol has a methoxy group at the 5 position of the aromatic ring, thus preventing the formation of β-5 and 5-5 C–C linkages. Indeed, a higher S/G ratio in lignin has been shown to produce higher monomer yields using reductive catalytic fractionation (RCF) as a depolymerization method, as can be seen in Supplementary Fig. 1 , where we compiled a variety of monomer yields found in the literature across a vast range of S/G ratios from different natural and genetically modified feedstocks. Van den Bosch et al. 26 showed that birch lignin (S/G = 3) subjected to RCF at 523 K produced a monomer yield of 50 C-mol%, while poplar (S/G = 1.5) and a softwood (S/G = 0.05) produced yields of 44 and 21 C-mol%, respectively. Similarly, reports on genetically modified poplar with high or low S/G ratios showed a slight correlation between S/G and monomer yields. Shuai et al. showed that genetically modified poplar (S/G = 58) 27 produced a monomer yield of 78 wt% when depolymerized using RCF 28 . Interestingly, Parsell et al. 29 observed a lower monomer yield of 36 wt% for an F5H-modified poplar with an S/G of 2.7 and Luo et al. 30 observed a yield of 32.5 wt% for a genetically modified low-S poplar (S/G = 0.51). Despite seeing trends across both different species and genetically modified poplar, it is difficult to isolate the effect of S/G ratio on monomer yields from the effects resulting from plant genotype and genetic engineering. We reasoned that the effect of S/G ratio within natural variants of poplar would allow us to better isolate the effect of S/G ratio from other factors. Although many active stabilization methods have been developed to extract and simultaneously depolymerize lignin into stabile aromatic units 6 , 31 , 32 , RCF is effective at achieving near-theoretical lignin monomer yields from β-O-4 bond cleavage. It works through a solvolytic extraction of biomass followed by reductive cleavage of ether linkages in lignin over a redox active catalyst 33 . Typically, RCF is performed in a polar protic solvent 34 , 35 with hydrogen gas or a hydrogen donor 36 – 38 as a reductant and either Ru 26 , Pd 39 , or Ni 36 , 40 , 41 catalysts at 180–250 °C. RCF depolymerizes lignin by selectively cleaving all β-O-4 linkages within the lignin polymer to produce a stable mixture of monomeric and oligomeric alkyl phenols while preserving carbohydrates as a solid 37 , 42 , 43 . Therefore, the distribution of monomeric and oligomeric phenols can be easily mapped to the native lignin structure of the plant. In this study, we use RCF to investigate the impact of the S/G ratio on the production of monomers using natural poplar variants with S/G contents ranging from 1.41 to 3.60, or percent S content of 58.5% to 78.3% (Table 1 ) 44 . These five samples capture the range of S/G ratios present in a naturally variant population of over 1000 poplar trees 45 . RCF experiments were performed in flow-through reactors to obtain time- and composition-resolved extraction profiles. Additionally, batch experiments were performed at near complete lignin extraction. The oligomeric fractions were analyzed by heteronuclear single quantum coherence (HSQC) NMR spectroscopy to obtain the time-averaged S/G ratio. Additionally, the oligomers were derivatized by silylation and analyzed by gas chromatography-mass spectrometry (GC-MS) to obtain a qualitative distribution of C–C linkages within the dimeric fraction. These time-resolved data on both the monomeric and oligomeric fractions led to insights on the C–C bonding patterns of S-units that result in a decreased dependency of monomer yields on the S/G ratio than was previously hypothesized in the literature 6 . Table 1 Compositional analysis of poplar natural genetic variants S/G % S % Lignin % Glucan % Xylan % Galactan % Arabinan % Mannan % Acetyl % Total 1.41 58.5 27.1 39.4 16.8 1.7 0 5.2 4.2 96.7 1.69 62.8 25.1 42.4 14.6 1.8 0 4.4 3.7 95.1 2.35 70.1 23.9 44.7 15.0 1.9 0 5.3 3.9 96.8 3.48 77.7 25.0 39.7 17.3 1.7 0 5.2 4.1 95.9 3.60 78.3 22.6 42.7 17.2 1.8 0 5.1 4.0 96.0",
"discussion": "Discussion The differences in monomer yields and lignin composition at different extraction times was unexpected. The initial surge of monomers might be linked to diffusion limitations associated with varying lignin fragment chain lengths. Specifically, the delayed release of larger lignin fragments could be due to incomplete depolymerization, recondensation reactions, or their inherent slower diffusivity through the biomass pores. HSQC-NMR spectroscopy showed little β-O-4 remaining in the lignin oil isolated at 1 h, and slightly more β-O-4 content in the later time points. Additionally, no peaks corresponding to β-O-4 dimers were observed, which previously were observed for incomplete hydrogenolysis 46 . Because of the low β-O-4 content in the RCF products, it is unlikely that incomplete depolymerization of lignin was the cause of the observed trends. Recondensation of lignin occurs through a dehydration step at the α-carbon of monomers, which generates a carbocation that is susceptible to attack from the aromatic ring of other lignin fragments. Indeed, stabilization strategies have been developed to protect this position and reduce recondensation. Lancefield et al. showed that lignin extracted in alcohols resulted in the addition of the alcohol to the α carbon of monomers to prevent condensation reactions 30 , 54 . Shuai et al. 28 also showed that formaldehyde treated lignin created a dioxane ring with the hydroxyls at the α and γ positions of monomers, which prevents deleterious dehydrations from occurring. Shuai et al. also showed that monomer yields from a formaldehyde-stabilized lignin were identical to those produced from direct RCF of the same biomass. In this work, we demonstrated similar trends using either Ni/C or commercial Ru/C catalysts, indicating the results obtained were not caused by the type of catalyst used. Therefore, the supercritical RCF experiments performed within this study should indicate a true monomer yield for each natural variant, ruling out recondensation as the cause of the observed trends. Comparing the data collected in supercritical batch and flow-through experiments in this work further supports the claim that variation in monomer yields due to condensation is unlikely. On average, the monomer-to-oil ratio from supercritical RCF experiments was 0.38, while flow-through RCF had a monomer-to oil ratio of 0.42 implying a similar amount of depolymerization. If lignin condensation had occurred, the monomer-to-oil ratio of 0.6 observed at 1 h on stream would be the expected result from the supercritical RCF runs. Additionally, the only condensation product observed by GC-MS in the dimer region was the intramolecular condensation of the resinol. Taken together, these data indicate that differences in the molecular weight distributions as a function of time are most likely not caused by repolymerization during extraction. The most probable factor influencing the temporal dependence of monomer yields and types of dimers observed is differences in chain length. Diffusion of lignin in the internal pores of wood particles would allow for small chains to appear early in the extraction while long chains would lag behind. The high occurrence of β-1 linkages observed at early times supports this hypothesis, because β-1 linkages can form from the fragmentation of a growing chain to start a new polymer chain 52 . As these fragments are formed later in the lignin synthesis, they are likely shorter than those chains formed at the beginning of the process. The increase in β-5 bonds over the course of the extraction also supports this hypothesis, since these are C–C linkages formed during chain growth. The high amount of β-β linkages present can only act as starting points to chains because β-β bonds cannot form on a growing chain. The hypothesis of chain diffusion also could explain the initially high monomer yields. The monomer yield will have some dependence on chain length. At two extremes, six 4-unit chains—each with an average of 50% β-O-4 content—will produce a 38–50% yield of monomers (depending on the location of the C–C bonds), while a single chain with 24 units and a 50% β-O-4 content will generate 25% of monomers. This concept was illustrated by Galkin et al. 55 and is based on the idea that monomer production requires β-O-4 bonds at both the 4- and β-position of the monomer unit in the polymer chain. In short polymers, terminal positions of the chain are important, while in longer chains their contribution is negligible. The consistent monomer yields between all of the different natural variants implies there may be some degree of control exerted by the plant that influences the types of linkages formed in the polymerization process. Lignin biosynthesis is dictated by the generation of radicals by peroxidase or laccase enzymes. Radical coupling is fast, and thus the only handle to manipulate the chemistry from a kinetically controlled polymerization is the concentration of the monomers in the cell wall 56 . In vitro experiments to generate dehydrogenase polymers (DHP) have been used to generate synthetic lignin. Batch experiments have been performed with coniferyl and sinapyl alcohol. High monomer concentrations led to the formation of C–C linked dimers with β-5 formed from coniferyl alcohol and β-β dimers from sinapyl alcohol 57 , 58 . In similar DHP experiments, coniferyl alcohol was slowly added to the reactor maintaining a low concentration. The low concentration led to nearly a 50% yield of β-O-4 linked dimers. Furthermore, when a dialysis membrane was used with both sinapyl and coniferyl alcohol, an insoluble polymer was generated exhibiting similar properties to lignin 59 , 60 . The importance of monomer delivery rate was shown computationally by van Parijs et al. 61 who demonstrated that decreasing the influx of monolignols in a lignin polymerization model increased the amount of β-O-4 formation by favoring the growth of longer chains and decreasing dimerization. Density functional theory calculations on the energetics of lignin monomer coupling align with these observations, showing that dimerization of two S units will kinetically favor the formation of β-β bonds 62 . Thus, during lignin biosynthesis, it is plausible that high sinapyl alcohol concentrations relative to the number of growing chains will lead to β-β formation, while a low sinapyl alcohol concentration during lignification would likely lead to formation of β-O-4 in a strictly syringyl polymer. This concept, illustrated in Fig. 6 , was also shown experimentally by Stewart et al. 27 who found that genetically modified poplar with 97.5% S showed a significant increase in levels of β-β linkages (relative to native S-levels in wild-type trees), but an insignificant increase in β-ether levels 27 . Therefore, the consistent monomer yields that we observed for a range of S/G values could be caused by manipulation of monomer concentrations by the plant to control lignin structure and composition leading to similar lignin in different natural variants. Broadly, monomer concentrations during lignification appear to be an important variable to consider when designing lignin for depolymerization in addition to the S/G ratio. Fig. 6 Illustration of monomer concentration influence on bond formation during lignification. In the case of fast monomer transport from the cytoplasm to the cell wall, monomers can couple together to form dimers, or add to growing lignin chains. In the case of slow monomer transport, if an S monomer can only add to a growing chain that already contains a β-β bond, it must form a β-O-4 ether bond The results of the study contained herein show that there is no correlation between S/G ratio and monomer yields within the range of S/G ratios in the naturally variant poplar population. All data were collected from only five poplar natural variants with differing S/G ratios. Indeed, several other parameters could influence lignin bond formation, including pore structure, plant cell wall microstructure, and lignin carbohydrate linkages 45 , 63 . In subsequent studies, we will employ a larger population of natural variants to perform multi-variate studies in order to understand to what extent other factors truly influence lignin depolymerization. The lignin in a series of natural poplar variants with lignin S/G ratios ranging from 1.41 to 3.60 was extracted and depolymerized using RCF in flow-through reactors. Surprisingly there was found to be no correlation between S/G ratio and monomer yields in a flow-through reaction at 50% lignin extraction. Furthermore, when operating at 80–90% lignin extraction, the monomer yields were similar between all poplar samples at approximately 32 wt%. HSQC-NMR spectroscopy, GPC, and silylated GC-MS were performed to understand differences in the high molecular weight fractions of each extracted lignin oil. GPC showed an increase in molecular weight of lignin oil extracted at later times on stream in the flow-through extraction. NMR spectroscopy indicated that these large molecular weight fragments consisted of primarily S lignin units. Analysis of the dimers produced at different times on stream showed an increase in S–G β-5 linkages over time as well as a high amount of S–S β-β linkages throughout the extraction. The similar monomer yields between a wide range of naturally variant wood samples is likely caused by the plants ability to regulate lignification."
} | 4,239 |
28246645 | PMC5310828 | pmc | 3,515 | {
"abstract": "Genome-wide assessment reveals opposing patterns of vertical connectivity in two depth-generalist coral species.",
"introduction": "INTRODUCTION Tropical coral reefs are in global decline, with many under immediate threat because of a rapidly changing climate and the accumulation of local stressors ( 1 , 2 ). Even in remote areas with strong legal protection, the increase in frequency and magnitude of large-scale bleaching and storm events has severely affected coral reefs ( 3 ). Coral reef persistence increasingly depends on local areas that can offer refuge against such disturbances ( 4 ) and provide propagules to recolonize affected areas ( 5 ). Deeper sections of coral reefs (below depths of 30 to 40 m), also referred to as “mesophotic coral ecosystems,” arguably represent potentially critical ecological refuges. Although certainly not immune to disturbance ( 6 – 9 ), they are commonly buffered from major bleaching and storm events ( 10 , 11 ) and have a near-ubiquitous presence directly adjacent to the world’s shallow coral reefs ( 12 ). Although their ability to escape disturbances has been relatively well documented, the hypothesis that deep reefs harbor reliable reproductive sources remains largely untested ( 11 , 13 – 15 ) and this was recently highlighted as a key knowledge gap in a report by the United Nations Environment Programme ( 16 ). The “reseed” potential of mesophotic coral reefs firstly depends on the extent of species overlap with their shallow-water counterparts. Although greater depths often confer greater protection from disturbances ( 11 ), the community similarity between shallow and deep reefs also decreases with depth, as a reflection of changing environmental conditions) ( 17 ). Given this trade-off, the “deep reef refuge” hypothesis (DRRH) pertains primarily to upper mesophotic depths (~30 to 60 m)—deep enough to escape disturbances but shallow enough to guarantee sufficient overlap in species composition ( 11 , 18 , 19 ). Species overlap between shallow and upper mesophotic communities can be substantial, with an estimated 25 to 40% of Caribbean coral species and ~20 to 40% of Acropora species (on the Great Barrier Reef) representing depth-generalist species ( 11 , 20 ). However, initial assessments have demonstrated that depth-generalist coral species are not necessarily composed of a single panmictic population over depth ( 14 , 21 – 25 ). The three coral species thus far assessed for shallow-mesophotic connectivity (which includes two brooding and one broadcasting species) exhibit ambiguous patterns across geographic regions, with genetic differentiation observed in certain locations but not others ( 14 , 22 , 23 , 25 ). This geographic variation has consequently hampered a more general assessment of the DRRH as the overall role of deep reefs in shallow reef recovery remains unclear ( 16 , 26 ). To understand the reseed potential on an ecosystem-wide scale (that is, across many species), it is essential to identify the nature of barriers to vertical connectivity and how these barriers may vary across coral species with distinct life history traits. The relative roles of depth and location have remained difficult to disentangle in past assessments of vertical connectivity, often because of geographic separation of sampled shallow and mesophotic populations ( 14 , 22 , 25 , 27 ). Nonetheless, depth-associated selection could pose an important ecological barrier to vertical connectivity ( 24 , 28 ) even in the absence of physical barriers (for instance, on those reef slopes where shallow and mesophotic communities are directly adjacent). Depth-associated selection does not necessarily restrict gene flow; however, in species with localized sperm and larval dispersal (such as in most brooding corals), localized dispersal could lead to assortative mating and therefore facilitate a divergence-with-gene-flow scenario ( 28 , 29 ). Similarly, assortative mating could facilitate divergence in broadcasting species if depth-associated selection is accompanied by temporal reproductive isolation, either through pleiotropic effects or through depth-related differences in spawning cues ( 24 , 27 , 30 ). The role of such ecological barriers to vertical connectivity remains poorly understood, and the impact of depth-associated selection on vertical connectivity has remained largely obscured because genomic sampling has been restricted to mostly to microsatellite sequencing in presumably neutral parts of the genome [but see the work of Brazeau et al. ( 23 )]. Reduced-representation approaches to genome sequencing [for example, restriction-site associated DNA sequencing (RAD-seq)] allow for sampling of both adaptive and neutral genetic variation across the genome ( 31 , 32 ). In combination with a replicated sampling design (sampling the same habitats across multiple locations in a region), such reduced-representation approaches have the potential to identify the nature of barriers (that is, ecological or physical) that might hamper cross-habitat connectivity. Nonetheless, uptake of such genotyping-by-sequencing approaches in symbiotic marine invertebrates has been slow [but see related studies ( 33 – 37 )] largely due to the nonspecific nature of these methods (for example, targeting universal restriction site motifs) in light of pervasive contamination by the obligate endosymbionts. In reef-building corals, this contamination stems from the fact that they are associated with a complex consortium of eukaryotes, prokaryotes, and viruses ( 38 ), with the hosted Symbiodinium representing a significant proportion of the overall biomass ( 39 ). Endosymbiont contamination might not necessarily affect inferred genetic patterns for the host; however, this can only be ascertained when comparing it against an aposymbiotic (that is, lacking symbionts) reference ( 36 ). Contamination is of particular concern when evaluating horizontal versus vertical connectivity, given the prevalence of geographic ( 40 ) and bathymetric ( 41 ) zonation of associated Symbiodinium . Unfortunately, published genomes for scleractinian corals are still scarce ( 42 , 43 ), and obtaining aposymbiotic tissue remains relatively difficult. In this study, we test the reseeding potential of deep reefs by sampling genome-wide variation in shallow and deep population pairs of two species with contrasting reproductive modes ( Agaricia fragilis and Stephanocoenia intersepta ). To overcome the contamination issue, we used a novel method that generated a subtraction reference genomic library of Symbiodinium symbionts (isolated from coral hosts using fluorescence-activated cell sorting). We focused on the reef system of Bermuda, where there is extensive upper mesophotic habitat adjacent to the shallow reef ( 44 ), and whose isolated geographic setting suggests the probability that Bermudan corals depend on local recruitment sources for recovery ( 14 , 25 , 45 ). In addition, previous studies highlighted the potential importance of the DRRH for at least some coral species at this location, given the observed vertical connectivity for two other Bermuda corals ( Montastraea cavernosa and Porites astreoides ) ( 14 , 25 ). Through sampling of replicate shallow and deep populations, we assess whether the vertical connectivity potential differs between the two species and how vertical connectivity might be affected by depth-associated selection. More broadly, we evaluate how these species-specific patterns affect the reseeding potential of deep reefs in Bermuda and discuss the wider implications and caveats in light of the DRRH.",
"discussion": "DISCUSSION The rapid decline of shallow-water coral communities has led to a growing interest in mesophotic coral ecosystems because of their potential to act as a refuge and aid in shallow reef recovery ( 10 , 11 , 19 , 46 ). Although the ability of deep reefs to escape disturbance events is relatively well established ( 11 ), their role as reproductive sources remains largely untested ( 13 – 16 ). Our study demonstrates that vertical connectivity can vary greatly between species within a single reef system. Using replicated sampling of adjacent shallow and mesophotic populations and a reduced-representation genome sequencing approach (which accounts for endosymbiont contamination), we demonstrate that selection drives depth differentiation in the brooding species A. fragilis . In contrast, extensive gene flow and lack of differentiation characterize the broadcasting species S. intersepta . Overall, the reseed potential of deep reefs appears restricted to only a subset of depth-generalist coral species; although ecologically relevant to these individual species, deep-water coral reseeding should not be assumed to be a broader ecosystem-wide phenomenon. The strong genome-wide differentiation between shallow and mesophotic populations of A. fragilis was linked to depth-associated selection for a select number of RAD loci ( Fig. 3 , A and B). A pattern of local-scale genetic differentiation is commonly observed among brooding species and is generally associated with the more localized dispersal of both sperm and larvae ( 45 , 47 – 50 ). In the case of A. fragilis , geographic distance cannot account for the strong observed differentiation between shallow and deep eastern populations, because differentiation between depths within each of the three locations consistently exceeded that observed for the same depth between locations (average horizontal distances of ~1.5 km versus ~12 km, respectively). Outlier analyses identified strong depth-associated selection across a number of RAD loci ( n = 12; 0.8% of those examined). In the presence of substantial gene flow, such divergent selection across opposing environments is expected to result in heterogeneous genomic divergence (or “genomic islands of divergence”) ( 51 ). Instead, for A. fragilis , the background levels of genetic differentiation were also observed to be high (with genetic structuring remaining after removing outliers), indicating that the depth-associated selection in A. fragilis has led to genome-wide “resistance” to shallow-deep gene flow. Given the observed lack of shallow-deep migrants, immigrant inviability (that is, selection against migrants between locally adapted populations) is likely an important contributor to this resistance to gene flow in A. fragilis , reducing the chances of between-ecotype mating encounters and leading to assortative mating ( 52 ). Despite this apparent gene flow barrier, we observed introgression between populations at shallow and mesophotic depths, which may be facilitated through ecologically rare events of migration and interbreeding that were detectable with our relatively small sample sizes. Alternatively, limited introgression could be facilitated by stepping-stone populations at intermediate depths (not sampled), effectively enabling “multigenerational vertical connectivity” ( 15 ). Either way, the extent of vertical connectivity does not appear to be ecologically relevant. Introgression between populations of A. fragilis was asymmetrical ( Fig. 5 ), matching observations from two other studies reporting directional gene flow from shallow to deep populations in the broadcasting species M. cavernosa and the octocoral Eunicea flexuosa ( 24 , 25 ). In the case of E. flexuosa , higher gamete production in shallow water (due to greater abundances and average colony sizes) was offered as a likely explanation for the asymmetric introgression ( 24 ). However, for A. fragilis , this explanation appears unlikely given that population densities at two of three locations ( Fig. 6 ) appear lower at shallow depths. Instead, the lower population densities and the near-complete or complete fixation of the major allele in 8 of 12 outlier RAD loci at shallow depths ( Fig. 3 ) indicate that selection against immigrants may be stronger at shallow depths (and, conversely, may not be as strong for downslope migration). Differential selective pressures could be a contributor to the observed asymmetric introgression ( Fig. 5 ) and associated higher genetic diversity at mesophotic depths (table S1). The observed introgression and the lack of alternatively fixed polymorphisms in A. fragilis appear to reflect a population-level rather than a species-level divergence. This is further corroborated by the observation that shallow-deep differentiation (eastern locations) is comparable to that observed between the western and eastern deep populations (fig. S2). Although distance and/or seascape resistance (for example, in relation to prevailing currents) may have played a role in the differentiation of this Western Blue Cut (WD) location, several outlier loci were identified in association with this specific population, perhaps indicative of environment-based selection (fig. S1). The western population also exhibited a phenotypic skeletal signature distinct from the other deep locations ( Fig. 1B ). Despite this study focusing on vertical connectivity, the identification of both neutral differentiation and outlier loci in association with this western location reiterates the broader role of environment-based selection in cross-habitat connectivity ( 21 , 28 , 53 , 54 ). For the gonochoric broadcasting species S. intersepta , no genetic structuring was observed across depths and locations ( Figs. 2 and 4 ). Strong genetic differentiation was only observed for the two out-group individuals from Curaçao ( Fig. 2 ), which is expected given their origin near the opposite end of the latitudinal distribution of this species. Eggs released by S. intersepta (as well as by M. cavernosa ) have been observed to exhibit high rates of fertilization immediately upon release ( 55 ), which appears to be facilitated by intratentacular fertilization ( 56 ). Although this rapid fertilization may not directly affect larval dispersal, it would result in more localized cross-fertilization similar to that in species with a brooding reproductive mode. Assuming such localized spermcasting, the observation of a single panmictic population across the reef platform of Bermuda indicates considerable larval migration between both depths and locations. Similar genetic connectivity over depth has been inferred in Bermuda for the broadcasting species M. cavernosa (4 to 58 m depth) ( 25 ). In contrast to the brooding species A. fragilis , no RAD loci were identified to be under depth-specific selection [although, of course, only a reduced proportion of the genome was queried ( 57 )]. In combination, the observed panmixia, apparent morphological plasticity, and association with a single endosymbiont highlight the depth-generalist nature of S. intersepta and the potential for deep populations of this species to act as a larval source for shallow reefs. The subtraction method allowed us to successfully apply a de novo reduced-representation genome sequencing approach (in this case, nextRAD) on endosymbiont-contaminated samples. Despite our attempt to reduce the endosymbiont contamination through multiple steps of centrifugation before DNA extraction, we still found 10 to 14% of RAD loci to be of Symbiodinium origin, highlighting the necessity of further steps to identify and remove such contamination. Only a small fraction of these sequences could be identified solely by comparison with Symbiodinium reference genomes, which is not unsurprising given their potential level of divergence (particularly in the case of the Symbiodinium minutum “B1” genome). This indicates that comparison with unrelated reference genomes alone is not a viable approach to eliminating endosymbiont contamination. Thus, although genotyping-by-sequencing approaches have demonstrated their de novo potential in nonmodel organisms, population genomic assessments of endosymbiotic marine invertebrates still require either an aposymbiotic host reference [by sequencing aposymbiotic sperm or larvae; for example, see related studies ( 35 – 37 )] or an endosymbiont subtraction reference (as in this study). Because obtaining aposymbiotic tissue can be difficult, particularly in brooding species, our subtraction method based on fluorescence-activated cell sorting provides an easy and cost-effective solution to this pervasive issue. Overall, we demonstrate that coral species with similar ecological distributions can exhibit very distinct potentials for vertical connectivity at a single location. Geographic congruence in patterns of vertical connectivity in M. cavernosa and P. astreoides (despite their opposing reproductive modes) highlighted the importance of location-specific extrinsic factors in determining vertical connectivity ( 14 ). Our study focused on a single geographic region to specifically evaluate the role of depth, controlling for location-specific effects (by sampling replicate pairs of proximate shallow and mesophotic populations) and sampling genetic variation across the genome. The observed pattern for A. fragilis demonstrates how depth-associated selection can hamper vertical connectivity in the absence of physical dispersal barriers and corroborates the emerging notion that divergence by depth is prevalent in species with a brooding reproductive mode ( 21 , 22 , 29 , 58 ). In contrast, we found that the broadcasting species S. intersepta formed a single panmictic population across the same depths and locations. Nonetheless, divergence by depth has been observed in both brooding and broadcasting species ( 14 , 21 – 25 ), and the hypothesis that vertical connectivity is more prevalent in broadcasters than in brooders ( 11 ) cannot be evaluated until more species and locations are assessed. Regardless, these results demonstrate that the reseeding potential of deep reefs only applies to a subset of depth-generalist species. Of the ~20 zooxanthellate scleractinian coral species described for Bermuda ( 59 ), 6 species appear to be depth-generalists occurring at both shallow and mesophotic depths ( 44 , 60 ). Among these species, A. fragilis , Madracis decactis , and Scolymia cubensis are brooders, while S. intersepta , M. cavernosa , and O. franksi are broadcasters ( 61 ). Although one of the brooders, A. fragilis , occurs in reasonable abundance at both shallow and mesophotic depths, our study showed little evidence for vertical connectivity between shallow and deeper-water populations. On the basis of their reproductive mode, such vertical connectivity may be similarly hampered for M. decactis and S. cubensis . Regardless, S. cubensis was rare at both depths, and M. decactis was only common at intermediate and mesophotic depths along the southern shore (and therefore not individually reported in Fig. 6A ), further indicating a limited importance of these species in a reseeding context. Similarly, P. astreoides (for which vertical connectivity was observed across 4 to 26 m depth) ( 14 ) was relatively rare at mesophotic depths (~40 m). In contrast, all three broadcasting species occurred in reasonable abundance at both shallow (12 m) and upper mesophotic (40 m) depths ( Fig. 6A ), and mesophotic populations of these species may represent a viable source of recruitment for the shallow reefs in this isolated reef system (although O. franksi has not been assessed). Nonetheless, although these three species constitute a considerable proportion of the diversity and composition of the upper mesophotic reef, they only represent a relatively small fraction of the diversity (~15% of species) and community (~15% of coral colonies) in shallow water ( Fig. 6B ). The DRRH postulates that deep reefs (i) are protected or dampened from disturbances that affect shallow reefs and (ii) can provide a reproductive source after disturbance [sensu Bongaerts et al. ( 11 )]. Although deep reefs can escape certain disturbances, they are not entirely immune to disturbance ( 6 – 9 , 26 ). In fact, even their ability to escape bleaching and storm events can vary spatially and temporally, and their greater depth does not confer long-term protection against disturbances associated with climate change ( 9 ). Similarly, disturbance events can vary strongly in their impact over depth and in their impact on specific species, which can vary in thermal susceptibility or structural fragility. Thus rather than the term “refugia” (as currently used in “deep reef refugia hypothesis”), which Keppel et al . ( 62 ) classify as large geographic areas offering an escape over evolutionary time scales (for example, climate change refugia), we propose that the term “refuge(s)” is probably more appropriate (that is, “deep reef refuge hypothesis”), reflecting the role of deep reefs in providing temporal and/or spatial protection from disturbances on ecological time scales. In addition, this study confirms that the active reseeding potential is specific to only a subset of species (~15% here) and does not apply to a significant proportion of the species (diversity) that make up shallow reefs. Therefore, we conclude that the “deep reef refuge hypothesis” should be considered as an ecological concept relevant to individual species, rather than being referred to as an ecosystem-wide phenomenon."
} | 5,335 |
40274865 | PMC12022357 | pmc | 3,516 | {
"abstract": "Off-policy learning exhibits greater instability when compared to on-policy learning in reinforcement learning (RL). The difference in probability distribution between the target policy ( \\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}$$\\pi$$\\end{document} ) and the behavior policy (b) is a major cause of instability. High variance also originates from distributional mismatch. The variation between the target policy’s distribution and the behavior policy’s distribution can be reduced using importance sampling (IS). However, importance sampling has high variance, which is exacerbated in sequential scenarios. We propose a smooth form of importance sampling, specifically relative importance sampling (RIS), which mitigates variance and stabilizes learning. To control variance, we alter the value of the smoothness parameter \\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}$$\\beta \\in [0, 1]$$\\end{document} in RIS. We develop the first model-free relative importance sampling off-policy actor-critic (RIS-off-PAC) algorithms in RL using this strategy. Our method uses a network to generate the target policy (actor) and evaluate the current policy ( \\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}$$\\pi$$\\end{document} ) using a value function (critic) based on behavior policy samples. Our algorithms are trained using behavior policy action values in the reward function, not target policy ones. Both the actor and critic are trained using deep neural networks. Our methods performed better than or equal to several state-of-the-art RL benchmarks on OpenAI Gym challenges and synthetic datasets.",
"introduction": "Introduction Various intricate challenges have been tackled using model-free deep RL methods 1 – 8 . Model-free RL learning encompasses both on-policy and off-policy approaches. Off-policy approaches enable the simultaneous learning of a target policy while observing and gathering data from another policy, known as the behavior policy. It means that an agent learns about a policy distinct from the one it is carrying out while there is a single policy (i.e., target policy) in on-policy methods. It means that the agent learns only about the policy it is carrying out. Simply put, if two policies are identical (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}$$\\pi =b$$\\end{document} ), then the arrangement is referred to as on-policy. Alternatively, the scenario is referred to as off-policy, if \\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}$$\\pi$$\\end{document} is not equal to b 8 – 13 . Figure 1 A comparison of on- and off-policy learning. Figure 1 a illustrates that off-policy learning primarily involves two policies: the behavioral policy (b), also known as the sampling distribution, and the target policy ( \\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}$$\\pi$$\\end{document} ), also known as the target distribution. Figure 1 a also shows that there is often a discrepancy between these two policies ( \\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}$$\\pi$$\\end{document} and b). This discrepancy makes off-policy unstable and introduces significant variance 14 – 20 ; a bigger difference between these policies, instability is also high, and a smaller difference between these policies, instability is also low in off-policy learning, whereas on-policy has a single policy (i.e., target policy), as shown in Fig. 1 b. The instability is not an issue for on-policy learning due to the sole policy. Therefore, compared to off-policy, on-policy is more stable. In addition to the aforementioned benefits, there are other advantages and disadvantages associated with off-policy and on-policy learning. On-policy approaches, while unbiased, frequently encounter challenges like sample inefficiency. Off-policy approaches are characterized by higher sampling efficiency and are safe, yet they may exhibit instability and add variance. Both on-policy and off-policy approaches have their limitations. Consequently, multiple approaches have been suggested to address the shortcomings of each strategy. For instance, it is possible for on-policy methods to attain a comparable level of sample efficiency as off-policy methods 5 , 6 , 8 , 21 , 22 . Similarly, off-policy methods can achieve a similar level of stability as on-policy methods 10 , 13 , 23 – 25 and mitigate the variance induced by distributional mismatch 20 , 26 . In practical reinforcement learning (RL) contexts, such as autonomous driving or robotic control, the policies that provide data frequently diverge substantially from the target policies. This distributional discrepancy might result in instability and elevated variance throughout the learning process, especially in continuous action spaces 14 , 16 , 20 . A key motivation for this study to mitigate instability and variance in off-policy learning. Recent improvements have underscored the advantages of divergent behaviour policy for exploration; nonetheless, these methods are frequently restricted in long-horizon tasks or offline reinforcement learning contexts, where exploration is restricted or impractical. Furthermore, numerous approaches depend on rigid assumptions about the behaviour policy or reward structure, hence limiting their application in intricate, real-world environments 15 , 27 . The suggested RIS-off-PAC algorithm seeks to address these constraints by dynamically adjusting for distribution discrepancies via relative importance sampling. This method guarantees a consistent learning process. RIS-off-PAC enhances reliability and scalability in off-policy learning by reducing instability and variance, which is crucial for real-world applications when data acquisition is costly or limited. Importance sampling is a well-known method to evaluate off-policy, permitting off-policy data to be used as if it was on-policy 12 . IS can be used to study one distribution while a sample is made from another distribution 28 . The degree of deviation of the target policy from the behavior policy at each time t is captured by the importance sampling ratio (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}$$IS = \\frac{\\pi (A_{t}|S_{t})}{b(A_{t}|S_{t})}$$\\end{document} ) 11 . IS is also considered as a technique for mitigating the variance of the estimate of an expectation by cautiously determining sampling distribution (b). Our new estimate has low variance, if b is chosen properly. The variance of an estimator relies on how much the sampling distribution and the target distribution are unlike 29 . For theory behind importance sampling that is presented here, we refer to see 28 , Chapter 9 for more details. An additional factor contributing to the instability of off-policy learning is the lack of uniformity in the values generated by importance sampling (IS) for different samples. The IS occasionally produces a high value for certain samples and a low value for other samples, hence amplifying the disparity between the two distributions. Authors 30 introduced a smooth version of importance sampling called the relative importance sample. This method was proposed to address the instability in semi-supervised learning. We apply this technique in deep reinforcement learning to alleviate the discrepancy between \\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}$$\\pi$$\\end{document} and b, hence diminishing the variation and instability associated with off-policy learning. Some notable methods based on Importance Sampling (IS) include: Weighted Importance Sampling (WIS) 23 , Sample Efficient Actor-Critic with Experience Replay (ACER) 24 , Retrace 16 , Q-prop 8 , Soft Actor-Critic (SAC) 25 , Off-Policy Actor-Critic (Off-PAC) 10 , The Reactor 13 , Guided Policy Search (GPS) 31 , Efficient Multiple Importance Sampling (MIS) 32 , and others. In summary, the following contributions are made by this paper: (i) We develop a simple Relative Importance Sampling (RIS) estimator that improves stability and diminishes the variance of off-policy approaches. (ii) We provide an off-policy actor-critic method, termed RIS-off-PAC, that utilises relative importance sampling in deep reinforcement learning. As far as we know, we are presenting the first instance of RIS with actor-critic. Furthermore, we investigate a variation of the actor-critic method termed the natural gradient actor-critic, which employs relative importance sampling. This form, known as the relative importance sampling-off-policy natural actor-critic (RIS-off-PNAC), substantially enhances our contributions. (iii) On benchmark problems, The RIS estimator exhibits performance that is either superior to or competitive with various state-of-the-art RL benchmarks, while maintaining stable learning. The remaining sections of the paper are organized as follows: The discussion of related works can be found in section 2. In section 3, we provide a preliminary. Sections 4 and 5 demonstrate the concepts of relative importance sampling and the actor-critic model, respectively. Section 7 provides a detailed account of the conducted experiments. Ultimately, we provide a conclusion in section 8.",
"discussion": "Discussion and conclusion We have demonstrated off-policy actor-critic reinforcement learning methods utilizing RIS. It has attained superior or comparable performance to state-of-the-art methods. This method mitigates the instability and variance typically associated with off-policy learning. Furthermore, our algorithm effectively addresses well-known RL challenges, including CartPole-v0, Humanoid-v2, MountainCar-v0, and Pendulum-v0. Our methodology can also be adapted to additional importance sampling methodologies with few modifications. For instance, Per-decision Importance Sampling (PDIS) can be transformed into Relative Per-decision Importance Sampling (RPDIS), and Weighted Importance Sampling (WIS) can be adjusted to Relative Weighted Importance Sampling (RWIS). We defer these extensions for future endeavours."
} | 2,968 |
38144681 | PMC10739613 | pmc | 3,517 | {
"abstract": "Nanofiber mats containing poly(3,4-ethylenedioxythiophene)\n(PEDOT)\nhold potential for use in wearable electronic applications. Unfortunately,\nthe use of PEDOT is often limited by the acidic nature of polystyrenesulfonate\n(PSS), a common dispersant for PEDOT. In this study, we explored the\nimpact of increasing the pH value of PEDOT:PSS/poly(vinyl alcohol)\n(PVA) precursors on the morphological and electrical properties of\nthe resultant electrospun fibers. Specifically, electrospun nanofibers\nwere analyzed using scanning electron microscopy, bright-field microscopy,\nand two-point probe measurements. We discovered that neutral and even\nslightly basic PEDOT:PSS/PVA precursors could be electrospun without\naffecting the resultant electrical properties. While cross-linking\neffectively stabilized the fibers, their electrical properties decreased\nafter exposure to solutions with pH values between 5 and 11, as well\nas with agitated soap washing tests. Additionally, we report that\nthe fiber mats maintained their stability after more than 3000 cycles\nof voltage application. These findings suggest that PEDOT:PSS-based\nfibers hold potential for use in wearable textile and sensor applications,\nwhere long-term durability is needed.",
"conclusion": "Conclusion In this study, we have electrospun PEDOT:PSS/PVA\nnanofibers by\nadjusting the pH of the precursors by adding NaOH. We found that electrospinnability\nand electrical properties, were not impacted when the pH became slightly\nbasic (pH = 7.3). We also explored the performance of the fibers after\ncross-linking, exposure to different pH value solutions, as well as\nwashing with agitated soap water. The electrical current was retained\nafter glutaraldehyde vapor cross-linking, but it decreased by a factor\nof 3 upon immersion in solutions with pH values of 5 to 11. Moreover,\nthe reduction in current after the first wash with soap water compared\nto cross-linked fiber mat was ∼50%. Our results indicate that\nPEDOT:PSS/PVA fibers are suitable for applications requiring exposure\nto acidic or basic environments. We also evaluated their stability\nfor long-term storage and found that fiber mats obtained from less\nacidic precursors are more stable. Future work to improve these fabrics\ntoward wearable devices could explore alternative cross-linking methods,\nwhich would provide more electrical stability postimmersion and multiple\nwashings of the fabrics.",
"introduction": "Introduction Electrically conductive fiber mats are\nideal for applications,\nsuch as sensors, 1 − 4 artificial skin, 5 − 7 and wearable textiles 8 due to their surface-to-volume ratio, porosity, stretchability,\nand mechanical strength. 9 Electrospinning\nis the most common method of fabricating fiber mats due to its ease\nof use, control over fiber diameter and morphology, production of\ncontinuous fibers, facile functionalization of fibers, and scalability. 10 − 12 To obtain conductive fibers via electrospinning, typically conductive\nadditives, such as carbon nanotubes, 13 , 14 metal nanowires, 15 and/or conjugated polymers, 16 , 17 are blended with a carrier polymer to facilitate fiber formation.\nIf the conductive additive is not as flexible as the polymer matrix,\nthen the composites can fail due to mechanical mismatch (i.e., interfacial\nslipping, cracking, etc.) 14 To avoid mismatch,\nan alternative method is to use a flexible conjugated polymer within\nthe composite, such as polyalanine, polypyrrole, or poly(3,4-ethylenedioxythiophene)\n(PEDOT). PEDOT has been widely explored due to its high and\nstable electrical\nconductivity, ease of doping, good photo- and electrical stability\nin air, as well as high work function and transmittance. 18 , 19 Notably, the synthesis procedure of PEDOT affects its conductivity\nand dispersibility. 20 While the most common\nmethod for synthesizing PEDOT is oxidative chemical polymerization,\nother methods, such as electrochemical polymerization and transition\nmetal-mediated coupling polymerization have also been used. 18 , 21 Unfortunately, working with PEDOT can be challenging because it\nhas limited dispersibility in solvents. Therefore, in most common\ncommercial syntheses, EDOT is polymerized in the presence of a water-soluble\npolyanion, i.e., polystyrenesulfonate (PSS). PSS disperses and stabilizes\nPEDOT by forming colloidal particles containing a higher number of\nPSS chains at the periphery and a higher number of PEDOT chains at\nthe core. 22 Due to its electrical properties,\nPEDOT:PSS-based electrospun fibers have been used in various applications,\nincluding gas sensors that detect organic vapors, 23 carbon monoxide, 24 ammonia, 25 and hydrogen gas, 26 as well as strain 27 , 28 and humidity sensors; 29 and thermoelectric devices. 30 − 32 In most of\nthese studies, PEDOT:PSS was blended with a non-conducting carrier\npolymer, such as poly(vinyl alcohol) (PVA), 33 − 35 polyvinylpyrrolidone\n(PVP), 30 or poly(ethylene oxide) (PEO) 33 , 36 to improve the electrospinnability. In one study by Huang et al., 23 the carrier polymer was avoided by blending\nPEDOT:PSS with a small amount of magnesium nitrate that acted as a\nthickener to facilitate electrospinning. There are other studies that\nused coaxial electrospinning; in these, the optical and electrical\nproperties of fibers containing PEDOT:PSS in the core 37 , 38 or in the shell 38 were evaluated. The electrical performance of single nozzle electrospun PEDOT:PSS-based\nfiber mats has been explored in many studies. For example, Park et\nal. 33 electrospun PEDOT:PSS-based fibers\nand evaluated the effect of the carrier polymers (PEO or PVA) and\ncosolvents (dimethyl sulfoxide or ethylene glycol) on the electrical\nconductivity. PEDOT:PSS/PEO fibers had a higher conductivity than\nthe PEDOT:PSS/PVA fibers because when PEO was used, the more conductive\nquinoid structure was in abundance, as indicated by a broad peak at\n1430 cm –1 in the Raman spectra. Additionally, while\nethylene glycol increased the conductivity more than dimethyl sulfoxide,\nusing either solvent improved the conductivity of the PEDOT:PSS/PVA\nand PEDOT:PSS/PEO fibers, which was also supported by a broader peak\nin the Raman spectra. The conductivity of the as-spun PEDOT:PSS/PEO\nfibers was also improved after they were immersed in ethanol 39 or ethylene glycol 39 , 40 due to the removal of PEO and change in PEDOT conformation. In these\nstudies, the use of PSS was essential in dispersing PEDOT in the appropriate\nsolvents. Due to the presence of sulfonic groups in PSS, commercial\nformulations\nof PEDOT:PSS dispersions have a pH value between 1.5 and 2.5. 41 This high acidity leads to corrosivity; for\nexample, when PEDOT:PSS was deposited and dried as a hole-injecting\nlayer on indium doped tin oxide (ITO) anode for light-emitting diodes,\nthe ITO etched into the PEDOT:PSS layer due to the acidic nature of\nthe latter. 42 The etching process has been\nobserved in other systems as well and decreases the lifetime and performance\nof the devices significantly. 43 , 44 Therefore, in many\nstudies, EDOT monomers are polymerized on the substrate producing\na thin film of PEDOT without involving acidic PSS. 45 Another method to eliminate the adverse effects of PEDOT:PSS’s\nacidity is to increase its pH value by mixing it with a base, such\nas sodium hydroxide (NaOH). For example, Mochizuki et al. 46 increased the pH value of PEDOT:PSS dispersions\nup to 13 by adding NaOH and found that the electrical conductivity\nof the cast thin films decreased as the pH value increased due to\ndedoping of PEDOT. While many strategies to overcome the adverse effects\ncaused by the acidity of PSS have been reported for thin films, 47 only one strategy has been explored for electrospun\nfibers. 48 − 50 This involved the polymerization of EDOT monomers\nwithin or onto fibers to enhance their biocompatibility as well as\ntheir electrical and mechanical properties. Moreover, for the more\ncommon PEDOT:PSS-based fibers, there are no studies that discuss how\naltering the pH value of these commercial precursors impacts the final\nmorphological or electrical properties of the electrospun fibers.\nFurthermore, many applications, such as in vivo sensors,\nrequire the conductive fibers to be used in different pH environments\nand to-date no studies have investigated how exposing PEDOT:PSS-based\nfibers to different environments impacts their properties. In\nthe current work, we examined the effect of increasing the\npH value of PEDOT:PSS-based precursors from ∼2.5 to ∼12\non the electrospinnability and electrical properties of the resulting\nfiber mats. For PEDOT:PSS-based fibers to be appropriate for applications,\nit is important to understand their long-term performance and stability\nin various environments in which they might be used. For this reason,\nwe also cross-linked the mats and evaluated if their electrical conductivity\nchanged after submersion in solutions with various pH values, as well\nas after multiple rounds of washing tests featuring soapy water. The\nelectrical conductivity of as-spun fiber mats over multiple months\nand 3000 current–voltage scans was also evaluated. This study\ndemonstrates that the electrical properties of PEDOT:PSS-based fibers\npersist when electrospun from a wide range of pH values and after\napplication of multiple current–voltage cycles, making them\nsuitable for a variety of applications, including wearable textiles\nand sensors.",
"discussion": "Results and Discussion Effect of Precursor pH on the Electrospinability and Morphology\nof PEDOT:PSS-Based Fiber Mats We have successfully obtained\nelectrospun fibers from PEDOT:PSS-based precursors as a function of\ndifferent pH values, as outlined in Table 1 . Uniform and smooth fibers with spherical\ncross sections (see Figure 2 and Table S1 ) that were electrospun\nfrom LowPEDOT3, LowPEDOT5, LowPEDOT7, and LowPEDOT8 precursors had\naverage diameters of 190 ± 30, 240 ± 33, 220 ± 36,\nand 290 ± 54 nm, respectively. According to a paired two tailed t test, the fiber diameters were statistically different\nwith p < 10 –3 . It was observed\nthat as the precursor’s pH value increased, so did the solution\nviscosity, especially for the precursor that had a pH value of 8.5\n(which also deposited fewer fibers). The increase in viscosity of\nPVA solutions with increasing pH value has been reported by Kasselkus\net al. 54 When the viscosity of a precursor\nsolution increases, it leads to increases in fiber diameter because\nthe resistance to stretching increases. 55 , 56 When the pH\nvalue of the LowPEDOT precursor was further increased to 11.4, the\nprecursor was gel-like ( Figure S2 ) and\ncould not be processed via electrospinning. Figure 2 SEM micrographs and fiber\ndiameter distribution for (A) LowPEDOT3,\n(B) LowPEDOT5, (C) LowPEDOT7, and (D) LowPEDOT8 fiber mats. The average\ndiameter is shown in upper-right corner of the distribution graphs\nand the pH value of each precursor is also provided. In contrast, the HighPEDOT precursors resulted\nin fibers with beads-on-a-string\nmorphology across all pH values tested ( Figure S3 and Table S1 ). When electrospun from a precursor with a\npH value of 2.4, the average bead and fiber diameter were 400 ±\n200 and 70 ± 13 nm, respectively (see Figure S4 ). The HighPEDOT precursors result in beads-on-string morphology\nwas likely due to the lower concentration of the carrier polymer PVA.\nTypically, uniform fibers are obtained only when the concentration\nof the polymer is above entanglement concentration. 12 , 57 For HighPEDOT, as observed for LowPEDOT precursors, the fiber collection\nrate visually decreased at a pH value of 8.5; at a pH value of 12,\nthe precursor was not electrospinnable due to gelation/precipitation\n( Figure S5 ). Effect of Fiber Mat Thickness on Their Electrical Properties The effect of fiber mat thickness on their electrical properties\nwas explored by electrospinning LowPEDOT3 fiber mats with different\nthicknesses directly onto IDEs. We categorically named the fiber mats\nas thin, medium, thick, and very thick based on their transparency\n(see Figure 3 ). To\nevaluate their electrical properties under vacuum, the current was\nmeasured as a function of time (see Figure S6 ) at each voltage between −2.5 to 2.5 V. As shown in Figure S6 , the variation in the current (∼0.002\nμA) with time was much smaller than the average current (∼3.66\nμA), which implies that the current was constant with time at\neach voltage. Next, the average current was plotted as a function\nof voltage, as shown in Figure 4 . Figure 3 Brightfield microscopy images of as-spun fiber mats on IDEs were\ncategorized as (A) thin, (B) medium, (C) thick, and (D) very thick\nbased on their transparency. All fibers were electrospun from precursor\nLowPEDOT3. For scale, we note that the spacing between the electrodes\nis 100 μm. Figure 4 Current ( I )–voltage ( V ) data and bright-field microscopy images of the LowPEDOT3 fiber\nmats tested. The inset shows current–voltage data for Very thick 2 and Thin 2 samples, as the\ncurrent was too low to be visible in the main plot. For scale, we\nnote that the spacing between the electrodes is 100 μm. Figure 4 shows that\nno correlation between the electrical current and the thickness of\nthe fiber mats was observed. For example, the electric current through\nthe Very thick 2 sample (0.02 μA at 2.4 V)\nwas orders of magnitude smaller than those through the Very\nthick 1 (4.4 μA at 2.4 V) and Thin 1 (1.2 μA at 2.4 V). For sample Thin 2 , the\ncurrent was too low to be measured by using our two-point probe setup.\nWe suggest that the reason for the disconnect between thickness and\nthe electric current is due to the variation in adhesion of the fibers\nto IDEs, as well as interfiber packing within the mats. Poor adhesion\nleads to high contact resistance between mats and IDEs. The\nadhesion of electrospun fibers on the collector depends on\nmany factors, such as applied voltage during electrospinning, the\nthickness of the fiber mats, the conductivity and roughness of the\ncollector, cross-linking process, thermal treatment, and surface interactions\nbetween the collector and fiber mats. 58 − 60 We observed that many\nof the thick fiber mats delaminated from the IDEs, possibly due to\nresidual electric charge between fiber layers. 59 , 61 For such fiber mats, both the interlayer resistance and contact\nresistance increase with thickness. Hence, for fiber mats, unlike\nthe films/sheet, the relationship between electrical current and thickness\nis not linear due to inconsistency in adhesion. Therefore, the calculations\nof true electrical conductivity require values for thickness and adhesion\nboth of which are challenging to acquire; many thickness measurement\ntechniques, such as ellipsometry and profilometry, cannot be used\non rough and porous surfaces like fiber mats and incorporating the\nfiber adhesion values into the calculation for contact resistance\nis beyond the scope of this study. We note that using a 4-point probe\nsetup would eliminate the need for contact resistance measurements;\nhowever, we did not have access to 4-point probe setup and the roughness\nof the fiber mats would remain a concern and lead to inaccurate electrical\nmeasurements. Nonetheless, the resistance of our well-adhered as-spun\nfiber mats (∼0.1–2 MΩ) was similar to what other\nstudies have obtained for similar PEDOT:PSS fiber mats. 35 , 39 Electrical Properties of PEDOT:PSS-Based Fiber Mats Electrospun\nfrom Different pH Value Precursors With the insight that\nproducing well-adhered fiber mats is more important than obtaining\nelectrospun mats with precisely the same thickness, we next selected\nwell-adhered fiber mats to evaluate the effect that the precursor\npH value had on the electrical current. Figures 5 and S7–S9 display the results for the LowPEDOT3, LowPEDOT5, and LowPEDOT7\nfiber mats. We note that the current–voltage data are not shown\nfor LowPEDOT8 fibers because, at this pH value, not enough fibers\ncould be obtained on the IDEs for current measurements. The current\nat 2.5 V varied from 1 to 15 μA irrespective of the pH value\nof the precursor, indicating that the pH value of the precursor did\nnot impact the electric current. This finding is consistent with prior\nstudies involving films cast from pure PEDOT:PSS, which found that\neven though the conductivity decreased as the pH increased between\npH 4 and 11, the decrement was not as significant as between pH values\nof 1 and 4 or above a pH value of 11. 46 However, directly comparing the conductivity of these fiber mats\nwas not carried out, because of the aforementioned challenges in measuring\nthickness and contact resistance. Our data suggest that we successfully\nelectrospun fiber mats from PEDOT:PSS-based precursors from acidic\nto neutral pH, without affecting their electrical properties. Figure 5 Current ( I )–voltage ( V ) data measured under\na vacuum for as-spun (A) LowPEDOT3, (B) LowPEDOT5,\nand (C) LowPEDOT7 fiber mats. The y and x axes represent electrical current ( I ) and voltage\n( V ), respectively. In each case, the current was\nmeasured for 4 distinct fiber mats with different thicknesses. The\nthickness increased from sample 1 to sample 4. See Figures S7–S9 for the microscopy images of the fiber\nmats. We also evaluated the electrical properties of\nHighPEDOT2, HighPEDOT5,\nand HighPEDOT7 fiber mats, as shown in Figure S10 . For the HighPEDOT2 fiber mats, the electric current varied\nfrom ∼0.01–1 μA, possibly due to variation in\nthickness and the adhesion of the fiber mats to the electrodes as\ndiscussed previously. However, when the pH value of the HighPEDOT\nprecursor increased to 5.3 and 7.3 (HighPEDOT5 and HighPEDOT7), the\nelectric current increased by 2 orders of magnitude (see Figures S10 and S11 ). We note that the reason\nfor this increase was likely the appearance of precursor droplets\non the electrodes. We suggest that the apparent precursor droplets\nflew off from the tip of the needle without developing Rayleigh instability\nor electrospraying. The distance between these droplets was too large\nto be observed using SEM. However, they significantly affected the\nelectric measurement, as shown in Figure S11 . This observation provides a pathway to remarkably increase the\nconductivity of the fiber mats by doping them with small droplets\nthat potentially would not impact the porosity or mechanical properties\nof the fiber mats. Electrical Properties of PEDOT:PSS-Based Fiber Mats as a Function\nof Relative Humidity Understanding the performance of PEDOT:PSS-based\nfibers after exposure to different relative humidities is critical\nto knowing the different environments in which they could be used.\nFor these experiments, we selected LowPEDOT3 fiber mats for preliminary\nstudies due to the ease of sample preparation (i.e., no pH value modification\nto the solution precursor). As can be seen from Figure 6 A, as the relative humidity increased up\nto 70%, the electrical current of the as-spun fiber mats decreased\nslightly, whereas it increased significantly at 90%. As the relative\nhumidity increased (up to 70%), the absorption of water molecules\nby PSS swells the PSS chains and thus increases the distance between\nPEDOT chains making it more difficult for electrons to hop. 62 , 63 However, when the relative humidity increased above 90%, the ambient\nwater condensed on the fiber mat and partially dissolved them due\nto their high content of water-soluble PVA, which led to a significant\nincrease in the electrical current. Figure 6 Current ( I )–voltage\n( V ) data for LowPEDOT3 fiber mats measured under\nvacuum. (A) Effect\nof relative humidity (RH) on as-spun fiber mats. (B) Effect of cross-linking\non the conductivity of four different fiber mats. Filled and open\ncircles represent as-spun and cross-linked fiber mats, respectively.\n(C) Conductivity of cross-linked fiber mats before and after immersion\nin aqueous solutions with pH values of 5, 7, 11, and 14. Thus, for situations where dissolution should be\navoided, we cross-linked\nthe LowPEDOT3 fibers with glutaraldehyde vapor and evaluated the effect\nof cross-linking. As can be seen from Figure S12 , the cross-linking did not have any impact on fiber mat appearance\nincluding area coverage, fibrous morphology, and transparency (i.e.,\nrelative thickness). Moreover, as can be seen from the overlaid data\nplotted in Figure 6 B for four different fiber mats, the electric current was equivalent\nfor as-spun (closed symbols) and cross-linked (open symbols) fiber\nmats indicating that the cross-linking process did not impact the\nelectrical conductivity either. Stability of PEDOT:PSS-Based Fiber Mats’ Electrical Properties\nPostwashing and Immersion Experiments We also explored if\nthere was any impact on electrical current after immersing the cross-linked\nLowPEDOT3 fiber mats in aqueous solutions with different pH values.\nWe note that the as-spun fiber mats could not be used in these experiments,\nas they immediately dissolved upon immersion in solution. As can be\nseen from Figures 6 C and S13 , the electrical current decreased\nby ∼60–70% compared to cross-linked fiber mats after\nimmersion in solutions that had pH values of 5, 7, and 11. However,\nafter immersion in a pH 14 solution, the current decreased by 3 orders\nof magnitude. Increasing the pH value could have potentially changed\nthe conformation and therefore, reduced the conductivity of the PEDOT\npolymer; 46 however, our Raman analysis\nfor fibers postimmersion in solutions with pH values from 5 to 11\ndid not support this. We note that because immersing the fibers in\npH = 14 solutions caused the cross-linked fibers to release from their\nsubstrates, they could not be analyzed using Raman. We hypothesize\nthat the electric current might have decreased due to a decrease in\nthe quantity of well-adhered fibers, diffusion of PEDOT:PSS from the\ncross-linked fibers into solution, and/or decrease in the conductivity\nof PEDOT postimmersion due to dedoping and/or change in conformation\nof PEDOT. We performed Raman spectroscopy on as-spun and cross-linked\nfiber mats, as well as on fiber mats postimmersion in solutions with\npH values from 5 to 11 (see Figure S14 ).\nFor all of the fiber mats, we observed a peak at 1430 cm –1 , which indicates the presence of PEDOT within these fiber mats.\nWe also observed that the width of the peak at 1430 cm –1 was the same for all fiber mats, regardless of the pH value of their\nprecursor or the pH value of the immersion solution. This result is\nconsistent with our electrical data as the pH value of precursor did\nnot affect the electrical properties of fiber mats. Moreover, as the\nwidth of the peak remained the same postimmersion in pH 5 to 11 solutions,\nwe suggest that the decrease in the current is most likely not due\nto the change in the conformation of PEDOT. We also evaluated\nthe electrical performance of the cross-linked\nLowPEDOT3 fiber mats before and after washing with soap water as shown\nin Figure S15 for two different samples.\nFor Sample 1 , the electrical current reduced by 40%\nafter the first wash compared to the cross-linked fibers. However,\nupon the second wash, the reduction in the current was 18% compared\nto the first washing. Similarly, for Sample 2 , the\nreduction in the electrical current was 55% and 20% for the first\nand second washes, respectively. Upon washing, it is possible that\nthe PEDOT:PSS might diffuse out of the fibers because it is most likely\nphysically trapped within the fibers and not chemically immobilized.\nThis will affect the electrical stability of the fibers upon washing\nas well as upon immersion in different pH solutions. While ideally\nthese fabrics could be washed many times without any loss, this preliminary\nevaluation suggests that the current decreased with subsequent washings\nand that the fabrics might be more appropriate for single wear applications\n(such as bandages). Notably, cross-linking (the PVA component of the\nfibers) enabled the ability to wash and reanalyze the same fiber mats.\nFuture work could explore alternative cross-linking methods that completely\navoid leakage of PEDOT from fibers because that would improve the\nmat’s durability and electrical stability postwashing. Electrical Properties after Long-Term Storage and Multiple Cycles\nof Current–Voltage Applications We also examined the\nstability of the as-spun LowPEDOT3 and LowPEDOT7 fiber mats. Even\nafter more than 3000 cycles of voltage applications, the current remained\nthe same for both the LowPEDOT3 or LowPEDOT7 fiber mats (see Figure 7 ). These fiber mats\nwere stored at room temperature and ambient relative humidity for\n3 months before being reevaluated. Even though electrical current\ndecreased for both LowPEDOT3 and LowPEDOT7 fiber mats, the reduction\nwas ∼50% for LowPEDOT7 fiber mats compared to >75% for LowPEDOT3\n(see Figure 7 , bottom).\nThis long-term storage study indicates that more stable fiber mats\nwere obtained from precursor electrospun from the less acidic precursors. Figure 7 Electrical\nstability of as-spun (A) LowPEDOT3 and (B) LowPEDOT7\nfiber mats (top) over 3000 cycles of voltage application as well as\n(bottom) multiple months of storage."
} | 6,286 |
22955625 | null | s2 | 3,518 | {
"abstract": "Hydrogels are used as scaffolds for tissue engineering, vehicles for drug delivery, actuators for optics and fluidics, and model extracellular matrices for biological studies. The scope of hydrogel applications, however, is often severely limited by their mechanical behaviour. Most hydrogels do not exhibit high stretchability; for example, an alginate hydrogel ruptures when stretched to about 1.2 times its original length. Some synthetic elastic hydrogels have achieved stretches in the range 10-20, but these values are markedly reduced in samples containing notches. Most hydrogels are brittle, with fracture energies of about 10 J m(-2) (ref. 8), as compared with ∼1,000 J m(-2) for cartilage and ∼10,000 J m(-2) for natural rubbers. Intense efforts are devoted to synthesizing hydrogels with improved mechanical properties; certain synthetic gels have reached fracture energies of 100-1,000 J m(-2) (refs 11, 14, 17). Here we report the synthesis of hydrogels from polymers forming ionically and covalently crosslinked networks. Although such gels contain ∼90% water, they can be stretched beyond 20 times their initial length, and have fracture energies of ∼9,000 J m(-2). Even for samples containing notches, a stretch of 17 is demonstrated. We attribute the gels' toughness to the synergy of two mechanisms: crack bridging by the network of covalent crosslinks, and hysteresis by unzipping the network of ionic crosslinks. Furthermore, the network of covalent crosslinks preserves the memory of the initial state, so that much of the large deformation is removed on unloading. The unzipped ionic crosslinks cause internal damage, which heals by re-zipping. These gels may serve as model systems to explore mechanisms of deformation and energy dissipation, and expand the scope of hydrogel applications."
} | 453 |
30918755 | PMC6428038 | pmc | 3,519 | {
"abstract": "Reef-forming cnidarians are extremely susceptible to the “bleaching” phenomenon caused by global warming. The effect of elevated seawater temperature has been extensively studied on Anthozoans; however, to date the impact of thermal stress on the expression of genes and proteins in Hydrozoan species has not been investigated. The present study aimed to determine the differential proteomic profile of Millepora alcicornis , which inhabits the Mexican Caribbean, in response to the El Niño-Southern Oscillation 2015–2016. Additionally, the cytolytic activity of the soluble proteomes obtained from normal and bleached M. alcicornis was assessed. Bleached specimens showed decreased symbiont’s density and chlorophyll a and c2 levels. After bleaching, we observed a differential expression of 17 key proteins, tentatively identified as related to exocytosis, calcium homeostasis, cytoskeletal organization, and potential toxins, including a metalloprotease, a phospholipase A2 (PLA2), and an actitoxin. Although, some of the differentially expressed proteins included potential toxins, the hemolytic, PLA2, and proteolytic activities elicited by the soluble proteomes from bleached and normal specimens were not significantly different. The present study provides heretofore-unknown evidence that thermal stress produces a differential expression of proteins involved in essential cellular processes of Hydrozoan species. Even though our results showed an over-expression of some potential toxin-related proteins, the cytolytic effect (as assessed by hemolytic, PLA2, and caseinolytic activities) was not increased in bleached M. alcicornis , which suggests that the cytolysis is mainly produced by toxins whose expression was not affected by temperature stress. These findings allow hypothesizing that this hydrocoral is able to prey heterotrophically when suffering from moderate bleaching, giving it a better chance to withstand the effects of high temperature.",
"conclusion": "Conclusions The present study represents the first report of the effect of thermal stress on the proteomic profile of a reef forming cnidarian of the class Hydrozoa. The decrease in symbiont’s density and chlorophyll a and c2 levels suggests that ENSO 2015–2016 induced a moderate bleaching in colonies of M. alcicornis that inhabit the Mexican Caribbean. Proteins involved in various important cellular processes, such as calcium homeostasis, exocytosis, and cytoskeleton organization were differentially expressed in bleached hydrocorals. Three of the differentially expressed proteins showed amino acid sequence similarity to potential toxins, however, the hemolytic, PLA2, and proteolytic activity of bleached M. alcicornis soluble proteome was not modified. These results suggests that the cytolytic effect induced by this hydrocoral is produced by toxins whose synthesis is not altered after bleaching, which allow us to hypothesize that M. alcicornis is able to prey heterotrophically when suffering from moderate bleaching, giving it a better chance to face the effects of high temperature.",
"introduction": "Introduction Coral reefs play a critical role in marine ecology and human sustainability ( Jain, Sonawane & Mandrekar, 2008 ; Mumby & Van Woesik, 2014 ; Anthony, 2016 ; Williams et al., 2016 ). Moreover, the organisms that constitute these ecosystems are considered a rich source of novel bioactive agents with considerable pharmaceutical and biotechnological potential, since they produce a great variety of molecules with unique structural features ( Jain, Sonawane & Mandrekar, 2008 ; Cragg & Newman, 2013 ; Mayer et al., 2016 ; Li, Ding & Li, 2017 ). Organisms of the genus Millepora , which belong to the class Hydrozoa , are considered the second most important reef forming species, after scleractinian hard corals of the class Anthozoa ( Lewis, 2006 ; Ruiz-Ramos, Weil & Schizas, 2014 ). In addition to their ecological importance, these organisms, commonly known as “fire corals,” induce local and systemic toxic effects in humans ( Lewis, 2006 ; Rojas-Molina et al., 2012 ). Unfortunately, coral reefs are extremely vulnerable to the stress related to greenhouse gas emissions, mainly ocean warming ( Lesser, 2007 ; Baird et al., 2009 ; Mumby & Van Woesik, 2014 ; Lough, 2016 ). Elevated seawater temperature provokes disturbances that can seriously affect and break down the homeostatic capacity of coral reefs to overcome stressors ( Mora, Graham & Nyström, 2016 ). One of the most devastating consequences of global warming is coral bleaching, in which corals and hydrocorals lose their photosynthetic symbiotic algae of the genus Symbiodinium or their pigments, which exposes the white exoskeleton composed of calcium carbonate ( Hoegh-Guldberg, 1999 ; Lesser, 2006 ; Bonesso, Leggat & Ainsworth, 2017 ; Neal et al., 2017 ). The frequency and seriousness with which this phenomenon occurs have increased in recent years. Massive bleaching events have been recorded in all the tropical regions of the world in 1987, 1998, 2003, 2005, and 2010 ( Eakin et al., 2010 ; Hoegh-Guldberg & Bruno, 2010 ; Heron et al., 2016 ). In fact, during 2014–2017, the worst documented bleaching event was observed ( Eakin et al., 2017 ; Hughes et al., 2017b , 2018 ). Numerous investigations have demonstrated that bleaching is deleterious to coral reefs. Bleaching events have caused massive damage to coral reefs around the world, with very severe effects on the balance of biodiversity in marine tropics. This phenomenon causes a general deterioration in reef health, as it provokes an increase in coral diseases, a decay in reef calcification, breakdown of reef framework by bioeroders, and loss of essential habitat for related reef organisms ( Baker, Glynn & Riegl, 2008 ; Lesser, 2011 ; Grottoli et al., 2014 ; Okazaki et al., 2016 ; Hughes et al., 2017a ). Some climate models predict that if CO 2 emissions continue at the current rate, bleaching events will increase its frequency and severity, seriously threatening the survival of coral reefs ( Houlbreque & Ferrier-Pagès, 2009 ; Lesser, 2011 ; Mumby & Van Woesik, 2014 ; Lough, 2016 ; Neal et al., 2017 ; Hughes et al., 2018 ). Studies that have addressed the etiology and effects of bleaching have focused on Anthozoan species. To date there are no reports regarding the impact of thermal stress on the expression of genes and proteins of Hydrozoan species. Genomic, transcriptomic, and proteomic approaches have revealed that heat stress causes a differential pattern in the expression of genes and proteins from various cellular processes, including oxidative stress response, Ca 2+ homeostasis, cytoskeletal organization, protein synthesis, apoptosis, endo-exophagocytosis, and immune response. Also, the differential expression of heat shock proteins and transcription factors has been observed ( Bay & Palumbi, 2015 ; Pinzón et al., 2015 ; Raina et al., 2015 ; Seneca & Palumbi, 2015 ; Weston et al., 2015 ; Maor-Landaw & Levy, 2016 ; Oakley et al., 2016 , 2017 ; Ricaurte et al., 2016 ; Sloan & Sawyer, 2016 ; Huang et al., 2017 ; Ruiz-Jones & Palumbi, 2017 ; Traylor-Knowles et al., 2017 ; Mayfield et al., 2018a ). It is generally accepted that cnidarian- Symbiodinium symbiosis is fundamental for the formation of coral structures, since algae carry out photosynthesis and transfer more than 50% of their photosynthetic products to the cnidarian host ( Houlbreque & Ferrier-Pagès, 2009 ; Davy, Allemand & Weis, 2012 ; Fransolet, Roberty & Plumier, 2012 ; Douglas, 2003 ; Venn, Loram & Douglas, 2008 ; Yellowlees, Rees & Leggat, 2008 ). Up to date, the role of heterotrophic carbon (C) input in the resilience and recovery of bleached reef forming cnidarians is still controversial. It has been observed that during bleaching certain species can increase their heterotrophic feeding to maintain and restore energy reserves ( Grottoli, Rodrigues & Palardy, 2006 ; Aichelman et al., 2016 ). However, some studies suggest that heterotrophically derived fixed carbon does not completely attenuate C budget imbalance ( Levas et al., 2016 ; Tremblay et al., 2016 ). Undoubtedly, after a bleaching event the central metabolism of both symbiotic partners is significantly altered ( Ricaurte et al., 2016 ; Hillyer et al., 2017 , 2018 ; Oakley et al., 2017 ; Mayfield et al., 2018a ), and the contribution of heterotrophy in the response of cnidarian- Symbiodinium symbiosis to elevated temperature is still unclear ( Grottoli, Rodrigues & Palardy, 2006 ; Levas et al., 2016 ; Aichelman et al., 2016 ; Tremblay et al., 2016 ). Considering that toxins are essential in prey capture and digestion, it is very likely that thermal stress, which provokes bleaching, alters their expression in cnidarians. Therefore, assessment of the effect of thermal stress on the expression of cnidarian toxins used to capture preys is fundamental to understand if heterotrophy constitutes a significant survival mechanism of reef-forming cnidarians after a bleaching phenomenon. In this context, the aim of the present study was to determine changes in the soluble proteomic profile and cytolytic activity of Millepora alcicornis (“fire coral”) that suffered bleaching during El Niño-Southern Oscillation (ENSO) 2015–2016 in the Mexican Caribbean Sea.",
"discussion": "Discussion In recent decades, increases in the frequency and intensity of coral bleaching events have resulted in declines in coral cover worldwide ( Roff, Zhao & Mumby, 2015 ; Lough, 2016 ). Accordingly, these damaged ecosystems have suffered a dramatic “phase shift,” in which habitats dominated by reef-forming organisms are now dominated by macroalgae and low-relief corals ( Jackson et al., 2014 ; Cramer et al., 2017 ). In fact, coral reefs are considered the most degraded ecosystems on the planet due to climate change ( Eakin et al., 2010 ; Roff, Zhao & Mumby, 2015 ; Cramer et al., 2017 ). During 2014–2017, record high temperatures triggered the third global scale event of coral bleaching ever registered ( Eakin et al., 2017 ; Hughes et al., 2017b , 2018 ). It is estimated that 30% of the reef areas that were monitored after the ENSO 2015–2016 showed severe bleaching ( Hughes et al., 2018 ), and the Caribbean coral reefs were seriously affected. Hydrocorals of the genus Millepora spp. are among the organisms most affected by bleaching ( Dias & Gondim, 2016 ). Particularly, M. alcicornis has been considered very vulnerable to temperature rise, since it has suffered severe bleaching episodes in different times and locations, such as in the Great Barrier Reef ( Marshall & Baird, 2000 ), where an 85% mortality of this species was observed during the summer of 2005; in the Florida Keys reef in the summers of 2006 and 2007 ( Wagner, Kramer & Van Woesik, 2010 ), and in Puerto Rico and the Caribbean during 1987, 1993, 1995, 1998, 2003, 2005, 2009, and 2010 ( Dias & Gondim, 2016 ). In the present study, normal and visibly bleached fragments of the “fire coral” M. alcicornis , ( Figs. 1A and 1B ) were collected in the Parque Nacional Arrecife de Puerto Morelos (Quintana Roo, México) in November 2016, the warmest year in NOAA’s 137-year series ( NOAA, 2017 ). Since bleaching is provoked by a decrease in symbionts population and a reduction in the concentration of their photosynthetic pigments ( Hoegh-Guldberg, 1999 ; Lesser, 2006 , 2011 ; Obura, 2009 ), chlorophyll contents, and symbiont’s density were determined in M. alcicornis . As was expected, chlorophyll a and chlorophyll c2 concentrations per cm 2 of hydrocoral were significantly lower in the bleached specimens ( Fig. 1C ), which indicates that warming provoked breakdown of the hydrocoral-algae symbiosis due to photoinhibition, in a similar way as that observed in Anthozoan species ( Warner, Fitt & Schmidt, 1999 ; Takahashi & Murata, 2008 ; Lesser, 2011 ). Additionally, bleached M. alcicornis specimens showed a decrease of 40% in the density of symbionts per square centimeter ( Fig. 1D ), which corresponds to a “moderate bleaching” according to the categories used in the ReefBase ( Oliver, Berkelmans & Eakin, 2009 ). Unlike what has been previously observed, the M. alcicornis colonies examined in the present study after the bleaching event of 2015–2016, did not show severe bleaching. This could indicate that this Caribbean hydrocoral is developing a certain degree of thermotolerance, which might be associated with the switching of symbionts, as reported in other species ( Silverstein, Cunning & Baker, 2015 ; Bay et al., 2016 ; Swain et al., 2018 ). However, this hypothesis should be proven in future studies. Comparative proteomic studies have been used to study the effect of heat stress on the total proteome from some Anthozoan species that include: Acropora palmata , Acropora microphthalma , Pocillopora acuta, Seriatopora hystrix , and Aiptasia pallida . The results derived from such investigations have shown that thermal stress, which is responsible for coral bleaching, provokes the differential expression of several proteins involved in different processes essential for the survival of the cnidarians, including cytoskeleton organization, thermal and UV stress response, redox state, immunity, calcium homeostasis, transcription factors, exocytosis, and metabolism ( Weston et al., 2015 ; Ricaurte et al., 2016 ; Oakley et al., 2017 ; Mayfield et al., 2018a , 2018b ). However, to date there are no reports about the effect of breakdown of the cnidarian- Symbiodinium symbiosis due to thermal stress on the expression of proteins, including toxins, in an organism of the class Hydrozoa. Therefore, in this study the differential expression of proteins in the soluble proteomes from normal and bleached M. alcicornis specimens was analyzed. The extraction of high-quality proteins from corals with calcareous exoskeleton is not an easy task ( Cheng et al., 2018 ). The most common methods reported for protein extraction include the use of trizol ( Garcia et al., 2016 ; Mayfield et al., 2018a ), glass bead-assisted extraction ( Weston et al., 2015 ), and sonication-assisted extraction with rehydration buffer ( Ricaurte et al., 2016 ). However, the method chosen for the extraction of M. alcicornis proteins involved osmotic shock in bidistilled water. This method causes the discharge of the nematocysts content ( Ibarra-Alvarado et al., 2007 ; García-Arredondo et al., 2015 , 2015 , 2016 ; Morabito et al., 2012 ; Hernández-Matehuala et al., 2015 ). Sodium dodecyl sulfate–polyacrylamide gel electrophoresis evidenced that soluble proteomes from normal and bleached M. alcicornis contain proteins with a broad range of molecular weights ( Fig. 2 ). Most of the bands detected in both electrophoretic profiles are within the range of 10–200 kDa. A similar electrophoretic pattern was observed for the proteome of isolated nematocysts from Hydra magnipapillata , (a well-known member of the Hydrozoa class), in which most of the protein bands are in the range of 10–70 kDa ( Balasubramanian et al., 2012 ). Some differences were observed in the SDS–PAGE profiles obtained from normal and bleached hydrocorals, indicating a differential protein expression between both types of specimens. Moreover, 2DE-PAGE showed that the soluble proteomes from normal and bleached specimens of M. alcicornis resolved 58 and 75 proteins, respectively. It is important to mention that in the case of the present investigation, we examined the effect of elevated seawater temperature on the soluble proteome of M. alcicornis , unlike what has been done in previous studies carried out on Anthozoan species, in which the impact of thermal stress on the total proteomes from Acropora palmata , A. microphthalma , Pocillopora acuta, Seriatopora hystrix , and Aiptasia pallida was analyzed ( Weston et al., 2015 ; Ricaurte et al., 2016 ; Oakley et al., 2017 ; Mayfield et al., 2018a , 2018b ). After bleaching, the differential expression of 17 proteins was observed in M. alcicornis , six proteins were up-regulated, while 11 were down-regulated. Calmodulin, actin, and collagen are some of the proteins differentially expressed in bleached specimens of M. alcicornis. Previous studies have reported the differential expression of these same proteins, which are involved in calcium homeostasis, cytoskeleton, and extracellular matrix (ECM) in Anthozoan species subjected to bleaching ( Bay & Palumbi, 2015 ; Pinzón et al., 2015 ; Raina et al., 2015 ; Seneca & Palumbi, 2015 ; Weston et al., 2015 ; Maor-Landaw & Levy, 2016 ; Oakley et al., 2016 , 2017 ; Ricaurte et al., 2016 ; Sloan & Sawyer, 2016 ; Huang et al., 2017 ; Ruiz-Jones & Palumbi, 2017 ; Traylor-Knowles et al., 2017 ; Mayfield et al., 2018a ). Calmodulin is a key Ca 2+ sensor, whose signaling is important in numerous cellular processes, such as cell cycle and calcium homeostasis ( Haeseleer et al., 2002 ; Desalvo et al., 2008 ; Reyes-Bermudez, Miller & Sprungala, 2012 ). Several studies have demonstrated that expression of genes and proteins involved in Ca 2+ homeostasis is modified after bleaching in Acropora palmata , Orbicella faveolata , Acropora millepora , and Galaxea astreata ( Desalvo et al., 2008 , 2010 ; Rodriguez-Lanetty, Harii & Hoegh-Guldberg, 2009 ; Moya et al., 2012 ; Ricaurte et al., 2016 ; Huang et al., 2018 ). In the case of M. alcicornis , CaM was up-regulated after bleaching, which implies that changes are occurring in biomineralization and other calcium-dependent processes related to the development of the hydrocorals. It has been shown that CaM overexpression and a rupture in Ca 2+ homeostasis are linked to an abnormal reorganization of the actin cytoskeleton ( Desalvo et al., 2008 ; Reyes-Bermudez, Miller & Sprungala, 2012 ). Actin was another protein that was up-regulated in bleached M. alcicornis specimens, in a similar way to what was found in scleractinian corals such as Acropora palmata, Stylophora pistillata , and Montastraea faveolata ( Desalvo et al., 2008 ; Kenkel, Meyer & Matz, 2013 ; Maor-Landaw et al., 2014 ; Ricaurte et al., 2016 ; Louis et al., 2017 ). Actin is implicated in the construction of filaments and supports the majority of motile events in eukaryotic cells ( Fletcher & Mullins, 2010 ). The increase in the expression of actin observed in bleached hydrocorals might be related to the relocation of symbionts as a mechanism of photoprotection ( Petrou, Ralph & Nielsen, 2017 ) and to the increase of heterotrophic feeding to mitigate the energetic imbalance by the departure of symbionts ( Wooldridge, 2014 ; Tremblay et al., 2016 ), which requires a significant contribution by the cytoskeleton to collect and transport nutrients. Among the proteins that showed a down-regulation in bleached hydrocorals was collagen. This result is in agreement with what was found in specimens of Acropora palmata, Stylophora pistillata , and Anemonia viridis that were subjected to bleaching conditions ( Moya et al., 2012 ; Maor-Landaw & Levy, 2016 ; Ricaurte et al., 2016 ). Collagen, abundantly found in the calicoblastic space, is central during the biomineralization process since it provides the structural support to which proteins of the subfamily CARP and analog proteins bind in nucleation and mineral growth sites ( Mohamed et al., 2016 ). Moreover, a decline in symbiont´s density has been related to changes in the active volume of calcification due to a reduction in dissolved inorganic carbon ( Drake et al., 2013 ; D’Olivo & McCulloch, 2017 ). Thus, down-regulation of collagen and reduced symbiont´s density suggest that thermal stress induces a decrease in the rate of calcification in the hydrocoral M. alcicornis . The exocytosis multiprotein complex provides spatial targeting of exocytotic vesicles to the plasma membrane ( Picco et al., 2017 ), and therefore is related to the process of symbionts expulsion ( Bieri et al., 2016 ). In fact, the exocytosis complex component 4 was up-regulated in Porites astreoides , and has been proposed as a biomarker of coral thermal stress ( Kenkel, Meyer & Matz, 2013 ; Kenkel et al., 2014 ). In contrast, exocytosis complex component 3 was down-regulated in bleached M. alcicornis , as previously reported for other Endo-exo phagocytosis-related proteins in Acropora palmata ( Ricaurte et al., 2016 ). Thus, the proposal that up-regulation of proteins that are part of the exocytosis complex represents a biomarker of thermal stress is still controversial. Three of the soluble proteins from M. alcicornis , which were differentially expressed in bleached hydrocorals, showed amino acid sequence similarity to potential toxins. Despite the ecological relevance of reef forming hydrocorals, the information regarding the chemical structure and mechanism of action of their toxins is scarce. Previous studies carried out by our research group showed that the aqueous extract of M. alcicornis caused hemolysis of rat erythrocytes and showed PLA2 activity ( Hernández-Matehuala et al., 2015 ). Zymographic analysis of this extract revealed that it contained ∼28–30 kDa cytolysins with PLA2 activity; ∼200 kDa cytoloysins, which do not have PLA2 activity ( Hernández-Matehuala et al., 2015 ) and proteins with proteolytic activity at ∼25, ∼40, and 80–200 kDa (Olguín-López Norma, Hernández-Elizarraga Víctor, Hernández-Matehuala Rosalina, Cruz-Hernández Andrés, Guevara-González Ramón, Caballero-Perez Juan, Ibarra-Alvarado César , and Rojas-Molina Alejandra, 2018, unpublished data). Interestingly, we found that bleaching induced up-regulation of proteins that showed sequence homology with DELTA-actitoxin-Ate1a-like protein from the sea anemone Exaiptasia pallida ( Sunagawa et al., 2009 ) and a putative PLA2 from Hydra magnipapillata ( Sher et al., 2005 ). In contrast, bleaching elicited down-regulation of a protein that has sequence similarity with a ATP-dependent Zinc metalloprotease from Aiptasia pallida ( Sunagawa et al., 2009 ). Cnidarian toxins are mainly proteins and peptides ( Anderluh & Maček, 2002 ; Nevalainen et al., 2004 ; Mariottini et al., 2015 ; Podobnik & Anderluh, 2017 ) and can be classified into three categories based on their mechanism of action: enzymes (PLA2 and metalloproteinases); pore-forming proteins (actinoporins, jellyfish toxins, hydralysins-related toxins, and membrane attack complex-perforins), and neurotoxins ( Jouiaei et al., 2015b ). The most studied pore forming toxins of cnidarians are actinoporins, that have been found in many species of marine anemones ( Anderluh & Maček, 2002 ; Kristan et al., 2009 ; Šuput, 2009 ). One of the proteins whose expression was up-regulated after bleaching displayed sequence similarity with actinoporin DELTA-actitoxin-Ate1a-like protein. This finding suggests that M. alcicornis produce PFTs, whose structure might be related to that of the anemone actinoporins, which is interesting considering that this type of toxins have been found in only one species of the class Hydrozoa, Hydra magnipapillata ( Jouiaei et al., 2015b ). On the other hand, the expression of a protein that exhibited sequence similarity to a putative PLA2 of 15.7 kDa from Hydra magnipapillata , which was found by a large-scale search at the Hydra magnipapillata EST Database, was up-regulated after bleaching. This Hydra PLA2 showed homology with PLA2s from Apis mellifera , the lizard Heloderma uspectum (“Gila Monster”), and the scorpions Mesobuthus tamulus and Pandinus imperator , mainly in the calcium binding domain and in the catalytic site. All of these PLA2s display neurotoxic activity and belong to the secreted PLA2 (sPLA2) group III (sPLA2-III) ( Sher et al., 2005 ). In contrast to snake venom sPLA2s ( Costa, Camargo & Antunes, 2015 ; Kordiš & Križaj, 2017 ; Tonello & Rigoni, 2017 ), cnidarian sPLA2s have been poorly characterized and their enzymatic action greatly differs between species ( Nevalainen et al., 2004 ). The present study also evidenced that proteins that displayed sequence similarity to disintegrin and metalloproteinase domain-containing protein 7 and an ATP-dependent Zinc metalloprotease showed down-regulation after bleaching. Metalloproteinases have been detected in some cnidarian nematocyst venoms, such as those of Stomolophus meleagris ( Li et al., 2014 ), Olindias sambaquiensis ( Knittel et al., 2016 ), Pelagia noctiluca ( Frazão et al., 2017 ), and Chironex fleckeri ( Jouiaei et al., 2015a ). This class of toxins shows a great diversity of effects. Snake venom metalloproteinases (SVMPs) affect the ECM in various ways, including the release of ECM-derived biologically-active peptides, which exerts either reparing or damaging effects on tissues ( Gutiérrez et al., 2016 ). These enzymes are also involved in the proteolytic processing of other venom proteins ( Da Silveira et al., 2007 ) and are capable of causing severe inflammation by disrupting capillary vessels and tissues ( Weston et al., 2013 ). However, the different kinds of cnidarian metalloproteinases and their effects are not yet clear. It has been demonstrated that SVMPs undergo various post-translational modifications, which contribute to their great functional diversity ( Moura-da-Silva et al., 2016 ). This could be the case of cnidarian metalloproteinases. If this is so, then it could be hypothesized that the imbalance of energy produced by symbionts expulsion or death induce cnidarians to optimize their energy expenditure, which impairs the metalloproteases synthesis that represents a great energy investment. Cnidarians employ toxins as a defense against predators and in prey capture to obtain heterotrophic nutrition ( Anderluh & Maček, 2002 ; Anderluh et al., 2011 ; Houlbrèque, Rodolfo-Metalpa & Ferrier-Pagès, 2015 ; Ben-Ari, Paz & Sher, 2018 ). Particularly, both, PFTs and PLA2s produce cytolysis and play a very important role in defensive and offensive actions, since they provoke the destruction of cell membranes ( Rojko et al., 2016 ; Macrander & Daly, 2016 ; Podobnik & Anderluh, 2017 ). It has been documented that certain corals, such as Oculina arbusculata , can increase their heterotrophic nutrition to maintain and restore energy reserves after a bleaching event ( Grottoli, Rodrigues & Palardy, 2006 ; Aichelman et al., 2016 ). In the case of the present study, the differential expression of potential toxins suggests that thermal stress alters heterotrophic competences to mitigate the energy imbalance in hydrocorals. However, heat stress did not affect the hemolytic, PLA2, and proteolytic activity of M. alcicornis soluble proteome, which suggests that the cytolysis induced by the toxins of this hydrocoral is mainly produced by enzymes and PFTs ( Nevalainen et al., 2004 ; Lee et al., 2011 ; Anderluh et al., 2011 ; Jouiaei et al., 2015a ; Rojko et al., 2016 ; Knittel et al., 2016 ), whose expression was not affected after suffering moderate bleaching. It is worth mentioning that in a previous study carried out on M. alcicornis specimens collected in the Mexican Caribbean in 2008, we found that thermal stress did not significantly modify the PLA2 activity of the aqueous extract prepared from bleached M. alcicornis ( García-Arredondo et al., 2015 ), which agrees with what we found in the present study with specimens collected in 2016. These findings seem to indicate that M. alcicornis subjected to moderate bleaching does not lose its capability to synthesize cytolysins. Evidently, it is important to continue investigating the impact of elevated ocean temperatures that provoke bleaching on the synthesis of toxins produced by reef forming cnidarians and the significance of heterotrophic feeding as a mechanism to counteract the deleterious effects of this phenomenon."
} | 6,971 |
35424748 | PMC8982352 | pmc | 3,520 | {
"abstract": "Multi-stimuli-responsive hydrogels are intelligent materials that present advantages for application in soft devices compared with conventional machines. In this paper, we prepared a bilayer hydrogel consisting of a poly(2-(dimethylamino)ethyl methacrylate) layer and a poly( N -isopropylacrylamide) layer. The hydrogel responded to temperature, pH, NaCl, and ethanol by undergoing bending deformation. At 40 °C, it only took 23 s for the hydrogel to bend nearly 300°. Carbon black was also introduced into the hydrogel network to render it conductive. Based on its multi-stimuli-responsive properties and conductivity, the hydrogel was used to construct a 4-arm gripper, thermistor, and finger movement monitor. The time required to grip and release an object was 141 s. The resistance changed with temperature, which affected the brightness of an LED. Finger motions were monitored, and the bending angle could be distinguished.",
"conclusion": "4. Conclusions PDMAEMA/PNIPAM bilayer hydrogel and the single-layer hydrogels corresponding to each layer were designed and prepared. The stimuli-responsive properties and deformation mechanisms of the bilayer hydrogel were studied via the swelling/de-swelling state of the single-layer hydrogels and bending of the bilayer hydrogel. The results indicated that the bilayer hydrogel was responsive to temperature, pH, salt, and ethanol and reverted without intense or prolonged simulation. Additionally, the stimulus that generated the fastest response was temperature, and the SDS concentration and temperature were also explored. A high concentration of SDS slowed the response speed, while a high temperature accelerated it. The bilayer hydrogel was rendered conductive by adding uniformly dispersed CB. When the hydrogel was deformed, the density of the conductive material per unit volume increased or decreased, which affected the resistance of the hydrogel. Due to the stimuli-responsive and conductive properties, we developed a 4-arm gripper, thermistor, and finger movement monitor using the hydrogel, which could grip and release, change their resistance with temperature, and monitor finger motions, respectively. These multi-stimuli-responsive bilayer hydrogels presented great potential in soft robots, multi-stimuli-dependent resistors and human body monitors.",
"introduction": "1. Introduction Hydrogels are three-dimensional networks of soft materials composed of polymers and water molecules that present the properties of both solids and fluids. 1–3 By absorbing or exuding water molecules, hydrogels can swell or de-swell, resulting in deformation. In recent years, smart hydrogels have become a research hotspot due to their stimuli–responsive properties. 4–6 After adding functional molecules or monomers, smart hydrogels become responsive to one or more stimuli such as temperature, pH, light, magnetic fields, etc. , 7–11 resulting in various deformation modes, e.g. , shrinkage/expansion, 12–14 bending, 15–17 twisting, 18,19 or composite behaviors. 20–22 In recent decades, bilayer hydrogels have garnered considerable attention from researchers because they can undergo multiple deformation modes. The different components of the two layers result in independent responses to various stimuli. Inspired by Mimosa, Zheng et al. 23 prepared a bilayer hydrogel via a moulding method that achieved multi-stimuli-responsiveness because it contained poly(acrylic acid) and poly( N -isopropylacrylamide). A soft gripper was fabricated by simply connecting a hydrogel flower to cotton fiber. Xiao et al. 24 synthesized a bilayer hydrogel consisting of poly([2-(methacryloyloxy)ethyl]trimethylammonium chloride) and poly[ N -(2-hydroxyethyl)acrylamide], which was responsive to different salt types and salt concentrations. Furthermore, an eight-arm gripper made of a bilayer hydrogel could grasp and release an object. Wang et al. 25 prepared a strong-interfacial bilayer hydrogel consisting of a shape-memory layer and an elastic layer. The bilayer hydrogel underwent different deformations via various heating-stretching modes. Most double-layer hydrogels have multi-stimuli-responsive properties and can be used for grasping functions. 23,24,26,27 To develop additional applications, conductivity is required. Similar to single-layer hydrogels, methods to improve the conductivity of bilayer hydrogels include introducing conductive materials into hydrogels. Navaei et al. 28 prepared conductive gelatin-based hydrogels where gold nanorods promoted electrical conductivity and enhanced the mechanical stiffness. Xia and co-workers 29 fabricated a hydrogel-based wearable strain sensor consisting of polyacrylamide (PAM) and chitosan (CS). The conductivity of the hydrogel was improved by increasing the amount of carboxyl-functionalized multi-walled carbon nanotubes (MWCNTs), which ionically cross-linked with CS. In the work of Chen et al. , 30 a multifunctional conductive hydrogel was prepared in which polyaniline (PANI) provided electronic conductivity, and Fe 3+ and poly(4-styrenesulfonate) (PSS) provided ionic conductivity. Ying et al. 31 designed and synthesized an AIskin with ionic conductivity, presenting high stretchability and sensitivity. In addition to strain sensors, a promising application as soft diode was initially prepared and functioned. In this paper, a bilayer hydrogel consisting of poly( N -isopropylacrylamide) (PNIPAM) and poly(2-(dimethylamino)ethyl methacrylate) (PDMAEMA) layers was prepared. The bilayer hydrogel was rendered multi-stimuli-responsive by adding PNIPAM and PDMAEMA as functional monomers. 32–35 The bilayer hydrogel was tested for its response to multiple stimuli, including temperature, pH, salt, and ethanol. These stimuli caused the composite to bend due to the deformation of the two individual layers. In addition, the responsive behaviors of the bilayer hydrogel and each single-layer hydrogel and their response mechanisms to multiple stimuli are described. Further, the conductivity of the bilayer hydrogel was improved. In the PDMAEMA layer (D layer), the addition of PSS, a polyelectrolyte, provided ionic conductivity and electrostatic interactions that enhanced the stiffness. While in the PNIPAM layer, carbon black was added, which offered electronic conductivity. Compared to the works mentioned above, for multiple stimuli-responsive properties and conductivity, two individual properties of hydrogels, 4-arm gripper and strain sensor as conventional applications were fabricated. Additionally, this paper attempted to prepare an organic thermistor, which was a combination of stimuli-responsiveness and conductivity.",
"discussion": "3. Results and discussion 3.1 Preparation of hydrogels In this work, PDMAEMA/PNIPAM bilayer hydrogels were designed and prepared via the casting method. As shown in Fig. 1 , the first step was to prepare the D layer, and when the polymerization degree of the D layer approached the gel point (the viscosity of the D layer increased, and the hydrogel presented solid properties), the mixture of the N layer was added. Then, the mould was sealed for 12 h. In the D layer, since sulfonyl and amino groups carry opposite charges, electrostatic interactions were formed between PDMAEMA and PSS, forming a semi-crosslinked network. In the N layer, the addition of PVA improved the strength and toughness by forming hydrogen bonds between acrylamides in PNIPAM and hydroxyls in PVA. TEMED was used to promote the homolysis of APS, reducing the initiation temperature. CB was added to further enhance the mechanical strength and provide electrically-conductive properties, along with SDS as the dispersant and stabilizer. Fig. 1 Schematic illustration of bilayer hydrogel preparation. SEM images provided detailed information about the D layer, N layer, and their interface. As shown in Fig. 2a and b , the D layer and the N layer possessed different morphologies. In Fig. 2c , the interface between the two layers was clear. Fig. 2d additionally presented the differences in color. Fig. 2 SEM images of (a) D layer, (b) N layer and (c) the interface; (d) photo of the prepared bilayer hydrogel cut in a size of 4.2 mm × 2.2 mm × 1.3 mm. 3.2 Deformation mechanism and stimuli-responsive behavior of bilayer hydrogels The bending deformation of bilayer hydrogels depends on the swelling ratio of the two layers, which respectively present their unique properties; however, the bending of the bilayer hydrogel depends on the combined effect of the two layers. Therefore, multi-stimuli tests were conducted, resulting in two bending directions. In Fig. 3a , the positive (+) bending and negative (−) bending were artificially specified. When bending, the shape of the hydrogel was treated as the arc of a circle, as shown in Fig. 3b , which simplifies the data processing but slightly decreases the data accuracy. Additionally, the corresponding single-layer hydrogels were synthesized and tested, which provided supplementary data, as shown in Fig. S1. † The thickness of each single-layer hydrogel was less than 1 mm; therefore, the area change was calculated by S / S 0 , where S 0 is the original area, and S is the area after the application of a stimulus or stimuli. Fig. 3 (a) Positive and negative bending directions; (b) simplified model for calculating the center angle and curvature of the bent hydrogel ( θ is the center angle corresponding to the arc, and r is the radius of the circle). In the thermal-responsive test, the bilayer hydrogel bent positively when immersed in water at 40 °C. As Fig. 4a shows, it only took 23 s for the bilayer hydrogel to bend to nearly 300°, and the curvature increased from 0.29 to 1.36. Both PNIPAM and PDMAEMA are thermal-responsive polymers with different lower critical solution temperatures (LCSTs). When the temperature exceeds the LCST, the hydrophobic interactions between polymer chains strengthen, and the hydrogen bonds weaken, resulting in the rapid collapse of polymer chains. Macroscopically, this causes the hydrogel to shrink. The LCST of PNIPAM and PDMAEMA were around 32 °C and 40 °C, respectively; thus, the de-swelling degree of the N layer was greater than that of the D layer at the same temperature, which was confirmed by Fig. S1; † however, the recovery time was much longer, as shown in Fig. 4b . It took 1800 s to recover at room temperature, which can be explained intuitively as a rapid squeeze but slow re-absorption. 36 By comparing the shape of the hydrogel before and after heating, the hydrogel both bent and shrank, as shown in Fig. 4 and S1. † It should be pointed out that when the gel was formed, the polymer did not dissolve; thus, the LCST could not be satisfied by definition. Because the polymer chains still shrank or relaxed, the volume phase transition temperature (VPTT) was used to represent the thermo-sensitive properties of the hydrogel. Fig. 4 Center angles and curvatures of the bilayer hydrogel at 40 °C (a) and its recovery (b). Two factors, the temperature and amount of SDS in the CB dispersion (the original concentration of SDS was 0.75 mg mL −1 ), were investigated for their influence on the thermal-sensitive properties of hydrogels. A series of temperature gradients was set, as shown in Fig. S2. † To display the data more intuitively, the average rates of the center angle ( R a ) and curvature ( R b ) were used. 1 where θ e is the final value of the center angle, θ 0 is the initial value of the center angle, and t is the total time. 2 where C e is the final curvature, C 0 is the initial curvature, and t is the total time. It is shown in Fig. 5a that as the temperature rose, the average rate of both parameters increased, presenting a positive correlation. A higher temperature led to a more rapid increase in the entropy of water molecules, which resulted in faster breaking of hydrogen bonds between the amide groups and water molecules, which facilitated hydrophobic interactions among isopropyl groups. In terms of recoverability, excessive stimulus caused irreversible damage to the hydrogels. Fig. 5 (a) Average rate of the two parameters; (b) center angles and curvature change of bilayer hydrogel in solution with a higher concentration of SDS (1.5 mg mL −1 ). Increasing the SDS concentration to 1.5 mg mL −1 extremely prolonged the response time. In Fig. 5b , the change in the center angle and curvature maintained the same trend as the experiment above at 40 °C, but the response time increased nearly ten times, and the bending degree decreased compared to the original value. As for the recovery (Fig. S3 † ), the time was much shorter than in the original concentration. One reason for this was that a large number of SDS molecules blocked the hydrophobic interactions between isopropyl groups with alkanes at the end of SDS, which increased the VPTT. 37 Another reason was that SDS created repulsive electrostatic interactions that inhibited the collapse of PNIPAM chains. 38 These two factors did not change the bending direction, but they did affect the rate. In the pH-responsive test, 1 M hydrochloric acid (HCl) and 1 M NaOH solution were used to represent acidic and alkaline environments, respectively. In contrast, 1 M NaCl solution is used as a neutral environment, while maintaining the same ion concentration. When placed in acidic, alkaline, and neutral environments with the same ion concentration, the bilayer hydrogel bent towards the positive direction, but at different rates. In an acidic environment, the bilayer hydrogel bent positively. The center angle rose to 182°, and the curvature rose to 0.79 over 570 s, as shown in Fig. 6a . In an alkaline environment ( Fig. 6c ), it took 200 s for the center angle to reach 300° and for the curvature to rise to 1.09, which shows a shorter response time and higher bending degree, while the bending direction was unchanged. In a neutral environment, the hydrogel also bent positively. In Fig. 6e , the center angle and curvature increased more slowly than the former two, with values of 187° and 0.86, respectively. Fig. 6 The center angles and curvatures of bilayer hydrogel in solutions of 1 M HCl (a), 1 M NaOH (c), 1 M NaCl (e) and their respective recovery (b), (d), (f). The two layers had different response mechanisms. In the D layer, the equilibrium swelling behavior was related to the degree of ion dissociation. The pH of the solution directly affected the degree of ion dissociation of the electrolyte. In the PDMAEMA chains N , N -dimethylaminoethyl, generally became more hydrophilic after the protonation of its many amino groups. At the same time, electrostatic repulsion between the same ions led to the diffusion of polymer chains, causing swelling. Under alkaline conditions, the polymer chains agglomerated due to the deprotonation of the tertiary amino groups and hydrophobic interactions, resulting in the de-swelling of the hydrogel. In the N layer, the PNIPAM chains were ion-responsive. Lee et al. showed that some salts decrease the LCST of thermal-responsive polymers (the “salting-in effect”), and others increase the LCST (the “salting-out effect”), which is true for NaCl. 39 On the one hand, the addition of NaCl increased the hydrogen bonds between water molecules, therefore decreasing those between water and hydrophilic chains. On the other hand, it increased the polarity of water molecules, enhancing the hydrophobic interactions between polymer chains. These combined factors caused the PNIPAM hydrogel to shrink in environments with a high ion concentration (Fig. S1b † ). It can be judged from the curves in Fig. 6a, c, and e that the response rate slowed down over time. We speculate that upon increasing the concentration of ions in the hydrogel, the osmotic pressure difference between the inside and outside of the hydrogel decreased gradually, which blocked the penetration of additional ions. There were also fewer unreacted responsive groups over time. During recovery procedure, whether in an acidic, alkaline, or neutral environment, the time was shorter than the response time, as shown in Fig. 6b, d, and f . We speculated that water molecules can engage in electrostatic interactions with ions. Additionally, the bilayer hydrogel was immersed in NaCl solution with different concentration (1 M, 2 M, 3 M, 4 M and 5 M) to further explore the salt-responsive property. The change in center angle and curvature of the bilayer hydrogel are shown in Fig. S4(a) and (b). † To display the data more intuitively, the average rates of the center angle and curvature were used as Fig. S4(c) † shows. It turned out that the average rate of the two parameters was positively correlated with the concentration of NaCl solution. Apart from temperature, pH, and ion concentration, the bilayer hydrogel also responded to ethanol. Different from the previous test results, when the hydrogel was immersed in ethanol, it bent towards the negative direction. In Fig. 7a , it took 1140 s for the center angle and curvature to increase to 150° and 0.58. The mechanism in the N layer was that ethanol was a better solvent for NIPAM; thus, the hydrogel swelled, as shown in Fig. S1b. † In the D layer, the mechanism is still unknown. We deduced that ethanol affects hydrogen bonds between water molecules and hydrophilic groups, which reduced the LCST. In Fig. 7b , the recovery time was approximately one-fifth that of the response time. We suspect that the attraction of a large number of water molecules to ethanol molecules was greater than to the hydrogel. Fig. 7 Center angles and curvatures of bilayer hydrogel in ethanol (a) and its recovery (b). In temperature, pH, and salt tests, the hydrogel bent positively, while in ethanol, it bent negatively. Among the stimuli-responsive tests, it can be deduced from the time axis of the figures that the temperature-response rate was the fastest. Additionally, the center angle and curvature presented the same trend because of the special stimuli response mechanisms of PDMAEMA and PNIPAM. This situation may be different in bilayer hydrogels prepared from different monomers. 3.3 4-Arm gripper Both ends of the bilayer hydrogel are regarded as particles whose motion can be de-composited into the horizontal and vertical directions. The midpoint of the hydrogel was set as the origin of the reference system, as Fig. 8a shows. When the hydrogel bent in response to stimuli, the moving track included the process of clamping; thus, a 4-arm gripper was manufactured. As concluded from the former chapter, the thermal response rate was the fastest, so the gripper was driven by temperature. Fig. 8b shows the gripping and release process of the 4-arm gripper. As Fig. 8c shows, for a faster and more convenient grip, the 4-arm gripper was first immersed in hot water (40 °C) and then subsequently adjoined close above the copper sheet. It took 20 s to generate an equal force for lifting an object against gravity. When the “arms” bent inwards and caught the object, the gripper was pulled up out of a high-temperature environment and then moved to a low-temperature environment (25 °C). Due to the thermal property of the bilayer hydrogel, the “arms” bent outwards, releasing the object within 142 s. Fig. 8 (a) Bending behavior of the bilayer hydrogel and reference system; schematic illustration (b) and digital images (c) of the gripping and release using the 4-arm gripper (the copper sheet is marked by the red circle). 3.4 Conductivity of bilayer hydrogels and thermistor The D layer was conductive for containing ionic polymers (PDMAEMA and PSS). The average resistance was around 44.8 kΩ. While in the N layer, conventional polymers are non-conductive; therefore, CB was added in the N layer to improve its electrical conductivity. The graphite phase in carbon black contains conjugated π-bonds that allow electrons to move freely. In addition, CB was well-dispersed, and in situ polymerization maintained the original locations of CB, forming a continuous conductive network; thus, the bilayer hydrogel was conductive. The bilayer hydrogel was added into a circuit as a thermistor to further prove its conductivity and thermal-responsive properties. This circuit included a regulated power supply (5 V), an LED, the bilayer hydrogel, and some copper wires, as shown in Fig. S5. † When the hydrogel was not connected to the circuit, the voltage of the LED tested by the multi-meter was very close to the total voltage due to line loss. Once the hydrogel was connected to the circuit, the voltage of the LED was divided, as Table 1 shows. Initially, the hydrogel was immersed in cold water (25 °C) with a divided voltage of nearly 1.91 V. Subsequently, the hydrogel was quickly immersed in hot water (50 °C) and then placed in cold water (25 °C), which changed the voltage to 1.91 V and then back to 3.05 V, demonstrating the thermal-responsive properties of the hydrogel. Apart from the value change in the multi-meter, the brightness of the LED was also used as an observation index, as Fig. 9b shows. The brightness of the LED changed in the order of dark–bright–dark ( Fig. 9a ) after immersing the hydrogel in water at different temperatures with the order: cold (25 °C) – hot (50 °C) – cold (25 °C). Voltage division (V) between the LED and bilayer hydrogel Working condition Hydrogel LED Rated voltage 25 °C (no hydrogel) × 4.99 5.00 25 °C 2.98 1.99 5.00 50 °C 1.91 3.02 5.00 25 °C 3.05 1.93 5.00 Fig. 9 (a) Brightness changes of the LED when the bilayer hydrogel was immersed in deionized water in the order of 25 °C–50 °C–25 °C. The correlation between (b) stretching degree and (c) resistance change rate of the bilayer hydrogel. During polymerization in the N layer, CB was well dispersed, and each position was fixed after gelling. As mentioned in the former section, a high environmental temperature caused the positive bending of the bilayer hydrogel and the shrinkage of the N hydrogel. The deformation reduced the distance between CB, increasing its density. In this way, the increase in the conductive material per unit volume reduced the resistance of the bilayer hydrogel; thus, the resistance of the hydrogel decreased when the temperature increased and vice versa , which explains the value changes in Table 1 . From a macroscopic view, the deformation caused by a force directly acting on the hydrogel also changed its resistance. Because axial compression is difficult to achieve, axial stretching was used. It is shown in Fig. 9c and d that the resistance of the hydrogel increased with the degree of stretching and vice versa . It can be concluded that the bilayer hydrogel is conductive, which was affected by volume changes, whether caused by a stimuli response or an external force. 3.5 Finger movement monitor A finger movement monitor was developed based on the conclusions drawn in the former chapter, in which the resistance of the bilayer hydrogel was correlated with stretching deformation. The assembled monitor is shown in Fig. 10a . The resistance changing along with finger movement was divided into different phases (bending 0°, 45°, and 90°). Initially, the finger was straightened, and the change in resistance was zero. As the bending degree of the finger increased and decreased, the resistance of the bilayer hydrogel changed correspondingly, and each phase was maintained for a short time, as Fig. 10b illustrates. In each phase, the resistance varied enough for it to be distinguished; therefore, according to the resistance data, when the finger completed at least one movement cycle, it is easy to judge which phase the finger is in. Furthermore, a repeatability test was carried out, as shown in Fig. 10c . Each phase of finger movement can be clearly distinguished, which ensures that the hydrogel can be used as a reliable monitor. Fig. 10 (a) Image of the assembled finger movement monitor; (b) the correlation between bending states of finger and resistance change rate of the bilayer hydrogel; (c) repeated finger movement monitoring (5 times)."
} | 6,023 |
23997654 | null | s2 | 3,521 | {
"abstract": "Contact with macroalgae often causes coral mortality, but the roles of abrasion versus shading versus allelopathy in these interactions are rarely clear and effects on gene expression are unknown. Identification of gene expression changes within corals in response to contact with macroalgae can provide insight into the mode of action of allelochemicals, as well as reveal transcriptional strategies of the coral that mitigate damage from this competitive interaction, enabling the coral to survive. Gene expression responses of the coral "
} | 135 |
21591680 | null | s2 | 3,522 | {
"abstract": "Living organisms have evolved a vast array of catalytic functions that make them ideally suited for the production of medicinally and industrially relevant small-molecule targets. Indeed, native metabolic pathways in microbial hosts have long been exploited and optimized for the scalable production of both fine and commodity chemicals. Our increasing capacity for DNA sequencing and synthesis has revealed the molecular basis for the biosynthesis of a variety of complex and useful metabolites and allows the de novo construction of novel metabolic pathways for the production of new and exotic molecular targets in genetically tractable microbes. However, the development of commercially viable processes for these engineered pathways is currently limited by our ability to quickly identify or engineer enzymes with the correct reaction and substrate selectivity as well as the speed by which metabolic bottlenecks can be determined and corrected. Efforts to understand the relationship among sequence, structure, and function in the basic biochemical sciences can advance these goals for synthetic biology applications while also serving as an experimental platform for elucidating the in vivo specificity and function of enzymes and reconstituting complex biochemical traits for study in a living model organism. Furthermore, the continuing discovery of natural mechanisms for the regulation of metabolic pathways has revealed new principles for the design of high-flux pathways with minimized metabolic burden and has inspired the development of new tools and approaches to engineering synthetic pathways in microbial hosts for chemical production."
} | 413 |
35641526 | PMC9156696 | pmc | 3,523 | {
"abstract": "Semi-conducting Fe oxide minerals, such as hematite, are well known to influence the fate of contaminants and nutrients in many environmental settings through sorption and release of Fe(II) resulting from microbial or abiotic reduction. Studies of Fe oxide reduction by adsorbed Fe(II) have demonstrated that reduction of Fe(III) at one mineral surface can result in the release of Fe(II) on a different one. This process is termed “Fe(II) catalyzed recrystallization” and is believed to be the result of electron transfer through semi-conducting Fe (hydr)oxides. While it is well understood that Fe(II) plays a central role in redox cycling of elements, the environmental implications of Fe(II) catalyzed recrystallization require further exploration. Here, we demonstrate that hematite links physically separated redox reactions by conducting the electrons involved in those reactions. This is shown using an electrochemical setup where Cr reduction is coupled with a potentiostat or Shewanella putrefaciens , a metal reducing microbe, where electrons donated to hematite produce Fe(II) that ultimately reduces Cr. This work demonstrates that mineral semi-conductivity may provide an additional avenue for redox chemistry to occur in natural soils and sediments, because these minerals can link redox active reactants that could not otherwise react due to physical separation.",
"introduction": "Introduction Metal (hydr)oxides are important substrates that constrain contaminant and nutrient fate in soil-sedimentary systems through sorption and redox reactions 1 – 11 . Hence, dissolution and transformation of these metal hydr(oxides) is an integral component of soil geochemistry, and necessary to understand major nutrient and contaminant cycling. For example, iron oxides may play a central role in the nutrient cycling for deep biosphere habitats, such as mines, where limited carbon and oxygen but abundant Fe favor the growth of metal reducing bacteria 12 . These solids are also well recognized sorbents for a variety of contaminants, including arsenic, radium, and chromium 10 , 11 , 13 , 14 .Hematite (Fe 2 O 3 ) has been thoroughly studied due to its environmental ubiquity, extensive sorption of many solutes of interest, and sensitivity to redox conditions. Reductive dissolution of hematite leads to the release of Fe(II), which can then react with organic matter or other metals in solution to drive further redox reactions 15 – 19 . Hematite reduction may occur through transport and sorption of reduced soluble constituents such as sulfide or quinones, which can be abiotic or biotic in origin, or by direct conduction of electrons from a dissimilatory metal reducing bacteria (DMRB) at the surface 20 – 22 . Reduction by DMRB is one of the most important controls on Fe redox cycling because DMRB rely on hematite and other Fe (hydr)oxides as a primary electron sink for their anoxic metabolism. A variety of mechanisms allow microbes to access these solid electron sinks, which include the use of microbial nanowires, release of electron carriers, such as cytochrome c , as well as direct contact with minerals by surface colonization 21 , 23 – 26 . Metal reducing bacteria use these mechanisms to access electron acceptors at ranges as large as 10 µm to 20 µm, as well as access nanopores that are too small for cells to enter 25 , 27 . DMRB have also been observed to use these mechanisms with conductive electrodes held at a reducing potential, which has provided the basis for development of microbial fuel cells 28 , 29 . While terminal electron accepting processes between DMRB and metal hydroxides are important from a metabolic perspective, they also facilitate mineralogical alterations that influence the cycling of metals/nutrients. Studies using isotopically labeled Fe(II) have demonstrated that the semi-conductivity of hematite (and other similar Fe oxides) results in a cycling of Fe between the aqueous and solid phase through atom exchange processes and mineral recrystallization 30 – 32 . Investigation of this phenomena revealed that following the sorption of Fe 57 enriched aqueous Fe(II) to one surface of the mineral, Fe 56 (II) would be released on another surface, resulting in isotopic equilibration of the solution with the mineral without alteration of the mineral structure or crystallinity 33 , 34 . This phenomena, termed Fe(II) catalyzed recrystallization, is postulated to arise from conduction of the electron donated by sorbed Fe(II) to another point on the mineral, which is facilitated by the mineral’s semi-conductivity 32 . This process is well understood to impact the environmental cycling of Fe, and may also impact the cycling of other metals and nutrients 35 . For example, Fe(II) catalyzed recrystallization of Fe solids in the presence of U boosted incorporation of U into the Fe solid over a 90 day period, while Fe(II) catalyzed recrystallization of Mn, Ni, or Zn doped Fe solids resulted in enhanced release of those trace metals over 5 to 10 days 36 , 37 . The Fe(II) produced by DMRB during bacterial metal reduction has also been shown to enable Fe(II) catalyzed recrystallization of Fe solids, further adding to the web of interactions between these solids and DMRB 38 . Furthermore, other redox active species produced by microbial respiration, such as nitrogen or sulfur species, may also readily drive this cycling of hematite 20 , 39 , 40 . These studies have collectively demonstrated that naturally occurring redox reactions will strongly influence the composition of Fe solids beyond the surface. While electron conduction, and its impact on the incorporation and release of elements, has been explored in detail for semi-conducting Fe (hydr)oxides such as hematite or goethite, it remains unclear whether electron transfer through Fe oxides may affect other redox reactions that occur on the surface or near an Fe oxide. Specifically, the current understanding of this process suggests that an electron donated to a semi-conductive Fe (hydr)oxide mineral surface (such as during bacterial metal reduction) could be conducted to another location where a subsequent redox reaction could occur. If so, the Fe (hydr)oxide would electrically link the two reactions, thus coupling the two reactions. This possibility was initially suggested in initial studies of Fe(II) catalyzed recrystallization, where dissolution and precipitation of hematite were observed on physically distinct crystallographic planes indicating the conduction of electrons from one surface to another 32 . A recent study has also explored this possibility with the model compound, cytochrome c , which was used to represent biotic Fe(II) oxidation, and found that electrons from Fe(II) sorbed to hematite would reduce cytochrome c by conduction through the hematite 41 . However, further investigation is required to understand if Fe (hydr)oxide conductivity could lead to the coupling of redox reactions via Fe oxide conductivity. To determine whether electron conduction through iron (hydr)oxides may link redox reactions, we investigate whether a DMRB or other electron source would be able to affect reduction of an electron acceptor, here, Cr, via conduction through hematite. Chromium is chosen as the terminal electron acceptor for these systems owing to its relevance for contaminated soils and sediments as well as its capacity to disambiguate direct microbial reduction from indirect reduction by Fe(II), thus serving as a chemical probe for the fate of electrons conducted by hematite. Cr(VI) is a carcinogenic and highly soluble metal, and is often produced and released to soils and groundwater through industrial activities, while Cr(III) is relatively insoluble and has minimal toxicity 42 – 46 . Hexavalent Cr can be directly reduced by metal reducing microbes to form Cr(OH) 3 , but can alternatively be reduced by Fe(II) to form mixed Fe/Cr hydroxide solids, which are typically in the form Cr 1–x Fe x (OH) 3 , where x is as large as 0.75 1 , 44 , 46 , 47 . The dependence of Cr solid composition on reductant has previously been leveraged to understand changes in Fe(II) activity during the redox cycling of goethite, and is used here to a similar effect 48 . Numerous studies have also shown redox coupling between pairs of these components (hematite, Cr, DMRB), however, it remains unclear how reduction would proceed if all three are present simultaneously. The overarching goal of this work, therefore, was to determine how hematite conductivity can link spatially segregated redox reactions, where the reactions are reduction of Fe by DMRB or abiotically, and reduction of Cr.",
"discussion": "Discussion In both the biotic experiments and abiotic experiments, Cr was removed from solution, though at differing rates. Since the control experiments clearly demonstrated that sorption was not affecting the Cr concentrations and is supported by the lack of Cr(VI) observed in XPS spectra for experiment C, the only driver for this observed behavior must be reductive precipitation, which must be associated with the current delivered. This follows from the normal speciation of Cr(VI) and Cr(III) at pH 7, where Cr(VI) will remain in solution while Cr(III) will be rapidly removed from solution by precipitation 1 , 44 , 47 . This is corroborated by the emergence of Fe(II) in potentiostatic experiments, which is a result of hematite reduction by the delivered current, and variations in that current yield variations in the observed Fe(II) released. The variations in turn are the result of the varied resistivity of the natural source hematite instead of a higher purity synthesized hematite. The difference in current magnitude between the biotic and potentiostatic experiments also suggest that the amount of current is linked with Cr reduction, though the linkage is less clear. In particular, the current differences are not able to discriminate between direct Cr reduction by electrons conducted through the hematite and indirect reduction by the produced Fe(II). Both cases will link total electrons delivered to the amount of Cr reduction observed, but these different mechanisms have markedly different implications for the mechanisms at play in these experiments. Cr, in addition to being a contaminant of concern in many groundwater systems, also then serves as a chemical probe for discriminating between direct reduction and indirect reduction by Fe(II), particularly when considering the ratios of Cr/Fe observed by XPS. This is readily calculated from the XPS data by comparing the ratios of Cr to Fe observed on the surface of the thin section hematite electrode used in experiment C (Fig. 3 ). The ratio varies throughout the sample, with Cr/Fe ratios ranging from 0.0 to 0.5 (SI figure S2). Previous studies of Cr reduction in the presence of Fe showed that Cr will form a mixed Cr/Fe solid when exposed to Fe(II), with Cr/Fe ratios as low as 0.33, as opposed to forming a pure phase Cr(OH) 3 when directly reduced, which here would result in a very large Cr/Fe ratio 46 – 48 . The general formulae for these minerals is Cr 1− x Fe x (OH) 3 · n H 2 O, corresponding to mixed Cr/Fe hydroxides. Multiple locations on the reacted hematite electrode (5 of 21) have Fe/Cr ratios within the range of these types of mixed Cr/Fe solid. The largest Cr/Fe ratio observed in the XPS results is 0.5, which would most likely represent these mixed Cr/Fe solid previously observed, rather than a pure Cr phase, thus suggesting that Cr was reduced by Fe(II) at all locations, rather than reduction by electrons conducted to surface associated Cr(VI). This solid would correspond to a stoichiometry of Cr 0.33 Fe 0.66 (OH) 3 · n H 2 O. Similarly, the peak fitting results indicate that all Cr on the surface is Cr(III), further reinforcing that co-precipitation with Fe has occurred, rather than sorption of Cr(VI). XPS studies of Cr reduction by Fe(II) have previously shown that it is difficult to discriminate between pure Cr(OH) 3 solids and mixed Fe/Cr minerals by shifts in the Cr 2p spectra alone, and generally found that during reduction of Cr by Fe(II), mixed Fe/Cr solids predominantly form 51 . These results are consistent with other studies as well, where Cr was observed to form mixed Cr/Fe solids following Cr(VI) reduction by Fe(II) in solution 46 , 47 . The indirect reduction process also explains why the timing of Fe(II) release and Cr reduction vary between experiments (i.e. that in expt. B Cr reduction precedes Fe(II) release, while in expt. C, they occur simultaneously). If the delivered current produces Fe(II) at a rate faster than Cr reduction by Fe(II) can occur, then both Cr removal and Fe(II) release would occur simultaneously. The variations in delivered current are also in alignment with this observed behavior. Lastly, indirect Fe reduction of Cr would explain why no dissolved Fe(II) is observed in the biotic experiment D, as the small amount of Fe(II) produced by the lower current has entirely driven reduction of Cr. To support a mechanistic conceptual model, Cr removal rates can be compared against the current to the hematite electrode. Since Cr reduction is the only pathway that results in removal from solution, the quantity of Cr removed is considering using the half reaction: 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}$${\\text{Cr}}\\left( {{\\text{VI}}} \\right)\\left( {{\\text{aq}}} \\right) + 3{\\text{e}}{ - } \\leftrightarrow {\\text{Cr}}\\left( {{\\text{III}}} \\right)\\left( {\\text{s}} \\right)$$\\end{document} Cr VI aq + 3 e - ↔ Cr III s Accordingly, each mol of Cr removed corresponds to 3 mol of electrons transferred to the hematite electrode. Multiplying the Cr removal rate given in Table 1 by 3 and comparing against the current should then give the fraction of current which contributes to Cr reduction. In potentiostatic experiments B, Cr reduction only accounts for 17.7% of the delivered current, while in experiment C it accounts for only 1.6% of the delivered current. The remaining current likely then results in Fe reduction, which is discussed below. In contrast with the abiotic experiments, the amount of Cr reduction observed in biotic experiments exceeds the delivered electrons, with the amount of Cr reduction being 157.1% of the delivered current. Based on the collected data, the reasons for such anomalously high Cr reduction in the biotic experiment remains unexplained. One possibility is that Fe/Cr solids that have formed due to reduction are able to uptake additional Cr through sorption. Sorption of Cr to Fe and mixed Fe/Cr solids has been observed in cyclic studies of Cr reduction by microbes in the presence of Fe 47 , 52 . While in those studies, DMRB were able to completely reduce all available Cr, some amount of surficial sorption was observed, and could readily explain the Cr discrepancies, since the discrepancy corresponds to only a few µmol of Cr. Similarly, the reduction of Fe will alter the surface of the hematite electrode, as evidenced by the electron backscatter images in Fig. 3 , which may further enhance sorption processes either by enhancing the surface area relative to the original polished surface, or by the formation of solids with higher sorption affinity for Cr. The reason this was likely not observed in potentiostatic experiment B and C is due to the large amount of current that was delivered to the electrode, which would lead to complete Cr reduction regardless of if it had time to sorb to the altered surface or not. A similar process can be applied to Fe to complete the balance of electrons transferred, in which each mol of Fe(II) measured accounts for a mol of the transferred electrons, following the half reaction: 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}$${\\text{Fe}}\\left( {{\\text{III}}} \\right)\\,\\left( {\\text{s}} \\right) + {\\text{e}}{ - } \\leftrightarrow {\\text{Fe}}\\left( {{\\text{II}}} \\right)$$\\end{document} Fe III s + e - ↔ Fe II For the biotic experiment D, no dissolved Fe is measured, and all current is accounted for. However, as evident in Table 1 , the Fe(II) production rate in potentiostatic experiments is significantly smaller than the total current. The only other potential redox reaction in this system would be splitting of water into hydrogen and oxygen, however, no gas evolution was observed in any experiment. Any measured current beyond that which contributes to Cr reduction must, therefore, result in Fe reduction. The electron backscatter images support this possibility, as the dramatic change in the surface between the reacted and unreacted electrodes could readily occur due to reductive dissolution of Fe. This does not, however, constrain the fate of all Fe(II) that would have been produced by the provided current. The fate of Fe(II) cannot be completely determined directly from the analyses performed here, but there are a few reasonable possibilities: First, Fe(II) generated remains sorbed to the hematite electrode surface. Fe(II) sorption to Fe (oxy)hydroxides is well documented and has previously been observed during cyclic oxidation and reduction of goethite 30 . In those experiments, reduced Fe(II) persisted even throughout oxic conditions by associating with the surface, so it is very likely at least some amount of Fe(II) would be retained on the surface 48 , 53 . Second, the dissolved Fe(II) may form a precipitate on the hematite surface. The formation of these types of solids has been observed in other studies of hematite redox cycling, and is reasonable to expect here 9 , 54 , 55 . Both sorption of reduced Fe as well as precipitation of other Fe minerals is reasonable here and would readily close the balance of electrons, however, further studies are needed that explicitly track Fe fate, such as through the use of an isotopically labeled Fe tracer. PIPES, the buffer used in these experiments, is well understood to enhance solution complexation of Fe(II) and would need to also be carefully considered in Fe tracking experiments 56 . Despite this limitation, the fate of Cr, Fe, and the current in this system is sufficiently constrained to propose a mechanism that explains the trends and behavior observed in these experiments. Figure 4 gives a proposed conceptual model that describes the fate of Cr, Fe, and electrons in this experimental system. Electrons that arrive at the hematite electrode, regardless of their origin, induce reductive dissolution of hematite, thereby producing Fe(II) at the interface of the hematite electrode and solution. The resulting Fe(II) reduces Cr to form mixed Fe/Cr solids, with perhaps some direct reduction of Cr(VI) to Cr(III). When the rate of Cr reduction is slower than the delivered current or all Cr has been reductively precipitated, excess Fe(II) is produced. This Fe can be found both in solution and associated with the hematite electrode. This proposed mechanism is a natural extension of the established mechanism behind Fe(II) catalyzed recrystallization; the main difference here from those works is that there is a separate source of electrons as well as a terminal electron acceptor instead of Fe(II) as electron source and terminal electron acceptor 31 – 33 , 41 . These results demonstrate that hematite, or other semi-conductive Fe minerals, may serve a broader role in controlling redox chemistry in natural soils. Figure 4 Schematic of conceptual model which illustrates how electrons interact with hematite to reduce Fe and Cr. Conducted electrons from an arbitrary source produce Fe(II) which can either reduce Cr to form a mixed Fe/Cr solid or remain as Fe(II) to distribute between sorbed and dissolved species. DMRB are well established to rely on Fe hydr(oxides), such as hematite, as an electron acceptor during anaerobic metabolism, and the results here expand the potential interactions between an Fe oxide and DMRB. The biotic experimental results clearly demonstrated that hematite created a coupling between microbial metabolism of lactate and Cr reduction, implying that semiconductive phases, such as hematite, may enable coupling of other redox processes. One prominent example where this will be relevant is direct interspecies electron transfer (DIET), where one bacterial species will donate the electrons from their metabolism to another species 57 , 58 . The observed coupling of reactions here supports the possibility that bacterial communities could use hematite as a mediator for these electron transfers. Frequently, hematite reduction is preceded by microbial reduction of sulfate or nitrate; the results here are not specific to microbial iron reduction, and semi-conductive hematite may also either mediate this microbial reduction, or facilitate redox reactions of the resulting nitrogen and sulfur species 39 , 40 , 59 . These results are also in good alignment with assertions that bacteria can use conductive substrates (i.e. minerals, other bacteria) to form bacterially active networks, as metal reducing bacteria here used hematite to access Cr as a terminal electron acceptor 21 , 58 , 60 – 62 . Further investigation is required to understand the prevalence of these networks in natural soils and their importance to bacterial growth in natural settings. Some of the earliest electrochemical studies, where sacrificial anodes were used to prevent oxidation of metal ship components, established that electrical contact was sufficient to couple redox reactions of different metals that were otherwise separated 63 , 64 . This work follows in the footsteps of that foundational work: hematite here has enabled spatially segregated redox reactions between a metal reducing bacteria or other electron source, and Cr. The main components of this system, a semiconductive Fe oxide, an Fe reducing reaction, and an Fe oxidizing reaction, are ubiquitous in the environment, suggesting that the prevalence of this kind of linkage is more prevalent than has been previously reported. The results of this work, however, outline only the first steps of broader understanding of how mineral conductivity may influence groundwater chemistry. Fe(II) forms favorable redox couples with other priority contaminants such as U or As, and this mechanism could play a role in the remediation or natural cycling of those contaminants 3 , 15 , 65 – 67 . Fe(II) catalyzed recrystallization has been demonstrated for goethite, another semi-conducting Fe mineral, thus the results here also imply that this conduction mechanism may be more broadly applicable wherever Fe cycling occurs. Given the ubiquity of semi-conductive Fe (oxide) solids in natural soils, it is likely that this type of coupling is likely widespread wherever redox gradients intersect these solid materials, though its importance may be affected by mineral or other organic coatings that ultimately constrain the conductivity of these minerals. One locale where this mechanism may be prevalent is in deep mines, where abundant semiconductive ores may allow native microbial communities to access distant electron acceptors and nutrients 12 , 68 . Further investigation is needed, however, to demonstrate the range of biogeochemical systems where this mechanism is relevant, further understand the influence of these processes on hematite surface chemistry and illustrate the importance of this mechanism in natural soils and groundwater."
} | 5,954 |
36335147 | PMC9637226 | pmc | 3,524 | {
"abstract": "Natural gels and biomimetic hydrogel materials have been able to achieve outstanding integrated mechanical properties due to the gain of natural biological structures. However, nearly every natural biological structure relies on water as solvents or carriers, which limits the possibility in extreme conditions, such as sub-zero temperatures and long-term application. Here, peptide-enhanced eutectic gels were synthesized by introducing α-helical “molecular spring” structure into deep eutectic solvent. The gel takes full advantage of the α-helical structure, achieving high tensile/compression, good resilience, superior fracture toughness, excellent fatigue resistance and strong adhesion, while it also inherits the benefits of the deep eutectic solvent and solves the problems of solvent volatilization and freezing. This enables unprecedentedly long and stable sensing of human motion or mechanical movement. The electrical signal shows almost no drift even after 10,000 deformations for 29 hours or in the −20 °C to 80 °C temperature range.",
"introduction": "Introduction Flexible strain/pressure sensors, mimicking human skin by detecting external stimuli and converting them into electrical signals, allow monitoring human physiological information in a noninvasive way, which are also promising in soft robotics, human-machine interfaces, etc 1 . Conductive hydrogels, a kind of traditional soft conductive materials, have been extensively investigated for this purpose. Compared with other soft conductive materials such as filler/elastomer composites, conductive hydrogels exhibit advantages including excellent biocompatibility, tunable mechanical properties, relatively high electrical conductivity, and low interfacial resistance. Therefore they have attracted more and more attention in recent years 2 . Significant progress has been made in the development of hydrogel-based flexible sensors, however, the reliability and stability of the sensors, particularly in long-term use and under extreme conditions, still remains a tremendous challenge 3 – 5 . To achieve this goal, a conductive hydrogel should meet a wide range of requirements simultaneously. First of all, the gel should be stretchable and tough enough. Unfortunately ordinary synthetic hydrogels are quite fragile because of their inhomogeneous network and the lack of mechanism for energy dissipation 6 . Their strength should be enhanced using strategies, such as double network gels 7 – 9 . nanocomposite gels 10 , and dual-cross-linked gels 11 . Secondly, the gels should be highly resilient to achieve reliable and stable response. Unfortunately for gel materials, high resilience and high toughness are seemingly contradictory 12 . High strength hydrogels usually have a low resilience 13 leading to low signal reproducibility and severe baseline drift of the sensors 7 , 14 – 17 . A third requirement is excellent tissue adhesiveness. This property simplifies the fixing of the gels 18 , 19 . More importantly, it allows conformal and intimate contact of the gel with the skin, which is essential for reliable signal transducing 4 , whereas ordinary hydrogels are usually non-adhesive. Anti-freezing and anti-drying properties are indispensable for their long-term use in various environments. However, ordinary hydrogels will freeze at sub-zero temperatures, leading to loss of flexibility and conductivity 7 , 20 . They also lose water gradually, leading to changes in their mechanical and electronic properties. Significant efforts have been made to design conductive hydrogels with improved toughness 7 , resilience 21 , adhesion 17 , 22 , anti-freezing 7 , 22 , 23 and anti-drying properties 21 , 24 . However, designing a gel with all these properties remains a huge challenge 22 . Here we report that conductive gels with all the properties required for high performance flexible strain/pressure sensors can be synthesized simply using a peptide cross-linker as cross-linker and deep eutectic solvent (DES) as solvent (Fig. 1 ). The peptide-crosslinked gels adopt a novel mechanism of energy dissipation through the breakage of intramolecular hydrogen bonds in the α-helical peptide chains. Like molecule-sized springs, these chains absorb energy when loading but return it back when unloading. Therefore the mechanical strength of the gels is improved significantly while maintains highly resilient simultaneously 25 . DESs are a new generation of green solvents sharing many characteristics and properties with ionic liquids but avoiding their major drawbacks, i.e., high toxicity, non-biodegradability, complex synthesis, and high cost 26 . Thanks to the good conductivity of DESs, the resulting eutectogels themselves are conductive without the need to add any other substances 27 , 28 . The low freezing point and low volatility of the solvent render the gels excellent anti-freezing and anti-drying properties. Interestingly, the gels also exhibit excellent adhesive property that are fully compatible with the adhesion to common surfaces. Combining all these properties, the gels give signals with unprecedented stability and repeatability when used as strain/pressure sensors. Fig. 1 Schematic diagram of the structure and properties of peptide-enhanced eutectogel. The gel exhibits high toughness, high resilience, high adhesion, anti-drying, and anti-freezing properties simultaneously.",
"discussion": "Discussion In summary, a conductive gel combining properties including excellent stretchability, high toughness, high resilience, excellent adhesive property, anti-freezing and anti-drying was successfully synthesized by photopolymerization of AAm in the presence of PCs in DES. Replacing of the common cross-linker with PC dramatically enhances the mechanical properties of the gel. Thanks for its unique mechanism for energy dissipation, the gel remains highly resilient. The gel is also highly crack-resistant, with the highest fracture toughness reported up to now for single network gels. DSC study revealed that the DES has a low freezing point. No glass transition or phase transition occurs for the gel in the temperature range from 80 °C to −50 °C. As a result little change in mechanical properties of the gel was found in the wide temperature range. Particularly the gel still remains flexible at −27 °C. The low volatility of DES renders the gel excellent solvent-retention properties. Only a very small mass loss was found for the eutectogel after 30 days storage under ambient conditions or 3 days under medium vacuum. The gel is also highly transparent and adhesive. In addition, it is conductive due to the high conductivity of DES. All these properties make the eutectogel excellent for wearable, flexible strain/pressure sensor. The sensor works well over a wide range of temperatures (from −20 °C to 80 °C). More importantly, it is highly reliable and durable. As a demonstration, in a test composed of 10,000 consecutive bending cycles which lasted for ~30 h under ambient conditions, the sensor gave almost the same resistance signals in each cycle. Finally, an intelligent obstacle avoidance function demonstrates potential applications including remote monitoring of robots and intelligent electronic skin."
} | 1,800 |
39905218 | PMC11794697 | pmc | 3,525 | {
"abstract": "Quorum sensing, first described in marine systems five decades ago, is a well-characterized chemical communication system used to coordinate bacterial gene expression and behavior; however, the impact of quorum sensing on interkingdom interactions has been vastly understudied. In this review, we examine how these molecules mediate communication between bacteria and marine eukaryotes; influencing processes such as development, disease pathogenesis, and microbiome regulation within marine ecosystems. We describe the varied mechanisms eukaryotes have evolved to interfere with bacterial quorum sensing signaling, the crucial role these signals play in host-virus interactions, and how their exchange may be governed by outer membrane vesicles, prevalent in marine systems. Here, we present a dynamic portrayal of the impact of quorum sensing signals beyond bacterial communication, laying the groundwork for future investigations on their roles in shaping marine ecosystem structure and function.",
"introduction": "Introduction Global biogeochemical cycling and marine ecosystem dynamics are dictated by molecular-level chemical exchanges between marine organisms 1 . Often, this chemical communication is orchestrated by bacteria, which secrete small diffusible quorum sensing (QS) signals 2 that induce density-dependent, population-wide changes in microbial behavior and community composition 3 . These self-produced chemical signals accumulate locally within the environment and attain concentrations that activate the transcription of genes critical for fitness 4 . Since the initial observation of cell density-dependent luminescence in Vibrio fischeri over fifty years ago 5 , our understanding of the diversity of the quorum sensing lexicon and factors controlling the expression of these molecules has expanded significantly. At the most rudimentary level, the molecular circuitry of quorum sensing relies on one or more synthase genes, which encode the production of the diffusible signal termed an autoinducer, and one or more receptors, which dictate transcription, including the synthesis of the signal itself 6 , 7 . Various classes of QS signals have been identified in marine bacteria. These signals are characterized by their ability to diffuse from the producing cell, be recognized by a receiving cell, elicit a response in the receiver that has co-evolved alongside signal production in the producer, and confer mutual benefits to both producer and receiver cells 4 . Acyl homoserine lactones (AHLs, autoinducer-1/AI-1), alkylquinolones (AQs), and α-hydroxyketones (cholera autoinducer-1/CAI-1) constitute the main classes of QS compounds in marine ecosystems 8 (Fig. 1 ). AHLs are the most abundant QS molecules in marine systems, produced by Gram-negative bacteria from representatives of Alpha -, Beta - and Gamma - Proteobacteria . More controversial, however, is the inclusion of furanosyl borate diesters (autoinducer-2/AI-2) in this group, as not all bacteria that produce AI-2 either have the sensor or trigger a response to AI-2 consistent with canonical QS systems 8 . Some believe AI-2 can be at times either a signal, a cue, or metabolic byproduct used opportunistically for signaling depending on the bacterial producer or receiver 4 , 9 , 10 . Similarly, tropodithietic acid (TDA) might fall into this autoinducer category as well, possessing some, but not all QS signal properties 11 . The production of TDA is density-dependent and induces transcription of tda genes which coincide with biofilm formation 12 . Expanding the vocabulary of QS molecules further, one could append this list to include the signal peptide arbitrium (Fig. 1 ). This peptide-based communication system used by viruses that infect bacteria is analogous to QS in bacteria 13 . Arbitrium-based signaling depends on the cell density-dependent accumulation of phage-encoded signaling molecules, which regulate host gene expression to promote a shift from lysis to lysogeny, thereby enhancing the long-term survival and propagation of the viral population within the host community. Fig. 1 Diversity of autoinducer molecules produced by marine representatives. The acylated homoserine lactone (AHL) family of QS signals with representatives including 3-oxo-C4-HSL ( N -3-oxo-butyryl-L-homoserine lactone), C6-HSL ( N -hexanoyl-DL-homoserine lactone), 3-OH-C14-HSL ( N -3-hydroxytetradecanoylhomoserine lactone). p -Coumaroyl-homoserine lactone ( p -coumaroyl-HSL) is a representative aryl-HSL in which the acyl side chain is replaced with p -coumaric acid moiety. The cholera autoinducer-1 (CAI-1, (S)-3-hydroxytridecan-4-one) family is produced only in Vibrio species. The autoinducer-2 (AI-2) family containing furanosyl-borate diester. The alkylquinolone family of QS signals including 2-heptyl-4-quinolone (HHQ). Finally, the QS-adjacent representatives including tropodithietic acid (TDA) and the viral peptide arbitrium encoded by the phage phi3T which infects Bacillus subtilis . QS-molecules are widespread among marine bacteria and are structurally diverse, yet together comprise a “language” allowing bacteria to behave collectively as a group. While the role of QS signaling in bacterial communication is well-established 8 , the exploration of QS molecules’ potential in mediating a wealth of cellular behaviors at the interkingdom level is still in the early stages 14 . Intricate recognition mechanisms between bacteria and eukaryotic cells may occur when host cells detect bacterial signals or when an autoinducer acts on the host cell without regulation 15 . Numerous studies in mammalian systems have identified targets for QS signals including nuclear receptors (aryl hydrocarbon receptor and peroxisome proliferator-activated receptor), G-protein coupled receptors, Toll-like receptors, transcription factors (NF-κB), protein kinases (p38 and p42/44 kinases), cytoskeletal proteins, and cell-surface lipid domains 16 . These targets broadly modulate processes like innate immune responses, lipid metabolism, inflammatory responses, cell structure, and intracellular calcium (Ca²⁺) signaling pathways (reviewed in ref. 14 ). In contrast, fewer studies have identified QS receptors in terrestrial plants. Notably, G-protein coupled receptors in Arabidopsis have been shown to be activated by AHLs, resulting in root elongation 17 , 18 . These studies in mammalian systems and terrestrial plants highlight the complexity of interkingdom QS between bacteria and eukaryotic host cells, and provide an excellent starting point for exploring cognate interkingdom QS receptors and associated regulatory pathways in marine eukaryotes. In this review, we examine the ability of bacterial QS signals to influence marine eukaryotic organisms, providing hosts with chemical contextual cues required for initiating developmental or physiological changes to enhance host fitness under distinct environmental conditions. Furthermore, the presence of QS signals could trigger cross-kingdom recognition, thereby allowing eukaryotic hosts to take stock of their bacterial consortia and respond by producing their own small molecules. The interdependence between hosts and their associated bacteria covers a spectrum of ecological interactions, from cooperative to competitive (Fig. 2 ). Here, we explore the emerging evidence that QS signals play a pivotal role in shaping co-evolutionary relationships in marine systems. Fig. 2 Overview of chemical cross-talk driven by quorum sensing signals liberated by bacteria in marine ecosystems. Here, we showcase the diversity of roles initiated by QS molecules and the consequences of these chemically-mediated interactions on marine invertebrates, macroalgae, phytoplankton, and viruses. For example, the quorum sensing alkylquinolone 2-heptyl-4-quinolone (HHQ) (1) arrests cell division, targets the inhibition of pyrimidine biosynthesis, and protects the coccolithophore Emiliania huxleyi from virus-induced mortality. Moreover, lumichrome (2), a derivative of the vitamin riboflavin, is a QS mimic whose release by exponentially growing E. huxleyi initiates antibiotic production in commensal bacteria, a likely strategy to thwart bacterial invaders. QS molecules can signal an impending viral threat to the community and coordinate the reduced expression of cell surface receptors critical for viral entry, and the induction of CRISPR-Cas systems to acquire immunity during viral invasion. Conversely, the viral peptide arbitrium, which is generated by bacteriophages during bacterial host infection, serves as a communication signal within the viral population, helping to determine the fate of the host cell. In the macroalgae Delisea pulchra , halogenated furanones (3) competitively inhibit acyl-homoserine lactone (AHL) (4) signalling, thereby controlling the bacterial species composition and halting AHL-induced virulence in macroalgal pathogens. The interaction between QS signalling and QS inhibition plays a crucial role in the progression of development and disease in marine invertebrates. Image credit: jenesesimre/ stock.adobe.com. Marine ecosystems provide an opportunity to explore the multifaceted roles of QS signals and autoinducers. These molecules act as a shared molecular language understood by both bacteria and eukaryotes, conveying vital information and directing foundational processes within marine environments. In this review, we aim to provide insight into the role of QS signals as drivers of eukaryotic fitness and determinates of trophic interactions in marine systems. Indeed, the long co-evolutionary relationship between microorganisms and their hosts has enabled marine eukaryotes to colonize all ocean habitats by forming holobionts 19 —an emerging concept that views a eukaryotic host and its associated microbes as a single collective organism 20 . Given the essential role of the holobiont in host biology and the widespread presence of QS signals in marine bacteria, the influence of interkingdom signaling on host evolution merits further exploration. Recent advances in our understanding of marine host-bacteria interactions illustrate how molecules implicated in bacterial QS contribute to disease progression, developmental trajectories, and the modification of marine eukaryotic physiology with examples from marine invertebrates 21 , 22 , macroalgae 23 , 24 , and phytoplankton 25 , 26 . Within marine ecosystems, eukaryotes both engage in and interfere with QS processes, actively participating in QS-mediated dialogs by producing compounds or proteins that disrupt bacterial QS signaling 22 , 27 . Mechanisms to disrupt QS systems are classified as either quorum quenching (QQ) molecules, which enzymatically degrade QS signals, or quorum sensing inhibitors (QSI), which interfere with QS signaling by competitively binding to QS molecule receptors on the bacterial cell surface 28 or by inactivating autoinducer synthases 29 . The production of these molecules implies that host-driven modulation of bacterial QS signals may have evolved as a mechanism for eukaryotes to regulate the composition of their associated microbial communities 22 , 27 . Interactions between marine viruses and microbes dictate the flow and fate of carbon in marine ecosystems. Viruses are the most abundant biological entities in the oceans, outnumbering bacteria by 10-fold, with the majority being bacteriophages—viruses that specifically target and infect bacteria 30 . Emerging research now reveals how QS molecules are influencing bacteriophage success. QS molecules provide critical information about bacterial population size, enabling phages to make lysis-lysogeny decisions by monitoring QS molecules. Likewise, viruses are responsible for mortality of ~20% of the oceanic photosynthetic biomass daily 31 . Recent work has uncovered a novel mechanistic role for a bacterial QS signal in mediating viral success within its phytoplankton host 25 , offering insight into a complex tri-trophic interaction governed by QS compounds. The rapid loss of QS compounds from their source due to diffusion or fluid flow in the ocean poses a major challenge to maintaining effective concentrations. Many chemical-mediated interactions are sustained by physical contact between the microbe and host, or in specialized host anatomical structures 32 . Additionally, biofilms produced by bacterial consortia act as sorptive sponges, trapping and concentrating organic molecules 33 . Another important mechanism involves outer membrane vesicles (OMVs), which play a critical role in both stimulating and distributing QS molecules. The packaging of QS signaling molecules into OMVs by Gram-negative bacteria appears to be more widespread in marine systems than previously recognized. This delivery system not only preserves the integrity of the signal during transport but also enables targeted delivery to specific sites via receptor-mediated processes. Considering the profound influence of QS signals on aquatic ecosystems and the wide range of hosts they affect, there is a pressing need for further investigation into how these signals shape interkingdom interactions in marine environments."
} | 3,294 |
32385386 | PMC7210978 | pmc | 3,526 | {
"abstract": "In microbial ecosystems, species not only compete for common resources but may also display mutualistic interactions as a result from metabolic cross-feeding. Such mutualism can lead to bistability. Depending on the initial population sizes, species will either survive or go extinct. Various phenomenological models have been suggested to describe bistability in mutualistic systems. However, these models do not account for interaction mediators such as nutrients. In contrast, nutrient-explicit models do not provide an intuitive understanding of what causes bistability. Here, we reduce a theoretical nutrient-explicit model of two mutualistic cross-feeders in a chemostat, uncovering an explicit relation to a growth model with an Allee effect. We show that the dilution rate in the chemostat leads to bistability by turning a weak Allee effect into a strong Allee effect. This happens as long as there is more production than consumption of cross-fed nutrients. Thanks to the explicit relationship of the reduced model with the underlying experimental parameters, these results allow to predict the biological conditions that sustain or prevent the survival of mutualistic species.",
"introduction": "Introduction Microbes play a fundamental role in different ecosystems on Earth. For example, they provide nutrients for plants in the rhizosphere via a symbiotic relationship 1 , contribute to the formation of planktonic communities in the ocean 2 , 3 and are used in the treatment of wastewater 4 . Even the human body is home to large ecosystems of microbial species, called microbiota, contributing to our health by providing essential nutrients and protecting us against potential threats or harmful microbial species 5 . To understand the dynamics of microbial ecosystems, the growth of microbes can be studied in vitro under well controlled environmental conditions 6 . This way, microbes also provide convenient model systems to study general ecological interactions 7 . A particularly suited laboratory device to experimentally study microbial growth is the chemostat 8 . Such a bioreactor allows to grow microbes in a chemically constant environment and to explictly monitor the consumption of metabolites. A chemostat consists of a well-mixed growth tank with a continuous inflow of nutrients and an outflow of the suspension with microbes and nutrients. It is a simplification of natural systems as the inflow and outflow occur at the same rate and the suspension is well-mixed so that spatial effects are ignored. Nevertheless, it constitutes an appropriate tool to probe the behavior of natural systems, which are typically open environments with a flux of energy. For instance, it has been shown that the human intestines can, to some extent, be modeled by chemostat equations 9 , which are particularly suitable to assess the correlation between perturbed microbiomes (dysbiosis) and diseases. Experimental as well as theoretical studies involving a chemostat thus provide an appropriate framework to predict behavior related to microbial interactions such as competition and mutualism in a natural environment. Although mutualism is thought to be less common than competition in microbial ecosystems because it tends to destabilize the community 10 , mutualism can arise via bi-directional cross-feeding of metabolites 11 . It has been shown that microbial diversity is promoted by cross-feeding, which can prevent competitive exclusion 12 . Furthermore, cross-feeding can be essential for different functions. For example, in the intestines metabolites are broken down in smaller components by some species for their consumption by other species 11 . This is necessary for the formation of health-promoting short-chain fatty acids 13 – 15 . Mutual cross-feeding has also been shown to reduce the energetic cost by dividing the labor for the utilization of metabolic pathways, for example for amino acid synthesis 16 . Besides the apparent benefits of mutualism, there is a downside: the interdependency increases the possibility of a collapse of the system due to a density threshold for survival, which has been observed experimentally 17 , 18 . A density threshold between two different states of the system, in this case survival and extinction, is related to the concept of bistability. In mutualism, this is generated by an Allee effect 19 – 23 . An Allee effect is characterized by a decreased fitness at low densities so that the individual growth rate reaches a maximum at an intermediate density due to cooperative behavior. In the case of obligate mutualism, the Allee effect leads to a survival dependency as one species cannot survive without the other 22 . This effect is in contrast with the prediction of the classical logistic growth which predicts that an increased population density limits the growth 24 . A distinction between a weak and a strong Allee effect is made 1 . Whereas a weak Allee effect leads to a single stable state (the species always survives), a strong Allee effect is characterized by bistability, whereby a density threshold for survival is present. Different models showing how mutualism causes bistability via the interdependence between the species have been proposed 25 – 28 . In some of these models the species are competitive as well as mutualistic, e.g. when mutualists become competitors for a resource or for available space at high density 29 – 31 . One study showed that the type of interaction could be modulated by varying the resource concentration 17 . Bistability is generated if the benefit of cooperation is counteracted by a cost, for example via the dispersal of the species. This increases the possibility of a collapse of the mutualistic system 23 . In order to predict when such disruptions occur and, if needed, to intervene to prevent the collapse of the community, a deeper understanding of mutualistic interactions and of the occurrence of thresholds in microbial communities is necessary. A currently unresolved problem is that phenomenological models where interaction mediators, like nutrients, are neglected can behave differently than models where these are explicitly incorporated 32 . As both types of models describe similar phenomena, it should be possible to reduce the mechanistic model into a phenomenological model 33 , 34 . Such reductions of mechanistic models already exist for the growth of a single species 8 , 35 , 36 or for competitive consumer-resource models 37 . In the case of mutualism, one approach described the occurrence of bistability by the saturation of the mutualistic strength at high densities 38 . Nevertheless, it remains unclear how the occurrence of bistability in a mutualistic cross-feeding community is related to nutrient concentrations and to their consumption and production kinetics. This is essential to quantify the effects of prebiotics or biological parameters on the survival of the species. In this theoretical work, we use a nutrient-explicit model for the growth of mutualistic species in a chemostat reactor and show how an Allee effect is created. This allows to interpret the effect of biological parameters on the dynamics in order to predict when bistability is created and to estimate the density threshold for survival. Nutrient-explicit models of a single species in the chemostat can be reduced to the logistic growth equation 8 , 35 , 36 . Using a similar approach, we reduce chemostat equations of a mutualistic system to an appropriate mechanistic model which only involves the species densities. This allows to relate the obtained equations to a generic growth model with an Allee effect. By establishing this analogy, we show that mutualism causes a weak Allee effect, which can be turned into a strong Allee effect under the influence of the dilution in the chemostat. Critical chemostat parameters such as the dilution rate and the influx of nutrients thus allow to manipulate the strength of the Allee effect and therefore of the survival threshold. As a consequence, it is possible to switch between regions of bistability, monostable survival, or monostable extinction. We also show that the production of cross-feeding nutrients needs to be larger than the consumption for an Allee effect to exist. This explicit relationship between experimental parameters and the Allee effect provides a way to bridge the gap between biological experiments and theoretical models.",
"discussion": "Discussion Microbial interaction networks are characterized by multiple positive and negative interactions. Species enter in competition for limited resources, but they can also display mutualistic relationships through cross-feeding. Through mutualistic interactions, both species benefit of each others presence. This may be seen as a stabilizing factor. However, mutualism carries the seed of its own instability: under the influence of dilution bistability may occur, causing a critical density threshold for the species to survive. Once the abundances of the species drop below this threshold, the community eventually collapses and the species become extinct. How biological parameters affect the survival threshold is often unclear. To provide an understanding of the effect of different parameters, we showed how nutrient-explicit equations for two mutualistic cross-feeding species can be reduced to a set of equations which only involve the densities of the species. These could then be related to a growth equation with an Allee effect, which can be analyzed to obtain a deeper understanding of the impact of the different biological parameters. Density thresholds for survival have previously been observed in mutualistic systems. For example, a survival threshold was found in a spatially cooperating microbial community 18 and in a cross-feeding system which could switch to other interactions like competition and parasitism, depending on the availability of nutrients 17 . Besides cross-feeding, mutualism can also arise via the protection of another species towards antibiotics. In such a cross-protection system it was found that periodic dilution drives oscillatory dynamics, potentially leading to extinction if the survival threshold was crossed 43 . Our results showed that the overall production rate of cross-fed nutrients needs to be larger than the overall consumption rate to create an Allee effect. The production and consumption rates can experimentally be altered by making use of synthetic cross-feeding systems 17 , 44 , 45 , which can be designed to display bistability, or on the contrary, to prevent bistability and thereby the risk of an abrupt collapse of the community. We obtained quantitative results for the case where both species have symmetric parameter values, but showed that this framework still applies to the case of asymmetric parameter values. By using time traces of microbial growth experiments it is possible to fit the experimental values of biological parameters 46 . This would allow the validation of these theoretical predictions if the microbes are obligate cross-feeders. If the system exhibits a sufficient production of cross-fed nutrients ( \\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}$${a}_{1}{a}_{2} > {\\nu }_{p1}{\\nu }_{p2}$$\\end{document} a 1 a 2 > ν p 1 ν p 2 ), then bistability is predicted for dilution rates within the range \\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}$${r}_{a}{C}_{a}a < \\,d < {r}_{a}{({C}_{a}+a)}^{2}\\mathrm{/4}$$\\end{document} r a C a a < d < r a ( C a + a ) 2 /4 . The dilution rate and the influx of nutrients are experimental parameters which can be tuned so that these allow to manipulate the behavior. As a result, this provides experimental guidelines to study the presence of bistability in a microbial system. Moreover, the influx of nutrients ( \\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}$$\\tilde{S}$$\\end{document} S ˜ 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}$$\\tilde{P}$$\\end{document} P ˜ ) allow to study the effect of prebiotics, in order to determine the effect of these growth-promoting nutrients on the survival of the species. Furthermore, if antibiotics are used, then determining whether a survival threshold exists would be essential to predict the survival or extinction of the species. The presence of bistability can significantly affect microbial behavior. For example, the existence of a survival threshold creates susceptibility to cheaters 28 . A cheater is an individual of the species which does not cooperate in the production of cross-feeding nutrients, thereby increasing the risk of a collapse of the system 47 . Addition of a third species can create global stability of the coexistence state if it is a facultative mutualist, providing a solution for the presence of the survival threshold 48 . Bistability also affects the behavior in the case of spatial expansions. Here, an Allee effect counteracts genetic drift of a species as it creates a pushed wave rather than a pulled wave corresponding to logistic growth 49 , 50 . Such behavior has been observed experimentally in a system of two cross-feeding species whereby the mutualistic strength was modulated by the inflow of nutrients 51 . Mutualistic species are usually part of a larger ecosystem. Therefore, when mutualists are decreased under the survival threshold, for example in response to antibiotics, the entire ecosystem can be destabilized 10 . Such critical effects on ecological interactions are often not well characterized 52 . Generalized Lotka-Volterra models are often used to study the interactions in microbial ecosystems 53 – 55 . However, Lotka-Volterra models for mutualism involve the addition of fitness effects 32 so that nonlinear growth rates are not incorporated 56 . We studied a theoretical microbial system that is mutualistic via cross-feeding, which is a necessary step to obtain a deeper understanding of complex behavior in microbial communities. The obtained phase plane and the nullclines, describing the dynamics of the species, are observed in different mutualistic models 25 – 28 , so that we can state this is a general phenomenon. Our results give insights into necessary conditions to obtain bistability in a model for microbial species which are obligate cross-feeders: there needs to be an Allee effect as well as a limiting function. In particular, we show that nonlinear growth rates significantly affect the dynamics of microbial communities."
} | 3,823 |
37068259 | PMC10151509 | pmc | 3,527 | {
"abstract": "Significance Designing sustainable artificial photosynthetic systems is a long-standing goal. Photoactive nanomaterials have been used to drive the respiration of microorganisms to produce biofuels. However, this strategy fails to capitalize on the excellent photocatalytic properties of nanomaterials, and the reaction products are limited to those compatible with the microorganism’s metabolism. Here, we use extracellular electron transfer from electrogenic bacteria to directly fuel photocatalytic hydrogen evolution by nanocrystals. This system introduces a way to capitalize on the catalytic properties of nanomaterials by providing an alternative to high-concentration sacrificial reductants. Given the high tunability of nanocrystalline photocatalysts, this method promises a sustainable and flexible route to a variety of solar fuels in the future.",
"discussion": "Results and Discussion CdSe QDs were synthesized as previously reported ( 25 ) with a diameter of 2.6 nm as defined by the wavelength of the peak in their lowest energy excitonic absorption feature (525 nm, SI Appendix , Fig. S1 ) ( 32 ). These QDs were made water soluble by exchanging their native ligands for 3-mercaptopropionic acid (MPA), yielding CdSe–MPA. CdSe–MPA QDs can directly catalyze the proton reduction reaction in water containing ascorbic acid as a sacrificial electron donor ( 25 ). In modified minimal medium (MM) used for MR-1 culture (see SI Appendix , for additional details), CdSe–MPA produced negligible H 2 under constant irradiation (<2 µmol), indicating that there are no suitable electron donors present in the medium to directly support photocatalytic H 2 production by CdSe–MPA ( SI Appendix , Fig. S2 ). Previous electrochemical studies of MR-1 indicate that EET from MR-1 to electrodes occurs at a more negative reduction potential ( 19 ) than the valence band potential of CdSe–MPA ( 18 , 33 ) ( Fig. 1 and SI Appendix , Fig. S3 ), suggesting that EET from MR-1 to photoexcited CdSe–MPA is thermodynamically favorable. Thus, we measured H 2 production by CdSe–MPA QDs in wild-type MR-1 (WT) cultures. Cultures were grown anaerobically from an initial OD 600 of 0.05 in MM at 25 °C with lactate (~20 mM) as an electron source. Irradiating CdSe–MPA (1.0 µM) in the culture with 530-nm light over the course of 1 wk (168 h) resulted in 18.2 ± 5.2 µmol H 2 . In the absence of CdSe–MPA, WT produced 6.2 ± 3.2 µmol of H 2 by the end of 1 wk ( Fig. 2 ). MR-1 has periplasmic hydrogenases (H 2 ase) that may reduce protons during anaerobic respiration ( 34 ). The H 2 produced by WT here is consistent with this activity. In the CdSe–MPA QD mixtures containing WT, most H 2 evolution activity was observed within the first 120 h, plateauing by the end of 1 wk. During the fastest period of H 2 evolution (between 12 to 72 h), a rate of 2.6 × 10 −1 µmol/h was observed. Notably, this is an order of magnitude faster than H 2 production by WT in MM in the same time frame, indicating that the H 2 evolution process in the presence of CdSe–MPA occurs with significantly different kinetics ( SI Appendix , Fig. S4 ). Fig. 2. Hydrogen evolution from CdSe–MPA (1.0 µM) with WT (initial OD 600 = 0.05, as labeled) and Δ hyaA MR-1 (initial OD 600 = 0.05, as labeled) was monitored under constant irradiation (530 nm) at 25 °C, pH 7. Hydrogen evolution in the absence of CdSe–MPA was also recorded. ( A ) Total H 2 evolved was measured after 1 wk (168 h). Error bars represent the SD from ten replicate experiments. ( B ) In addition, H 2 was monitored over time (0 to 168 h). Error bars represent the range of values observed at each time point for triplicate experiments. The increased H 2 evolution activity of CdSe–MPA and WT over WT alone is consistent with WT supporting CdSe–MPA photocatalysis, but the background H 2 evolution activity of WT (from H 2 ase activity) is a complicating factor. To test the hypothesis that the enhanced H 2 production in the complete system results from bacterial respiration of CdSe–MPA, rather than an increase of MR-1 H 2 production activity in the presence of CdSe–MPA, an MR-1 strain suppressing hyaA expression ( Δ hyaA ) was used. The hyaA gene encodes the small subunit of periplasmic [Ni-Fe]-H 2 ase. Consistent with impaired H 2 ase activity, Δ hyaA produced no detectable H 2 over the course of 1 wk when cultured in MM. However, irradiation of CdSe–MPA within a culture of Δ hyaA results in H 2 evolution ( Fig. 2 ). The total H 2 evolved from CdSe–MPA and Δ hyaA was 20.8 ± 5.8 µmol H 2 after 1 wk, which is similar to the activity of CdSe–MPA in the presence of WT ( Fig. 2 ). The overall rate of H 2 evolution is similar for CdSe–MPA with WT or Δ hyaA. The highest rate achieved for CdSe–MPA with Δ hyaA (1.6 × 10 −1 µmol/h) is of the same order of magnitude as the system containing CdSe–MPA and WT, and both slow after approximately 120 h ( SI Appendix , Fig. S5 and Fig. 2 ). If the bacteria are killed via heating to 100 °C, H 2 production plateaus ( SI Appendix , Fig. S6 ). This result further demonstrates that, without functioning electron transfer pathways from Δ hyaA , no H 2 evolution activity is observed. The production of H 2 in the system using Δ hyaA supports the hypothesis that MR-1 is providing electrons to CdSe–MPA through respiration. We next sought to gauge the dependence of H 2 production on light. With CdSe–MPA present in a WT culture, H 2 evolution was monitored in the light and dark over the course of 1 wk ( SI Appendix , Fig. S7 ). In the dark, 4.1 µmol less H 2 on average was produced compared to samples cultured in the light ( Fig. 3 ). The amount of H 2 produced in the dark by CdSe–MPA and WT (9.8 µmol ± 0.6 µmol) is similar to the amount of H 2 produced by WT without QDs present and is attributed to H 2 production by WT. The dependence of activity on irradiation was tested with CdSe–MPA in cultures of Δ hyaA, monitoring the system’s H 2 evolution in the light and dark ( SI Appendix , Fig. S7 ). A mixture of CdSe–MPA and Δ hyaA in the dark resulted in negligible H 2 evolution (<1 µmol), whereas in the light, 17.4 ± 1.4 µmol H 2 was produced ( Fig. 3 ). The light-dependent evolution of H 2 supports the hypothesis that the QDs act as photocatalysts with activity sustained by bacterial respiration. Fig. 3. ( A ) Total H 2 evolution from WT (initial OD 600 = 0.05) and Δ hyaA MR-1 (initial OD 600 = 0.05) with CdSe–MPA (1.0 µM) as monitored in the light (530 nm) and dark after 1 wk at 25 °C, pH 7. Error bars represent the range observed in triplicate experiments. ( B ) After a week-long catalysis with WT (initial OD 600 = 0.05) and Δ hyaA MR-1 (initial OD 600 = 0.05) with and without CdSe–MPA (1.0 µM), the mass accumulated was measured. We also examined the effect of QD concentration on system activity. As QD concentration is increased from 0.5 to 1.0 to 1.5 µM, H 2 production increases. However, at 2.0 µM CdSe–MPA, no activity is observed, suggesting that MR-1 is sensitive to increased QD concentration in these experimental conditions ( SI Appendix , Fig. S8 ). These results indicate that further system optimization is possible, especially if MR-1 can be tuned to accommodate increased CdSe–MPA concentrations. Other studies have shown that CdSe NCs exhibit toxicity to MR-1 ( 35 ). We demonstrate that MR-1 survives in the presence of the catalytic concentrations of CdSe–MPA reported here (<1.5 µM, versus >200 µM in toxicity studies). In the presence of some exogenous electron acceptors, the metabolism of MR-1 is altered ( 36 ). We hypothesized that interacting with CdSe–MPA QDs as an electron acceptor may impact the properties of MR-1. To estimate the amount of biomass present, after 1 wk of catalysis, vials containing CdSe–MPA paired with WT or Δ hyaA were centrifuged and the total wet mass of the resulting pellets was measured. The same procedure was performed on solutions of WT or Δ hyaA cultured in the absence of QDs for 1 wk. Systems containing CdSe–MPA showed higher wet mass than those without ( Fig. 3 ). With CdSe–MPA, wet masses from cultures of WT and Δ hyaA were 11.1 ± 1.4 mg and 11.3 ± 1.1 mg, respectively. Without CdSe–MPA, the respective wet masses for WT and Δ hyaA were 6.7 ± 1.3 mg and 7.0 ± 0.8 mg. The differences in overall mass are not accounted for by CdSe–MPA alone (0.5 ± 0.2 mg). We also measured differences in OD 600 before and after 1 wk of catalysis. The results reveal higher OD for the culture in the presence of CdSe–MPA, showing the same trend as the biomass measurements ( SI Appendix , Table S5 ). It is important to note that the increase in biomass may be a result of exopolysaccharide formation as a defense mechanism ( 37 ). Increased biomass in the presence of CdSe–MPA was observed in both light and dark conditions ( SI Appendix , Fig. S9 ). Water-soluble carbon nanomaterials, carbon dots (CDs), were recently shown to impact the metabolism of MR-1 ( 36 ). The presence of CDs was found to boost the metabolic rate of MR-1, increase biofilm formation, and possibly increase intercellular signaling. Similarly, interactions between MR-1 and nanoparticulate TiO 2 promoted increased riboflavin secretion, possibly boosting the electrogenic behavior of MR-1 ( 38 ). The interaction between MR-1 and nanomaterials is complex, varying by nanomaterial concentration, composition, and charge ( 39 ). However, given the enhanced biomass with CdSe–MPA observed in both the light and the dark here, it is possible that the presence of the QDs could promote biofilm formation. Consistent with biofilm formation, bacteria and QDs formed MR-1:CdSe–MPA aggregates during experiments ( SI Appendix , Fig. S10 ). As aforementioned, H 2 evolution activity in MR-1:CdSe–MPA systems plateaued after approximately 1 wk. MR-1 utilizes lactate in the growth medium as a primary carbon source. Prior to irradiation of Δ hyaA and CdSe–MPA, we used a commercial lactate assay (see SI Appendix , Fig. S11 and SI Appendix for additional details) to determine an initial lactate concentration of 20.8 ± 0.9 mM. After 212 h of reaction with irradiation, the lactate concentration was depleted to 0.6 ± 0.2 mM. Notably, lactate concentration and H 2 production over time are anticorrelated, with both leveling off after approximately 150 h ( SI Appendix , Fig. S11 ). We hypothesized that depletion of nutrients may contribute to the end of H 2 evolution and sought to assess whether the MR-1:CdSe–MPA aggregate retains catalytic activity by replenishing the bacterial growth medium. Bacteria–QD aggregates were separated from MM containing planktonic cells and colloidal CdSe–MPA after 1 wk, and fresh MM was added to the aggregates to a total amount of 5.0 mL in each vial. Upon replenishing the medium and irradiating the mixtures, H 2 evolution activity resumed ( Fig. 4 ). Combined with WT, CdSe–MPA sustained 95% of the original activity (measured as total H 2 evolved) over the second week. When combined with Δ hyaA , 41% of the original activity was recovered. After replenishing the medium, MR-1:CdSe–MPA aggregates typically reform within 12 h. In the absence of CdSe–MPA during the second week, WT in the presence of fresh MM resumed evolving H 2 , while Δ hyaA produced no H 2 over the course of 2 wk ( SI Appendix , Fig. S12 ). These results suggest that these systems combining MR-1 and CdSe–MPA retain their activity after a week of catalysis, suggesting that the bacteria continue to provide electrons to CdSe–MPA within the reformed aggregates. To further test this hypothesis, colony-forming unit (CFU) analysis of the aggregates was performed at specific timepoints with results showing that cells maintain viability over time ( SI Appendix , Table S6 ). Fig. 4. Hydrogen produced by the MR-1:CdSe–MPA system in MM, with fresh MM added to ( A ) CdSe–MPA (1.0 µM) with WT (initial OD 600 = 0.05, as labeled) and ( B ) CdSe-MPA (1.0 μ M) with Δ hyaA MR-1 (initial OD 600 = 0.05, as labeled) after 168 h. Hydrogen evolution was monitored for an additional 168 h under constant irradiation (530 nm) at 25 °C, pH 7. Error bars represent the range of values observed at each time point for triplicate experiments. In many cases, error bars are small relative to the data points. The MR-1:CdSe–MPA aggregates retain the bright orange color of CdSe–MPA ( SI Appendix , Fig. S10 ). This is also true for cell pellets formed by centrifugation postexperiment, where cells in systems containing CdSe–MPA retain the orange coloration of the QDs ( SI Appendix , Fig. S13 ). The presence of both MR-1 and CdSe–MPA in the MR-1:CdSe–MPA aggregate was confirmed with fluorescence microscopy ( SI Appendix , Fig. S14 ). Furthermore, electron microscopy images of the aggregate reveal bacteria surrounded by QDs, with contacts between QDs and the outer membrane ( SI Appendix , Fig. S15 ). Thiol-coated CdSe NCs are reported to precipitate due to photocatalytic oxidation of the thiol ligands as well as photooxidation of the NCs ( 40 ). This is supported by the observation of a red-brown precipitated solid for CdSe–MPA held under irradiation in MM in the absence of bacteria, which precipitates within 72 h ( SI Appendix , Fig. S16 ). To test whether the precipitated CdSe–MPA were still catalytically active, Δ hyaA was added to a solution of precipitated QDs in MM. H 2 evolution from the precipitated CdSe–MPA after Δ hyaA is introduced begins to plateau after only 48 h ( SI Appendix , Fig. S17 ). The total H 2 evolution from precipitated CdSe–MPA in the presence of Δ hyaA was minimal, achieving only 18% of the total H 2 evolved from colloidal CdSe–MPA with Δ hyaA . The decrease in activity is likely due to photooxidation of the QDs. We hypothesize that the interaction between CdSe–MPA and bacteria slows down the photooxidation of CdSe–MPA, helping to sustain activity for over 2 wk. Sustained by EET from bacteria, a quantum yield of 0.2% is achieved here, along with a TON per QD (TON QD ) of 4,160 at neutral pH (see SI Appendix , Table S7 for further details). Previously, TON QD for photocatalytic H 2 evolution from CdSe–MPA using a large excess of ascorbic acid (≥500 mM) as an electron donor has ranged from 10,000 to 20,000, with external quantum yields of approximately 1.5 to 9.5% at an acidic pH of 4.5 ( 25 ). Here, MR-1 survives and supports photochemical H 2 production by CdSe-MPA with a relatively low (~20 mM) concentration of lactate as a primary carbon source. Notably, lactate is abundant in wastewater, present in a variety of food-processing industrial liquid waste, and also as a product of fermentation ( 41 ). Ranging from approximately 3 mM to upward of 100 mM, lactate is already present in these wastewater sources at a concentration necessary to sustain MR-1 growth ( 41 , 42 ). These initial TON QD and quantum yield values from CdSe–MPA sustained by MR-1 are obtained without optimization of the system. As evidenced by previous work involving CdSe–MPA ( 43 ), further optimizing hole transfer to MR-1 will enable an even more active QD system. Herein, we have demonstrated a fully light-driven system that utilizes the natural respiratory pathway of an electrogenic bacterium, Shewanella oneidensis MR-1, to provide electrons to QD photocatalysts. Using CdSe–MPA QDs for H 2 evolution via this route is a promising step toward the sustainable production of chemical fuels. This strategy has a number of benefits, among them that MR-1 uses lactate, an industrial waste product, as an energy source to sustain cell viability and supply respiratory electrons for QD-catalyzed H 2 evolution for over a week. We demonstrate that the viability of MR-1 is not diminished by QDs in catalytic concentrations. A major advantage of a nanocrystalline photocatalysis system driven by electrogenic bacteria is its design flexibility. The interaction between MR-1 and CdSe–MPA may set a foundation to expand to the combination of different electrogenic bacteria and various NCs to enhance electron transfer and biofilm formation ( 44 ). While possible routes of improvement could be focused on genetically engineering bacteria, this work can also be expanded by tailoring NCs. Adding light-harvesting components with the ability to absorb a greater portion of the solar flux may improve quantum yields. In addition, NCs can be tailored to carry out a variety of catalytic reactions. Rich chemistry at the interface of NCs and enzymes for catalysis continues to be explored ( 45 – 47 ). Catalytic cascades taking advantage of both NC photocatalysis and reactions carried out by microbes also can be envisioned. The robust nature of the H 2 evolution system described herein demonstrates the expanding possibilities at the interface of bio- and nano-technology."
} | 4,204 |
26904007 | PMC4748050 | pmc | 3,528 | {
"abstract": "Quorum sensing is known to play a major role in the regulation of secondary metabolite production, especially, antibiotics, and morphogenesis in the phylum Actinobacteria . Although it is one of the largest bacterial phylum, only 25 of the 342 genera have been reported to use quorum sensing. Of these, only nine have accompanying experimental evidence; the rest are only known through bioinformatic analysis of gene/genome sequences. It is evident that this important communication mechanism is not extensively explored in Actinobacteria. In this review, we summarize the different quorum sensing systems while identifying the limitations of the existing screening strategies and addressing the improvements that have taken place in this field in recent years. The γ -butyrolactone system turned out to be almost exclusively limited to this phylum. In addition, methylenomycin furans, AI-2 and other putative AHL-like signaling molecules are also reported in Actinobacteria. The lack of existing screening systems in detecting minute quantities and of a wider range of signaling molecules was a major reason behind the limited information available on quorum sensing in this phylum. However, recent improvements in screening strategies hold a promising future and are likely to increase the discovery of new signaling molecules. Further, the quorum quenching ability in many Actinobacteria has a great potential in controlling the spread of plant and animal pathogens. A systematic and coordinated effort is required to screen and exploit the enormous potential that quorum sensing in the phylum Actinobacteria has to offer for human benefit.",
"conclusion": "Conclusion The enormous metabolic and phylogenetic diversity that exists in Actinobacteria offers a unique opportunity to explore its multifactorial abilities for biotechnological applications. Quorum sensing is one such property that is evidently under-explored in this phylum. Based on the limited information that is known, quorum sensing systems in Actinobacteria show considerable diversity in terms of the types of signals and the mechanisms it controls. However, there exists a taxa specific segregation within the phylum. For instance, GBL-mediated regulation is not only limited to Streptomyces but is also species specific. Interspecific signaling is therefore likely to expand the list of compounds and mechanisms involved in quorum sensing. The lack of good detection systems is a major limitation for further exploration of the communication system in Actinobacteria. Developing newer systems which can respond to a wider range of signals and that too at very low quantities are the need of the hour. Further exploration using these systems within and between multiple taxa is likely to reveal an even greater diversity of signals. Similarly, the quorum quenching ability of Actinobacteria exhibit a great potential, especially through their use as bio-control agents for plant pathogens and in controlling the spread of antibiotic-resistant organisms. However, systematic screening of specific ecosystems is required to fully exploit the quorum quenching potential. Using the knowledge gained from an in-depth understanding of the existing quorum sensing systems, Actinobacteria are likely to exhibit a wider array of properties that are likely to have significant implications for plant, animal and human health.",
"introduction": "Introduction Cell-to-cell communication in bacteria via quorum sensing is a density-dependent regulation of gene expression. The system relies on two major components, a signaling molecule and a transcriptional activator protein. In many Gram-negative bacteria, a member of the N-acylhomoserine lactone (AHL) family acts as a diffusible signal molecule, the synthesis of which is controlled by the members of the LuxI family of synthases ( Figure 1 ). Above a threshold concentration, this signal molecule activates target genes in conjunction with a member of the LuxR family of transcriptional activators ( Fuqua et al., 1996 ). The AHL-based quorum sensing system plays major role in regulating multiple functions such as bioluminescence ( Nealson and Hastings, 1979 ), synthesis of antibiotics ( Bainton et al., 1992 ), the production of virulence factors ( Barber et al., 1997 ), exopolysaccharide biosynthesis ( Beck von Bodman and Farrand, 1995 ), bacterial swarming ( Eberl et al., 1996 ), and plasmid conjugal transfer ( Fuqua and Winans, 1994 ). In contrast, most Gram-positive bacteria use processed oligo-peptides for signaling and communication ( Kleerebezem et al., 1997 ; Sturme et al., 2002 ). These signals, referred to as autoinducing polypeptides (AIPs) are produced in the cytoplasm as precursor peptides and are subsequently cleaved, modified, and exported. The AIP-based quorum-sensing systems are known to regulate the expression of many factors such as genetic competence ( Solomon et al., 1995 ), sporulation ( Magnuson et al., 1994 ), and virulence factor expression ( Qin et al., 2000 ). While it may seem that the differentiation in the type of signaling compound is a consequence of the structural differences in the cell wall between the two bacterial types; however, this is not the case. For instance, certain Actinobacteria (Gram-positive) are known to use γ-butyrolactones for signaling, whereas most Gram-negative bacteria are known to possess signaling peptides as part of their genome ( Lyon and Novick, 2004 ). Regardless of the cell type, quorum sensing is a near universal mode of cell-to-cell communication amongst pathogenic bacteria. Hence, it is now considered an important target for controlling their spread, especially antibiotic resistant bacteria. FIGURE 1 Schematics of quorum sensing systems in bacteria. (A) Gram-negative bacteria: at threshold concentrations the diffusible autoinducer signaling molecule, typically a homoserine lactone, binds to its cognate receptor inside the cell forming the autoinducer-receptor complex which then regulates the expression of target genes through its binding to the target promoter; and (B) Gram-positive bacteria: at threshold concentrations the autoinducer peptide molecule which is actively transported activates the sensor kinase protein inside the cell which phosphorylates the response regulator protein, which then regulates the expression of target genes through its binding to the target promoter. Adapted from Koh et al. (2013) . Despite the diversity and importance of the phenotypes that are regulated by the quorum sensing network, the information on their environmental distribution is very limited. Further, those that are available, only focus on the AHL-mediated gene expression systems. A survey by Manefield and Turner (2002) showed that merely 2.2% (21 bacterial genera) of the total number of bacterial genera listed in the Bergey’s Manual of Systematic Bacteriology ( Garrity et al., 2001 ) are known to harbor the AHL producing species, and all of which belong to the alpha, beta and gamma proteobacteria only. At the species level, the percentage of AHL producers drops to a fraction of a percent. Although the estimate is more than a decade old, it still reflects on the state of the information available on quorum sensing in bacteria. Motivated by this lack of information, our screening for luxRI homologs and AHL production in the genus Aeromonas not only revealed that the homologs are universally present in this genus, but also reported that a wide diversity of AHLs are secreted by the species in the genus ( Jangid et al., 2007 , 2012 ). This study only points to the fact that quorum sensing is indeed a widespread phenomenon among bacteria, however, a systematic evidence is lacking. Thus, there is a need to survey the existence and study the taxonomic distribution of the quorum sensing systems amongst bacterial taxa. The phylum Actinobacteria is one of the largest phyla within domain Bacteria and consists of six classes, 23 orders including one Incerta sedis and 53 families ( Ludwig et al., 2012 ). As of December 2015, there were 342 genera in this phylum with standing in nomenclature as determined from the LPSN database ( Parte, 2015 ). Actinobacteria are typically Gram-positive but at times stain-variable and have a rigid cell wall that contains muramic acid with some containing wall teichoic acids. The phylum comprises of a plethora of phenotypically diverse organisms, with widespread distribution in nature and exhibiting varied oxygen, nutritional, temperature, and pH requirements for growth, making it an important phylum. Their diverse physiological potential makes Actinobacteria a dominant role player in the biotechnology industry. Their applications are widespread and vary from agroindustry, pharmaceuticals, bioremediation among numerous others. They play a key role in natural geochemical cycles, especially through their ability to decompose organic matter. Actinobacteria are also abundant in the rhizosphere and produce a wide range of biologically active metabolites, thereby influencing plant development ( Selvakumar et al., 2014 ). Many Actinobacteria are also known pathogens of plants and animals. However, amongst the most important potential of Actinobacteria, it is the production of a significant number of secondary metabolites like antibiotics and other compounds of biotechnological interest that has been exploited most. For instance, among the polyene macrolides, a class of polyketides which are antifungal compounds, are synthesized by more than 100 different species of actinomycetes ( Recio et al., 2004 ). In addition, members of the genus Streptomyces are known to produce more than 70% of commercially available antibiotics ( Weber et al., 2003 ). The expression of virulence determinants, production of secondary metabolites, and morphogenesis is associated with high cell densities and typically controlled by diffusible low molecular weight chemical substances, similar to the Gram-negative autoinducer, suggesting a role of quorum sensing in regulating these mechanisms ( Takano, 2006 ; Santos et al., 2012 ). Further exploration of novel phenotypes under quorum sensing regulations is likely to contribute to the advancement in medical, biotechnological and ecological fields. Hence, there is a need of studying quorum sensing in Actinobacteria. Most of what is known about quorum sensing in Actinobacteria, comes from the study of antibiotic production in this taxa. While it is indeed the most important phenomenon, the aim of this review is not to present an overview on the quorum sensing regulation of antibiotic production. The reader is therefore directed to read Takano (2006) , Liu et al. (2013) and references within. Further, for clarity Actinobacteria means all species within the phylum Actinobacteria , unless otherwise stated as class Actinobacteria. In this review, we present an overview of quorum sensing systems described so far for the phylum Actinobacteria , indicating the limitations of existing screening strategies and addressing improvements in newer technologies for the discovery of quorum sensing in more taxa. In addition, we summarize the current status of known quorum quenching activity in this phylum."
} | 2,806 |
37379422 | PMC10134890 | pmc | 3,529 | {
"abstract": "Abstract Sponges perform important ecosystem functions, host diverse microbial symbiont communities (microbiomes), and have been increasing in density on Caribbean coral reefs over the last decade. Sponges compete for space in coral reef communities through both morphological and allelopathic strategies, but no studies of microbiome impacts during these interactions have been conducted. Microbiome alterations mediate spatial competition in other coral reef invertebrates and may similarly impact competitive outcomes for sponges. In this study, we characterized the microbiomes of three common Caribbean sponges ( Agelas tubulata , Iotrochota birotulata , and Xestospongia muta ) observed to naturally interact spatially in Key Largo, Florida (USA). For each species, replicate samples were collected from sponges in contact with neighbors at the site of contact (contact) and distant from the site of contact (no contact), and from sponges spatially isolated from neighbors (control). Next‐generation amplicon sequencing (V4 region of 16S rRNA) revealed significant differences in microbial community structure and diversity among sponge species, but no significant effects were observed within sponge species across all contact states and competitor pairings, indicating no large community shifts in response to direct contact. At a finer scale, particular symbiont taxa (operational taxonomic units at 97% sequence identity, OTUs) were shown to decrease significantly in some interaction pairings, suggesting localized effects for specific sponge competitors. Overall, these results revealed that direct contact during spatial competition does not significantly alter microbial community composition or structure of interacting sponges, suggesting that allelopathic interactions and competitive outcomes are not mediated by microbiome damage or destabilization.",
"conclusion": "5 CONCLUSION In summary, our data revealed that sponge‐to‐sponge spatial contact does not affect overall microbial community composition and structure for the investigated Caribbean host species, with only minor shifts (changes in individual OTUs) occurring from indirect impacts of the interaction (e.g., shading). These results highlight the stability of sponge microbial communities during spatial interactions and suggest that microbiome disruption is not the main mechanism of host damage from allelopathic interactions and has minimal impacts on spatial competition outcomes among Caribbean sponges. Future investigations targeting additional host sponge species and experimenting with forced interactions among sponges will provide additional insight into the interplay between microbiome structure, chemical defenses, and spatial competition in coral reef invertebrates. Ultimately, a clear understanding of sponge‐to‐sponge competition may yield insights into which species of sponges will dominate and shape future Caribbean reefs as corals continue to decline.",
"introduction": "1 INTRODUCTION Sponges represent a diverse invertebrate lineage containing over 8,500 identified species (Van Soest et al., 2012 ). They are sessile organisms recognized for their remarkable filtering processes (Milanese et al., 2003 ) and are known to possess symbiotic microorganisms capable of nutrient cycling within their mesohyl (Bayer et al., 2014 ; Hoffmann et al., 2009 ; Schläppy et al., 2010 ). Accordingly, the sponge and its complex microbial community (“microbiome”) have been used to address ecological questions pertaining to the sponge host and its environmental impacts. Furthermore, sponges have shown trends of increasing in biomass on Caribbean coral reefs over the last decade (McMurray et al., 2010 ), are already dominant community members (60% of reef cover) on some Caribbean reefs (Loh et al., 2015 ), and are known to compete allelopathically with other coral reef species (Chadwick & Morrow, 2011 ; Slattery & Lesser, 2021 ). Thus, understanding how sponges interact with other coral reef community members, including other sponge species, is important for future conservation efforts and predictive forecasts in coral reef communities. Previous work on Caribbean coral reefs has shown that most sponges grow in contact with (28.6%) or in proximity to (31.0%) other sponges, with the remaining individuals (40.4%) observed growing in isolation (Engel & Pawlik, 2005 ). Such direct contact or proximity to neighboring sponge individuals can result in tissue damage, impeded growth, and over‐growth as the organisms compete for space and ambient resources. These interactions are mediated by differential sponge growth rates, chemical defenses (i.e., allelopathy), and ambient predation pressures. Indeed, predation levels can interact with spatial competition pressures, such as physical or chemical defenses utilized by the host to deter sponge predation (Chanas & Pawlik, 1996 ; Pawlik et al., 1995 ; Uriz et al., 1996 ) may also represent allelopathic chemicals that assist in spatial competition with other sponge species or corals (Engel & Pawlik, 2000 ; Pawlik et al., 2007 ; Porter & Targett, 1988 ). Previous work has suggested that sponges vary in allelopathic chemical defenses on a species‐by‐species basis (Assmann et al., 2004 ; Engel & Pawlik, 2000 ; Proksch, 1994 ; Waddell & Pawlik, 2000 ) and that intraspecific variation in defenses occurs within some sponge species (Chanas & Pawlik, 1997 ). These intraspecific variations in allelopathy do not correlate with the sponge's size or its ability to compete spatially (Chanas & Pawlik, 1997 ) but may be affected instead by the level of predation within the sponge's environment. Sponge microbiome characterization is increasingly being incorporated into the study of sponge health and ecological function, facilitated by more affordable and rapid DNA sequencing technologies. Previous research has shown that sponges host abundant and complex microbial communities (Taylor et al., 2007 , 2013 ; Thomas et al., 2010 , 2016 ) that are distinct from the free‐living microbial assemblages (Gantt et al., 2017 ; Hentschel et al., 2002 ; Weigel & Erwin, 2016 ). These microbial communities are generally sponge species‐specific, even across great distances (Hentschel et al., 2002 ; Lee et al., 2011 ), and host sponges exist in two main categories based on the abundance and diversity of their associated microbes (Gloeckner et al., 2014 ; Poppell et al., 2014 ; Schöttner et al., 2013 ): high microbial abundance (HMA) sponges that contain 10 8 –10 10 bacteria cells per gram of sponge (2–4 orders of magnitude greater than seawater, Hentschel et al., 2006 ) and low microbial abundance (LMA) sponges that host microbial communities at concentrations similar to seawater (10 6 –10 8 bacteria cells per gram of sponge, Hentschel et al., 2006 ). Previous studies have applied microbial community analyses to assess sponge health (Webster et al., 2002 ), the role of sponges in nutrient cycling within coral reef communities (Gantt et al., 2019 ; Hoffmann et al., 2009 ; Rix et al., 2016 ), and climate change impacts on sponge functioning and survival (Lemoine et al., 2007 ; Lesser et al., 2016 ). Despite the importance of the sponge microbiome to host health and ecology (Pita et al., 2018 ; Slaby et al., 2019 ) and contributions to secondary metabolite synthesis (Helber et al., 2019 ; Liu et al., 2019 ), the impact of spatial competition on the structure of microbial communities in sponges has not been investigated. To characterize the ecological aspects of sponge communities more fully in coral reef ecosystems, the current study investigated microbial community effects from the spatial competition (i.e., direct contact) among interacting sponge species. In this study, we posed three hypotheses: (1) sponge microbiomes will differ among host species, (2) sponge microbiomes will differ within each host species between tissue in contact with neighboring sponges versus no contact and control tissues, and (3) intraspecific microbiome shifts will vary by host and competitor pairing. To test these hypotheses, sponge tissue was sampled from three common Caribbean sponges ( Agelas tubulata , Iotrochota birotulata , and Xestospongia muta ) observed to naturally interact spatially and microbial communities were characterized using partial 16S rRNA gene sequences (V4 region). Differences in sponge microbiomes were assessed at the community and operational taxonomic unit (97% sequence identity, OTU) levels for each interaction type and sponge competitor to assess the effects of sponge‐to‐sponge contact on the composition and structure of host‐associated microbial communities.",
"discussion": "4 DISCUSSION The microbial communities from each sponge species investigated ( A. tubulata , I. birotulata , and X. muta ) differed significantly from each other and from free‐living microbial communities in the surrounding environment, consistent with previous studies (Hentschel et al., 2012 ; Jackson et al., 2012 ). The LMA sponge, I. birotulata , hosted microbial communities more similar to seawater communities than those in HMA sponges ( A. tubulata and X. muta ), with significantly lower diversity and increased Proteobacteria presence than HMA counterparts, supporting results from past HMA‐LMA sponge studies and comparisons (Gantt et al., 2019 ; Giles et al., 2013 ; Gloeckner et al., 2014 ; Poppell et al., 2014 ). These characteristic patterns in sponge microbial communities are likely related to differences in host physiology and pumping rates between HMA and LMA sponges (Poppell et al., 2014 ; Weisz et al., 2008 ). Notably, distinct microbiomes were also observed between the two HMA hosts, indicating a strong influence of host species on microbial community structure. Within each host sponge species, microbial community analyses revealed high microbiome stability across interaction types, suggesting that allelopathic interactions and spatial competition outcomes among coral reef sponges were not mediated by microbiome damage or destabilization. This is surprising since the microbial communities of some sponges are involved in allelopathic chemical production (Rust et al., 2020 ; Tianero et al., 2019 ) and microbiome disruption from spatial competition has been documented in corals (Pawlik et al., 2007 ; Thinesh et al., 2020 ; Vega Thurber et al., 2012 ). Of the sponges investigated herein, both A. tubulata and X. muta utilize allelopathy during the competition (Assmann et al., 2004 ; Kelly et al., 2003 ; Waddell & Pawlik, 2000 ), while I. birotulata is not known to utilize chemical defenses to deter predation or overgrowth (Engel & Pawlik, 2000 ; Pawlik et al., 1995 ). Our data show little effect of direct tissue contact on sponge microbial communities and suggest that direct damage to host cells is the primary mechanism of allelopathic interactions among sponges. In contrast, coral microbiome disruption appears to be an important mechanism of spatial interactions on reefs. For example, some macroalgae utilize dissolved organic carbon release to disrupt coral microbiomes (Smith et al., 2006 ; Vega Thurber et al., 2012 ) and harbor coral pathogens in their microbial communities that may aid in spatial competition (Barott et al., 2012 ). Direct contact with sponges can also disrupt coral microbiomes (Thinesh et al., 2020 ), further evidence linking microbiome stability and spatial interaction mechanisms on coral reefs. These results follow a general pattern of greater microbiome stability in sponge hosts compared to coral counterparts. Coral microbiomes have been shown to change in response to temperature fluctuations (Maher et al., 2019 ), season (Glasl et al., 2020 ), pollution (Joyner et al., 2015 ), and disease (Slaby et al., 2019 ), among other factors. Sponge microbiomes (at least in shallow water habitats) have shown resistance to perturbations and stability in response to stressors such as pollution (Gantt et al., 2017 ), elevated temperatures (Luter et al., 2012 ; Pita et al., 2013 ), and ocean acidification (Kandler et al., 2018 ). Sponges also show strong stability in community composition and structure across seasons (Erwin et al., 2012 , 2015 ), when losing or acquiring photosymbionts (Britstein et al., 2020 ) and during periods of food shortage stress (Pita et al., 2013 ). Results from the current study add allelopathic interactions to the growing list of environmental factors that do not alter sponge microbiomes and highlight differences in microbiome stability across coral reef species that may inform predictions of host success under changing environmental conditions. While sponge microbiomes were stable at the community level, fine‐scale analyses revealed changes in the relative abundance of some symbiont OTUs across interaction types, primarily at contact sites. These trends indicate minor shifts in microbial communities when sponge tissue is in contact with other sponge competitors, as most OTUs that varied across interaction types were rare members of the sponge microbiomes. Only nine symbiont OTUs (of the >9,000 identified) varied significantly at contact sites and exhibited >1% relative abundance: two OTUs in the phylum Chloroflexi and seven OTUs in the phylum Proteobacteria. Some of these microbial shifts may result from altered physical (e.g., shading) and chemical conditions at contact sites. In A. tubulata , OTU010 (Gammaproteobacteria) decreased in relative abundance in contact tissue, while OTU014 (Chloroflexi) increased two‐fold in contact tissue. OTU010 was affiliated with Ectothiorhodospiraceae, a family of purple sulfur bacteria that utilizes photosynthesis (Henry & Cogdell, 2013 ), thus shifts in the abundance of this OTU may result from shading at contact sites. OTU014 was affiliated with SAR202, a group of heterotrophic, free‐living cells known to be sulfite‐oxidizers (Mehrshad et al., 2018 ). In X. muta , two proteobacterial OTUs (OTU029, 094) decreased in relative abundance in contact tissue, with OTU094 classified as Rhodospirillaceae, a family of purple non‐sulfur photosynthetic bacteria (Kim et al., 2011 ). Therefore, while the overall stability of microbial communities across interaction types supports minimal impacts of allelopathic competition on sponge microbiomes, particular OTUs may exhibit shifts from localized indirect impacts during direct tissue contact with competitors, such as shading and abrasion."
} | 3,620 |
38489250 | PMC10942071 | pmc | 3,530 | {
"abstract": "The networks proposed here show how neurons can be connected to form flip-flops, the basic building blocks in sequential logic systems. The novel neural flip-flops (NFFs) are explicit, dynamic, and can generate known phenomena of short-term memory. For each network design, all neurons, connections, and types of synapses are shown explicitly. The neurons’ operation depends only on explicitly stated, minimal properties of excitement and inhibition. This operation is dynamic in the sense that the level of neuron activity is the only cellular change, making the NFFs’ operation consistent with the speed of most brain functions. Memory tests have shown that certain neurons fire continuously at a high frequency while information is held in short-term memory. These neurons exhibit seven characteristics associated with memory formation, retention, retrieval, termination, and errors. One of the neurons in each of the NFFs produces all of the characteristics. This neuron and a second neighboring neuron together predict eight unknown phenomena. These predictions can be tested by the same methods that led to the discovery of the first seven phenomena. NFFs, together with a decoder from a previous paper, suggest a resolution to the longstanding controversy of whether short-term memory depends on neurons firing persistently or in brief, coordinated bursts. Two novel NFFs are composed of two and four neurons. Their designs follow directly from a standard electronic flip-flop design by moving each negation symbol from one end of the connection to the other. This does not affect the logic of the network, but it changes the logic of each component to a logic function that can be implemented by a single neuron. This transformation is reversible and is apparently new to engineering as well as neuroscience.",
"introduction": "1. Introduction This article is the fourth in a series of articles that show how neurons can be connected to process information. The first three articles [ 1 – 3 ] explored the analog properties of neuron signals in combinational logic operations, whose outputs depend only on the current state of the inputs. A fuzzy logic decoder was shown to generate the major phenomena of both olfaction and color vision (such as color mixing, mutually exclusive colors, and the shape of perceived color space), including the brain’s shortcomings (such as the Bezold-Brücke hue shift) [ 1 , 2 ]. The decoder’s design is radically different from a standard electronic digital (Boolean logic) decoder [ 2 , 4 , 5 ]. If implemented with electronic components and given digital inputs, the decoder performs the same Boolean function as the standard digital design more efficiently. It was shown that a single neuron with one excitatory input and one inhibitory input, with signal strengths X and Y, respectively, can function as a logic primitive, X AND NOT Y [ 1 , 2 ]. In simplest terms, this is because the neuron is active when it has excitatory input and does not have inhibitory input. It was also shown that an AND-NOT gate can be configured to function as an inverter (i.e., a NOT X logic primitive). The AND-NOT gate together with a NOT gate make up a functionally complete set, meaning any logic function can be performed by a network of such components. The neuron AND-NOT gate will be reviewed here and used in the proposed networks. The present article considers the Boolean logic properties of neuron signals in sequential logic operations, whose outputs are functions of both the current inputs and the past sequence of inputs. That a neuron can operate as a functionally complete logic gate, analog or digital, provides a framework for the brain’s processing of information—analog and digital, combinational and sequential. Flip-flops are the basic building blocks of sequential logic systems. A flip-flop is a mechanism that can be set repeatedly to either one of two stable states, commonly labeled 0 and 1. A flip-flop can be used as a memory mechanism to store one bit of information. It is shown here that a few AND-NOT gates can be connected to perform the same function as two standard electronic flip-flops, an active low and an active high Set-Reset (SR) flip-flop. These are not the only flip-flops that can be constructed with AND-NOT gates, but they may be the simplest. The network designs are modifications of standard electronic logic circuit designs. It is shown here that the NFF designs are derived directly from the standard electronic designs simply by moving each negation circle from one end of the connection to the other. This changes the logic of each component, but it does not materially affect the logic of the network. The modifications are necessary to implement the circuits with neurons because the AND-NOT gate is virtually never used as a building block in electronic computational systems. The NFFs produce both known and testable, unknown phenomena of short-term memory. With inputs from the outputs of NFFs, neural decoders proposed in [ 2 ] can retrieve encoded information that is held in NFFs. That is, a memory can be recalled. The NFFs’ robust operation in the presence of noise is demonstrated here by simulation, but the properties can be proven directly from the explicit network connections and minimal neuron properties of excitation and inhibition. In [ 6 ] it was shown that NFFs, together with a network that can produce the oscillations commonly known as brainwaves, suggest a resolution to the longstanding controversy of whether short-term memory depends on neurons firing persistently or in brief, coordinated bursts [ 7 , 8 ]. The NFFs’ operation is dynamic, meaning the only changes are the levels of neuron activity. No structural change is required, such as neurogenesis, synaptogenesis, or pruning, nor is any change required in the way neurons function, such as a change in synaptic strength or the strength of action potentials. This makes the networks’ speed consistent with the “real time” of most brain functions (a few milliseconds). The NFFs’ architectures are explicit, meaning all neurons, connections, and types of synapses are shown explicitly, and all assumptions of neuron capabilities are stated explicitly. Only minimal neuron capabilities are assumed, and no network capabilities are assumed. It was shown in [ 9 ] that designing a simple logic circuit that can perform a single, biologically advantageous task can lead to a discovery of how neurons are connected to process information. This is the method that was used to find the networks proposed here and in [ 1 – 6 ]. Besides performing a biologically useful task, the networks are dynamic, explicit, and able to generate phenomena that are central to a particular brain function. These four properties are characteristics that networks in the brain must have. The neuron properties used to achieve the results for these networks—excitation, inhibition, and sigmoid neuron responses—have been known a long time."
} | 1,742 |
33150067 | PMC7585372 | pmc | 3,531 | {
"abstract": "Plant-microbe associations are increasingly recognized as an inextricable part of plant biology and biogeochemistry. Microbes play an essential role in the survival and development of plants, allowing them to thrive in diverse environments. The composition of the rhizosphere soil microbial communities is largely influenced by edaphic conditions and plant species. In order to decipher how environmental conditions on a mine site can influence the dynamics of microbial communities, we characterized the rhizosphere soil microbial communities associated with paper birch, speckled alder, and spruce that had naturally colonized an acidogenic mine tailings deposit containing heavy metals. The study site, which had been largely undisturbed for five decades, had highly variable vegetation density; with some areas remaining almost barren, and others having a few stands or large thickets of mature trees. Using Illumina sequencing and ordination analyses (redundancy analysis and principal coordinate analysis), our study showed that soil bacterial and fungal community structures correlated mainly with vegetation density, and plant species. Tailings without any vegetation were the most different in bacterial community structure, compared to all other areas on the mine site, as well as an adjacent natural forest (comparison plot). The bacterial genera Acidiferrobacter and Leptospirillum were more abundant in tailings without vegetation than in any of the other sites, while Bradyrhizobium sp. were more abundant in areas of the tailings deposit having higher vegetation density. Frankia sp. is equally represented in each of the vegetation densities and Pseudomonas sp. present a greater relative abundance in boreal forest. Furthermore, alder rhizosphere showed a greater relative abundance of Bradyrhizobium sp. (in comparison with birch and spruce) as well as Haliangium sp. (in comparison with birch). In contrast, fungal community structures were similar across the tailings deposit regardless of vegetation density, showing a greater relative abundance of Hypocrea sp. Tailings deposit fungal communities were distinct from those found in boreal forest soils. Alder rhizosphere had greater relative abundances of Hypocrea sp. and Thelephora sp., while birch rhizosphere were more often associated with Mollisia sp. Our results indicate that, with increasing vegetation density on the mine site, the bacterial communities associated with the individual deciduous or coniferous species studied were increasingly similar to the bacterial communities found in the adjacent forest. In order to properly assess and restore disturbed sites, it is important to characterize and understand the plant-microbe associations that occur since they likely improve plant fitness in these harsh environments.",
"conclusion": "Conclusions Our results indicate that vegetation density classes (VDC) and plant species best explained the bacterial and fungal community structures. Indeed, PCoA showed significant differences in total microbial communities between the various VDC as well as plant species. These two parameters are intimately related, since the plants, by their root exudates, are able to select the type of microorganisms composing their rhizospheres. Moreover, the plants that colonize these soils change them differently, from one VDC to another, thus producing different physical and chemical properties. These may or may not stimulate plants to structure their rhizosphere soil microbial communities in different ways. The capability of complex plant assemblages to colonize a tailings storage area with these properties is a demonstration of autonomous recovery following human disturbance. We observed that as vegetation density increased and began in some sectors to resemble that of the adjacent natural forest, microbial communities associated with alder, spruce, and birch varied. Indeed, while plant species is a known determinant of microbial community composition in the rhizosphere, this study revealed that the parameter of vegetation density also explained community structure. This is an important observation since the capability of nutrient-poor substrates to sustain plant growth can be improved by the presence of bacterial and fungal communities capable of weathering minerals and increase water and nutrient acquisition in plants. Microbial communities and more microbial diversity can improve survival and development of plants. They can contribute directly and indirectly to resilience of these recovering environments. By extension, it is worthwhile to consider planting biodiverse plant thickets in reclamation efforts. While our study begins to shed light into the associations, and specific fungal and bacterial taxa, that were found to thrive in this particular environment, subsequent studies will be required to deepen our understanding of the specific plant-microbe associations we observed. Ultimately, this will help us better harness plant-microbe and plant-plant associations to improve and accelerate the ecological restoration of anthropized environments.",
"introduction": "Introduction In Québec, the first gold-mining operation began in the 1840s ( Ministère de l’Énergie et des Ressources naturelles, 2018 ). At the time, regulations concerning mine site restoration did not exist, and once mining operations were completed, sites were generally abandoned. It is only recently that legislation concerning the obligation to restore mine sites has been put in place. In the meantime, 499 mining sites are considered abandoned in Québec ( Ministère de l’Énergie et des Ressources naturelles, 2019 ). The abandoned sites have several common environmental problems, such as acidification and general soil degradation through the oxidation of sulfide minerals (such as iron pyrite (FeS 2 )) in the presence of water, oxygen and microorganisms and eventually through the production of acid mine drainage (AMD) ( Kefeni, Msagati & Mamba, 2017 ; Simate & Ndlovu, 2014 ). The nature of acid mine drainage (acidic pH and high metal(loid) concentration) alter the physical, chemical and biological structure of the ecosystem. This, combined with low organic matter content (top soil removed by mining activities), hinders plant colonization ( Fellet et al., 2011 ; Wang et al., 2017 ). Moreover, the acidic pH that characterizes mine tailings, in addition to low plant density, result in low clay-humic complexes in these soils, leading to weak or almost no absorption, aggregation and sedimentation of metal(loid)s ( Agbenyeku, Muzenda & Msibi, 2016 ; Akcil & Koldas, 2006 ; Pandey, Pandey & Misra, 2000 ). Thus, free metals in soils are able to migrate with runoff to streams and nearby forests unaffected by mining activities and this transfer of heavy metals into plants, water and the food chain can cause a public health problem ( Ali, Khan & Sajad, 2013 ; Baalousha, Motelica-Heino & Le, 2006 ; Simate & Ndlovu, 2014 ). Several remediation techniques including mechanical separation, pyrometallurgical separation, and chemical soil treatments are possible, but these techniques are costly in addition to being slow to implement ( Mulligan, Yong & Gibbs, 2001 ). The use of plants to remedy mining sites is an increasingly popular measure, due to its low cost and ease of implementation. One such plant remedy is phytostabilization, which consists of reducing the mobility of soil contaminants using plants and their associated microorganisms ( Ali, Khan & Sajad, 2013 ; Ma et al., 2011 ). In a mining context, the soils make the establishment of plants very difficult, since the plants are lacking in macroelements (i.e., C, N, P) essential for growth and they are confronted with constant oxidative stress caused by the excessive concentration of heavy metals ( Sheoran, Sheoran & Poonia, 2010 ; Simate & Ndlovu, 2014 ). The critical period of establishment of a plant could be greatly facilitated by the inoculation of roots by microorganisms that can increase the adaptability and resilience of the plant and therefore enhance the success of phytostabilization ( Grossnickel, 2005 ; Phieler, Voit & Kothe, 2013 ; Zhuang et al., 2007 ). Certain microorganisms in the rhizosphere, including PGPR bacteria (Plant Growth-Promoting Rhizobacteria) and mycorrhizal fungi, promote plant growth and allow the plant to adapt more easily to harsh environments ( Bulgarelli et al., 2013 ; Sasse, Martinoia & Northen, 2018 ; Thijs et al., 2016 ). As an example, PGPR organisms can fix atmospheric nitrogen or solubilize phosphates, making these limiting macronutrients available to the plant. PGPR organisms could also secrete phytohormones (mostly auxins), that could enhance the germination and growth of the plant, and siderophores that promote metal mobility as they increase solubility ( Berendsen, Pieterse & Bakker, 2012 ; Rajkumar et al., 2012 ). Moreover, mycorrhizal fungi in association with plant roots extend their mycelial networks in the rhizosphere soil which leads to higher nutrient (i.e., P, N) and water capture than plant roots themselves ( Jung et al., 2012 ). Indeed, Gamalero et al. (2009) reported that arbuscular mycorrhizal fungi could modify the root architecture of plants (depending on the host) thus resulting in more branching roots and eventually by increasing the nutrient uptake by the plant. Some arbuscular fungi (AM) and ectomycorrhizal (ECM) fungi act as buffers between the soil and the plant roots which could decrease heavy metal bioavailability by chelating them in their vacuoles, improving plant resistance to heavy metal stress in soil ( Colpaert et al., 2011 ; Khan, 2005 ; Miransari, 2010 ; Thijs et al., 2016 ). The mine site studied was characterized as being naturally colonized by trees, shrubs, and grasses. The vegetation density was quite variable on site, from a lack of plants to dense groves composed of several species ( Tardif et al., 2019 ). The mine tailings had a pH ranging from 3 to 9, and arsenic concentrations ranging from 1 to 1951 ppm. The objectives of this study were to determine the main environmental parameters that structured the microbial communities in soil and the rhizosphere, and begin to decipher how environmental conditions on a mine site can influence the dynamics of microbial communities, which are a key factor in plant survival and development in the natural environment. Understanding how plant-microbial associations developed in a mining context could help develop tailored microbial inocula to prepare seedlings used in mine site restoration and facilitate their establishment.",
"discussion": "Discussion Vegetation density classes and plant species as drivers of microbial communities Our results showed that the vegetation density classes (VDC) and plant species (p_s) significantly influenced the structure of the microbial (bacterial and fungal) communities present on the study site ( Figs. 2 and 3 ). The vegetation density classes presented in this paper ranged from a total absence of plants (VDC-1) to a boreal forest, unaffected by human activities (VDC-6). In a mining context, acidic and poor-nutrient tailings make plant establishment difficult, leading to a low carbon concentration in soil, which was observed on tailings (VDC-1, 2, 3 and 4). These VDC had an average organic matter content of 2.1% compared to 22.39% for VDC-5 and 6 ( Table S2 ). The presence of plants is known to change soil properties through the secretion of root exudates composed of organic acids, H + , chelators, and phytosiderophores, which influence pH, the stability of aggregates, and nutrient availability ( Anger & Caron, 1998 ; Gould et al., 2016 ; Philippot et al., 2009 ; Van der Putten et al., 2013 ). The absence of vegetation in VDC-1 leads to a different soil structure and lack of root exudates, which are truly important in the structure of the bacterial communities ( Chaparro et al., 2013 ; Marschner et al., 2001 ). All these factors could explain why VDC and plant species were the main factors structuring microbial communities in RDAs, as well as the results observed in PCoAs where VDC-1 showed the greatest difference in the bacterial community structure in comparison to all other VDC ( Figs. 2 and 3A and Table 1A ). Our study also showed that fungal communities were different in the boreal forest soils (VDC-5 and 6) compared to those found on the tailings deposit (VDC-1 to 4) ( Fig. 3B and Table 1B ). This might be due to the differences in soil carbon levels. A study conducted by Liu et al. (2015a) as well as Sun et al. (2016) reported that carbon concentration in soil was indeed one of the major determinants of fungal community structure. Several studies have shown that plant species is one of the major determinants of the microbial composition in the rhizosphere soil ( Cavaglieri, Orlando & Etcheverry, 2009 ; Chaparro et al., 2013 ). Root exudates vary according to plant species, age of the plant, its state of stress, as well as its stage of development ( Cavaglieri, Orlando & Etcheverry, 2009 ; Chaparro et al., 2013 ; Turner, James & Poole, 2013 ). It was also showed by Prescott & Grayston (2013) , as well as Urbanovà, Šnajdr & Baldrian (2015) that trees mostly influenced composition of fungal and bacterial communities. Our results also showed differences in the rhizospheres of different plant species. PCoA results showed significant differences of bacterial communities in the rhizosphere soil of alder compared to birch and different fungal communities in the rhizosphere soil of alder compared to spruce ( Fig. 3 ). We measured the highest relative abundance of Actinobacteria in alder rhizosphere (4.68%) in comparison to birch (3.30%) and spruce (3.74%) rhizospheres ( Fig. 4 and Table S7A ). As mentioned previously, plant species can harbour a rhizomicrobiome that is specifically beneficial ( Berendsen, Pieterse & Bakker, 2012 ). The fact that alders establish a root nodule-forming symbiosis with Actinobacteria of the genus Frankia might explain the higher relative abundance of Actinobacteria in their rhizosphere. Additionally, this symbiosis requires specific nutrients (i.e., Mo, Mg, Fe and P) to support its nitrogen fixation activity, which might in turn influence the plant-microbe associations that tend to develop ( Bélanger, Bellenger & Roy, 2013 ; Kabata-Pendias, 2001 ; Ngom et al., 2016 ). Bacteria associated with vegetation density classes and plant species This study revealed that the six most abundant classes of bacteria for each VDC and p_s were Gammaproteobacteria, Deltaproteobacteria, Alphaproteobacteria, Planctomycetacia, Acidobacteria, and Actinobacteria and represented approximately 50% of the microbial community classes ( Fig. 4 and Table S7A ). A study conducted by Mendez, Neilson & Maier (2008) reported that the main bacterial classes associated with iron-rich tailings with low pH (2.7) to moderate pH (5.7) were Gammaproteobacteria, Alphaproteobacteria, Acidobacteria, Actinobacteria, Nitrospira and Firmicutes, which corroborates some of our results. Hur et al. (2011) showed that zinc mine tailings with low pH, low organic matter as well as high Zn, Cd and Pb concentrations were dominated by Proteobacteria, Actinobacteria and Acidobacteria. The studies done by Hur et al. (2011) as well as Mendez, Neilson & Maier (2008) corroborate what was observed on the studied mine site, that Proteobacteria, Acidobacteria and Actinobacteria were the dominant taxa colonizing mine tailings. As mentioned above, the rhizosphere soil is the nutrient rich, high biodiversity zone surrounding plant roots. The rhizosphere soil is a beneficial niche for many key microorganisms ( Prashar, Kapoor & Sachdeva, 2014 ). In VDC-1 where there was little to no plant colonization and where pH was the lowest (3.9) compared to other VDC, we found the highest relative abundance of the chemoautotrophs Acidiferrobacter sp. and Leptospirillum sp. This could be expected since Acidiferrobacter sp. derive their energy through the oxidation of various inorganic compounds and Leptospirillum sp. does so through the oxidation of iron, which is present in high concentrations (31,000 ppm) in VDC-1 ( Table S2 ) ( Tyson et al., 2000 ). The low organic matter content of the tailings (average of 2.1% in (VDC1 to 4) ( Gagnon et al., 2020 ) likely limited the development of heterotrophic microbial communities in tailings. This could have also contributed to the relative abundance of chemoautotrophs in VDC-1. Such chemoautotrophic organisms tend to be less active in the presence of plants because the latter secrete many carbon-rich molecules, which allows heterotrophs to dominate in their presence ( Bais et al., 2006 ; Turner, James & Poole, 2013 ). It is known that PGPR organisms facilitate plant growth in abiotic-stressed environments ( Grandlic et al., 2008 ; Sasse, Martinoia & Northen, 2018 ; Thijs et al., 2016 ; Weyens et al., 2009 ; Zhuang et al., 2007 ). Berendsen, Pieterse & Bakker (2012) , as well as Santoyo et al. (2016) , showed that plants could select and shape their rhizomicrobiome as a function of their homeostatic states and for meeting their needs. Plants growing in metal-contaminated soils have different rhizomicrobiomes compared to plants growing in non-contaminated soils; they recruit beneficial microorganisms to support their growth and limit their metal uptake ( Sasse, Martinoia & Northen, 2018 ; Thijs et al., 2016 ). Microorganisms in the rhizosphere soil can stimulate plant growth by providing nitrogen, solubilizing phosphorus, producing phytohormones (i.e., auxins, cytokinins), synthesizing siderophores, and protecting plants from pathogens ( Glick, 2010 ; Prashar, Kapoor & Sachdeva, 2014 ). We observed the presence of nitrogen-fixing, symbiotic organisms such as Bradyrhizobium sp., Frankia sp., and Rhizobium sp. in VDC-2, 3, and 4. These nitrogen fixers may have improved the competitiveness of plants harbored in these nitrogen-poor substrates ( Bais et al., 2006 ; Ma et al., 2011 ; Mus et al., 2016 ; Wani, Khan & Zaidi, 2007 ) as VDC-2 to 4 contained less than 0.25 ppm N ( Table S2 ). Bradyrhizobium sp. have been reported to benefit plants growing in metal-contaminated soil. For example, Bradyrhizodium sp. ( vigna ) was shown to improve growth, nodulation, nitrogen content, and reduce the uptake of Ni and Zn in green gram ( Vigna radiata) . This occurred through the production of indole-3-acetic acid, siderophores and ammonia ( Ma et al., 2011 ; Wani, Khan & Zaidi, 2007 ). Despite many PGPR traits reported in Pseudomonas sp. in contaminated soils (i.e., mine tailings), our study showed ( Fig. 6E ) that the relative abundance of Pseudomonas sp. was higher in boreal forest soils (VDC-5 and 6) compared to that of the tailings storage area (VDC-1 to 4) ( Glick, 2010 ; Ma et al., 2011 ). Pseudomonas sp. are known for their production of organic acids that solubilize mineral phosphorus, and the production of phosphatase/phytase that hydrolyzes phosphate-organic compounds, increasing the availability of this limiting nutrient in soils ( Hunter, Teakle & Bending, 2014 ; Richardson et al., 2009 ; Vacheron et al., 2013 ). In our study, average phosphate concentrations were not systematically higher in boreal forest soils (VDC-5 and VDC-6; 90.8 and 57.9 ppm, respectively) compared to the tailings (VDC-1 to VDC-4; 18.7, 25.8, 62.2, and 35.1 ppm, respectively) however, higher phosphate concentrations did tend to occur in denser vegetation (VDC-3 to VDC-6) ( Table S2 ). Soil phosphate concentrations in VDC-6 might be influenced by the rapid uptake of phosphate by mature trees, as the nutrient becomes available. Our observations are therefore not contradictory; the likely higher demand for phosphorus in the boreal forest vegetation could explain the higher abundance of Pseudomonas sp. in these soil (VDC-5 and 6), compared to mine tailings (VDC-1 to 4). The nitrogen demand of alders is higher than that of other plant species. Despite other tree species, the absorption of nutrient by alder (N and P) is approximately in the same proportion, which could explain the high nitrogen demands by this species to support growth. To meet this high nitrogen demand, plant recruitment of beneficial microorganisms in the rhizosphere soil is primary ( Lõhmus et al., 2006 ). Bradyrhizobium sp. were known for their symbiotic association with legume plant, and to promote nodulation in these plants. In addition, like other rhizobia, Bradyrhizobium sp. also have the capacity to fix atmospheric nitrogen and to make this nutrient available for other organisms ( Saharan & Nehra, 2011 ). There is little literature on recruitment of Bradyrhizobium sp. by alder, but it is possible that this microorganism could be favorably recruited by alder (in comparison to birch and spruce) to improve its nitrogen acquisition. Furthermore, few authors have reported the possible role of Haliangium sp. as PGPR organisms. However, Kundim et al. (2003) as well as Ma et al. (2018) do mention that Haliangium sp. secrete antifungal molecules that limit the development of phytopathogens. The highest relative abundance of Bradyrhizobium sp. and Haliangium sp. in alder rhizosphere soil could also be explained by the different bacterial selection between different plant species ( Lei et al., 2018 ; Urbanovà, Šnajdr & Baldrian, 2015 ). In essence, our results indicate that bacterial richness was significantly lower in VDC-1 compared to all other VDC, including those of the natural environment (VDC-5 and VDC-6). Considering the increase of vegetation density as an indicator of environmental recovery, these results strongly suggest that bacterial diversity promptly recovers to natural levels following the establishment of vegetation. Fungi associated with vegetation density classes and plant species The ectomycorrhizal (ECM) fungi, arbuscular fungi (AM) and dark septate endophytes (DSE) are reported to enhance the phytoremediation potential of plants through elemental cycling and fungal-metal interactions ( Deng & Cao, 2017 ; Pirttilä & Frank, 2011 ). ECM fungi include an estimated 6000 species (mostly Basidiomycetes), and establish symbiosis with only 5% of terrestrial plants, in few woody plant families and genera (i.e., Pinaceae , Fagaceae , Betula sp., Populus sp., and Alnus sp.), AM fungi comprise approximately 150 species of Zygomycetes ( Deng & Cao, 2017 ). They are associated with herbaceous plants, and various woody plant families ( Deng & Cao, 2017 ; Fortin, Plenchette & Piché, 2015 ; Landeweert et al., 2001 ; Mandyam & Jumpponen, 2005 ). The DSE fungi are reported to establish symbiosis with over 600 plant species, including plants that were not reported to be mycorrhizal. They are present in the rhizosphere soil and roots of plants colonizing metal-contaminated sites, and they are characterised as conidial and sterile fungal endophytes, which form melanised inter- and intra-hyphal structures ( Mandyam & Jumpponen, 2005 ; Upson et al., 2009 ). Many fungal taxa are reported to establish symbiosis with plants colonizing metal-laden sites. For example, strains of genera Trichoderma sp., Fusarium sp., Aspergillus sp., and Cladosporium sp., colonized Portulaca plant, a heavy metal hyperaccumulator ( Deng et al., 2014 ). Penicillum spp. and Trichoderma spp. are also the most frequently isolated fungi that attenuate heavy metal stress in plants ( Babu et al., 2014a ; Babu et al., 2014b ; Deng & Cao, 2017 ; Khan et al., 2014 ). ECM fungi can play a role in metal housekeeping through mechanisms such as precipitation, chelation, cell-wall binding, and the binding of metals by organic acids, polyphosphates, peptides and their transport through intracellular compartments ( Colpaert et al., 2011 ). Moreover, glomalin (a glycoprotein) produced by some AM fungi increase heavy metal binding, reducing the uptake of heavy metals by the host plants ( Bano & Ashfaq, 2013 ). In addition, ECM and AM fungi are involved in contaminant detoxification and mediate the nutritional status of heavy metals in plants ( Barea, Azcon & Azcon-Aguilar, 2002 ; Harms, Schlosser & Wick, 2011 ; Luo et al., 2014 ; Thijs et al., 2016 ). DSE are known to enhance mineral uptake of host plants, increase the utilization of various organic pools and modification of host water uptake. Indeed, the high melanisation of DSE allows them to resist severe drought and heat and increase plant growth under such conditions ( Kilvin, Emery & Rudgers, 2013 ; Mandyam & Jumpponen, 2005 ; Pirttilä & Frank, 2011 ; Zhang et al., 2008 ). Liu et al. (2015a) reported a shift in fungal communities as a function of organic carbon content in soils: the abundance of Agaricomycetes decreased when more organic carbon was present, and when there was an increase of Incertae sedis. Our results corroborate this; the relative abundance of Agaricomycetes in mine tailings (VDC-1 to 4) (average ∼20%) is higher than its relative abundance in boreal forest soil (VDC-5 and 6) (average ∼15%). We also observed, as Liu et al. (2015a) , a higher relative abundance of Incertae sedis in boreal forest soils (VDC-5 and 6) (∼16%) compared to that in mine tailings (VDC-1 to 4) (∼5%) ( Fig. 4 and Table S7B ) . The higher organic matter content in boreal forest could also explain our PCoA results that showed a difference in fungal communities in VDC-5 and 6 in comparison to those in mine tailings soils (VDC-1 to 4) ( Fig. 3B and Table 1B ). In our study, ECM Hypocrea sp. were associated more with plant rhizospheres from VDC-2 to 4, and could potentially be beneficial to plants on such sites. As an example, Morales-Barrera & Cristiani-Urbina (2015) found that Hypocrea tawa could be useful to detoxify Cr(VI)-contaminated wastewaters because of its capacity to reduce hexavalent chromium (Cr(VI)) to a much less toxic trivalent chromium form (Cr(III)). Hypocrea can solubilize metals (i.e., Cr, Ni, Cu and Pb) and, with other microorganisms, could immobilize these same heavy metals following their solubilisation ( Congeevaram et al., 2007 ; Kumari et al., 2015 ). Based on the previous studies, Hypocrea sp. could have alleviated heavy metal stress, helping host plants grow on the tailings storage site. Our study revealed a higher relative abundance of Mollisia sp. in VDC-1 and 2. This could be explained by the fact that Mollisia sp. is known for its high metal tolerance ( Pirttilä & Frank, 2011 ), and that the low organic matter content in these VDC likely provided conditions allowing high bioavailability of metals ( Table S2 ). Our study also found a greater relative abundance of Zygomycota fungi ( Dissophora sp. and Mortierella sp.) in boreal forest soil (VDC-5 and 6) compared to tailings (VDC-1 to 4). These observations corroborate those of Liu et al. (2015a) and Sun et al. (2016) according to which the relative abundance of Zygomycota (and specifically Mortierella sp.) increased with soil carbon content. The organic matter content of VDC-5 and 6 were on average 22.4%, while they were 2.1% on the tailings storage area VDC ( Fig. 7 and Table S2 ). The capability of plants species to shape their rhizosphere soil microbiome to adapt to various environmental conditions can lead to significant differences in the relative abundance of diverse fungi and bacteria ( Berendsen, Pieterse & Bakker, 2012 ; Sasse, Martinoia & Northen, 2018 ). Prescott & Grayston (2013) and Urbanovà, Šnajdr & Baldrian (2015) reported that fungal communities in soil were largely explained by dominant trees. Urbanovà, Šnajdr & Baldrian (2015) reported that different proportions of arbuscular and ectomycorrhizal fungi were found under different tree species. In our study, we showed that birch rhizosphere present a higher relative abundance of the DSE fungus Mollisia sp. compared to the rhizosphere soil of alder and spruce ( Fig. 9A ). This finding is similar to that reported by Fernández-Miranda Cagigal (2017) , where DSE were the dominant fungi of healthy fine root of, amongst other species, Betula pubescens . Globally, we observed that fungal richness was not as dependent upon the presence or density of vegetation as was bacterial diversity. Fungal diversity did not differ significantly between VDC-1 to VDC-4, compared to that of boreal forest soils (VDC-5 and VDC-6). While bacterial richness in diversity increased following vegetation establishment, that of fungal communities remained relatively stable. These observations corroborate those of Sun et al. (2017) which found that bacterial diversity was 7.0- to 7.5-fold greater compared to fungal diversity in a recovering forest ecosystem."
} | 7,205 |
38257937 | PMC10821162 | pmc | 3,532 | {
"abstract": "The use of microalgae as a raw material for biogas production is promising. Macroalgae were mixed with cattle manure, wheat straw, and an inoculant from sewage sludge. Mixing macroalgae with co-substrates increased biogas and methane yield. The research was carried out using a three-stage bioreactor. During biogas production, the dynamics of the composition of the microbiota in the anaerobic chamber of the bioreactor was evaluated. The microbiota composition at different organic load rates (OLRs) of the bioreactor was evaluated. This study also demonstrated that in a three-stage bioreactor, a higher yield of methane in biogas was obtained compared to a single-stage bioreactor. It was found that the most active functional pathway of methane biosynthesis is PWY-6969, which proceeds via the TCA cycle V (2-oxoglutarate synthase). Microbiota composition and methane yield depended on added volatile solids (VS added ). During the research, it was found that after reducing the ORL from 2.44 to 1.09 kg VS/d, the methane yield increased from 175.2 L CH 4 /kg VS added to 323.5 L CH 4 /kg VS added .",
"conclusion": "4. Conclusions This study demonstrated that in a three-stage bioreactor, both a higher yield and concentration of methane in biogas are obtained compared to a one-stage bioreactor. The research shows that the production of biogas increases and correlates with the number of methanogens in the anaerobic reactor. Studies have shown that microbial composition changes in conjunction with changes in the OLR. It was identified that the most active functional pathway of methane biosynthesis is PWY-6969, which proceeds via the TCA cycle V (2-oxoglutarate synthase). It was found that after reducing the OLR from 2.44 to 1.09 kg VS/d, the methane yield increased from 175 L CH 4 /kg Vs added to 324 L CH 4 /kg VS added . This change was due to the increased performance of the RT and Euryarchaeota family. By digesting macroalgae mixed with co-substrates in a three-stage bioreactor, methane-producing microorganisms can work effectively and achieve high methane yields.",
"introduction": "1. Introduction The production of biogas using agricultural and eutrophication products such as the algae Saccorhiza polyschides allows the simultaneous digestion of organic waste and the generation of a multipurpose energy carrier (methane) that is further converted into electricity and heat [ 1 , 2 , 3 ]. Biogas is produced during the anaerobic digestion of organic matter. Anaerobic digestion consists of four stages: hydrolysis, acidification, acetate production, and methanogenesis. All steps are performed by highly specialized and complex microbial communities, and the roles of each member in a consortium are different [ 4 ]. These complex interspecies relationships hamper the investigation of microbial communities using traditional microbiological methods [ 5 ]. To solve this problem, it is necessary to go beyond the simple identification of microbial species in the methanogenic microbiota to reveal their functional roles during biogas production. The use of macroalgal biomass for biogas production also has its challenges. Salinity can be a serious barrier to microbiota activity. Also, seasonal dispersion may affect the growth and quality of macroalgal biomass. The use of macroalgae found in freshwater bodies or low-salinity seas, such as the Sea of Azov, for biogas production could reduce this risk. It is also necessary to ensure proper biomass collection. Anaerobic digestion of a substrate for biogas production is affected by many factors. One key factor is organic load rate (OLR). OLR depends on the physicochemical parameters of the substrate, such as the C:N ratio, total solids (TS), volatile solids (VS), and hydraulic retention time (HRT) [ 6 ]. Therefore, the optimal OLR allows for controlling biogas production [ 7 ]. There are many studies showing that macroalgae can be used for biogas production in biopower plants due to their nutritional properties. Due to the high carbon-nitrogen ratio (C:N), they should be mixed with materials with a low C:N ratio, such as cattle manure, to achieve the optimal C:N ratio. The optimal C:N for biogas production should be 20:1 to 30:1 [ 8 , 9 ]. Biogas produced using macroalgae with co-substrates can be a good alternative to fossil fuels. Macroalgae contain high levels of cellulose, which would affect the yield of biogas during its production. According to previous experiments performed, the most promising approach to pretreatment, at least for brown algae, is a hydrothermal one. Using this method, high amounts of methane were produced, and a positive energy balance was obtained [ 10 ]. Another type of experiment revealed a high yield of methane in biogas produced using macroalgae, Spirogyra varians . The untreated methane yield was 64.7%, and the pretreatment even increased the methane yield percentage to 69.4%. Due to their rapid biomass production and their suitable composition, these macroalgae are widely used for biogas production [ 11 ]. Organic matter is mineralized during anaerobic digestion. They are mineralized in the absence of inorganic oxidants such as nitrate, sulfate, iron, etc. Over 67% of acetaclastic and up to 33% of hydrogenotrophic methanogens produce methane when cellulose degrades [ 12 ]. Methanogenic archaea reduce to methane formate, CO 2 , acetate, methylamines, methanol, methyl sulfides, and C1 and C2 compounds. Methanogenic archaea obtain energy during the reduction process. Methanogens also use ingredients such as secondary alcohols and methoxylated aromatic substances (in methoxydotrophic methanogenesis). The methanogenic lineages of Euryarchaea are broadly characterized as hydrogenotrophic (using H 2 and CO 2 ), acetalastic (acetate), methylotrophic (X-CH 3 ), and methylotrophic (using H 2 and X-CH 3 ) lineages. The orders Methanococcales, Methanobacteriales, Methanocellales, Methanomicrobiales, and Methanopyrales consist of strict hydrogenotrophic methanogens [ 13 ]. Numerous microbial communities regulate the stages of anaerobic digestion: hydrolysis, acidogenesis, acetogenesis, and matenogenesis. They work in a symbiotic relationship. These communities change as parameters change. They depend on the substrates used for anaerobic digestion. For biogas production using macroalgae and co-substrates, single- or two-stage bioreactors are usually used [ 14 , 15 ]. A two-stage bioreactor can produce about 30% more energy than a single-stage bioreactor [ 16 ]. High methane yields can be achieved using a three-stage bioreactor. To obtain good methane production, the anaerobic conditions should be stable, and oxygen levels should be as low as possible. Thus, the intermediate chambers of three-stage bioreactors increase the dissolved oxygen reduction efficiency of the substrate by 80.5% [ 17 ]. In the first chamber of the bioreactor, the primary crushing and mixing of biomass are carried out. The prepared substrate then enters the second chamber along with excess oxygen. In the second chamber, the biomass is preheated, and one-third of the biomass always remains, ensuring that oxygen does not enter the third chamber and that anaerobic conditions are maintained in it. The third chamber is used for the digestion process and biogas production. The aim of this study is to determine and evaluate the composition of the microbiota in a three-stage bioreactor when a substrate, consisting of a mixture of macroalgae, cattle manure with wheat straw, and sewage sludge, is anaerobically treated at different OLRs. Currently, there is insufficient information on changes in microbiota in a three-stage bioreactor when a substrate, consisting of a mixture of macroalgae, cattle manure, and sewage sludge, is anaerobically treated at different organic loadings. The present research would allow a better understanding of the microbiological processes taking place in a three-stage bioreactor, where macroalgae participate in anaerobic digestion.",
"discussion": "3. Results and Discussion 3.1. Changes in Substrate Composition during Biogas Production The concentrations of proteins, lipids, and polysaccharides play an important role during the methanogenic process. We determined their concentrations in the substrate before incubation in the bioreactor and after stabilization of the process at 5 days of retention (when the OLR was 2.44 kg VS/d) and at 10 days of retention (when the OLR was 1.09 kg VS/d) ( Table 3 ). Before methanogenesis, the total lipid content of the substrate was recorded at 1.49%. Subsequently, during the retention period of 5 days, when the OLR was 2.44 kg VS/d, there was a notable reduction of 1.3% in the total lipid content, which further decreased by 0.4% after 10 days, when the OLR was 1.09 kg VS/d. It is noteworthy that the initial substrate exhibited the highest total lipid content. This observation raises the possibility that the initial homogenization process in the first bioreactor stage may have been less effective. A possible reason may be the substrate fraction size, which ranged from 0 to 1 mm, hindering the even distribution of lipids. Therefore, it is imperative to improve the homogenization process, as our data indicates that lipid content varies among the samples by up to 1.3%. The homogeneity of the substrate had no significant effect on glucose or total protein determination. The microorganisms can use the nutrients as shown in Table 3 ; the concentration of glucose was about 9.68% before anaerobic digestion, 6.68% after 5 days of retention, and about 7.19% after 10 days of retention. The difference between the last two values is not statistically significant. At the same time, the concentration of total protein decreased during retention, being about 9.5% before anaerobic digestion and then 1.7% after biogas synthesis. This demonstrates that proteins are digested most efficiently during methanogenesis. As shown previously [ 31 ], a concentration of long-chain fatty acids in bioreactor filler that is too high can inhibit the synthesis of methane. It is also possible that glucose is being released slowly from cellulose, which is present in the CM and MA. 3.2. Structure of the Microbial Community during Methanogenesis A microbial community of anaerobic sludge from a city sewage (CS) treatment plant was used for biogas production. The composition of the microbiota was determined beforehand and then again at 5 and 10 days into the anaerobic digestion process, as demonstrated in Figure 2 . Sequencing results demonstrated decreased richness and diversity of the microbial community, as well as a decreased relative abundance of bacteria in relation to archaea, similar to what has been observed previously [ 32 ]. At the same time, the presence of eukaryotic microorganisms and viruses was not observed, but another study found eukaryotes and viruses contributing 0.52% and 0.02% of abundance [ 33 ]. In sewage sludge samples (1U and 1L), the abundance of the phylum Euryarchaeota increased from 8.95% to 22.55% after the fourth retention phase. Actinobacteria, initially present in the starter mix at 10.25%, decreased to 3.4% (OLR was 2.44 kg VS/d), but rebounded to 9.6% in the final sample when OLR was 1.09 kg VS/d. The variation in Bacteroidetes was approximately 4% higher compared to samples taken before methanogenesis. The percentage of the bacteria Candidatus cloacimonetes varied in the samples from 4.85% initially to 6.65% and 2.80% after the fermentation process. Percentages of another group of bacteria, Phylum Firmicutes , decreased from 27.35% to about 19.23% and were stable at different OLR phases (2.44 and 1.09 kg VS/d) under anaerobic conditions. The bacteria Phylum Synergistetes were present at 6.35% prior to anaerobic digestion and decreased to 0.15% by the end of the process. At the order level, the abundance of Euryarchaeota OFGB 9269 increased to 6.8%, and that of the Methanomicrobiales increased to 14.4%. The abundance of Methanocorpusculaceae increased to 14.2%, while that of Methanomicrobiaceae decreased from 9.8% to 0.3%. Another study reported at the order level comparing a one-stage bioreactor with substrate was sludge from a local wastewater treatment plant in Beijing, China [ 33 ], wherein the abundance of Methanobacteriales was 1,66%, the phylum of Euryarchaeota was 2.02%, and the genus level was total. Archeal sequencing showed that when the solid anaerobic digestion batch [ 34 ] increased, Methanoculleus increased from 4.60% to 83.0%, but in our research, this methanogens decreased from 8.94% to 0%. In our research, Methanocorpusculum (taxonomy genus level) was increased from 0 to 13.65%, as noted for the species Methanocorpusculum bavaricum. Compared to different types of bioreactors with different types of sludge, microbiota was formed, which can produce biogas, but the main genus was not the same. 3.3. α and β Diversity The evaluation of α-diversity revealed a significant difference prior to anaerobic digestion at OLRs of 2.44 and 1.09 kg VS/d. Figure 3 shows the α and β diversity of microorganism species before and after different OLRs. The highest Shannon diversity index was observed to be 3.2 at the substrate before, and it increased by 9.6% over time and through the phases of methanogenesis when ORL was 2.44 and 1.09 kg VS/d. Changes in the bacterial community in the three-stage bioreactor resulted in a potential rate-limiting step during anaerobic digestion. Principal Coordinate Analysis (PCoA) performed with the representative OTUs showed microbiota separation before biogas synthesis and at the different OLRs (2.44 kg VS/d and 1.09 kg VS/d) and indicated the similar composition of the bacterial microbiota ( Figure 4 ). Figure 4 shows that the microbiota in samples 1U and 1L before biogas production are different compared to samples taken during biogas synthesis in 2U and 2L and 3U and 3L when ORL was 2.44 kg VS/d and 1.09 kg VS/d, respectively. However, in samples 2U and 2L and 3U and 3L when the ORL was 2.44 kg VS/d and 1.09 kg VS/d, respectively the microbiota was very similar and produced methane. 3.4. Methane Production at the Three-Stage Bioreactor Methanogenesis is the biological production of methane, a process of anaerobic digestion performed by a group of methanogens belonging to the Archaea domain of single-celled organisms. Because methanogenesis is the final step in the anaerobic degradation of organic carbon, the functional pathways performed by methanogens that convert acetate to CO 2 and CH 4 and oxidize H 2 to H 2 O were sorted. The correlation between methane production and the abundance of Euryarchaeota is shown in Figure 5 . Cell metabolism depends on the biochemical reactions performed by various enzymes. The expression and activity of enzymes are closely related to the rate of biochemical reactions. It was found that a 77.8% concentration of methane, when OLR was 2.44 kg VS/m 3 , correlated with a 22.65% abundance of Euryarchaeota, which was 13.65% greater than that in the starter mix. Reducing OLR to 1.09 kg VS/m 3 decreased methane concentration to 73.8% without any significant decrease in the abundance of Euryarchaeota. During the adaptation period, the yield of biogas and methane was 450.9 L/Kg VS added and 329.6 L CH 4 /Kg VS added , respectively. Sufficiently high yields of biogas and methane were achieved due to the long HRT, which was 20 days. CH 4 /m 3 -d was produced between 1100 and 1374 L per day in a three-stage bioreactor at different OLRs. Although the OLR differed 2.2 times between the samples, the amount of Euryarchaeota remained similar. It was determined that when the OLR was 2.44 kg VS/d, the maximum biogas and methane yield was 223 L/kg VS added and 175 L CH 4 /kg VS added , respectively. After the OLR was reduced to 1.09 kg VS/d, the maximum methane yield increased to 324 L CH 4 /kg VS added and the biogas yield increased to 429 L/kg VS added . Although the concentrations of Archea Euryarchaeota at different OLRs were similar, the methane yield from 1 kg VS added was higher when the OLR was 1.09 kg VS/d. This was caused by a higher VS retention time in the bioreactor, which reached 10 days. It has been previously demonstrated [ 35 ] that 408 mL of biogas can be obtained by mixing macroalgae cultures of Ulva rigida with anaerobic sludge and water in a single-stage bioreactor under mesophilic conditions. During those tests, a biogas yield of 114 L/kg VS added was achieved in a single-stage bioreactor, and the methane concentration reached 75%. A maximum biogas yield of 1200 L/kg VS added was obtained when Ulva rigida was mixed with the sugar waste. In that study, the OLR reached 1.66 g VS/L-d. By mixing macroalgae with co-substrates, higher yields of biogas and methane can be obtained. A previous study [ 3 ] has shown that a maximum methane yield of 146 L/kg VS is achieved using the marine macroalgae Saccorhiza polyschides . The maximum concentration of methane in the biogas was 64.5% when the experiments were carried out in a single-stage bioreactor under mesophilic conditions. In order to demonstrate the two-stage bioreactor effect [ 16 ], methane yield studies in a bioreactor using substrate from Napier grass ( Pennisetum purpureum ) were performed. During the research, it was determined that a methane yield of 282 L CH 4 /kg VS was produced, which was 30% higher than that produced by a single-stage bioreactor. Research shows that mixing macroalgae with co-substrates such as cattle manure with straw can produce higher methane yields of up to 324 L CH 4 /kg VS added in a three-stage bioreactor. CO 2 , O 2 , and H 2 S concentrations in biogas were determined during the adaptation period and at different OLRs. CO 2 concentrations were in the range of 20 to 30% during the research. O 2 and H 2 S concentrations were 0–2% and 0–10 ppm, respectively. 3.5. Functional Pathways of Methane Production at the Three-Stage Bioreactor Methanogenesis is the biological production of methane, a process of anaerobic respiration performed by a group of methanogens belonging to the Archaea domain of single-celled organisms. Because methanogenesis is the final step in the anaerobic degradation of organic carbon, the functional pathways performed by methanogens that convert acetate to CO 2 and CH 4 and oxidize H 2 to H 2 O were sorted ( Figure 6 ). 1200 functional pathways in total were identified; however, only 16 are related to methane biosynthesis. The relationship between the results and the data obtained during the investigation is shown in Figure 6 . Methanogens (anaerobic archaea) do not have the ability to perform a complete TCA cycle, and it is proposed that biosynthesis intermediates are synthesized through an incomplete cycle. TCA plays an important role in producing electron carriers, such as NADH and FADH2, for energy production. One of the methanogenic subsystems has a reductive tricarboxylic acid cycle (RTCA). However, most anaerobic species use the RTCA cycle, which reduces CO 2 and H 2 O in order to synthesize carbon compounds. Methanogenic anaerobes also have RTCA cycles. In addition, their TCA cycles are incomplete due to a lack of several steps and enzymes [ 36 ]. The abundance of pathways in different metabolic subsystems is shown in Figure 6 . The dominant pathway in the city sewage sludge was the incomplete reductive TCA cycle P42-PWY. The current Macrogen database describes the P42-PWY cycle, which has seven functions provided by six enzymes ( Table 4 ). In the order Methanomicrobiales, three families were identified: Methanomicrobiaceae, Methanocorpusculaceae, and Methanospirillaceae. All members of this order are able to produce methane from CO 2 and H 2 . In many species, formate and secondary alcohols are used as alternative electron donors [ 37 , 38 ]. In this study, we identified that the most active functional pathway of methane biosynthesis is PWY-6969, which proceeds via the TCA cycle V (2-oxoglutarate synthase) ( Table 4 ). Methanocorpusculum bavaricum was found to be present at 14.2% (OLR was 2.44 kg VS/d) and at 14.15% (OLR was 1.09 kg VS/d) when methane concentration and yield were highest. However, this microorganism was not found in the inoculant (city sewage). While comparing the functional pathways between the starter culture and phases of the bioreactor (OLRs were 2.44 and 1.09 kg VS/d), it was demonstrated that methane synthesis pathways differ ( Figure 6 ). Examples of these differing functional pathways are COA-PWY-1, the super-pathway of coenzyme A biosynthesis III (mammals); PWY-5209, methyl-coenzyme M oxidation to CO 2 ; PWY-5659, GDP-mannose biosynthesis; and PWY-7851, coenzyme A biosynthesis II (eukaryotic). These pathways could be found in bacteria or in eukaryotic organisms. In the starter culture, archaea made up 9% of the total microbiota, and three species were identified. By comparison, during anaerobic digestion, when the yield of methane was highest, archaea made up 22.65% of the total microbiota, and five species were identified. At this stage, the number of functional pathways decreased by twofold compared to the starter cultures, though these pathways were more directed toward methane biosynthesis."
} | 5,310 |
33935995 | PMC8079732 | pmc | 3,533 | {
"abstract": "Quorum sensing (QS) is a signaling mechanism governed by bacteria used to converse at inter- and intra-species levels through small self-produced chemicals called N-acylhomoserine lactones (AHLs). Through QS, bacteria regulate and organize the virulence factors’ production, including biofilm formation. AHLs can be degraded by an action called quorum quenching (QQ) and hence QQ strategy can effectively be employed to combat biofilm-associated bacterial pathogenesis. The present study aimed to identify novel bacterial species with QQ potential. Screening of Palk Bay marine sediment bacteria for QQ activity ended up with the identification of marine bacterial isolate 28 (MSB-28), which exhibited a profound QQ activity against QS biomarker strain Chromobacterium violaceum ATCC 12472. The isolate MSB-28 was identified as Psychrobacter sp. through 16S-rRNA sequencing. Psychrobacter sp. also demonstrated a pronounced activity in controlling the biofilm formation in different bacteria and biofilm-associated virulence factors’ production in P. aeruginosa PAO1. Solvent extraction, heat inactivation, and proteinase K treatment assays clearly evidence the enzymatic nature of the bioactive lead. Furthermore, AHL’s lactone ring cleavage was confirmed with experiments including ring closure assay and chromatographic analysis, and thus the AHL-lactonase enzyme production in Psychrobacter sp. To conclude, this is the first report stating the AHL-lactonase mediated QQ activity from marine sediment bacteria Psychrobacter sp. Future work deals with the characterization, purification, and mass cultivation of the purified protein and should pave the way to assessing the feasibility of the identified protein in controlling QS and biofilm-mediated multidrug resistant bacterial infections in mono or multi-species conditions.",
"conclusion": "Conclusion Quorum quenching is of great concern in controlling infectious pathogens without interfering with growth, thus avoiding the selection pressure that often results in the emergence of resistance strains. In this study, we found marine sediment bacteria with QQ potential, identified as Psychrobacter sp., that is able to degrade AHLs and thereby inhibit the QS mechanism and biofilm formation of diverse bacterial pathogens. Moreover, lactonolysis and chromatograpic analysis revealed the presence of AHL-lactonase in the CFS of Psychrobacter sp. Thus, the attained results emphasize that the QQ activity of Psychrobacter sp. could potentially be used as a biocontrol agent to combat multidrug resistant bacterial infections caused by Gram-negative human as well aquatic pathogenic bacteria.",
"introduction": "Introduction Biofilms are a complex aggregation of mono or mixed species of microbial populations embedded on the biotic or abiotic surfaces by a self-produced extracellular polymeric matrix ( Sharma et al., 2019 ). Biofilm forming microorganisms are responsible for a cluster of common hospital-acquired ailments including lung infection in patients with cystic fibrosis (CF), otitis media, periodontitis, burn wound infections caused by a variety of surgical implants, endocarditis, and urinary tract infections. National Institutes of Health (NIH) recommended that approximately 60% of human infections are the consequence of biofilm formation on human mucosa. Initial attachment and subsequent maturation of biofilm are the two important steps in host tissue colonization and subsequent persistent infections ( Costerton et al., 1999 ). Bacteria living inside biofilms are habitually able to tolerate host immune responses and are distinctly highly resistant to different antibiotics ( Costerton et al., 1999 ). Comparatively, this level will frequently surpass the maximum dosage level, and hence limit the efficient treatment available to control bacterial infections. The underlying mechanism of resistance is multi-factorial which includes restricted penetration, heterogeneous metabolic activity, and expression of certain genes conferring enhanced resistance to antibiotics. Hence, identification of such compounds with potential antibiofilm activity is imperative to combat the pathogenesis of these detrimental pathogens. In most of the bacterial pathogens, quorum sensing (QS) mechanism regulates the biofilm formation and other virulence factors’ production, in order to establish pathogenesis in the host. This QS mechanism is also called the cell-to-cell communication system, as the bacteria communicate with each other at inter- and intra-species levels using small diffusible signal molecules called autoinducers (AIs). In Gram negative bacteria, N-Acyl Homoserine Lactone (AHL) is the prime AI responsible for QS ( Papenfort and Bassler, 2016 ), which bind to their cognate receptor proteins that together activate the expression of QS-controlled genes ( Whitehead et al., 2001 ). In lux I/R QS system, the LuxI family protein synthesizes AHL. The LuxR family protein binds with AHL and regulates the expression of many genes responsible for their coordinated behavior including motility, antibiotic biosynthesis, virulence factor production, and biofilm formation ( Davies et al., 1998 ). Most importantly, the QS mechanism governs the biofilm formation in most of the bacterial pathogens; the QS inhibitory process termed as quorum quenching (QQ) has offered a novel target to control the biofilm-associated infections ( Dong and Zhang, 2005 ; Costerton et al., 2007 ). Contrasting to antibiotics, QS inhibitors will not set bacteria under strong selective pressure to develop drug-resistance ( Zhao et al., 2020 ). Besides QS, flagellar motility and exopolysaccharide (EPS) have also been found to be essential for bacterial aggregation and biofilm formation ( Pratt and Kolter, 1998 ; Jaisi et al., 2007 ). Secondary metabolites from marine organisms are considered an important source of biomolecules for drug discovery ( Newman and Cragg, 2004 ; Borges and Simões, 2019 ). Bacteria associated with corals, sponges, and other organisms have been recognized as the factual producers of many bioactive compounds ( Kelman et al., 2006 ; Freckelton et al., 2018 ). Though AHL degradation enzymes from bacterial isolates have been identified from different sources ( Dong et al., 2002 ; Uroz et al., 2003 ; Ulrich, 2004 ; Hassan et al., 2016 ), the investigation of antibiofilm activity of bacteria, particularly from marine resources, are expected to act against antibiotic resistant bacterial pathogens ( Huang et al., 2019 ; Zhou et al., 2019 ). Recently, a marine isolate showing a promising antibiofilm activity against Pseudomonas aeruginosa has been reported from red sea sediment ( Rehman and Leiknes, 2018 ). Also, the literature evidenced these marine bacteria as one of the sources of secondary metabolites and other extracellular hydrolytic enzymes ( Romano et al., 2017 ; Borges and Simões, 2019 ). Hence it is believed that bacteria from marine sediments (MSB) may also have the ability to produce several secondary metabolites that target bacterial QS mechanism. In light of this view, the present study aimed to isolate marine sediment bacteria that target the bacterial QS mechanism, and to divulge the mechanism of QS inhibition.",
"discussion": "Discussion The emergence of antimicrobial resistance is responsible for the failure of current antibiotics treatment on biofilm-based bacterial infections and has emphasized the urgent need for developing new alternate strategies. Bacterial QS mechanism seems to be an attractive target to develop an alternative therapeutic approach, as inhibition of QS hinders the virulence production, biofilm formation, and subsequent infection by many bacterial pathogens. Regardless, the marine bacterial organisms are being extensively investigated for their antimicrobial potentials; studies on their QQ potential are meager ( Musthafa et al., 2011 ; Borges and Simões, 2019 ). Therefore, in this study, bacteria were isolated from Palk Bay sediment samples and screened against QS biomarker strain C. violaceum . Except the CFS of marine isolate MSB-28, none of the isolates showed pronounced QQ activity against C. violaceum 12472. Similar reports were made by Shepherd and Lindow (2009) in which QQ activity was observed in CFS of P. syringae B728a which enabled the bacteria to degrade QS signals and thus block the expression of QS-regulated traits. The isolate showing profound QQ activity was found to be Psychrobacter sp. and belongs to Proteobacteria. These results corroborate well with earlier studies, where the bacteria isolated from a marine environment with QQ potential were identified as Proteobacteria ( Tan et al., 2015 ; Torres et al., 2016 ; Rehman and Leiknes, 2018 ). QS governs the biofilm formation and maturation in several bacterial species. The marine bacterial isolate Psychrobacter sp. exhibited biofilm inhibition against various Gram-negative pathogens such as PAO1, S. marcescens , V. vulnificus , and V. parahaemolyticus at 20% v/v without growth retardation ( Supplementary Table S1 ). The degree of variation in the QQ activity of Psychrobacter sp. could be due to the involvement of AHLs produced by the target bacteria with varied lengths of acyl side-chain. Flagella and pili aid to initiate the biofilm formation by reversible and irreversible attachment followed by microcolony formation ( Shi and Sun, 2002 ). Hence, any interference in their expression by the metabolites of Psychrobacter sp. would result in the failure of biofilm formation. Development of distinctive biofilm architecture is the most important stage in biofilm formation ( Kannappan et al., 2017 ). The attained results of CLSM analysis suggest that biofilm formation of the target pathogens was inhibited at the early stages of biofilm development. Moreover, the result of COMSTAT analysis ascertained the biofilm quantification assay. Altogether, these results suggest that the QQ agent present in Psychrobacter sp. might possibly interrupt the biofilm development without any negative effect on the bacterial growth. Bacteria are known to secrete antibacterial compounds. The growth analysis of pathogens clearly portrayed that the CFS of Psychrobacter sp. had no antibacterial activity towards the target pathogens. QS inhibition without any growth reduction is considered as the best alternative strategy to control the virulence factors’ production and pathogenesis of bacterial pathogens, and leaves no scope for the development of antibiotics resistance ( Rasmussen and Givskov, 2006 ). In this light, CFS of Psychrobacter sp. showed a profound QQ activity without any growth inhibition against the representative Gram-negative pathogens, which holds great clinical significance. We also examined the QQ ability of Psychrobacter sp. to control other virulence factors associated with biofilms produced by PAO1. Production of EPS is known to maintain the biofilm architecture, and also correlates with an increased resistance of the biofilm-residing cells to biocides and host immune response ( Kannappan et al., 2017 ). Hence, inhibiting EPS secretion by marine bacterium Psychrobacter sp. would loosen the biofilm architecture; thus, it is possible to reintroduce the use of antibiotics in treating biofilm cells along with active leads produced by Psychrobacter sp. In P. aeruginosa , several QS regulated phenotypic behaviors have been reported to be a part of the biofilm formation ( Fong et al., 2019 ). In this study, treatment of Psychrobacter sp. would result in the reduced production of rhamnolipid; an important factor enhances the swarming motility by reducing the surface tension. In contrast, deficiency in surfactant production alters the swarming migration pattern and the altered bacteria would fail to colonize over the surface. The possible mode of action of the Psychrobacter sp. to block biofilm development is interfering either with C4-HSL signaling pathway accountable for surfactant production and swarming motility or blockage of 3-oxo-C 12 HSL signals which have a direct control over biofilm formation. As signal-mediated QS regulates the virulence factors’ production and biofilm formation, a remarkable reduction in biofilm formation and associated behaviors by Psychrobacter sp. might result from an effective hindrance of signal molecules by the secondary metabolites from Psychrobacter sp. Consistent with this result, a marine isolate VG-12 from red sea sediment inhibited the biofilm formation of PAO1 via QS signal degradation ( Rehman and Leiknes, 2018 ). The CFS of Psychrobacter sp. that lost its QQ activity upon being subjected to solvent extraction ( Figure 3A ), heat ( Figure 3B ), and Proteinase k ( Figure 3C ) treatments indicate that bioactive lead produced by Psychrobacter sp. is enzymatic in nature. Moreover, it is suggested that these QQ enzymes are heat sensitive. Hence, it is speculated that the loss of AHLs signaling was either because of AHL acylase or AHL lactonase activity. It is known that QQ bacteria able to degrade small-chain AHLs can also degrade medium and long-chain AHLs ( Tan et al., 2015 ; Torres et al., 2016 ; Rehman and Leiknes, 2018 ). Hence, it is envisaged that investigation of QQ bacteria should have a focal point on identifying bacteria that targets small-chain AHLs, as recommended previously ( Rehman and Leiknes, 2018 ). Interestingly, in this study the marine isolate was able to degrade the external C6-AHLs ( Figure 5 ). AHL lactonases and AHL acylases are the best-known examples of AHL degrading so far reported and studied. Though the activity of AHL-acylases on short-chain AHLs remains unclear ( Shepherd and Lindow, 2009 ; Czajkowski et al., 2011 ), it would be factual if the degraded C6-AHLs will be restored after acidification. In the present study, the degradation and restoration of short chain C6-AHLs suggest that the observed QQ activity of Psychrobacter sp. in attenuating the QS-mediated biofilm formation by bacterial pathogens such as S. marcescens , P. aeruginosa , V. parahemolyticus , and V. vulnificus is possibly due to the presence of an AHL lactonase. Altogether, the obtained results from the present study evidence that the AHL lactonase produced by the marine bacterium Psychrobacter sp. is heat-liable and active against different AHLs produced by other pathogens. Moreover, this bacterium was not found to produce any AHL signal molecule and hence could be used as a potent source for AHL degrading lactonase enzyme. For the first time, the present report divulged the quorum quenching lactonase enzymes production from Psychrobacter sp."
} | 3,663 |
39774301 | PMC11741664 | pmc | 3,534 | {
"abstract": "The denitrifying bacterium Thauera sp . MZ1T, a common member of microbial communities in wastewater treatment facilities, can produce different compounds from a range of carbon (C) and nitrogen (N) sources under aerobic and anaerobic conditions. In these different conditions, Thauera modifies its metabolism to produce different compounds that influence the microbial community. In particular, Thauera sp . MZ1T produces different exopolysaccharides with floc-forming properties, impacting the physical disposition of wastewater consortia and the efficiency of nutrient assimilation by the microbial community. Under N-limiting conditions, Thauera sp . MZ1T decreases its growth rate and accelerates the accumulation of polyhydroxyalkanoate-related (PHA) compounds including polyhydroxybutyrate (PHB), which plays a fundamental role as C and energy storage in this β-proteobacterium. However, the metabolic mechanisms employed by Thauera sp . MZ1T to assimilate and catabolize many of the different C and N sources under aerobic and anaerobic conditions remain unknown. Systems biology approaches such as genome-scale metabolic modeling have been successfully used to unveil complex metabolic mechanisms for various microorganisms. Here, we developed a comprehensive metabolic model (M-model) for Thauera sp . MZ1T ( i Thauera861), consisting of 1,744 metabolites, 2,384 reactions, and 861 genes. We validated the model experimentally using over 70 different C and N sources under both aerobic and anaerobic conditions. i Thauera861 achieved a prediction accuracy of 95% for growth on various C and N sources and close to 85% for assimilation of aromatic compounds under denitrifying conditions. The M-model was subsequently deployed to determine the effects of substrates, oxygen presence, and the C:N ratio on the production of PHB and exopolysaccharides (EPS), showing the highest polymer yields are achieved with nucleotides and amino acids under aerobic conditions. This comprehensive M-model will help reveal the metabolic processes by which this ubiquitous species influences communities in wastewater treatment systems and natural environments.",
"introduction": "1. Introduction Thauera sp. MZ1T is a floc-forming Gram-negative bacterium frequently found in wet soil, polluted freshwater, and in wastewater treatment facilities [ 1 ]. This facultative anaerobic bacterium belongs to the family Rhodocyclaceae within the β-proteobacteria and performs versatile metabolic processes that influence the environments it grows in [ 2 ]. Other members of Rhodocyclaceae are similarly abundant in soil, sediments, and aquatic systems [ 3 ]. The Thauera genus plays a crucial role in the nitrogen (N) cycle, converting inorganic N sources (ammonium, nitrate, and nitrite) into molecular N (N 2 ) through denitrification [ 4 , 5 ]. The Thauera genus contains over 80 fully sequenced members (Refseq and Genbank) [ 6 – 9 ]. Most members of this genus can perform denitrification [ 4 – 6 ] and degrade aromatic compounds in the presence or absence of oxygen [ 10 , 11 ]. Thauera members such as T . selenatis , T . aromatica , and Thauera sp . MZ1T, can produce various polymers, e.g. polyhydroxyalkanoates (PHA), which play a key role as a carbon (C) storage molecule [ 9 , 12 ]. These bacteria can also make exopolysaccharides (EPS) that are involved in floc formation in wastewater treatment systems [ 13 – 16 ]. Thauera sp. MZ1T in particular contains multiple specialized oxygen-sensitive enzymes for the production of N 2 and denitrification intermediates such as nitrous and nitric oxide [ 1 , 12 , 15 , 17 ]. Additionally, it can perform ammonification via dissimilatory nitrate reduction to ammonium (DNRA) using nitrite reductase under anoxic and limited N conditions, switching from oxygen to nitrate as a terminal electron acceptor [ 18 , 19 ]. At elevated oxygen levels, denitrification and DNRA pathways are partially or totally repressed and the organism switches back to aerobic respiration with oxygen as the terminal electron acceptor. In Thauera sp. MZ1T’s metabolism, oxygen and N compounds (like nitrate and nitrite) serve as terminal electron acceptors under varying environmental conditions, facilitating energy generation and enabling metabolic flexibility. This adaptability allows the bacterium to thrive in fluctuating oxygen environments common in wastewater systems. Low N conditions also trigger the biosynthesis of PHA-related compounds including polyhydroxybutyrate (PHB) as an important C storage molecule. Under high C:N ratio conditions (>5:1), Thauera sp. MZ1T produces PHB to sequester C within intracellular granules [ 12 , 15 ]. This behavior is sharply elevated with acetate as the primary C source, permitting selective PHA-related compound generation from Thauera sp. MZ1T [ 1 , 12 , 15 ]. Thauera sp. MZ1T is metabolically highly versatile and capable of growing heterotrophically with multiple C sources under aerobic and anaerobic conditions. The bacterium can assimilate various carbohydrates (e.g. glucose, fructose, galactose, sucrose), organic acids (e.g. acetate, lactate, citrate, and formic acid), alcohols (methanol, ethanol, and butanol), and aromatic compounds (e.g. toluene, xylene, various phenolic compounds, and benzoate) [ 1 , 12 , 15 – 18 , 20 ]. Under hypoxic or anoxic conditions, it can simultaneously denitrify and remove these aromatic compounds through assimilation and degradation [ 1 , 12 , 15 – 17 ]. Furthermore, it can assimilate important organic N compounds like urea, various amino acids (e.g. alanine, aspartate, glutamine) and nucleotide-related compounds [ 15 , 16 ]. In wastewater systems Thauera sp MZ1T produces abundant amounts of floc-forming EPS via four main intermediates: dTDP-D-N-acetylfucosamine, dTDP-L-rhamnose, UDP-D-galactose, and UDP-N-acetylglucosamine. These floc-forming EPS can significantly impact the structure and total biomass of the wastewater community [ 13 , 14 , 21 ]. Thauera sp MZ1T also has the versatility to produce EPS from inositol, but only under particular scenarios [ 12 , 16 , 17 ]. Its capability to assimilate various C and N compounds and promote floc formation contributes to Thauera sp. MZ1T’s abundance and importance in wastewater microbial communities [ 21 , 22 ]. Bioinformatics tools have been previously employed to elucidate genomic features and identify functional genes involved in denitrification, DNRA, PHA and EPS biosynthesis, and aromatic compound degradation to infer the metabolic potential of this versatile Thauera species [ 1 , 12 , 15 , 16 ]. However, the mechanisms by which this strain regulates these capabilities in response to specific resource conditions, and thereby influences the surrounding microbial community and its environment are still poorly characterized. Genome-scale metabolic models (GEMs) represent a fundamental approach to explore microbial metabolic functions across various conditions, allowing accurate predictions of metabolic trade-offs in response to specific environmental and resource constraints [ 23 – 27 ]. By simulating condition-specific metabolic pathways, GEMs support the investigation of complex microbial processes such as denitrification, polysaccharide production, and degradation of environmental pollutants [ 23 – 27 ]. To address this, we reconstructed a GEM for Thauera sp. MZ1T ( i Thauera861) using semi-automated methods. i Thauera861 contains 1,744 metabolites, 2,384 reactions, and 861 genes. The initial draft model was manually refined to improve the quality of the phenotypic predictions. Using experimental data, the model was constrained under heterotrophic conditions. Over 70 C and N sources were evaluated under aerobic and anaerobic conditions to assess the accuracy of the model. Additionally, i Thauera861 was evaluated under oxygen- and N-limiting conditions to quantify the changes in production of PHB and EPS using multiple C substrates. The model is the first refined and validated M-model for any Thauera member and will aid in unraveling the impact these bacteria have on the surrounding environments. iThauera861 offers a valuable platform for studying Thauera sp. MZ1T metabolic pathways, with applications in optimizing PHB and EPS production critical for microbial flocculation and C storage in wastewater treatment. It enables detailed exploration of C and N utilization strategies, enhancing our understanding of the organism’s ability to degrade diverse organic compounds under varying conditions. By simulating diverse environmental scenarios, i Thauera861 provides insights for improving bioreactor efficiency and advancing theoretical studies on microbial community dynamics. This refined GEM serves as a relevant tool for both applied and fundamental research into the roles of Thauera sp. MZ1T in wastewater ecosystems.",
"discussion": "3. Discussion We have reconstructed a comprehensive GEM for an important floc-forming and denitrifying wastewater bacterium using semiautomatic strategies. The high-quality, manually refined, and validated metabolic model of Thauera sp . MZ1T unravels the metabolic capabilities of Thauera sp . MZ1T under aerobic and anaerobic conditions. Initially, the M-model was reconstructed based on three Gram-negative bacterial reference models from BiGG [ 28 ] selected according to their metabolic and physiological similarities ( Escherichia coli K-12 substr. MG1655 [ 29 ], Klebsiella pneumoniae subsp. pneumoniae MGH 78578 [ 30 ], and Pseudomonas putida KT2440 [ 31 ]). Unlike other semi-automatic reconstruction approaches for metabolic modeling, we employed a reconstruction strategy considering multiple BLASTp optimal parameters depending on the number of template models. There is no clear consensus on the optimal BLASTp criteria in the metabolic reconstruction process since the BLASTp parameter values directly depend on how similar the studied microorganisms are to the reference models. However, multiple studies have reported BLASTp cutoffs of e-value between 1e-15 and 1e-5, query length of 50 to 150 amino acids, and identity percentage of 20–40% for bacteria [ 24 – 27 ], and even for eukaryotic organisms [ 35 ]. Selecting a unique set of BLASTp parameter values can significantly impact the number of false positive (wrong gene assignments) and false negative (missing genes) calls in the GPR associations. In this study, we reduced the number of wrong gene assignments by more than 20%, directly impacting the time required for manual refinement. Additionally, having multiple BLASTp parameters reduced the number of genes from the reference models by almost 30%. While the chosen parameters provided sufficient results for the model of Thauera sp . MZ1T, further analyses will be necessary to identify the accuracy of multiple BLASTp parameters optimization for other organisms and template models. i Thauera861 contains close to 22% of the annotated proteins assigned to GPR associations. Compared to the metabolic genes of the template models, i Thauera861 only contains a higher percentage of metabolic genes compared to i JN746 (14%). However, the updated version of Pseudomonas putida KT2440 M-model, i JN1463 (27%), as well as i ML1515 (31%) and i YL1228 (25%) display greater percentages of metabolic genes. Considering the updated template model, the average percentage of metabolic genes in the reference organisms surpasses the percentage in Thauera sp . MZ1T by almost 6%. i JN1463 and i ML1515 include over 400 reactions more than i Thauera861, meanwhile i YL1228 contains 122 less reactions ( S2 Material ). The difference among metabolic genes could be related to the lower amount of orphan reactions identified in the template models (20% compared to almost 28% in i Thauera861) and the large number of reactions with less than three TMZ genes in the GPR associations (almost 50% excluding the pseudo reactions of the M-model). Further analysis must be performed to determine the metabolic function of multiple hypothetical and putative proteins from the Thauera sp . MZ1T annotation to incorporate the new findings in the GPR associations of the model. In addition, we compared the metabolic features of i Thauera861 to three extensively validated M-Models of Gram-negative bacteria, i.e. Azotobacter vinelandii DJ, i DT1278 [ 26 ]; Nitrosomonas europaea ATCC19718, i GC535 [ 24 ]; and Rhodopseudomonas palustris Bis A53, i DT1294 [ 25 ], which are involved in the degradation of aromatic compounds, play a role in nitrogen metabolism, PHB production, or are present in wastewater or soil environments. i Thauera861 contains a similar number of reactions compared to the A . vinelandii model i DT1278 (2,432), sharing the core metabolic pathways for aerobic metabolism and a similar metabolic mechanism regarding PHB production. R . palustris Bis A53’s metabolic model, which contains 2,123 metabolites and 2,721 reactions, surpasses i Thauera861 by almost 350 metabolites and 400 reactions. However, the Thauera sp . MZ1T M-model shares multiple subsystems and metabolic features with i DT1294. Both M-models encompass specific reactions for the degradation of multiple aromatic compounds in the presence and absence of oxygen, denitrification, PHB production, and the capability to consume a wide range of C and N substrates. The significant difference in model size is related to R . palustris’ capability to perform anoxygenic photosynthesis, nitrogen fixation, and to grow under four metabolic states (chemoautotrophy, chemoheterotrophy, photoautotrophy, and photoheterotrophy). i Thauera861 can grow under aerobic and anaerobic conditions, but solely heterotrophically. The largest difference in the quantity of metabolites and reactions was observed when i Thauera861 was compared against i GC535 ( N . europaea ). i GC535 contains only 1,114 metabolites (36% less than i Thauera861) and 1,149 reactions (52% less). Additionally, N . europaea is capable of growing only on a few C and N substrates under strict aerobic conditions [ 24 ]. Despite the low metabolic similarities between N . europaea and Thauera sp . MZ1T, these microorganisms can be found in wastewater environments as part of a microbial community, interacting with other bacteria to remove most of the C and N pollutants [ 65 ]. Thauera organisms play a key role in wastewater treatment contaminated with aromatic compounds, removing these toxic compounds and thus enabling growth of N . europaea and in turn guaranteeing nitrification [ 12 , 16 , 17 , 24 , 56 ]. Different studies have reported a negative impact of aromatic compounds such as benzene and toluene, on N . europaea . These aromatics inhibit ammonium oxidation in N . europaea and trigger an energy drain [ 66 , 67 ]. Our GEM encompasses well-detailed metabolic pathways for degradation of aromatic compounds, both under aerobic and anaerobic conditions. i Thauera861 comprehends the consumption of aromatics through the catechol meta-cleavage pathway in the presence of oxygen and the benzoyl degradation pathway when oxygen is not available ( S8 Material ). This M-model contains more aromatic degradation pathways and metabolic mechanisms than other well-curated models such as P . putida KT2440 ( i JN746 and i JN1463) [ 31 ], R . palustris Bis A53 ( i DT1294) [ 25 ], N . europaea ATCC19718 ( i GC535) [ 24 ], or Geobacter metallireducens GS-15 ( i AF987) [ 33 ]. i Thauera861 shares most of the aerobic aromatic degradation with i JN1463, i.e. transformation of most aromatic metabolites into catechol and further catabolization to pyruvate. i JN1463 exclusively contains the metabolic pathways for the aerobic consumption of 2,4,6-trinitrotoluene, o -xylene, p- xylene, vanillin, and vanillate. However, i Thauera861 is the first M-model that includes the metabolic pathway for the degradation of aniline in presence of oxygen. Under anaerobic conditions, i Thauera861 employs the benzoyl degradation pathway, also found in i DT1294 and i AF987. However, the specific mechanism to catabolize benzoyl-CoA into acetyl-CoA utilizing the exact same enzymes is shared between Thauera sp . MZ1T and G . metallireducens GS-15 [ 48 ]. i DT1294 utilizes a slightly different metabolic pathway still channeling metabolites into benzoyl-CoA and producing acetyl-CoA but using different enzymes and intermediates. A similar metabolic strategy has also been characterized for Azoarcus sp . CIB [ 48 ]. Among all these M-models, i Thauera861 stands as the unique metabolic model capable of degrading multiple aromatic compounds under aerobic and anaerobic conditions. Thus, the model developed in this study enables the study of aromatic compound metabolism under various oxygen concentrations, often encountered in wastewater systems. Thauera sp . MZ1T’s metabolic model was carefully employed to identify the impact of different biological parameters involved in the production of PHB and EPSs. Three parameters (C substrate, oxygen presence, and C:N ratio) were shown to have a significant effect on the yield of these polymers. For example, nucleotides as C sources significantly increased the production of PHB and EPS, while substrates lacking N in their chemical structure such as alcohols, carbohydrates, and organic acids, contributed significantly to polymer production. Our results match with experimental results reported previously that found a positive correlation of C substrates without N to the production of PHB and EPS under aerobic conditions [ 54 , 60 , 61 , 68 ]. Additionally, i Thauera861 predicted the C:N ratio to have a greater impact on the PHB production independently of the C source employed in presence of oxygen than on the other polymers produced. This has been observed previously under different experimental conditions, where the C:N ratio impacts PHB production by Thauera sp MZ1T independent of its EPS synthesis [ 12 , 16 , 17 , 56 , 57 ]. Different studies have linked this phenomenon to biological parameters, since PHB biosynthesis is triggered by N-limiting conditions to generate C storage compounds, whereas EPS production is not dependent on the available N substrate concentrations [ 12 , 16 , 17 , 56 , 57 ]. Oxygen appears to have a lesser influence on polymer production than C source and C:N ratio. Under aerobic conditions, i Thauera861 only predicted an increase in the production of PHB, EPS3, and EPS4. Our modeling analysis was limited to predict the biosynthesis of only six EPS variants of the wide spectrum of EPS compositions. Many diverse EPS compositions and varying yields have been reported for Thauera sp . MZ1T, which results in different structures that influence flocculation [ 16 , 17 , 20 ]. To date, there is not enough data to conclusively determine which specific EPS variants and C substrates have an influence on the consistency of bacterial flocs [ 16 , 17 , 20 ]. Deciphering the impact of the main metabolic parameters evaluated in the current study to produce PHB and EPSs can potentially play a role to understand the development and quality of the wastewater microbial community biofilms and granular sludges. The variations in PHB and EPS production identified in our study have meaningful implications for wastewater treatment and bioremediation. Specifically, the composition and concentration of PHB and EPS directly impact the quality of wastewater sludge granules by affecting both the physical integrity and metabolic capabilities of the microbial community. The production of PHB and specific EPS variants aids in flocculation, enhancing the settling and compactness of biomass, which is essential for efficient wastewater treatment. Additionally, these polymers contribute to sludge permeability, which could facilitate nutrient transfer and removal efficiency. This structural support provided by PHB and EPS improves sludge stability and potentially enables more efficient operation of treatment facilities, especially under conditions where nutrient availability fluctuates. Thus, optimizing conditions to enhance PHB and EPS production could offer valuable strategies for improving the quality and performance of wastewater treatment systems. The GEM reconstructed, refined, and validated in the present study ( i Thauera861) provides new insights into how key parameters, including C source, oxygen levels, and C:N ratio, influence PHB and EPS production, both essential biopolymers for floc formation in wastewater treatment. Notably, i Thauera861 confirms a positive correlation between N-deficient carbon sources and enhanced polymer synthesis, which aligns with experimental data. Our findings also demonstrate that while oxygen presence has limited impact, the C:N ratio plays a dominant role in PHB yield. With applications extending to wastewater treatment, i Thauera861 offers a predictive tool for enhancing microbial granule quality by optimizing PHB and EPS production to improve sludge granule compactness, settling, and permeability. This metabolic model thus advances our understanding of Thauera sp . MZ1T and its role in wastewater treatment systems, providing a foundation for further environmental and bioremediation applications."
} | 5,322 |
31684861 | PMC6829849 | pmc | 3,535 | {
"abstract": "Background Experimental evolution of microbes often involves a serial transfer protocol, where microbes are repeatedly diluted by transfer to a fresh medium, starting a new growth cycle. This has revealed that evolution can be remarkably reproducible, where microbes show parallel adaptations both on the level of the phenotype as well as the genotype. However, these studies also reveal a strong potential for divergent evolution, leading to diversity both between and within replicate populations. We here study how in silico evolved Virtual Microbe “wild types” (WTs) adapt to a serial transfer protocol to investigate generic evolutionary adaptations, and how these adaptations can be manifested by a variety of different mechanisms. Results We show that all WTs evolve to anticipate the regularity of the serial transfer protocol by adopting a fine-tuned balance of growth and survival. This anticipation is done by evolving either a high yield mode, or a high growth rate mode. We find that both modes of anticipation can be achieved by individual lineages and by collectives of microbes. Moreover, these different outcomes can be achieved with or without regulation, although the individual-based anticipation without regulation is less well adapted in the high growth rate mode. Conclusions All our in silico WTs evolve to trust the hand that feeds by evolving to anticipate the periodicity of a serial transfer protocol, but can do so by evolving two distinct growth strategies. Furthermore, both these growth strategies can be accomplished by gene regulation, a variety of different polymorphisms, and combinations thereof. Our work reveals that, even under controlled conditions like those in the lab, it may not be possible to predict individual evolutionary trajectories, but repeated experiments may well result in only a limited number of possible outcomes.",
"conclusion": "Conclusions We have studied how in silico WTs of Virtual Microbes adapt to a serial transfer protocol like that of the LTEE. The LTEE has shown a sustained increase in competitive fitness, and intensive research displays how the evolved clones are still improving their growth rates with respect to their ancestor as to this day [ 66 – 68 ]. Our experiments have generated a novel hypothesis that microbes in a serial transfer protocol will eventually evolve to anticipate the regular resource interval, and can do so by evolving either a high growth rate mode, or a high yield mode. Both these modes can be achieved by a single individual lineage, or by a collective of two strains which both have their own temporal niche. Taken together, our results reveal important insights into the dynamics and relevant selection pressures in experimental evolution, advancing our understanding of the eco-evolutionary dynamics of microbes.",
"discussion": "Discussion In this study we have taken a serendipitous approach to study how microbes adapt to a serial transfer protocol, and to what extent this is determined by their evolutionary history. The Virtual Microbe modelling framework serves this goal by building biology from the bottom up, i.e. implementing basic biological features and their interactions. We observe that regardless of their evolutionary history, all WTs learn to anticipate the regularity of the serial transfer protocol by evolving a fine-tuned balance between high growth rate and yield. Long-term survival without nutrients, which is now masked from natural selection, always deteriorates after prolonged exposure to such a protocol. Furthermore, this anticipation is done in two distinct ways. The high yield mode makes sure that the cells are ready to divide as soon as transferred to a fresh medium, whereas the high growth rate mode maximally exploits the medium but results in a poor performance during the stationary phase. We next show that WTs have similar trajectories towards a growth versus yield trade-off, but may subsequently diverge along it. Polymorphisms within populations are frequently observed, which can happen by means of cross-feeding interactions, resource specialisation, or by means of growth vs. yield specialisation. We furthermore find that these evolved collectives are dependent on one another, as both lineages perform better in the presence of the other. Finally, we show that regulated gene expression allows for an individual lineage to fill both temporal niches by itself, but that populations without regulated gene expression can still be well adapted to the protocol by diverging into two strains. In general, our results are robust to details in the serial transfer protocol, such as using only a single resource, or varying the interval between transfers (see Additional file 1 : Table S2). The anticipation effects therefore appear to be generic features of microbes exposed to prolonged evolution in a serial transfer protocol. How do our results map onto experimental evolution in the lab? E. coli REL606 has been subjected to a daily serial transfer protocol for over 30 years (∼70.000 generations) in the LTEE. Many of our observations are very similar to the LTEE, such as the improved growth rate and cell sizes during the log phase[ 33 ], the (quasi-)stable dynamics of coexisting lineages[ 20 ], and “leapfrogging” dynamics (e.g. Fig. 5 a-b) where an abundant lineage is overtaken by another lineage before rising to fixation [ 38 , 39 ]. The comparison with respect to the growth rates, yield, and the anticipation effects discussed in this work, is however less straightforward. We have observed how all our WTs quickly evolve to be maximally efficient given our artificial chemistry, and only subsequently diverge along the apparent growth versus yield trade-off (see Additional file 1 : Figure S6). In the LTEE, growth and yield have continued to improve so far, and although a trade-off has been observed within the populations[ 40 ], no growth versus yield trade-off between the replicate populations has been observed yet. Nevertheless, we propose that anticipation of periodic environmental change, and a growth versus yield trade-off, provides testable hypotheses for the LTEE and similar experimental studies. More similarities with empirical studies are found in the surprising number of experiments that result in balanced polymorphisms. A repeatedly observed mechanism for such a polymorphism is cross-feeding [ 11 , 13 , 16 , 17 ], where modeling has shown that this adaptive diversification involves character displacement and strong niche construction[ 18 ], and furthermore strongly depend on the regularity of a serial transfer protocol [ 19 ]. We however also found balanced polymorphisms that did not include cross-feeding, involving one lineage with high growth rates during log phase and a slower growing lineage which performs better in stationary phase. Similar mechanisms of coexistence has been observed in respiratory and fermenting strains of Saccharomyces cerevisiae in chemostat [ 34 ], and single nucleotide mapping has furthermore revealed the existence of this trade-off [ 35 ]. These results are directly related to r/K selection theory [ 41 ], which describes an inherent conflict between the quantity and quality of ones offspring. Indeed, these dynamics have been shown to lead to two species stably coexisting in microbial populations [ 36 , 42 , 43 ]. Manhart & Shakhnovich [ 44 ] furthermore show that an unlimited number of species can theoretically coexist within a serial transfer protocol, occupying any niche on a trade-off continuum. Here we show that these dynamics can emerge from a more complex eco-evolutionary setting. However, our results suggest that the trade-off between growth and yield is not continuous, as intermediate solutions rarely evolve. This is caused by the fact that as soon as the volume-at-transfer for our digital microbes is smaller than the division volume (i.o.w. something else than the main nutrient becomes limiting for division), a cell may as well exploit its resources fully. Experimental evolution of Pseudomonas fluorescens has shown that different evolutionary paths can lead to the same phenotypic adaptations in a new environment [ 45 , 46 ]. On the other hand, many studies have also suggested that adaptation can often entail mutations in the same genes [ 47 , 48 ]. In our experiments, prior adaptations can in some cases strongly shape the way subsequent evolution plays out, but these evolutionary constraints can strongly differ between WTs (Additional file 1 : Figure S6). Furthermore, these data show that these evolutionary constraints may or may not diminish after prolonged evolution. There is a lot of variation on the predictability during the serial transfer experiment, revealing that evolutionary constraints by means of historical contingencies, are themselves the result of contingencies. A factor that has been hypothesised to strongly impact the predictability and evolvability of biological systems are their GRNs [ 6 , 49 – 51 ], where for example global transcription factors could serve as mutational targets with large-scale phenotypic effects [ 8 ]. While our results (Fig. 6 b) clearly show an example where similar mutations result in similar adaptive changes, other regulating WTs showed much less predictability. For example, WT #09 is another strong regulating WT, but showed different outcomes with respect to diversification and regulation in all 3 cases. In other words, while the GRN appears to add knobs and buttons for evolution to push, other mechanisms are clearly available to adapt and be fit in a serial transfer protocol. One such mechanism could be ‘metabolic regulation’, which has recently been shown to be able to achieve very high levels of robustness without leading to a loss in adaptive degrees of freedom [ 52 ]. Because all the kinetic parameters of enzymes (K m , V max , etc.) in the Virtual Microbes are freely evolvable, it is likely that this metabolic regulation of homeostasis plays a very important role in Virtual Microbes. This could furthermore explain why the differences in evolvability between regulating and non-regulating populations were smaller than we initially expected. We have indeed observed that, for certain WTs, a change in metabolism could bypass regulated protein expression by means of kinetic neofunctionalistaion of importer proteins, that evolved to be sensitive to different concentrations. Although such a solution does waste more building blocks on the continuous production of importer proteins, it is also much more responsive to environmental changes. It is possible that subtle differences like this explain, for example, why two of our WTs were much more sensitive to extinction by over-exploiting the medium than others. Furthermore, although the phenotypes that are reachable can be limited by prior evolution [ 53 ], the trajectories of evolution may be much less predictable on the long-term [ 54 ]. The role of metabolic regulation, and how this interplays with the repeatability and timescales of evolution, is a promising endeavour for future studies. Who is anticipating what? Our experiments reveal how populations of microbes can evolve to anticipate the regularity of a serial transfer protocol, trusting that new resources will be delivered on time. The concept of microbial populations anticipating predictable changes is frequently observed in nature [ 29 , 29 , 55 ], and is supported by theoretical models [ 30 , 56 ]. This form of anticipation however typically entails an environmental cue, where a preceding unrelated signal is used to anticipate environmental changes, usually followed by individuals taking some form of action. Without the necessity of such a cue, we show that anticipation can readily emerge in many different ways from an eco-evolutionary process. Although our form of anticipation is more passive, where not an individual but the system as a whole has temporal dynamics that accurately fit the protocol, this does not necessarily exclude individual-based anticipation. Like WT#07, most of the evolved regulating populations actually did not evolve to down-regulate their resource importers during the stationary phase, despite having repeatedly evolved to down-regulate other catabolic and anabolic enzymes (illustrated in Fig. 6 b). Since no more resource is available, and building blocks are consumed in order to keep expressing these importer proteins, this clearly does not have a positive impact during the late stationary phase. One can wonder why these individuals seem to keep the engine running. Whereas bet-hedging strategies have been shown to be a way to deal with irregular environmental changes [ 24 , 26 – 28 , 57 , 58 ], this passive form of anticipation can be a way deal with regular, predictable changes in the environment. Furthermore, this could potentially be the first step towards active anticipation by means of a circadian rhythm, such as the sunflower heliotropism [ 59 ] and the diurnal migration of life in lakes and oceans [ 60 – 62 ]. Moving towards an eco-evolutionary understanding The dynamics of Virtual Microbes expose that even a simple serial transfer protocol entails much more than sequentially evolving higher and higher growth rates. Instead, adaptation is an eco-evolutionary process that strongly depends on prior evolution, timescales, the presence of other competitors and mutants, and transient fitness effects. Although we found that competition experiments generally favoured the evolved population over the ancestral WTs, there were exceptions to this rule. It is therefore possible that the ancestral WTs perform better in such an experiment, but that this does not describe the stable eco-evolutionary attractor. Indeed, survival of the fittest is an eco-evolutionary process where any emerging lineage interacts with other lineages (or with other mutants) through changes in the environment, often resulting in a collective, community-based solution rather than the winner of all pair-wise interactions [ 44 ]. Furthermore, faster growth becomes less and less important as populations become better adapted to the serial transfer protocol, perhaps making the aforementioned interactions between lineages increasingly relevant. Other recent studies have recently elucidated the importance of eco-evolutionary dynamics [ 44 , 63 ], and how this can readily give rise to coexistence of multiple strains which could not have formed from a classical adaptive dynamics perspective [ 64 , 65 ]. Indeed, metagenomics have revealed much more diversity in the LTEE than previously anticipated [ 20 ]. Shifting focus from competition experiments towards the ever-changing selection pressures that emerge from the eco-evolutionary dynamics and interactions, will make the field of experimental evolution harder, but more intriguing, to study."
} | 3,715 |
37555068 | PMC10405931 | pmc | 3,536 | {
"abstract": "Microorganisms in subsurface sediments live from recalcitrant organic matter deposited thousands or millions of years ago. Their catabolic activities are low, but the deep biosphere is of global importance due to its volume. The stability of deeply buried sediments provides a natural laboratory where prokaryotic communities that live in steady state with their environments can be studied over long time scales. We tested if a balance is established between the flow of energy, the microbial community size, and the basal power requirement needed to maintain cells in sediments buried meters below the sea floor. We measured rates of carbon oxidation by sulfate reduction and counted the microbial cells throughout ten carefully selected sediment cores with ages from years to millions of years. The rates of carbon oxidation were converted to power (J s −1 i.e., Watt) using the Gibbs free energy of the anaerobic oxidation of complex organic carbon. We separated energy dissipation by fermentation from sulfate reduction. Similarly, we separated the community into sulfate reducers and non-sulfate reducers based on the dsrB gene, so that sulfate reduction could be related to sulfate reducers. We found that the per-cell sulfate reduction rate was stable near 10 −2 fmol C cell −1 day −1 right below the zone of bioturbation and did not decrease with increasing depth and sediment age. The corresponding power dissipation rate was 10 −17 W sulfate-reducing cell −1 . The cell-specific power dissipation of sulfate reducers in old sediments was similar to the slowest growing anaerobic cultures. The energy from mineralization of organic matter that was not dissipated by sulfate reduction was distributed evenly to all cells that did not possess the dsrB gene, i.e., cells operationally defined as fermenting. In contrast to sulfate reducers, the fermenting cells had decreasing catabolism as the sediment aged. A vast difference in power requirement between fermenters and sulfate reducers caused the microbial community in old sediments to consist of a minute fraction of sulfate reducers and a vast majority of fermenters.",
"introduction": "1. Introduction As continued deposition gradually buries marine sediments, they become increasingly isolated from the surface world. Dissolved electron acceptors can be supplied to the isolated microbial community from above, but the pool of organic carbon that fuels microbial life is mostly constrained to the stationary solid phase (e.g., Komada et al., 2013 ). Thus, a microbial community must live from the finite amount of organic carbon buried with it in the sediment. The anaerobic food chain is inefficient, and repeated cycles of cell death and reassimilation of necromass would lead to rapid loss of carbon ( Orsi et al., 2020 ). Nevertheless, we find microorganisms in ancient sediments that are still thriving and are slowly degrading the old and refractory organic matter ( Røy et al., 2012 ). This implies that the rates of mineralization are exceedingly low. Indeed, the reactivity of organic matter decreases steeply in aging sediment ( Middelburg, 1989 ; Boudreau and Ruddick, 1991 ; Shang, 2023 ). As the rates of carbon mineralization decrease with increasing age, so does the size of the microbial community ( Røy et al., 2012 ). Although the assembly of the deep biosphere community conserves a part of the surface community ( Starnawski et al., 2017 ), the decreasing community size with increasing age and depth in the sediment is not merely due to a slow death of the surface community. This can be seen by the continuous production of dead microbial cells (necromass) far in excess of the size of the original community ( Lomstein et al., 2012 ), and by the fact that the estimated biomass turnover times of the sedimentary microbes is much shorter than the age of the sediment they live in Biddle et al. (2006) , Hoehler and Jørgensen (2013) , Braun et al. (2017) . Thus, the microbial community in the deep sediment column must largely be in steady state with respect to their basal power requirement and the local availability of energy at any time. As the rate of liberation of labile carbon substrates from the refractory organic matter decreases with time, the size of the microbial community is decreasing accordingly. If the energy supply were in excess, the community would grow and thereby reduce the energy flux available per cell to approach the basal power requirement. If the energy supply falls below the basal power requirement, some cells will die, and this increases the per-cell energy availability ( LaRowe and Amend, 2015b ). Thus, we expect that the ever-decreasing energy turnover in aging marine sediments will force cells in the deep biosphere to constantly exist at the lowest power dissipation that will sustain their community ( Hoehler and Jørgensen, 2013 ; Lever et al., 2015 ). The lower limit, i.e., the basal power requirement, of prokaryotic cells is, most likely, set by physical and chemical decay processes in the cells such as the rate of leakage of membrane potential, the rate of depurination of nucleic acids, and the rate of racemization of amino acids. None of these processes are, however, constrained well enough to confidently calculate the basal power requirement of the individual cells ( Lever et al., 2015 ). Yet, the intrinsic rate of racemization of aspartic acid, which is the amino acid with the highest rate of racemization, indicate that this process leads to the largest unavoidable loss of energy ( Brinton et al., 2002 ; Onstott et al., 2014 ). Indeed, a gene encoding the enzyme (Protein-L-iso aspartate(D-aspartate) O-methyltransferase), which recognizes damaged L-isoapartyl and D-aspartyl residues in proteins and catalyzes their repair while still within the protein, was found widely distributed and expressed in deeply buried sediments in the Baltic Sea ( Mhatre et al., 2019 ). The lowest basal power requirement for prokaryotes in the deep biosphere is difficult to determine in the laboratory yet experiments with axenic bacterial cultures maintained without addition of substrates for prolonged time under so-called long-term stationary phase have shown general mechanisms of adaptation in cell respiration to extreme nutrient limitation (e.g., Riedel et al., 2013 ; Robador et al., 2019 ). To avoid artifacts related to laboratory cultivation, we have searched for the basal power requirement of cells in the natural environment by relating the catabolic rate of a community to the community size ( Hoehler and Jørgensen, 2013 ) directly in sediments of varying age, ranging from tens to millions of years old. To limit the number of variables in the data, we focused on the sulfatic zone ( Canfield and Thamdrup, 2009 ; Jørgensen, 2021 ), where anaerobic respiration is dominated by sulfate reduction. The goal of our study was to identify if, and how, the availability of energy controlled the microbial community size and the per-cell metabolic rate in the energy-starved deep biosphere. In addition, we compared the community size of the two main metabolic guilds, fermenters and sulfate reducers, living syntrophically in sulfate-rich sediments, with the power available to each of the guilds.",
"discussion": "4. Discussion 4.1. Reaction rate and reactivity of organic matter in old sediments The careful selection of sites and methods allowed us to determine the rate of mineralization of organic matter in sulfatic sediments ( Canfield and Thamdrup, 2009 ; Jørgensen, 2021 ) with ages from 10 years to 3,750,000 years. The data extended the power law of organic matter decay constants, first presented by Middelburg (1989) , to even older sediments and overcame the original need for a site-specific “initial age,” possibly because our data were more uniform with respect to temperature and geochemical zone. In contrast to previous data syntheses [e.g., Middelburg (1989) and Katsev and Crowe (2015) ], our approach did not rely on a measurable decrease in the concentration of organic matter with increasing depth and age of the sediment. In fact, we found little correlation between depth in the sediment column and the concentration of organic matter ( Figure 3 ). Our correlation predicted a slightly faster loss of reactivity with k = 0.180 × sediment age -1.11 ( Eq. 13 ), compared to models based on, or verified from, vertical profiles of organic carbon, which have exponents ranging from −0.8 to −1 ( Shang, 2023 ). Power-law models with exponents close to −1 demonstrate that the overwhelming factor responsible for decreasing rates of microbial catabolism in ageing sediments is not so much the loss of organic carbon, but rather the loss of organic carbon reactivity ( Figures 3 , 4D ). Conceptually, this corresponds well with the increasing molecular complexity of degrading organic matter (e.g., Estes et al., 2019 ; Hach et al., 2020 ), but poorly with models that assume concurrent degradation of multiple pools of organic matter with individual degradability ( Jørgensen, 1978b ; Shang, 2023 ). Note, however, that the two types of models provide equally good fits to empirical data ( Arndt et al., 2013 ). 4.2. The link between sediment age, community size, and respiration rate Cell numbers decreased as a function of sediment depth and age, as seen in several previous studies ( Whitman et al., 1998 ; Kallmeyer et al., 2012 ). The decrease in total cell abundance with increasing age of the sediment was much slower than the decrease in carbon oxidation rates ( Figure 4A vs. Figure 4B ), which implied a large drop in the mean cell-specific rates of catabolism ( Figure 4D ). The degradation of organic matter in sulfatic marine sediments is, however, divided between two major guilds of cells, one guild that hydrolyses and ferments complex organic matter to volatile fatty acids and hydrogen, and a second guild that oxidizes these fermentation products to CO 2 and water while reducing sulfate to sulfide. The rate-limiting step in mineralization of complex organic matter is the initial hydrolysis ( Beulig et al., 2018 ) and the fermenters will take up the resulting monomers so efficiently that their concentrations are mostly too low to even detect. The community that oxidizes the fermentation products via anaerobic respiration is equally efficient, and fermentation products do not normally accumulate above a few μM ( Postma and Jakobsen, 1996 ; Wang et al., 2010 ; Glombitza et al., 2015 ). The metabolic guild that has the most energetic respiratory metabolism will deplete the fermentation products to such low concentrations that energy conservation via proton translocation across the cell membrane is only barely possible. This excludes energy conservation from respiration with less energetic electron acceptors and causes the characteristic geochemical redox-zonation with limited overlap between utilization of different external electron acceptors ( Postma and Jakobsen, 1996 ). We can, therefore, assume a tight link between fermentation and sulfate reduction regardless of which specific fermentation processes that were active: The fermenters oxidize some organic carbon to CO 2 , while concurrently producing more reduced carbon or H 2 in the process. When these electron-rich substrates are then used by the sulfate reducers, it balances the stoichiometric ratio between sulfate consumption and total CO 2 production to the same value as if the sulfate reducer had mineralized the original organic matter with no fermentation involved ( Eq. 2 ). This way, the electrons from complex organic matter must pass through both fermentation and sulfate reduction for the carbon to be funneled fully to CO 2 , and the rate of sulfate reduction is a measure of the flow of carbon and electrons through both processes. Even syntrophic interactions with direct electron transfer do not change the overall stoichiometry between mineralized organic carbon/CO 2 and sulfate. The fermentation product acetate is instrumental in the transfer of reducing equivalents from fermentation to sulfate reduction, and oxidation of acetate accounts for 30–65% of the sulfate reduction rate in sulfatic sediments ( Christensen and Blackburn, 1982 ; Finke et al., 2007 ; Beulig et al., 2018 ). We specifically selected sediments for the study where all other electron accepters than sulfate and CO 2 had already been depleted. Under these conditions, sulfate reduction is the primary terminal electron accepting process, and all known prokaryotes that respire via sulfate use the dsrB gene. The gene is also used in sulfide oxidation, but the sediment strata that we analyzed did not contain suitable electron accepters for this process. We therefore assumed that all cells that contained the dsrB gene were potential sulfate reducers, and as they were found in sulfate-reducing sediment, we assumed that all potential sulfate reducers were active. This allowed us to calculate the per-cell sulfate reduction rate. We found elevated rates of per-cell sulfate reduction in upper centimeters of sediment with active bioturbation. But when both the decreasing cell numbers and the decreasing proportion of sulfate reducers were considered, the cell-specific carbon oxidation rate by sulfate-reducing microorganisms in the subsurface stayed remarkably constant on the order of 10 −2 fmol C cell −1 day −1 , despite increasing depth and a drop in sulfate reduction rates of four orders of magnitude ( Figure 5 ). Previous studies have shown relatively high cell-specific rates of sulfate reduction in the very surface of coastal sediments, gradually decreasing to an organic carbon oxidation rate of 10 −3 to 10 −2 fmol C cell −1 day −1 between 30 and 100 cm below seafloor ( Hoehler and Jørgensen, 2013 ; Petro et al., 2019 ). Interpretation of the data deeper in the sediment from these coastal sites is difficult because of the transition from the sulfatic zone and into the methanic zone, and because the early qPCR protocols and primers used ( Leloup et al., 2007 , 2009 ) lacked the necessary resolution. In an attempt to predict the cell specific rates of sulfate reduction deeper in the sediment, Lever et al. (2015) extended the estimate of Hoehler and Jørgensen (2013) into deeply buried sediments of the Peru Margin by assuming that sulfate reducers constituted 10% of the total microbial community regardless of sediment depth and age. This assumption resulted in calculation of constantly decreasing per-cell rates of sulfate reduction in deeper sediment, but later studies have not confirmed the proportion of sulfate reducers to be constant ( Webster et al., 2006 ; Petro et al., 2019 ). Thus, the constant cell specific sulfate reduction rates across depth and age up to 20,000 years seen in our study do not contradict the more recent studies and might imply that the size of the sulfate-reducing community is indeed controlled by a fixed minimum energy requirement of these microorganisms, as hypothesized by Hoehler and Jørgensen, (2013) . Further, this minimum metabolic rate of sulfate reducers in deeply buried marine sediments might not be far from that seen below the depth of bioturbation in sediment that is buried only 30 cm deep and is only 300 years old ( Petro et al., 2019 ). Our attempt to detect sulfate-reducing prokaryotes in sediments older than 26,000 years was not successful as the dsrB gene copy numbers fell below our limit of quantification of ~10 4 gene copies cm −3 . Thus, it awaits more sensitive experimental methods to see if the proportion of sulfate reducers continue to drop with increasing sediment age beyond 26,000 years, which is necessary for the cell specific sulfate reduction rates to stay at the constant level of 10 −2 fmol C cell −1 day −1 that we observed down to this depth. There are, however, indications that this could indeed be the case: Studies have revealed (a) low abundance of functional genes related to sulfate reduction ( dsrAB genes) or methanogenesis (methyl-coenzyme M reductase, mcr genes) in deep sediments (<1% of total community, Schippers and Neretin, 2006 ; Lever, 2013 ), (b) fermentation-related genes that were much more abundant than dsr or mcr genes in metagenomes ( Kirchman et al., 2014 ; Gaboyer et al., 2015 ), and (c) apparent virtual absence of genes related to sulfate reduction in metagenomes from deep sulfatic sediments of the Bering Sea ( Biddle et al., 2008 ). The apparent difficulty in detection and quantification of dsr genes in the diverse pool of DNA extracted form sediment from deep below the sea floor on the Peru Margin indicate that terminal-oxidizing prokaryotes are present at low abundance ( Webster et al., 2006 ), while the consistent detection of mRNA transcript of dsr ( Orsi et al., 2013 , 2016 ) indicate that sulfate reduction play a larger role in community activity than the low proportion of sulfate-reducing prokaryotes suggests. 4.3. Division of Gibbs free energy between fermentation and sulfate reduction To assess the division of Gibbs free energy (ΔG r ) between fermenters and sulfate reducers, we calculated the Gibbs free energy (ΔG r ) by complete oxidation of organic matter to CO 2 via Eqs 11 , 12 according to LaRowe and Van Cappellen (2011) . We also calculated ΔG r for oxidation of acetate to CO 2 with sulfate as electron acceptor based on measured concentrations of reactants and products (Eqs 12 , 13 ). By subtracting the energy yield of this terminal oxidation from the total energy yield, we could estimate the energy yield of fermentation under in situ conditions without needing to know the molecular identity or the concentrations of fermentation substrates (i.e., the products of hydrolysis). This is a rather crude approximation, as fermentation processes produce other products than acetate, and sulfate reducers oxidize H 2 and other volatile fatty acids, such as formate or propionate, in addition to acetate ( Glombitza et al., 2015 ). Moreover, this calculation overestimates the energy available for the fermenters, as the free energy from extracellular hydrolysis cannot be coupled to energy conservation (i.e., to ATP). Note also that our operational definition of “fermenters” includes all organisms involved in production of acetate regardless of the actual biochemical pathway. Thus, the purpose was not to calculate the accurate ΔG r of the processes or to compare energy yields close to thermodynamic thresholds. But to test if the overwhelming dominance of fermenting cells could be explained by an inequal sharing of energy between fermentation and sulfate reduction. Thus, we calculated the division of energy between fermentation and sulfate reduction on the four Greenland cores and in the oldest core from the Eastern Equatorial Pacific, where we had the most complete data on pore water chemistry that is needed. In the core from the Greenland Continental Shelf (SA13-ST3-20G), the ΔG r liberated by fermentation 1 m below seafloor was −43 kJ (mol C) −1 , while the ΔG r liberated by acetate oxidation was −24 kJ (mol C) −1 . At 5.9 meters below seafloor, the values changed only slightly to −42 kJ (mol C) −1 and − 18 kJ (mol C) −1 , respectively. The remaining Greenlandic cores (SA13-ST5-30G, SA13-ST6-40G, SA13-ST8-47G) were in the same range, with even less variation down-core. In the core from the Eastern Equatorial Pacific (ODP Leg 201, Hole 1,226-B), the fermenters and the terminal oxidizers shared the energy in a similar manner with −42 to −43 kJ (mol C) −1 for the fermenters and − 35 to −26 kJ (mol C) −1 for the sulfate reducers. Similar values of −42.9 ± 2.7 kJ mol acetate −1 have previously been reported for sulfate reduction coupled to acetate oxidation for most of the sediment column at ODP Site 1226 ( Wang et al., 2010 ), and − 32 kJ mol C −1 for global marine sediments in general ( Bradley et al., 2020 ). The assumptions and approximations in the thermodynamic calculations above were coarse. Most notably the assumed temperature of 25°C, while the effects of the pressure between one bar and in situ were negligible ( Helgeson, 1969 ). But effects of pressure and temperature on thermodynamic calculations are much less severe than the effects on kinetics ( Jannasch, 1997 ), and even the temperature effect does not influence the conclusions drawn here: The standard Gibbs energy of sulfate reduction per mole acetate, for example, only change from −48.1 to −44.5 kJ (mol acetate) −1 between reference temperature and pressure (25°C, 1 bar) and in situ conditions (2°C, 50 bar) at the Greenlandic sites. Thus, the coarse calculations still allow us to conclude that ΔG r from mineralization of organic carbon to CO 2 was shared between the guilds of fermenters and terminal oxidizers in a ratio of roughly 1:1 regardless of site and sediment age. Since the fermentation products do not accumulate in the sediment, there must be a balance between fermentation and sulfate reduction, whereby rates of fermentation limit and control the rates of sulfate reduction ( Jørgensen, 2021 ). Thus, the carbon flow through fermentation must be similar to the carbon flow through sulfate reduction. As non-sulfate-reducing bacteria outnumbered the sulfate reducers by >100-fold, the energy dissipation (i.e., power) available to the individual fermenters was much lower than the power available to the individual sulfate reducers. Volume-specific power of reactions in the sediment (Watt cm −3 or Joule s −1 cm −3 ) was calculated by multiplying ΔG r of the reaction (J mol −1 \\C) by the rates of reaction (mol C cm −3 s −1 ). The mean cell-specific sulfate reduction rates in the Greenland cores were in the order of 10 −2 fmol C cell −1 day −1 , which was not far outside the range from 10 −1 to 10 1 fmol C cell −1 day −1 seen in cultures of mesophilic and psychrophilic sulfate-reducing bacteria ( Canfield et al., 2000 ). The 10 −2 fmol C cell −1 day −1 translates to a power dissipation of 10 −17 W cell −1 if we assume an energetic yield of −42 kJ (mol C) −1 (see above and note that J/s equals W). This rate of energy dissipation is similar to the value calculated for coastal surface sediments by Lever et al. (2015) , and similar to the per-cell power dissipation in slow-growing axenic cultures of Desulfotomaculum putei ( Davidson et al., 2009 ). It is also similar to the power dissipation of extremely low-light adapted green sulfur bacteria in the Black Sea (1.9 × 10 −17 W cell −1 ; Marschall et al., 2010 ). Conversely, the values are far above the per-cell power dissipation calculated for deeply buried sulfate reducers on the Peru Margin by Lever et al. (2015) , even though our estimation of the carbon oxidation rates in the same deep Pacific sediment based on NH 4 + agrees well with calculations based on sulfate ( Wang et al., 2008 ). The difference between our relatively high and invariant estimate of power dissipation by deeply buried sulfate reducers, and the low and age-dependent power dissipation reported by Lever et al. (2015) is that these authors assumed a constant 10% proportion of sulfate-reducing microorganisms (see above). The fact that the power dissipation we calculated per sulfate-reducing cell did not continue to drop with depth and age in the sediment, is because the estimated number of sulfate reducers decreased in near perfect balance with the decrease in sulfate reduction rates. This could indicate that 10 −17 W cell −1 approaches a minimum power requirement of sulfate reducers in deeply buried sediments. As the same range of power dissipation can be observed only 30 cm below the sediment–water interface in eutrophic coastal sediments and in axenic cultures, this suggests that the sulfate reducers in the deep biosphere do not necessarily have any unique physiological adaptations to low energy availability. The cell-specific power dissipation of extremely old oxic sediments has been calculated to 5 × 10 −18 W cell −1 in the North Pacific Gyre, and within the range from 3.5 × 10 −18 down to 4.9 × 10 −20 in the South Pacific Gyre ( LaRowe and Amend, 2015a ). These calculations assume an energetic yield of −443 kJ (mol C) −1 for aerobic oxidation of organic matter all the way to CO 2 in a single step ( Lever et al., 2015 ). The low values from the SPG were calculated based on an assumed constant rate of carbon burial across the 75 million years and is, most likely, less accurate than the data from the North Pacific Gyre (NPG) that was based directly on modelled oxygen consumption rates ( Røy et al., 2012 ). Thus, the apparent maintenance power of aerobic and anaerobic respiration come within one order of magnitude of each other. And in contrast to the per cell power dissipation of the “fermenters” (or the total carbon oxidation per total cells), both aerobic and anaerobic respiration appears to converge at fixed rates of per cell power dissipation rather than dropping continuously with increasing sediment age. The difference in the two levels is not understood, as the higher energetic cost of biomolecule synthesis aerobes ( McCollom and Amend, 2005 ) would suggest that the minimum maintenance power of aerobes should be larger than that of anaerobes. It is possible to estimate the power dissipation by the guild of fermenters in the deep sulfatic sediments older than 26,000 years that we studied, even though the proportion of fermenters vs. terminal oxidizers was not known. This is because the number of fermenters is essentially equal to the total cell numbers, only offset by 1% sulfate reducers or less. In contrast to the situation for sulfate reducers, the per cell power dissipation for the fermenters did not reach a plateau that could indicate a minimum power requirement of fermenting cells, although such a limit must exist. The lowest values reached in our dataset was 7 × 10 −22 W cell −1 . This rate of power dissipation is so low that the cells would just barely cover the lowest estimate of the repair cost associated with spontaneous racemization of aspartic acid ( Lever et al., 2015 ). But there are no indications in the data that 10 −6 fmol C cell −1 d −1 , corresponding to 10 −22 W cell −1 is the lower limit, it is simply where our dataset ends. In principle, the correlation could extend further back in time. But extrapolation of log–log plots back in time from 4 million years would rapidly reach the age of Planet Earth. In conclusion, we found that the population size of sulfate-reducing microorganisms was in balance with the rate of sulfate reduction in sub-surface sediments. The minimum per-cell sulfate reduction rate, and thus the minimum per cell power dissipation, in 10,000 to 100,000 years old sediment was not radically different from the much shallower sub-surface. Thus, we do not expect sulfate reducers in the deep subsurface to have a fundamentally different maintenance metabolism compared to organisms thriving less than 1 m below the seafloor in coastal seas ( Starnawski et al., 2017 ). We did not see a lower threshold for metabolic activity of cells not involved in terminal oxidation, although there was a clear log–log correlation between the total number of cells and the rate of carbon mineralization. It is unknown which physiological properties make the fermenters able to apparently subside with a minute fraction of the power needed for sulfate reducers. The large number of cells relative to the extremely low rates of mineralization in the deep biosphere therefore remain enigmatic."
} | 6,899 |
29308273 | PMC5750040 | pmc | 3,537 | {
"abstract": "We present an effective dynamical model for the onset of bacterial bioluminescence, one of the most studied quorum sensing-mediated traits. Our model is built upon simple equations that describe the growth of the bacterial colony, the production and accumulation of autoinducer signal molecules, their sensing within bacterial cells, and the ensuing quorum activation mechanism that triggers bioluminescent emission. The model is directly tested to quantitatively reproduce the experimental distributions of photon emission times, previously measured for bacterial colonies of Vibrio jasicida , a luminescent bacterium belonging to the Harveyi clade, growing in a highly drying environment. A distinctive and novel feature of the proposed model is bioluminescence ‘quenching’ after a given time elapsed from activation. Using an advanced fitting procedure based on the simulated annealing algorithm, we are able to infer from the experimental observations the biochemical parameters used in the model. Such parameters are in good agreement with the literature data. As a further result, we find that, at least in our experimental conditions, light emission in bioluminescent bacteria appears to originate from a subtle balance between colony growth and quorum activation due to autoinducers diffusion, with the two phenomena occurring on the same time scale. This finding is consistent with a negative feedback mechanism previously reported for Vibrio harveyi .",
"conclusion": "6. Conclusion We proposed a model for the QS mechanism that closely reproduces experimental data on bioluminescence by V. jasicida , within the set-up of Delle Side et al . [ 30 ]. By using a simulated annealing procedure, we obtained biochemical and growth parameters that are in good agreement with the literature data. The results of our numerical fit clearly suggest that the activation of the quorum mechanism is intertwined with the rapid growth of the bacterial population. In other words, the bioluminescence response acts exactly as a proxy to sense the rapid growth of the colony. This concerted trade-off could be simply due to the choice of the initial bacterial cell density in the experiments, or rather enforced by a feedback mechanism that tunes the sensitivity of the QS response to AI concentration, as already reported for V. harveyi [ 53 ]. Understanding whether the latter mechanism operates also in the experimental set-up considered in this paper is a subject that deserves further investigation. Moreover, a distinctive novel feature of the proposed model is the presence of a short (approx. 30 s) bioluminescence ‘quenching’ time, crucially needed to explain the observed data. Although the quenching time in our numerical fits is assumed to be the same for all activated cells in the bacterial colony, heterogeneous quenching behaviour could as well be introduced within the framework proposed here. The nature of single cell quenching behaviour and its hypothesized connection with the protective role of bacterial luciferase against reactive oxygen species are issues that call for further experiments.",
"introduction": "1. Introduction Many living organisms are able to transform chemical energy into visible light, an ability known as bioluminescence [ 1 ]. Light emission is due to a reaction involving molecular oxygen, occurring on a substrate (luciferin, in most cases) and catalysed by an enzyme (luciferase). Substrate and enzyme properties change significantly across different bioluminescent systems, the sole common features being light emission and the requirement for molecular oxygen. Bioluminescent bacteria are the most abundant and widely distributed light-emitting organisms [ 2 , 3 ]. In cases when bacteria grow as symbionts with fishes or squids, the function of light emission relates to the use of photogenic organs by the host, whereas bacteria receive nutrients. In these organisms, the light-emitting reaction involves a luciferase-catalysed oxidation of reduced flavin mononucleotide, with the concomitant oxidation of a long-chain aliphatic aldehyde. This leads to the emission of blue–green light from an electronically excited species. The study of the onset of bioluminescence in bacterial colonies has led to the finding of the fascinating mechanism now generally called ‘quorum sensing’ (QS) [ 4 ]. QS is the mechanism by which bacteria are able to ‘sense’ their environment, by activating complex collective actions that result beneficial to bacterial cells only when carried out by a group. Besides bioluminescence, these actions involve the expression of genes that control biofilm development, virulence and several other traits [ 5 ]. There is a large knowledge about the biochemistry of QS [ 6 ]. It is widely known that bacteria realize this mechanism through the synthesis, secretion and detection of some chemical signalling molecules known as autoinducers (AIs). In particular, a key role is played by the AI concentration profile that is generated across the colony. QS-responsive bacteria are able to detect the attainment of a ‘quorum’ threshold concentration of AIs through their binding to dedicated receptors. This triggers an intracellular signalling cascade that results in a phenotypic switch. In this way, all cells in the colony can become ‘quorum-active’ and start new collective behaviours in a coordinate manner. The evolutionary origin of QS is still being hotly debated. AI concentration was initially considered as a direct proxy for cell density. In fact, this concept is at the origin of the expression ‘quorum sensing’ [ 7 ]. However, it is now clear [ 5 , 8 – 16 ] that AI concentration could be affected by several environmental factors other than cell density, such as AI diffusion, advection, the spatial arrangement of bacteria and the colony extension from an adsorbing boundary [ 17 ]. Furthermore, it has been suggested that AIs could work also as transducers of host cues [ 18 ]. Lately, the QS mechanism attracted significant interest as a possible target for the development of drugs that interfere with AIs’ signalling in order to prevent biofilm development or the expression of virulence factors. In particular, it has been considered as a target for next-generation drugs [ 19 , 20 ] able to overcome the problems arising from the rapid increase of antibiotic-resistant bacterial diseases which have been observed over the last decades [ 21 ]. Despite the important scientific achievements obtained in the field of QS biochemistry, we still know very little about the generic features of the collective QS behaviour, and especially of its interplay with the growth of a bacterial colony. For example, it is practically impossible to answer the simple question ‘when will a growing bacterial colony reach QS?’. Several groups proposed interesting dynamical models to describe QS [ 22 – 27 ]. All such approaches, however, are based on a detailed description of the biochemical processes underlying QS, thus requiring a very large number of parameters, most of which are extremely difficult to know with a reasonable precision. Furthermore, some of the models available so far were developed when the biochemical details about QS had not been fully assessed yet. In this work, we present a simplified effective dynamical model for the onset of bacterial bioluminescence. Our main aim is to disentangle the roles of colony growth and of quorum activation in shaping the bioluminescent signal. The model is directly tested to quantitatively reproduce the experimental distributions of emission times for photons emitted by growing bacterial colonies of Vibrio jasicida , a member of the Harveyi clade [ 28 , 29 ]. We use data that were previously measured for bacterial colonies growing in a highly drying environment [ 30 ]. This minimizes swarming and allows us to model bacteria as fixed on the growing substrate. This is a crucial simplification for our model that is missing or is unjustified in other approaches. The same experimental data were observed to follow the extreme value Gumbel statistics [ 30 ]. However, the connection between this kind of statistical distribution and bioluminescence activation was hypothetical, and the Gumbel distribution parameters used to fit the data had no clear interpretation. The simplified model that we propose here is able to fit experimental bioluminescence data with a slightly better quality than the Gumbel statistics. Most importantly, all parameters used in the model have a definite biochemical meaning and the values we estimate for them are consistent with what is known in the literature. Moreover, a good fitting of experimental data crucially requires that all cells cease light emission after a typical ‘quenching time’ (approx. 30 s) elapsed since their own activation. Intriguingly, this novel feature might be part of an oxygen quenching mechanism that establishes a tight metabolic control over the amount of reactive oxygen species in the bacterial ‘milieu’ [ 31 , 32 ]. Finally, our model allows to show that the overall bioluminescent signal can be split into two roughly equally weighted contributions, one related to the growth of the colony and the other to the increase in the number of bioluminescent cells due to quorum activation. This rationalizes the highly nonlinear relationship between the total number of emitted photons and the number of bacterial cells previously observed during growth [ 33 ].",
"discussion": "5. Discussion It was previously observed that the same experimental data studied here can be fitted with a Gumbel function [ 30 ]. Yet, that was just an empirical fit and could not be explained by a predictive model such as the one we propose in this paper. Although the Gumbel distribution occurs in extreme value statistics, in bacterial QS no underlying extremal process is known to take place. On the other hand, it had been conjectured [ 62 ] that Gumbel statistics can originate in systems with spatially averaged critical properties. One could then speculate that bacterial colonies may represent an example of such a system [ 30 ]. Our present results show instead that a simple modelling of the QS mechanism is sufficient to reproduce the experimental data with a slightly higher accuracy, without relying on any unproven conjecture or unknown mechanism. The connection of the experimental curve with the Gumbel function can be rationalized by observing that the bacterial growth curve is modelled by a Gompertz function, e.g. the integral of the Gumbel function [ 33 ]. Having established a model that relies on a simple biochemical interpretation and quantitatively reproduces experimental data, we can then put forward some interesting observations. First, in our simple dynamical model several features, well established for QS systems, were neglected, such as the positive feedback on AI production [ 6 ], the integration of signals from different AIs in the QS cascade [ 56 ], and the strongly heterogeneous bioluminescent response at single cell level [ 67 ]. The ability of our model to quantitatively and consistently reproduce the experimental signal is therefore remarkable. Although the above features should be considered in a more realistic and sophisticated model, they are not necessary to rationalize the time evolution of the bioluminescent signal produced by a growing bacterial colony, at least within the set-up of Delle Side et al . [ 30 ]. Second, it has to be stressed that the concerted behaviour observed in figure 2 , with the growth of the bacterial colony and the increase in the fraction of active cells contributing with roughly equal amplitude and timing to the bioluminescent signal, is not the only a priori possible outcome of the experiments. Indeed, one could have, in principle, two alternative scenarios as limiting cases. In the first one, say at very high effective AI production rate α ~ , AI concentration would reach the threshold c * at a time t 1 smaller than the lag time λ. In such a case, the overall signal would present a first smaller peak at t 1 (because n ( t ) would still be small) followed by a main second one at around λ due to g ( t ), the contribution of h ( t ) being negligible (when the Hill function is already nearly constant and close to 1, for t ≫ t 1 ). Conversely, at low enough effective AI production rate α ~ , the first increase in the concentration of signal molecules is achieved because of the growth of the colony, prior to quorum activation. This peak would thus be small, because h ( t )≪1. The main peak in the radiation flux should then occur at t 2 >λ, driven by h ( t ), the contribution of g ( t ) being negligible due to the saturation of the bacteria population n ( t ) for t >λ. For even lower values of the effective AI production rate, no radiation flux would be eventually detected, when the saturation AI concentration αn (0)×10 K /( βV ) at the end of the colony growth is still below the quorum threshold c * and the colony never gets activated. For the practical realizations of the above-described scenarios, one should consider that the distinction between active/inactive cells (for AI concentration above/below the quorum threshold) becomes less sharp in the case of essentially non-cooperative AI/receptor binding, as is ours. Finally, we highlight that according to the model that we present here, different cells start bioluminescent emission, upon QS activation, at different times. Each cell emits photons with the same rate for approximately 30 s after its own activation and then ceases further emission. The details of how we modelled the quenching mechanism, in particular the choice of a step-wise function equation ( 3.6 ) for the ‘memory function’, are somehow arbitrary, and were chosen to optimize the computational effort in the fitting procedure. An alternative choice, more computationally costly, could have been, for example, an exponentially decaying memory function after activation ( m ( t , t ′ ) = exp [ − ( t − t ′ ) / τ ] ). However, the very existence of a quenching mechanism with a short characteristic time is crucial to obtain a reliable fit to the experimental data. Note that the quenching time is much smaller than the other time scales in the model. Therefore, we argue that any memory function with a unique short characteristic time would fit the experimental data equally well. On the other hand, the interpretation of the memory function discussed in §3.2 cannot be decided based solely on our fitting. Accordingly, the quenching behaviour could either be the same or vary across different cells in the colony. Equation ( 3.6 ) implies the former possibility, whereas the latter would correspond to an exponentially decaying memory function that describes a reduction in the number of bioluminescent cells. To discriminate between the two scenarios remains an interesting open question. From a biological perspective, the very small value of the ‘quenching time’ τ makes a quenching mechanism based on luciferase turnover unlikely. On the other hand, it is intriguing to hypothesize that the bioluminescence quenching that we observe here could be due to a switch in the type of luciferase activity, driven by the high amount of reactive oxygen species in the highly drying environment where the experiments were carried out [ 30 ]. As a matter of fact, it had been previously reported that V. harveyi luciferase, but not necessarily the process of light emission, may be involved in the detoxification of reactive oxygen species, thus playing a role in the protection of cells against oxidative stress [ 32 ]. This role had also been suggested to be at the origin of the development of marine bioluminescence [ 31 ]."
} | 3,929 |
24385858 | PMC3873186 | pmc | 3,538 | {
"abstract": "Lignin is a major cell wall component of vascular plants that provides mechanical strength and hydrophobicity to vascular vessels. However, the presence of lignin limits the effective use of crop straw in many agroindustrial processes. Here, we generated transgenic maize plants in which the expression of a lignin biosynthetic gene encoding CCoAOMT, a key enzyme involved in the lignin biosynthesis pathway was downregulated by RNA interference (RNAi). RNAi of CCoAOMT led to significantly downregulated expression of this gene in transgenic maize compared with WT plants. These transgenic plants exhibited a 22.4% decrease in Klason lignin content and a 23.3% increase in cellulose content compared with WT plants, which may reflect compensatory regulation of lignin and cellulose deposition. We also measured the lignin monomer composition of the RNAi plants by GC-MS and determined that transgenic plants had a 57.08% higher S/G ratio than WT plants. In addition, histological staining of lignin with Wiesner reagent produced slightly more coloration in the xylem and sclerenchyma than WT plants. These results provide a foundation for breeding maize with low-lignin content and reveal novel insights about lignin regulation via genetic manipulation of CCoAOMT expression.",
"introduction": "Introduction Lignin is a major component of plant cell walls and accounts for approximately 30% of the organic carbon in the biosphere ( Boerjan et al. , 2003 ). However, lignin in the plant cell walls hinders many agroindustrial processes, such as paper manufacturing and cellulosic biofuel production. Thus, many studies have focused on the repression or alteration of lignin biosynthesis to permit more efficient utilization of plant cell walls ( Hu et al. , 1999 ; Blee et al. , 2001 ). Lignins from angiosperms are mainly polymerized from three cinnamyl alcohols (also called monolignols), including p -coumaryl, coniferyl and sinapyl alcohols, which form hydroxyphenyl (H), guaiacyl (G) and syringyl (S) lignin, respectively. Accordingly, lignin biosynthesis in plants comprises two major steps, including monolignol biosynthesis and the subsequent crosslinking of lignin monomers to form different polymers ( Boudet et al. , 2003 ). Considerable effort has been directed towards understanding the mechanisms of monolignol biosynthesis. The biochemical pathways of monolignol biosynthesis are highly conserved throughout vascular plants, and enzymes involved in these biosynthetic pathways have been isolated and characterized. Recent studies have shown that at least 10 enzymes are required for monolignol biosynthesis, including phenylalanine ammonia-lyase (PAL), cinnamic acid 4-hydroxylase (C4H), 4-coumarate CoA ligase (4CL), p-coumarate 3-hydroxylase (C3H), p-hydro-xycinnamoyl-CoA (HCT), caffeoyl-CoA O-methyltransferase (CCoAOMT), hydroxycinnamoyl-CoA reductase (CCR), cinnamyl alcohol dehydrogenase (CAD), ferulate 5-hydroxylase (F5H) and caffeic acid/5-hydroxyferulic acid O-methyltransferase (COMT) ( Boudet et al. , 2003 ). Lignin biosynthesis begins with the amino acid phenylalanine. PAL is one of the most intensively studied enzymes in plant secondary metabolism due to the key role that this enzyme plays in catalyzing the deamination reaction to produce cinnamic acid, which is then converted into p -coumaric acid by C4H. The downregulation of PAL and C4H gene expression in transgenic tobacco lead to a significant reduction in lignin content, which is consistent with the crucial roles of PAL and C4H in phenylpropanoid biosynthesis ( Bate et al. , 1994 ; Elkind et al. , 1990 ; Sewalt, et al. , 1997 ). In addition, studies have also shown that PAL is mainly responsible for the biosynthesis of G-lignin, while the downregulation of PAL expression in plants mainly leads to a reduction in the G units of lignin, whereas the downregulation C4H expression mainly leads to a reduction in S units of lignin. In addition, many studies have shown that lignin content can by altered by modifying the expression of other key enzymes in the lignin biosynthesis pathway. For example, increasing evidence suggested that CCoAOMT is involved in a parallel pathway for lignin monomer formation ( Ye et al. , 1994 ; Zhong et al. , 1998 ). The downregulation of CCoAOMT results in a reduction in lignin production, along with an increase in the S/G ratio due to a reduction in G units ( Ye et al. , 1994 ; Guo et al. , 2001 ; Pincon et al. , 2001 ; Boerjan et al. , 2003 ). Studies from transgenic plants have shown that the downregulation of COMT expression leads the reduced production of S units, which suggests that COMT is mainly responsible for the biosynthesis of S monomers ( Tsai et al. , 1998 ; Lapierre et al. , 1999 ; Doorsselaere et al. , 2003 ). Maize ( Zea mays L. ) straw is one of the most important leading forage crops. However, maize straw contains considerable amounts of lignin, which seriously affects the digestion and nutrient absorption of maize straw by livestock. Moreover, high lignin content in plant cell walls has a negative impact on forage quality ( Marita et al. , 2003 ). As mentioned above, many studies have examined the effect of enzymes at key positions in the monolignol biosynthesis pathway such as CCoAOMT and COMT. In this study, a 229 bp fragment corresponding to the fifth exon of maize CCoAOMT was generated by PCR amplification to construct RNA interference (RNAi) expression vector, which was transferred into maize by Agrobacterium -mediated transformation. The results showed that the repressed expression of CCoAOM T in maize largely reduces the lignin content of maize straw and significantly increases the S/G ratio and cellulose content. The results of this study reveal new details about the effects of the downregulation of CCoAOMT on lignin regulation and provide a basis for further studies aimed at breeding plants with low lignin content.",
"discussion": "Discussion Plants with low lignin content can be obtained through breeding and selection, or by purposely altering lignin content through genetic engineering. Since lignin biosynthesis pathways are highly conserved in most plant species ( Boerjan et al. , 2003 ; Umezawa, 2010 ), genetic engineering can be employed to directly manipulate the target genes involved in the lignin biosynthesis pathway by down-regulating the expression of these genes, or by manipulating transcription factors that regulate the expression of key lignin synthesis gene(s) ( Sánchez et al. , 2005 ; Zhou et al. , 2009 ; Xu et al. , 2011 ;). To date, at least ten genes encoding the key enzymes in the lignin biosynthesis pathway have been isolated and characterized. Various enzymes in the lignin biosynthesis pathway have been shown to play important roles in regulating lignin content and quality. Amongst the genes encoding these enzymes, CCoAOMT was discovered only recently, and has mainly been studied in tobacco, poplar and Medicago ( Martz et al. , 1998 ; Meyermans et al. , 2000 ; Guo et al. , 2001 ). Despite the importance of maize for forage and industrial materials, few studies have focused on the functional characterization of lignin synthesis-related genes in maize. In this study, a 229 bp fragment of maize CCoAOMT corresponding to the fifth exon of this gene was used to construct an RNAi vector, which was then transferred into maize. The results of this study confirmed many features of CCoAOMT that were previously reported, but also revealed some new aspects of this gene that differ from those observed in other species. In transgenic poplar and alfalfa with repressed CCoAOMT activity, down-regulation of CCoAOMT leads to a general reduction in Klason lignin content but an increase in the S/G ratio ( Meyermans et al. , 2000 ; Zhong et al. , 2000 ; Guo et al. , 2001 ). In addition, transgenic studies have indicated that the increase in the S/G ratio results mainly from a decrease in the amount of G units, with no reduction in S lignin. Our results showed that down-regulation of CCoAOMT in maize can largely reduce the Klason lignin content and significantly increase the S/G ratio, which are consistent with the previous studies. Meyermans et al. (2000) showed that introducing the antisense of CCoAOMT into transgenic poplar resulted in an 11% increase in the S/G ratio. By contrast, our results showed that the S/G ratio in plants with downregulated CCoAOMT expression was 57.08% higher than that of WT plant, possibly due to the different plant species or methods used to repress gene expression in these two studies. More interestingly, transgenic plants showed a concomitant increase in cellulose content in the lignin-reduced transgenic plants, suggesting that lignin and cellulose deposition may be regulated in a compensatory fashion in maize, which may contribute to metabolic flexibility and a growth advantage to help sustain mechanical strength in lignin-deficient straw. However, the mechanism by which this compensatory phenomenon in lignin and cellulose deposition occurs is still unknown. To obtain plants with reduced lignin, studies should focus not only on lowing lignin content and composition, but also on the selection of transgenic plants with normal development. Many studies have focused on altering the expression of various enzymes in the lignin biosynthesis pathway, some of which have had important effects on plant growth and development. For example, down-regulation of 4CL in transgenic aspen leads to reduced lignin content, but also to the production of plants with enhanced growth phenotypes such as thicker stems, longer internodes, and larger leaves than control plants ( Hu et al. , 1999 ). On the contrary, down-regulation of CCR in Leucaena leucocephala produced plants with significant changes in phenotypes, such as stunted growth and development, wrinkled leaves, and delayed senescence, with a 24.7% reduction of Klason lignin vs. WT plants ( Prashant et al. , 2011 ). In the current study, transgenic maize plants with downregulated CCoAOMT expression exhibited significant reductions in Klason lignin, but displayed a normal phenotype compared with control plants, besides a slightly delayed growth, suggesting that CCoAOMT is an effective and ideal target gene for the regulation of the lignin biosynthesis pathway. However, there is an inherent relationship between lignin content and plant lodging resistance. Although our results show that plants with down-regulated CCoAOMT expression had almost normal phenotypes, except the slightly delayed growth, we did not adequately address the effect of reduced lignin content on lodging resistance in these lines, due to the small number of independent transgenic lines obtained in this study. Thus, more transgenic lines will be required to examine the effects of reduced lignin content on lodging resistance in maize in a future study. In addition, lignin deposition is influenced by abiotic and biotic stresses ( Halpin, 2004 ; Boudet, 2007 ), it will be interesting to compare the lignin content of CCoAOMT down-regulated plants under field and greenhouse conditions."
} | 2,793 |
37239950 | PMC10218252 | pmc | 3,539 | {
"abstract": "The extraordinary potential of hydrogen as a clean and sustainable fuel has sparked the interest of the scientific community to find environmentally friendly methods for its production. Biological catalysts are the most attractive solution, as they usually operate under mild conditions and do not produce carbon-containing byproducts. Hydrogenases promote reversible proton reduction to hydrogen in a variety of anoxic bacteria and algae, displaying unparallel catalytic performances. Attempts to use these sophisticated enzymes in scalable hydrogen production have been hampered by limitations associated with their production and stability. Inspired by nature, significant efforts have been made in the development of artificial systems able to promote the hydrogen evolution reaction, via either electrochemical or light-driven catalysis. Starting from small-molecule coordination compounds, peptide- and protein-based architectures have been constructed around the catalytic center with the aim of reproducing hydrogenase function into robust, efficient, and cost-effective catalysts. In this review, we first provide an overview of the structural and functional properties of hydrogenases, along with their integration in devices for hydrogen and energy production. Then, we describe the most recent advances in the development of homogeneous hydrogen evolution catalysts envisioned to mimic hydrogenases.",
"conclusion": "6. Conclusions and Future Perspectives In the last few decades, research interest toward hydrogen production has greatly exploded, prompted by the urgency to face the current energy crisis. Hydrogen has a huge potential as a clean fuel, due to its incredibly high energy density and the possibility to be combusted without generating CO 2 . However, the development of cost-effective and environmentally friendly methods for hydrogen production is far from trivial, and much effort has been devoted to the elaboration of enzyme-based devices. These typically involve the use of hydrogenases either for light-driven hydrogen production (photoelectrochemical cells) or its conversion into electricity, upon matching of hydrogen oxidation with a reductive chemical reaction (biofuel cells). Despite hydrogenases showing impressive catalytic properties, their application in large-scale processes is currently hampered by difficulties with their expression and their limited tolerance to ambient oxygen. Extensive research has been conducted to elucidate the catalytic mechanism of hydrogenases, with the aim of replicating their function into artificial and tailorable systems. Since the late 1990s, a plethora of molecular catalysts have been proposed as hydrogenase mimics. While some have focused on reproducing the structure of the [FeFe]-hydrogenase active site, a wide variety of HECs have also been developed, combining redox non-innocent ligands with nickel or cobalt ions. Considerable improvements in catalytic properties have been accomplished through the incorporation of secondary coordination sphere interactions and, particularly, proton shuttle functionalities. Even though remarkably high catalytic performances have been achieved in some cases, the majority of these complexes only function in organic solvents and in the presence of strong acids, further highlighting the essential role of a bioinspired ligand framework in imparting enzyme-like properties. Incorporation of catalytic centers into natural or designed protein scaffolds conceivably represents the most promising strategy to obtain artificial hydrogenases capable of paralleling their natural counterparts. Indeed, protein design methods allow precisely controlling the dielectric properties and the interactions within the residues in the active site, enabling a fine modulation of the metal redox properties. Furthermore, protein scaffolds can also serve as a platform to bind together catalytic and light-harvesting units, as reported in prominent examples discussed within the review. In summary, biohybrid and bioinspired catalysts are promising candidates for sustainable hydrogen and energy production; nevertheless, some important challenges still need to be addressed. In particular, most artificial biomolecular catalysts are still fragile and display limited durability under operative conditions, especially in light-driven catalysis. One possible approach to overcome this problem is the immobilization of catalysts onto nanomaterials, which could extend their durability, allowing for easy catalyst recycling. Furthermore, as the majority of HECs reported so far contain transition metal ions such as nickel and cobalt, evaluation of their toxicity is a crucial aspect to consider, especially when scaling up the process to the industrial scale [ 277 ]. Indeed, the toxicity of metal complexes greatly depends on the nature of the ligands; thus, it is not trivial to predict [ 278 ]. A step further in the development of sustainable methods for hydrogen production would consist of the construction of completely protein or peptide-based materials, opening the way for the assembly of biosynthetic nanoreactors for energy-related catalysis."
} | 1,285 |
39537773 | PMC11560975 | pmc | 3,540 | {
"abstract": "A chemical discrimination system based on photonic reservoir computing is demonstrated experimentally for the first time. The system is inspired by the way humans perceive and process visual sensory information. The electro-optical reservoir computing system is a photonic analogue of the human nervous system with the read-out layer acting as the ‘brain’, and the sensor that of the human eye. A task-specific optimisation of the system is implemented, and the performance of the system for the discrimination between three chemicals is presented. The results are compared to the previously published numerical simulation (Anufriev et al. in Opt Mater Express 12:1767–1783, 2022, 10.1364/OME.449036). This publication provides a feasibility assessment and a demonstration of a practical realisation of photonic reservoir computing for a new neuromorphic sensing system - the next generation sensor with a built-in ‘intelligence’ which can be trained to ‘understand’ and to make a real time sensing decision based on the training data.",
"conclusion": "Conclusions This paper has demonstrated for the first time an experimental implementation of a neuromorphic sensing system. This system consists of four distinct parts, namely the control layer, the sensing and pre-processing layer, the EORC kernel, and the post-processing and read-out layer. A methodology has been demonstrated for the automatic optimisation of the sensing system for the chemical classification task. The performance of the system has been evaluated for a group of three aliphatic alcohols and a group of three essential oils. It has been shown that using a thresholding limit of just ± 10%, with a training set of as low as 48 essential oils, yields a classification success of 94%; a perfect classification was also achieved for aliphatic alcohols with a training set of 72 samples. The bifurcation of the system was studied and validated by the numerical simulated results. Finally, the stability of the system was studied, and a range of operational parameters suggested for optimal stability.",
"introduction": "Introduction The human nervous system possesses remarkable computational abilities 1 , 2 . It is an incredibly powerful biological computer capable of performing pattern recognition, regression and forecasting on massively parallel information in real time 3 , 4 . Inspired by this, artificial neural networks (ANNs) aim to replicate the neural functions of humans as a computational framework. Such systems are called ‘neuromorphic’, as they are inspired by the computing architecture of the human nervous system and brain. ANNs and neuromorphic systems have shown effectiveness for tasks, like pattern recognition and inference, and found applications in bioinformatics, medical image processing, stock market forecasting, and telecom signal recognition 5 – 15 . Numerous architectures, implementations, and applications of ANNs have been demonstrated – both in software and as hardware systems 16 – 18 . Among those, photonic implementations have been demonstrated to be suitable for high-speed processing due to the larger bandwidths offered by optical signals and components 18 , 19 . Photonic reservoir computing (PhRC) is one such implementation. This offers an alternative approach to the architecture and functionality of photonic neural networks 20 – 22 , which are based on the conventional feed-forward neural networks. In reservoir computing, which was first demonstrated as a software implementation, training is exclusively carried out at the read-out layer, allowing the kernel to remain semi-random and untrained 18 , 20 – 22 . The reservoir kernel (see Fig. 1 (b)) performs complex temporal dynamics and nonlinear transformations which processes input data, remapping it to a new higher-dimensional representation space 23 – 25 . This higher-dimension representations allows for a final linear discrimination to be performed by the read-out layer. PhRC systems have been demonstrated in an optical-fibre setup, including electro-optical feedback 24 , 25 , all-optical feedback 26 , 27 and all-optical stimulated Brillouin scattering systems 28 , in a chip-scale integrated photonic devices, including as network of complex interconnected waveguides 29 , stochastic photonic field 30 , 31 , and lasers 17 , 32 . Using these PhRC platforms, applications for telecommunications, quantum computing, and chaotic time series generation and prediction have already been reported 20 , 21 . In the present work, we considered the practical implementation of electro-optical reservoir computing (EORC), with an optical fibre delay-line and Mach-Zehnder modulator, which provide the memory and nonlinear effects respectively 24 , 25 , 33 . The work presented here focuses on the first experimental demonstration of PhRC as neuromorphic chemical sensing system - the next generation of sensing with built-in intelligence. Such a system can be trained to make real time sensing decisions based on training data - inspired by the fact that the human computing capacity is predominantly used for processing of sensory information 34 . The sensory part of the neuromorphic sensing system was implemented in the infrared, mimicking the properties of human eyes, namely the discrete and broadband response of the cone cells in the retina. The application of this neuromorphic sensory system was demonstrated to the discrimination of different chemicals. This paper is structured as follows: Sect. \" Methods \" describes the neuromorphic sensory system. This starts by outlining the experimental operation of the pyroelectric sensing apparatus, describing the experimental implementation of the EORC, and the pre- and post-processing of data that was carried out. Section Results and discussion presents and discusses the results obtained and their comparison to the earlier work 33 . Section \" Results and discussion \", further, reports the dynamics of the EORC kernel, before presenting the results on stability and performance for the chemical discrimination tasks. The impact of the optimisation parameters on system performance is also briefly discussed. The conclusions are provided in Sect. Conclusions .",
"discussion": "Results and discussion This Section starts with the characterisation of the EORC with no sensory input present. Section \" Chemical discrimination by the neuromorphic sensing system \" demonstrates the application of the neuromorphic sensing system to the discrimination of chemicals, describes the performance optimisation of the EORC, and the system stability. EORC operation states and bifurcation The EORC used in the present work was based on a delay feedback system 24 , 25 , 33 . Such a system features parameter dependent behaviour and was characterised by the bifurcation diagram. System bifurcation was achieved in our setup by varying input laser power ( \\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}$$\\:P$$\\end{document} ) and the gain of the RF-amplifier ( \\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} ), labelled (1) and (14) in Fig. 1 (c) respectively. First, consider Fig. 5 (b), which depicts the bifurcation of the monitored optical signal at the read-out photodetector (labelled (6) in Fig. 1 (c)). To illustrate this bifurcation phenomenon, the states of the optical signal, i.e., 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}$$\\:\\text{m}\\text{a}\\text{x}\\left[x\\left(t\\right)\\right]-\\text{m}\\text{i}\\text{n}\\left[x\\left(t\\right)\\right]$$\\end{document} , have been plotted. For a fixed input laser power 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}$$\\:{P}_{0} = 55\\:\\text{m}\\text{W}$$\\end{document} , and low amplification, the states of the optical signal remained constant (single-valued), 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}$$\\:\\text{m}\\text{a}\\text{x}\\left[x\\left(t\\right)\\right]\\approx\\:\\text{m}\\text{i}\\text{n}\\left[x\\left(t\\right)\\right]$$\\end{document} , as the gain of the RF-amplifier, \\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} , increased. This is because, although the gain \\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} increased, the optical power provided by the laser module remained constant. In the single-valued region, the small fluctuation observed was caused by random noise, which was observed to be around 0.2 V. However, after a specific RF-amplifier gain 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}$$\\:\\gamma\\:=23.4\\:\\text{d}\\text{B}$$\\end{document} , the states of the optical signal distinctly split denoting the bifurcation. Figure 5 (c) shows the bifurcation of the optical signal as a function of the input laser power ( \\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}$$\\:P$$\\end{document} ) for a fixed RF-amplifier gain \\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}$$\\:\\gamma = 25\\:\\text{d}\\text{B}$$\\end{document} . Figure 5 (c) demonstrates a similar bifurcation behaviour, but with an overall positive gradient due to the increase of overall input laser power. To further illustrate the bifurcation phenomenon, examples of the temporal signals of the system are shown in Fig. 5 (d)–(f) for the specific input laser powers ( \\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}$$\\:P$$\\end{document} ) marked in Fig. 5 (c). To gain an overall picture of the bifurcation phenomenon, the surface plot in Fig. 5 (a) shows the state of the optical signal as a function of both the input laser power ( \\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}$$\\:P$$\\end{document} ) and the RF-amplifier gain ( \\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} ). Figure 5 (a) shows two distinct regions: the region where the system response was single-valued (dark blue) and the region where the system response was oscillating (not-blue). This bifurcation behaviour, single-valued for both low input laser power ( \\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}$$\\:P$$\\end{document} ) and the RF-amplifier gain ( \\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} ); and oscillating for high input laser power ( \\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}$$\\:P$$\\end{document} ) and the RF-amplifier gain ( \\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} ), is consistent with previously published simulation work 33 . The transition between the two states, the bifurcation line, is indicated by the light blue boundary between dark blue and green colours in Fig. 5 (a). Furthermore, the specific parameters used for Fig. 5 (b),(c), namely \\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}$$\\:{P}_{0} = 55\\:\\text{m}\\text{W}$$\\end{document} 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}$$\\:\\gamma = 25\\:\\text{d}\\text{B}$$\\end{document} , are marked in Fig. 5 (a) by dashed white lines. The method used in characterising the different regions of operation was a two-parameter sweep across laser power \\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}$$\\:\\left(P\\right)$$\\end{document} and RF-amplifier gain \\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}$$\\:\\left(\\gamma\\:\\right)$$\\end{document} . This only allowed the location of the first order bifurcation points (the first instance of the system response becoming multi-valued). It is possible that higher order bifurcations exist, but they were not considered in this paper. \n Fig. 5 Operational states of EORC kernel. ( a ) The range of reservoir activation states, \\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}$$\\:x\\left(t\\right)$$\\end{document} , observed at the read-out photodetector as function of EORC parameters: laser power \\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}$$\\:P$$\\end{document} and RF-amplifier gain \\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} . The single-valued system response is shown in dark blue and the region where the system response was oscillating is shown with other colours. The colour bar depicts the difference in Volts between the maximum value of x(t) and the minimum value of x(t) (i.e. the states of the signal) observed on the oscilloscope. The dashed lines mark the specific input laser power \\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}$$\\:{P}_{0}=55\\:\\text{m}\\text{W}$$\\end{document} and RF-amplifier gain \\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}$$\\:\\gamma\\:=25\\:\\text{d}\\text{B}$$\\end{document} , for the bifurcation diagram ( b ) and ( c ), respectively. The star marks the optimum operation region obtained using particle swarm optimisation for the discrimination of a group of aliphatic alcohols. The hexagon marks the optimum operation region obtained using particle swarm optimisation for the discrimination of a group of essential oils (see Sect. \" Chemical discrimination by the neuromorphic sensing system \"). ( d – f ) depict the temporal signals \\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}$$\\:x\\left(t\\right)$$\\end{document} for the specific operation parameters at amplifier gain \\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}$$\\:\\gamma\\:=25\\:\\text{d}\\text{B}$$\\end{document} , \\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}$$\\:P=20\\:\\text{m}\\text{W},50\\:\\text{m}\\text{W},$$\\end{document} 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}$$\\:90\\:\\text{m}\\text{W}$$\\end{document} as marked in ( c ). \n Chemical discrimination by the neuromorphic sensing system A group of three aliphatic alcohols, i.e. ethanol, methanol, and isopropanol, was first used for chemical discrimination to demonstrate the accuracy of the trained system, allowing for direct comparison with previously published simulation work 33 . Furthermore, to show its universal application of handling other chemical samples, the sensing system was then trained to classify a group of essential oils - eucalyptus, lavender, and rapeseed oils. A dataset of 90 spectral responses was used during the discrimination of the group of aliphatic alcohols (30 spectra for each type of alcohol), and 60 spectra during the discrimination of the group of essential oils (20 spectra for each type of oil). In both cases, 80% of the spectra available were used for training and 20% for testing. As described in Sect. \" Post-processing: time-demultiplexing, training and testing at the read-out layer \", one-hot encoding was used to represent the chemical sample-under-test, here, it is defined as in Table 1 . \n Table 1 One-hot-encoding for discrimination of groups of chemicals. Discrimination of: Aliphatic alcohols Essential oils \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}$${\\text{Y}}_{{{\\text{target}}}} = \\left\\{ {\\begin{array}{*{20}c} {[1,0,0]} \\\\ {[0,1,0]} \\\\ {[0,0,1]} \\\\ \\end{array} } \\right.$$\\end{document} \n \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}$$\\:\\text{E}\\text{t}\\text{h}\\text{a}\\text{n}\\text{o}\\text{l}$$\\end{document} \n \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}$$\\:\\text{E}\\text{u}\\text{c}\\text{a}\\text{l}\\text{y}\\text{p}\\text{t}\\text{u}\\text{s}\\:\\text{o}\\text{i}\\text{l}$$\\end{document} \n \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}$$\\:\\text{M}\\text{e}\\text{t}\\text{h}\\text{a}\\text{n}\\text{o}\\text{l}$$\\end{document} \n \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}$$\\:\\text{R}\\text{a}\\text{p}\\text{e}\\text{s}\\text{e}\\text{e}\\text{d}\\:\\text{o}\\text{i}\\text{l}$$\\end{document} \n \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}$$\\:\\text{I}\\text{s}\\text{o}\\text{p}\\text{r}\\text{o}\\text{p}\\text{a}\\text{n}\\text{o}\\text{l}$$\\end{document} \n \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}$$\\:\\text{L}\\text{a}\\text{v}\\text{e}\\text{n}\\text{d}\\text{e}\\text{r}\\:\\text{o}\\text{i}\\text{l}\\:\\:\\:\\:$$\\end{document} \n \n Particle swarm optimisation is a very common optimisation used in engineering 40 , 41 . The details of the particle swarm optimisation approach can be found in the documentation for the particle optimisation toolbox 38 . Heuristic optimisation using the particle swarm method was employed to find the optimum EORC operational parameters for accurate sample classification. This approach yielded a minimised the NMSE between the target ( \\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}$$\\:{\\mathbf{Y}}_{\\text{t}\\text{a}\\text{r}\\text{g}\\text{e}\\text{t}}$$\\end{document} ) and the predicted ( \\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}$$\\:{\\mathbf{Y}}_{\\text{p}\\text{r}\\text{e}\\text{d}\\text{i}\\text{c}\\text{t}\\text{i}\\text{o}\\text{n}}$$\\end{document} ) signals. The particle swarm optimisation for this minimisation was set to use 200 particles at each iteration. At each iteration, the set of EORC operation parameters, \\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}$$\\:P$$\\end{document} 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}$$\\:\\gamma\\:$$\\end{document} , which yielded the lowest value of NMSE was recorded. For the aliphatic alcohol discrimination, the optimisation concluded after 50 iterations (Fig. 6 (a)) and after 20 iterations for the discrimination of essential oils (Fig. 6 (b)). The EORC operation parameters, \\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}$$\\:P$$\\end{document} 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}$$\\:\\gamma\\:$$\\end{document} , that corresponded to the lowest recorded NMSE are marked in Fig. 5 (a) by a star for discrimination within a group of aliphatic alcohols and a hexagon for the discrimination within a group of essential oils. The NMSE for classification of chemicals in each group at optimal parameters is shown in Fig. 5 (a), \\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}$$\\:{N}_{x} = 50$$\\end{document} mask values were used for this task. It is noted that the optimum operation point for aliphatic alcohol discrimination was not in line with the previously carried out simulation 33 , where it was shown to be near the bifurcation points. We believe that this is due to our use here of the particle swarming method, which is based on stochastic optimisation, converging to a local minimum, in contrast to the exhaustive search conducted in the simulation work 33 . This suggests that further system performance improvements could be possible. \n Fig. 6 EORC operational parameter optimisation by the particle swarm method. Averaged classification NMSE with each particle swarm optimisation iteration for: ( a ) the group of aliphatic alcohols and ( b ) the group of essential oils. \n Figure 7 (a) and (b) show a bar chart of the predicted outputs, \\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}$$\\:{\\mathbf{Y}}_{\\text{p}\\text{r}\\text{e}\\text{d}\\text{i}\\text{c}\\text{t}\\text{i}\\text{o}\\text{n}}$$\\end{document} , for the chemical discrimination tasks for a group of aliphatic alcohols and a group of essential oil respectively. Distinct discriminations among testing samples have been achieved with less than ± 10% error standard deviation as is shown by the error bars in Fig. 7 . Using ± 10% as a thresholding condition, a 100% classification accuracy was achieved within the group of alcohols and a classification accuracy of 94% was achieved within the group of essential oils. The confusion matrices for these classifications are shown in Fig. 8 (a) and (b) for the group of aliphatic alcohols and the groups of essential oils respectively. No testing samples were misclassified by the optimised neuromorphic sensing system, and only a single sample was unclassified due to a \\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}$$\\:{\\mathbf{Y}}_{\\text{p}\\text{r}\\text{e}\\text{d}\\text{i}\\text{c}\\text{t}\\text{i}\\text{o}\\text{n}}$$\\end{document} value outside the thresholding limits. \n Fig. 7 ( a ) and ( b ) depict the discrimination results for groups of aliphatic alcohols and essential oils, respectively. Within this classification ± 10% thresholding limits were applied in order to judge the system classification quality. \n \n Fig. 8 Confusion matrices of the discrimination results for neuromorphic sensing of ( a ) the group of aliphatic alcohols and ( b ) the group of essential oils. \n Here, we also report the observation that the neuromorphic sensing system developed here was influenced by ambient changes of the environment, including noise (electrical noise, thermal noise, acoustic and mechanical vibrations) and thermal component drifts (of the laser and MZM). To exemplify this, Fig. 9 presents a histogram of the NMSE values for 100 independent instances of classification for the group of aliphatic alcohols, carried out using a fixed set of optimised EORC parameters, \\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}$$\\:P$$\\end{document} 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}$$\\:\\gamma\\:$$\\end{document} (see Fig. 5 (a)), at randomly selected times over the duration of a week, with the same training and testing datasets used throughout. The NMSE mean for an optimised neuromorphic sensing system applied to classification of the group of aliphatic alcohols was 0.0148 and had an error standard deviation \\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}$$\\:\\sigma\\:=\\:8.73\\times\\:{10}^{-4}$$\\end{document} . We believe that the classification robustness of the neuromorphic sensing system could be significantly improved by using low-noise photodetectors and RF-amplifiers, performing signal averaging to enhance the signal-to-noise ratio of the signal at the readout layer, \\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}$$\\:x\\left(t\\right)$$\\end{document} , and by employing a more stable (less jittering) control system unit for the VOA and RF-amplifier. \n Fig. 9 A histogram depicting the NMSE for 100 runs of aliphatic alcohol discrimination with the constant optimised parameters. \n Furthermore, we also investigated the impact of the number of masks, \\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}$$\\:{N}_{x}$$\\end{document} , used to multiplex the input signal, \\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}$$\\:u\\left(t\\right)$$\\end{document} , on the system performance, NMSE . Figure 10 shows the NMSE for the discrimination of the group of aliphatic alcohols for various numbers of masks, \\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}$$\\:{N}_{x}$$\\end{document} , and the error bars depict the standard deviation from the mean values. It confirms that a higher number of masks, \\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}$$\\:{N}_{x}$$\\end{document} , improved the accuracy of the neuromorphic sensing system as suggested by previous publications 24 , 25 , 33 . The available number of masks was limited by equipment employed in the experiment and a steady increase in the performance of the neuromorphic sensing system is observed until \\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}$$\\:{N}_{x}=50$$\\end{document} . Better performance may be possible, however the standard deviation of the NMSE increases after \\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}$$\\:{N}_{x}=50$$\\end{document} , likely due to the ability of the AWG to resolve individual masks becoming increasingly compromised. The rate of increase of system performance as a function 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}$$\\:{N}_{x}$$\\end{document} until \\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}$$\\:{N}_{x}=50$$\\end{document} was observed to be in agreement with the numerical analysis of the same system in previously published simulation work 33 . \n Fig. 10 The NMSE for the classification of the group of aliphatic alcohols as a function of the number of masks, \\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}$$\\:{N}_{x}$$\\end{document} . The error bars denote the standard deviation of the NMSE over 10 runs from the mean values and the solid line connects the mean values of NMSE for each value 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}$$\\:{N}_{x}$$\\end{document} ."
} | 9,221 |
29652587 | PMC6118408 | pmc | 3,541 | {
"abstract": "A vast majority of computation in the brain is performed by spiking neural networks. Despite the ubiquity of such spiking, we currently lack an understanding of how biological spiking neural circuits learn and compute in vivo, as well as how we can instantiate such capabilities in artificial spiking circuits in silico. Here we revisit the problem of supervised learning in temporally coding multilayer spiking neural networks. First, by using a surrogate gradient approach, we derive SuperSpike, a nonlinear voltage-based three-factor learning rule capable of training multilayer networks of deterministic integrate-and-fire neurons to perform nonlinear computations on spatiotemporal spike patterns. Second, inspired by recent results on feedback alignment, we compare the performance of our learning rule under different credit assignment strategies for propagating output errors to hidden units. Specifically, we test uniform, symmetric, and random feedback, finding that simpler tasks can be solved with any type of feedback, while more complex tasks require symmetric feedback. In summary, our results open the door to obtaining a better scientific understanding of learning and computation in spiking neural networks by advancing our ability to train them to solve nonlinear problems involving transformations between different spatiotemporal spike time patterns.",
"introduction": "1 Introduction Neurons in biological circuits form intricate networks in which the primary mode of communication occurs through spikes. The theoretical basis for how such networks are sculpted by experience to give rise to emergent computations remains poorly understood. Consequently, building meaningful spiking models of brain-like neural networks in silico is a largely unsolved problem. In contrast, the field of deep learning has made remarkable progress in building nonspiking convolutional networks that often achieve human-level performance at solving difficult tasks (Schmidhuber, 2015 ; LeCun, Bengio, & Hinton, 2015 ). Although the details of how these artificial rate–based networks are trained may arguably be different from how the brain learns, several studies have begun to draw interesting parallels between the internal representations formed by deep neural networks and the recorded activity from different brain regions (Yamins et al., 2014 ; McClure & Kriegeskorte, 2016 ; McIntosh, Maheswaranathan, Nayebi, Ganguli, & Baccus, 2016 ; Marblestone, Wayne, & Kording, 2016 ). A major impediment to deriving a similar comparison at the spiking level is that we currently lack efficient ways of training spiking neural network (SNNs), thereby limiting their applications to mostly small toy problems that do not fundamentally involve spatiotemporal spike time computations. For instance, only recently have some groups begun to train SNNs on data sets such as MNIST (Diehl & Cook, 2015 ; Guerguiev, Lillicrap, & Richards, 2017 ; Neftci, Augustine, Paul, & Detorakis, 2016 ; Petrovici et al., 2017 ), whereas most previous studies have used smaller artificial data sets. The difficulty in simulating and training SNNs originates from multiple factors. First, time is an indispensable component of the functional form of a SNN, as even individual stimuli and their associated outputs are spatiotemporal spike patterns rather than simple spatial activation vectors. This fundamental difference necessitates the use of different cost functions from the ones commonly encountered in deep learning. Second, most spiking neuron models are inherently nondifferentiable at spike time, and the derivative of their output with respect to synaptic weights is zero at all other times. Third, the intrinsic self-memory of most spiking neurons introduced by the spike reset is difficult to treat analytically. Finally, credit assignment in hidden layers is problematic for two reasons: (1) it is technically challenging because efficient autodifferentiation tools are not available for most event-based spiking neural network frameworks, and (2) the method of weight updates implemented by the standard backpropagation of error algorithm (Backprop) is thought to be biologically implausible (Grossberg, 1987 ; Crick, 1989 ). Several studies of multilayer networks that build on the notion of feedback alignment (Lillicrap, Cownden, Tweed, & Akerman, 2016 ) have recently illustrated that the strict requirements imposed on the feedback by backpropagation of error signals can be loosened substantially without a large loss of performance on standard benchmarks like MNIST (Lillicrap et al., 2016 ; Guergiuev, Lillicrap, & Richards, 2016 ; Neftci et al., 2016 ; Baldi, Sadowski, & Lu, 2016 ; Liao & Carneiro, 2015 ). While some of these studies have been performed using spiking networks, they still use effectively a rate-based approach in which a given input activity vector is interpreted as the firing rate of a set of input neurons (Eliasmith et al., 2012 ; Diehl & Cook, 2015 ; Guergiuev et al., 2016 ; Neftci et al., 2016 ; Mesnard, Gerstner, & Brea, 2016 ). While this approach is appealing because it can often be related directly to equivalent rate-based models with stationary neuronal transfer functions, it also largely ignores the idea that individual spike timing may carry additional information that could be crucial for efficient coding (Thalmeier, Uhlmann, Kappen, & Memmesheimer, 2016 ; Denève & Machens, 2016 ; Abbott, DePasquale, & Memmesheimer, 2016 ; Brendel, Bourdoukan, Vertechi, Machens, & Denéve, 2017 ) and fast computation (Thorpe, Fize, & Marlot, 1996 ; Gollisch & Meister, 2008 ). In this article, we develop a novel learning rule to train multilayer SNNs of deterministic leaky integrate-and-fire (LIF) neurons on tasks that fundamentally involve spatiotemporal spike pattern transformations. In doing so, we go beyond the purely spatial rate-based activation vectors prevalent in deep learning. We further study how biologically more plausible strategies for deep credit assignment across multiple layers generalize to the enhanced context of more complex spatiotemporal spike pattern transformations. 1.1 Prior Work Supervised learning of precisely timed spikes in single neurons and networks without hidden units has been studied extensively. Pfister, Toyoizumi, Barber, and Gerstner ( 2006 ) have used a probabilistic escape rate model to deal with the hard nonlinearity of the spike. Similar probabilistic approaches have also been used to derive spike timing dependent plasticity (STDP) from information-maximizing principles (Bohte & Mozer, 2007 ; Toyoizumi, Pfister, Aihara, & Gerstner, 2005 ). In contrast to that, ReSuMe (Ponulak & Kasiński, 2009 ) and SPAN (Mohemmed, Schliebs, Matsuda, & Kasabov, 2012 ) are deterministic approaches that can be seen as generalizations of the Widrow-Hoff rule to spiking neurons. In a similar vein, the Chronotron (Florian, 2012 ) learns precisely timed output spikes by minimizing the Victor-Pupura distance (Victor & Purpura, 1997 ) to a given target output spike train. Similarly, Gardner and Grüning ( 2016 ) and Albers, Westkott, and Pawelzik ( 2016 ) have studied the convergence properties of rules that reduce the van Rossum distance by gradient descent. Moreover, Memmesheimer, Rubin, Ölveczky, & Sompolinsky ( 2014 ) proposed a learning algorithm that achieves high capacity in learning long, precisely timed spike trains in single units and recurrent networks. The problem of sequence learning in recurrent neural networks has also been studied as a variational learning problem (Brea, Senn, & Pfister, 2013 ; Jimenez Rezende & Gerstner, 2014 ) and by combining adaptive control theory with heterogeneous neurons (Gilra & Gerstner, 2017 ). Supervised learning in SNNs without hidden units has also been studied for classification problems. For instance, Maass, Natschläger, and Markram ( 2002 ) have used the p-delta rule (Auer, Burgsteiner, & Maass, 2008 ) to train the readout layer of a liquid state machine. Moreover, the tempotron (Gütig & Sompolinsky, 2006 ; Gütig, 2016 ), which can be derived as a gradient-based approach (Urbanczik & Senn, 2009 ), classifies large numbers of temporally coded spike patterns without explicitly specifying a target firing time. Only a few works have embarked on the problem of training SNNs with hidden units to process precisely timed input and output spike trains by porting backprop to the spiking domain. The main analytical difficulty in these approaches arises from partial derivatives of the form where is the spike train of the hidden neuron and is a hidden weight. SpikeProp (Bohte, Kok, & La Poutre, 2002 ) sidesteps this problem by defining a differentiable expression on the firing times instead, on which standard gradient descent can be performed. While the original approach was limited to a single spike per neuron, multiple extensions of the algorithm exist, some of which also improve its convergence properties (McKennoch, Liu, & Bushnell, 2006 ; Booij & tat Nguyen, 2005 ; Shrestha & Song, 2015 , 2017 ; de Montigny & Mêsse, 2016 ; Banerjee, 2016 ). However, one caveat of such spike timing–based methods is that they cannot learn starting from a quiescent state of no spiking, as the spike time is then ill defined. Some algorithms, however, do not suffer from this limitation. For instance, an extension of ReSuMe to multiple layers was proposed (Sporea & Grüning, 2013 ) in which error signals were backpropagated linearly. More recently, the same group proposed a more principled generalization of Backprop to SNNs in Gardner, Sporea, and Grüning ( 2015 ) using a stochastic approach, which can be seen as an extension of Pfister et al. ( 2006 ) to multiple layers. In a similar flavor as Fremaux, Sprekeler, and Gerstner (2010), Gardner et al. ( 2015 ) substitute the partial derivative of hidden spike trains by a point estimate of their expectation value. Although, theoretically, stochastic approaches avoid problems arising from quiescent neurons, convergence can be slow, and the injected noise may become a major impediment to learning in practice. Instead of approximating partial derivatives of spike trains by their expectation value, in Bohte ( 2011 ), the corresponding partial derivative is approximated as a scaled Heaviside function of the membrane voltage. However, due to the use of the Heaviside function, this approach has a vanishing surrogate gradient for subthreshold activations, which limits the algorithm's applicability to cases in which hidden units are not quiescent. Finally, Huh and Sejnowski ( 2017 ) proposed another interesting approach in which instead of approximating partial derivatives for a hard spiking nonlinearity, instead a “soft” spiking threshold is used, for which by design, standard techniques of gradient descent are applicable. In contrast to this previous work, our method permits training multilayer networks of deterministic LIF neurons to solve tasks involving spatiotemporal spike pattern transformations without the need for injecting noise even when hidden units are initially completely silent. To achieve this, we approximate the partial derivative of the hidden unit outputs as the product of the filtered presynaptic spike train and a nonlinear function of the postsynaptic voltage instead of the postsynaptic spike train. In the following section, we explain the details of our approach.",
"discussion": "5 Discussion In this article, we have derived a three-factor learning rule to train deterministic multilayer SNNs of LIF neurons. Moreover, we have assessed the impact of different types of feedback credit assignment strategies for the hidden units, notably symmetric, random, and uniform. In contrast to previous work (Pfister et al., 2006 ; Fremaux et al., 2010 ; Gardner et al., 2015 ), we have used a deterministic surrogate gradient approach instead of the commonly used stochastic gradient approximations. By combining this rule with ideas of straight-through estimators (Hinton, 2012 ; Bengio et al., 2013 ) and feedback alignment (Lillicrap et al., 2016 ; Baldi et al., 2016 ), we could efficiently train and study precisely timed spiking dynamics in multilayer networks of deterministic LIF neurons without relying on the introduction of extraneous and unavoidable noise present in stochastic models, noise that generally impedes the ability to learn precise spatiotemporal spike pattern transformations. The weight update equation of SuperSpike constitutes a voltage-based nonlinear Hebbian three-factor rule with individual synaptic eligibility traces. Each of these aspects has direct biological interpretations. For instance, a nonlinear voltage dependence has been reported ubiquitously by numerous studies on Hebbian long-term plasticity induction in hippocampus and cortex (Artola, Bröcher, & Singer, 1990 ; Feldman, 2012 ). Also, the window of temporal coincidence detection in our model is in good agreement with that of STDP (Feldman, 2012 ). Moreover, the time course of the eligibility traces could be interpreted as a local calcium transient at the synaptic spine level. Finally, the multiplicative coupling of the error signal with the eligibility trace could arise from neuromodulators (Izhikevich, 2007 ; Pawlak, Wickens, Kirkwood, & Kerr, 2010 ; Frémaux & Gerstner, 2016 ; Kusmierz et al., 2017 ). However, instead of only one global feedback signal, our work highlights the necessity of a higher-dimensional neuromodulatory or electrical feedback signal for learning potentially with some knowledge of the feedforward pathway. The biological exploration of such intelligent neuromodulation, as well as extensions of our approach to deeper and recurrent SNNs, are left as intriguing directions for future work."
} | 3,445 |
30787397 | null | s2 | 3,542 | {
"abstract": "A remaining challenge within microbial ecology is to understand the determinants of richness and diversity observed in environmental microbial communities. In a range of systems, including activated sludge bioreactors, the microbial residence time (MRT) has been previously shown to shape the microbial community composition. However, the physiological and ecological mechanisms driving this influence have remained unclear. Here, this relationship is explored by analyzing an activated sludge system fed with municipal wastewater. Using a model designed in this study based on Monod-growth kinetics, longer MRTs were shown to increase the range of growth parameters that enable persistence, resulting in increased richness and diversity in the modeled community. In laboratory experiments, six sequencing batch reactors treating domestic wastewater were operated in parallel at MRTs between 1 and 15 days. The communities were characterized using both 16S ribosomal RNA and non-target messenger RNA sequencing (metatranscriptomic analysis), and model-predicted monotonic increases in richness were confirmed in both profiles. Accordingly, taxonomic Shannon diversity also increased with MRT. In contrast, the diversity in enzyme class annotations resulting from the metatranscriptomic analysis displayed a non-monotonic trend over the MRT gradient. Disproportionately high abundances of transcripts encoding for rarer enzymes occur at longer MRTs and lead to the disconnect between taxonomic and functional diversity profiles."
} | 381 |
26908346 | PMC4764831 | pmc | 3,543 | {
"abstract": "Water decontamination and oil/water separation are principal motives in the surge to develop novel means for sustainability. In this prospect, supplying clean water for the ecosystems is as important as the recovery of the oil spills since the supplies are scarce. Inspired to design an engineering material which not only serves this purpose, but can also be altered for other applications to preserve natural resources, a facile template-free process is suggested to fabricate a superporous, superhydrophobic ultra-thin graphite sponge. Moreover, the process is designed to be inexpensive and scalable. The fabricated sponge can be used to clean up different types of oil, organic solvents, toxic and corrosive contaminants. This versatile microstructure can retain its functionality even when pulverized. The sponge is applicable for targeted sorption and collection due to its ferromagnetic properties. We hope that such a cost-effective process can be embraced and implemented widely.",
"conclusion": "Conclusions In summary, we have successfully demonstrated the synthesis of a scalable, multifunctional ultra-thin graphite sponge with exceptional porosity and surface area as well as superhydrophobic and oleophilic properties via a modified sol-gel assisted process. The structure and synthesis process are designed to be effortlessly inexpensive and highly scalable to be widely adopted and used. The synthesized sponge has ferromagnetic properties as a result of graphene-wrapped α-Fe nanoparticles with average diameter of about 20 nm. We have also illustrated the versatility of the architecture enabling the required modifications such as direct CNT growth or incorporation of other metallic nanoparticles within the structure which expands its applicable venues. The obtained sponge has an exceptional contact angle of 154.72° with water as well as oleophilic properties. It offers a remarkable surface area of 823.77 m 2 .g −1 and an average pore diameter of 1.4 nm without chemical activations. Such surface area is due to the presence of macro, meso and micropores and channels in the structure. The sponge is effectively applicable to be used underwater and even when pulverized. Selective and targeted delivery and collection of the sponge to contaminated coordinates is also possible as a result of embedded ferromagnetic α-Fe nanoparticles. The achieved architecture offers a promising ability environmental cleaning applications such as oil-water separation and water decontamination. Owing to its versatile microstructure, it may also be used for gas filtration and sensing as well as energy storage."
} | 651 |
33491685 | null | s2 | 3,544 | {
"abstract": "Methanotrophic bacteria represent a potential route to methane utilization and mitigation of methane emissions. In the first step of their metabolic pathway, aerobic methanotrophs use methane monooxygenases (MMOs) to activate methane, oxidizing it to methanol. There are two types of MMOs: a particulate, membrane-bound enzyme (pMMO) and a soluble, cytoplasmic enzyme (sMMO). The two MMOs are completely unrelated, with different architectures, metal cofactors, and mechanisms. The more prevalent of the two, pMMO, is copper-dependent, but the identity of its copper active site remains unclear. By contrast, sMMO uses a diiron active site, the catalytic cycle of which is well understood. Here we review the current state of knowledge for both MMOs, with an emphasis on recent developments and emerging hypotheses. In addition, we discuss obstacles to developing expression systems, which are needed to address outstanding questions and to facilitate future protein engineering efforts."
} | 246 |
35520851 | PMC9057580 | pmc | 3,545 | {
"abstract": "Inspired by natural surfaces such as butterfly wings, cactus leaves, or the Nepenthes alata plant, synthetic materials may be engineered to directionally transport liquids on their surface without external energy input. This advantageous feature has been adopted for various mechanical and chemical processes, e.g. fog harvesting, lubrication, lossless chemical reactions, etc. Many studies have focused on the manipulation and transport of water or aqueous droplets, but significantly fewer have extended their work to low surface tension (LST) liquids, although these fluids are involved in numerous industrial and everyday processes. LST liquids completely wet most surfaces which makes spontaneous transportation an active challenge. This review focuses on recently developed strategies for passively and directionally transporting LST liquids.",
"conclusion": "7 Conclusion and perspectives Bio-mimetic surfaces have been developed to enable the directional transportation of liquid droplets. However, LST liquids either partially or completely wet all surface chemistries which limits the functionality of these surfaces. In this review, the developed strategies to passively transport LST liquids are categorized in either the spread or slip mode of transportation. In the spread mode of transportation, a droplet spreads preferentially in one direction and is pinned in the reverse direction, with potential applications in self-lubrication, microfluidic devices, etc. The slip mode can potentially enable lossless transportation of LST liquids as the contact line recedes from the trailing end of the droplet. Droplet transportation velocity and maximum travel distance were used to compare the performance of these strategies. The wettability behavior of LST liquids brings up challenges that limit the functionality of each method; the oleophobic background for patterned channels on smooth surfaces, the complex fabrication of wedge corner structures along with their low transportation velocities, and the limited travel distance for conical structure and re-entrant patterns. The development of a versatile method that satisfies all the requirements for LST liquids transportation remains an active challenge. However, current strategies have shown promise towards the guided transportation of LST liquids.",
"introduction": "1 Introduction The ability to directionally transport liquid droplets has naturally developed in several instances. For example, desert beetles, spider silk, and cactus stems are all capable of collecting water from humid air. 1–7 Due to recent advances in micro/nanofabrication methods, scientists have developed bio-inspired surface structures that also achieve directional liquid transportation. These surfaces have shown promise in various fields, such as micro-fluidics, oil–water separation, thin-film lubrication, and heat management. 8–15 The transportation of water droplets on these engineered surfaces is well studied, both experimentally and theoretically. 16–22 Considerably fewer studies have demonstrated directional transportation of Low Surface Tension (LST) liquids. Here we review surfaces that utilize physical texture gradients to uniquely enable the directional transportation of LST liquids. A droplet deposited on a solid surface creates a solid/liquid/vapor contact line as a result of an equilibrium between interfacial tensions. In the absence of any anisotropy, the droplet will move/expand equally in all directions in response to external pressure. Wetting is caused by a capillary force, which scales with the product of surface tension and a characteristic length. Accordingly, wetting may be manipulated by a gradient in either the surface tension or the characteristic physical length. A surface tension gradient may be induced by intrinsic surface chemistry variation 23–27 or by external stimuli such as an electric field, light, temperature, vibration, or a magnetic field. 28–35 In comparison, engineered surfaces exhibiting anisotropic surface texture are beneficial for power-free, spontaneous liquid transportation. As LST liquids generally spread on most surface chemistries, rather than forming droplets, achieving a chemical wettability gradient for the wetting driving force is difficult. However, droplet formation (non-zero contact angles) can occur in the case of underwater deposition, on very low surface energy materials, and on specially engineered rough surfaces. Therefore, this review mainly discusses surfaces with textural anisotropy, although the limited studies utilizing a chemical gradient are also discussed."
} | 1,145 |
23307738 | null | s2 | 3,547 | {
"abstract": "Here we describe the development of a water-responsive polymer film. Combining both a rigid matrix (polypyrrole) and a dynamic network (polyol-borate), strong and flexible polymer films were developed that can exchange water with the environment to induce film expansion and contraction, resulting in rapid and continuous locomotion. The film actuator can generate contractile stress up to 27 megapascals, lift objects 380 times heavier than itself, and transport cargo 10 times heavier than itself. We have assembled a generator by associating this actuator with a piezoelectric element. Driven by water gradients, this generator outputs alternating electricity at ~0.3 hertz, with a peak voltage of ~1.0 volt. The electrical energy is stored in capacitors that could power micro- and nanoelectronic devices."
} | 202 |
31032842 | PMC6487390 | pmc | 3,548 | {
"abstract": "Abstract Progress in genome sequencing and bioinformatics opens up new possibilities, including that of correlating genome annotations with functional information such as metabolic pathways. Thanks to the development of functional annotation databases, scientists are able to link genome annotations with functional annotations. We present MetAboliC pAthways DAtabase for Microbial taxonomic groups (MACADAM) here, a user-friendly database that makes it possible to find presence/absence/completeness statistics for metabolic pathways at a given microbial taxonomic position. For each prokaryotic ‘RefSeq complete genome’, MACADAM builds a pathway genome database (PGDB) using Pathway Tools software based on MetaCyc data that includes metabolic pathways as well as associated metabolites, reactions and enzymes. To ensure the highest quality of the genome functional annotation data, MACADAM also contains MicroCyc, a manually curated collection of PGDBs; Functional Annotation of Prokaryotic Taxa (FAPROTAX), a manually curated functional annotation database; and the IJSEM phenotypic database. The MACADAM database contains 13 509 PGDBs (13 195 bacterial and 314 archaeal), 1260 unique metabolic pathways, completed with 82 functional annotations from FAPROTAX and 16 from the IJSEM phenotypic database. MACADAM contains a total of 7921 metabolites, 592 enzymatic reactions, 2134 EC numbers and 7440 enzymes. MACADAM can be queried at any rank of the NCBI taxonomy (from phyla to species). It provides the possibility to explore functional information completed with metabolites, enzymes, enzymatic reactions and EC numbers. MACADAM returns a tabulated file containing a list of pathways with two scores (pathway score and pathway frequency score) that are present in the queried taxa. The file also contains the names of the organisms in which the pathways are found and the metabolic hierarchy associated with the pathways. Finally, MACADAM can be downloaded as a single file and queried with SQLite or python command lines or explored through a web interface.",
"conclusion": "Conclusions MACADAM was designed for the microbiology community as a functional annotation information database based on multiple sources of data on functional annotations and on metabolic pathways (MetaCyc, MicroCyc, FAPROTAX and the IJSEM phenotypic database). The database is also based on the complete and interoperable NCBI taxonomy. MACADAM covers all known bacterial and archaeal phyla, as of February 2019. A standardized score enables quick comparison and comprehension of the potential presence of a pathway. If there is no functional information on the taxonomy entered, MACADAM automatically checks the upper taxonomic rank in order to provide functional information associated with related organisms to users. MACADAM can be explored via metabolites, reactions, enzymes, EC numbers or specific pathways. A user-friendly web interface makes querying easy. MACADAM will be useful to all biologists who need to determine the functional potential of a prokaryotic species or any other taxonomic rank. Since the source code to build MACADAM is available to everyone (GitHub URL: https://github.com/maloleboulch/MACADAM-Database ), MACADAM can be included in any functional inference tool able to integrate the abundance tables of complete microbial communities generated, among others, we plan to include MACADAM in the FROGS software ( 39 ) to analyze amplicon metagenomics data.",
"introduction": "Introduction For many years, Bergey’s Manual of Determinative Bacteriology ( 1 ) and its successor, Bergey’s Manual of Systematic Bacteriology ( 2–6 ), which provides descriptions of the taxonomy, systematics, ecology, physiology and other biological properties of all described prokaryotic taxa, has been the best consensus for an official prokaryotic classification and the best source of information for prokaryotic organisms and taxa. Thanks to advances in genome sequencing and bioinformatics, it is now possible to link genome annotations and functional information. To make this possible, databases have been built and contain metabolic pathways, e.g. series of chemical reactions catalyzed by enzymes within a cell. For instance, the KEGG database ( 7 ) can display any of these pathways in a graphical environment. The Human Metabolome DataBase (HMDB) ( 8 ) and Reactome ( 9 ) are highly curated and complete databases specializing in human metabolism. WikiPathways ( 10 ) is an open access collaborative platform containing metabolic pathways across different species. PATRIC ( 11 ) is a bacterial database containing >201 000 prokaryotic genomes, each associated with functional information. BioCyc ( 12 ) links the genome sequence of an organism to its functional annotation in >14 560 eukaryotic, bacteria and archaea species. All these databases are referred to as pathway genome databases (PGDBs); i.e. they associate the genome sequences with metabolic pathways. Currently, among available databases, some databases are highly curated, including the EcoCyc ( 13 ), BsubCyc ( 14 ) and HumanCyc ( 15 ) databases devoted to Escherichia coli K-12, Bacillus subtilis or human metabolic pathways, respectively. They are based on the MetaCyc ( 16 ) database, which is a highly curated database containing >2666 metabolic pathways throughout the living world. The MicroCyc database ( 17 ), based on the MetaCyc database, has been improved by automatic and manual curation by specialized biologists. Finally, some other databases are also curated using functional information from the literature, e.g. FAPROTAX ( 18 ) or the IJSEM phenotypic database ( 19 , 20 ). Each of these latter databases has limits to link microbial taxonomy to functional information and is not easily downloadable. HMDB, Reactome and WikiPathways are manually and highly curated, but, despite the high quality of their functional information, they cover a small number of organisms [1, 1 and 31 (3 of which are microbes), respectively]. KEGG, despite having >5299 prokaryotic organisms, has moved to a subscription mode and cannot be accessed offline. PATRIC depends on the KEGG pathway data and cannot be downloaded ( Table 1 ). The BioCyc database is a microbial genome web portal that combines thousands of genomes with pathway information, but the BioCyc website uses a subscription model, free access to the derived BioCyc database is limited to a 2-year-old collection of PGDBs and online consultation is limited to a specific number of times per month. Further access requires a paid subscription. MetaCyc contains a greater number of metabolic pathways than KEGG ( 21 ) (2666 vs. 530 metabolic pathways, respectively), and it is freely available for academics, but only a few metabolic pathways are retrievable via a taxonomy or an organism name. With existing databases, it is difficult to request up-to-date functional annotations and up-to-date taxonomic lineages for a taxon (e.g. a whole family). But one MACADAM feature is not provided by these databases: the possibility to infer functional information for prokaryotic taxa with no functional information associated with it. That is why we built MACADAM (MetAboliC pAthways DAtabase for Microbial taxonomic groups), a user-friendly database that makes it possible to find presence/absence/completeness statistics for metabolic pathways at a given archaeal and bacterial taxonomic rank or organism and to be able to infer functional information using the taxonomy for taxa without functional information. MACADAM is not intended to replace existing databases but provides additional information for scientists wishing to better characterize the functional information of all taxonomic groups, from phyla to species. Table 1 Overview of MACADAM, BioCyc, PATRIC and KEGG database features with a focus on metabolic pathway and functional information among prokaryotic organisms \n MACADAM \n \n BioCyc \n \n PATRIC \n \n KEGG \n \n Microbial taxonomy used for requests \n NCBI taxonomy NCBI taxonomy NCBI taxonomy KEGG taxonomy (with cross-link to NCBI) \n Query possibilities \n On one or several taxonomies or organisms, with few filters On one organism, with multiple filters On one or several taxonomies or organisms, with multiple filters On one organism, with no filters \n Number of bacterial organisms \n 13 195 ~13 400 198 855 5014 \n Number of archaeal organisms \n 314 ~400 3069 285 \n Number of unique metabolic pathways \n 1260 2666 143 530 \n Genome origins \n RefSeq (complete genomes) Genbank and RefSeq Genbank and RefSeq Genbank and RefSeq \n Functional annotation sources \n RefSeq (functional annotations), MetaCyc (metabolic pathways), MicroCyc (metabolic pathways), FAPROTAX (functional features), IJSME PhenoDB (phenotypic data) Genbank/RefSeq (annotations) and MetaCyc (metabolic pathways) Genbank/RefSeq, KEGG (metabolic pathways) KEGG \n Analysis tools and metrics \n PS and PFS SmartTables, genome browser, omics data analysis, metabolic models and routes and comparative analysis KEGG pathway map, comparative pathway heatmap, multiple sequence alignment, enzymes and genes conservation in pathway KEGG mapper tools \n Output data types \n Metabolic pathway name, pathway class hierarchy, hyperlink to MetaCyc pathway and functional information of the upper rank for taxa without data Metabolic pathway name, pathway class hierarchy, pathway map including enzymes and metabolites, associated genes, protein associated with pathways and literature references Metabolic pathway name, pathway class hierarchy, KEGG pathway map including enzymes and metabolites, associated genes, enzyme and gene evolution data Metabolic pathway name, pathway class hierarchy, KEGG pathway map including enzymes and metabolites, associated genes and literature references \n Database flat files downloadable \n Yes Yes for academics No Yes with license \n Results downloadable \n Yes Yes Yes Available on the web only \n Command line interrogation \n SQLite or python script Application Programming Interface (API) API + free command line software API \n Frequency of updates \n 6 months 2–6 months 6 months 3 months Figure 1 MACADAM building workflow.",
"discussion": "Utility and discussion \n MACADAM is designed to collect bacterial and archaeal genomes of the highest quality associated with the highest quality annotations \n Reliable data are needed to infer the functional potential of complex prokaryotic communities. Indeed, obsolete functional information can lead to inaccurate insights ( 36 ). As far as possible, MACADAM avoids sequentially spurious annotations. In the MACADAM database, reliability is ensured by filters on genome quality, meaning that only complete genomes have been taken from RefSeq. Moreover, to ensure up-to-date annotations, we compute PGDBs with Pathway Tools at each release, based on RefSeq, NCBI taxonomy, MetaCyc, MicroCyc, FAPROTAX and the IJSEM phenotypic database. Figure 7 Phyla distribution in the MACADAM database according to their database of origin. ( A ) MetaCyc and MicroCyc for bacterial organisms, ( B ) MetaCyc and MicroCyc for archaea organisms, ( C ) FAPROTAX for bacterial organisms, ( D ) FAPROTAX for archaea organisms, ( E ) IJSEM phenotypic database for bacterial organisms and ( F ) IJSEM phenotypic database for archaea organisms. \n MACADAM PGDBs cover all phyla recognized by the List of Prokaryotic names with Standing in Nomenclature \n (LPSN; http://www.bacterio.net/ ) and the 10 other newly proposed phyla from the NCBI taxonomy ( Figure 7 ). Proteobacteria is the most prevalent phylum and accounts for >55% of the genomes collected in MACADAM, followed by the Firmicutes and Actinobacteria phyla that account for >22% and 11% of the collected genomes, respectively. This pre-eminence is probably explained by the research effort devoted to these phyla by biologists. Accordingly, Escherichia and Salmonella are the prevalent genera in the database (5.1% and 4.3% of the database, respectively). Interestingly, according to Figure 7 , the Acidobacteria phylum is weakly represented in MACADAM (only 21 PGDBs, 19 of which are from MicroCyc). But these PGDBs include the highest number of metabolic pathways (mean = 230; min = 138; max = 298; Figure 8 ). Acidobacteria is one of the most widespread phyla, but few organisms of this phylum are cultivated and sequenced ( 37 ) and may explain the low representation of available high-quality genome that could be included in MACADAM. The high number of metabolic pathway present in these widespread phyla may indicate that, despite few sequenced organisms, these are subject to particular care in terms of functional annotation. In parallel, we should also recall that the most common metabolic pathways in MACADAM belong to cofactor biosynthesis (bacteria: 16.4%; archaea: 18.06%) and the metabolism of amino acids (bacteria: 17.7%; archaea: 22.86%). Figure 8 MACADAM functional diversity for phyla with >10 PGDBs in MACADA ( A ) among bacterial phyla, ( B ) among archaea phyla, ( C ) the 10 main hierarchical groups of pathways in all bacterial organisms and ( D ) the 10 main hierarchical groups of pathways in all archaea organisms. Table 4 Statistics on the L-lysine fermentation to acetate and butanoate pathway (MetaCyc ID is P163-PWY; values in brackets are minimum and maximum values \n MetaCyc ID: P163-PWY characteristics \n Number of reactions in the complete pathway * 10 Number of bacteria in which this pathway is present 756 Median number of unique enzymes present in this pathway in organisms 4 [1–10] Median number of enzymes present in this pathway in organisms 7 [1–70] Key reaction * 1 (enzyme classification: 5.4.3.2) PS 0.4 [0.1–1] PFS 0.7 [0.1–7] \n * Value found in a MetaCyc flat file; a key reaction is a reaction that is specific to a single pathway, i.e. a reaction that is not found in any other pathway. \n MACADAM allows users to refine queries using PSs \n PSs range from 0 to 1 ( Table 2 ). A value of 1 indicates that all the enzymes required for the pathway are present in the genome of the organism. A value of 0 indicates that none of the enzymes are present in the targeted genome. In the latter case, the pathway is still present because of the functional inference applied by MetaCyc experts, information that they cross-reference with phenotypic evidence described in the literature. MACADAM contains eight pathways in 1519 organisms with a PS equal to 0. The PFS ranges from 0 to 77 ( Table 2 ). Nocardia nova SH22a has the maximum PFS for its long-chain fatty acid activation pathway. This pathway comprises only one reaction and is part of a longer lipid biosynthesis pathway. \n Table 4 is an example of statistics performed on L-lysine fermentation to acetate and butanoate pathway output (MetaCyc ID: P163-PWY), a pathway of interest in understanding the interactions between host and gut microbiota in health and disease. In fact, butyrate is a microbial fermentation product that is used as an energy source by enterocytes and whose signaling properties are involved in multiple functions of enterocytes, including cell differentiation, gut tissue development, immune modulation, oxidative stress reduction and diarrhea control ( 38 ). Ten enzymes are involved in the L-lysine fermentation to acetate and butanoate pathway. The pathway is present in 992 different organisms in MACADAM. This complete pathway is composed of 10 enzymes. If the median value of the PS is equal to 0.4, this means that four reactions have at least one annotated enzyme in the pathway. If the median value of the PFS is equal to 0.7, this means these 4 reactions have >1 associated enzymes, 7 enzymes for 10 reactions in all ( Figure 2 ). According to MetaCyc, the 5.4.3.2 reaction is a key reaction in this pathway. Therefore, if the organism contains this reaction, the whole pathway will be identified for the organism. These PS and PFS data are useful to biologists for data mining. An important feature of MACADAM is its ability to infer functional annotations for taxa with no associated genomic sequences using the taxonomy . In the case of a taxon with no functional information in MACADAM, i.e. missing in MetaCyc, MicroCyc, FAPROTAX and the IJSEM phenotypic database, we provide information for the upper taxonomy rank. For example, if a species has no functional information, MACADAM automatically requests functional information at the genus level, just like FAPROTAX. However, MACADAM can do this at any taxonomic rank, while FAPROTAX is limited to the order rank. In MACADAM, in this case, all annotations for organisms belonging to this genus are shown in the output file. As for FAPROTAX, it indicates functional information described in the literature at the genus rank if all described species of the genus have been shown to exhibit the given functional information and not the addition of functional information about all of the organisms belonging to this rank. Thanks to the column that gives the number of organisms with the targeted pathway (column 2 in Figure 6B ), it is possible to see pathways that are more or less conserved in the taxon of interest. For example, ‘urea cycle’ is a pathway conserved in most of Lactobacillus , unlike the ‘urea degradation I or II’ pathways. Thus, this feature provides functional information about organisms with no functional information based on related taxonomic species."
} | 4,339 |
37795899 | PMC10826995 | pmc | 3,551 | {
"abstract": "Summary Grass lignocelluloses feature complex compositions and structures. In addition to the presence of conventional lignin units from monolignols, acylated monolignols and flavonoid tricin also incorporate into lignin polymer; moreover, hydroxycinnamates, particularly ferulate, cross‐link arabinoxylan chains with each other and/or with lignin polymers. These structural complexities make grass lignocellulosics difficult to optimize for effective agro‐industrial applications. In the present study, we assess the applications of two engineered monolignol 4‐ O ‐methyltransferases (MOMTs) in modifying rice lignocellulosic properties. Two MOMTs confer regiospecific para ‐methylation of monolignols but with different catalytic preferences. The expression of MOMTs in rice resulted in differential but drastic suppression of lignin deposition, showing more than 50% decrease in guaiacyl lignin and up to an 90% reduction in syringyl lignin in transgenic lines. Moreover, the levels of arabinoxylan‐bound ferulate were reduced by up to 50%, and the levels of tricin in lignin fraction were also substantially reduced. Concomitantly, up to 11 μmol/g of the methanol‐extractable 4‐ O ‐methylated ferulic acid and 5–7 μmol/g 4‐ O ‐methylated sinapic acid were accumulated in MOMT transgenic lines. Both MOMTs in vitro displayed discernible substrate promiscuity towards a range of phenolics in addition to the dominant substrate monolignols, which partially explains their broad effects on grass phenolic biosynthesis. The cell wall structural and compositional changes resulted in up to 30% increase in saccharification yield of the de‐starched rice straw biomass after diluted acid‐pretreatment. These results demonstrate an effective strategy to tailor complex grass cell walls to generate improved cellulosic feedstocks for the fermentable sugar‐based production of biofuel and bio‐chemicals.",
"introduction": "Introduction Grass crops such as rice ( Oryza sativa ), maize ( Zea mays ), wheat ( Triticum aestivum ), sugarcane ( Saccharum spp.) and sorghum ( Sorghum spp.) yield not only edible grains but also abundant lignocellulosic materials, such as straw, stover and bagasse, as agricultural residues. Approximately 44 billion tons of such non‐wood lignocellulosics are available annually world‐wide (Tye et al ., 2016 ), which could potentially produce 491 GL/year bioethanol. Among them rice straw can contribute to 205 GL/year, representing the largest amount from single biomass feedstock (Kim and Dale, 2004 ). However, as a common obstacle, lignocellulosic utilization is largely impeded by the associated aromatic constituents, including lignin and low‐molecular‐weight phenolics, within the grass fibers (Halpin, 2019 ). Lignocellulosic biomass with low lignin content and/or easily removable lignin has been desired (Liu et al ., 2014 ). Grass stems account for the majority of secondary wall‐forming sclerenchyma tissues and the interfascicular fibers that develop thick secondary walls conferring mechanical strength for the upright stem (Coomey et al ., 2020 ). While the architecture of eudicot and monocot cell walls in general is similar, i.e., both consist of a network of cellulose fibers surrounded by a matrix of non‐cellulosic polysaccharides and impregnated with lignin, grasses have many special cell‐wall features distinct from eudicots and other plants. One of the defining features is the presence of abundant hydroxycinnamates, namely p ‐coumarate ( p CA) and ferulate (FA). The amounts of these phenolics accumulate up to 3%–4% of dry grass biomass (Hatfield et al ., 1999 ). p CA primarily links to lignin through the incorporation of acylated lignin monomers (monolignols) (Karlen et al ., 2018 ), while FA predominantly links to the heteroxylan (arabinoxylan and glucuronoarabinoxylan) through ester linkage between the carboxylic acid group of FA and the primary alcohol on the C5 of arabinosyl side‐chain of heteroxylan (Bunzel et al ., 2004 ; de Oliveira et al ., 2015 ). The oxidation of FA in the cell wall generates esterified (dehydro‐)diferulate (diFA) and/or oligoferulate bridges, which serve as linkages between two arabinoxylan polymers. In lignified tissues, esterified FAs on arabinoxylan can also be etherified to lignin polymer units via oxidative radical coupling, forming covalent linkages between hemicellulosic arabinoxylan and lignin, which represents an important mechanism for cell wall reinforcement (Hatfield et al ., 2016 ). The esterified FA on arabinoxylan is also considered as a nucleation site of lignification in grass cell walls (Ralph et al ., 1995 ). Nevertheless, the FA bridges possess a negative impact on the cell wall digestibility (de Oliveira et al ., 2015 ; Grabber et al ., 1998 ). Lignin in both eudicots and monocots is mainly composed of guaiacyl (G) and syringyl (S) units, along with a lower level of p ‐hydroxyphenyl (H) units, which are synthesized via oxidative radical coupling of monolignols, coniferyl alcohol, sinapyl alcohol and p ‐coumaryl alcohol, respectively (Tobimatsu and Schuetz, 2019 ; Vanholme et al ., 2010 ). Although sharing these conventional components with eudicot lignin, grass lignin has several special features, including the incorporations of flavone tricin (Lan et al ., 2015 ) and γ‐ p ‐coumaroylated monolignols such as coniferyl p ‐coumarate and sinapyl p ‐coumarate (Karlen et al ., 2018 ) as well as γ‐feruloylated monolignols albeit at much lower levels (Karlen et al ., 2016 ). Tricin has been revealed to act as a real lignin monomer and incorporate into grass lignin in varying amounts across grass species (del Rio et al ., 2012 ; Lan et al ., 2016b ). Its incorporation is mainly established via 4′‐ O ‐β coupling, thus being proposed to serve as another possible nucleation site for lignification in addition to arabinoxylan‐bound FA (Lan et al ., 2015 , 2018 ). In monolignol biosynthesis, the aromatic ring of phenolic intermediate is methoxylated at meta (3′and/or 5′) positions for the formation of coniferyl alcohol and sinapyl alcohol. The 3′‐ and 5′‐ O ‐methylations of benzene ring are catalysed by caffeoyl coenzyme A 3‐ O ‐methyltransferase (CCoAOMT, EC. 2.1.1.104) and caffeic acid/5‐hydroxyferulic acid 3/5‐ O ‐methyltransferase (COMT, EC. 2.1.1.6), respectively. CCoAOMT preferentially catalyses the methylation of the 3‐hydroxyl of 3,4‐dihydroxy (caffeoyl) precursors, while COMT prefers 5‐hydroxyconiferaldehyde and 5‐hydroxyconiferyl alcohol substrates (Guo et al ., 2001 ; Parvathi et al ., 2001 ). However, many studies reveal that COMT catalyses a broad range of substrates, including not only lignin biosynthetic precursors such as 5‐hydroxyconiferaldehyde, 5‐hydroxyconiferyl alcohol, caffeoyl aldehyde, caffeoyl alcohol (Osakabe et al ., 1999 ; Parvathi et al ., 2001 ), but also, flavones luteolin and selgin, leading to the formation of methylated flavone tricin in rice, maize and sorghum (Eudes et al ., 2017 ; Fornale et al ., 2017 ; Lam et al ., 2019 ). Despite the broad substrate range, the activities of both CCoAOMT and COMT are confined to strict regio‐specific methylations, i.e., the methylation occurs exclusively at the meta (3′ or 5′) ‐hydroxyl (Bhuiya and Liu, 2010 ; Osakabe et al ., 1999 ; Parvathi et al ., 2001 ). The para ‐hydroxyl of lignin monomer is free from methyl etherification and is critically important for radical generation and oxidative polymerization (Bhuiya and Liu, 2010 ; Davin and Lewis, 2005 ; Ralph et al ., 2004 ). Previously, we have engineered a set of monolignol 4‐ O ‐methyltransferases (MOMTs) through structure‐guided iterative saturation mutagenesis on a parent enzyme, isoeugenol 4‐ O ‐methyltransferase, an ortholog evolutionarily derived from COMT (Bhuiya and Liu, 2010 ). The obtained MOMTs confer substantial ability in methylating the para ‐hydroxyl of monolignols. The para ‐hydroxyl methylation deprives the propensity of the modified lignin monomer for further dehydrogenation and incorporation into lignin polymer, which consequently reduces the availability of lignin monomers for polymerization, thus reducing total lignin content or altering lignin composition if a particular type of monolignol is preferentially etherified by MOMTs. One of the mutant variants, MOMT4, which bears four amino acid substitutions (T133L/E165I/F175I/ /H169F) in its active site, effectively catalyses the 4‐ O ‐methylation of both coniferyl alcohol and sinapyl alcohol, with a slight preference for the latter (Bhuiya and Liu, 2010 ; Zhang et al ., 2012 ). The variant MOMT9, which possesses five additional substitutions (M26H/S30R/V33S/F166W/T319M), however, preferentially etherifies coniferyl alcohol over sinapyl alcohol in vitro , due to its significantly reduced substrate binding pocket (Cai et al ., 2015 ). The expression of MOMT4 in eudicots ( Arabidopsis and poplar) significantly impeded lignin formation and/or altered composition, consequently enhanced lignocellulosic digestibility (Cai et al ., 2016 ; Zhang et al ., 2012 ). Interestingly, the effects of MOMT4 in Arabidopsis and poplar on lignin and the related phenylpropanoid biosynthesis were different. The expression of MOMT4 in Arabidopsis resulted in the reduction of both G‐ and S‐lignin subunits without alteration in the S/G ratio (Zhang et al ., 2012 ), whereas when MOMT4 was expressed in poplar, it drastically reduced S‐lignin accumulation without impairing G‐lignin subunits and, consequently, drastically reversed the S/G ratio of poplar lignin (Cai et al ., 2016 ). These data validate the salient diversity and plasticity of lignin biosynthesis in different plant species and the functional plasticity of MOMTs on lignin formation in different plants. However, this set of engineered MOMTs has not been assessed in grass species. In the present study, we evaluate the effects of not only MOMT4 but also MOMT9 on lignin biosynthesis and cell wall feruloylation in rice. We reveal that the activity of either MOMT variants results in much broader effects in grass than in eudicot species. Not only S‐ and G‐lignin monomers are drastically reduced in rice transgenic lines, the wall‐bound p CA and FA, and lignin‐bound tricin are all dramatically decreased. Concomitantly, significant amounts of novel 4‐ O ‐methylated phenolics, i.e., the 4‐ O ‐methylated FA and/or 4‐ O ‐methylated sinapate derivatives, are accumulated in the methanolic extractable fraction of the MOMT‐overexpressing rice. Furthermore, consistent with their in vitro catalytic properties, MOMT4 and MOMT9 in the rice transgenic lines exhibit discernible differential effects on the syntheses of lignin and the wall‐bound and methanol‐soluble phenolics. As a consequence of the cell wall compositional changes, lignocellulosic biomass from MOMT‐overexpressing rice shows 15%–30% increases in simple sugar releases after diluted acid pretreatment. These data signify that applying MOMTs could effectively tailor lignin and the related cell wall phenolics biosynthesis in grass species, which allows agricultural residues from grass crops such as rice straw to be more efficiently utilized by mitigating the presence of lignin and cell wall cross‐linkers for biofuels production.",
"discussion": "Discussion Grass lignocelluloses represent an abundant renewable biomass source for chemical pulping and fermentable sugar‐based applications such as biofuel and biochemical productions (Tye et al ., 2016 ; Umezawa, 2018 ). However, as the common obstacle, lignin present in the grass cell walls confers significant recalcitrance to enzymatic hydrolysis (Moore and Jung, 2001 ). Moreover, the cell wall matrix of grass is much more complicated than that in eudicots and gymnosperms, featuring the presence of the cell‐wall‐cross‐linking FA and the lignin decorations by p CA and flavone tricin (Ralph et al ., 2019 ). These complexities render substantial difficulties in optimizing grass lignocellulosics for agro‐industrial applications. The expression of either MOMT4 or MOMT9 in rice resulted in drastic reduction of both lignin and wall‐bound phenolics; concomitantly, a large quantity of methanolic extractable 4‐ O ‐methylaled phenolics were produced (Figures 2 , 3 , 5 and 6 ). These results demonstrate the effectiveness and robustness of both MOMTs in tailoring grass cell wall chemical and structural properties. The expression of MOMTs in rice led to a drastic reduction of both G and S lignin units, which differs from what occurred when MOMT4 was expressed in eudicot poplar, where S lignin unit was predominantly suppressed (Cai et al ., 2016 ). This discrepancy might reflect the diversity and plasticity of the phenylpropanoid and lignin biosynthesis among different plant species. Poplar wood particularly its fibre walls enriches S‐lignin with S:G ratio around 2 (Zhao et al ., 2021 ), while lignin of grass species contains approximately equal amount of G and S subunits with S:G ratio around 0.6–0.7 (Lam et al ., 2017 , 2019 ). In addition, the promoters used to drive MOMT transgene expression in poplar and rice are also different; in the former, it was a French bean PAL2 gene promoter, while in the latter a rice C4H gene promoter was adopted, which might lead to differential spatiotemporal expression patterns of MOMT transgenes, ultimately result in different chemical phenotypes. Furthermore, two tested MOMT variants exhibit different effects on lignin and phenolic biosynthesis in planta. MOMT4, which favours methylating the S‐type sinapyl alcohol over the G‐type coniferyl alcohol in vitro (Bhuiya and Liu, 2010 ; Zhang et al ., 2012 ), suppresses more S lignin deposition in transgenic rice, resulting in large reductions in the S/G lignin ratios as determined by both thioacidolysis (Figure 5c ) and 2D NMR (Figure 6c ), whereas expressing MOMT9, a variant that more prefers modifying G monomers over S monomers in vitro due to its smaller substrate binding pocket (Cai et al ., 2015 ), leads to a less prominent effect on the S/G lignin ratio reductions (Figures 5c and 6c ), and concomitantly the sole production of the soluble 4‐ O ‐methylated FA‐derived metabolites (Figure 2 ). These results demonstrate that both MOMT variants function properly in grass species and are in line with their in vitro structural and catalytic properties. The expression of both MOMTs not only suppressed lignin biosynthesis but also drastically reduced wall‐bound FA and p CA (Figure 3 ). The FA and p CA esters, particularly the FA‐arabinoxylan esters, have been demonstrated to be involved in radical coupling reactions to cross‐link cell wall polysaccharide (arabinoxylan) and lignin polymer chains. The resulting arabinoxylan‐FA‐arabinoxylan and arabinoxylan–FA–lignin complexes represent the key structures that stabilize grass cell walls and maintain cell wall rigidity (Hatfield et al ., 2016 ). In vitro assays suggested that the wall‐bound FA is negatively correlated with cell wall digestibility (de Oliveira et al ., 2015 ). Suppression of wall‐bound FA in Setaria viridis via the downregulation of a BAHD acyltransferase that potentially functions for conjugating FA to arabinoxylan significantly improved biomass saccharification efficiency (de Souza et al ., 2018 ). The mild acidolysis, a procedure for determining the relative acylation of grass arabinoxylan (Eugene et al ., 2020 ; Lapierre et al ., 2019 ), revealed about 60% reduction of arabinoxylan‐bound FA in the cell walls of the MOMT transgenic lines (Figure 4 ). Such drastic loss of the arabinoxylan‐bound FA expectedly could result in the decrease of cross‐links between the matrix heteroxylan and lignin polymer chains, contributing to the improvement of enzymatic cell wall digestibility. Significant suppression of the wall‐bound phenolics, particularly the polysaccharide‐linked FA, in the MOMT overexpression lines is somehow unexpected. The formation of wall‐bound hydroxycinnamates is catalysed by particular BAHD family acyltransferases that transfer FA or p CA moieties from their corresponding CoA thioesters to the acceptors (e.g., arabinose and/or monolignols); then the conjugates are incorporated into polysaccharides or lignin (Bartley et al ., 2013 ; de Souza et al ., 2018 ; Li et al ., 2018 ; Petrik et al ., 2014 ; Withers et al ., 2012 ). While MOMTs possess strict regiospecificity, viz . specifically catalysing para ‐hydroxyl methylation of aromatic ring, they show a certain extent of substrate promiscuity and recognize a range of small molecule phenolics, including FA in addition to the monolignols (Table 1 ). Therefore, one possibility is that MOMTs, when over‐accumulated in planta, also methylate the accessible hydroxycinnamic acids besides monolignols. The resulting 4‐ O ‐methoxycinnamic acids are likely unable to be efficiently catalysed by the endogenous 4‐coumaroyl CoA ligases to transform to the active CoA thioesters, or the produced 4‐ O ‐methoxycinnamoyl CoAs can't be further utilized by BAHD family acyltransferases. Consequently, the availability of FA incorporating into the cell wall fraction is restricted (See the depiction in Figure S8 ). This assumption is consistent with the observation that both MOMT4 and MOMT9 exhibit detectable activities in methylating the examined FA (Table 1 ) and of the hyperaccumulation of the methanolic soluble 4‐ O ‐methylated FA metabolites in the MOMT transgenic plants (Figure 2 ). Interestingly, along with FA, the cell‐wall‐bound p CA was largely reduced in the MOMT transgenic plants (Figures 3 and 6 ) despite the 4‐ O ‐methylation activity towards p CA was barely detectable for MOMTs (Table 1 ). Given that cell‐wall‐bound p CA in grass cell walls is mainly originated from the incorporations of monolignol‐ p CA conjugates (i.e., sinapyl p ‐coumarate) in lignification (Karlen et al ., 2018 ), it is plausible that the reductions of cell‐wall‐bound p CA in the MOMT transgenic lines are mainly due to the disruptions of the incorporations of monolignol‐ p CA conjugates via 4‐ O ‐methylations of their monolignol moieties by MOMTs. Alternatively, the 4‐ O ‐methylation of monolignols might divert a large portion of metabolic flux away from the phenylpropanoid–lignin pathway, resulting in the overall reduction in the amount of both lignin biosynthetic precursors and pathway intermediates, including hydroxycinnamates, such as p CA and FA, thus limiting the availability of monomeric precursors for lignin polymerization and for wall‐bound phenolic ester synthesis. The productions of soluble 4‐ O ‐methylated FA and 4‐ O ‐methylated sinapate glucose esters suggest that rice has a versatile phenolic ester biosynthetic pathway. In Arabidopsis , a hydroxycinnamaldehyde dehydrogenase (aldehyde dehydrogenase, ALDH, also known as REF1) converts lignin precursors coniferaldehyde and sinapaldehyde into the corresponding hydroxycinnamates, FA and sinapate, respectively, which can then be transformed into the corresponding esters via conjugation with glucose, malate, or choline through the activities of specific glycosyltransferases and/or acyltransferases (Nair et al ., 2004 ). This intrinsic phenolic ester biosynthetic pathway exhibits considerable flexibility and able to accommodate and transform the 4‐ O ‐methylated monolignols or their aldehydes into the corresponding 4‐ O ‐methylated feruloyl and sinapoyl malate or choline in the MOMT4 transgenic Arabidopsis (Zhang et al ., 2012 ). Expectedly, the similar ALDH activity and the glucose ester biosynthetic pathway might also (partially) exist in the grass species, which are able to convert the 4‐ O ‐methylated monolignols (and/or the corresponding aldehydes) into the corresponding 4‐ O ‐methylated FA and sinapte, then to the corresponding glucose esters (Supplementary Figure 8 ). Tricin is the first flavonoid being recognized as an authentic lignin monomer involved in lignification in grasses (del Rio et al ., 2012 ; Lan et al ., 2015 ). It is part of the native lignin polymer in wheat ( Triticum aestivum ; (del Rio et al ., 2012 ; Zeng et al ., 2013 ), bamboo ( Phyllostachys pubescens ; Wen et al ., 2013 , maize ( Zea mays ) (Lan et al ., 2016a ), sugarcane ( Saccharum officinarum ; (del Río et al ., 2015 ) and rice (Lam et al ., 2017 ). Tricin incorporates into lignin polymer via 4′‐ O ‐β cross‐coupling with normal (non‐acylated) and/or acylated monolignols, and thus appears to function as a nucleation/initiation site for lignification (Lan et al ., 2015 , 2016b , 2018 ). Disturbing this initiation site might alter lignin content or structure, thus affecting cell wall digestibility. Indeed, some rice mutants deficient in tricin biosynthetic genes have reduced lignin content and/or altered lignin structures with enhanced saccharification efficiency (Lam et al ., 2017 , 2021 ), although the effects of the disruptions of tricin biosynthetic genes on lignin content and structure appeared to show some deviations (Lam et al ., 2022 ). The overexpression of both MOMT variants in rice resulted in notable reduction in both the methanol‐soluble and lignin‐bound tricin (Figures 2 and 6 ). Although it remains to be further determined how the expression of MOMTs suppresses tricin biosynthesis and its incorporation into lignin in transgenic rice, MOMTs exhibit discernable catalytic activities for the 4′‐ O ‐methyaltion of tricin and its biosynthetic precursors chrysoeriol and/or luteolin (Table 1 ). Such property is reminiscent of lignin biosynthetic enzyme COMTs in grass species, which has been demonstrated with dual functionality, methylating 5‐hydroxyconiferaldehyde/5‐hydroxyferulic acid for monolignol biosynthesis and the 3′‐ or 5′‐hydroxyl of luteolin, tricetin, and selgin, the precursors of tricin biosynthesis (Eudes et al ., 2017 ; Kim et al ., 2006 ; Lam et al ., 2019 ; Zhou et al ., 2006 ). Interestingly, the 4′‐ O ‐methylated flavones were not dominantly detected in MOMT transgenic lines. Although it remains to be further determined one possibility is that the 4‐ O ‐methylated phenolics ferualate/sinapate resulted from disturbance of monolignol biosynthesis and/or even the trace amount of 4′‐ O ‐methylated flavones might act as catalytic inhibitors to the tricin biosynthetic enzymes such as OMTs thus suppressing tricin overall biosynthesis. This assumption apparently is consistent with the observation that accumulation levels of tricin in both the soluble fraction and lignin‐bound fraction were substantially reduced in MOMT transgenic lines (Figures 2 and 6 ). Engineered plants with low lignin levels have enhanced release of cell wall sugars for biofuel production (Chen and Dixon, 2007 ). However, modifying lignin in transgenic plants can negatively affect the growth. This phenomenon has been referred to as lignin modification‐induced dwarfism (LMID), which can be attributed directly to the lack of lignin as a structural component required to support water transport, but LMID can also associate with the lack or hyperaccumulation of pathway intermediates or end products or the activation of cell wall integrity surveillance mechanisms (Muro‐Villanueva et al ., 2019 ). So far, the actual cause(s) of the LMID phenotype remain unknown (Muro‐Villanueva et al ., 2019 ). In the present study, both MOMT4 and MOMT9 genes were expressed in rice under the control of a strong lignin biosynthetic C4H gene promoter. It has been demonstrated that C4H promoter overperforms the constitutive cauliflower 35S promoter in driving expression cassette to manipulate lignin biosynthesis (Stewart et al ., 2009 ). The expression of MOMT4/9 resulted in substantial reduction of both G and S lignin content, the lignin‐bound tricin and the wall‐bound phenolics. It is not surprising that such drastic alterations in lignin biosynthesis and phenylpropanoid metabolism could render plant growth defects. The observation of more obvious dwarfism in asexually propagated transgenic offspring probably is due to the build‐up of toxic phenolic substances or the accumulation of detrimental effects from generation to generation. In addition to dwarfism, MOMT4/9 transgenic lines also failed in developing viable pollen grains and consequently were sterile. It has been documented that the formation of pollen wall exine (laid by the sporopollenin precursors) and pollen intine necessarily requires the participation of phenylpropanoid metabolites such as FA, p CA, p ‐hydroxybenzoate and G‐lignin units (Xu et al ., 2017 ; Xue et al ., 2020 ; Zhang et al ., 2023 ). The expression of MOMT4/9 driven by OsC4H promoter not only suppressed lignin formation but also substantially reduced cell wall FA and p CA. These chemical compositional changes might impair pollen wall formation thus causing the failure in pollen grain development. Previously C4H was shown to highly express in the tapetum layer of anthers and is involved in exine formation (Xue et al ., 2020 ). The C4H promoter driven MOMT gene expression might have rendered more severe disturbance on phenylpropanoid metabolism in anther tapetum layer/pollen wall. Several recent studies suggest that tailoring lignin biosynthesis in tissue‐ or cell type‐specific manner enables to disentangle lignin modification from perturbations in plant development (De Meester et al ., 2018 , 2021 ; Gui et al ., 2020 ). With validation of MOMT4/9 functioning in rice in the present study, the tissue‐specific expression of MOMTs, e.g., in fibers of grasses, might be able to effectively perturb lignin biosynthesis and phenylpropanoid metabolism while mitigating the deleterious effects on plant growth. Biomass to biofuel conversion technologies aim to access the much greater quantities of sugars in the less accessible structural carbohydrate fraction of plant biomass (Somerville, 2006 ), which benefits from modification of biomass chemistry (Chen and Dixon, 2007 ). The range of changes in lignin and wall‐bound phenolics in MOMT transgenic rice resulted in a measurable improvement of saccharification of the pretreated cell wall biomass. Glucose yield from the diluted acid‐treated cell wall residues of transgenic rice was enhanced up to 30% compared to the WT rice materials, which benefits the fermentable sugar‐based biofuel production. Interestingly, when raw biomass of rice stem was pre‐treated and digested under the same conditions, the improvement in saccharification efficiency was not easily perceived. This is probably due to the hyper‐accumulation of methanol extractable 4‐ O ‐methylated phenolics that might be toxic to the digestive enzymes. Nevertheless, the accumulation of facile extractable phenolics in transgenic biomass could facilitate lignin valorization process and benefit the biological upgrading of aromatics to the desired higher value end products such sustainable aviation fuels."
} | 6,758 |
31811154 | PMC6898677 | pmc | 3,552 | {
"abstract": "Illumina-MiSeq next-generation sequencing of ITS 5.8S rRNA gene demonstrated the transgenerational transmission of fungal seed-endophytes (mycobiome) across three consecutive wheat host generations under standard-control and drought conditions in the greenhouse. Drought-stressed plants experienced a positive shift in the seed mycobiome’s composition, moderated by the external acquisition of endophytic Penicillium (E+) at the seed level. Untreated (E−) and unstressed plants harbor a maximal fungal diversity of non-equilibrium ecological communities. While fungal composition in drought-stressed E− plants experienced important fluctuation, E+ plants maintained fungal ecological communities in phase equilibrium across generations. E+ plants hosted a relatively higher abundance of Ascomycota in the 2 nd and 3 rd seed generations of wheat, whereas higher abundance of Basidiomycota was detected in 1 st generation seeds. The dynamic response of ecological communities to environmental stress is conducive to E+ plants’ active recruitment of endosymbiotic consortia in seeds, benefiting host stress resilience and phenotype. In contrast, E− plants showed an erratic distribution of detected OTUs with an increased occurrence of phytopathogens and diminished plant performance under stress. The present study gives insight into the understanding of the seed-mycobiome composition and dynamics with the potential to improve plant host traits in an adverse environment.",
"introduction": "Introduction The seed is considered as a plant reproductive unit which harbors a diversity of fungal endophytes 1 . Seed-associated fungal microbiomes (mycobiomes) are ubiquitous and thus potentially important for plant growth and sustainable crop production 2 . Although there has been a wide acceptance of the beneficial role of plant-root endophytes in enhancing plant defense and host growth promotion systems, seed-borne endophytes have been poorly explored 2 , 3 . Recent studies have found that mycovitality or the endosymbiotic seed–fungus relationship can improve stress resilience in addition to providing prenatal care to plant 3 via coleorhiza (initial root)-specific dormancy release and enhanced biological seed stratification 4 . Also, there is evidence that some seed endophytes can influence plant post-germination phenophases 3 , 5 , 6 due to the presence of fungal microbiome. Hence, it is foreseeable that seed-fungal microbiome coevolution directed towards interactive plant physiology could influence the host’s growth and stress mitigation 7 under global warming 8 , 9 . Many fungi are seed-borne endophytes and plant pathogens, henceforth, the seed is considered as an important vector in transmitting mycobiomes influencing plant host evolution as well as crop domestication over the last 10,000 years 10 . Endophytic fungi reveal a broad variation in their mode of transmission from one host to another 2 and allow vertical transmission from one generation to the next 11 . Thus, it becomes essential to generate new knowledge and provide deeper insights into the beneficial seed-borne mycobiome, as well as stress-mitigating mechanisms in symbiotic plants having a possible translational effect on crop productivity 7 . Frequent drought stress particularly limits the performance of agriculture crops 12 that may also negatively affect fungal and bacterial composition in stressed plants 13 , 14 . However, little is known about how seed selects mycobiome communities to establish efficient symbiotic associations in securing seed germination, plant growth, and reproduction, particularly under climate change 15 . The plant’s growth and crop reproduction success might be the result of genotype adaptation and associated evolutionary traits modulated by frequent shifts in host-associated fungal and bacterial microbiomes 16 under fluctuating environmental conditions. Under continuous pressure of increasing climatic alterations, it becomes imperative to better understand plant-microbiome relationships 7 and to unravel the role of fungal microbiomes in conferring plant protection against abiotic stresses. Particularly, drought is limiting climatic factor to plant physiology and adaptation that might be palliated by the external acquisition of fungal endosymbionts at the seed level. Fungal endosymbionts control seed vitality via mycovitality, improved hydrothermal time of germination and myco-stratification 4 , 5 , 17 to initiate plant growth promotion (PGP) under stress conditions 3 , 18 . Hence, the fungal endophytes with PGP trait bear tangible benefits to plant host functioning 19 . However, it is still unknown whether reintroduced PGP can induce shifts in plant mycobiome composition to improve host resistance in coping with adverse changing climate 20 , 21 . Therefore, characterizing the innate seed-mycobiome of Triticum with and without the external (horizontal) re-acquisition of the PGP fungus under drought stress can be a crucial step towards the understanding transgenerational transmission of seed endophytes and host-fungi interactions under climate change. While the factors that drive plant microbiome composition are now better understood at the host-root level 22 , a major challenge for seed microbiology is to link microbial community composition and dynamics with ecological functions. As many fungi are seed-borne, we hypothesized that native endophytic seed fungal microbiome might be greatly influenced by local-drought environment and biological-acquisition of new fungal endophyte into seed conditions. The goal of this study was to figure out whether the stress-induced seed mycobiome profile may be an important effector of the plant adaptive phenotypes relative to transgenerational seed fungal composition. The Next Generation Sequencing (NGS) technology was employed for mapping core seed endophyte mycobiomes in wheat including host-mediated and subsequent-generations mycobiome selection under drought stress modified by Penicillium inoculation into seed. To understand complex interactions among these factors in seed, three specific hypotheses have been postulated. Firstly, the early inoculation of beneficial fungal endosymbiont at the first-generation wheat seed could result in the dynamically equilibrated flow of mycobiome composition across generations. Secondly, the transgenerational re-establishment of the seed-associated fungal communities depicts important points in mycobiome selection under stress towards the climate-resilient plant. Thirdly, seed endophytic consortia could play multiple roles in plant growth and fitness including host phenotypic traits.",
"discussion": "Discussion The true motivation to study the seed-inhabiting mycobiome lies on the particular fungal traits to be transmitted vertically to the progeny in promoting host-specific germination and improved transgenerational host phenotypes. In the present study, we depicted some temporal dynamics of the seed-born fungal endophytic communities after drought disturbance events. It appears that the more resilient community recovers to its initial state after perturbation, particularly when the newly reintroduced fungal endophyte PGP - Penicillium was acquired into first-generation seed. The resilient fungal community members were recruited from the environmental/native microbial pool and so in a sustainable way across generations. The present study revealed the mycobiome transgenerational transfer as an integral part of the evolution of the seed-micro ecosystem. According to Orozco-Mosqueda et al . 32 , the predictability of the multi-generation inheritance of endophytic genome components in plants is one of the major challenges involved in the use of endophytes. In that regard, we discovered that if the plant growth benefits are provided early at the seed germination level by the endophytic inoculum, the resilient holobiome (a host plus its microbiome acting alongside as an ecological community of organisms) may still be able to enhance crop productivity under stress similar to unstressed plants across generations. To mitigate stress while not compromising plant growth and yield has been suggested to be an ideal strategy. To the best of our knowledge, this is the first report of the microbiome shift in the presence of endophyte with desired seed ecological characteristics showing the potentially significant impact on host phenotypical traits under drought stress. These findings reinforce the concepts of holobiont-mediated adaptation via host-mediated and microbiome-mediated selection processes working in a concert to better respond to external environmental factors. The modern concept of evolution considers holobiont and its hologenome-composed of plant genome and associated microbiomes 33 . Hence, multicellular organisms can no longer be considered individuals by the classical definitions of the term 34 . Microbial symbionts, including multicellular fungi, can be reinstalled or acquired by plant and transmitted from parent to offspring by horizontal and vertical pathways using vectors and the environment. Several studies have reported that these symbionts contribute to the anatomy, physiology, development, innate or adaptive immunity, and behavior 35 , 36 , thus leading to genetic variation, origin, and evolution of species 37 . It seems that the acquisition of microbes and microbial genes is a powerful mechanism for driving the evolution of complexity. Evolution of the human microbiome proceeds via both cooperation and competition working in parallel 38 , which could be a conceivable way to drive a structure of microbiomes in the plant. Indeed, the MiSeq Illumina sequence data depicted significant changes in both wheat resilience and phenotype/biomass which coincide with changes in the composition and diversity of the seed endophytic fungal communities. In some instances, endophytic fungal symbionts undergo mixed horizontal and vertical transmission, also named an “imperfect transmission” when the symbiont is not transmitted to all offspring 39 . In this study, MiSeq-ITS sequences analyses depicted a highly dynamic process in fluctuating fungal endophytes over three seed generations. The results also revealed a continuum in the transmission of fungal communities from the first seed generation to subsequent second and third generations, also implying a selective systemic fungal transmission. Using the ITS sequence-based approach, Ganley and Newcombe 40 were able to discriminate the transmission of diverse fungal endophytes in coniferous seeds. The results of the present study revealed that seed-endosymbiotic fungal mutualism over the generations was greatly affected by both plant- Penicillium sp. acquisition and plant-drought exposure regimes. The diversity of seed endophytes under drought or shortage of water was mediated by the Penicillium inoculant. Notably, the occurrence of the Trichomaceae including Penicillium, Aspergillus , and Paecilomyces predominated over generations, particularly in inoculated drought exposed seeds. In contrast, Trichomaceae taxa were rare in first-generation standard-control seeds. According to Barret et al . 41 a plant’s genotype has a strong effect on the dynamics of the seed microbiota. An increased presence of the Penicilllium inoculant on seed could induce a shift in concurrence of r- vs. K-strategists 42 due to the possible effect of Penicillium on the fungal microbiota’s structure during seed germination and seedling emergence. This fungal inoculant has a fast-growing ability (r-strategist) that may explain its rapid colonization of seedlings. Inoculated and drought exposed G1 seeds showed an initial presence of Penicillium inoculant, and the predominance of Cladosporium sp., a member of the Davidiellaceae (Fig. 6 ). The maximum shift of Penicillium over Cladosporium was detected in G2 drought-stressed generation. It resulted in re-equilibrated Penicillium : Cladosporium dynamic state associated with G3 drought-stressed generation. Despite that equilibrium, the diversity of Penicillium species increased from drought-stressed G1 (2-species including Penicillium sp.-inoculant) to drought-stressed G2 and G3 (7-species including Penicillium sp.-inoculant) generations. In the same time, the number of Cladosporium, Aspergillus , and Paecilomyces species remained stable, 1–2 taxa, throughout all three drought-stressed generations. This phenomenon supports the first hypothesis that the external inoculant recruitment could dictate a process of both endosymbiont transmission and retention thus equilibrating composition in the wheat seeds over generations. In addition, several studies showed evidence of vertical transmission of seed associated endosymbionts up to the second generation of the host from different plant organs via vascular connections or through gametes (systemic entry) 43 – 47 . Further, the seed allows coexistence of various fungal endopsheric consortia. When exposed to the environmental stress/stressors, the fungal taxa (species, guilds and communities) can alter in diversity and abundance. It also appears that the Penicillium inoculant can modulate abundance of different functional groups of seed-born fungi. Its increased abundance coincides with the reduction of certain common molds ( Alternaria and Aspergillus ) as well as phytopathogens such as Colletotrichum (anthracnose), Fusarium (crown rot and head blight) and Staganospora (leaf spot). Seed is an environmental niche which creates conducive environment for hosting various functional groups of fungi. Nevertheless, the protective effect of the inoculant against molds/phytopathogens may be related to Penicillium ’s competitive ability for occupying most of the seed microenvironments. In terms of the biocontrol consortia, Acremonium / Sarocladium and Trichoderma species appeared to gradually increase abundance from G1-inoculant to G3 inoculant treatment. The highest biocontrol abundance coincides with the lowest abundance of the Giberella / Fusarium across all generations. This finding may open a new avenue to better understand the influence of the endophytic inoculant on seed mycobiome as well as transmission and persistence of the fungal species, and consortia from generation to generation. Subsequently, results may encourage plant microbiologists to work more closely with plant ecologists, pathologists and breeders to take advantage of these findings. Wheat preferentially chose distinct endophytic taxa into seeds depending on the type of the environmental stresses. Kingdom Fungi provided all phyla with a predominance of endophytic Ascomycota and Basidiomycota occupying wheat seed followed by Incertae sedis , Chytridiomycota , and Zygomycota . Chytridiomycota and Zygomycota represent the earliest terrestrial divergences from ancestral aquatic fungi 48 . Typical environmental conditions, optimal for plant growth, dictate normal or Gaussian fungal distribution in the seed. Only PGP-inoculum treatment combined with drought stress continuously increased the presence of Ascomycota . This phylum showed an increased cumulative proportion over time, while Basidiomycota preferentially displayed increasing proportion and so exclusively related to the combination of inoculant and drought conditions. The endmost indicates the existence of the three most equilibrated states in dynamic of seed-mycobiome communities, named “standard”- control, “induced”- drought stress alone or inoculant acquisition alone, and “combined” inoculant plus drought intervention. The dynamic changes in both fungal richness and evenness induced by “combined” intervention seemed to induce a substantial shift in host-mycobiome interactions to reinstate host resilience and phenotypic trait. The induced stresses followed by the alteration in fungal phylogenetic groups from the r -growers to the K-growers 49 also indicate a possible fluctuation in seed nutrients throughout generations. It has been reported for human gut microbiome that fungal diversity correlated with diet, nutrient values in the habitat, as well as associated bacterial inhabitants 50 . The fungal distribution in standard-control plants served as initial diversity or a pool of non-equilibrated microbial communities dispersed over generations. Imposed drought stress and symbiont inoculation moved the overall system toward asymmetric left-curve in a new preferential equilibrium (Fig. 3 ) and thus a more desirable system for seed when exposed to the biotic or abiotic external influences. Consequently, the asymmetric movement of the sequence values in seed endophytes influenced by biotic and abiotic factors was presumably correlated with the abundance of two fungal E (1 and 2) and B (2) clusters or communities (Fig. 2 ). They were detected in G1-G2-inoculant and inoculant plus drought exposed seeds characterized by positive skewness fungal distribution from the normal-symmetry of the curve. Interestingly, standard-control seeds were not harbored the majority of those fungal taxa and its mycobiome’s curve followed a normal frequency distribution model. It indicates the shift in specific taxa selection occurred in response to the tested environmental influences. For instance, the members of family Pleosporaceae shifts in drought stress related diversity and abundance (Supplement Fig. 3S ). This family contains both endophytic fungi and opportunistic plant pathogens, such as Alternaria, Bipolaris,Curvularia, Exserohilum, Pleospora and Pyrenophora , which requests further investigation and proper explanation 51 . Similarly, this can be applied on the members of family Davidiellaceae , i.e. genus Cladosporium/Davidiella might be considered as potential biocontrol ( C. cladosporioides , C. tenuissima and C. uredinicola ) against wheat diseases on the mature plants, although it is often categorized (as in this paper) as a potential mold on the seed level. Further, some F (1) and F (2) fungal groups were undetectable in G3-generation seeds (Fig. 2 ) regardless of biotic and abiotic influences as possibly stress-unspecific taxa. However, both specific and unspecific direction changes in the distribution of the fungal taxa may be the result of the cross-talk between fungi, microbes, and seed host (occupied plant ecological niche) towards phase equilibrium in mycobiome composition and dynamics over time. Hence, our second hypothesis supports the prediction of the changes in the outcome distribution of the fungal OTUs 52 , based on MiSeq Illumina NGS technology 53 that facilitates combining information from phylogenetic branching events and bell-curves, as an effective means to find out how fungal communities behave to adapt and disperse over plant-seed generations. In this study, the motivational direction of the third hypothesis was to verify the potential phenotypical changes in wheat during drought exposure with and without endophytic intervention. Current results indicate that the persistence of the introduced endophytic inoculant over three seed generations has an influence on the mycobiome composition towards an improved plant biomass production under drought conditions rather than standard-control growth conditions (Fig. 5 ). It seems that the phenotype of the symbiotic (E+) G1 plant response to drought persisted beyond the first seed generation. The phenotypic changes were typically associated with preferential endophytic consortia transferred via seed which can assist the altered wheat traits over generations. It implies that plants under stress may actively recruit members of the mycobiome community for steady state systems where substantial insight into energy balance and plant metabolic rates can be achieved 21 . Selection of suitable fungal endosymbionts 54 – 56 promote plant phenotype and ameliorate resilience to water scarcity sustaining fungal ecological niche. Some possible mechanisms driving drought tolerance may be linked to reduced stomatal conductance 57 and favorable biochemical processes leading to balanced water regime and equilibrated nutritional status of the plant 58 – 60 . Moreover, the observation by Molina-Montenegro et al . 61 is in a partial agreement with our findings. They reported that the net photosynthesis, fresh and dry biomass production, as well as water use efficiency were significantly higher in the treatments with the presence of inoculated root-endophytes at 25% reduced water regimes, compared to endophyte-free lettuce cultivar. Further research in that direction on seeds from different host genotypes and throughout world’s ecoregions is highly merited. Our results provide important insights into the dynamics of seed endosymbionts over generations, as a possible regulatory mechanism of plant phenotypic adaptation and resilience under an external constraint. In summary, the evolutionary mechanisms of seed microbiome distribution and dynamics can be exploited for the selection and implementation of endophytic microbial inoculants, specific to the plant genotype(s) and environmental condition(s). Some endophytes appeared to be commonly distributed, others limited to biotic and abiotic influences in one or more seed generations. The ITS 5.8S gene distribution curve model depicted a distinctive shift in a fine scale of the core seed mycobiome variations consisting of native fungal taxa. It allows appreciating an important and intriguing role of endophyte horizontal acquisition in seed via artificial seed coating further shaping the recruitment and dynamic of native endosymbionts over generations. The ever-changing dynamics of seed mycobiome seem to be driven simultaneously by inoculant and drought influences resulting in subsequent changes in plant phenotype and biomass production. However, actual mechanism and driving factors in inoculant transmission process over seed generations and subsequent role in triggering biomass accumulation under water-stressed plant tissue conditions were not clearly understood. The manipulation of the mycobiome by re-introducing an endophytic strain in plant system, such as endophytic Penicillium at seed germination stage 58 , is apparent. It provides an opportunity to promote specific consortia of fungal endophytes in seed as reproductive unit coping with climate change. The occurrence of the stress-induced mycobiome profile may be an important effector in reaching seed phase equilibrium maintained under drought stress across generations. Although no causal relationship between different observed parameters was fully established, a considerable correlation between the structure and dynamics of the seed-endosymbionts, transmitted fungal inoculant to seed and shifted phenotypic changes were unveiled. These changes resulted in improved plant biomass and resilience under abiotic stress. In conclusion, the new seed mycobiome taxa discovered herein will spur further questions associated with the functional dynamics and genetic network of host-endophyte interactions under stressed conditions. Our current understanding-based on transgenerational assessment of seed-mycobiome relationships is built on research experiments under controlled conditions; it warrants further experimentation in field settings for a deeper insight into the practical aspects of sustainable food production and new breeding programs for symbiotic fungal endophytes in staple crops."
} | 5,869 |
39758973 | PMC11700267 | pmc | 3,554 | {
"abstract": "Highlights • Microalgae's multifaceted role in agroecosystems explored. • Interdisciplinary insights integrated for comprehensive understanding. • Emerging trends in microalgae technology identified and analyzed. • Challenges and opportunities critically assessed for practical implications.",
"introduction": "1 Introduction The surging global demand for food presents formidable challenges to agricultural systems worldwide, urging a transition towards sustainable methodologies to curb environmental degradation and uphold enduring food security [ 1 , 2 ]. Conventional agricultural practices, marked by excessive use of chemical inputs, monocropping, and deforestation, exacerbate soil erosion, loss of biodiversity, and contamination of water bodies [ 3 , 4 ]. The extensive application of synthetic fertilizers and pesticides compounds these environmental woes, with fertilizer runoff causing eutrophication and consequent harm to aquatic habitats [ 5 , 6 ]. In addressing these pressing concerns, sustainable agricultural techniques emerge as indispensable for combating water pollution, rectifying ecosystem imbalances, and mitigating biodiversity decline [ 7 ]. Microalgae have emerged as a promising solution for sustainable agriculture, offering unique advantages such as efficient nutrient extraction from wastewater and potential as a biofertilizer [ 1 , 5 ]. Harnessing the properties of microalgae, particularly their ability to thrive on wastewater and extract nutrients, presents a dual solution for wastewater treatment and soil fertility enhancement in agricultural ecosystems. The exploration of microalgae in agriculture dates back to the 1960s, with research indicating their potential to enhance soil fertility through the activity of micronutrients and metabolites [ 8 , 9 ]. Microalgae-derived materials contribute to improved soil structure and fertility, containing a rich source of macronutrients and biologically active compounds [ 10 , 11 ]. Such advancements underscore the significance of microalgae as a sustainable resource in the quest for eco-friendly agricultural practices. Furthermore, microalgae-based biofertilizers or biostimulants offer promising avenues for sustainable agriculture by promoting plant growth and enhancing soil fertility [ 12 , 13 ]. Specific strains, such as Chlorella vulgaris, Scenedesmus quadricauda, Desmodesmus subspicatus and Spirulina platensis , are particularly effective in absorbing nitrogen, phosphorus, and trace minerals from various wastewater sources, including agricultural runoff and municipal wastewater, transforming pollutants into a nutrient-rich biomass [ 5 ]. This biomass can then be applied to soils, enhancing fertility, and supporting plant growth, as it contains essential nutrients along with bioactive compounds like amino acids, phytohormones, and vitamins [ 1 ]. For instance, applying Scenedesmus quadricauda, Spirulina platensis and Chlorella vulgaris to beetroot resulted in beneficial effects on root architecture, such as increases in root length and lateral root number, which in turn increased the root surface area available for nutrient uptake [ 14 ]. In a greenhouse, Desmodesmus subspicatus aqueous extracts and lyophilized biomass boosted germination in vitro and sped up development during the transplanting and acclimatization phase [ 15 ]. The exopolysaccharides that Spirulina platensis releases into the soil also serve the crucial purpose of sequestering sodium and metal ions, which lowers their uptake by maize plants and promotes their growth in saline or contaminated soils [ 16 ]. As part of the Sustainable Development Goals (SDGs), sustainable agriculture plays a crucial role in addressing goals related to food security, terrestrial ecosystem preservation, and water quality [ 17 , 18 ]. However, meeting the growing global food demand while upholding sustainability goals remains a complex challenge, necessitating innovative solutions like microalgae-based agriculture [ 19 ]. The integration of microalgae into agricultural practices holds promise not only for increasing crop productivity but also for mitigating environmental impacts, thus aligning with the overarching objectives of sustainable development. The existence of microalgae as a biofertilizer can be observed in Fig. 1 , illustrating the documents compiled from the Scopus database. Employing a meticulous search criterion focusing on documents featuring the terms \"microalgae\" and \"biofertilizer\" in \"all fields,\" and refining the search to review papers and articles, the research team systematically collected data. This rigorous selection process aimed to provide a targeted and comprehensive analysis of the literature, offering insights into the relationship between microalgae and biofertilizers within the academic discourse. Fig. 1 Documents contained keywords “microalgae” and “biofertilizer” collected from scopus database. Fig 1 The compiled data, spanning from 2000 to 2023, showcases a progressive increase in research output, with the number of documents rising from 2 in 2000 to a peak of 466 in 2023. This upward trajectory reflects the growing scholarly interest in the topic, indicating sustained and heightened attention to the exploration of microalgae as a biofertilizer. With a total of 1,834 documents found, the thorough analysis of this data, grounded in the specificity of search terms and document types, adds credibility and rigor to the meta-analysis, establishing a robust foundation for understanding the evolving landscape of microalgae-based biofertilizer research within the academic domain. Despite their potential, challenges such as the high cost of nutrient provision for microalgae cultivation hinder their widespread adoption in agriculture [ 20 , 21 ]. Addressing these challenges and exploring cost-effective strategies for integrating microalgae into agricultural practices are crucial steps towards realizing their full potential as sustainable and resilient components of agricultural ecosystems. This review is dedicated to evaluating the opportunities and challenges inherent in leveraging microalgae as a biofertilizer in agroecosystems. With the rising global food demand and the environmental degradation caused by the overuse of chemical fertilizers, there is a pressing need for sustainable alternatives. Current agricultural practices contribute to soil depletion, eutrophication, and biodiversity loss, challenging long-term food security. Despite its potential as a biofertilizer, the widespread use of microalgae in agriculture faces significant hurdles such as high production costs and scalability issues. This review aims to explore the potential of microalgae as a biofertilizer and its role in phycoremediation for wastewater treatment in agroecosystems. It evaluates both the opportunities and the challenges that hinder large-scale implementation, providing a comprehensive assessment of the current landscape and potential solutions to promote sustainability in agriculture."
} | 1,754 |
37101533 | PMC10124587 | pmc | 3,556 | {
"abstract": "Lignin is the dominant aromatic renewable polymer on earth. Generally, its complex and heterogeneous structure hinders its high-value utilization. Catechyl lignin (C-lignin), a novel lignin discovered in the seed coats of vanilla and several members of Cactaceae, has received increasing attention due to its unique homogeneous linear structure. Obtaining substantial amounts of C-lignin either by gene regulation or effective isolation is essential to advance C-lignin's valorization. Through a fundamental understanding of the biosynthesis process, genetic engineering to promote the accumulation of C-lignin in certain plants was developed to facilitate C-lignin valorization. Various isolation methods were also developed to isolate C-lignin, among which deep eutectic solvents (DESs) treatment is one of the most promising approaches to fractionate C-lignin from biomass materials. Since C-lignin is composed of homogeneous catechyl units, depolymerization to produce catechol monomers demonstrates a promising way for value-added utilization of C-lignin. Reductive catalytic fractionation (RCF) represents another emerging technology for effective depolymerizing C-lignin, leading to a narrow distribution of lignin-derived aromatic products ( e.g. , propyl and propenyl catechol). Meanwhile, the linear molecular structure predisposes C-lignin as a potential promising feedstock for preparing carbon fiber materials. In this review, the biosynthesis of this unique C-lignin in plants is summarized. C-lignin isolation from plants and various depolymerization approaches to obtaining aromatic products are overviewed with highlights on RCF process. Exploring new application areas based on C-lignin's unique homogeneous linear structure is also discussed with its potential for high-value utilization in the future.",
"conclusion": "Conclusion and outlook C-lignin is receiving increasing attention recently due to its homogeneous linear structure and narrow monomer product distribution through depolymerization. Expanding C-lignin feedstock sources and developing green and efficient extraction methods hold promise for promoting C-lignin valorization. The hope for reducing the cost of C-lignin feedstock is to regulate the biosynthesis of C-lignin in plants by genetic means ( Fig. 6 ). Genetic engineering to regulate C-lignin synthesis requires systematic investigation of plant methyl homeostasis and caffeyl alcohol polymerization mechanism to address the issue that suppression of COMT and CcOAOMT genes affects plant growth and reduces total lignin contents. Traditional lignin isolation methods are not effective for C-lignin. Usually, hydrophilic polar solvents can be used to extract C-lignin from seed coats. DES treatment represents an effective method to isolate C-lignin, but the interaction between DES and C-lignin during the isolation process remains unclear. More efficient isolation methods need to be explored to provide C-lignin. Fig. 6 Summary of C-lignin biosynthesis, isolation, depolymerization, and application. Currently, the utilization of C-lignin is mainly metal-catalyzed hydrolysis to produce catechol monomers and their derivatives. The narrow monomer product distribution makes C-lignin promising for producing fine chemicals. In addition, the good acid and thermal stability as well as the linear structure of C-lignin provides unique strength for the development of new polymeric materials, such as carbon fibers. The catechol compound is one of the key intermediates during the lignin bioconversion. The structure of depolymerized monomers from C-lignin is much like the catechol molecule, suggesting that C-lignin may be more suitable for biological valorization compared to the traditional G/S-type lignin. Therefore, the application of C-lignin in biotransformation holds promise for enhancing the biological lignin valorization performance. Exploration of a broader range of downstream applications is essential and promising for the future high-value utilization of C-lignin.",
"introduction": "Introduction Lignin is a complex polymer that widely exists in various types of plants in nature. Its abundant functional groups ( e.g. , hydroxyl and carboxyl groups) and aromatic nature offer great potential for high-value utilization. 1 Generally, plants transform phenylalanine through aromatic hydroxylation and O -methylation to produce lignin monomers with different degrees of methoxylation, which are classified into syringyl units (S), guaiacyl units (G), and p -hydroxyphenyl units (H). These lignin monomers are conjugated through various ether or carbon–carbon bonds to form lignin macromolecules. In recent years, the application areas of lignin have been expanded to adsorbents, 2 fertilizers, 3 epoxy resin curing agents, 4 lipids, 5 polyhydroxyalkanoates (PHA), 6 and polyurethanes, 7 etc. Despite these developments, it is estimated that only 2% of the total industrial lignin stream is currently used for preparing derivative products, while the major part is subjected to combustion or abandoned in landfills. The low efficient utilization of lignin mainly owes to its complex and inhomogeneous molecular structure. 8,9 It has been reported that various phenolic compounds are possible substrates that can be transformed into lignin units that are different from those typical lignin subunits ( e.g. , S, G, and H). Ralph and coworkers elucidated the presence of a COMT (caffeic acid O -methyltransferase) defect poplar by NMR analysis of the lignin. 10 In 2012, Chen and coworkers detected the presence of a novel lignin, catechyl lignin (C-lignin), in the seed coats of Vanilla planifolia and several members of the Cactaceae ( e.g. , Melocactus obtusipetalus ). 11 C-lignin is formed by the free coupling of oxidized radicals, resulting in a linear polymer composed of caffeyl alcohol ( Fig. 1 and 2 ). 11–14 During plant growth, the lack of O -methyltransferase (OMT) activity leads to the selective formation of caffeyl alcohol monomers. The C 5 –OH in caffeyl alcohol facilitates monomer coupling with β- O -4 radicals to form intramolecular closed loops, resulting in homopolymers of C-lignin without condensation units (almost exclusively linked by benzodioxane bonds). 15–17 Compared to the typical G/S lignin, C-lignin has a lower molecular weight, probably due to the weak polymerization ability of caffeyl alcohols compared to G/S lignin units. 11,13 According to the analysis of the 3D structure by all-atom molecular dynamics simulation, C-lignin was found to be more dense and rigid. 18 In addition, C-lignin shows good acid stability due to the stable benzodioxane structure. 17 Fig. 1 Proposed structure of H/G/S lignin versus C-lignin revealed by NMR analysis. 15"
} | 1,682 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.