pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
37635855 | null | s2 | 1,826 | {
"abstract": "A variety of artificial silk spinning approaches have been attempted to mimic the natural spinning process found in silkworms and spiders, yet instantaneous silk fiber formation with hierarchical structure under physiological and ambient conditions without post-treatment procedures remains unaddressed. Here, we report a new strategy to fabricate silk protein-based aerosols and silk fibers instantaneously (< 1 s) "
} | 104 |
29158553 | null | s2 | 1,828 | {
"abstract": "Mutualistic networks have been shown to involve complex patterns of interactions among animal and plant species, including a widespread presence of nestedness. The nested structure of these webs seems to be positively correlated with higher diversity and resilience. Moreover, these webs exhibit marked measurable structural patterns, including broad distributions of connectivity, strongly asymmetrical interactions and hierarchical organization. Hierarchical organization is an especially interesting property, since it is positively correlated with biodiversity and network resilience, thus suggesting potential selection processes favouring the observed web organization. However, here we show that all these structural quantitative patterns-and nestedness in particular-can be properly explained by means of a very simple dynamical model of speciation and divergence with no selection-driven coevolution of traits. The agreement between observed and modelled networks suggests that the patterns displayed by real mutualistic webs might actually represent evolutionary spandrels."
} | 270 |
27418359 | PMC5119850 | pmc | 1,829 | {
"abstract": "The selection of microbes by enrichment on plant biomass has been proposed as an efficient way to develop new strategies for lignocellulose saccharification. Here, we report an in-depth analysis of soil-derived microbial consortia that were trained to degrade once-used wheat straw (WS1-M), switchgrass (SG-M) and corn stover (CS-M) under aerobic and mesophilic conditions. Molecular fingerprintings, bacterial 16S ribosomal RNA (rRNA) gene amplicon sequencing and metagenomic analyses showed that the three microbial consortia were taxonomically distinct. Based on the taxonomic affiliation of protein-encoding sequences, members of the Bacteroidetes (e.g. Chryseobacterium , Weeksella , Flavobacterium and Sphingobacterium ) were preferentially selected on WS1-M, whereas SG-M and CS-M favoured members of the Proteobacteria (e.g. Caulobacter , Brevundimonas , Stenotrophomonas and Xanthomonas ). The highest degradation rates of lignin (~59 %) were observed with SG-M, whereas CS-M showed a high consumption of cellulose and hemicellulose. Analyses of the carbohydrate-active enzymes in the three microbial consortia showed the dominance of glycosyl hydrolases (e.g. of families GH3, GH43, GH13, GH10, GH29, GH28, GH16, GH4 and GH92). In addition, proteins of families AA6, AA10 and AA2 were detected. Analysis of secreted protein fractions (metasecretome) for each selected microbial consortium mainly showed the presence of enzymes able to degrade arabinan, arabinoxylan, xylan, β-glucan, galactomannan and rhamnogalacturonan. Notably, these metasecretomes contain enzymes that enable us to produce oligosaccharides directly from wheat straw, sugarcane bagasse and willow. Thus, the underlying microbial consortia constitute valuable resources for the production of enzyme cocktails for the efficient saccharification of plant biomass. Electronic supplementary material The online version of this article (doi:10.1007/s00253-016-7713-3) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction Plant biomass is an important source of energy that is stored in the form of complex polysaccharides, primarily hemicelluloses and cellulose. The transformation of these polymers into sugars enables downstream applications such as the production of biofuels. The saccharification process is currently carried out by (thermochemical) pretreatment followed by the use of a mixture of microbial enzymes (e.g. lytic polysaccharide monooxygenases, xylanases, arabinofuranosidases, cellobiohydrolases, endoglucanases and β-glucosidases) that can work synergistically (Meyer et al. 2009 ; Hasunuma et al. 2013 ). Plant waste sources that are used for the production of second generation of biofuels include agricultural by-products (e.g. sugarcane bagasse), wood residues and non-food energy crops, such as switchgrass. Such are attractive as they do not seem to compete with food production (Sims et al. 2010 ; Limayem and Ricke 2012 ). The leading industrial source of cellulase cocktails is Trichoderma reesei . Several strains exist and their secretomes have been widely used to develop commercial cocktails for plant biomass hydrolysis (e.g. Celluclast 1.5 L, Cellic CTec2 and HTec2 from Novozymes). However, T. reesei secretomes are dominated by cellobiohydrolases (CBHs) and endoglucanases, with only low quantities of xylanases, lytic polysaccharide monooxygenases (LPMOs), and β-glucosidases being produced. Hence, addition of such enzymes is thought to improve the hydrolytic efficiency (Mohanram et al. 2013 ). For instance, Gao et al. ( 2011 ) showed that the addition of defined hemicellulases (e.g. β-xylosidases, α-arabinofuranosidases and α-glucuronidases) from Clostridium thermocellum , Geobacillus thermodenitrificans , Geobacillus stearothermophilus and Dictyoglomus turgidum , to a core cellulase cocktail from T. reesei and Aspergillus niger , enhances the saccharification of pretreated corn stover. Typically, in biorefinery processes, Celluclast 1.5 L (1,4-(1,3:1,4)-β-D-glucan 4-glucano-hydrolase) is supplemented with a β-glucosidase from A. niger (Merino and Cherry 2007 ). Moreover, Cellic CTec2 includes cellulases, high levels of improved β-glucosidases with less glucose inhibition, hemicellulases and LPMOs. In industry, it is recommended to dose the Cellic CTec2 in accordance with the level of cellulose in the substrate. If (pretreated) plant biomass contains an appreciable amount of hemicellulose, it is advised to combine Cellic CTec2 with HTec2 (endoxylanases) to boost cellulose hydrolysis (Cannella and Jørgensen 2014 ; Rodrigues et al. 2015 ). Given the complexity of the required enzymes, efficient plant biomass hydrolysis by microbial consortia, instead of single strains, has been proposed (Cheng and Zhu 2012 ). One disadvantage of this strategy is that the monosaccharides released from plant biomass are often rapidly assimilated by co-occurring microorganisms. To overcome this hurdle, extracellular enzymes may be harvested from the microbial consortia and applied directly onto the plant biomass (Gladden et al. 2011a ; Park et al. 2012 ). Enrichments of lignocellulolytic microbes from soils have been performed with switchgrass (SG), wheat straw (WS) and corn stover (CS) as the sole sources of carbon (DeAngelis et al. 2013 ; Jiménez et al. 2014a ; Brossi et al. 2015 ). Such plant biomass is known to not only contain recalcitrant polysaccharides, but also (easily degradable) small soluble substrates (e.g. oligosaccharides). These increase the proliferation of opportunistic microorganisms that cannot deconstruct the lignocellulosic structures. To remove such soluble substrates, washes of the plant biomass with water and ethanol have been proposed (Gladden et al. 2011a ). Moreover, biological pretreatment can be based on living organisms or on enzyme cocktails. The former is exemplified by the use of white-rot basidiomycetes such as Phanerochaete chrysosporium and Trametes versicolor (Pinto et al. 2012 ; Wan and Li 2012 ). The latter makes use of commercial enzyme cocktails (as explained earlier). However, biological pretreatments using (enzymes from) microbial consortia offer alternatives that have so far been poorly explored. Metagenomics- and metatranscriptomics-based approaches have been increasingly used to study lignocellulolytic microbial consortia (Wongwilaiwalin et al. 2013 ; Simmons et al. 2014 ). Comparison of metagenomic sequences with data stored in the “Carbohydrate-Active Enzyme database” (CAZy) (Lombard et al. 2014 ) allows for evaluation of the metabolic potential in the deconstruction of plant polysaccharides. Recently, Jiménez et al. ( 2015a ) unveiled such potential in two microbial consortia selected on wheat straw. Significant enrichments of genes encoding GH2, GH43, GH92 and GH95 family proteins were found. In taxonomic terms, the genes were mostly affiliated with those present on the genomes of Sphingobacterium , Bacteroides , Flavobacterium and Pedobacter species. Here, we used an enrichment process in two stages, i.e. (1) enriching biodegrader soil-derived microbial consortia on wheat straw, switchgrass and corn stover (Brossi et al. 2015 ) and then (2) re-using the partially degraded substrate as the carbon source for a second growth step with the same microbial consortia. We hypothesised that the once-used plant biomass specifically selected for microbes with high capacities to degrade the more complex plant polysaccharides as well as lignin. We thus presumed the biological pretreatment removed the easily degradable substrates from the three plant biomass materials and studied how the microbial consortia changed along the two steps in the enrichment process. The main aim of this study was to characterize these selected “second-phase” microbial consortia by lignocellulose consumption profiles, metagenomics (taxonomic and CAZy profiling) and extracellular enzymatic activities using a new generation of versatile chromogenic substrates (Kračun et al. 2015 ).",
"discussion": "Discussion Recent work has successfully enriched microbial consortia using plant waste as sole carbon source (Gladden et al. 2011a ). Such microbial consortia were shown to have ample capacities to degrade plant biomass (D’haeseleer et al. 2013 ), being promising for lignocellulose saccharification (Park et al. 2012 ). Here, we used an innovative enrichment strategy, with two stages, based on partially degraded plant biomass (once-used) as sole carbon source. A similar approach was recently used in an anaerobic enrichment from lake sediment using switchgrass (Korenblum et al. 2016 ). Such a methodology is thought to enhance the prevalence of microbes acting on the most recalcitrant part of the lignocellulose (e.g. complex hemicellulose structures, crystalline cellulose and lignin). Additionally, this procedure may maintain plant biomass complexity, which decreases upon chemical and/or enzymatic pretreatments due to generation of more defined substrates (Lazuka et al. 2015 ). Clearly, a better picture of the consortial behaviour will be obtained by the evaluation of the time points along the microbial growth and incubation. However, it is known that the expression and secretion of the enzymes involved in the lignocellulose degradation are often more frequent at the final stages of growth (start of the stationary phase). Based on this premise, we decide to compare across three microbial consortia at one end-point of incubation (6 days), that is, between the exponential and stationary phase of growth. Based on the results, we postulated that substrate type is the main driver of the structure of microbial consortia developing in enrichments. Recently, Cortes-Tolalpa et al. ( 2016 ) reported that inoculum source is also a key factor that strongly influences the composition of plant biomass-degrading microbial consortia. However, stochastic factors (“first come, first bite”) might also have affected the selection process and so driven the microbial diversity in the consortia. Then, the growth on the partially degraded plant biomass clearly changed the structure of the original consortia (Fig. 2 a; Supplementary Fig. S1 ), suggesting that substrate structure and composition indeed drove the communities. In this respect, different proportions of lignin, cellulose and hemicellulose were observed after the first growth step as compared to the untreated plant biomass (Fig. 1 b). Regarding lignocellulose utilization, the degradation rates of hemicellulose were higher in the original microbial consortia compared with the selected ones, suggesting a higher availability of hemicellulose in the original substrates. Thus, this polymer could support, to a large degree, the growth of the consortia. Here, we used 16S rRNA-based PCR-DGGE coupled to 16S rRNA gene amplicon sequencing in order to determine the bacterial community structures along the enrichment experiment. In this respect, distinct structures were observed between the microbial source and the first and second enrichment steps, suggesting that, indeed, the consortia were strongly driven by the nature of the substrate, i.e. fresh versus once-used (Fig. 2 a). Although bacterial 16S rRNA gene (and fungal ITS1) surveys constitute powerful techniques to evaluate the diversity of microbial consortia (Jiménez et al. 2014b ), the here used “gene-centric” metagenomics approach may be regarded as superior, since it allows for the simultaneous characterization of microbial community structure and its metabolic potential. The 16S rRNA gene and the total metagenomic data are complementary approaches. However, it is not possible to perform a direct comparison between them due to differences in numbers of 16S rRNA gene copies, the database used and the genome sizes between the consortium members. The metagenomics-based analyses were performed using unassembled sequences, as this is presumed to cause minimal disturbance with respect to the representation of sequences of the abundant genera in the dataset (Teeling and Glöckner 2012 ). Moreover, on the basis of previously reported ITS1 versus bacterial 16S rRNA gene copy numbers (Brossi et al. 2015 ), next to the annotation of our metagenomic sequences, we postulate that the microbial consortia were dominated by bacteria. A comparison of the relative abundance values of the most abundant genera (>2 %) in our selected microbial consortia with the ones reported from forest soil metagenomics data (similar inoculum as used in this study; Jiménez et al. 2015a ) showed a fold increase of approximately 200 and 165 for Brevundimonas spp. in SG-M and CS-M, respectively. In contrast, Weeksella was the most enriched genus in WS1-M (~350-fold increase) (Supplementary Fig. S5 ). These organisms were undetectable by culture-based approaches applied to the original consortia (Brossi et al. 2015 ), suggesting their preferential growth on the once-used plant biomass. Based on the assumption that mainly microbes active in plant biomass degradation were enriched, it is reasonable to propose that such abundant consortium members contain enzymatic machineries that allow the deconstruction of lignocellulosic structures. The SG-M consortium that contained high abundances of Brevundimonas , Caulobacter , Pseudomonas , Citrobacter and Aeromonas , showed a high lignin degradation rate (~59 %). Caulobacter- like organisms were undetectable, by culture-based approaches, in the SG consortium (Brossi et al. 2015 ), which is consistent with a presumed selection of these microbes by the second growth step. However, the 16S rRNA gene amplicon-sequencing data showed a slight decrease of the abundance of Caulobacteriales -like organisms from the SG to SG-M consortia (Fig. 2 c). Otherwise, DeAngelis et al. ( 2011a ) reported enrichments of Caulobacter and Brevundimonas types (catalase producers) in lignin-amended soils compared with unamended ones. Moreover, it has been shown that Pseudomonas and Aeromonas have high capacities to transform lignin (Prabhakaran et al. 2015 ; Wu et al. 2010 ). For instance, Wang et al. ( 2013 ) reported a bacterial consortium that could break down 60.9 % of lignin in reeds at 30 °C under conditions of static culture within 15 days. This consortium was dominated by Pseudomonas species. In addition, Abhishek et al. ( 2015 ) showed that Citrobacter freundii can co-metabolize model and kraft lignin. These studies reflect the relevance of such taxa in lignin bioconversion by the SG-M consortium. Notably, Pseudomonas was the most abundant taxon in the WS1-M and CS-M consortia. However, the lignin degradation rates were significantly lower than those in the SG-M consortium ( p < 0.01), suggesting that Brevundimonas and Caulobacter species in SG-M may be the more relevant lignin degraders. Considering the latter, it is still unclear whether the lignin was completely metabolized or is present as modified acid-precipitable polymeric lignin (a water-soluble catabolite) in the culture supernatant, as has been observed for a compost-derived microbial consortium cultivated on pretreated switchgrass (Eichorst et al. 2014 ). One possible reason for the high degradation of lignin in SG-M might relate to a lower lignin recalcitrance in switchgrass, as compared to wheat straw and corn stover. Alternatively, the SG-M consortium might have developed a higher synergism between the degraders. In terms of cellulose and hemicellulose degradation, the CS-M consortium showed significantly higher degradation rates than the SG-M and WS1-M consortia ( p < 0.01). This CS-M consortium was mostly composed of Pseudomonas , Brevundimona s and Caulobacter types, but members of Stenotrophomonas , Xanthomonas , Pseudoxanthomonas , Achromobacter and Paenibacillus were also preferably selected (Fig. 3 b). Previous genome sequence analyses revealed that Caulobacter crescentus has the potential to degrade plant polysaccharides through the production of exo-enzymes, including cellulases, xylosidases and polysaccharide deacetylases (Nierman et al. 2001 ). Song et al. ( 2013 ) have shown degradation of cellulose by the mesophilic Caulobacter sp. FMC1 under aerobic and anaerobic conditions. Moreover, Eichorst and Kuske ( 2012 ) found that members of the Caulobacteriales and Xanthomonadales became prevalent in soil microcosms amended with [ 13 C] cellulose. Besides, Talia et al. ( 2012 ) reported the presence of Brevundimonas , Caulobacter , Pseudomonas , Xanthomonas , Stenotrophomonas , Achromobacter and Paenibacillus species in carboxymethylcellulose (CMC) and filter paper enrichment cultures from soil. Additionally, several strains of Pseudomonas , Stenotrophomonas and Paenibacillus retrieved from the CS consortium showed CMC-ase activity (Brossi et al. 2015 ). These studies reinforce our results, suggesting that the CS-M microbial consortium contains key members that were highly relevant in the degradation of (hemi)cellulose. In this study, the WS1-M consortium was dominated by Pseudomonas species that could be related with lignin bioconversion. As we also observed a strong selection of Bacteroidetes (e.g. Flavobacteriales and Sphingobacteriales ) (Fig. 2 b), similar to previous results (Jiménez et al. 2014b ), these data suggest that polysaccharides present in wheat straw selected for Bacteroidetes instead of Proteobacteria . Bacteroidetes like Sphingobacterium species can secrete enzymes such as endo-β-1,4-xylanases, α-L-arabinofuranosidases, β-glucosidases, α-glucuronidases and α-L-fucosidases when grown in the presence of wheat straw (Jiménez et al. 2015b ). Interestingly, organisms belonging to the Enterobacteriales (e.g. Klebsiella , Kluyvera and Enterobacter species) were most abundant in the WS1-M consortium as compared with WS1. The high abundance of Enterobacteriales in WS1-M was in line with the high frequency of strains belonging to this class retrieved from WS1 (Brossi et al. 2015 ). This suggested that, in this scenario, key organisms of the Enterobacteriales are strongly involved in the deconstruction of complex and recalcitrant plant polysaccharides. Degradation of lignin by Enterobacter and Klebsiella species has indeed been reported in recent papers (DeAngelis et al. 2011b ; Woo et al. 2014 ). Regarding the carbohydrate-active enzyme profiles, CAZy families GH10 (endoxylanases), GH3 and GH43 contain enzymes mainly involved in xylan, arabinan or arabinoxylan degradation, whereas families GH13 and GH28 are often active on pectin and rhamnogalacturonan, respectively. In addition, families GH3 and GH4 have broad substrate specificities and proteins of these families have β-D-glucosidase (GH3 and GH4), N -acetyl-β-D-glucosaminidase (GH3), α-glucosidase, α-galactosidase and α-glucuronidase (GH4) activities. The GH3 family was found to be highly abundant (Fig. 4 ). Similar results were reported in a rice straw-degrading microbial consortium (Wongwilaiwalin et al. 2013 ). Moreover, family GH16 enzymes cleave β-1,4 or β-1,3 glycosidic bonds in various glucans and galactans. Finally, families GH29 (α-L-fucosidases) and GH92 (α-mannosidases) contain exo-acting enzymes that can release fucose and mannose, respectively, from hemicellulose structures. Based on these considerations, we suggest that the three selected microbial consortia contain a wide genomic capacity to deconstruct different classes of plant polysaccharides, including hemicellulosic polymers. Although relative gene abundances do not report on actual enzymatic activities, we found relations between the abundance of particular metabolic potential (in terms of GH relative abundances) and the defined extracellular enzymatic activities. For example, high frequencies of genes encoding proteins of CAZy families GH10, GH3, GH43, GH28 and GH16 were found in the WS1-M, SG-M and CS-M metagenomes. Proteins of these families could be related to the enzymatic activities detected on CPH-xylan, CPH-arabinan, CPH-arabinoxylan, CPH-rhamnogalacturonan, CPH-galactomannan and CPH-β-glucan. Moreover, the low abundance of enzymes involved in cellulose (e.g. CBHs and endoglucanases) and lignin degradation (e.g. AA2) is not a signal that the underlying genes cannot be expressed. However, we did not find endo-activity on CPH-HE-cellulose, suggesting that the xylo-oligosaccharides released from the hemicellulose structures could strongly inhibit the activity of endoglucanases (Kont et al. 2013 ). Alternatively, endo-cellulases might be more active at lower pH, where we have tested only at pH 7.0. Notably, Jiménez et al. ( 2014b ) also reported a low activity of CBHs, compared with β-xylosidases and β-galactosidases, in the metasecretome of microbial consortia cultivated on wheat straw. Thus, the high activity of endo/exoglucanases, in plant biomass-degrading microbial consortia, may not be common. For instance, Gladden et al. ( 2011b ) found low activities of CBHs and β-glucosidases in a microbial consortium bred on acid-pretreated switchgrass. Also, D’haeseleer et al. ( 2013 ) reported the absence of CBHs in the metasecretome of a thermophilic bacterial consortium adapted to deconstruct switchgrass. Indeed, the majority of secreted GHs were associated with the deconstruction of hemicellulose (e.g. GH3, GH10 and GH51) or α-glucan polysaccharides (GH13 and GH31). Moreover, genes for enzymes of families GH5 and GH9 (endoglucanases) were highly abundant in a mesophilic cellulose-converting consortium (Wang et al. 2015 ). Based on these studies, we posit that the low abundance of genes for CBHs and endoglucanases in our metagenomes, next to the low activities in the metasecretomes, are in some way related with the differential response to the composition of the substrate (Gladden et al. 2011b ). There is increasing interest in auxiliary enzymes acting on cellulose by a non-hydrolytic mechanism of depolymerization. Among these enzymes, LPMOs (CAZy family AA10) represent the most promising class due to their capability of enhancing the efficiency of lignocellulose degradation by acting on polysaccharides that are recalcitrant to cellulases within highly crystalline cellulose (Dimarogona et al. 2013 ; Beeson et al. 2015 ). With the recent discovery of AA10 enzymes, a new model for enzymatic cellulose depolymerization has been proposed. Thus these enzymes, which oxidatively cleave endoglycosidic bonds in crystalline cellulose, may create new chain ends that can be attacked by CBHs and this synergistic effect probably improves the overall hydrolysis yield (Horn et al. 2012 ). The presence of this gene type in all three consortia provides evidence of the capacity to degrade cellulose or increase the deconstruction of other plant polysaccharides by this new oxidative mechanism. Regarding the AA6 family, these are intracellular enzymes involved in the biodegradation of aromatic compounds. Benzoquinone reductases are involved in a quinone redox cycle that generates extracellular Fenton reagents. In addition, these enzymes are involved in lignin degradation by fungi (Levasseur et al. 2013 ; Dashtban et al. 2010 ). However, we still do not know the actual role of these proteins in a lignocellulolytic bacterium dominated consortium. Finally, the production of oligosaccharides from plant biomass was detected using the ICB substrates. These findings suggest that the WS1-M, SG-M and CS-M microbial consortia have a high capacity to deconstruct plant biomass and convert complex polysaccharides into oligo and/or monosaccharides useful for downstream applications. The enzymatic activities detected on CPH and ICB substrates (Fig. 5 ) allowed to catalogue the three consortia as microbial enzyme “factories” that constitute excellent sources of efficient enzyme cocktails for the saccharification of plant biomass. Future experiments that combine the metasecretomes with available commercial cellulases can assist in raising the efficiency of plant biomass degradation for second-generation biofuel production. In addition, metatranscriptomics, metaproteomics and two-dimensional nuclear magnetic resonance spectroscopy (2D-NMR) analyses would help to better understand the lignocellulose degradation process (in particular the lignin bioconversion)."
} | 6,099 |
30746486 | PMC6358320 | pmc | 1,830 | {
"abstract": "By constructing a conjoined-network, we tackled the challenge in developing stiff and tough hydrogels of biogenic molecules.",
"introduction": "INTRODUCTION The design and construction of hydrogels for wide applications in biomedical fields ( 1 , 2 ), such as tissue engineering ( 3 , 4 ), drug delivery ( 5 ), wound dressing ( 6 ), and structural implants ( 7 , 8 ), have attracted intensive attention. Because of the inherent biocompatibility and biodegradability of biogenic molecules, hydrogels based on polysaccharides ( 9 – 11 ), proteins ( 12 – 14 ), peptides ( 15 , 16 ), and DNA ( 17 ) are paving the way toward versatile application scenario. Under such a circumstance, onion-like chitosan hydrogels with intermembrane spaces for cell introduction ( 9 ), biostable d -amino acid residue hydrogels for intratumoral chemotherapy ( 18 ), and tunable alginate microgels encapsulating single cells for therapeutic delivery ( 19 ) have been demonstrated. Among these, little attention has been paid to biomedical load-bearing materials from biogenic hydrogels, for example, ligament, cartilage, and cell culture scaffold, because they are normally soft and brittle. Effective strategies to improve their mechanical properties are urgently needed. Generally speaking, improving homogeneity and incorporating energy dissipation mechanisms have been recognized as effective methodologies to enhance mechanical performances of hydrogels. The ideally homogeneous network (involving spatial, topological, connectivity, and motility homogeneities) can behave cooperatively to avoid stress concentration, thus increasing the whole mechanical strength to some extent ( 20 , 21 ). However, due to the lack of effective energy dissipation mechanisms, significant improvement in mechanical properties cannot be achieved. On the other hand, incorporating energy dissipation mechanisms has proved to be more prominent ( 22 – 31 ). A typical case is the well-known double-network (DN) hydrogels ( 23 , 26 , 29 , 30 ), which combine two networks with heterogeneous structure and complementary properties and are denoted to have excellent mechanical properties. The first rigid and brittle network of DN hydrogels can effectively dissipate energy by scission of bonds, while the second soft and ductile network can withstand large strain to maintain the integrity of the hydrogel. Currently, DN hydrogels have been more dominating, and some derivatives are designed to improve them further, for instance, establishing connections between the two networks to distribute stress ( 30 ). However, because of the inhomogeneity between these two networks, the energy dissipation is mainly borne by the first network; therefore, increasing stiffness often deteriorates the toughness and vice versa ( 32 , 33 ), i.e., improving stiffness and toughness simultaneously remains a challenge. Different from the above two approaches, we designed conjoined-network hydrogels with synergistic energy dissipation mechanism to balance the inverse relation of stiffness and toughness ( Fig. 1 ). Conjoined-network hydrogels stand for a class of hydrogels consisting of two or more networks that are connected by sharing interconnection points to collaborate and featured as follows: (i) All the composed networks had similar or equal energy dissipation mechanism, and (ii) these networks were intertwined to effectively distribute stress in the whole system. In detail, within conjoined-network hydrogels, at the initial deformation stage, bonds in all composed networks randomly broke to absorb energy, increasing the initial modulus significantly; at the larger deformation stage, one of these networks can withstand large strain attributed to the reduction of cross-linking density, promoting toughness effectively. Fig. 1 Structure features of a conjoined-network hydrogel. ( A and B ) A conjoined-network hydrogel composed of (A) the first network(s), which used to effectively consume energy by bond rupture, and (B) the second network(s) with similar energy dissipation mechanism to the first network(s) but lower cross-linking density, which involved dual functions: partaking energy dissipation with the first network(s) at the initial stage and maintaining the integrity of the hydrogel at the large deformation stage. ( C ) These networks were further intertwined with each other to form a conjoined-network, which effectively distributed stress in the whole system. Here, we reported a particular example of conjoined-network hydrogels composed of biogenic molecules, including ionic cross-links of the first and second networks. In detail, a feasible “gelation and soaking” method was applied to convert weak chitosan-gelatin (C-G) composite hydrogels to stiff and tough chitosan-gelatin-phytate (C-G-P) conjoined-network hydrogels in sodium phytate solution. Sodium phytate, which has six phosphate groups, can interact with a dozen of amino groups of chitosan or gelatin chains to form two individual networks, and also couple these two networks by interacting with amino groups of both chitosan and gelatin simultaneously, resulting in a compact physical conjoined-network. The amine-phytate electrostatic domains played a dual role in enhancing the mechanical properties of hydrogels: reinforcing hydrogels by bond rupture and enhancing self-recovery and anti-fatigue capability by bond reformation. The gelation and soaking strategy was also a straightforward methodology to modulate the mechanical strength of hydrogels over broad ranges. Moreover, hydrogels were composed of biogenic molecules (chitosan, gelatin, and sodium phytate), which are well known to be biocompatible and biodegradable. A focus toward developing the conjoined-network should offer new directions for stiff and tough biomedical hydrogels based on other recoverable energy-dissipating mechanisms, such as host-guest interaction, hydrophobic interaction, and hydrogen bonding.",
"discussion": "DISCUSSION To conclude, the conjoined-network hydrogel was constructed, which differed from the homogeneous network or the DN hydrogel in concept and topological structure: All the composed networks owned similar energy dissipation mechanism, and these networks were intertwined to effectively distribute stress in the whole system, which demonstrated a significant improvement in both modulus and toughness. As an example, a conjoined-network hydrogel composed of biogenic molecules using a gelation and soaking method was reported. The C-G-P hydrogels achieved an unusual combination of high modulus and toughness owing to the formation of the conjoined-network by chitosan, gelatin, and phytate. Moreover, the physical amino-phytate domains endowed the C-G-P hydrogels remarkable self-recovery and anti-fatigue capacity. It provided various facile strategies to tune mechanical properties over a wide range by adjusting the concentration of soaking media, the functionality of cross-linker, or the diverse ionic combination for various applications. In general, our results illustrate an approach to stiff and tough conjoined-network hydrogels with multifunction based on electrostatic interaction or other recoverable energy-dissipating mechanisms, such as hydrogen bonding and host-guest interaction. We believe that these conjoined-network hydrogels from biogenic sources can be applied in the fields of tissue engineering, vibration absorbers, soft robotics, and smart wearable devices."
} | 1,855 |
34478597 | PMC9293205 | pmc | 1,832 | {
"abstract": "Abstract Microbial communities are hugely diverse, but we do not yet understand how species invasions and extinctions drive and limit their diversity. On the one hand, the ecological limits hypothesis posits that diversity is primarily limited by environmental resources. On the other hand, the diversity‐begets‐diversity hypothesis posits that such limits can be easily lifted when new ecological niches are created by biotic interactions. To find out which hypothesis better explains the assembly of microbial communities, we used metabolic modelling. We represent each microbial species by a metabolic network that harbours thousands of biochemical reactions. Together, these reactions determine which carbon and energy sources a species can use, and which metabolic by‐products—potential nutrients for other species—it can excrete in a given environment. We assemble communities by modelling thousands of species invasions in a chemostat‐like environment. We find that early during the assembly process, diversity begets diversity. By‐product excretion transforms a simple environment into one that can sustain dozens of species. During later assembly stages, the creation of new niches slows down, existing niches become filled, successful invasions become rare, and species diversity plateaus. Thus, ecological limitations dominate the late assembly process. We conclude that each hypothesis captures a different stage of the assembly process. Species interactions can raise a community's diversity ceiling dramatically, but only within limits imposed by the environment.",
"introduction": "1 INTRODUCTION Bacterial life on Earth is extraordinarily diverse. Our planet is inhabited by an estimated 1.4–1.9 million bacterial lineages (Louca et al., 2018 ), and every gram of soil hosts between 2000 and 18,000 distinct such lineages (Daniel, 2005 ). Understanding how Earth holds all this biodiversity is a fundamental challenge to ecology and evolution. Part of the difficulty in meeting this challenge is that planet Earth offers limited resources, which may impose a ceiling on biodiversity. In this vein, the prominent “ecological limits” hypothesis of biodiversity (Rabosky & Hurlbert, 2015 ; Schluter & Pennell, 2017 ) posits that rates of diversification should decrease as diversity increases and fills available ecological niches. In contrast, the “diversity begets diversity” hypothesis posits that diversity could stimulate further diversification (Calcagno et al., 2017 ; Whittaker, 1972 ) as species–species interactions get more complex and novel niches are created (Erwin, 2008 ; Laland et al., 1999 ), or as existing niches are partitioned more and more finely as a result of competition (Bailey et al., 2013 ; Dieckmann & Doebeli, 1999 ). Evidence for either hypothesis in bacteria, animals and plants is mixed. For example, a combination of field work and phylogenetic analysis of Himalayan songbirds suggests that an observed slowdown in their speciation rate is explained by niche filling. Their species distributions are well explained by resource abundance (Price et al., 2014 ). Two evolution experiments with Pseudomonas fluorescens strains also support the ecological limits hypothesis (Brockhurst et al., 2007 ; Gómez & Buckling, 2013 ). In these experiments a strain's diversification slowed down as more strains were included in a coculture. Conversely, similar experiments with different strains of P . fluorescens support the “diversity begets diversity” perspective. These experiments tracked the diversification of a focal lineage of P . fluorescens in bacterial communities that harboured from one to eight lineages of P . fluorescens . More novel morphotypes evolved when communities were initially more diverse (Jousset et al., 2016 ). Further support for the “diversity begets diversity” hypothesis comes from experiments with crops and weeds, which show that weed diversity is high whenever crop diversity is high (Palmer & Maurer, 1997 ). It also comes from data on plant and arthropod diversity on the Canary and Hawaiian islands, where the proportion of endemic species increases with increasing species numbers (Emerson & Kolm, 2005 ). It also comes from inferences of microbial diversification rates from several taxonomic ratios, such as the species to genus ratio (Madi et al., 2020 ). In addition, theoretical modelling using adaptive dynamics also suggests that some initial diversity can facilitate further diversification (Calcagno et al., 2017 ). The diversity of a community may increase both through ecological processes, as species disperse and invade a community, or through evolutionary processes, as existing member species diversify and speciate. Pertinent evidence on the ecological limits and diversity‐begets‐diversity hypotheses comes in part from experiments or field work that only quantify diversification rates, making it difficult to disentangle ecological from evolutionary limitations. However, the hypotheses should ideally be distinguished by considering both evolutionary and ecological processes. In a scenario where one can control the appearance of new species in a community, either because they evolve in the community or disperse into the community, will environmental resources constrain community diversity, or will species–species interactions promote ever‐increasing diversity? This is the question we aim to answer here. To this end, we used a very simple community assembly strategy inspired by the environmental filter metaphor (Levy & Borenstein, 2013 ; Thakur & Wright, 2017 ) and by trait‐based models (Mcgill et al., 2006 ). Such models organize ecological processes around traits—species properties that impact species survival. Typically, species show trade‐offs in these traits, and a trait that gives a species an advantage in one environment might be disadvantageous in another. The environment selects (filters) species according to their traits. This modelling strategy has been widely successful, for example to predict community composition along environmental gradients (Allison, 2012 ; Laughlin et al., 2012 ; Litchman & Klausmeier, 2008 ; Mcgill et al., 2006 ; Thakur & Wright, 2017 ). Here, we use it to model the assembly of microbial communities, where a trait refers to a species’ ability to thrive on a specific source of carbon and energy. We simulated the assembly of microbial communities in which we represent each individual species by a metabolic network that comprises thousands of metabolic reactions needed for an organism's survival in a given chemical environment. The advantage of this approach is that it allows fundamental metabolic traits to emerge from first biochemical principles, which are embodied in the metabolic reaction network of an organism. Especially important traits for our purpose include the ability to survive on a given source of carbon and energy, and on the ability to excrete specific by‐products of metabolism. Traits like these can be computationally predicted with flux balance analysis (FBA; Orth et al., 2010 ), an experimentally validated (Orth et al., 2011 ; Varma & Palsson, 1994 ) computational method that can determine the flux of matter through every reaction in a metabolic network when cells are in a metabolic steady‐state, and when they grow their biomass is at the maximum possible rate given the biochemical reactions they can catalyse. FBA has been successfully used to predict growth and by‐product secretion of bacteria in different media (Edwards et al., 2001 ; Ibarra et al., 2002 ; Varma & Palsson, 1994 ). Metabolic networks have been characterized for hundreds of organisms (Gu et al., 2019a ). They are highly valuable to understand metabolic biology and evolution because they reflect an organism's evolutionary history. For the same reason, however, they are of limited use for studies like ours, which aim to understand how community assembly may be affected by specific metabolic properties of the assembled species, such as the number of biochemical reactions any one species harbours, and the number of carbon sources it can utilize. For this purpose, one needs to vary these properties systematically, but in the network of any one organism, they are fixed. To circumvent this limitation, we started our community assembly not from previously characterized metabolic networks of microbial species. Instead, we randomly sampled (with a Markov chain Monte Carlo [MCMC] method) thousands of metabolic networks from a much larger “universe” or “pan‐metabolism” of biochemical reactions, such that each network fulfilled specific requirements. These requirements include the ability to sustain life on specific carbon sources such as glucose, while containing an otherwise random complement of chemical reactions. We refer to such metabolic networks as random viable networks. For the purpose of our analysis, each such network represents a different “species,” and we used these “species” to simulate the assembly of a community in a well‐mixed chemostat‐like environment where resources are supplied by the environment at a constant rate. Our community assembly procedure consists of three iterated steps (Figure 1 ). First, a random species invades the environment/community. Second, the environment acts as a filter that selects those species that persist in the community. Third, the persisting species may change the environment by excreting metabolic by‐products. We repeated these three assembly steps for thousands of species invasions in any one community, and simulated hundreds of community assemblies in environments with different amounts of resources and varying strengths of competition. FIGURE 1 Modelling community assembly. (a) The three steps we used to model community assembly. (I) Species (modelled as random viable metabolisms) are added at random to a standing community. (II) The environment acts as a filter, selecting successful species, that is those best at growing on at least one of the carbon sources available in the environment. Successful species persist in the community. (III) The species that comprise the community modify the environment through the excretion of metabolites that can serve as carbon sources for other species. (b) Example assembly trajectory of one community up to the invasion of the fourth species. Simulations begin by initializing the composition of the environment, shown as a purple rectangle at the top. The presence of each of the 223 potential carbon sources is represented as a purple vertical bar, such that each location on the horizontal x ‐axis corresponds to one carbon source. In this example and in most of our simulations (unless otherwise stated), the initial environment contains glucose as the sole carbon source, which is indicated in the top rectangle by the single vertical purple line. After establishing the initial environment, we perform the first invasion with “species 1” chosen at random from our MCMC‐derived sample of random viable networks. The species’ ability to grow on each potential carbon source (whether or not the carbon source is present in the environment) is shown in the green rectangle by a green line in the x ‐position that corresponds to the specific carbon source. Darker green lines indicate higher growth. Species 1 can grow on the only available carbon source (glucose), and because no other species is yet present, we consider species 1 successful. Species 1 modifies the environment with the excretion of by‐products of its growth on glucose, as shown in the second purple rectangle from the top. These by‐products are available as potential nutrients for the second round of assembly. In this round, a randomly chosen species 2 invades the community. Its growth rate on each carbon source is shown below that of species 1. Species 2 cannot grow faster than species 1 on any of the available carbon sources. It therefore goes extinct and only species 1 persists in the community. The carbon sources it excretes are available for the third assembly step. Each such step consists of a new round of (I) species invasion, (II) environmental filtering and (III) niche construction During community assembly, we found that many persisting species excrete metabolic by‐products, which can sustain species that invade later through cross‐feeding interactions in which one organism consumes the excretion of another. In other words, species create opportunities for cross‐feeding, which begets further diversity. Further, we find that communities grow more diverse when competition among species is stronger. Together, these two observations show that biotic interactions between species are critical to establishing diverse communities. At the same time, communities show limits in the number of species they accommodate. As community richness increases, the probability that a new species successfully invades a community decreases. The reason is that by‐product excretion cannot create new niches ad infinitum. Eventually all new niches have been created, and subsequent invasions mostly fill the existing niches. At this late stage, the assembly dynamics enters an “ecological limits” regime.",
"discussion": "4 DISCUSSION The “ecological limits” and “diversity begets diversity” hypotheses were proposed to explain the forces that shape biodiversity on Earth. Several analyses aiming to validate these hypotheses have examined the rate of evolutionary diversification to provide support for one or other hypothesis (Brockhurst et al., 2007 ; Gómez & Buckling, 2013 ; Jousset et al., 2016 ). However, evolution is not the only means by which community diversity can increase, because new species can also increase diversity by invading a community. We therefore feel that the hypotheses should be distinguished by considering a scenario that considers the contributions of both ecological and evolutionary processes to community diversity. A modelling approach like ours is well suited for this, because the species we simulate could be thought as being new evolved species or species that invade a community as they disperse from elsewhere. With these considerations in mind, we here simulate the assembly of thousands of communities in which we can control the number of new species allowed to invade a community and monitor the community's response to each invasion. Our results suggest that the dichotomy between the diversity‐begets‐diversity and the ecological limits hypothesis is a false one. Both hypotheses capture important aspects of diversification. At initial stages of community assembly, species–species interactions in the form of cross‐feeding and competition facilitate diversification. Species “construct” new niches through their metabolic excretions, which creates opportunities for the invasion of new species that thrive on these excretions (Figure 2 ). Competition helps partition niche space into more and narrower niches, which can result in larger diversity (Figure 4 ). At later stages in the assembly process, no new niches can be created, niche space cannot be partitioned further, successful species invasions become rare and diversity reaches a ceiling. In sum, the diversity‐begets‐diversity and the ecological limits scenarios best capture different stages—early vs. late—of community assembly. Community diversity has limits, but these limits are not just externally imposed. They can be modified by the species themselves. Two other conclusions emerge from our results. First, even when all niches have become filled during community assembly, diversity may not have reached its ceiling. Competition can further increase community richness even in this case, because it can help narrow the realized niche of individual species (Figure S2 ). Second, our conclusions reinforce the argument that it is best to consider ecological processes separately from evolutionary processes when studying diversification. The reason is that the probability of successful invasions decreases even early during community assembly, while species are still creating new niches during the “diversity begets diversity” regime (Figure 2d ; Figures S3, S5 ). This observation, which agrees with existing theory (Case, 1990 ; Tilman, 2004 ; Vila et al., 2019 ), experimental observations (Fargione et al., 2003 ; Stachowicz et al., 1999 ) and the empirical observation that diversification also slows down on evolutionary timescales (Condamine et al., 2019 ), implies that a slowdown in diversification (or speciation) is by itself not sufficient proof for the ecological limits hypothesis. Competition is commonly thought to stimulate the diversification of species (Abrams, 2006 ; Bailey et al., 2013 ; Dieckmann & Doebeli, 1999 ; Meyer et al., 2016 ; Meyer & Kassen, 2007 ; Yoder et al., 2010 ), although it can also prevent such diversification (Bailey et al., 2013 ; Brockhurst et al., 2007 ; Gómez & Buckling, 2013 ). In our work, a successful invasion increases competition, because available resources have to be shared between more species. After such an invasion, communities may show an increase, no change or a net decrease in diversity if multiple species go extinct as a result of the invasion (Figure 2a ), in agreement with empirical observations from the literature (Abrams, 2006 ; Bailey et al., 2013 ; Brockhurst et al., 2007 ; Dieckmann & Doebeli, 1999 ; Gómez & Buckling, 2013 ; Meyer et al., 2016 ; Meyer & Kassen, 2007 ; Yoder et al., 2010 ). However, such varying short‐term consequences of competition for individual invasions need to be distinguished from the long‐term consequences for an entire assembly process. At the end of this process (once community richness no longer increases), we find that communities with stronger competition have become more diverse (Figure 4 ). In other words, our work predicts that increased competition results in a net increase of diversity. In an organism, many traits are interconnected. Traits that are advantageous for one aspect of fitness may trade‐off with traits that affect other aspects. Such trade‐offs are thought to be important for determining community structure (Begon et al., 2006 ; Johnson et al., 2012 ; Litchman & Klausmeier, 2008 ; Tilman, 1982 ). Our results also suggest—albeit indirectly—that trade‐offs in resource use are important for the communities we assemble. If there were no such trade‐offs, we would expect that one or a few species consume all available resources. Instead, we find a great diversity of species, which suggests the presence of trade‐offs in resource use. It is important to point out that our modelling framework does not assume such trade‐offs to exist. They arise from elementary biochemical principles embodied in the set of biochemical reactions that constitute a metabolic reaction network. Our work has several limitations. Addressing them offers opportunities for future research. First, we study a very constrained subspace of Hutchinsonian niche space, the multidimensional space comprising all environmental factors required for the survival of a species (Hutchinson, 1957 ). Specifically, we only consider the sources of carbon and energy that sustain heterotrophic organisms, because they are frequently cross‐fed between bacteria (D’Souza et al., 2018 ; McNally & Borenstein, 2018 ). Our modelling framework does not account for other abiotic factors, such as temperature and pH, nor does it account for non‐nutrient‐mediated species interactions, such as those that arise from cell‐to‐cell communication, or the secretion of toxic molecules. Community richness could increase above the limit we observe if such interactions help create even more niches. In addition, our knowledge about pan‐metabolism is increasing, and newly characterized metabolic reactions could further increase a community's predicted potential to create more niches and diverse communities. Second, we observe that different species grow on similar sets of carbon sources–their fundamental niches are similar—which may result from the method we used to create metabolic networks for community assembly. Specifically, we constrained these networks to be viable on glucose, which may have increased their niche similarity, because these networks must share subsets of biochemical reactions required for growth on glucose (Barve et al., 2012 ). To explore how this choice may have affected our analysis, we performed two complementary analyses. In a first analysis, we compared how the fundamental niche overlap of random viable networks compares to that in real organisms. For that, we used bacteria from the human gut (Magnúsdóttir et al., 2017 ). Specifically, we identified in the metabolic networks of 818 species those metabolites that could serve as sources of carbon, and quantified fundamental niche overlap based on the results. We found that the fundamental niche overlap between these species can range from zero (no overlap) and up to one (full overlap), like the range of overlap explored with random viable networks (see Text S1, Figures S7, S8 ). In a second analysis, we created random metabolic networks required to be viable on carbon sources other than glucose (acetate, pyruvate, serine, alanine and lactose) (Figure S6 ). Communities assembled using these networks differed little in community diversity, even when we assembled communities from species required to be viable on different carbon sources. Therefore, this limitation is not likely to affect our observations dramatically. Third, we observed that communities can be composed of up to 80 species whose persistence depended, directly or indirectly, on the species consuming the initial resource, glucose. Unfortunately, our current modelling approach cannot answer how abundant these species are, and how much glucose would be required to sustain a highly diverse community. To answer these questions remains an important task for future work. Fourth, we use the growth rate of species on single carbon sources to predict growth in a complex environment with multiple carbon sources. In such an environment, a microbial species growing in isolation may either consume a single (preferred) carbon source, several carbon sources, or all available carbon sources. Which of these strategies it pursues depends on multiple factors, including the species, the cultivation method (batch or chemostat) and the similarity between the carbon sources (Aidelberg et al., 2014 ; Egli et al., 1993 ; Kovárová‐Kovar & Egli, 1998 ). The picture is even more complex in cocultures, where species interactions may affect consumption patterns. Our decision to use maximal growth on a single carbon source to model a species’ persistence is motivated by recent experiments which show that the assembly of communities in mixtures of nutrients can be predicted from assembly in single‐nutrient environments (Estrela, Sanchez‐Gorostiaga et al., 2020 ; Fu et al., 2020 ). For example, the taxonomic structure of a community assembled in glucose medium (Goldford et al., 2018 ) can be explained by the growth of its constituent species in isolation, either on the externally provided glucose, or on the cross‐fed metabolites acetate, succinate and lactate (Estrela, Vila, et al., 2020 ). Metabolic modelling is a powerful modelling framework. It has been successfully used to study microbial physiology (Notebaart et al., 2008 ; Segrè et al., 2002 ; Shlomi et al., 2005 ; Varma & Palsson, 1994 ), to design strains for industrial and medical applications (Gu et al., 2019b ; Lun et al., 2009 ; Mishra et al., 2018 ), and to explore questions in the field of evolution (Bajić et al., 2018 ; Barve & Wagner, 2013 ; William Harcombe et al., 2013 ; Ibarra et al., 2002 ; Notebaart et al., 2014 ; San Roman & Wagner, 2018 , 2020 ; Sandberg et al., 2017 ) and ecology (Estrela, Sanchez‐Gorostiaga et al., 2020 ; Harcombe et al., 2014 ; Levy & Borenstein, 2013 ; Machado et al., 2021 ; McNally & Borenstein, 2018 ; Zelezniak et al., 2015 ), as reviewed previously (García‐Jiménez et al., 2021 ; Gu et al., 2019b ; Mardinoglu & Nielsen, 2012 ). Though outside the scope of our work, the modelling framework used here can in principle also be used to study other ecological phenomena, such as successional dynamics (Bell & Pascual, 2020 ; Chase, 2003 ; Dini‐Andreote et al., 2014 ; Lockwood et al., 1997 ; Nemergut et al., 2007 ), and the role of historical contingency (or priority effects) in the assembly process (Chase, 2003 ; Fukami, 2015 ). What is more, whole‐community genome sequencing, together with semi‐automatic methods for metabolic reconstruction, facilitate the creation of genome‐scale metabolic models for not just one organism, but for multiple organisms in a community (DeJongh et al., 2007 ; Dias et al., 2015 ; Mendoza et al., 2019 ; Wang et al., 2018 ). Thus, it may soon be possible to conduct an analysis like ours with metabolic networks characterized in a community from the wild. While such an analysis cannot control all the variables that our computational work can control, it can go beyond a proof of principle, explain actual limits on community diversity and identify rules of community assembly important in nature. It will be especially suited for the analysis of those complex communities where cross‐feeding interactions are important, such as that of the human gut (Magnúsdóttir et al., 2017 ), the soil (Baran et al., 2015 ) or planktonic organisms (Enke et al., 2019 )."
} | 6,387 |
34113833 | PMC8170003 | pmc | 1,833 | {
"abstract": "Summary Harvesting mechanical energy via a triboelectric nanogenerator (TENG) is a promising strategy for solving energy problems. However, it is necessary to develop an effective and safe energy managing circuit for preventing high voltage breaking electronic devices. Here, a universal managing circuit is developed to optimize TENG's output performance, which for the first time allows the TENG to safely power various sensor systems with a safe and stable voltage. Based on the circuit, TENG's output can be transformed into a stable voltage with tunable amplitude, while an enhanced short-circuit current of 94 mA with an energy loss lower than 5% is achieved. For demonstrations, three different types of TENGs, respectively, targeting at ocean energy, wind energy, and walking energy have been prepared to reveal the capability of the circuit. This study offers a strategy to greatly enhance the output performance of TENGs to provide useful guidance for constructing self-powered and distributed sensor systems.",
"conclusion": "Conclusions In summary, we have developed a different universal managing circuit as a link between TENG and functional electronics devices, such as wireless environmental sensors or microcontrollers. The universal managing circuit integrated with the TENG, well matching the high output impedance of TENGs, can provide an output current in the milliampere level, while the stable output voltage from this universal managing circuit can be stabilized at a designed value to power different wireless devices for preventing high voltage breakdown. The energy managing process of this circuit can be achieved automatically without the need of extra battery source, and the energy loss within this managing circuit can be maintained to be lower than 5%. The universality and efficiency of this universal managing circuit is verified by using the TENG with different working modes and structures. Three environmental sensor systems, which can harvest different micro-energies (water wave energy, windy energy, and walking energy), are demonstrated to verify the stability, versatility, and practicality of the universal managing circuit. The proposed universal managing circuit can be used for constructing wireless and distributed sensor systems with minimized energy consumption to meet challenges arose from robotics, wearable devices, environmental monitors, and the era of internet of things.",
"introduction": "Introduction Entering the era of internet of things, wearable and implantable electronic devices are in rapid development, leading to an urgent demand of clear, sustainable, and distributed energy supply ( Tarancón, 2019 ; Say et al., 2020 ; Wang, 2018 ; Tang et al., 2020 ). Traditionally, batteries are employed to power these devices, while their limited capacity and big volume cannot fully satisfy the fast-rising demands ( Liang et al., 2020a , 2020b , 2020c ; Nyholm, 2020 ). Harvesting energy from ambient environment or human motion is one of most promising strategies to compensate the deficiencies of battery ( Li et al., 2020a , 2020b ; Zhang et al., 2018 ; Liu et al., 2020a , 2020b ; García Núñez et al., 2019 ; Xue et al., 2017 ; Seol et al., 2015 ; Kim et al., 2018 ). First proposed by Wang Group in 2012 ( Fan et al., 2012 ), the triboelectric nanogenerator (TENG, also called Wang generator), which can convert all kinds of mechanical energy into electricity, has so far been applied in many fields, including the exploitation of ocean wave energy ( Liang et al., 2020a , 2020b , 2020c ), harvesting human motions energy ( Zhao and You, 2014 ; Ren et al., 2020a , 2020b ; Miao et al., 2019 ), and even heartbeat energy ( Ouyang et al., 2019 ; Liu et al., 2019a , 2019b ). Driven by the Maxwell's displacement current, TENGs have significant high-voltage output, which can reach thousands of volts, while the current output of TENGs is very low within the microampere level, and the internal resistance is very large within the megohm level ( Wang et al., 2020a , 2020b , 2020c ; Liu et al., 2019a , 2019b ; Xia et al., 2019 ; Zi et al., 2016a , 2016b ; Li et al., 2020a , 2020b ; Mao et al., 2017 ). Accordingly, an effective and safe energy managing system is quite necessary for TENG devices. The working voltage and impedance of electronic device and energy storage unit are relatively low, which cannot match with the high voltage and internal resistance of TENGs, and thus, it is not possible to directly use TENGs as a power source for the electronic system ( Harmon et al., 2020 ; Wang et al., 2020a , 2020b , 2020c ; Xia et al., 2020 ; Wang et al., 2018 ). Meanwhile, mechanical energies in environment usually come from random mechanical motions, and the output signal from the TENGs has pulsed waveform with random amplitude and frequency. In order to regulate the output of TENGs, current boosting, buck converting, and energy storage are three important modules need to be considered ( Cheng et al., 2019 ; Zhang et al., 2019 ). In the past few years, even though many significant improvements have been achieved in these directions, several unsolved issues are still hindering the practical applications of TENGs. First of all, the controlling circuits in most of managing circuits of TENGs are powered by another circuit, which requires external power supply or manual operation ( Song et al., 2019 ; Cheng et al., 2017 ; Zhang et al., 2020 ). Therefore, a fully active power management (PMM) is quite necessary for TENGs. Secondly, traditional buck converting for managing output signal of TENGs including inductive transformer or passive switch cannot provide a stable output voltage signal, and it is difficult to be directly connected with electronic units ( Zi et al., 2016a , 2016b ; Zi et al., 2017 ; Xu et al., 2018 ; Pu et al., 2016 ). So far, a PMM specially designed for TENGs is still one of the most important tasks for the study of TENGs ( Liu et al., 2020a , 2020b ; Niu et al., 2015 ). Here, we proposed for the first time a universal managing circuit (UMC) for TENGs, which is able to provide a stable and optional output voltage for matching standard electronic voltage, while its controllable output current can be up to 94 mA. This managing circuit with stable output voltage can directly supply power to the sensor system, avoiding the breakdown of electronic devices due to TENG's high voltage, and can connect directly with high-voltage TENGs and well match the high output impedance of TENGs with no additional power supply. Both the voltage and the current can also be regulated by the inner capacitor of the managing circuit which allows charge to be accumulated on an input capacitor until the buck converter can efficiently transfer a portion of the stored charge to the output. Meanwhile, the charging speed of a TENG-capacitor system is increased 2.5 times with the help of this managing circuit. Three demonstrations with different types of TENGs have been prepared to reveal the capability of this managing circuit, and the output signal from this managing circuit can be directly used to power various low power devices, such as an ocean temperature monitor, humidity sensor, Bluetooth transmitter for position detecting, and so on.",
"discussion": "Results and discussion A framework for a self-powered sensor system based on TENGs is illustrated in Figure 1 A. A fully self-powered wireless environment sensor system requires microcontrollers and transducers that collect energy from environments. The link between an energy generator and energy consumer is the energy managing circuit. The typical working principle of a TENG is shown in Figure 1 B, where a contact-separation TENG is selected as the example, and the universal managing circuit is aiming at serving all kinds of TENGs. Usually, nanostructures are designed on the surface of triboelectric materials to further enhance the output performance of the TENG. For example, the poly tetra fluoroethylene (PTFE) film with nano-patterns on its surface (water contact angle of 156°) is modified by the inductively coupled plasma etching treatment, as shown in Figure 1 C including ⅰ and ⅱ. As shown in Figure S1 , the nano-structured PTFE improved the TENG's output performance. The detailed fabrication process is presented in STAR Methods section. The performance of TENGs can be enhanced and regulated by the managing circuit, while the stable voltage the direct current (DC) outputs provided by the managing circuit can power various functional devices with a standard electronic voltage, such as sensors, alarming devices, displaying devices, and so on. As shown in Figure 1 D, the universal managing circuit can transfer the AC output of TENGs into a stable DC output with high current value and stable voltage value. For improving the wearable performance, a flexible and soft circuit is also designed to meet different application scenarios as shown in Figure 1 E. The detailed fabrication process of managing circuit is presented in the Transparent Methods Supplemental information . Figures 1 F–1H show the superior output performance of TENGs under the management of the managing circuit. With the managing circuit, the high voltage pulses of TENG can be gathered, while the collected charges can be released at once. Then, a current pulse signal with the amplitude of dozens of milliampere can be produced. Moreover, the voltage stabilizing module in the managing circuit can regulate the voltage of the output signal to be a stable value, not in the form of pulses, as shown in Figure 1 F. The voltage value can be designed to be 1.8V, 2.5V, 3.3V, or 3.6V, which are the common operation voltage of various electronics devices. The TENG with the managing circuit can offer a current output in the milliampere scale, lighting up a very bright light-emitting diode (LED) (see Figure 1 G ⅱ and Video S1 ), while the same TENG without universal managing circuit can only provide current output in the microampere scale (see Figure 1 G i). With high current and stable voltage, the UMC-based TENG with milliampere level current can light UV LED very bright, which can be used as a self-powered currency detector (see Figure S2 and Video S2 ). Moreover, the amount of single charge transfer of the TENG with universal managing circuit is about 2.88 times of that the TENG just with a simple AC-DC rectifier bridge, as shown in Figure 1 H. Figure 1 Structure design of the self-powered system based on the universal managing circuit and TENG (A) Framework for the self-powered sensor system achieved by a universal managing circuit with TENGs harvesting micro-energy. (B) Working principle of each TENG unit. (C) Scanning electron microscope (SEM) images of the PTFE polymer nanostructure (including i) and (ii) its surface with water contact angle of 156°. (D) Photograph of a hard universal managing circuit with its size and pins. (E) Photograph of flexible and soft universal managing circuits (the scale bar is 1 cm). (i and ii) Photographs show its softness and flexibility. The superior output performance of the TENG under the universal managing circuit in (F) output voltage, (G) current, and (H) single charge transfer, respectively. Video S1. LED with UMC-TENG_spl_, related to Figure 1 Video S2. UV LED with UMC-TENG, related to Figure 1 The managing circuit is diagrammed in Figure 2 A, mainly consisting of two full bridge rectifiers, a voltage regulator, and some capacitors. As explained in above part, the circuit has two major functions. The first one is to collect and store the output charges from the TENG, in order to realize a high current output. The second is related to the voltage stabilizing function, which can release the stored charge at the designed voltage value. The high efficiency voltage regulator and one buck converter are contained in the chip LTC 3588-1 (see Figure S3 ). It should be emphasized that the commercial chip LTC 3588 cannot directly connect with high-voltage TENGs because the chip's input voltage cannot stand with the voltage signal larger than 20 V. Hence, it is necessary to establish an external circuit to manage the high voltage signal of the TENG and only use the reaction loop of the chip LTC-3588. This management circuit is specially designed for the TENG, which is totally different from traditional application of the chip LTC-3588. The first rectifier D1 can transform the alternating current output of the TENG into direct current. The second one is to further rectify the output current to ensure that the function circuit is working properly. The charges are stored in the input capacitor C1 and then transferred by the voltage regular to the output capacitor C4. The target output voltage can be set to 1.8V, 2.5V, 3.3V, or 3.6V, by connecting pin D0 and D1 to pin of input voltage 2 (Vin2) or ground (GND) (see Table S1 ). Therefore, the pin D0 and D1 are connected to Vin2 and GND, respectively, to match the working voltage of the electronic device in the next experiments. Figure 2 Circuit structure and stimulation of the universal managing circuit (A) Circuit diagram of the universal managing circuit. (B) Stimulation result of output current with different external Cap (C1). (C) The voltage of external Cap (C1) charging by TENG and the output current of the universal managing circuit. (D) Simulation result of internal time between two current peaks with respect to various low frequencies. LTspice software is used to simulate the universal managing circuit; related parameters are shown in Table 1 . The simulation result with respect to various external Cap (C1) is plotted as Figure 2 B. The output current increases with the increase of external Cap (C1). As shown in Figure 2 C, when the voltage of C1 charging by TENGs increase to 5.2V, the voltage regulator controls the C1 to discharge then the voltage of C1 decreases to 3.8V, meanwhile the output current is significantly increased, with a peak value above 15.57 mA. Moreover, with the increase of frequency of input signal, the interval time between current peaks becomes smaller (see in Figure 2 D). Table 1 Parameters used for simulating the universal managing circuit Electronic device Quantity Unit C2 1 μF C3 4.7 μF C4 10 μF L1 10 μH To explore the performance of this universal managing circuit, the grating-structured freestanding TENG (GF-TENG) is applied to work with the managing circuit, and a basic operating unit is sketched in Figure 3 A. It has a PTFE film with a metal electrode deposited on the back side (back electrode). On the front side, the PTFE film makes relative motion with another metal electrode (contact electrode). During contact, electrons are transferred from metal into PTFE, leading to the accumulation of negative charges on PTFE surface and positive ones on metal surface. COMSOL Multiphysics software based on finite-element simulation is employed to calculate the potential distribution across the two electrodes at different states, as shown in Figure 3 B, and the variation of this potential difference induces displacement current in the external circuit. Figures 3 C–3E show the output performance of the GF-TENG at 2Hz without universal managing circuit. The open-circuit voltage can reach to 100 V, and the short-circuit current is about 40 μA. With the AC-DC rectifier bridge, the amount of single charge transfer is about 7.52 μC. The stable value of the output voltage as well as the output current is shown in the table of Figures 3 F and 3G, and a video demonstration with LED light is shown in Supporting Videos S1 and S2 . The output current of the TENG integrated with the universal managing circuit is significantly increased, with a peak value above 14.3 mA under C1 of 3.3 μF. The amount of transferred charges corresponding to one current peak is over 21.67 μC, as shown in Figure 3 H. The charge transfer efficiency (η) of the universal managing circuit can be calculated based on the following equation: η = Q o u t n Q T E N G here, Q out from managing circuit is about 21.67 μC, n = 3 is charging cycles of the TENG, and Q TENG is 7.52 μC. Hence, the efficiency η of the universal managing circuit is calculated to be about 96.1%, which means the energy loss within this managing circuit is lower than 5%. As shown in Figure 3 I, when the voltage of external Cap (C1) is charging by TENG increase to about 5.2 V, the voltage regulator controls the C1 to discharge; then, the voltage of C1 decreases to 3.8V, meanwhile the output current is significantly increased to a peak value, which is basically similar to the previous circuit simulation. The amplitude of the external Cap (C1) charging curve variation is basically maintained in the range of 3.8V–5.2V, as shown in Figure S4 . For exploring the influence of the external Cap (C1) on the output current, the output short-circuit current is measured under different capacitors. When the external Cap (C1) is 47 μF, the short-circuit current is about 94 mA (see Figure 3 J). As shown in Figure S5 , the interval time of output current decreases with the increase of force, but the peak value remains basically unchanged. Finally, we use the same TENG with and without managing circuit to charge a capacitor of 470 μF. It has been found that the charging speed of the TENG with a managing circuit is 2.5 times faster than that without a managing circuit, as shown in Figure 3 K. When the capacitor voltage reaches 3.3 V, the voltage will maintain stable at 3.3 V, which can protect the subsequent electronic devices or sensors from high voltage breakdown. As shown in Figures S6 A and S6B, the best matching resistance of UMC-based GF-TENG (500Ω) is much smaller than the best matching resistance of GF-TENG (1MΩ), which is helpful for UMC to be used in practical functional circuits. Moreover, the maximum power of UMC-based GF-TENG is three times of the maximum power of GF-TENG, as shown in Figure S7 . Figure 3 Comparison of the results of electric measurements for a GF-TENG and UMC-based TENG (A) Schematic working principle and material composition of GF-TENG. (B) The simulated potential distributions for the TENG at three different rolling displacements by COMSOL software employing the finite element method (FEM). Measurement results of the (C) open-circuit voltage, (D)short-circuit current, and (E) amount of transferred charges of the GF-TENG (E), respectively. Inset of (E): connection with rectifier bridge AC-DC for the measurement of amount of transferred charges. Measurement results of the (F) open-circuit voltage, (G) short-circuit current, and (H) amount of transferred charges of the UMC-based TENG, respectively. Inset of (H): connection with rectifier bridge AC-DC for the measurement of amount of transferred charges. (I) The voltage of external Cap (C1) charging by UMC-based TENG and the output current of UMC-based TENG. (J) Measured output current with variable external Cap (C1). (K) Comparison for the charging performance of the UMC-based GF-TENG and GF-TENG. The managing circuit can be applied for facilitating the energy harvesting of ocean waves. As can be seen in Figure 4 A, a wireless ocean sensory system based on rolling-structured TENG (RS-TENG) and universal managing circuit is prepared. The schematic diagram of two polymethyl methacrylate spherical shells as well as the related TENG device is shown is Figures 4 B and 4C. The inner shell is an RS-TENG consisting of a rolling PTFE ball and two stationary Cu electrodes as the electrification materials as shown in Figure 4 B (i) and (ii). Driven by wave vibrations, the freestanding ball can roll back and forth between the two electrodes, providing alternating current to the external load. The outer shell can isolate electronic devices from seawater and keep stable environmental humidity for the operation of generators. The universal managing circuit and a Bluetooth low energy (BLE) sensor beacon are placed in the gap between the two polymethyl methacrylate spherical shells, where poly lactic acid made by 3D printing can help to fix the position shift. The BLE sensor beacon and its receiver are shown in Figure S8 . In detail, diameters of the outside shell sphere and inside one are 9 cm and 7 cm, respectively, and the diameter of the PTFE ball is 2 cm, as shown in Figure 4 C i, ii, and iii. The operating principle of the proposed RS-TENG is based on the conjugation of the triboelectric effect and electrostatic induction. When the freestanding PTFE ball rolls on the top of the right-hand Cu electrode, equal amounts of charges with different polarities generate on the top surface of the Cu and the surface of the PTFE ball ( Figure 4 B (i)). When the PTFE ball rolls from the right electrode toward the left electrode, the negative charges flow from the left electrode to the right electrode via the external circuit due to the electrostatic induction ( Figure 4 B (ii)). COMSOL Multiphysics software based on finite-element simulation is employed to calculate the potential distribution across the two electrodes at five different states, as shown in Figure 4 D i-v, and the variation of this potential difference induces displacement current in the external circuit. Figure 4 A demonstration of a wireless distribution ocean sensor system based on RS-TENG and UMC (A) Photograph of wireless distribution ocean sensor balls floating on seawater. (B) Structure and basic operations of an ocean senor system ball integrated with RS-TENG and UMC. (C) (i) Photograph of one integral ocean senor system ball as fabricated. (ii) Photograph of an RS-TENG, sensor, and UMC, and (iii) photograph of one PTFE ball (the three scale bars represent 1 cm.). (D) The simulated potential distributions for an RS-TENG at five different rolling displacements by COMSOL software employing the finite element method (FEM). (E) Photograph showing three integral ocean senor system balls sending data to the receiver. (F) System diagram of a wireless distribution ocean sensor system based on RS-TENG and UMC. (G) Charging curves comparison of 400 μF under RS-TENG with UMC and without UMC. (H) Temperature and humidity data from three wireless distribution ocean sensor balls working for 7 hr. A simplified condition with continuous water ware is prepared, while three wireless ocean sensors are tested with real water ware, as shown in Figure 4 E and Video S3 of the Supplemental information . The temperature of the water and the humidity of the inside ball can be recorded by these wireless ocean sensors, and the related data can be sent to the receiver on the computer side via wireless BLE sensor beacon (see computer screen). The whole process of data recording and data sending is powered by the shell TENG with universal matching circuit. In detail, as shown in Figure 4 F, the universal managing circuit integrated in a spherical TENG can power a BLE sensor beacon to monitor the temperature and humidity change. The energy from the TENG with UMC can support the BLE sensor beacon to send the hexadecimal data to the receiver on the computer, while the stable output voltage of UMC-based TENGs can match the voltage of the BLE sensor beacon. Finally, the computer translates hexadecimal data from all three BLE sensors into decimal data information and shows all these data in the form of tables and color cloud diagrams. The details of the charging processes and waveform of BLE sensor are shown in Figure S9 . In order to further verify the contribution of the universal managing circuit to the operation of ocean wave TENGs, we connected the output terminal of the universal managing circuit to a capacitor (400 μF). The capacitor can be charged to 3.26 V within 4 min with the help of the universal managing circuit. Meanwhile, if we remove the universal managing circuit, the same TENG needs about 8.5 min to charge the similar capacitor, as shown in Figure 4 G. The output current of RS-TENG and UMC-based RS-TENG is shown in Figures S10 and S11 , respectively. The long-term stability of the device is a very important parameter for the real operation of these self-powered systems. As shown in Figure 4 H, the wireless distribution ocean sensor system is put in the actual water wave environment for 7 hr, and the three BLE sensor beacons can stably send their data to the computer receiver. As shown in Figure S12 , the time required to collect the charges by UMC is about 1.5 min. Hence, the TENG combined with universal managing circuit can offer a feasible power solution to the long-term, wide-area, and near real-time monitoring of climate change in the open sea. Video S3. Ocean environmental sensor, related to Figure 4 The similar environmental sensor system can also be driven by flexible leaf-shaped TENGs (LS-TENGs) and natural wind energy. As shown in Figure 5 A, LS-TENGs collect wind energy, and the universal managing circuit converts the energy into suitable electrical signals that can directly drive environmental sensors. Then, environmental information data can be transmitted to the mobile phone through low-power Bluetooth, as shown in Video S4 and Figure S13 . An environmental sensor system based on flexible LS-TENGs is exhibited in Figure 5 B, which is consisted of three parts: TENGs, universal managing circuit, and BLE sensor beacon. The TENG is a typical leaf-shaped device with six strips made by a PTFE film, Cu foil, and polyethylene terephthalate (PET) film, while the detailed design concept can be found in the previous studies ( Ren et al., 2020a , 2020b ; Zheng et al., 2018 ; Wang et al., 2020a , 2020b , 2020c ; Xu et al., 2020 ). The contact-separation motions of six strips orderly can harvest environmental wind energy ( Ren et al., 2019 ; Zhang et al., 2016 ). PET is selected as the backbone of the strip, mainly owning to its decent strength, low cost, and good machinability. In order to enhance the effective contact area of PTFE polymer and Cu, the nanostructure of PTFE polymer surface is created on the exposed PTFE surface by a top-down method of reactive ion etching. In order to consolidate the working principle of the TENG, COMSOL is employed to simulate the periodic potential variation between the two electrodes during contact-separation motions, as demonstrated in Figure 5 C i and ii. The long-term stability of the device is shown in Figure 5 D, where the wireless environmental sensor system is put in the actual windy environment for more than 7 hr and the BLE sensor beacon can stably send environmental information data to the computer receiver. Figure 5 Two demonstrations of a wireless environmental sensor system and a self-powered indoor position system (A) Photograph of a wireless environmental sensor system harvesting wind energy and sending environmental sensor data to the mobile phone. (B) Structure and basic operations of a wireless environmental sensor system integrated with LS-TENGs and the universal managing circuit. (C) The simulated potential distributions for an LS-TENGs unit by COMSOL software employing the finite element method (FEM). (D) Temperature and humidity data from a wireless environmental sensor system working for 7 hr. (E) The TENG and BLE Bluetooth assembled into the shoes to serve as a self-powered indoor position system. (F) Simulation of a realistic human walking by using a linear motor for drive the layer TENG. Inset (i) is the photography of BLE beacon. (G) Illustration of the working mechanisms of the TENG, and potential distribution for the distance between two triboelectric layers (or air gap) changes from 2 to 0.1 mm. (H) Charging curves comparison of 470 μF under the layer TENG with UMC and without UMC. (I) The variation of RSSI with different distances. Video S4. Harvesting windy energy, related to Figure 5 In addition to this environmental energy harvesting, the universal managing circuit can also work with the TENG for collecting energy from human daily motions. For example, as shown in Figure 5 E, the multilayer TENG integrated with human shoes can harvest energy from walking, while the universal managing circuit helps to achieve a sustainable power supply for mobile BLE beacon. The stable output voltage of UMC-based TENGs can match the voltage of the BLE beacon. Accordingly, the self-powered indoor position system based on this multilayer TENG is illustrated in Figure S14 . During the measurement, the multilayer TENG is driven by linear motor with a motion frequency of 1 Hz, as shown in Figures 5 F and Video S5 , while the universal managing circuit is made by soft substrate, in order to work with this wearable system. The detailed fabrication of the soft universal managing circuit is shown in Experiment part and Figure S15 . As shown in Figure 5 G, the electrification layers (PTFE and Cu) of the multilayer TENG show continuous contact-separate motions during human walking. This contact-separation working mechanism ensures little abrasion of the two surfaces, which helps to enhance the durability of the TENG. The basic working principle of this contact-separation mode of TENGs is simulated by using COMSOL software (see Figure 5 G i and ii), which utilizes the conjugation between contact electrification and electrostatic induction. The capacitor of 470 μF should be charged to 3.3 V, in order to power the BLE beacon for transmitting the signal to receivers. It has been found that the charging speed of this multilayer TENG with universal managing circuit is increased about 2.4 times, as shown in Figure 5 H. The peak of output current of the multilayer TENG is about 7.2 μA, while this value can reach 15.3 mA with the help of the universal managing circuit, as shown in Figures S16 and S17 . Video S5. Powering position system, related to Figure 5 The value of received signal strength indicator (RSSI) represents the distance from the receiver to the transmitter (BLE beacon). To estimate the distance based on the signal from a BLE beacon, a path loss model is needed. Here, we adopt the path loss model ( Zuo et al., 2018 ): RSSI = − ( 10 n l o g 10 d + A ) where parameter A is the absolute RSSI value represented by dBm at 1 m away from the beacon, n is a parameter related to the signal propagation environment, and d is the distance from the beacon. In our implementation, we use this method to estimate the parameters. The RSSI of BLE beacon is measured under different distances between Bluetooth and receiver (mobile phone) as shown in Figure 5 I. Finally, the path loss model with pre-defined parameter in our implementation is RSSI = − ( 10 ∗ 3.3 ∗ l o g 10 d + 69 ) In order to determine the specific coordinates of a point in a two-dimensional plane, three receivers are used to get three distances from the point. The detail of how to position the person wearing the position system is shown in Figure S18 A and S18B ( Supplemental information ). This demonstration not only offers a strategy to power the BLE beacon by TENGs but also provides useful guidance for constructing a self-powered position platform used in special environment like mine and underground parking. Conclusions In summary, we have developed a different universal managing circuit as a link between TENG and functional electronics devices, such as wireless environmental sensors or microcontrollers. The universal managing circuit integrated with the TENG, well matching the high output impedance of TENGs, can provide an output current in the milliampere level, while the stable output voltage from this universal managing circuit can be stabilized at a designed value to power different wireless devices for preventing high voltage breakdown. The energy managing process of this circuit can be achieved automatically without the need of extra battery source, and the energy loss within this managing circuit can be maintained to be lower than 5%. The universality and efficiency of this universal managing circuit is verified by using the TENG with different working modes and structures. Three environmental sensor systems, which can harvest different micro-energies (water wave energy, windy energy, and walking energy), are demonstrated to verify the stability, versatility, and practicality of the universal managing circuit. The proposed universal managing circuit can be used for constructing wireless and distributed sensor systems with minimized energy consumption to meet challenges arose from robotics, wearable devices, environmental monitors, and the era of internet of things."
} | 8,140 |
23144897 | PMC3492453 | pmc | 1,834 | {
"abstract": "Division of labor, an adaptation in which individuals specialize in performing tasks necessary to the colony, such as nest defense and foraging, is believed key to eusocial insects' remarkable ecological success. Here we report, for the first time, a completely novel specialization in a eusocial insect, namely the ability of Cataglyphis cursor ants to rescue a trapped nestmate using precisely targeted behavior. Labeled “precision rescue”, this behavior involves the ability of rescuers not only to detect what, exactly, holds the victim in place, but also to direct specific actions to this obstacle. Individual ants, sampled from each of C. cursor's three castes, namely foragers, nurses and inactives, were experimentally ensnared (the “victim”) and exposed to a caste-specific group of potential “rescuers.” The data reveal that foragers were able to administer, and obtain, the most help while members of the youngest, inactive caste not only failed to respond to victims, but also received virtually no help from potential rescuers, regardless of caste. Nurses performed intermediate levels of aid, mirroring their intermediate caste status. Our results demonstrate that division of labor, which controls foraging, defense and brood care in C. cursor , also regulates a newly discovered behavior in this species, namely a sophisticated form of rescue, a highly adaptive specialization that is finely tuned to a caste member's probability of becoming, or encountering, a victim in need of rescue.",
"introduction": "Introduction Division of labor, one of the most prominent and widely studied features of colony behavior in social insects [1] – [6] , takes one of two general forms: morphological polyethism , in which workers' size and/or shape determines what tasks they will perform; and, temporal polyethism , in which individuals perform different tasks as they mature [1] – [7] . Temporal polyethism is widespread in social insects and typically follows the pattern of younger workers performing tasks within the nest and older workers performing tasks outside, such as foraging and defense [1] – [7] . Presumably, this behavioral specialization, which is thought responsible for social insects' enormous ecological success, increases the overall efficiency of the colony because workers that focus on and repeat a particular task will perform it more reliably [2] , [6] . Cataglyphis cursor , a sand-dwelling Mediterranean ant, exhibits temporal polyethism in which foragers, typically the oldest members of the colony, are responsible for securing food, nurses specialize in brood care, and inactives, the youngest workers, remain near the brood but almost never tend them [8] , [9] . Nowbahari et al. [10] have shown that C. cursor ants also are capable of highly sophisticated rescue behavior. That is, when an individual becomes entrapped, as often happens in nature when it is caught under collapsing sand or debris, or falls into a predatory antlion larvae pit [11] , [12] , nearby nestmates begin by digging near the victim and pulling on its limbs, a very simple form of rescue behavior observed in several ant species [13] – [15] . In addition, however, C. cursor rescuers somehow are able to identify exactly what holds the victim in place, to transport sand away from that obstacle, and then, as illustrated in Figure 1 , to target their bites precisely to it alone, excavating sand as necessary to expose the obstacle further [10] . Carefully aimed, biting at the obstacle never is misplaced, even though it may be in direct contact with the victim's body. 10.1371/journal.pone.0048516.g001 Figure 1 Precision rescue behavior. In addition to digging near an entrapped nestmate and pulling on its limbs, a very simple form of rescue behavior observed in several ant species, C. cursor ants somehow are able to identify exactly what obstacle holds the victim in place, to transport sand away from that obstacle, and then to target their bites precisely to it alone, excavating sand as necessary to expose the obstacle further. We have labeled this behavior “precision rescue.” Here, a C. cursor rescuer already has transported sufficient sand away from the victim, exposing the nylon thread snare holding its nestmate in place (part of the white filter paper has been exposed as well), and is pictured biting the snare that holds the victim to the paper. Carefully aimed, snare biting never is misplaced, even though the snare has been tied snugly around the pedicel (waist) of the victim and is in direct contact with the victim's body. Photograph by Paul Devienne. Subsequent observations of C. cursor revealed, however, that not all adults administered help and not all victims were able to elicit help – differences that might reflect other aspects of individuals' division of labor, their physiological maturation or both. That is, because division of labor in C. cursor follows an age polyethism pattern [16] , C. cursor foragers, as in all Cataglyphis ant species, are among the older colony members. Foragers, which are capable of high individual nestmate discrimination abilities, are physiologically more mature and, thus, for these reasons we predicted that they would be more likely both to give and to receive aid. Nest-bound inactives, on the other hand, younger individuals that are less physiologically mature, might be less able not only to call for help, but also to provide aid. Finally, nurses, specialized for brood care, might require some of the same behavioral patterns needed by efficient rescuers. Consequently, in the present study, we examined the role of polyethism in the rescue behavior of C. cursor ants by conducting tests of rescue behavior in which we systematically varied the caste of both victim and rescuers. Based on their physiological maturation and their other specializations, we predicted that foragers and inactives would differ substantially in their ability both to give and receive aid, with nurses possibly intermediate between these two castes. As predicted, our results show that temporal polyethism, which controls foraging, defense and brood care in C. cursor , also regulates the capacity of these ants to deliver precision rescue behavior. We suggest that this highly adaptive specialization has been finely tuned through evolution to match a caste member's probability of becoming, or encountering, a victim in need of rescue.",
"discussion": "Discussion Our data reveal a novel behavioral specialization in a eusocial insect, a specialization never before reported in the literature, either in ants or in any other eusocial insect. In support of our hypothesis, the expression of specialized rescue behavior in C. cursor is characterized by a form of temporal polyethism in which, as workers mature and assume the duties of nurses and foragers, they are more likely to respond quickly to a nestmate in distress and to persist in that behavior for a longer time. Furthermore, our results suggest that caste membership determines not only the ability to provide aid, but also to receive it. That is, foragers were able both to administer, and to obtain, the most help; inactives were incapable of responding to victims, as well as incapable of eliciting help from potential rescuers, regardless of caste; and, nurses generally performed intermediate levels of aid, mirroring their intermediate caste status. According to Retana and Cerdá [8] , division of labor in C. cursor adult workers is based on a well-defined temporal polyethism in which young workers are initially inactive, and then, as they mature, perform various nest-related tasks, and finally leave the nest to become foragers. This same pattern is exhibited by several other Cataglyphis species, including Cataglyphis bicolor \n [17] , and Cataglyphis niger \n [18] . The differences we observed in C. cursor ants' ability to rescue nestmates map easily on this pattern of temporal polyethism, a pattern that places some castes of workers at greater risk of entrapment and, thus, in greater need of the capacity to give and receive aid. For example, C. cursor foragers, like other Cataglyphis ants, do not form ant trails to food, but search individually, relying on their highly-developed orientation abilities [1] , [19] – [22] . Under the hot desert conditions experienced by these ants, entrapment easily could be lethal. Because foragers are the sole providers of food, but represent only 14.6% of the workers [8] , a trapped forager represents a potentially large cost to the colony. Natural selection can ameliorate this cost, however, through specialized rescue behavior, a mechanism that enables foragers both to call for help, and to respond to the call of another forager. Like foragers in some ways, nurses also must respond to nestmates, in their case larvae, frequently moving them using nearly the same pulling behavior involved in rescue [1] , [8] . Thus, nurses' ability to rescue a trapped nestmate is not surprising. Nonetheless, our results show that they are not as expert in rescue behavior as foragers, a difference that could reflect their slightly less mature development, their inexperience with trapped nestmates, or both. Finally, because inactives, the youngest workers, also never leave the nest but, unlike nurses, have no responsibility for brood care, they have no need of a capacity to rescue nestmates. This age-dependent division of labor in C. cursor almost certainly reflects workers' physiological maturation, including both brain development, as has been demonstrated in another Cataglyphis species, C. albicans \n [23] , as well as glandular development [3] . For example, in Myrmica rubra ant workers, the volume of secretions produced by the Dufour and poison glands, which are used to signal alarm, increases with the age of workers [24] . In a study of C. cursor pheromones, we found some evidence that these same two glands are involved in rescue behavior (unpublished data). Thus, because C. cursor foragers are the oldest workers, they would be expected to possess more developed glands, which would enable them to emit a more intense alarm signal than less-developed nurses; in turn, nurses would be expected to signal more strongly than even less-developed inactives. Because, in the current study, the ability to receive aid from rescuers – which reflects ants' ability to signal their distress – generally matched their ability to provide aid to victims, our results suggest that both sending and perceiving the distress signal develops concurrently. In sum, our study shows that precision rescue behavior, a highly developed and complex behavior in which C. cursor ants somehow are able to identify exactly what holds an entrapped nestmate in place and then to target their behavior to it alone, is regulated by temporal polyethism, a form of division of labor in which adult workers perform different tasks as they mature. Although several anecdotal reports of rescue behavior exist in the scientific literature [25] , the ability to perform specifically targeted rescue behavior – what we call precision rescue – has been studied experimentally in only two species, namely ants [10] and, very recently, in rats [26] . Although researchers have yet to determine why some species possess this capability and others do not, the ability of C. cursor to rescue its nestmates appears to have evolved to meet the particular risks it faces in its harsh environments."
} | 2,874 |
36540214 | PMC9733714 | pmc | 1,835 | {
"abstract": "Self-healing materials have attracted widespread attention owing to their capacity to extend the lifetime of materials and improve resource utilization. However, achieving superior mechanical performance and high self-healable capability simultaneously under moderate conditions remains a long-standing challenge. Integrating multiple dynamic interactions in waterborne polyurethane (WPU) systems can overcome the above-mentioned issue. Herein, environmentally friendly WPU systems containing multiple hydrogen bonds and boronic ester bonds in their polymer backbones were synthesized, where 2,6-diaminopyridine (DAP) and boric acid (BA) served as a dynamic chain extender and reversible cross-linking agent, respectively. The chain structure of the polymer was adjusted by controlling the ratio (DAP/BA) of hard segments, which could effectively meet the requirement of mechanical robustness and desirable self-healable efficiency. Benefiting from multiple dynamic interactions, the prepared WPU elastomer exhibited good mechanical properties, such as tensile strength (from 18.89 MPa to 30.78 MPa), elongation (about 900%) and toughness (from 54.82 MJ m −3 to 92.74 MJ m −3 ). Driven by water and heat, the IP-DAP 40 -BA 10 -WPU film cut in the middle exhibited good self-healing ability, with healing efficiencies of tensile stress of 90.74% and elongation of 91.29% after self-healing at 80 °C for 36 h. Meanwhile, the synthesized WPU elastomer exhibited good water resistance and thermal stability. This work presents a novel way to design robust self-healable materials, which will have wide promising applications in flexible electronics, smart coatings and adhesives.",
"conclusion": "4 Conclusions In summary, a series of environmentally friendly waterborne polyurethanes (IP-DAP x -BA y -WPU) with desirable mechanical strength and self-healable capacity at mild temperature were fabricated by employing 2,6-diaminopyridine (DAP) as a dynamic chain extender and boric acid (BA) as a dynamic cross-linking agent, which contained multiple hydrogen bonds and boronic ester bonds in their polymer backbone. The prepared waterborne polyurethanes (IP-DAP x -BA y -WPU) exhibited excellent storage stability, water resistance and thermal stability. Varying the ratio n (DAP/BA) of hard segments and adjusting the chain structure of the polymer could effectively overcome the contradiction between mechanical strength and healing efficiency. IP-DAP 30 -BA 20 -WPU exhibited good mechanical properties, such as tensile stress (30.78 ± 1.56 MPa), stretchability (930.12 ± 48.54%) and toughness (92.74 ± 2.42 MJ m −3 ), respectively. It was found that a wide scratch on IP-DAP 40 -BA 10 -WPU film healed at 80 °C with the support of water, and after 1 h was almost no visible fracture trace. Driven by both water and heat, they shortened the healing time and improved the healing efficiency. The IP-DAP 40 -BA 10 -WPU film was cut in the middle and healed at 80 °C for 36 h with the aid of water, exhibiting desirable self-healing properties and the healing efficiency of tensile strength and breaking elongation were above 90.0%, which could hold a weight of 2.25 kg easily. Thus, it shows broad application prospects in flexible electronics, smart coatings and adhesives.",
"introduction": "1 Introduction Self-healable polymers are materials that can repair their mechanical damage either autonomously or under specific stimuli, 1 endowing these materials with self-healing ability, which can not only significantly improve their safety, stability and extend their lifetime but can also effectively reduce resource wastage and maintenance costs. 1–3 In recent years, self-healing polymer materials as primary media have shown considerable application prospects in flexible electronics and wearable devices, such as flexible electronic skin, intelligent coatings and durable sensors. 4–8 The polyurethane elastomer matrix with good flexibility and abrasion resistance, which can appropriately satisfy the robustness requirement, has attracted wide attention in scientific and engineering circles. 9,10 Healable materials can be divided into intrinsic and extrinsic self-healable materials according to their healing mechanism. 1,11 A typical characteristic of extrinsic self-healable polymers is releasing the healing agent embedded in them to complete damage repair. 12–14 However, their preparation process is usually complicated and their low healing efficiency severely hinders their further development. 5 By contrast, intrinsic self-healable polymers mainly rely on the interactions of dynamic covalent bonds or noncovalent bonds in the polymer backbone. Dynamic covalent bonds mostly include urea bonds, 15,16 Diels–Alder bonds, 11,17 boronic ester, 12,18–21 disulfide bonds, 22–24 diselenide bonds 25 and phenol–carbamate bonds. 26 Noncovalent interactions include hydrogen bonds, 27–32 metal–ligand coordination, 33–37 ionic interactions, 38,39 host–guest interactions, 40 and π–π stacking interactions. 41,42 However, healing via DA ring addition and phenol–carbamate bonds requires high temperatures of up to 120 °C, which consumes a lot of energy. 43 Meanwhile, self-healing PU embedded with dynamic metal-coordinated bonds can heal at room temperature, but this usually takes a long time and affords poor mechanical properties. Bao et al. 44 prepared polyurethanes with dynamic metal–ligand coordination and hydrogen bonds simultaneously, which exhibited high a healing efficiency of 98% and healed at ambient temperature for two days, but its tensile strength was 1.8 MPa. Zhang et al. 5 prepared polyurethanes with dynamic Cu–DOU coordination and hydrogen bonds for healing under ambient conditions, which exhibited excellent toughness (87 MJ m −3 ), but healing took about six days, limiting its practical application. At present, the research hot spot is how to overcome the conflict in mechanical performance and self-healing efficiency, striving to obtain elastomer materials with outstanding mechanical properties and desirable self-healing ability under mild conditions. Particularly, hydrogen bonds play a significant role in constructing self-healable polyurethanes because of their directivity and reversibility. Recently, Xia et al. 45 introduced hierarchical H-bonds in novel self-healable polyurethane backbones, which possessed high tensile stress (34.1 MPa) and superior toughness (127.3 MJ m −3 ). Boric acids can form diverse reversible covalent and noncovalent bonds, and thus widely employed to construct self-healing polymer materials. The orientation of boric acid and boronic ester/boric acid can be adjusted easily by heating, adding water or Lewis base. 46,47 For example, Kai et al. 12 incorporated reversible boronic ester bonds, hydrogen bonds and B–N coordination in polyurethane healed at ambient temperature, which exhibited excellent tensile stress of 10.5 MPa and superior stretchability of 3120%. Previous researchers have done much work on solvent-based polyurethanes with self-healable capacity. However, little research has been done on self-healable WPU through the combination of dynamic hydrogen bonds and boronic ester bonds. Also, the trend of environmental friendliness, avoiding the use of harmful volatile organic compounds (VOCs), and taking full advantage of waterborne polyurethane should be considered. Herein, we introduce multiple hydrogen bonds and boronic ester bonds in the molecular structure of WPU to prepare a self-healable WPU elastomer. PTMG with flexible chain mobility for better self-healing, acted as the soft segment. IPDI was chosen as the hard segment with asymmetric, non-rigid and bulky structure, which can restrain the crystallization and facilitate the chain mobility. 5 DAP was selected as the chain extender and the primary amine group produced a urea group after chain extension, resulting in dihydrogen bonding, which may have a positive influence on the simultaneous improvement of mechanical performance and self-healing ability. BA was used as a dynamic crosslinker to form cross-linking points based on reversible boronic ester bonds in the polyurethane system, which not only improved the mechanical strength of the material, but also induced a reversible hydrolysis reaction, promoting the self-healing ability. Designing the recipe, the chain structure of the polymer was adjusted by controlling the ratio n (DAP/BA) of hard segments, the combination of multiple hydrogen bonds and boronic ester bonds, which could effectively satisfy the demands of good mechanical performance and desirable self-healing efficiency. Finally, we obtained water-borne polyurethane with good storage stability, water resistance, high strength and toughness, which also exhibited good self-healing ability by surface wetting and heating at 80 °C.",
"discussion": "3 Results and discussion 3.1 Particle size and stability of IP-DAP x -BA y -WPU emulsions The particle size and distribution are important physical parameters of the WPU emulsion, which are strongly associated with its appearance and dispersion stability. When water was added during the emulsification process, the polyurethane could form colloidal particles through phase inversion to achieve efficient dispersion in the aqueous phase. As shown in Fig. 1a , with an increase in boric acid content, the average particle size of the IP-DAP x -BA y -WPU emulsions increased gradually, which was 29.42 nm, 30.30 nm, 30.77 nm, 32.47 nm, 44.71 nm, 50.92 nm, respectively. Meanwhile, the distribution was uniform for all the samples, where the polydispersity (PDI) was in the range of 0.128 to 0.225. Usually, the particle size of a WPU emulsion is closely related to its chain stiffness, crystallinity, hydrophilic groups and cross-linking structure. 23 With an increase in the content of the boric acid cross-linking agent, the waterborne polyurethane molecular junction increased and the molecular chains had poor mobility, which could not assemble quickly to form latex particles in the emulsification, and thus the tightness of the formed latex particles became worse, and more boric acid also made it difficult to disperse in water, leading to larger particle sizes. Fig. 1b shows the color and appearance of the IP-DAP x -BA y -WPU emulsions in natural light, which were bright and atrovirens. As the particle size increased, the color deepened. Zeta potential is an important indicator to characterize the stability of the dispersion system, where a high absolute zeta potential value indicates that sufficient electrostatic repulsion exists on the surface of the WPU particles. The absolute value was in the range of 35 to 50 mV for all the IP-DAP x -BA y -WPU emulsions ( Fig. 2b ), which exhibited distinct values. 50 As shown in Fig. 1c , the IP-DAP x -BA y -WPU emulsions left for 40 days were centrifuged at 3000 rpm for 0.5 h, which all displayed good storage stability without evident precipitation or stratification. Fig. 1 (a) Particle size and distribution of the as-prepared WPU particles. Color and appearance of IP-DAP x -BA y -WPU emulsions (b) before and (c) after centrifugation in natural light. Fig. 2 (a) Water contact angle (WCA) and water absorption (WA) of different IP-DAP x -BA y -WPU samples. (b) Zeta potential values of different IP-DAP x -BA y -WPU emulsions. (c) FT-IR spectra of BA, DAP and IP-DAP 40 -BA 10 -WPU in the frequency range of 650–4000 cm −1 . (d) FTIR spectra of IP-DAP 40 -BA 10 -WPU in the C \n \n\n<svg xmlns=\"http://www.w3.org/2000/svg\" version=\"1.0\" width=\"13.200000pt\" height=\"16.000000pt\" viewBox=\"0 0 13.200000 16.000000\" preserveAspectRatio=\"xMidYMid meet\"><metadata>\nCreated by potrace 1.16, written by Peter Selinger 2001-2019\n</metadata><g transform=\"translate(1.000000,15.000000) scale(0.017500,-0.017500)\" fill=\"currentColor\" stroke=\"none\"><path d=\"M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z\"/></g></svg>\n\n O stretching vibration region. 3.2 Surface wettability and water resistance As shown in Fig. 2a , with an increase in the amount of boric acid, the water contact angle (WCA) increased from 86.09° to 102.75°. Many polar hydrophilic groups dissociated on the surface of the IP-DAP 50 -BA 0 -WPU film containing only the chain extender, showing hydrophilicity. The crosslinking agent of BA will react with the hydrophilic groups to form cross-linking points inside, reduce the hydrophilic groups on the surface, and form a smooth film with lower surface energy, resulting in stronger hydrophobicity. When the amount of boric acid increased to a certain level, the residual hydrophilic polar groups on the surface increased, the WCA of the IP-DAP 25 -BA 25 -WPU film decreased to 93.65° (Fig. S1 † ). Generally, the water absorption (WA) is influenced by the hydrophilic composition content and film surface structure. In our study, the water absorption of the specimens decreased with an increase in the BA/DAP molar ratio, which was only 7.9% of IP-DAP 35 -BA 15 -WPU. As the degree of cross-linking in the system gradually increased, the film surface formed was denser, which left fewer polar groups on the surface and exhibited better water resistance. When the cross-linking density and the emulsion particle size were too large, more defects were left on the surface, making it easy for water to invade from these defects. The fractured surface morphology of the representative WPU films was further observed by SEM. As shown in Fig. S4, † with an increase in the content of BA crosslinking agent, the fractured surface gradually became rougher, which showed a typical ductile fracture. 10 The surface of IP-DAP 25 -BA 25 -WPU film was irregular and coarse, and thus the water absorption increased to 20.85%, which exhibited poor water resistance. In general, the synthetic polyurethane film had good water resistance, which overcame the disadvantage of poor water resistance of traditional waterborne polyurethane. 3.3 Chemical structure of IP-DAP x -BA y -WPU films The chemical structure of the IP-DAP x -BA y -WPU samples was verified by FT-IR analysis. As depicted in Fig. 2b and S2, † the peaks at around 3320 cm −1 and 1549 cm −1 correspond to the –NH stretching and bending vibration of the urethane groups, respectively. The peaks at 2856 cm −1 and 2942 cm −1 correspond to the stretching vibration of –CH 2 –. 22 The peak at 1110 cm −1 is assigned to the stretching vibration of C–O–C. The characteristic peaks at 1592 cm −1 and 1370 cm −1 are attributed to the C N vibration mode of the free pyridine of 2,6-diaminopyridine (DAP) and B–O vibration mode of boric acid (BA), respectively, demonstrating that the DAP and BA moieties were introduced in the polymer backbones of IP-DAP x -BA y -WPU effectively. 51 Moreover, the disappearance of the typical –NCO peak at around 2260–2280 cm −1 indicates the completion of the reaction. 10,52 The peaks at 1740 cm −1 to 1620 cm −1 are attributed to the different vibrations of C O, which were further analyzed by peak fitting and differentiation, as shown in Fig. 2d . The C O peaks were divided into five sub-peaks in the spectrum of IP-DAP 40 -BA 10 -WPU. The peak at 1714 cm −1 is attributed to the free C O of urethane, and that at 1700 cm −1 attributed to the vibration of the ordered H-bond. The peak at 1673 cm −1 is assigned to the free C O of urea and that at 1652 cm −1 to the disordered H-bond vibration of urea. 53 The vibration of the ordered H-bond in the urea appeared at 1637 cm −1 . These results indicate that the content of H-bonded C O in the IP-DAP 40 -BA 10 -WPU was higher than 53.80%. Notably, the proportion of hydrogen bonds in IP-DAP 50 -BA 0 -WPU, IP-DAP 45 -BA 5 -WPU, IP-DAP 35 -BA 15 -WPU, IP-DAP 30 -BA 20 -WPU, and IP-DAP 25 -BA 25 -WPU are 46.82%, 51.62%, 43.79%, 39.75% and 31.25%, respectively (Fig. S3 † ). The formation of hydrogen bonds in C O is mainly due urethane–urethane, urethane–urea, and urea–urea, in which the bonding energy of H-bond constructed by urea–urea is much higher than that with urethane–urethane. 32,45 Some of the boronic ester replaced the urea bond, where a bond with a lower bonding energy is easier to form, which may be more conducive to the formation of ordered hydrogen bonds. The presence of multiple hydrogen bonds had a positive effect on promoting the mechanical performance and self-healing capacity. 3.4 Thermal properties of the IP-DAP x -BA y -WPU films The thermal stability of polyurethane is critical for its practical application, which is closely connected with its rigid groups, cross-linked structure and crosslinking density. 54 The thermal degradation process was determined by TGA, as shown in Fig. 3a and b , where it can be observed that the representative IP-DAP 50 -BA 0 -WPU, IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU films all displayed two-stage decomposition steps. The initial weight loss of the film occurred in the first stage at about 230–320 °C, which is attributed to the decomposition of the hard segment. The decomposition temperature in the second stage was 360–438 °C, which is attributed to the degradation of the soft segments. In addition, it was observed that the mass loss of 10% occurred 297.36 °C, 300.62 °C and 303.18 °C, while the mass loss of 50% occurred at 405.19 °C, 407.12 °C and 410.34 °C for the IP-DAP 50 -BA 0 -WPU, IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU samples, respectively. With an increase in the cross-linking density, the thermal decomposition temperature increased slightly, but the effect was not significant, where all the samples exhibited good thermal stability below 200 °C. Fig. 3 (a) TGA curves, (b) DTG curves, (c) DSC curves, and (d) loss factor (tan θ ) of the representative IP-DAP 50 -BA 0 -WPU, IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU samples. According to the DSC analysis, as shown in Fig. 3c , the glass transition temperature ( T g ) of WPU was not distinct in the DSC curve, and thus a DMA test was further employed to determine it visually. As shown in Fig. 3d , the measured glass transition temperatures were −54.47 °C, −55.15 °C and −53.46 °C, respectively. Adjusting the ratio n (DAP/BA) of hard segments, the composition of hydrogen bonds to boronic ester bonds in the polymer chain further changed, but the three films had the same ratio of soft segments and hard segments, which had little effect on T g . This further shows that the samples have good chain mobility, which will have a positive impact on the self-healing process. 3.5 Mechanical properties of IP-DAP x -BA y -WPU films Mechanical properties are important parameters to measure for practical applications, which are affected by various factors such as intermolecular forces, physicochemical cross-linking, and rigidity in the molecular chains. 27 In this work, the molecular chain structure was designed by adjusting the ratio n (DAP/BA) of hard segments, aiming to obtain the desired mechanical robustness and chain mobility synchronously. As shown in Fig. 4a and b , the tensile stress, elongation and toughness were 18.89 ± 0.92 MPa, 898.16 ± 46.32% and 54.82 ± 1.93 MJ m −3 for the pure IP-DAP 50 -BA 0 -WPU film, respectively. With an increase in the boric acid content, the tensile stress and toughness gradually increased, while the elongation remained basically unchanged (Table S2 † ). Particularly, the IP-DAP 30 -BA 20 -WPU film exhibited excellent mechanical properties, which were 30.78 ± 1.56 MPa, 930.12 ± 48.54%, 92.74 ± 2.42 MJ m −3 , respectively. In the case of self-healable waterborne polyurethane, there are few similar high strength and toughness elastomers in other studies. The pristine IP-DAP 50 -BA 0 -WPU elastomer only contained urea doublet hydrogen bonding and urethane singlet hydrogen bonding in its hard segment domains, which serve as physical cross-linking junctions. However, the ratio of n (DAP/BA) could promote the degree of phase separation, the dynamic boronic ester bond could enhance intermolecular force as a point of cross-linking, and also served as sacrificial bonds with hydrogen bonding to increase the hidden length of the folded polymer chain, thus enhancing the mechanical properties. 12 Meanwhile, excessive cross-linking will generate stress concentration, which destroys the specimen before it can disperse more stress, resulting in mechanical deficiency. 26 Undesirably, when using pure boric acid, there were many film-forming defects in IP-DAP 0 -BA 50 -WPU, which exhibited terrible mechanical properties and poor self-healing ability (Table S2 † ). It failed to yield further application value. Thus, based on the above-mentioned results, tuning the moderate degree of cross-linking and components of the hard segments could promote the mechanical properties of self-healable WPU elastomers efficiently. Fig. 4 (a) Stress–strain curves. (b) Tensile strength and toughness of different IP-DAP x -BA y -WPU films. (c) Mechanical performance of the representative IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU films in this work compared with those of other studies. 3.6 Cyclic tensile tests of IP-DAP x -BA y -WPU films The multiple dynamic hydrogen bonds and covalent boronic ester bonds in the films were expected to endow them with high toughness and good elasticity, respectively. Thus, to further assess their performance, we conducted a cyclic tensile test. Taking the IP-DAP 30 -BA 20 -WPU film as an example, there was no waiting period in the continuous cyclic tensile test ( Fig. 5a ), which involved two successive loadings. Following the first small-strain (100%) stretching cycle, there was an entropic drive for the covalent network and the unbroken reversible bonds (hydrogen and boronic ester bonds) to return the network almost to its pristine status. Under sequential cyclic larger strain (≥200%), large hysteresis loops could be seen clearly owing to the energy dissipation resulting from the reformation of broken dynamic bonds at new sites during stretching, but the residual strain remained low. This revealed that IP-DAP 30 -BA 20 -WPU had good elasticity with a strain of 200%, making it suitable for many practical applications. Fig. 5 (a) Sequential cyclic tensile curves of the IP-DAP 30 -BA 20 -WPU film at different strains without waiting time. (b) Cyclic tensile curves of IP-DAP x -BA y -WPU films with a maximum strain of 400%. (c) Residual strain and hysteresis energy of IP-DAP x -BA y -WPU films during the cyclic tensile processes. To further reveal how the hydrogen bonds and boronic ester bonds affected the mechanical properties of the films, cyclic tensile tests were carried out. Fig. 5b and c exhibit the loading–unloading cycle curve at a fixed strain of 400% and comparison with the hysteresis loop, respectively. The IP-DAP 50 -BA 0 -WPU film had the maximum area hysteresis loop area (mechanical hysteresis) compared to the other elastomers. The mechanical hysteresis resulted in the delayed recovery of its configuration and conformation caused by the friction between the polymer chains. In the case of the IP-DAP 50 -BA 0 -WPU elastomer, the hydrogen bonding between the urea group, urethane group and pyridine was the main interchain force that restricted the polymer segment from reforming. For the elastomer with boronic ester bonds, the presence of boronic ester bonds significantly interfered with the formation of hydrogen bonds between the urea groups, and the cross-linking point limited the slip of the molecular chain as well, which was beneficial for the recovery of the polymer chain configuration and conformation, resulting in a slight lag curve. In conclusion, the high elasticity of the elastomers could be significantly improved by adjusting the composition of hard segments reasonably. 3.7 Self-healing properties of IP-DAP x -BA y -WPU films The self-healing performance was firstly studied at 60 °C for different healing times. As shown in Fig. 6 , it can be observed that the strength of the representative samples healed for 36 h of IP-DAP 50 -BA 0 -WPU, IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU was 5.63 MPa, 9.41 MPa, 9.46 MPa, respectively. Although IP-DAP 40 -BA 10 -WPU exhibited the higher healing efficiency of elongation of 72.30%, its strength was undesirable. The motility of the molecular chains was not strong at the lower temperature was 60 °C, and thus higher activation energy would be required. In our study, the self-healing systems via multi-dynamic interactions needed to be exposed to higher external stimulus. Fig. 6 Original and self-healed tensile stress–strain curves (a–c) and self-healing efficiency (d) of IP-DAP 50 -BA 0 -WPU, IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU films after healing at 60 °C for 12 h, 24 h, and 36 h. To further investigate the self-healing behavior of the IP-DAP x -BA y -WPU films, their healing efficiency was studied at 80 °C by the tensile test taking the IP-DAP 50 -BA 0 -WPU, IP-DAP 40 -BA 10 -WPU and IP-DAP 30 -BA 20 -WPU films as examples. As depicted in Fig. 7(a–c) and (e) , after undergoing thermal repair at 80 °C for 40 h, the strength was only 7.32 MPa for the IP-DAP 50 -BA 0 -WPU film, which exhibited low self-healing efficiency of 38.61%. By contrast, the IP-DAP 40 -BA 10 -WPU film exhibited stress of 21.26 MPa and strain of 679.52%, which were 89.07% and 79.02% self-healing efficiency, respectively. Compared with the pure IP-DAP 50 -BA 0 -WPU film, its self-repairing ability had been greatly improved. Therefore, the existence of a too high DAP content with the structural rigidity of its hard segments and lower hydrogen bonding will weaken the mobility of the segments, which is adverse to the healing process. The combination of dynamic reversible non-covalent bonds (multiple hydrogen bonds) and covalent bonds (boronic ester bond) was more favorable for self-healing. The IP-DAP 30 -BA 20 -WPU film had stress of 17.34 MPa and strain of 627.76%, which were 56.33% and 69.85% of the self-healing efficiency, respectively. The higher content of hard segments with boric acid content vastly enhanced its strength, but sharply weakened its chain mobility, giving rise to reduced self-healable capacity. Undesirably, the IP-DAP 25 -BA 25 -WPU film had stress of 8.73 MPa and strain of 422.70%, which were 36.74% and 55.12% of the self-healing efficiency, respectively (Table S3 † ). The research on self-healing materials is focused on how to achieve a satisfactory balance between mechanical performance and healing capacity. Appropriate dynamic crosslinking points not only can strengthen and toughen materials, but also ensure better chain mobility, aiming to effectively repair them under mild conditions. However, too many cross-linking points constrain the molecular chain movement, requiring a higher energy to activate the chain segment and mobility, thus showing poor repairing ability. Consequently, the healing efficiency can be significantly improved by reasonably adjusting the composition of the hard segments. Fig. 7 Original and self-healed tensile stress–strain curves (a–c) of IP-DAP x -BA y -WPU films after healing at 80 °C for 40 h without water. (d) Representative original and self-healed tensile stress–strain curves of IP-DAP 40 -BA 10 -WPU films under different conditions. (e) Self-healing efficiency of IP-DAP x -BA y -WPU films healed at 80 °C for 40 h without water and (f) healed at 80 °C for 36 h with the aid of water. More interestingly, the self-healing behaviors of the IP-DAP x -BA y -WPU films was significantly influenced by water. Specifically, a cut section was soaked in water for 30 s, and then placed in an oven at 80 °C. At regular intervals, water was dripped on the contact surface with a dropper until the healing time was almost over, and then the surface was dried with filter paper. The results show that the healing efficiency improved dramatically with the support of water. It was observed that the typical IP-DAP 40 -BA 10 -WPU film healed for 36 h with the aid of water could be easily stretched to several times of its original length, as presented in Fig. 8b . Furthermore, the IP-DAP 40 -BA 10 -WPU film could hold a weight of 2.25 kg and also showed good self-recovery ability, as shown in Fig. 8c . Fig. 7d and f depicted the representative stress–strain curves and healing efficiency of the IP-DAP 40 -BA 10 -WPU film, which recovered for diverse lengths of time without water or with the aid of water, respectively. The IP-DAP 40 -BA 10 -WPU film healed at 80 °C for 36 h with the support of water, which exhibited the stress of 21.67 MPa and strain of 787.29%, which were 90.74% and 91.29% healing efficiency, respectively. Driven by both water and heat, they not only shortened the repair time, but also improved the healing efficiency (Table S4 † ). Fig. 8 (a) Crack healing photos of IP-DAP x -BA y -WPU film at 80 °C with the surface wetting of water for 1 h observed using a 200× optical microscope: 1-IP-DAP 50 -BA 0 -WPU, 2-IP-DAP 45 -BA 5 -WPU, 3-IP-DAP 40 -BA 10 -WPU, 4-IP-DAP 35 -BA 15 -WPU, 5-IP-DAP 30 -BA 20 -WPU, and 6-IP-DAP 25 -BA 25 -WPU. (b) Stretching picture of IP-DAP 40 -BA 10 -WPU film (75 mm × 4 mm × 1 mm) healed for 36 h at 80 °C with water and (c) subjected to a 2.25 kg weight lifting test. The self-healing ability of the film could be intuitively evaluated through the recovery of its surface scratches by heating on a hot plate at 80 °C with surface wetting for 1 h. A polarizing microscope was employed to record the entire healing process. As shown in Fig. 8a , the IP-DAP 50 -BA 0 -WPU film retained a deep, wide scratch. With an increase in the amount of boric acid, which replaced pyridine in the hard segments, the scratches on its surface became lighter and its color became weaker. It was clear that the IP-DAP 40 -BA 10 -WPU film exhibited the most desirable thermal self-healing capacity, where the contact surface exhibited no obvious fracture trace. When the cross-linking density was too large, the surface scratches became deeper and the healing effect worsened. In contrast, the scratch in the IP-DAP 25 -BA 25 -WPU film was visible and barely healed, which displayed poor self-healing ability. 3.8 Self-healing mechanism The desirable mechanical strength and self-healable capacity of the IP-DAP x -BA y -WPU film should be attributed to the multiple dynamic networks, as shown in Fig. 9 , which involved three categories of dynamic non-covalent bonds (hydrogen bonds of urethane/urea and hydrogen-bonded pyridine) and one reversible covalent bond (boronic ester bond). The primary amine group of 2,6-diaminopyridine produced a urea group after chain extension, resulting in dihydrogen bonding, where its cohesive energy was higher than that of the urethane group, which was conducive to enhancing the mechanical performance of the elastomer. However, the pure-IP-DAP 50 -BA 0 -WPU film exhibited low self-healable capacity, which just relies on dynamic hydrogen bonds. Introducing new active cross-linking agents and adjusting the ratio n (DAP/BA) of hard segments could effectively promote the mechanical properties and not compromise the self-healing ability. The IP-DAP 40 -BA 10 -WPU film had strength of 23.88 ± 0.69 MPa, breaking strain of 857.98 ± 26.88% and toughness of 73.91 ± 1.29 MJ m −3 , which showed a good level compared to other studies on waterborne polyurethane. Predominantly, driven by both water and heat, it showed excellent self-healing efficiency of above 90.0%. When water was dripped on the damaged surface, the boronic esters were hydrolyzed into boronic acids. Meanwhile, the hydrogen bonds of urethane and hydrogen-bonded pyridine were also partially broken. 12,18,55 The dissociation of the boronic ester bonds and hydrogen bonds enhanced the chain mobility dramatically. When the two sides of the fractured sample were contacted, the cross-linking structure at the sample section was undocked under the joint action of heating and water, and the molecular chains at both sides of the fracture surface could move, diffuse and tangle freely, gradually completing the section repair. Fig. 9 Schematic diagram of self-healable mechanism of IP-DAP x -BA y -WPU film."
} | 8,061 |
29382768 | PMC5816194 | pmc | 1,837 | {
"abstract": "Significance The origin of mitochondria is a challenging and intensely debated issue. Mitochondria are ancestrally present in eukaryotes, and their endosymbiotic inclusion was an extremely important step during the transition from prokaryotes to eukaryotes. However, because of the unknown order of eukaryotic inventions (e.g., cytoskeleton, phagocytosis, and endomembranes), it is unknown whether they led to or followed the acquisition of mitochondria. According to the farming hypothesis, the mitochondrial ancestor was captured by a phagocytotic host, but the advantage was not direct metabolic help provided by the symbiont; rather, it was provisioning captured prey to farmers in poor times, like humans farm pigs. Our analytical and computational models prove that farming could lead to stable endosymbiosis without any further benefit assumed between partners.",
"discussion": "Discussion We have found in both the analytical and the computational models that no explicit benefit is required from the partners for a stable integration to evolve, provided parties receive implicit benefit (food for the host and shelter for the symbiont in poor times). Farming is a form of bet hedging: the host applies different strategies in good and hard times to minimize its overall risk of extinction. In consequence, relative fitness becomes higher in poor environments and overall temporal variance of fitness is reduced in expense of reduced fitness in rich environments ( 26 ). Although no examples of bet hedging are known in Archaea, it is prevalent and well documented in Bacteria and Eukarya. Because there is no need to assume any further metabolic interaction, our bet-hedging strategy can explain stable integration of endosymbiont with host without preexisting metabolic coupling. Given that in both of our models, there is a wide range of parameters where farmers can spread and dominate the population, we claim that ours is a general result that could explain many cases of evolution towards stable endosymbiosis. Whereas neither phagocytotic archaea nor endosymbionts in archaea are known ( 30 ) [the association of haloarchaean genes with a hypothetical early bacterial endosymbiont ( 31 ) seems to be an artifact ( 32 )], recall that no intermediate of any stage toward eukaryotes is known (neither mitochondriate prokaryotes nor primarily amitochondriate eukaryotes are known to exist). It is obvious, however, that some must have existed. An early appearance of phagocytosis in Archaea is increasingly, albeit as yet inconclusively, supported by finding the necessary components ( 14 , 20 – 24 ). Assumption of phagocytosis implies that the farming strategy can be applied to the establishment of mitochondria. What makes our models specific to mitochondrial origins are ( i ) the complete lack of any preexisting metabolic interaction or preadaptation (that could provide any explicit benefit) that are certainly there in any modern eukaryote harboring endosymbionts; ( ii ) the lack of synchronized cell cycles for host and symbiont; ( iii ) phagocytosis; and ( iv ) farming and delayed digestion of the farm. Consequently, the lack of relevant examples makes it very hard to estimate parameter values for an association that occurred ∼1–2 Ga ago ( 33 ). If the host was increased in size (required for effective phagocytosis), it is not unreasonable to assume an internal endosymbiotic population size in the order of 10 2 –10 3 [e.g., in ciliates ( 34 ); modern amoeba might contain as many as 10 6 ( 35 )]. Some Korarchaea and Thaumarchaea (in fact in symbiosis with epibiotic bacteria) can reach sizes of 10–100 μm ( 36 ). For a detailed explanation on our parameter choices, see SI Appendix , Parameterization . An analogy of our proposed mechanism lies in the farming behavior of Dictyostelium discoidum . Some clones of this slime mold can establish a symbiosis with members of the Burkholderia bacterial genus. By incorporating the bacteria in the fruiting body instead of eating them all (prudent predation), the amoebae can ensure that the germinating spores find themselves equipped with edible food in a nutrient-poor medium ( 28 ). The bacteria can confer the farming behavior on the amoebae ( 29 ). Remarkably, another bacterium, Pseudomonas fluorescens is also associated with this phenomenon with two different strains: one is inedible but produces useful small molecules to the farmers (and possibly harmful to nonfarmers) and the other one is edible. The difference between the two strains is caused by a mutation in a regulatory gene ( 37 ). A recent experimental paper ( 27 ) also shows that full exploitation is a feasible route to the establishment of endosymbiosis (among eukaryotic partners). In the Paramecium – Chlorella symbiosis, the algal partner never gains net benefit from endosymbiosis, whereas the ciliate host benefits from engulfed algae under high light irradiance, and suffers from a cost in dim light. This seems to be the closest analogy to the interim evolutionary phase considered here. Taken together, these farming phenomena support our scenario for mitochondrial origins, with the caveat that living examples always involve highly evolved eukaryotic hosts. They also prompt us to hypothesize that some extra benefits (not considered in our model, such as resistance against other bacteria) may have provided advantage to host cells even when eating of ingested prey would have been still inefficient. A serious problem of farming is divisional dilution: even in good times (when the farm is not supposed to be culled), the actual farm size will reduce in successive divisions, unless something counters it. (This is even more pronounced in case of nonsynchronous host and symbiont cell cycles.) Storing more prey does not help; it reduces the relative reproduction rate of farmers compared with nonfarmers because farmers can eat less in unit time (see above). The only factor that can counter divisional dilution at the start of the partnership is autonomous growth of the farm ( SI Appendix , Fig. S11 ). Accordingly, the farm’s ability to grow inside the host must have been paramount in countering occasional culling and halving at every division (in the minimal model, this is implicitly assumed). Furthermore, farmed bacteria directly depend on the external resource (i.e., the environment), so in poor periods they can only grow very slowly (or not at all). Therefore, the farm will not last indefinitely in poor times because the host will ultimately consume faster than it can grow. If the poor period is any longer than the provisions, it means a death sentence for farmers (even if nonfarmers have already been starved to death). Thus, farmers must balance between building up a farm, paying costs, and competing with nonfarmers in good periods, and rationing their farms in poor periods such that in the long run they outperform nonfarmers. To model a worst-case scenario, we deliberately implied costs on everything the farmer does to prevent any trivial advantage over nonfarmers, so as not to beg the question. Farming has an explicit cost, dependent on farm size, that the host must pay in the form of reduced growth. Furthermore, we also added an implicit cost of farming: farmers cannot store and cull in the same time step (modeling increased handling time). Because any food stored is not consumed right away, it means that farming equals giving up eating. This ensures that farming is not a zero-sum process and has a disadvantage in good periods: farmers grow less in unit time compared with nonfarmers (even if farming has zero explicit cost). Consequently, farmers must have superior reproductive rates compared with nonfarmers in poor environments, otherwise they will go extinct or cannot invade. This is achieved by delaying death in poor periods (or even being able to reproduce in the individual-based model) when nonfarmers simply starve. Thus, according to our models, hosts do not receive free lunch; nevertheless, protomitochondria are able to stably integrate. In our model, captured bacteria end up enclosed in the host’s phagosome, just like in real and laboratory conditions ( 35 ) (though this is not always the case; see ref. 34 ). Wrapping symbionts in a phagosomal membrane has consequences. It might limit the symbiont metabolism if it relies on extensive transport of large molecules. However, this does not seem to be a problem if the symbiont is phototrophic (as we assume). Many modern examples attest to this, most prominently the various plastids, which sometimes retain extra membranes. Although losing the phagosomal membrane could have been a late invention (followed by the integration of ANT), the reproduction of symbionts within a phagosome provides a natural solution as to how they could have ended up in the host’s cytoplasm: the phagosome simply burst due to overpopulation. No phagosomal remnants can be found around modern mitochondria, whose outer membrane is partly of bacterial origin. The most important consequence of the phagosomal membrane is, however, that the symbiont could only reproduce clonally. Unless the host was sexual (discussed later), the symbiont is also exclusive to the host’s lineage. As a result, the symbiont genome becomes closely linked with the host’s genome, even before nuclear transfer of any genes. In the individual-based model, we assume asexual hosts. We also associate the farming ability with the host’s genome (instead of the symbionts), as is the case with Dictyostelium , where carrying a farm is a clone-specific trait ( 28 ). Consequently, all evolutionary traits presently associated with the asexual host could equally be associated with its clonal symbiont, i.e., the farm allocation rate of the host could in fact be the digestion-evasion rate of the symbiont. The evolved trait of culling can thus be interpreted as the ability to slowly overcome this evasive strategy. If, however, bacteria can evade host’s digestion, nonfarmers lacking the culling ability can end up with internal bacteria that only imply costs and do not provide any benefit. This would be a parasitic scenario. Although the above argument holds, ancient Archaea might have practiced sex and fused to share genes and farms. Eury- and Crenarchaeota are known to undergo fusion and fission ( 38 , 39 ). We presented a simple game theoretical model that nevertheless captures the essence of the situation. Assuming that fusion is triggered by starvation, we find that farmer–farmer interactions are less critical because nonfarmers can practically “steal” part of the stock when fusing (and splitting) with farmers. We show that nonfarmers can never build up stock larger than farmers if diffusion is responsible for exchange. In other words, farmers in the poor period always have more stock than nonfarmers, which maintains their selective advantage (in terms of survival) in poor times. There are other important considerations that favor farmers in the long-term. Repeated fusion–fission is costly (draining stocks), and leads to selective death of those running out of their reserves. Remorseless decline in population density entails an Allee effect that favors farmers: as densities drop (there is no reproduction in the poor period), mating probability (also of farmers and nonfarmers) decreases hyperbolically. Ultimately, internal stock levels will decide who survives the poor period and in what density. That is, farmers must maintain a farm large enough to survive the poor period with an end period density that prohibits nonfarmers to outgrow them in the following good period. Furthermore, if members of the farm can occasionally escape the host (not modeled), farmers can (re)colonize habitats where prey is missing (like Dictyostelium does) ( 28 ). Consequently, farm escape, or other means that facilitate farm-sharing, renders the prey as a public good in poor times (and farming/seeding ability a useful asset in the population), which minimizes risk for all population members, thus stabilizing the overall population. We have not modeled this, but the effect was proven in a eukaryotic example ( 40 ). Because rare escape of farmed bacteria is enough to reseed habitats and has only a marginal cost, we believe that this beneficial mechanism could have been additional to provisioning. There also seems to be a natural way to lose the farm-allocation ability—something we indeed cannot see in established mitochondrial dynamics. If the farm autonomously grows within the host, allocation becomes a neutral trait ( SI Appendix , Fig. S9 ). Our results specifically show that there is no real need for an allocation mechanism at all from the host’s side: any process that serves to delay digestion up to the point where the symbiont can reproduce is sufficient. Furthermore, we found that when symbionts provide more benefit than what they inflict as cost on their host, and the host receives more benefit than what it pays to feed its farm, then maintaining the farm becomes more profitable than eating it ( SI Appendix , Fig. S12 ). Consequently, culling is abandoned when the farm can provide more explicit benefit [e.g., photosynthate ( 41 ), metabolite ( 42 , 43 ), or ATP ( 11 , 44 )] than what the host would receive by eating it. The route to synchronized cell cycles opens up. The farming hypothesis of Maynard Smith and Szathmáry ( 7 , 8 ) is a plausible scenario for the origin of mitochondria. Our models provide strong support for the farming hypothesis and, consequently, for the origin of mitochondria right after phagocytosis and before any metabolic coupling, especially before the invention of ANT. Our work explicitly tests a mitochondrial origin hypothesis in a dynamical model, and is intended to bridge the gap between telltale evolutionary scenarios and ecological assumptions within the origin of eukaryotes. Our theory of mitochondrial origin, as any other, has to explain the same questions raised previously ( 3 ). Although our scenario does not explain all of the issues of eukaryogenesis (neither of the hypotheses do) ( 3 ), it provides a plausible explanation for the early endosymbiotic relationship between partners (not just in the mitochondrial context) and the emergence of relevant evolutionary innovations (farming, prudent predation)."
} | 3,584 |
31141846 | PMC6559207 | pmc | 1,838 | {
"abstract": "Summary Biomass recalcitrance is still a main challenge for the production of biofuels and high‐value products. Here, an alternative Miscanthus pretreatment method by using lignin‐degrading bacteria was developed. Six efficient Miscanthus ‐degrading bacteria were first cultured to produce laccase by using 0.5% Miscanthus biomass as carbon source. After 1–5 days of incubation, the maximum laccase activities induced by Miscanthus in the six strains were ranged from 103 to 8091 U l −1 . Then, the crude enzymes were directly diluted by equal volumes of citrate buffer and added Miscanthus biomass to a solid concentration at 4% (w/v). The results showed that all bacterial pretreatments significantly decreased the lignin content, especially in the presence of two laccase mediators ( ABTS and HBT ). The lignin removal directly correlated with increases in total sugar and glucose yields after enzymatic hydrolysis. When ABTS was used as a mediator, the best lignin‐degrading bacteria ( Pseudomonas sp. AS 1) can remove up to 50.1% lignin of Miscanthus by obtaining 2.2‐fold glucose yield, compared with that of untreated biomass. Therefore, this study provided an effective Miscanthus pretreatment method by using lignin‐degrading bacteria, which may be potentially used in improving enzymatic hydrolysability of biomass.",
"conclusion": "Conclusions This study showed that the enzymatic digestibility of Miscanthus can be improved by laccase‐secreted bacteria. Six efficient Miscanthus ‐degrading bacteria secreted abundant laccase when Miscanthus was used as sole carbon source. The crude enzymes induced by Miscanthus markedly decreased the content of lignin, especially in the presence of two laccase mediators, with the lignin removal percentage of 29.7–59.5%. After enzymatic hydrolysis, the glucose released from bacterial pretreated Miscanthus was 1.3–2.2‐fold higher than that from untreated biomass. These results will provide a two‐step pretreatment method to further studies on the use of laccase‐producing bacteria for bioenergy production.",
"introduction": "Introduction Various sources of lignocellulosic feedstocks have been regarded as emerging biofuels to partly replace the fossil fuels in the near future. However, their complicated cell wall structure makes the production of biofuels more difficult and costly due to the existence of biomass recalcitrance (Pauly and Keegstra, 2008 ). Among the three main polymeric components (cellulose, hemicellulose and lignin) constituting lignocellulosic biomass, lignin, which covers the cellulose and hemicellulose in various ways, is the most stable and complex polymer and acts as an obstacle in the enzymatic hydrolysis of cellulose (Mussatto et al ., 2008 ; Lee et al ., 2014 ). To make more cellulose and hemicellulose available before enzymatic hydrolysis, a pretreatment step is necessary to weaken and remove lignin (Lee et al ., 2014 ). The lignin can be disrupted via physical, chemical and biological processes (Alvira et al ., 2010 ). Various pretreatments methods through physical and chemical processes (Hu and Wen, 2008 ), alkaline and dilute acid (Si et al ., 2015 ), organic solvent (Sun et al ., 2000 ), steam explosion (Öhgren et al ., 2007 ) and wet oxidation (Hendriks and Zeeman, 2009 ) are available to remove lignin from lignocellulosic biomass. However, high energy input, high costs of chemicals and increased environmental risk make these methods challenging to scale up (Millati et al ., 2011 ). In addition, the lignin degraded by physical and chemical processes will release some inhibiting compounds, such as furfurals, 5‐hydroxymethyl furfurals and other volatile products, which inhibit the next stages of enzymatic hydrolysis and yeast fermentation (Hendriks and Zeeman, 2009 ; Camesasca et al ., 2015 ). Organosolv pretreatment may cause severe damage like fire explosions in the absence of proper safety measures, although it can achieve a biomass fractionation to lignin, hemicellulosic sugars and a relatively pure cellulose fraction (Kumar and Sharma, 2017 ; Li et al ., 2017 ). These are the disadvantages for large‐scale application. Lignin removal via biological pretreatment has received increasing attention as an alternative to physical and chemical pretreatment for enhancing cellulose digestibility due to its environmental and economical benefits (Wan and Li, 2012 ; Sindhu et al ., 2016 ). Among the various biological pretreatments adopted for lignin removal from lignocellulosic biomass, fungal pretreatments have been widely studied because fungi, especially white‐rot fungi, were thought to be excellent laccase and peroxidase producers (Wesenberg et al ., 2003 ; Wan and Li, 2012 ). However, fungi usually grow slowly and require a long time to produce lignolytic enzymes like laccase, manganese peroxidase (MnP) and lignin peroxidase (LiP), which means that long pretreatment time is necessary for lignin removal by fungi (Leonowicz et al ., 1999 ; Wan and Li, 2012 ). Moreover, large loss of cellulose and hemicellulose is another major weakness of fungal pretreatment, although remarkable enhancement of cellulose digestibility has been obtained from lignocellulosic biomass pretreated by white‐rot fungi (Wu et al ., 2005 ; Lee et al ., 2007 ). A number of bacteria are able to produce various lignocellulolytic enzymes like endoglucanase (CMCase), xylanase and laccase (Bugg et al ., 2011 ; Guo et al ., 2017a ). Of these bacteria, Bacillus sp., Pseudomonas sp., Streptomyces sp. and Aeromonas sp. strains have been reported to break down lignin (Bugg et al ., 2011 ; Chang et al ., 2014 ). Bacillus sp. CS‐1 isolated from forest soils was used to degrade up to 20% lignin in rice straw with remarkable laccase production (Chang et al ., 2014 ). The laccase from Bacillus licheniformis showed strong oxidation capacity towards substrates 2,2‐azino‐bis (3‐ethylbenzothiazoline‐6‐sulphonic acid) (ABTS), syringaldazine, 2,6‐dimethoxyphenol and phenolic acids (Koschorreck et al ., 2008 ). A soil bacterium, Pseudomonas putida , known as aromatic degrader, was capable of removing lignin from various lignocellulose biomasses accompanied by producing monocyclic phenolic products (Ahmad et al ., 2010 ). However, the laccase production yields of bacteria are much lower than those of fungi and thus impeding their application in industry. In our previous study, it was found that the laccase production of bacteria can be significantly induced by lignocellulosic biomass, such as algae and Miscanthus . The high activity of laccase induced by Miscanthus may be due to the high components of total ester‐bound phenolics, which can stimulate the secretion of laccase (Guo et al ., 2017b ). In addition, the delignification efficiency of laccase can be largely improved by the laccase mediators, which can increase the oxidation of polymeric lignin by transferring hydrogen atoms (Baiocco et al ., 2003 ; Yao and Ji, 2014 ). Therefore, a two‐step pretreatment procedure was developed in this study. A low concentration Miscanthus biomass (0.5%, w/v) was first used to induce the production of laccase, and then, the crude enzymes of these bacteria induced by 0.5% Miscanthus biomass were directly mixed with equal volumes of citrate buffer and Miscanthus at solid concentration of 4% (w/v) to debilitate the lignin. The effects of two laccase mediators, 2,2‐azino‐bis(3‐ethylbenzothiazoline‐6‐sulphonic acid) (ABTS) and 1‐hydroxybenzotriazole (HBT), on the degradation of lignin were also determined in the process of pretreatment. Then, the pretreatment effect was evaluated by measuring the sugar released after enzymatic hydrolysis.",
"discussion": "Discussion Miscanthus biomass is a good candidate for laccase production Twelve biomass‐degrading bacteria (seven Bacillus , two Pseudomonas , one Exiguobacterium , one Aeromonas and one Raoultella species) were isolated on plates using carboxymethyl cellulose as the sole carbon source by our previously laboratory members (Bugg et al ., 2011 ; Maki et al ., 2012 ; Paudel and Qin, 2015 ). It has been reported that the degradation ability was extremely different using different carbon sources in both fungi (Olsson et al ., 2003 ; Sipos et al ., 2010 ) and bacteria (Waghmare et al ., 2014 ; Mohammadkazemi et al ., 2015 ). In this study, the degradation abilities of the twelve bacteria to Miscanthus biomass were determined using the Gram's iodine method, while only six bacteria strains (A4, AS1, AS2B, K1, X4 and X8) showed higher hydrolytic potential than that of the industrial strain C. xylanilytica when Miscanthus was used as the sole carbon source. To further evaluate the characteristics of these six strains, laccase activities were monitored in the mineral salt medium containing 0.5% (w/v) Miscanthus . The results showed the strains A4, AS1, AS2B, K1, X4 and X8 can efficiently secrete laccase into the medium, with the significantly different among the bacteria strains. Previous studies have shown the production of bacterial lignocellulolytic enzymes differed markedly using different carbon sources even if the same strain was used (Mohammadkazemi et al ., 2015 ; Guo et al ., 2017a ). The maximum laccase activities induced by wheat bran in strains A4, AS2B, K1 and X4 were 246.7, 67.0, 82.4 and 137.0 U l −1 , respectively (Guo et al ., 2017a ), and they were 501, 1149, 586 and 378 U l −1 in the presence of Miscanthus respectively. The lignocellulolytic enzyme activity produced by Klebsiella sp. PRW‐1 was higher when grass power and sugarcane were used as carbon source compared with that of other agricultural wastes like sorghum husks, corn straw and paddy straw (Waghmare et al ., 2014 ). The biodegradation and enzyme production of lignocellulosic biomass by fungi and bacteria mainly depended on the porosity of biomass materials and crystallinity of cellulosic fibre (Zhang et al ., 2006 ; Kumar et al ., 2009 ). A series of lignocellulosic enzymes were produced by microorganisms using lignocellulosic biomass as carbon source, and certain lignocellulosic biomasses possess abundant soluble carbohydrates and enzyme synthesis‐related inducers, which can efficiently increase the production of lignocellulosic enzymes (Rosales et al ., 2005 ; Singhania et al ., 2010 ). Moreover, our results showed that in the presence of Miscanthus, the Pseudomonas strain AS1 specifically produced a large amount of laccase, which has been regarded as the main lignin‐degrading enzymes in the process of delignification of biomass (Martinez et al ., 2009 ; Pollegioni et al ., 2015 ). Delignification of Miscanthus and improved saccharification by bacterial pretreatment and laccase mediators The polymers of lignin closely overlay the other polymers (mainly including cellulose and hemicellulose) in various ways in the lignocellulosics biomass. Lignin is very impervious and hard to degrade by enzymes or chemicals, and this is the main barrier to make cellulose and hemicellulose more susceptible to enzymatic hydrolysis in industrial processes (Brebu and Vasile, 2010 ). The degradation of lignin by using fungi has been well‐studied, especially by using white‐rot and brown‐rot fungi (Leonowicz et al ., 1999 ; Guillén et al ., 2005 ). But research about the degradation of lignin by using bacteria is scarce. In this study, six Miscanthus ‐degrading bacteria were screened to degrade the lignin of Miscanthus biomass. The results showed that the crude enzymes produced by all the tested bacteria efficiently decreased the lignin content of Miscanthus , especially in the presence of two laccase mediators. The maximum lignin removal percentage (59.5%) was observed by the Pseudomonas sp. AS1 strain with the addition of 0.5% HBT (g g −1 dry biomass). Our previous study reported that the Pseudomonas sp. AS1 isolated from municipal waste showed good potential for black liquor decolourization and industrial degradation of lignocellulosic biomass (Maki et al ., 2012 ). Moreover, the other four Bacillus sp. strains and one Exiguobacterium similarly reduced the content of lignin even if only their crude enzymes were used, which obtained the lignin removal percentage range from 18.6% to 42.3%. It has been reported that some of bacterial stains were able to degrade lignin via producing laccase and other lignin peroxidase (Bugg et al ., 2011 ; Chang et al ., 2014 ). The Bacillus sp. CS‐1 strain isolated from forest soils can degrade at least 61% alkali lignin within 48 h, and pretreatment using this strain followed by lactic acid bacteria removed 61.9% lignin of rice straw and enhanced cellulase performance (Chang et al ., 2014 ). The Pseudomonas sp. LD002 and Bacillus sp. LD003 were isolated from beneath decomposing wood logs and exhibited good growth on lignin fractions and excellent dye‐decolourizing abilities (Bandounas et al ., 2011 ). The chemicals of ABTS and HBT have been regarded as the most efficient laccase mediators, which play as an electron carrier between laccase and oxidized substrate (Munk et al ., 2015 ). In the current study, the average lignin contents of Miscanthus pretreated by six bacteria strains with two laccase mediators (ABTS and HBT) were 20.2% and 30.1% lower than that without laccase mediator. This was consistent with the higher lignin removal percentage obtained by laccase pretreatment in the presence of various laccase mediators (Rico et al ., 2014 ; Rencoret et al ., 2016 ). Furthermore, all bacteria pretreatment in the citrate buffer (pH 3.0) selectively degraded the lignin and improved the cellulose and hemicellulose contents of Miscanthus . The results were similar to that of pretreatment of feedstock with commercial laccase, which removed up to 50% of lignin and significantly increased the content of glucose and xylose in the presence of laccase mediators (Rico et al ., 2014 ; Rencoret et al ., 2016 ). A previous study showed Miscanthus biomass saccharification largely depends on the effective lignin removal rather than the hemicellulose extraction under various alkali and acid pretreatments (Si et al ., 2015 ). In this study, total sugar and hexoses released after enzymatic hydrolysis were significantly improved by all pretreatments and showed a positive correlation with the lignin removal percentage, which indicated that lignin ratio was the key factor that positively influenced Miscanthus digestibility after bacterial pretreatments. In addition, the maximum enzymatic digestibility (calculated as glucose released) from bacterial pretreatment was up to 87%, which was comparable with that of most fungal pretreatments, such as Trameteshirsute yj9 (73.99%; Sun et al ., 2011 ) and Pycnoporus sp. SYBC‐13 (90%; Liu et al ., 2013 ). Therefore, the bacterial pretreatment in this study might be more promising in the removal of lignin on Miscanthus compared with that of direct incubation biomass with fungi from the following three aspects: first, the laccase secreted by most of the tested bacteria in this study can rapidly reach the maximum after 1–3 days of incubation, which was easier to obtain crude laccases than most of fungi (Wesenberg et al ., 2003 ; Liu et al ., 2013 ). Second, even though the lower redox potential was found in most bacterial laccases (Bugg et al ., 2011 ), the lignolytic enzyme activities induced by biomass in these bacteria were much higher than that in fungi. For example, the highest laccase activity of AS1 strain in this study was 1618 U g −1 dry biomass, while the maximum of it in reported fungi was 935.4 U g −1 dry biomass (Chang et al ., 2012 ; Yang et al ., 2012 ). Third, the process of lignin removal by fungal pretreatment was accompanied by large losses of cellulose and hemicellulose, which may be utilized as a carbon source to support fungal growth and metabolism (Wan and Li, 2010 ; Bugg et al ., 2011 )."
} | 4,004 |
38784799 | PMC11111985 | pmc | 1,839 | {
"abstract": "Lead (Pb) is a hazardous heavy metal that accumulates in many environments. Phytoremediation of Pb polluted soil is an environmentally friendly method, and a better understanding of mycorrhizal symbiosis under Pb stress can promote its efficiency and application. This study aims to evaluate the impact of two ectomycorrhizal fungi ( Suillus grevillei and Suillus luteus ) on the performance of Pinus tabulaeformis under Pb stress, and the biomineralization of metallic Pb in vitro . A pot experiment using substrate with 0 and 1,000 mg/kg Pb 2+ was conducted to evaluate the growth, photosynthetic pigments, oxidative damage, and Pb accumulation of P. tabulaeformis with or without ectomycorrhizal fungi. In vitro co-cultivation of ectomycorrhizal fungi and Pb shots was used to evaluate Pb biomineralization. The results showed that colonization by the two ectomycorrhizal fungi promoted plant growth, increased the content of photosynthetic pigments, reduced oxidative damage, and caused massive accumulation of Pb in plant roots. The structural characteristics of the Pb secondary minerals formed in the presence of fungi demonstrated significant differences from the minerals formed in the control plates and these minerals were identified as pyromorphite (Pb 5 (PO 4 ) 3 Cl). Ectomycorrhizal fungi promoted the performance of P. tabulaeformis under Pb stress and suggested a potential role of mycorrhizal symbiosis in Pb phytoremediation. This observation also represents the first discovery of such Pb biomineralization induced by ectomycorrhizal fungi. Ectomycorrhizal fungi induced Pb biomineralization is also relevant to the phytostabilization and new approaches in the bioremediation of polluted environments.",
"conclusion": "5 Conclusion Our results revealed a notable, positive impact of inoculating S. grevillei and S. luteus on P. tabulaeformis growth and Pb tolerance, which enhances the establishment of remediation plants under Pb-contaminated soil. Our in vitro experiments confirmed that ECM fungi participated in the bio-mineralization of Pb into pyromorphite. As such, our corroboration of pyromorphite formation represents the first discovery of such Pb biomineralization induced by ECM fungi. This observation increases our understanding of the biogeochemical cycle of Pb. ECM fungi induced Pb biomineralization is also relevant to the new approaches in the bioremediation of polluted environments. Predictably, the space between the soil particles, which is unreachable for the roots, is now filled with hyphae of ECM fungi, meaning more extensive and efficient Pb phytostabilization.",
"introduction": "1 Introduction Lead (Pb) is a hazardous heavy metal that causes severe environmental and human health problems ( Rehman et al., 2017 ; Monchanin et al., 2021 ). Pb is mainly introduced into soils by human activities such as lead acid battery production, paint production, mining, and leaded petrol production ( Jarup, 2003 ; Li et al., 2014 ). Pb pollution restricts soil usage and fertility due to the non-degradable and toxic characteristics of Pb ( Adriano, 2001 ). Phytoremediation is an effective and viable method for Pb polluted soils ( Sarwar et al., 2017 ). However, Pb harms plant development and survival due to the production of an excessive amount of reactive oxygen species (ROS) ( Reddy et al., 2005 ; Zhang et al., 2020 ). Consequently, phytoremediation has some drawbacks, including sluggish development of the accumulator plants, poor biomass production, and low heavy metal absorption, which makes it a lengthy and inefficient procedure ( Vergara et al., 2020 ). Ectomycorrhizal (ECM) symbiosis is widespread in numerous ecosystems between fungi from Basidiomycota, Ascomycota, and Zygomycota, and the ecologically and economically most important forest trees, including Pinaceae, Fagaceae, Salicaceae, Betulaceae, Caesalpinioideae, Dipterocarpaceae, and Phyllanthaceae ( Tedersoo et al., 2010 ). ECM fungi form Hartig nets inside plant roots, form sheath-like mantles around lateral roots, and form extrametrical mycelia to explore, absorb, and translocate nutrients and water from the surrounding soil. The ECM symbiosis promotes plant nutrition and water uptake, increases plant growth performance, and facilitates the establishment of host plants in harsh environments ( Arocena and Glowa, 2000 ; Baum et al., 2006 ; Szuba et al., 2017 ; Wen et al., 2017 ; Liu et al., 2020 ). Microbe-enhanced phytoremediation that using ECM is an effective measure for remediating metal-contaminated soils ( Shi et al., 2019 ; Liu et al., 2020 ). The ECM symbiosis results in enhanced host plants’ tolerance to heavy metals, including alleviation of inhibition of plant photosynthesis caused by heavy metals and a beneficial impact on reducing the metal-induced oxidative stress on plants ( Schützendübel and Polle, 2002 ; Canton et al., 2016 ; Fernández-Fuego et al., 2017 ; Mohammadhasani et al., 2017 ). In addition to improving the growth and enhancing the heavy metal tolerance of accumulator plants, ECM fungi may play a role in biomineralization. Biomineralization is the process of living organisms’ induced mineral formation. The majority of fungal-involved biomineralization is the consequence of metabolic processes that affect the external environment in a way that facilitates mineral precipitation. Examples of these processes include changes in pH, O 2 , redox potential, redox transformations of metal species, and excretion of organic and inorganic metabolites like CO 2 , H + , or organic acids ( Gadd, 2021 ). Fomina et al. (2007) showed that Beauveria caledonica causes uranyl phosphate minerals formation via biomineralization. Bacillus cereus 12-2, which was isolated from lead-zinc mine tailings, could transform the Pb into rod-shaped Ca 2.5 Pb 7.5 (OH) 2 (PO4) 6 nanocrystal ( Chen et al., 2016 ). Povedano-Priego et al. (2016) described lead phosphate formation via biomineralization in the interaction of Penicillium chrysogenum with metallic Pb. Phanerochaete chrysoporium participates in Pb biomineralization and transforms Pb into Pb 5 (PO 4 ) 3 OH through fungal phosphatase ( Zhao et al., 2020 ). Pinus tabulaeformis Carr. is one of the most widely distributed pines in northern China ( Chen et al., 2008 ) and it has remarkable drought endurance and great adaptation to poor soil ( Wang and Guo, 2010 ). The well-known ectomycorrhizal species P. tabuliformis , characterized by its strong mycorrhizal dependency and potential colonization by various fungal species ( Allen, 1991 ), can be effectively used for restoration in post-mining areas ( Yan et al., 2020 ; Zhang et al., 2023 ). It was found that the dominant ECM fungi in the rhizosphere soil of P. tabuliformis , which belongs to the genus Suillus ( Wang et al., 2022 ), serves as a model system for understanding mycorrhizal fungal metal tolerance ( Branco et al., 2022 ). Although there were studies showing that pines with ectomycorrhizal fungi promoted tolerance against heavy metals ( Bizo et al., 2017 ; Liu et al., 2020 ; Ouatiki et al., 2022 ), the response of the symbiosis between P. tabulaeformis and ectomycorrhizal fungi to Pb pollution was rarely reported. In this study, we used two ectomycorrhizal fungi ( Suillus luteus and Suillus grevillei ) to evaluate their ability to (1) form symbiosis with P. tabuleaformis , (2) improve plant growth, (3) promote the activity of antioxidant enzymes and photosynthetic pigment content, (4) regulate Pb uptake and distribution, and (5) biomineralize Pb in vitro .",
"discussion": "4 Discussion Mycorrhizal plants acquire water and mineral nutrients via fungi hyphae, while supply organic nutrients to fungi ( Brundrett, 2002 ; Kaiser et al., 2010 ; Pena and Polle, 2014 ). In this study, we observed an increase in the biomass of P. tabulaeformis inoculated with S. grevillea under both Pb-treated and Pb-free treatment. This improvement caused by S. grevillea can be explained by greater mineral absorption ( Canton et al., 2016 ; Sun et al., 2021 ) and Pb tolerance ( Baum et al., 2006 ). However, the effect of S. luteus on the growth of P. tabulaeformis was not obvious. Compared with ECM fungi sensitive to heavy metals, the tolerant strains are reported to be more effective in enhancing the metal tolerance and growth of host plants ( Adriaensen et al., 2006 ; Szuba et al., 2017 ). Thus, we speculated that S. luteus exhibit lower Pb tolerance compared with S. grevillea , which can be corroborated by the fact that the S. luteus colonization rate significantly decreased under Pb Stress ( Table 1 ). Meanwhile, slower growth of S. luteus than S. grevillea was observed on MMN agar medium with lead shots ( Figure 3 ). The Pb stress response in both ECM fungi treatments showed improvement in P. tabulaeformis , with a decrease in lipid peroxidation (shown by MDA level) and reactive oxygen species (shown by H 2 O 2 concentration) and a significant improvement in antioxidant defense (shown by SOD, POD, and CAT activity). Our findings support the early studies that discovered the promotion of ECM symbiosis through the rise in antioxidant activities in their host plants against Cb and Al stress ( Liu et al., 2020 ; Sun et al., 2021 ). In this work, S. grevillea inoculation resulted in a considerable rise for carotenes or chlorophylls but not S. luteus , which could be related to the different properties of fungi, as mentioned above. This result is supported by the findings of Szuba et al. (2017) , with more chlorophylls and carotenes increments of Populus × canescens colonized by the Pb-tolerant Paxillus involutus . Our results supported the idea that inoculating host plants with ECM fungus considerably increases their Pb tolerance and shields them from heavy metal stress ( Szuba et al., 2017 ). Pb is a weakly translocated metal. Up to 90% of the Pb absorbed by plants is accumulated in their roots ( Kumar et al., 1995 ; Fahr et al., 2013 ). The accumulation of Pb in plant roots has been found to be higher than in shoots, particularly in mycorrhizal seedlings where Pb accumulation in roots was significantly elevated compared to non-mycorrhizal seedlings ( Table 2 ). This finding is consistent with previous studies ( Bojarczuk et al., 2015 ; Gu et al., 2017 ; Hachani et al., 2020 ; Liu et al., 2020 ) which have suggested that excess Pb is sequestered through biomineralization in fungus hyphae, often along with colonized roots for analysis. ECM fungi play a crucial role in lead homeostasis within ecosystems by accumulating a substantial amount of lead through their mycelium ( Liu et al., 2020 ; Branco et al., 2022 ). These mechanisms involve the storage of excessive metals in compartments within cells ( González-Guerrero et al., 2008 ; Ruytinx et al., 2013 ), and the capture of these metals by proteins or metabolites in ECM fungi ( González-Guerrero et al., 2007 ; Leonhardt et al., 2014 ). Additionally, excess metals can be exclusion by ECM fungi ( Ruytinx et al., 2013 ; Majorel et al., 2014 ). Various metal transporters and chelating proteins in mycorrhizal fungi have been functionally studied, providing insight into their role in reducing the harmful effects of excessive environmental metals ( Leonhardt et al., 2014 ; Coninx et al., 2017 ; Ruytinx et al., 2017 ; Gómez-Gallego et al., 2019 ). To investigate potential interactions between metallic Pb and the ECM fungi, two ECM fungi were incubated with a Pb shot. It was clear from the structural variations between the Pb secondary minerals generated in the presence of the fungus and those formed in the control plates that the fungi had a role in their formation. Subsequently, the particular mineral was identified as pyromorphite (Pb 5 (PO4) 3 Cl). Under general geochemical conditions, pyromorphite is the most stable environmental Pb compound (Ksp = 10 –79.6 ± 0.15 ) ( Topolska et al., 2016 ) and pyromorphite deposition has consequently been proposed as a remediation treatment for Pb-contained soil ( Miretzky and Fernandez-Cirelli, 2008 ). In the natural environment, pyromorphite generation depends on various chemical and biological processes ( Rhee et al., 2012 ). A previous study has demonstrated that Paecilomyces javanicus participates in the biomineralization of Pb metal and transforms Pb into pyromorphite and the authors speculated that this phenomenon is linked with microbial transformations of inorganic and organic phosphorus by extracellular acid phosphatases secreted by fungal ( Rhee et al., 2012 ; Liang et al., 2015 ). This process produced the main ligands (PO 4 3− ) of Pb, which may contribute to biologically induced pyromorphite formation. The variation of pH could change the dissolution of the Pb and P sources and oxidation–reduction potential (Eh), which affect the efficacy of pyromorphite formation ( Chen et al., 2003 ; Ehrlich and Newman, 2009 ). Fungi can produce organic acids, which can mobilize metal ions while changing pH ( Adeyemi and Gadd, 2005 ). Similarly, we observed less Pb precipitation in the control medium, which may be attributed to the mobilization of PbO, Pb 2 O 3 , and Pb 3 (CO 3 ) 2 (OH) 2 by organic acids secreted by fungal hyphae. In vivo and in vitro growth conditions may alter the nature of the ligands provided by fungal hyphae. However, here we are lacking information on ECM fungi’ effects on Pb biomineralization in the symbiotic state, but Fomina et al. (2005) reported that ECM fungus provides the same ligands of functional groups (e.g., phosphate, carboxylate, sulfhydryl) in both pure culture and ectomycorrhizal association. Thus, it is likely that ECM fungi hyphae in a symbiotic relationship would also provide functional ligands for Pb precipitation."
} | 3,444 |
35381088 | PMC9338886 | pmc | 1,840 | {
"abstract": "Abstract Microbial biofilms are ubiquitous. In marine and freshwater ecosystems, microbe–mineral interactions sustain biogeochemical cycles, while biofilms found on plants and animals can range from pathogens to commensals. Moreover, biofouling and biocorrosion represent significant challenges to industry. Bioprocessing is an opportunity to take advantage of biofilms and harness their utility as a chassis for biocommodity production. Electrochemical bioreactors have numerous potential applications, including wastewater treatment and commodity production. The literature examining these applications has demonstrated that the cell–surface interface is vital to facilitating these processes. Therefore, it is necessary to understand the state of knowledge regarding biofilms’ role in bioprocessing. This mini-review discusses bacterial biofilm formation, cell–surface redox interactions, and the role of microbial electron transfer in bioprocesses. It also highlights some current goals and challenges with respect to microbe-mediated bioprocessing and future perspectives.",
"conclusion": "Conclusion Numerous efforts to use microbes as the centerpiece of bioprocessing technologies have revealed the importance of biofilms. Among other characteristics, biofilms facilitate redox reactions and are resilient against otherwise harmful products. Our understanding of the mechanisms underlying these traits in the context of bioprocessing is ongoing. As described in this review, there is a need for improvements with respect to the biology (e.g. bioprospecting and genetic engineering) and bioreactor design (e.g. electrode modification) (Fig. 1 ). As we turn toward transitioning these applications from the lab into industry, it is essential to address the barriers described here and elsewhere (Conners & Bose, 2021 ; Conners et al., 2022 ; Jourdin & Burdyny, 2021 ; Karthikeyan et al., 2019 ; Lee et al., 2021 ). Fig. 1 Relevant areas of consideration for optimizing electroactive biofilms in bioelectrochemical applications. Efficient design and operation of bioelectrochemical systems rest on fine-tuning bioelectrochemical cell parameters using tools such as mathematical modeling to allow for scalability; investigation of microbe–electrode interactions and the effect of electrode type, electrode modifications, microbial composition, and biofilm formation on performance; and understanding the cellular processes underlying electron exchange and product formation. Future efforts to improve bioelectrochemical cell performance should focus on improving Coulombic efficiency, optimizing biofilm formation, enhancing bioproduct formation, and bioprospecting for novel electroactive strains and electron exchange mechanisms."
} | 680 |
33462162 | PMC8506656 | pmc | 1,841 | {
"abstract": "Bacterial biofilms are communities of bacteria that exist as aggregates that can adhere to surfaces or be free-standing. This complex, social mode of cellular organization is fundamental to the physiology of microbes and often exhibits surprising behavior. Bacterial biofilms are more than the sum of their parts: single-cell behavior has a complex relation to collective community behavior, in a manner perhaps cognate to the complex relation between atomic physics and condensed matter physics. Biofilm microbiology is a relatively young field by biology standards, but it has already attracted intense attention from physicists. Sometimes, this attention takes the form of seeing biofilms as inspiration for new physics. In this roadmap, we highlight the work of those who have taken the opposite strategy: we highlight the work of physicists and physical scientists who use physics to engage fundamental concepts in bacterial biofilm microbiology, including adhesion, sensing, motility, signaling, memory, energy flow, community formation and cooperativity. These contributions are juxtaposed with microbiologists who have made recent important discoveries on bacterial biofilms using state-of-the-art physical methods. The contributions to this roadmap exemplify how well physics and biology can be combined to achieve a new synthesis, rather than just a division of labor.",
"introduction": "1. Introduction Bacterial biofilms are integrated communities of cells that adhere to surfaces and are fundamental to the ecology and biology of bacteria. Bacterial biofilm communities can be harmful, such as those that contribute to lethal airway infections in cystic fibrosis. However, bacterial communities can also be beneficial, and help train your immune system or digest your vegetables, as well as break down hydrocarbons in oil spills. Recent collaborative work between physicists and microbiologists has shown that bacteria employ surprisingly sophisticated physics and chemistry in order to organize these biofilm communities on a surface. How does one get started in this multidisciplinary field? One of the most common questions from incoming graduate students is whether they have to master biology before doing biophysics. The answer is not a simple one. Adapting an idea from Karl Kraus may begin to answer this question: instead of being someone who masters a language, an artist is rather a servant of the word. Besides depth of inquiry, what unites the contributors in this multidisciplinary roadmap is a cognate sense of service to the field of bacterial biofilm microbiology. Rather than using microbiology as a mere context for new physics, each contributor from physics in this roadmap is interested in microbiology itself, and uses different aspects of physics to discover new microbiology. Their contributions are juxtaposed with those of well-known microbiologists who have made recent important discoveries on bacterial biofilms using state-of-the-art physical methods. Using these organizing principles for this roadmap, we hope it can live up to the onomastic promise of physical biology. Bacteria have developed various strategies to move, sense, and organize in low Reynolds number environments; these often involve bacterial motility appendages such as flagella. Antani and Lele review the role of the flagellum in motility and mechanosensing: obstructions in the rotation of the flagellar motor will drive recruitment of additional stator units to the motor to increase torque. Kühn and Persat review the mechanics and dynamics of type IV pili (TFP), which are extension–retraction appendages often compared to grappling hooks. In particular, they examine how TFP are coordinated by considering them from the perspective of non-equilibrium systems. Chen and Nan review ‘gliding’ motility, where bacteria do not use appendage technology at all for motility, and employ force-generating complexes along helical tracks instead. Bru, Høyland-Kroghsbo and Siryaporn review how stress responses can redirect movement of bacterial populations and ultimately control bacterial spatial organization, via quorum sensing (QS) and stress signals. The roadmap also contains sections on how bacteria adapt their existence to complex environments. Conrad explores bacterial mechanisms for controlling adhesion on real, heterogeneous interfaces, both solid and liquid, including for example oil droplets, which are particularly important for mitigating oil spills. Marine microbial environments are often characterized by heterogeneous and transient nutrient fluctuations, which can lead to interesting bacterial ecologies in different environmental niches. Carrara, Yawata and Stocker describe how bacteria solve these problems by gene expression and energetic investments. The first step in the formation of a bacterial biofilm is contact with the surface on which the community will eventually form, raising the intriguing question: ‘how does a microbe know it is on a surface?’ Intracellular second messengers such as cyclic-AMP (cAMP) and cdiGMP play key roles in this process, and have emerged as a kind of master regulator of bacterial behavior. Brun reviews how TFP are used to surface sense, using labeling and visualization of pili dynamics in live cells. Lee, de Anda, Schmidt, Golestanian, O’Toole and Wong review the signal processing of surface sensing and how it is propagated from mother cell to daughter cell via a kind of multigenerational memory. cdiGMP signaling and downstream biosynthesis of the exopolysaccharide biofilm matrix are pivotal events in bacterial community development. Floyd and Yildiz review the consequences of cdiGMP signaling in Vibrio cholerae using an elegant method based on an mRNA riboswitch-based biosensor to determine changes in cdiGMP, and on visualization of pili in live cells. ‘What I cannot create, I do not understand’ was found written on Richard Feynman’s blackboard at the time of his death in 1988. In this spirit, Yang and Jin take a completely different approach to surface sensing based on synthetic biology: they show how we can reprogram bacterial surface sensing behavior using the chemical language of second messengers via optogenetic control of bacterial cdiGMP production. All bacteria have to solve their energy problems in order to survive. Electron transfer couples the oxidation of electron donors to the reduction of electron acceptors, and constitutes the basis of bacterial respiration. However, bacteria are not limited to electron donors (such as organic molecules in growth media) or electron acceptors (such as oxygen) that exist in solution. They can solve their ‘life or death’ electron transfer problems by coupling directly to a solid surface via extracellular electron transfer (EET), a process that allows metal-reducing and oxidizing bacteria to catalyze generation of electricity and waste degradation. There has been great recent progress in EET, specifically in understanding bacterial nanowires, which were previously thought to be composed of protein-based pilin units: the situation is considerably more complex and diverse. Zacharoff and El-Naggar show that in Shewanella , bacterial nanowires take the form of membrane extensions studded with cytochromes. Yalcin and Malvankar show that in Geobacter , the nanowires that provide a continuous path for electron flow are polymerized six-heme cytochrome OmcS. What happens when bacterial communities become progressively more crowded? Ideas about QS have now spread well beyond microbiology. Toyofuku, Eberl and Nomura offer a new perspective. QS signals are often amphiphilic molecules. It turns out that bacteria can use membrane vesicles (MV) rather than solvated signal molecules to mediate a kind of quantized QS signaling. Yan, Stone, Wingreen and Bassler developed methods to image living biofilms with single-cell resolution, and show how V. cholerae grew from the founder cell to clusters of different morphologies to biofilms of ~10 000 cells. Using new quantitative imaging techniques, Rojas-Andrade and Hochbaum map out bacterial metabolism in communities, with heterogeneity that fluctuates in space and time. In the review from Wu and Xu, we come full circle, and examine motility, now in the form of self-organized synchronized collective motion of strongly interacting bacteria. In a forward looking review, Drescher and Dunkel examine how data science and machine learning may be used to help formulate the next generation of models for understanding key mechanisms and discovering general principles for biofilm formation. The excellent individual roadmap sections collected here will attract and reward the attention of beginners and experts alike."
} | 2,177 |
30965888 | PMC6418621 | pmc | 1,843 | {
"abstract": "Advances in flexible and multifunctional electronic devices have enabled the realization of sophisticated skin for robotics applications. In this paper, a large-scale, flexible and self-powered tactile sensing array (TSA) for sensitive robot skin is demonstrated based on the triboelectric effect. The device, with 4 × 4 sensing units, was composed of a top triboelectric polyethylene terephthalate (PET) layer, a bottom triboelectric copper (Cu) layer and a bottom PET substrate. A low-cost roll-to-roll ultraviolet embossing fabrication process was induced to pattern the large-scale top PET film with microstructures for high-output performance. The working mechanism and output performance of the triboelectric TSA were demonstrated and characterized, exhibiting good stability and high sensitivity. By integrating a tactile feedback system, the large-scale TSA, acting as intelligent skin for an industrial robot, was able to realize emergency avoidance and safety stop for various unknown obstacles under various working conditions. The system also has good real-time performance. By using a large-scale roll-to-roll fabrication method, this work pushes forward a significant step to self-powered triboelectric TSA and its potential applications in intelligent robot skin.",
"conclusion": "4. Conclusions A flexible large-scale TSA based on the triboelectric effect was designed and characterized in this paper. It is able to transform the applied force into voltage signal without an external power supply. Moreover, a low-cost roll-to-roll embossing fabrication method was utilized to realize mass surface treatment of the PET film, and to enhance the sensitivity of the TSA. The tactile sensing characteristics indicated that the microstructured patterns on PET film fabricated by large-scale roll-to-roll embossing process contributed to the excellent output performance of the TSA. We also demonstrated a tactile feedback system for the application of emergency avoidance and safety stop towards industrial robots. The flexible self-powered TSA showed high detection capability and fast response in an unconstructed environment, and would be of benefit for the future secure interactions of humans and robots.",
"introduction": "1. Introduction Tactile sensing is a key technology for intelligent robotics able to operate in unstructured environments and conduct safe interactions with humans and objects. A flexible robot skin capable of tactile sensing over a large area is expected to broaden the cognitive capability of robots, and to enhance the autonomous movement capability within an unknown environment. In comparison to tactile sensors, non-contact sensors are more susceptible to the ambient environment. Owing to harsh environments and complex industrial requirements, non-contact sensors are easily confused, and make poor decisions, which could lead to a hidden risk of accidents. Flexible tactile sensors have been widely investigated in the past few decades, and many sensing mechanisms have been reported in the literature, including capacitive [ 1 , 2 , 3 , 4 ], inductive, resistive [ 5 , 6 , 7 , 8 ], piezoelectric [ 9 , 10 , 11 ], thermoelectric [ 12 ], and other functional materials with sensing characteristics [ 13 , 14 , 15 , 16 ]. However, the complicated fabrication process and low scalability of these methods still remain critical obstacles to the mass production of large-scale robot skin. Moreover, most of these sensors rely on an external power supply to work, thus potentially making the electrical wiring of a large number of sensor arrays on existing industrial robots more complicated. In this paper, we introduce a high-performance self-powered TSA based on a triboelectric mechanism. The contact and separation process between two triboelectric layers introduces surface charge transfer, i.e., conversion of the external mechanical force into an electrical output signal without an external power source. It also has the advantages of easy fabrication, high efficiency and a wider choice of materials. To achieve high sensitivity of the TSA, the surface morphology of the triboelectric contact layers is crucial for charge generation in the process of contact electrification [ 17 , 18 , 19 ]. Therefore, domestic and foreign researchers have figured out different methods to improve the surface morphology and effectiveness, such as block copolymer self-assembly [ 20 , 21 ], soft lithography [ 22 , 23 ], nanoparticle deposition [ 24 ], chemical treatment [ 25 , 26 , 27 ] and plasma etching [ 28 , 29 , 30 ]. However, these methods are critically limited by the chamber or wafer size, and hence can only fabricate relatively small-sized samples at high cost. They are not beneficial for mass production and commercialization of flexible triboelectric TSA. We demonstrated a process flow of roll-to-roll UV embossing for patterning large-sized PET film with mass replicate microstructures. The mass surface treatment of the PET friction material could effectively enhance the triboelectrification effect of the TSA [ 31 ]. The flexible TSA as artificial electronic skin is expected to accommodate irregular surfaces, and imitate the human somatosensory system for potential applications in health monitoring, artificial prosthetics and advanced robotics [ 32 ]. Mei et al. applied their flexible tactile sensor on a prosthetic hand for measuring the grasping force [ 33 ]. They designed a typical signal-processing circuit for the application, including a microcontroller unit (MCU) and an analog digital conversion unit [ 34 , 35 , 36 ]. In this work, we successfully demonstrated a tactile feedback system targeting industrial robot, in which the signal-processing circuit worked as the interface connecting the TSA to the robot controller. The tactile feedback system has good real-time performance, and the emergency avoidance and stop functions were implemented efficiently at the minimum impacting velocity of 0.12 m/s.",
"discussion": "3. Results and Discussion 3.1. Characterization Figure 2 a shows the operating mechanism of the triboelectric TSA by the coupling of contact electrification and electrostatic induction. In the non-contact state, the patterned top PET film is separated from the bottom Cu film by a spacer, and both layers are uncharged. When an external mechanical force is applied, such as finger tapping, both the triboelectric layers are brought into full contact. According to the triboelectric series, electrons are transferred onto the patterned PET film from the Cu film because the PET tends to be more triboelectrically negative than Cu in the contact electrification process. The generated triboelectric charges with opposite polarities then approach balance, leading to there being no current in the external circuit. When the applied force is released, the patterned PET film and Cu film start to separate, leading to electric potential difference, which generates current flowing from the Cu film to the reference ground. When these two films separate to the maximum position, there will be no current through the load resistor connected between the Cu film and the ground. When the external force is applied again, the PET film begins approaching the Cu film. This causes a reverse-orientation electric current in the load resistor, until the patterned PET film and Cu films are fully in contact with each other again. This is a working cycle of electricity generation in contact-separation mode due to external force applied on the TSA. To study the effect of embossed patterns on the performance of triboelectric mechanism, both the fabricated TSA with and without embossed patterns were characterized. Figure 2 b shows the output voltages of the triboelectric TSA by finger tapping using unpatterned and patterned top PET triboelectric layers. The corresponding peak-to-peak voltages were measured to be 119 and 454 V, respectively. It can be seen that the embossed patterns of the PET film led to improved triboelectric performance due to easier charge separation, as demonstrated by Fan et al. [ 19 ]. Figure 2 c,d illustrates the variations of voltage and power transferred to different load resistors for the TSA with unpatterned and patterned microstructures, respectively. The measured peak-to-peak voltage increased continuously as the value of the connected load resistance increased. The maximum power of the TSA with patterned microstructures was obtained to be 2.82 mW at a matched load resistance of 35 MΩ, corresponding to an output power density of 0.47 mW/cm 2 , which was much higher than the values for the TSA without microstructure patterns. In the dynamic sensing process, the speed and frequency of the applied force may constantly change, resulting in variable output performance of the triboelectric TSA. The dynamic characterization was conducted by a linear motor system (Linmot E1100, NTI AG, Spreitenbach, Switzerland), consisting of a slider, a stator and a motion controller, as shown in Figure 2 e. The end effector of the linear motor is able to be accurately positioned at any point of the full-stroke at various speeds. A force sensor (LSZ-F08, SZOBTE, Suzhou city, China) was fixed to the bottom of the slide bar of the linear motor to measure the force applied to the TSA, as shown in Figure 2 g. The TSA of 4 × 4 sensing units was placed under the linear motor, and each unit was applied a fixed dynamic force with different speeds ranging from 0.02 m/s to 0.4 m/s. Figure 2 f shows the variation of the peak-to-peak voltages of the TSA under different speeds of the external force. It is fairly obvious that the output performance of the TSA correlates well with the speed of the applied probe. As the probe speed was increased to 0.02, 0.05, 0.1, 0.2 and 0.4 m/s, the corresponding peak-to-peak voltage increased to 50, 125, 189, 219 and 261 V, respectively. The output increment is caused by faster electron flow, while the charge transfer is maintained under higher dynamic probe speed [ 37 ]. In the next, the experimental setup shown in Figure 2 g was utilized to measure the output voltage of the TSA against the amplitude of the applied force. It can be observed from Figure 2 h that the output voltage increases linearly as the force applied to the sensing unit increases. The force sensitivity of the sensing unit was characterized to be 1.14 V kPa −1 . The output increment can be attributed to the increment of contact area due to larger contact force [ 38 ]. Figure 2 i demonstrates that each patterned triboelectric unit of the TSA was able to simultaneously light up 105 LEDs by finger tapping. 3.2. Tactile Feedback System The fabricated large-scale TSA acting as flexible electronic skin has wide application prospects in advanced robotics. For instance, industrial robots with multiple degrees of freedom can realize complex working procedures driven by servo system. Because industrial robots cannot recognize the unknown or unexpected obstacles, there is a huge security risk in the production line. Although there are strict safety procedures and protective fences on site, robot injury accidents occur frequently due to erroneous operation or unexpected intrusion of people or objects. In order to avoid such accidents, a tactile feedback system integrating flexible TSA on a robotic arm was developed for use in an obstacle-avoidance safety application. When an unexpected intrusion of people or objects runs into an arbitrary sensing unit of the TSA, the working robot should be able to detect the tactile force and respond quickly to avoid adverse consequences. Figure 3 illustrates the diagram of the tactile feedback system. It consists of four parts: self-powered TSA, signal acquisition module, relay control module and industrial robot. The flexible TSA of 4 × 4 units was attached on the robotic arm in order for precise self-sensing of external tactile force to be possible. The output voltage of any triboelectric sensing unit can be detected by the signal acquisition module, and an interrupt program can be triggered accordingly. The interrupt command enables the relay control module to open its switch and cut off the servo power supply, resulting in emergency stop of the industrial robot. The signal acquisition module is mainly comprised of a MCU and a peripheral circuit. The MCU is employed for data acquiring, processing, sending, and program commands processing. The relay control module is composed of an electromagnetic relay (SRD-05VDC-SL-C, Songle, Ningbo city, China) and some auxiliary electronic components. When an external force is applied to the triboelectric sensing unit, which indicates that the industrial robot has encountered an unknown obstacle during movement, the output voltage of the TSA is transmitted to the MCU, triggering the interrupt program. While executing the interrupt service program, a high-level signal is sent to the relay control module. Thereafter, the interrupt program is completed, and the main program continued until the signal-acquisition module stops running. The output terminal of the relay control module is connected to the controller of the industrial robot so as to turn the servo power supply on/off. 3.3. Application on Industrial Robot The tactile feedback system was successfully demonstrated on an industrial robot from Jiangsu Huibo Robot Co. Ltd, which was composed of base, arm, forearm, joints and electrical cables, and can realize multi-degree-of-freedom movement. The overall height of the arm and the base was 157 cm and the length of the forearm was 89 cm. The integrated tactile feedback circuit modules were encapsulated in a curved plate whose surface was covered by the TSA. The whole tactile feedback system was fixed on the forearm of the industrial robot to carry out the experiments. According to the robot safety operation specifications, industrial robots should include emergency stop control circuits. The emergency stop circuit is used to switch on and off the power supply of the servo driver. Once the power supply of the servo driver is cut off, the braking device of the servo motor starts to work, and the robot stops. The emergency stop control circuit was powered by 24 V DC power supply. The COM terminal of the relay control module was connected to the 24 V DC power supply, while the NC terminal was connected to the emergency stop circuit. When the COM terminal is changed to NO terminal, the power supply for servo driver is cut off, and the robot runs to emergency stop. Based on the prescribed running procedure, the robot forearm repeated a horizontal swing, i.e., moving forward from position A, to B, to C, and then backward from position C, to B, to A, as illustrated in Figure 4 a–c. The red arrow represents the direction of movement. Once an unexpected obstacle suddenly appears in the movement path of the robot forearm, the obstacle encounters the sensitive TSA. Consequently, the output voltage of the sensing unit triggers the interrupt program. Hence, the signal-acquisition module controls the relay to cut off the servo power supply. Thus, the robot achieves emergency stop, and avoids further adverse consequences. In Figure 4 d–f, we carried out a series of tactile feedback experiments by using different obstacles, such as a human hand, a rubber hammer and a metal plate. It was found that the TSA was very sensitive for immediately detecting various objects and feedback under unknown circumstances. ( Videos S1, S2 and S3 show the tactile feedback experiments of the industrial robot with sensitive TSA by using different obstacles, being a human hand, rubber hammer and metal plate, respectively. The implementation of the emergency stop can be observed clearly.) To analyze the real-time response of the tactile feedback system, the output signal of the sensing array and the corresponding on/off states of the power supply for the robot emergency stop were recorded when the TSA encountered different obstacles as shown in Figure 4 g–i. The pulse signal of the TSA indicated that it had encountered an obstacle. The sudden increase in voltage of the emergency stop circuit represented a sudden change in the movement state of the robot, i.e., from running to stopped. From these figures, it can be concluded that when the running robot experiences an unexpected force from different obstacles at any point in time, the tactile feedback system will respond quickly. The tactile feedback system has good real-time performance. To measure the performance of the tactile feedback system, the emergency stop experiment of the robot forearm was conducted at different running speeds, which means that the TSA approached the obstacles at different dynamic forces. After repeated testing, the TSA was seen to be sensitive and effective between a minimum speed of 0.12 m/s and a maximum speed of 2.37 m/s relative to an obstacle. The TSA as a sensitive robot skin maintains high detection capability at a running speed of 0.12 m/s, which proves the tactile feedback system using the TSA is able to realize an emergency stop function even at very low impact speeds. ( Videos S4, S5 and S6 show the tactile feedback experiments of the industrial robot with the sensitive TSA by using a human hand at different running speeds, which are low speed, medium speed and high speed, respectively. The implementation of emergency stop can be observed clearly even at extremely low running speeds.)"
} | 4,355 |
36421079 | PMC9687489 | pmc | 1,846 | {
"abstract": "The effects of the inoculum origin, temperature or operational changes on ex situ biomethanation by complex microbial communities have been investigated; however, it remains unclear how the diversity of the inoculum influences the process and its stability. We explored the effect of microbial diversity of four inocula (coded as PF, WW, S37 and Nrich) on methane production, process stability and the formation of volatile fatty acids as by-products. The highest methane amounts produced were 3.38 ± 0.37 mmol, 3.20 ± 0.07 mmol, 3.07 ± 0.27 mmol and 3.14 ± 0.06 mmol for PF, WW, S37 and Nrich, respectively. The highest acetate concentration was found in less diverse cultures (1679 mg L −1 and 1397 mg L −1 for S37 and Nrich, respectively), whereas the acetate concentrations remained below 30 mg L −1 in the more diverse cultures. The maximum concentration of propionate was observed in less diverse cultures (240 mg L −1 and 37 mg L −1 for S37 and Nrich cultures, respectively). The highly diverse cultures outperformed the medium and low diversity cultures in the long-term operation. Methanogenic communities were mainly composed of hydrogenotrophic methanogens in all cultures. Aceticlastic methanogenesis was only active in the highly diverse sludge community throughout the experiment. The more diverse the inocula, the more methane was produced and the less volatile fatty acids accumulated, which could be attributed to the high number of microbial functions working together to keep a stable and balanced process. It is concluded that the inoculum origin and its diversity are very important factors to consider when the biomethanation process is performed with complex microbial communities.",
"conclusion": "5. Conclusions This study showed that microbial diversity is an important parameter to consider when performing biomethanation with mixed cultures. We report that a complex inoculum such as sludge from a wastewater treatment plant is suitable for biomethanation of hydrogen based on the observation of stable process performance and low amounts of by-products. The potential benefits of using sludge biomass to perform biomethanation also offer the possibility to withstand idle operation periods created by the intermittent nature of renewable energy. Additionally, the availability of large amounts of microbial biocatalyst in industrial scale reactors makes anaerobic sludge from wastewater treatment plants suitable for further exploration of the advantages and pitfalls of large-scale trials. This would reduce the risk in the technological advancement of biological P2G so that the renewable energy sector could be coupled to the wastewater treatment sector. Future research may explore the potential benefit of existing processes from bioaugmentation and re-inoculation strategies to increase the microbial diversity and functional redundancy in order to improve the biomethanation process.",
"introduction": "1. Introduction Renewables such as photovoltaics and wind power may produce temporary surplus electricity that needs an energy storage solution. Power-to-gas (P2G) has gained attention to enable the storage of surplus electricity in the form of a storable gas such as hydrogen or methane [ 1 ]. Hydrogen can be used as a fuel and chemical feedstock or stored in the natural gas grid up to certain limits [ 2 ] according to country regulations. Methane, on the other hand, is more advantageous because it can be used in the same applications as hydrogen with the advantage of higher volumetric energy content [ 3 ]. Methane may be a better option because it is more compatible with the existing gas infrastructure allowing direct grid injection and the higher energy density of methane makes its transport and storage easier [ 3 ]. However, CH 4 is a more potent greenhouse gas than CO 2 ; therefore, its leakage should be avoided during the process and distribution. Moreover, if CH 4 is oxidized again to CO 2 in an industrial setup, the resulting CO 2 could be reconverted to CH 4 with renewable hydrogen once again in a circular process. Considering methane as the energy carrier of choice for the P2G concept, the first step of the process is hydrogen production through water electrolysis (Equation (1)) followed by a methanation step in which hydrogen drives the reduction of carbon dioxide to methane (Equation (2)) [ 4 ].\n (1) 4 H 2 O → 2 O 2 + 4 H 2 \n (2) 4 H 2 + C O 2 → C H 4 + 2 H 2 O The methanation step is either a catalyst-based chemical reaction (Sabatier process) or a biochemical reaction (hydrogenotrophic methanogenesis). The biochemical reaction can be performed with pure cultures [ 5 ] or mixed cultures [ 6 ], and depending on the type of biocatalyst the bioprocess can be classified as in situ, ex situ or hybrid biomethanation as described in detail by [ 7 ]. Several studies reporting the use of mixed cultures in the aforementioned bioprocesses have been summarized in review papers [ 6 , 8 , 9 ]. Injecting hydrogen to reactors containing mixed cultures often leads to the accumulation of volatile fatty acids (VFA) such as acetate, propionate, butyrate or even longer-chain or branched C4 and C5 organic acids. This has been observed during in situ [ 10 , 11 , 12 , 13 , 14 ], ex situ [ 7 , 15 , 16 , 17 , 18 ] or hybrid biomethanation [ 19 ]. Additionally, it was recently shown that formate is also produced and consumed during H 2 /CO 2 metabolism with mixed cultures [ 15 , 20 ]. The production of acetate or longer-chain organic acids is due to the enhancement of homoacetogenesis and/or chain elongation reactions. The produced acetate is converted to methane if aceticlastic methanogens are present in the mixed cultures [ 6 ]. However, acetate can be used to build microbial biomass as well. Another possibility for acetate consumption is mediated by syntrophic acetate-oxidizing bacteria (SAOB) that convert acetate into H 2 and CO 2 if the H 2 partial pressure is low enough to make the reaction thermodynamically feasible [ 21 ]. Different aspects, such as reactor configuration [ 7 ], process operation mode [ 18 , 22 , 23 , 24 ], methods to improve gas delivery [ 25 , 26 , 27 , 28 ], temperature, inoculum [ 29 ], and pH [ 30 , 31 ], affect the process. Since temperature is a deterministic factor influencing the microbial community structure in anaerobic digestion [ 32 , 33 , 34 ] or methanogenic activity and diversity of methanogenic communities in natural environments [ 35 ], it has also been widely studied in engineered systems. Recent studies have investigated the effect of temperature in gas fermentation of H 2 /CO 2 or syngas [ 18 , 31 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 ], and one study modeled syngas biomethanation under mesophilic and thermophilic conditions and found that thermophilic communities showed higher specific methane productivity (18.8 mmol/g VSS/d) than mesophilic counterparts and that modulating the partial pressure of CO 2 can boost the product selectivity towards methane [ 44 ]. Another study found that psychrophilic conditions can inhibit methanogenic activity but either mesophilic or psychrophilic conditions can enrich homoacetogens [ 45 ]. To better understand the microbial communities and their functioning, omics techniques such as metagenomics and metatranscriptomics have been exploited to unravel in more detail the community members and their metabolic functions during biomethanation of hydrogen [ 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 ]. Recent biomethanation studies have identified the pathways under mesophilic and thermophilic conditions [ 54 ], combined experimental and model data to dissect the competition between methanogens and homoacetogens [ 55 ], determined how functional redundancy leads to quick recovery of the methane production rate [ 24 ], analyzed the carbon flow during methanogenesis inhibition [ 20 ], and revealed microbial community changes during syngas biomethanation in trickle bed reactors with different nutrient sources, including non-sterile digestates [ 56 ]. The enhancement of the process can be achieved by addition of zero valent iron (ZVI) [ 57 ], by introducing micro-porous materials as enhancer of biofilm immobilization and hydrogen mass transfer during ex situ biomethanation in trickle bed reactors [ 58 ] or via bioaugmentation with hydrogenotrophic methanogens ( Methanoculleus bourgensis and Methanothermobacter thermautotrophicus) as active microbial resource management during mesophilic and thermophilic in situ biomethanation [ 59 ]. Moreover, the enrichment of hydrogenotrophic methanogens during in situ biomethanation [ 60 ] and the microbial successions and carbon flow were investigated in standard continuous stirred tank reactors [ 61 ]. A recent study indicated a link between the inoculum origin and the acetate consumption rate [ 10 ], while other studies have found that inoculum sources (mesophilic and thermophilic) or sludge inocula from different types of reactors [ 17 , 62 ] play an important role during biomethanation of hydrogen. However, it remains unclear how the diversity of the inoculum affects the biomethanation process. Hitherto, little attention has been given to the influence of the inoculum diversity on the stability and performance of the biomethanation process. This study fills this gap by testing cultures of high, medium and low diversity under similar operational conditions. The present study compared inocula of different diversity regarding the process performance in biomethanation of hydrogen. It was hypothesized that the inoculum diversity influences the process, especially the unwanted accumulation of acetate. Process parameters were closely monitored and the microbial community composition was studied via terminal restriction fragment length polymorphism (T-RFLP) of mcr A (for methanogens) and 16S rRNA genes (for bacteria).",
"discussion": "4. Discussion Although it is well known that inoculation of biomethanation reactors plays an important role in process performance and stability, the particular role of microbial diversity of the inoculum has not been thoroughly studied. To clarify the answer to this question, we selected two sludge inocula of different origin and two enrichment cultures generated in our laboratory. One enrichment culture was specialized in straw degradation [ 63 ] and the second one was hydrogenotrophic enrichment culture [ 15 ]. To exclude the effect of temperature as a driver of process performance change, we selected inocula that originated from mesophilic conditions only. In the beginning of the experiment, the methane amount was comparable for all inocula. However, the methane produced in batch cycles of 24 h decreased dramatically in the cultures inoculated with the enrichment cultures. In contrast, the methane amounts produced by the cultures inoculated with sludge remained in a similar range throughout the entire experiment. In fact, the methane amounts produced initially were similar to the values found in our previous studies [ 15 , 62 ], which indicates that the process results were comparable. Furthermore, the methane concentration of ≥90% was in the range of values found in the literature [ 6 , 7 , 26 , 69 ]. It was unexpected that the most diverse inocula outperformed an already accommodated hydrogenotrophic enrichment culture. Based on our observations it can be stated that the diversity of the inoculum played an important role in maintaining a high performing and stable process. Therefore, it is an important factor to consider for practical applications, especially when mixed cultures from biogas plants or wastewater treatment plants are available to perform biomethanation. Hydrogenotrophic enrichment cultures suffering from low production of methane and VFA accumulation can be remediated by applying microbial resource management measures such as refining the growth medium and using sodium sulfide as reducing agent as described recently in our study [ 15 ]. Furthermore, bioaugmentation with a diverse methanogenic community (e.g., sludge/anaerobic granules from WWTP) could be a simple and practical microbial resource management measure. Acetate production and accumulation during biomethanation of hydrogen can be problematic if homoacetogenesis is stimulated and acetoclastic methanogenesis is not functioning. Previous studies have found acetate as a major by-product during biomethanation of hydrogen [ 7 , 11 , 12 , 26 , 29 , 70 , 71 ], whereas other studies have also found propionate, iso-butyrate and n-butyrate [ 13 , 18 ]. The occurrence of these by-products can be explained by the activity of homoacetogenic bacteria as they can produce acetate and small amounts of butyrate [ 72 ]. Propionate could be produced from amino acids during the recycling of biomass and its accumulation can be explained by the inhibition of propionate degraders due to high hydrogen partial pressure. The production of organic acids of higher carbon chain length can be also explained by chain elongation of acetate, which is enhanced by feeding hydrogen [ 73 , 74 ]. In this study, the cultures S37 and Nrich produced significantly more acetate than PF and WW. For S37 and Nrich cultures, this observation could be explained by (i) the high hydrogen partial pressure, which created a selective advantage for acetate production, (ii) the inhibition of acetate and propionate metabolizing syntrophic bacteria or low relative abundance of this functional group, (iii) the absence of acetoclastic methanogens and (iv) the lower diversity of S37 and Nrich communities compared with their WW and PF. A previous study confirmed that the inoculum and the predominant methanogens are important for the performance of biomethanation [ 62 ]. We found that in enrichment cultures, only one methanogen was the most dominant (S37: Methanoculleus , Nrich: Methanobacterium ), while the sludge derived cultures were more diverse. Our results are consistent with our previous results and the reports found in literature [ 7 , 15 , 29 , 47 , 62 ]. Thus, it can be inferred that these types of methanogens have a selective advantage when hydrogen is highly available. In the light of the aforementioned results, it seems that both Methanoculleus and Methanobacterium are needed for the most efficient process (WW and PF), while only one of them (as in S37 and Nrich) is not sufficient. Diversity analyses based on ecological indices have been poorly reported in biomethanation of hydrogen research. One study quantified the α-diversity as means to ensure that the sequencing depth was sufficient to cover the microbial richness [ 7 ]. The same authors highlighted that the community complexity in serial upflow and bubble column reactors showed distinct ordination patterns between the different reactor types based on PCoA analysis [ 7 ]. Another study quantified the α-diversity based on amplicon sequencing data and found higher diversity under mesophilic conditions than under thermophilic conditions [ 75 ]. A similar finding was described in a recent publication but ecological indices (α- or β-diversity) were not calculated [ 29 ]. Hydrogen addition to reactors acts as a strong selection factor favoring hydrogenotrophic metabolism, resulting in a decrease in the diversity as reported previously [ 15 , 75 ]. In the current study, we observed that the richness of the methanogenic communities decreased after long-term feeding of hydrogen. We used only mesophilic inocula and the highest richness was found in the sludge inocula (PF and WW). This was expected, since the other two inocula (S37 and Nrich) originated from enriched communities utilizing a single substrate—either wheat straw (S337) or H 2 /CO 2 (Nrich). Figeac and colleagues highlighted that high microbial diversity enabled faster adaptation to changes in temperature during biomethanation of hydrogen [ 29 ]. Therefore, it can be inferred that diversity plays an important role in adaptation to new conditions. In the current study, the most diverse cultures produced higher amounts of methane than the less diverse ones, a finding supported by the linear relationship between the richness of the inocula and the amount of methane produced, especially in long term operation ( Figure 6 ). The advantages of mixed cultures such as sludge are: (i) their complex microbial communities harboring several metabolic functions and allowing stable process performance; (ii) the ability to withstand starvation periods, which can be explained by the dominant methanogens present in the inoculum; (iii) functional redundancy as recently demonstrated by [ 24 , 62 ]; and (iv) fewer concerns with regard to the purity of the input gas. To explore our findings in more detail, further investigations with amplicon sequencing and metagenomics are needed to identify the taxonomy and potential metabolic functions of the microbial communities."
} | 4,220 |
20485560 | PMC2869309 | pmc | 1,848 | {
"abstract": "Enterobacter sp. 638 is an endophytic plant growth promoting gamma-proteobacterium that was isolated from the stem of poplar ( Populus trichocarpa × deltoides cv. H11-11), a potentially important biofuel feed stock plant. The Enterobacter sp. 638 genome sequence reveals the presence of a 4,518,712 bp chromosome and a 157,749 bp plasmid (pENT638-1). Genome annotation and comparative genomics allowed the identification of an extended set of genes specific to the plant niche adaptation of this bacterium. This includes genes that code for putative proteins involved in survival in the rhizosphere (to cope with oxidative stress or uptake of nutrients released by plant roots), root adhesion (pili, adhesion, hemagglutinin, cellulose biosynthesis), colonization/establishment inside the plant (chemiotaxis, flagella, cellobiose phosphorylase), plant protection against fungal and bacterial infections (siderophore production and synthesis of the antimicrobial compounds 4-hydroxybenzoate and 2-phenylethanol), and improved poplar growth and development through the production of the phytohormones indole acetic acid, acetoin, and 2,3-butanediol. Metabolite analysis confirmed by quantitative RT–PCR showed that, the production of acetoin and 2,3-butanediol is induced by the presence of sucrose in the growth medium. Interestingly, both the genetic determinants required for sucrose metabolism and the synthesis of acetoin and 2,3-butanediol are clustered on a genomic island. These findings point to a close interaction between Enterobacter sp. 638 and its poplar host, where the availability of sucrose, a major plant sugar, affects the synthesis of plant growth promoting phytohormones by the endophytic bacterium. The availability of the genome sequence, combined with metabolome and transcriptome analysis, will provide a better understanding of the synergistic interactions between poplar and its growth promoting endophyte Enterobacter sp. 638. This information can be further exploited to improve establishment and sustainable production of poplar as an energy feedstock on marginal, non-agricultural soils using endophytic bacteria as growth promoting agents.",
"introduction": "Introduction Endophytic bacteria reside inside living plant tissue without harming it [1] . Endophytic colonization is considered as a sign of a healthy plant system, as many endophytes will promote the growth, health and development of their host plant. Any organ of the plant can by colonized with a broad diversity of bacterial endophytes, many of which are closely related to common soil bacteria representative of genera such as Enterobacter , Pseudomonas , Burkholderia , Bacillus , and Azospirillum \n [2] – [4] . The diversity of endophytes is dependent on plant species, cultivar and probably cultivation conditions [5] , [6] . Plant roots are the main site of endophytic colonization. Root colonization by bacteria was described to involve several stages [7] . In the initial step bacteria move towards the plant roots either passively via soil water fluxes, or actively via specific induction of flagellar activity by plant-released compounds (chemotaxis). Second, a non-specific adsorption of bacteria to roots occurs, followed by anchoring (3rd step) that results in firm attachment of bacteria to the root surface. Specific or complex interactions between the bacterium and the host plant, such as the secretion of root exudates, may arise resulting in the induction of bacterial gene expression. Finally, endophytic bacteria can enter the plant at sites of tissue damage, which naturally arise as the result of plant growth, through root hairs and at epidermal conjunctions [8] . In addition, plant exudates given off through these wounds provide a nutrient source for the colonizing bacteria and thus create favorable conditions. This model of endophytic root colonization was confirmed by several microscopic studies for a number of plants [9] – [11] , including poplar [12] , [13] . Alternatively, endophytic bacteria can use vector organisms (e.g. arbuscular mycorrhizae and insects) to gain entrance to the apoplastic spaces to colonize the host plant [14] – [16] . Although likely to occur for many plant species, the involvement of specific vector organisms for endophytic colonization has not been demonstrated to play a role in poplar. After entering the plant, endophytic bacteria must establish themselves. Once established, they can enhance plant health and/or growth by producing plant growth regulating compounds such as indole acetic acid (IAA), acetoin (3-hydroxy-2-butanone), 2,3-butanediol and cytokinins, or metabolize compounds like phenyl acetic acid (PAA), gamma-aminobutyrate (GABA) or the stress ethylene precursor 1-aminocyclopropane-1-carboxylic acid (ACC) [17] – [23] . Endophytic bacteria can also protect the plant from fungal, microbial or insect infections by producing chitinases [24] , mannitol dehydrogenase [25] , volatile organic compounds and other molecules with antimicrobial activity [26] , and by inducing ultrastructural modifications in plant tissues that hinder their penetration by plant pathogens [27] . \n Enterobacter sp. 638, which was isolated from the stem of a 10-year-old hybrid poplar ( Populus trichocarpa×P. deltoids cv. H11-11) [12] , belongs to the family Enterobacteriacea whose endophytic members were identified in several plants species including cucumber [28] – [30] , grapevine [31] , maize [30] , [32] and potato [33] , [34] . Enterobacter sp. 638 is able to increase by up to 40% the growth of several species of poplar, including the Populus deltoides × P. nigra cultivars DN-34 [12] and OP367 (unpublished data). Enterobacter sp. 638 was also found to provide systemic drought resistance to poplar (S. Taghavi and L. Newman, unpublished). In this study we describe the analysis of the Enterobacter sp. 638 genome sequence and use metabolite analysis to confirm the production of phytohormones and antimicrobial compounds. Using quantitative RT-PCR we confirmed the dependence of the production of acetoin and 2,3-butanediol on the presence of sucrose, a major plant sugar, in the growth medium. Exploitation of the Enterobacter sp. 638 genome sequence presents a major path forward to identify via a systems biology approach the key functions in plant growth promotion, plant protection and more generally to validate the model describing endophytic colonization, establishment and interaction with the host plant. These findings can be further translated into comprehensive strategies to increase plant establishment and biomass production, which can be used to improve sustainable agriculture, bioenergy feedstock production on marginal lands, or fight desertification of arid areas.",
"discussion": "Discussion Analysis of the whole genome of the plant growth promoting endophytic bacteria Enterobacter sp. 638, which was isolated from the stem of poplar, reveals several features that reflect the dualistic lifestyle of this bacterium: survival in the rhizosphere, which represents a harsh and competitive environment, and endophytic colonization of the relatively protected plant environment. This is represented by the metabolic capabilities of this strain, as we discussed in the results section, and by the architecture of its genome. The Enterobacter sp. 638 genome contains 8 IS elements, which is relatively low compared to other endophytic bacteria such as K. pneumoniae 342 and S. proteamaculans 568 whose genomes contain 20 [52] and 12 full-length IS elements, respectively, or compared with the 66 IS-like genes found on the genome of E. coli K12 [80] . More generally, an endophytic life style may protect the bacteria from the outside environment and therefore requires less genome plasticity. This idea is supported by the example of the clinical isolate Stenotrophomonas maltophilia K279a [81] , which has with 30 full-length IS elements twice the number of IS elements as found in strain R551-3 [82] , an endophytic bacterium to which it is closely related. The chromosome of Enterobacter sp. 638 encodes four toxin/anti-toxin (T/A) systems ( relE/B , Ent638_0434-0435; yeeV/U , Ent638_0476-0477; hipA/B , Ent638_2033-2034; and chpA/R , Ent638_2066-2067). This low number is representative for host-associated organisms [37] . In contrast, the relative high number of six T/A systems (five relBE and one parED ) found on plasmid pENT638-1 seems to reflect the dualistic life style, rhizosphere and endophytic, of Enterobacter sp. 638 [37] . When residing in the rhizosphere the presence of the low copy number (1 to 2 copies per chromosome) 157 kb pENT638-1 plasmid, which besides arsenic resistance does not seem to code for any function that provides a competitive advantage for rhizosphere survival, causes a burden on strain 638. However, the architecture of pENT638-1 seems to reflect an essential role for this plasmid in the endophytic lifestyle of this strain. In addition to the active parAB partitioning system, the T/A systems make it virtually impossible to lose this plasmid under non-selective conditions. The various T/A systems are strategically positioned in the proximity of four regions of pENT638-1 that encode putative genes related to plant adhesion and colonization ( Figure 2 ), such as a region coding for a putative hemagglutinin-related autotransporter (Ent638_4267). Based on their different G/C content, these regions were likely acquired via horizontal gene transfer. The T/A systems ensure their stable integration in pENT638-1 under non-selective conditions. In comparison the stable maintenance of plasmid pG8786, the Y. pestis pFra virulence-associated plasmid, is enforced by the presence of a single parAB partitioning system [39] . \n Enterobacter sp. 638 belongs to the Enterobacteriaceae , a genus that contains both beneficial plant associated microorganisms as well as opportunistic pathogens. Other examples of this dualistic life style include K. pneumoniae , which both have endophyte (strain 342) and opportunistic pathogen (strain MGH78578) [52] or S. maltophilia with the endophytic strain R551-3 and the opportunistic pathogens K279a strain [82] . It is therefore important to assess the pathogenic potential of strain 638. The Enterobacter sp. 638 genome annotation and its comparison with other endophytic bacteria and closely related (opportunistic) pathogenic bacteria revealed that strain 638 is lacking a type III secretion system, which is considered a prerequisite for an active virulent life style typical for plant pathogens such as Erwinia and P. syringae . Enterobacter sp. 638 and other well-known endophytic bacteria share many genetic determinants for adhesion and even hemolysin with opportunistic pathogenic bacteria. In fact, plasmid pENT638-1 alone carries four putative genes involved in plant adhesion. These genes are likely essential for colonization and not related to pathogenicity. Therefore, although functions putatively involved in virulence were found, we feel confident to conclude that Enterobacter sp. 638 is in overall a beneficial organism and not an opportunistic pathogen. This is supported by the broad plant growth promoting host range of Enterobacter sp. 638, which includes tomato, tobacco, poplar, and sunflower. Our study reveals that the genome of Enterobacter sp. 638 carries many genes that make it an interesting candidate for agricultural application, and to improve the growth of biofuel feedstocks such as poplar. The plant growth promoting properties of Enterobacter sp. 638 depend on different routes of interactions. Enterobacter sp. 638 can indirectly stimulate plant health by protecting its host against bacterial and fungal infections, via the competition for essential nutrients such as iron, and the production of the antimicrobial compounds. For example, production of 2-phenylethanol gives a competitive advantage to Enterobacter sp. 638 in the rhizosphere but will also protect its host plant against pathogen infection. Although Enterobacter sp. 638 is able to produce low levels of the phytohormone IAA [12] , its major pathway to directly promote plant growth and development seems to rely on the production of acetoin and 2,3-butanediol. Increased levels of these phytohormones stimulate root development and will result in better access to nutrients and water, which will consequently increase growth and establishment of the host plant and better management of available water. In addition, the production of acetoin and 2,3-butanediol by plant growth promoting bacteria was reported to increase systemic disease resistance [83] and drought tolerance [84] . This is consistent with our preliminary observations that Enterobacter sp. 638 increases drought tolerance in Populus deltoides × P. nigra OP-367. On the metabolome and transcriptome level the production of acetoin and 2,3-butanediol by Enterobacter sp. 638 was induced by the presence of sucrose, a sugar abundant in plants. Together with genes coding for sucrose transport and utilization, the budABC operon for acetoin and 2,3-butanediol synthesis is located on region 29. Such clustering of genes further supports a relation between sucrose utilization and the inducible synthesis of these phytohormones. It should also be stated that the genome of Populus trichocarpa lacks the genes for the biosynthesis of acetoin from pyruvate, but that the gene responsible for the conversion of acetoin into 2,3-butanediol was identified. This points to a remarkable interaction between Enterobacter sp. 638 and its poplar host with the endophyte responsible of the production of a phytohormone, and a precursor for another, that poplar is unable to synthesize, and where the production of the plant growth promoting compounds depends on the presence of plant synthesized compounds, such as sucrose, in the growth medium. So far, the budRABC operon from Enterobacter sp. 638 was found syntenic to the following genomes: Enterobacter cancerogenus ATCC 35316, K. pneumoniae MGH 78578 and 342, Enterobacter sakazakii ATCC BAA-894, Vibrio alginolyticus 12G01, and Vibrio cholerae N16961, MZO-3, B33, O395, V52, 2740-80, 1587, MAK 757, NCTC 8457, MZO-2 and 623-39. Other compounds, known to be involved in plant growth promotion, were predicted by genome analysis to be synthesized by Enterobacter sp. 638. Since these compounds, which include putrescine, spermidine and cadaverine, are also produced by a large variety of non-endophytic bacteria, their role in plant growth promotion by Enterobacter sp. 638 remains to be demonstrated. Another example of a plant growth promoting compound is 4-aminobutyrate (GABA), an important phytohormone and eukaryotic neurotransmitter. GABA is produced by the plant in response to parasite infection and can interact with the brain neurotransmitters of insects. Like E. coli K12, Enterobacter sp. 638 possesses the genes required for GABA synthesis, but cannot utilize it as a sole carbon or nitrogen source [12] . Annotation reveals that the Enterobacter sp. 638 genome lacks the gene encoding for the GABA permease, which makes it questionable if Enterobacter sp. 638 can produce and export GABA as a protecting agent. Overall, the Enterobacter sp. 638 genome sequence presents a major tool to identify via a systematic approach the key functions in plant growth promotion, plant protection and more generally to validate the model describing endophytic colonization, establishment and interaction with the host plant. A combination of transcriptomics, proteomics, metabolomics and mutagenesis to study the plant colonization process will be of invaluable help in this respect: it will allow assigning new functions to putative genes and pathways, help to detect new proteins, and confirm the metabolic potential of the strain. As a first step to better understand the endophytic interactions between Enterobacter sp. 638 and poplar, a whole genome Enterobacter sp. 638 microarray was designed that is currently being validated to study changes in gene expression that occur in strain 638 during the various steps of the endophytic colonization process of poplar. Simultaneously, microarrays studies can provide valuable information on how endophytic colonization affects gene expression in poplar. Changes in gene expression for Enterobacter sp. 638 and poplar will provide strong support to identify genes involved in the successful endophytic colonization process, including those putative genes related to plant adhesion and colonization, many of which were coded for by plasmid pENT638-1. Furthermore, gene expression studies combined with metabolite analysis and proteomics will help to better understand mechanisms for the inducible synthesis of phytohormones, signaling compounds and other secondary metabolites that play a role in endophytic colonization and plant growth promotion and development, as was already proven for acetoin and 2,3-butanediol synthesis. Comparative transcriptome and proteome analysis will also provide valuable insights in which other genes and pathways are affected during the endophytic colonization process and the observed stimulation in plant growth and development. Genetic engineering and mutation analysis of Enterobacter sp. 638 and poplar should confirm the role of genes and metabolic pathways in successful endophytic colonization and plant growth and development. These basic finding can eventually be translated into comprehensive strategies to exploit the use of endophytic bacteria to improve plant establishment and biomass production, which can be applied in sustainable agriculture, bioenergy feedstock production on marginal lands, or fight desertification of arid areas."
} | 4,462 |
29607036 | PMC5869360 | pmc | 1,849 | {
"abstract": "Abstract Community assembly may not follow predictable successional stages, with a large fraction of the species pool constituted by potential pioneering species and successful founders defined through lottery. In such systems, priority effects may be relevant in the determination of trajectories of developing communities and hence diversity and assemblage structure at later advanced states. In order to assess how different founder species may trigger variable community trajectories and structures, we conducted an experimental study using subtidal sessile assemblages as model. We manipulated the identity of functionally different founders and initial colony size (a proxy of the time lag before the arrival of later species), and followed trajectories. We did not observe any effects of colony size on response variables, suggesting that priority effects take place even when the time lag between the establishment of pioneering species and late colonizers is very short. Late community structure at experimental panels that started either with the colonial ascidian Botrylloides nigrum , or the arborescent bryozoan Bugula neritina , was similar to control panels allowed natural assembling. In spite of high potential for fast space domination, and hence negative priority effects, B. nigrum suffered high mortality and did not persist throughout succession. Bugula neritina provided complex physical microhabitats through conspecific clustering that have enhanced larval settlement of late species arrivals, but no apparent facilitation was observed. Differently, panels founded by the encrusting bryozoan Schizoporella errata led to different and less diverse communities compared to naturally assembled panels, evidencing strong negative priority effects through higher persistence and space preemption. Schizoporella errata founder colonies inhibited further conspecific settlement, which may greatly relax intraspecific competition, allowing resource allocation to colony growth and space domination, thus reducing the chances for the establishment of other species.",
"introduction": "1 INTRODUCTION The assembly of ecological communities is recognized today as a combined result of deterministic niche‐based mechanisms and neutral stochastic processes (Adler, HilleRisLambers, & Levine, 2007 ; Chase, 2007 ; Chase & Myers, 2011 ; Vellend, 2010 ). While species‐specific tolerance to environmental conditions and species interactions (e.g., competition and predation) may underlie convergent community dynamics at different habitat patches (Berlow, 1997 ; Caro, Navarrete, & Castilla, 2010 ; Louette, De Meester, & Declerck, 2008 ), several different studies have shown considerable temporal and spatial variation in community structure across sites prone to similar environmental conditions (Almany, 2004 ; De Meester, Vanoverbeke, Kilsdonk, & Urban, 2016 ; Edmunds, 2014 ; Hart, 1992 ; Klingbeil & Willig, 2016 ; Sutherland & Karlson, 1977 ; Trowbridge, 2007 ). Therefore, processes other than the ones above may play important, although overlooked, roles in the regulation of species assemblages (Fukami, 2015 ; Irving, Tanner, & McDonald, 2007 ). Historical contingencies caused by differences in immigration, birth, death, and local species extinctions have been recognized for a long period as important factors affecting the organization of many communities (Allen, VanDyke, & Cáceres, 2011 ; Chang & Marshall, 2017 ; De Meester et al., 2016 ; Fukami, 2015 ; Sutherland, 1974 ; Sutherland & Karlson, 1977 ). One class of such contingencies is referred to as priority effects, that is, the effects of early arriving species on the subsequent establishment of other species (Connell & Slatyer, 1977 ; De Meester et al., 2016 ; Fukami, 2015 ). Depending on the identity of first colonizers, species assemblages may evolve through different trajectories and develop to multiple stable states (Chase, 2003 ; Fukami & Nakajima, 2011 ; Jiang, Joshi, Flakes, & Jung, 2011 ; Osman, Munguia, & Zajac, 2010 ; Petraits & Dudgeon, 2015 ; Sutherland, 1974 ; Sutherland, 1978 ). Their history may thus have long‐lasting effects on species composition and abundance (Stier, Geange, Hanson, & Bolker, 2013 ; Weslien, Djupström, Schroeder, & Widenfalk, 2011 ), regulating the access to available resources (Blaustein & Margalit, 1996 ; Zuo, Li, Ma, & Callaway, 2016 ), productivity (Martin & Wilsey, 2012 ), energy flow, and nutrient cycling (Fukami et al., 2010 ). From the conservationist point of view, priority effects may even affect resistance to invasive species (De Meester et al., 2016 ; Dickson, Hopwood, & Wilsey, 2012 ; Stuble & Souza, 2016 ) and disturbance events (Symons & Arnott, 2014 ). Considering longer‐term effects, historical contingency may underlie patterns of genetic structure (Sefbom, Sassenhagen, Rengefors, & Godhe, 2016 ) and ultimately species evolution (De Meester et al., 2016 ; Fukami, Beaumont, Zhang, & Rainey, 2007 ). Priority effects are thus deeply linked to species coexistence and the maintenance of biodiversity (Adler et al., 2007 ; Chesson, 2000 ; Sutherland, 1974 ; Vellend, 2010 ). While historical contingencies can restrict our ability to forecast patterns of species assembling (Dickie, Fukami, Wilkie, Allen, & Buchanan, 2012 ; Sutherland, 1974 ), life history traits of some early colonizers may help predicting the consequences of priority effects (Cifuentes, Krueger, Dumont, Lenz, & Thiel, 2010 ; Cleland, Esch, & McKinney, 2015 ; Sutherland, 1978 ). Depending on traits of the early colonizers, priority effects may occur either by negative or positive species interactions, largely dictated by inhibition or facilitation during early stages of the development of communities (Fukami, 2015 ; Gerla & Mooij, 2014 ; Weslien et al., 2011 ). The best known inhibitory effects occur through niche preemption, when early founders monopolize important resources that would otherwise be available to other species (De Meester et al., 2016 ; Fukami, 2015 ; Sutherland, 1978 ), or when first colonizers act as ecosystem engineers, modifying their habitat and preventing the establishment of other organisms (Bonnici, Evans, Borg, & Schembri, 2012 ; Jones, Lawton, & Shachak, 1994 ). Depending on intrinsic species traits, however, ecosystem engineering may otherwise play an opposite role and facilitate species arriving later (Fukami, 2015 ; Jones et al., 1994 ), for example, by providing tridimensional substrates and therefore increasing settlement grounds (Russ, 1980 ), or through the mitigation of abiotic stress by supplying more benign microhabitats (Jurgens & Gaylord, 2016 ; Perea & Gil, 2014 ; Vogt et al., 2014 ), ultimately leading to species coexistence and an increase of biodiversity. The extent to which founders affect the trajectory of developing communities may depend on how conspecifics of early‐colonizer species interact. In the case of species with complex life cycles, the effects of adult cues to propagules or juveniles are particularly important, because they may deeply affect local abundance and spatial distribution patterns of individuals (Lara‐Romero, Cruz, Escribano‐Ávila, García‐Fernández, & Iriondo, 2016 ). Attraction through chemical cues may constitute an efficient way to promote aggregation in suitable habitats (Pawlik, 1992 ; Robinson, Larsen, & Kerr, 2011 ; Silva‐Filho, Bailez, & Viana‐Bailez, 2012 ), indirectly increasing the strength of density‐dependent regulation mechanisms of founder populations. Positive density‐dependent interactions include the mitigation of abiotic stress in crowding intertidal invertebrates (Jurgens & Gaylord, 2016 ; Minchinton, 1997 ), caterpillars (Klok & Chown, 1999 ), and plants (Vogt et al., 2014 ); enhanced reproduction, such as fruit dispersal in plants (Blendinger, Loiselle, & Blake, 2008 ) and fertilization in marine invertebrates (Kent, Hawkins, & Doncaster, 2003 ; Levitan, Sewell, & Chia, 1992 ) and terrestrial woodlice (Broly, Deneubourg, & Devigne, 2013 ); and diminishing of predation risk in invertebrates (Denno & Benrey, 1997 ; Turchin & Kareiva, 1989 ) and vertebrates (Blumstein & Daniel, 2003 ; Carrascal, Alonso, & Alonso, 1990 ). Negative interactions usually lie in some sort of intraspecific competition, which may reach unsustainable levels under conditions of very high‐population density (Branch, 1975 ; Chisholm & Muller‐Landau, 2011 ; Gerla & Mooij, 2014 ; Hart & Marshall, 2009 ; Robins & Reid, 1997 ). Therefore, the potential for propagule dispersal and the strength of intraspecific competitive interactions after establishment may ultimately have shaped the selection of chemical responses to conspecific cues, either positively or negatively. To establish such a link, there is a need to assess conspecific responses in light of species‐specific strategies on resource use and competitive hierarchy. In this sense, we could expect species in which individuals grow fast and tend to monopolize resources to repel conspecifics (and thus reduce intraspecific competition), and species in which isolated individuals are poor competitors, and rely on the attraction of conspecifics to form more resistant aggregations (Hart & Marshall, 2009 ; Porensky, Vaughn, & Young, 2012 ). Careful experimental manipulation of founder assemblages, followed by the examination of community trajectories, may help understanding how priority effects modulate species diversity and assemblage structure at advanced successional states (Chang & Marshall, 2017 ; Cifuentes et al., 2010 ; Edmunds, 2014 ). Such methodology, however, has been used majorly for simple organisms, and effects measured in terms of individual counts for species that contributed the most to contrasts between treatments (Geange, Poulos, Stier, & McCormick, 2017 ; Irving et al., 2007 ; Jiang et al., 2011 ). Probably because manipulating the abundance of more complex organisms in natural conditions is more challenging, effects on natural assemblages dominated by clonal or colonial plants or animals are less documented (De Meester et al., 2016 ; Fukami, 2015 ). For decades, the construction of marine facilities, such as harbors and marinas, has promoted several modifications in coastal ecosystems, increasing the availability of hard substrata, which may support a diverse community of sessile organisms (Bulleri & Chapman, 2010 ). While some studies show a clear competition–colonization trade‐off, and thus, sessile community succession being determined by niche‐based processes (Buss and Jackson, 1979 ; Edwards and Stachowicz, 2010 ) suggested that nontransitive competitive relationships among species may explain alternative community states across or within habitats under similar conditions. Under these circumstances, priority effects may be important, because common species cannot be classified a priori as either early or late successional species in any obvious way, and because propagules of several species belonging to distinct functional groups may be available anytime. The importance of this historical contingency has been extensively discussed in descriptive studies (Osman, 1977 ; Sebens, 1986 ), but as long as we know this is one of the few studies addressing experimentally in the field how the identity and survivor of founder species affect community assembling and diversity. To do so, we triggered different history contingencies by manipulating both the identity of early colonizers and the time lag between the founder arrival and the subsequent establishment of other species (Urban & De Meester, 2009 ) in sessile communities from shallow subtidal artificial substrates. We expected to find (i) founder‐specific priority effects, which would be (ii) more intense when the time elapsed between first colonization and the arrival of later species arrivals was longer. We also predicted (iii) lower diversity of advanced assemblages first colonized by strong competitors for bare space.",
"discussion": "4 DISCUSSION As predicted, our study indicates that the identity of founder species has an important role in the subsequent organization of sessile communities. However, contrasting to our predictions, significant departures from the natural assembling process, in which chances of priority effects were not increased by any means, were only evident for only one founding species, the space‐monopolizing species Schizoporella errata , regardless of initial colony size. When arriving first, S. errata decreased the chances of later species to settle nearby by preempting space, resulting in less diverse communities over time, majorly dominated by the founder itself. Differently, B. neritina , which do not monopolizes space, and Botrylloides nigrum , that is potentially able to fast space preemption but did not persist longer, led to similar richness and structure compared to natural untouched communities, restraining growth rates of clonal species and ensuring baseline species coexistence. Negative priority effects of S. errata were shown to be pervasive over time, possibly owing to this species resistance to predators (Oricchio, Pastro et al., 2016 ), and long colony duration (Sutherland, 1978 ; Sutherland & Karlson, 1977 ), and should only relax when aged colonies eventually detach, creating new space available for the establishment of other species. Our results differ from those reported in other studies that suggest larger priority effects when the founding population is either numerous or constituted by larger individuals, ensuring more effective resource preemption (De Meester et al., 2016 ; Fukami, 2015 ; Poulos & McCormick, 2014 ). This mechanism is evident, for instance, when comparing exotic and native pioneering plants, with exotic species commonly showing strong priority effects owing to their higher ability of resource use, which ensures faster biomass accumulation both in terms of increasing individual numbers and size. Such effects are still important when well‐adapted native species settle shortly after pioneers (De Meester et al., 2016 ; Dickson et al., 2012 ). Similarly, large coral reef fishes exhibited stronger priority effects with only 3 hr of residence before the arrival of other fishes (Poulos & McCormick, 2014 ). We ensured a differential chance of advantage for arriving first by allowing founding colonies of two size categories (a proxy of time lag) to more (for long‐lag/large colonies) or less (for short‐lag/small colonies) relaxed competition for resources before the establishment of later species (Urban & De Meester, 2009 ). Although we expected that colonies of different sizes would show different potential of monopolizing or providing resources, we surprisingly did not observe differences on the extent of outcomes between the two lags tested. This means that priority effects, such as the ones found for S. errata , take place very fast, probably right after settlement (<15 days) when colonies are just a few zooids large. In spite of its potential to rapidly spread over flat surfaces (Kay & Keough, 1981 ; Nandakumar, 1996 ), and thus to impose priority effects through resource preemption when arriving first (De Meester et al., 2016 ; Fukami, 2015 ), we observed no effect for the founding colonial ascidian B. nigrum on community structure. This species exhibited the lowest persistence, likely as a result of low‐survival rate and the lack of conspecific cues leading to colony clusters and thus the formation of larger and more resistant patches. Low resistance to predation, as previously observed for ascidians in the study area (Oricchio, Flores, & Dias, 2016 ; Vieira et al., 2012 , 2016 ) and elsewhere (Jurgens, Freestone, Ruiz, & Torchin, 2017 ; Osman & Whitlatch, 2004 ) may have played an important role. As such, we conclude that potentially strong competitors, as B. nigrum , may impose very limited priority effects on community structure because they fail to persist under average environmental conditions. Differently, the encrusting bryozoan S. errata exerted strong priority effects through space preemption, deeply affecting recruitment and consequently community trajectory and structure, resulting in low diversity at advanced stages when compared to the other treatments, which showed abundant species other than the founders themselves covering larger areas. After 5 months, ascidians were virtually absent on panels founded by S. errata while covering 10%–15% of area in panels founded by B. nigrum and B. neritina . Besides, S. errata was the dominant encrusting bryozoan in S. errata founded panels, covering around 65% of the area, while panels founded by B. nigrum and B. neritina had other species contributing to the encrusting bryozoan coverage, with Schizoporella covering only around 20% and 40%, respectively. Fast growth led to rapid space monopolization, preventing the establishment of late‐arriving species close to S. errata colonies (Jackson & Hughes, 1985 ; Kay & Keough, 1981 ), which face a high risk of being dislodged or overgrown (Nandakumar, 1996 ; Russ, 1982 ). The high‐survival rate observed here coupled with the well‐known resistance against predators (Lidgard, 2008 ; Oricchio, Pastro et al., 2016 ) resulted in remarkable persistence of S. errata growing colonies, rendering pervasive effects on community structure, which may last until any disturbance event results in the detachment of aged colonies. The capacity of Schizoporella to efficiently monopolize space and inhibit the arrival of other species was previously suggested in classic studies during the 70's, based on both observational and experimental approaches (Sutherland, 1978 ; Sutherland & Karlson, 1977 ). By forcing sets of experimental assemblages to be founded by functionally different sessile invertebrates, we provide here independent and novel evidence for the unique role of Schizoporella species worldwide in the determination of benthic community dynamics. In our study area, S. errata seems to be the strongest competitor in artificial hard substrates, owing both to its fast growth and resistance to predators (Vieira et al., 2016 ; Oricchio, Flores, & Dias, 2016 ; Oricchio, Pastro et al., 2016 ). Differently from earlier studies showing later Schizoporella die‐off and subsequent species replacement after 1 year (Sutherland, 1978 ; Sutherland & Karlson, 1977 ), high year‐round recruitment of S. errata likely allows this species to regain space and persist longer in the study area. Our results strongly suggest that large monospecific Schizoporella stands found on pier pilings at the Yacht Club of Ilhabela, and other boating facilities in the region (personal observations), may be primarily a result of priority effects, with selective predation on soft‐bodied invertebrates playing a complementary role. We predicted that founders such as B. neritina , which does not monopolize resources but rather increase environmental complexity, could facilitate a large number of species, and therefore increase diversity when arriving first (Fukami, 2015 ). Our results do support overall settlement facilitation, and previous studies had indicated that intricate branches of B. neritina can provide settlers of different species valuable refuge from predation (Breitburg, 1985 ; Russ, 1980 ; Walters & Wethey, 1996 ). However, contrary to expectations, we found no priority effects in panels started with this species. In spite of creating physically complex substrates that could provide settlement secondary habitat and shelter from predators to an array of other species, the overall abundance of recruits in B. neritina plots did not differ from control ones. It is possible that gregarious settlement has led to exceedingly high levels of intraspecific competition (Chisholm & Muller‐Landau, 2011 ; Gerla & Mooij, 2014 ), ultimately decreasing founder persistence and thus any eventual long‐term positive effects on species diversity. Despite of showing a higher survival rate after 30 days when compared to B. nigrum , B. neritina did not persist in advanced community stages, allowing S. errata to colonize and occupy substantial space in panels. This nullified any possible priority effects, and shifted communities to states similar to the ones observed in naturally assembled panels. Overall, our results highlight that priority effects caused by species capable of fast resource monopolization and that persist through time, such as S. errata , may change recruitment patterns and consequently further community assembly. Schizoporella errata remarkably precluded the establishment of several other species through space preemption, leading to relatively impoverished assemblages. While our results do not support avoidance as a mechanism underlying poor recruitment of the whole suite of species, they do strongly suggest the inhibition of conspecific recruitment near developing S. errata colonies. Such a response may relax intraspecific competition, allowing the allocation of resources to clonal growth resulting in fast resource monopolization. In contrast, poor competitors such as B. neritina may potentially create novel habitat for recruits of many different species, but still not cause longer‐term effects on community assembling because of their low persistence in the community. We conclude that founder effects ultimately depend on life history strategies of pioneering species, with density‐dependent effects on benthic stages playing a crucial role."
} | 5,407 |
29790348 | null | s2 | 1,851 | {
"abstract": "By combining antifouling shark-skin patterns with antibacterial titanium dioxide (TiO"
} | 21 |
32503910 | PMC7414962 | pmc | 1,852 | {
"abstract": "HMFO is the only enzyme described to date that can catalyze by itself the three consecutive oxidation steps to produce FDCA from HMF. Unfortunately, only one HMFO enzyme is currently available for biotechnological application. This availability is enlarged here by the identification, heterologous production, purification, and characterization of two new HMFOs, one from Pseudomonas nitroreducens and one from an unidentified Pseudomonas species. Compared to the previously known Methylovorus HMFO, the new enzyme from P. nitroreducens exhibits better performance for FDCA production in wider pH and temperature ranges, with higher tolerance for the hydrogen peroxide formed, longer half-life during oxidation, and higher yield and total turnover numbers in long-term conversions under optimized conditions. All these features are relevant properties for the industrial production of FDCA. In summary, gene screening and heterologous expression can facilitate the selection and improvement of HMFO enzymes as biocatalysts for the enzymatic synthesis of renewable building blocks in the production of bioplastics.",
"conclusion": "Concluding remarks. Two enzymes from two Pseudomonas species have been added to the repertoire of HMFOs available, where only one member was purified and characterized to date. The enzyme from P. nitroreducens is a promising candidate for HMF oxidation due to its higher catalytic efficiency for FFCA oxidation, the bottleneck for FDCA production, compared to that of the previously described Metsp HMFO. More importantly, in the context of industrial production of FDCA, additional advantages of the P. nitroreducens enzyme are its stability under exposure to H 2 O 2 and its robustness for long-term incubations, as shown by its high t 1/2 values under wider ranges of pH and temperature conditions. This feature would facilitate the enzyme’s reuse and applicability in continuous operation for the production of FDCA as a building block for the production of bioplastics.",
"introduction": "INTRODUCTION 5-Hydroxymethylfurfural oxidases (HMFOs; EC 1.1.3.47) are flavoenzymes of biotechnological interest, classified in the superfamily of glucose-methanol-choline oxidases and dehydrogenases (GMC) ( 1 ). The importance of these enzymes resides in their ability to catalyze the three consecutive oxidation steps for the production of 2,5-furandicarboxylic acid (FDCA) from 5-hydroxymethylfurfural (HMF) with only the need of molecular oxygen as a cosubstrate. The final product, FDCA, is reported as a building block for the production of renewable and biodegradable plastics through its polymerization into poly(ethylene-2,5-furandicarboxylate) (PEF) ( 2 ), which is expected to substitute for petroleum-derived poly(ethylene-terephthalate) (PET) plastics in the near future. HMF-oxidizing activity was first identified in Cupriavidus basilensis strain HMF14 ( 3 , 4 ) and more recently in Pseudomonas putida strain ALS1267 ( 5 ), with their metabolic pathways being characterized. The hmfH gene of C. basilensis encodes an HMFO (UniProt D5KB61) that contributes to the conversion of HMF into FDCA, while the lack of this gene in P. putida suggests that, in this species, the conversion of HMF is catalyzed by different enzymes. The only HMFO characterized to date is that from Methylovorus sp. strain MP688 ( Metsp HMFO), a homologue of the enzyme encoded by the hmfH gene of C. basilensis , which was produced in Escherichia coli ( 6 ). This flavoenzyme, with a flavin adenine dinucleotide (FAD) molecule as a prosthetic group, is active on primary alcohols, primary thiols, and hydrated aldehydes ( 6 , 7 ). Its catalytic mechanism, similar to that of other GMC oxidases ( 8 – 11 ), involves a proton transfer from the hydroxyl (or thiol) group to a conserved catalytic base, Metsp HMFO His467 ( 12 ), and hydride abstraction from the substrate α-carbon by the oxidized flavin. The reduced flavin is then reoxidized by molecular oxygen, yielding hydrogen peroxide as a by-product ( 13 ). The reported activity of HMFO on furfuryl alcohols and aldehydes makes this enzyme a suitable biocatalyst for the production of FDCA from HMF, which takes place through the hydrated 2,5-diformylfuran (DFF) and 2,5-formylfurancarboxylic acid (FFCA) gem -diol intermediates ( Fig. 1 ) ( 14 ). Other oxidases, such as galactose oxidase (GAO) from Dactylium dendroides and glyoxal oxidase (GLOX) from Pycnoporus cinnabarinus , have also been reported to act on furfural derivatives, but their activity is restricted to the alcohol or the aldehyde groups, respectively ( 15 ). A recent study has demonstrated the ability of aryl-alcohol oxidase (AAO) from Pleurotus eryngii to catalyze the separate oxidation of HMF, DFF, and FFCA, although the complete oxidation of HMF by AAO stops at the FFCA level, due to inhibition of the last oxidation step by the peroxide generated in the first two reactions ( 16 ). More recently, a new copper radical oxidase from Colletotrichum graminicola has been described as an AAO-type enzyme and reported to oxidize HMF into DFF and 5-hydroxymethylfurancarboxylic acid (HMFCA) into FFCA without detected activity on the aldehyde groups of these compounds ( 17 ). Due to the above-named characteristics, efficient production of FDCA by the latter oxidases requires the combination of several enzymes, such as the AAO/chloroperoxidase ( 18 ), AAO/peroxygenase ( 15 , 19 ), GAO/peroxygenase ( 15 , 20 ), GLOX/AAO ( 21 ), or AAO/catalase ( 16 ) couples already reported. Therefore, Metsp HMFO is the only enzyme described to date able to perform by itself (i.e., without the concourse of a second enzyme) the three oxidation steps for FDCA production from HMF. In recent years, several strategies have been followed to increase the performance of Metsp HMFO for FDCA bioproduction, such as its coexpression with the HmfH enzyme, which oxidizes HMF by the HMFCA route ( 3 ), in whole-cell systems ( 22 ), directed mutagenesis of Metsp HMFO ( 23 ), or its combination with lipase in an enzymatic cascade ( 24 ). FIG 1 Reactions for conversion of HMF into FDCA by Metsp HMFO. First, the alcohol group of HMF is oxidized, resulting in DFF. Then, oxidation of the hydrated form ( gem -diol) of DFF gives FFCA. Finally, hydrated FFCA is converted into FDCA. Taking into account the good performance and promising use of HMFO for FDCA production, in the present study, we focused first on the search for new HMFOs in databases. Then, as a result of Escherichia coli expression trials with four HMFO genes, two new enzymes, from Pseudomonas nitroreducens and an unidentified Pseudomonas strain ( 25 ), were purified and characterized, and their oxidative activity on HMF was analyzed in short- and long-term reactions.",
"discussion": "DISCUSSION HMFO screening, heterologous expression, and purification. The HMFO from Methylovorus sp. ( Metsp HMFO) is the only enzyme described to date that is directly related to HMF oxidation. This connection was established due to its homology with the related HmfH protein that forms part of a gene cluster involved in HMF metabolism in C. basilensis ( 3 , 6 ). In the present study, new HMFOs from P. nitroreducens and a Pseudomonas sp. strain were described, following their production as recombinant proteins. These two HMFOs, together with others from a Xanthobacter sp. strain and B. arachidis that could not be produced in E. coli , were selected from a total of 41 HMFO-like sequences analyzed after a database search for homology with Metsp HMFO. Although an earlier attempt to express the gene responsible for HMF oxidation by C. basilensis and a homologous gene from Paraburkholderia phytofirmans was made, only Metsp HMFO had been successfully produced as a functional protein ( 6 ) to date. Here, the two sequences from Pseudomonas strains, with 63 to 64% identity with the Metsp HMFO sequence, were heterologously expressed in E. coli as active, soluble enzymes, together with Metsp HMFO. The three recombinant HMFOs appeared as monomeric proteins, in agreement with a previous report on the latter enzyme ( 12 ), and were purified with good yields, increasing to three the number of members of the HMFO family currently available. New information on Metsp HMFO. Comparison of the results obtained here for the Methylovorus enzyme purified without and with a His tag ( Metsp HMFO and Metsp HMFO His ) indicated that the tag does not perturb the incorporation of the cofactor, as shown by the UV-visible spectra. Moreover, deconvolution of the CD spectra indicated an increase in the unordered structure, consistent with 90% of the His tag being disordered ( 27 ), and a slight decrease in the α-helix content. Neither activity nor the main physicochemical parameters were affected by the presence of the His tag, but a positive effect on stability under exposure to H 2 O 2 was found for the tagged Metsp HMFO enzyme. This would represent an advantage for enzymes generating H 2 O 2 , which, as reported for other oxidases of the same superfamily, can inhibit the enzyme reactions or the enzymes themselves ( 28 , 29 ). The kinetic parameters for vanillyl alcohol oxidation by Metsp HMFO were similar to those previously reported for the same enzyme ( 14 ). However, the reported k cat values for HMF and DFF were significantly higher (594 min −1 and 96 min −1 , respectively) than those obtained in the present study. Considering the steady-state kinetic parameters obtained here, the estimated times for HMF and DFF consumption would be 7 and 98 min, respectively, which is consistent with the results obtained in 1-h reactions (1.5 mM HMF and 2.5 μM HMFO). In contrast, Dijkman et al. ( 14 ) still detected DFF after 1 h of reaction between 2 mM HMF or DFF and 1 μM HMFO, while according to their k cat value, all DFF should be consumed in ∼20 min. In any case, the k cat and K m values for oxidation of these furfural derivatives by the Metsp HMFO produced here are of the same order as those previously reported for other oxidases acting on these substrates, such as AAO ( 19 ) and GLOX ( 21 ). Regarding FFCA, the catalytic efficiency, estimated as k obs /[FFCA], was of the same order as that previously reported for Metsp HMFO ( 12 ), although the strong inhibition by FFCA (above 3 mM) has not been described before. Properties of two new HMFOs. The new enzymes from Pseudomonas strains were able to directly produce FDCA from HMF, with slight differences in their optimal reaction conditions, such as the higher activity and stability of Pseni HMFO under slightly alkaline conditions. Although the new HMFOs do not show more activity on HMF and DFF than Metsp HMFO, it is important to mention that they show higher efficiencies for FFCA oxidation, the rate-limiting step in FDCA production by HMFO ( 14 ) and other oxidases ( 16 ). Strikingly, in terms of efficiency, Psesp HMFO showed values for FFCA oxidation similar to those observed for DFF oxidation. The cooperative effect for DFF oxidation by this enzyme can be attributed to slow conformational changes that accompany substrate binding or product release. Similar cooperative effects have been described for other monomeric enzymes with single ligand-binding sites ( 30 , 31 ). In the case of Pseni HMFO, the efficiency for FFCA oxidation was only half of that of DFF oxidation. The results described above contrast with the much lower efficiency observed for Metsp HMFO (FFCA oxidation 10-fold lower than DFF oxidation). The comparatively higher speed in rate-limiting FFCA oxidation by Pseni HMFO results in higher final yields of FDCA in HMF bioconversions under optimized conditions. Moreover, its higher tolerance of the H 2 O 2 produced during the three oxidation steps results in a higher enzymatic stability during turnover, increasing its half-life. This stability provides an important advantage, since oxidases are often inhibited by large amounts of H 2 O 2 . Thus, our data reveal Pseni HMFO as a robust and suitable biocatalyst for prolonged incubations, due to its longer half-life in wider ranges of pHs, temperatures, and H 2 O 2 concentrations. Analysis of HMFO molecular models. Molecular modeling of Psesp HMFO and Pseni HMFO was undertaken at the Swiss-Model server ( 32 ), using the Metsp HMFO crystal structure (PDB code 4UDP ) ( 12 ) as the template. The predicted secondary structures were similar, as expected from their far-UV CD spectra. However, inspection of their active sites revealed subtle differences in the environment (less than 6 Å) of the reactive N5 of FAD. Interestingly, in the Metsp HMFO crystal structure, a water molecule is located at 3.7 Å from the above-named N atom and 4.6 Å from the catalytic His467 ( Fig. 9a ), at a position that is most probably occupied by the substrates during catalysis. Concerning neighbor residues, the pair Met103-Val104 in Metsp HMFO ( Fig. 9a ) is replaced by Met103-Met104 in Psesp HMFO ( Fig. 9b ) and by Val104-Leu105 in Pseni HMFO ( Fig. 9c ). These two positions follow an active-site asparagine that is totally conserved in the GMC superfamily ( 1 ) as part of the conserved PS00623 sequence. Directed mutagenesis of Metsp HMFO revealed how an M103A substitution (a changed of Met to Ala at position 103) significantly modified the enzyme turnover (for vanillyl alcohol) due to substrate pocket modification ( 12 ). A free space, potentially affecting the position of HMF and its partially oxidized derivatives (DFF and FFCA) with respect to the flavin cofactor, could also exist in Pseni HMFO with a valine at this position. FIG 9 Details of the active sites of Metsp HMFO (a, d), Psesp HMFO (b, e), and Pseni HMFO (c, f). (a to c) Residues at less than 6 Å from the reactive N5 atom of the FAD cofactor (CPK-colored sticks) are indicated, with those differing from Metsp HMFO underlined. Distances from His467 and flavin ring to water2176 (red sphere) are shown in panel a. (d to f) Semitransparent solvent-accessible surfaces colored by electrostatic potential (red, negative; blue, positive) around the flavin access channel. Water2176 inside the channel is shown in panel d. From the crystal structure with PDB code 4UDP ( 12 ) (a and d) and molecular homology models ( 32 ) (b, c, e, and f). The above-described differences do not include the presence of residues that could promote carbonyl activation in FFCA ( 33 ), as found in classical aldehyde-oxidizing oxidoreductases. However, changes in the flavin environment of Psesp HMFO and Pseni HMFO could modify the interaction of the flavin with the protein. Such interactions can modulate the redox potential of the oxidized isoalloxazine ring, promoting the enzyme’s activity on less reactive substrates ( 34 ). Pointing to the active-site cavity, Val367 is found in Metsp HMFO. This residue is conserved in Psesp HMFO, while Leu367 replaces the valine in Pseni HMFO ( Fig. 3 ). Interestingly, it has been reported that mutating this valine to arginine in Metsp HMFO results in higher activity on FFCA ( 12 ), as found for Pseni HMFO. However, the effect in the Metsp HMFO variant is due to the introduction of a basic (arginine) residue, contributing to adequate positioning of the above-mentioned acidic substrate, while the natural change in Pseni HMFO would only slightly reduce the pocket size. On the other hand, the electrostatic surface potentials of the three HMFOs showed differences in charge distribution around the entrance of the flavin-access channel ( Fig. 9d to f ), which could also modulate the access of substrates. Namely, the electronegative region partially covering the channel entrance in Metsp HMFO could impede the access of acidic FFCA to the active site. However, the less electronegative entrances in Psesp HMFO and especially in Pseni HMFO would facilitate the entrance of this substrate, resulting in the observed higher activity of the new enzymes in catalyzing the rate-limiting step in HMF conversion to FDCA. The solvent access surfaces also showed that, although the O 2 substrate would more easily diffuse to attain the flavin ring of the buried cofactor in HMFOs, the entrance of the furfuryl-reducing substrates through the narrow access channel would most probably involve some side chain rearrangement at the residues forming the channel, as described in a related AAO ( 35 , 36 ). Concluding remarks. Two enzymes from two Pseudomonas species have been added to the repertoire of HMFOs available, where only one member was purified and characterized to date. The enzyme from P. nitroreducens is a promising candidate for HMF oxidation due to its higher catalytic efficiency for FFCA oxidation, the bottleneck for FDCA production, compared to that of the previously described Metsp HMFO. More importantly, in the context of industrial production of FDCA, additional advantages of the P. nitroreducens enzyme are its stability under exposure to H 2 O 2 and its robustness for long-term incubations, as shown by its high t 1/2 values under wider ranges of pH and temperature conditions. This feature would facilitate the enzyme’s reuse and applicability in continuous operation for the production of FDCA as a building block for the production of bioplastics."
} | 4,326 |
26648839 | PMC4664726 | pmc | 1,854 | {
"abstract": "Neuromorphic hardware are designed by drawing inspiration from biology to overcome limitations of current computer architectures while forging the development of a new class of autonomous systems that can exhibit adaptive behaviors. Several designs in the recent past are capable of emulating large scale networks but avoid complexity in network dynamics by minimizing the number of dynamic variables that are supported and tunable in hardware. We believe that this is due to the lack of a clear understanding of how to design self-tuning complex systems. It has been widely demonstrated that criticality appears to be the default state of the brain and manifests in the form of spontaneous scale-invariant cascades of neural activity. Experiment, theory and recent models have shown that neuronal networks at criticality demonstrate optimal information transfer, learning and information processing capabilities that affect behavior. In this perspective article, we argue that understanding how large scale neuromorphic electronics can be designed to enable emergent adaptive behavior will require an understanding of how networks emulated by such hardware can self-tune local parameters to maintain criticality as a set-point. We believe that such capability will enable the design of truly scalable intelligent systems using neuromorphic hardware that embrace complexity in network dynamics rather than avoiding it.",
"introduction": "1. Introduction What role does the brain serve for producing adaptive behavior? This intriguing question is a long-standing one. So far, most attempts to understand brain function for adaptive behavior have primarily described it as the computation of behavioral responses from internal representations of stimuli and stored representations of past experience, a description we will take issue with below. As computational systems have grown in functional complexity, the analogy between computers and the brain began to be widely adopted. The basic premise for this analogy was that both computers and the brain received information and acted upon it in complex ways to produce an output. This analogy between computers and the brain (also known as the computer metaphor) has provided a candidate mechanism for cognition, equating it with a digital computer program that can manipulate internal representation according to a set of rules. The extensive use of the computer metaphor has resulted in the applied notions of symbolic computations and serial processing to construct human-like adaptive behaviors. The task of brain science has become focused on answering the question of how the brain computes (Piccinini and Shagrir, 2014 ). The key issues, such as serial vs. parallel processing, analog vs. digital coding, and symbolic vs. non-symbolic representations, are being addressed using the computer metaphor wherein perception, action, and cognition are taken to be input, output, and computation. Traditional algorithms that are derived by adopting the computer metaphor have yielded very limited utility in complex, real-world environments, despite several decades of research to develop machines that exhibit adaptive behaviors. This impasse has forced us to rethink the notion of how adaptive behavior might be realized in machines. One inspiration comes from a key observation that was made in 1950s by Ashby ( 1947 ), when he designed a machine called the homeostat (Ashby, 1960 ). According to him, animals are driven by survival as the objective function and animals that survive are very successful in keeping their essential variables within physiological limits. The term homeostasis dates from 1926, when Cannon ( 1929 ) used it to describe the specialized mechanisms unique to living systems which preserve internal equilibrium in the case of an inconstant world (Moore-Ede, 1986 ). These variables and their limits are fixed through evolution. For example, in humans, if the systolic blood-pressure (which is an example of an essential variable) drops from 120 mm of mercury to 30, the change will result in death. Ashby's thesis then was that systems that exhibit adaptive behaviors are striving to keep their essential variables within limits. Our take-away was that machines that exhibit adaptive behaviors are like a control system that strives to keep a variable within or around a set point. A related observation is that the complexity of adaptive behaviors increases with the number of physiological parameters that are to be maintained within their limits. We believe that this includes collective essential variables that are learned during the animal's interaction with the environment. However, approaching this from a control system point of view, it can be interpreted as the system striving to maintain stability across a complex set of interacting or coupled control loops with several set points. It is known that maintaining stability of such complex networks with multiple set points is a non-trivial task (Buldyrev et al., 2010 ). A second non-triviality is due to the inherent delays associated with homeostatic mechanisms. These delays require that homeostatic processes are also predictive. Indeed, analysis of periodic variations in essential variables, such as plasma cortisol levels in humans, shows that the so-called responses of homeostatic mechanisms are largely anticipatory (Moore-Ede, 1986 ). The homeostat can be thought of as a simple kind of self-organization. Once its structure is set, external forces interact with various internal forces that automatically balance and create feedback that affects those same external forces. A special kind of dynamic balance, one that persists through and because of constant change, is known as criticality (for the notion of criticality intended, see Bak et al., 1987 ; Beggs and Plenz, 2003 ; Legenstein and Maass, 2007 ). A critical system is poised to react quickly to deviations or perturbations, because of a different balance at the system level—between decay and explosion. Homeostatic balance is a balance of forces, while critical balance is a balance of the dynamics themselves. A system that contains both kinds of balance is typical of so-called self-organized criticality (SOC). A slight reorganization of SOC puts criticality under the control of a homeostatic balance. Such a system would be driven toward criticality."
} | 1,589 |
36354322 | PMC9769770 | pmc | 1,856 | {
"abstract": "ABSTRACT l -Lactic acid (LA) is a three-carbon hydroxycarboxylic acid with extensive applications in food, cosmetic, agricultural, pharmaceutical, and bioplastic industries. However, microbial LA production is limited by its intrinsic inefficiency of cellular metabolism. Here, pathway engineering was used to rewire the biosynthetic pathway for LA production in Saccharomyces cerevisiae by screening heterologous l -lactate dehydrogenase, reducing ethanol accumulation, and introducing a bacterial acetyl coenzyme A (acetyl-CoA) synthesis pathway. To improve its intrinsic efficiency of LA export, transporter engineering was conducted by screening the monocarboxylate transporters and then strengthening the capacity of LA export, leading to LA production up to 51.4 g/L. To further enhance its intrinsic efficiency of acid tolerance, adaptive evolution was adopted by cultivating yeast cells with a gradual increase in LA levels during 12 serial subcultures, resulting in a 17.5% increase in LA production to 60.4 g/L. Finally, the engineered strain S.c-NO.2-100 was able to produce 121.5 g/L LA, with a yield of up to 0.81 g/g in a 5-L batch bioreactor. The strategy described here provides a guide for developing efficient cell factories for the production of the other industrially useful organic acids. IMPORTANCE \n Saccharomyces cerevisiae is one of the most widely engineered cell factories for the production of organic acids. However, microbial production of l -lactic acid is limited by its intrinsic inefficiency of cellular metabolism in S. cerevisiae . Here, the transmission efficiency of the biosynthetic pathway was improved by pathway optimization to increase l -lactic acid production. Then, the synthetic ability for l -lactic acid was further enhanced by adaptive evolution to improve acid tolerance of S. cerevisiae . Based on these strategies, the final engineered S. cerevisiae strain achieved high efficiency of l -lactic acid production. These findings provide new insight into improving the intrinsic efficiency of cellular metabolism and will help to construct superior industrial yeast strains for high-level production of other organic acids.",
"conclusion": "Conclusions. In this study, the biosynthetic pathway for LA production was successfully rewired in S. cerevisiae by combining pathway construction with product transport. Then, the potential bottlenecks for LA production were rationally identified and removed by screening and strengthening the monocarboxylate transporters. Finally, the performance of the engineered strain in the production of LA was improved by adaptive evolution to promote cell tolerance to the high concentration of LA. Based on these strategies, the final concentration of LA with strain S.c-NO.2-100 was increased to 121.5 g/L. Although LA production with strain S.c-NO.2-100 in our study is lower than that of S. cerevisiae SP1130 in the previous study ( 24 ), our study offers an alternative strategy based on the combination of metabolic engineering and adaptive evolution. This strategy has great potential for developing efficient microbial cell factories for production of the other industrially useful organic acids.",
"introduction": "INTRODUCTION L-Lactic acid (LA) is a three-carbon hydroxycarboxylic acid with extensive applications in food, cosmetic, agricultural, pharmaceutical, and bioplastic industries ( 1 , 2 ). Current industrial LA fermentations are based on different species of LA bacteria ( 3 ), but these bacteria are sensitive to low pH, and large amounts of neutralizing agents such as CaCO 3 and NaOH are necessary for industrial LA production ( 4 ). Thus, LA production with LA bacteria is limited by its high production cost due to the regeneration of precipitate lactate salts ( 5 ). Thus, yeast is an attractive alternative for production of LA, due to its advantages such as growing and surviving in low pH. Various yeast species have been metabolically engineered for LA production, such as Saccharomyces cerevisiae ( 4 , 6 , 7 ), Kluyveromyces lactis ( 8 , 9 ), Pichia stipitis ( 10 ), Zygosaccharomyces bailii ( 11 ), Candida utilis ( 12 ), and Candida boidinii ( 13 ). Among these, S. cerevisiae was the most widely engineered for LA production. Six metabolic engineering strategies have been investigated for LA production in S. cerevisiae ( Table 1 ). The first is to introduce heterologous lactate dehydrogenase (LDH) genes to redirect carbon flux from pyruvate to LA. When LDH from Lactobacillus plantarum and monocarboxylate transporters (JEN1) were overexpressed in S. cerevisiae , LA yield showed a large increase, to 0.52 g/g ( 14 ). The second strategy is to delete pyruvate decarboxylase genes ( PDC1 , - 5 , and - 6 ) or alcohol dehydrogenase genes ( ADH1 to - 5 ) to reduce ethanol accumulation. When the PDC1 and ADH1 genes were deleted, LA yield was significantly improved, to 0.75 g/g ( 15 ). The third strategy is to screen highly acid-tolerant yeasts to maintain a neutral intracellular pH. Based on the hypothesis that the better LA-producing strain has a higher intracellular pH, the high-LA-producing strain S. cerevisiae CEN.PK m850 was obtained by three consecutive rounds of cell sorting from the UV-mutagenized populations of S. cerevisiae Z26, and its LA production was increased to 70 g/L ( 16 ). In addition, by adaptive laboratory evolution of the LA-producing S. cerevisiae SR8LDH, the evolved S. cerevisiae BK01 was able to produce 119 g/L LA without the use of pH neutralizers ( 17 ). The fourth strategy is to express monocarboxylate transporters (JEN1, ADY2, or ESBP6) to export LA. The JEN1 and ADY2 genes were constitutively expressed in S. cerevisiae \n jen1 Δ-LDH and S. cerevisiae \n ady2 Δ-LDH, respectively, leading to a higher external LA concentration ( 5 ). The fifth strategy is to delete the S -adenosylmethionine synthetase (SAM2) gene to remodel the cell membrane during acid stress. When SAM2 was deleted in S. cerevisiae CEN.PK m850, LA production was increased by 5.4%, to 69.2 g/L, compared with no SAM2 deletion ( 18 ). The sixth strategy is to delete NADH-consuming enzymes (NDE1/2) to enhance the cofactor availability of intracellular redox. LA was produced at 117 g/L, with a yield of up to 0.58 g/g, under low-pH conditions by deleting NDE1 and NDE2 in S. cerevisiae SP3 ( 19 ). In summary, LA production has been improved by metabolic engineering strategies ( 20 ), but LA productivity still needs to be enhanced to improve the intrinsic efficiency of cellular metabolism. TABLE 1 Comparison of LA production by S. cerevisiae strains S. cerevisiae strain Titer (g/L) Yield (g/g glucose) Productivity (g/L/h) Reference SPP 17.4 0.30 0.15 \n 4 \n PK27 37.9 0.66 0.79 \n 48 \n YIBO-7A 55.6 0.62 0.77 \n 21 \n CEN.PK m850 sam2Δ 69.2 0.88 0.96 \n 18 \n Z26 70 0.93 1.00 \n 16 \n AF297C 75 0.75 0.75 \n 15 \n YIBL- pdc1/5Δ 82.3 0.38 0.81 \n 22 \n SP7 117 0.58 2.39 \n 19 \n BK01 119 0.72 1.24 \n 17 \n SP1130 142 0.89 3.55 \n 24 \n NO.2-100 121.5 0.81 1.69 This study In this study, S. cerevisiae was used as a model system to rewire the biosynthetic pathway for LA production ( Fig. 1 ). Transporter engineering was conducted to improve LA export, and adaptive evolution was used to enhance acid tolerance. Based on these strategies, LA productivity was improved, and the final engineered strain, S.c-NO.2-100, was able to produce 121.5 g/L LA. FIG 1 Major metabolic pathways for the formation of LA in S. cerevisiae . PDC1, 5, and 6, pyruvate decarboxylase; ADH1, 2, 3, 4, and 5, alcohol dehydrogenase; ALD, acetaldehyde dehydrogenase; ACS, acetyl-CoA synthetase; A-ALD, acetylating acetaldehyde dehydrogenase; l -LDH, l -lactate dehydrogenase; JEN1, monocarboxylate transporter; TCA, tricarboxylic acid.",
"discussion": "RESULTS AND DISCUSSION Rewiring the biosynthetic pathway for LA production. In S. cerevisiae , ethanol is the main by-product of l -lactic acid (LA) production ( 21 ). Three pyruvate decarboxylase (PDC) genes, PDC1 , PDC5 , and PDC6 , contribute directly to ethanol production, but PDC activity is mainly from PDC1 and PDC5 genes ( 22 ). To enhance LA production and reduce ethanol accumulation simultaneously, three l -LDH genes, from Lactobacillus casei ( LcLDH ), bovines ( BoLDH ), and Rhizopus oryzae ( RoLDH ), were used to replace the coding region of PDC1 in the chromosome of S. cerevisiae through homologous recombination. When LcLDH , BoLDH , and RoLDH were expressed, LA production was increased to 12.4 g/L, 15.3 g/L, and 9.8 g/L, respectively, which were 21.5-, 26.8-, and 16.8-fold higher than yields of the control strain S.c-0 ( Fig. 2A ). Ethanol accumulation was decreased by 31.5%, 40.7%, and 24.5%, but its titers were still as much as 18.7 g/L, 16.2 g/L, and 20.6 g/L, respectively ( Fig. 2B ). In addition, cell growth was reduced compared with that of the control strain S.c-0 (see Fig. S2 in the supplemental material), but there was no significant difference among Lc LDH, Bo LDH, and Ro LDH activities (Fig. S1). These results indicated that ethanol accumulation was not significantly reduced by deleting PDC1 . FIG 2 Rewiring the biosynthetic pathway for LA production. (A) Effect of gene expression or deletion on LA accumulation. (B) Effect of gene expression or deletion on ethanol accumulation. PDC1, pyruvate decarboxylase; ADH1, alcohol dehydrogenase; LDH, lactate dehydrogenase. In S. cerevisiae , the cytosolic alcohol dehydrogenase (ADH1) gene contributes most of the catalytic activity for converting acetaldehyde to ethanol ( 15 ). To further reduce ethanol accumulation, we deleted the ADH1 gene in strain S.c-PΔ-B. The resulting strain, S.c-PΔAΔ-B, produced only 7.4 g/L ethanol, which was 54.3% lower than that of strain S.c-PΔ-B ( Fig. 2B ). At the same time, the concentration of LA was increased by 104.6%, to 31.3 g/L, with its yield from glucose (Y lac ) being 0.35 g/g ( Fig. 2A ). However, strain S.c-PΔAΔ-B showed growth retardation, leading to a 34.1% decrease in optical density at 600 nm (OD 600 ) compared with that of strain S.c-PΔ-B (Fig. S2), possibly due to the accumulation of intracellular acetaldehyde caused by the ADH1 deletion. This accumulation could affect the activity of acetaldehyde dehydrogenases (ALDs) by substrate inhibition ( 23 ), resulting in the deficient supply of acetyl coenzyme A (acetyl-CoA) for normal cell growth ( 24 ). To overcome this limitation of the endogenous pathway, the heterogenous pathway was selected to replace or support the function of ALD–acetyl-CoA synthetase (ACS). The bacterially produced acetylating acetaldehyde dehydrogenase (A-ALD) can directly convert acetaldehyde to acetyl-CoA without energy consumption ( 25 ). Thus, the eutE gene from Escherichia coli was introduced and expressed under the control of the ADH1 promoter in strain S.c-PΔAΔ-B, resulting in strain S.c-PΔAΔ-BE. The specific activity of A-ALD in strain S.c-PΔAΔ-BE was increased by 2.1-fold compared with that of strain S.c-PΔAΔ-B (Fig. S3). In addition, cell growth of strain S.c-PΔAΔ-BE (OD 600 = 10.4) was not significantly different from that of the control strain S.c-0 (Fig. S2). Further, LA titer (43.6 g/L) was increased by 39.3% compared with that of strain S.c-PΔAΔ-B ( Fig. 2A ). However, ethanol accumulation was similar with and without eutE expression ( Fig. 2B ). These results indicated that cell growth could be improved by introducing the heterogenous pathway to supply acetyl-CoA, increasing LA production. The increased acetyl-CoA levels might serve as a driving force to increase the synthesis of acetyl-CoA-originated building blocks such as amino acids, fatty acids, and sterols ( 26 ). Based on this, cellular metabolic activities might be activated to redirect the major metabolic flux from ethanol accumulation to LA production ( 24 ). Improving LA production by transporter engineering. Strain S.c-PΔAΔ-BE showed a large increase in LA production, but its Y lac was only 0.48 g/g, possibly due to the fact that the accumulation of the intracellular LA causes intracellular acidification and l -LDH inhibition, leading to a decrease in Y lac ( 14 ). To demonstrate this possibility, the intracellular LA and pH were determined for strains S.c-0 and S.c-PΔAΔ-BE. The concentration of intracellular LA in strain S.c-PΔAΔ-BE was increased by 152.2%, compared with that of the control strain S.c-0 ( Fig. 3A ). In addition, the intracellular pH in strain S.c-PΔAΔ-BE was 11.7% lower than that of the control strain S.c-0 ( Fig. 3B ). Furthermore, l -LDH activity in strain S.c-PΔAΔ-BE without CaCO 3 as a neutralizing agent was reduced by 18.5% compared with that of CaCO 3 addition ( Fig. 3C ). These results indicated that the accumulation of the intracellular LA exerted toxic effects on LA production, possibly suggesting that the transport capacity of LA needed to be enhanced to transport LA out of S. cerevisiae . FIG 3 Improving LA production by transporter engineering. (A to C) Effects of the monocarboxylate transporters on intracellular LA concentrations (A), intracellular pH (C), and l -LDH activity. (D) A series of JEN1 , ADY2 , and ESBP6 expression cassettes were designed with different combinations. (E) Concentrations of LA achieved by different JEN1 , ADY2 , and ESBP6 expression cassettes. The genes JEN1 , ADY2 , and ESBP6 encode the native monocarboxylate permeases, which have been used to export LA, acetic acid, formic acid ( 5 , 27 ). First, we tested the effect of JEN1 , ADY2 , and ESBP6 individually on LA production, and the highest concentration of LA (51.4 g/L) was obtained with strain S.c-NO.2 ( Fig. 3D ). At the same time, the intracellular LA was decreased by 29.3% compared with that of strain S.c-PΔAΔ-BE ( Fig. 3A ), and the intracellular pH was increased by 5.6% ( Fig. 3B ). In addition, its l -LDH activity without CaCO 3 as a neutralizing agent was reduced by 10.3% compared with that seen with CaCO 3 ( Fig. 3C ). Next, we analyzed the effect of combinations of two of these genes on LA production. When JEN1 and ESBP6 were overexpressed simultaneously, the LA titer increased to 49.6 g/L, which was similar to that of JEN1 overexpression ( Fig. 3D and E ). Finally, the concentration of LA was increased to 50.3 g/L by simultaneously overexpressing JEN1 , ADY2 , and ESBP6 , which was also similar to that of JEN1 overexpression ( Fig. 3D and E ). These results indicated that the monocarboxylate transporters effectively enabled the export of LA, especially for JEN1. JEN1 is a member of the sialate-proton symporter subfamily in the major facilitator superfamily ( 28 ). JEN1 can be induced to take up LA, but when LA is accumulated inside yeast cells ( 29 ), it also can mediate the efflux of LA ( 5 , 14 ) due to the fact that the pK a value of LA is much lower than the cytoplasmic pH value in yeast cells ( 14 ). Thus, a large proportion of the accumulated LA in cytosol is in the dissociated form and has to be actively transported outside yeast cells ( 14 ). Enhancing LA production by adaptive evolution. Although strain S.c-NO.2 showed good performance in LA production, growth limitation by the inhibitory effect of LA is still a major bottleneck for high production of LA ( 30 ). Thus, to further enhance LA productivity, we carried out adaptive evolution by cultivating the cells, with a gradual increase in LA levels from 10 to 60 g/L during 12 serial subcultures ( Fig. 4A ). Among 100 LA-tolerant candidate strains, strain S.c-NO.2-100 was selected based on glucose consumption ability and LA production level ( Fig. 4B ). In addition, strain S.c-NO.2-100 showed better growth than the unevolved strain S.c-NO.2 on medium A containing LA ( Fig. 4C ). FIG 4 Enhancing LA production by adaptive evolution. (A) Schematic illustration of the adaptation process. (B to G) Effects of adaptive evolution on LA production (B), acid tolerance (C), glucose consumption (D), cell growth (E), LA production (F), and ethanol formation (G). When the evolved strain S.c-NO.2-100 was used for LA fermentation, its final titer, yield, and LA production were increased to 60.4 g/L, 0.67 g/g, and 0.84 g/L/h, which are 17.5%, 17.5%, and 75.0% higher, respectively, than those of the unevolved strain S.c-NO.2 ( Fig. 4D to G ). In addition, its cell growth (OD 600 ) and the average glucose consumption rate showed 42.1% and 50.6% increases compared with those of the unevolved strain ( Fig. 4D and E ). However, ethanol formation was increased by 27.2% compared with that of the unevolved strain ( Fig. 4G ). These results indicated that adaptive evolution was efficient for screening strains with good performance in the production of LA. Adaptive evolution is a powerful tool for strain development in industrial applications ( 30 , 31 ). Generally, adaptive evolution is performed by progressively increasing stress to screen microbes with the corresponding phenotype in batch cultivation, commonly by means of tube culture, flask culture, and plate culture ( 32 ). In this process, spontaneous mutations accumulate, thus yielding the desired phenotype ( 33 ). Further, multiple experimental purposes can be achieved by combining adaptive evolution with metabolic engineering, which have great potential in strain development. In this study, the evolved strain S.c-NO.2-100 was obtained, which showed good LA production and cell growth. However, the physiological mechanism underlying LA tolerance of S. cerevisiae still needs to be more thoroughly understood. Although LA tolerance appears to be a very complex trait, there is already substantial knowledge about this mechanism. The generation of LA tolerance has been demonstrated to be closely related to various cellular metabolism and regulation processes ( 34 ), as follows. (i) The first such process is transcriptional regulation. The transcriptional response upon LA stress is largely regulated by the HAA1 regulon ( 35 , 36 ). LA productivity was increased by overexpressing HAA1 in an LA-producing S. cerevisiae strain ( 31 ). (ii) Second is intracellular pH (pHi) homeostasis. pHi homeostasis is tightly regulated by the H + -ATPase pump (PMA1) in the plasma membrane and the V-ATPase pump in the vacuolar membrane ( 37 ). PMA1 overexpression was used as a candidate method for improving organic acids and low pH tolerance in yeast. (iii) The third process is anion transport. To counteract lactate anion accumulation in S. cerevisiae , anions have to be exported out of S. cerevisiae cells by lactate anion transporters such as JEN1 and ADY2. Overexpression of JEN1 and ADY2 could increase LA production in S. cerevisiae ( 5 ). (iv) The fourth process is reactive oxygen species (ROS) scavenging. ROS formed in S. cerevisiae under aerobic conditions not only can cause lipid, protein, and nucleic acid oxidative damage but also can act as second messengers to induce various cellular processes. To deal with this issue, S. cerevisiae could be metabolically engineered to increase the formation of ROS scavengers such as glutathione (GSH) ( 38 ) and ascorbic acid ( 39 ). (v) Next is cell envelope rearrangements. To counter weak acid stress or low external pH, cell envelope rearrangements can be achieved by reinforcing the cell wall structure to decrease porosity and altering the lipid composition of the plasma membrane to increase membrane rigidity ( 40 , 41 ). As it is a key enzyme responsible for S -adenosylmethionine synthesis involved in phospholipid biosynthesis, deletion of SAM2 could further enhance acid tolerance and LA production in LA-producing S. cerevisiae ( 18 ). (vi) Last is amino acid, iron, and energy metabolism ( 34 ). LA stress can lead to a substantial decrease in intracellular amino acids by disrupting the proton gradient to affect the amino acid transporters and disturbing vacuolar integrity to affect amino acid storage in the vacuole. Metal cation homeostasis upon LA stress is regulated mainly by the transcription factor AFT1, which can alter the expression levels of many iron-related proteins. LA stress has a negative influence on energy metabolism through disruption of the electron transport chain, the ATP-generating metabolic pathways, and the energy-requiring export of protons and anions. Production of LA in a 5-L bioreactor. We next tested LA production of the evolved strain S.c-NO.2-100 in a 5-L batch bioreactor. In this batch culture, glucose was rapidly consumed during cell growth and LA synthesis and was depleted completely at 72 h ( Fig. 5 ). Strain S.c-NO.2-100 grew continuously from 0 to 72 h and attained a maximal OD 600 of 15.3 at 72 h ( Fig. 5 ). LA accumulated gradually in the broth from 0 to 72 h, and the maximal titer, yield, and productivity of LA were 121.5 g/L, 0.81 g/g, and 1.69 g/L/h, respectively, at 72 h ( Fig. 5 ). These results suggest that the final strain S.c-NO.2-100 can be utilized for efficient production of LA in fermentation. FIG 5 Production of LA by strain S.c-NO.2-100 in a 5-L batch bioreactor. Conclusions. In this study, the biosynthetic pathway for LA production was successfully rewired in S. cerevisiae by combining pathway construction with product transport. Then, the potential bottlenecks for LA production were rationally identified and removed by screening and strengthening the monocarboxylate transporters. Finally, the performance of the engineered strain in the production of LA was improved by adaptive evolution to promote cell tolerance to the high concentration of LA. Based on these strategies, the final concentration of LA with strain S.c-NO.2-100 was increased to 121.5 g/L. Although LA production with strain S.c-NO.2-100 in our study is lower than that of S. cerevisiae SP1130 in the previous study ( 24 ), our study offers an alternative strategy based on the combination of metabolic engineering and adaptive evolution. This strategy has great potential for developing efficient microbial cell factories for production of the other industrially useful organic acids."
} | 5,520 |
35855980 | PMC9287189 | pmc | 1,857 | {
"abstract": "Microbial fuel cells (MFCs) are a technology that can be applied to both the wastewater treatment and bioenergy generation. This work discusses the contribution of improvements regarding the configurations, electrode materials, membrane materials, electron transfer mechanisms, and materials cost on the current and future development of MFCs. Analysis of the most recent scientific publications on the field denotes that dual-chamber MFCs configuration offers the greatest potential due to the excellent ability to be adapted to different operating environments. Carbon-based materials show the best performance, biocompatibility of carbon-brush anode favors the formation of the biofilm in a mixed consortium and in wastewater as a substrate resembles the conditions of real scenarios. Carbon-cloth cathode modified with nanotechnology favors the conductive properties of the electrode. Ceramic clay membranes emerge as an interesting low-cost membrane with a proton conductivity of 0.0817 S cm −1 , close to that obtained with the Nafion membrane. The use of nanotechnology in the electrodes also enhances electron transfer in MFCs. It increases the active sites at the anode and improves the interface with microorganisms. At the cathode, it favors its catalytic properties and the oxygen reduction reaction. These features together favor MFCs performance through energy production and substrate degradation with values above 2.0 W m −2 and 90% respectively. All the recent advances in MFCs are gradually contributing to enable technological alternatives that, in addition to wastewater treatment, generate energy in a sustainable manner. It is important to continue the research efforts worldwide to make MFCs an available and affordable technology for industry and society.",
"conclusion": "7 Conclusions In this work, review and analysis of most recent scientific publications on MFCs technology have been performed in order to assess the contribution of configurations, electrode and membrane materials, electron transfer mechanisms, and cost of components on the current and future development of MFCs. Dual-chamber MFCs stand out for their excellent capability to adapt of the needs of the users. Likewise, the use of carbon-based electrode materials stands out as the ideal choice. In addition, it is possible to enhance their performance through the application of nanotechnology. The economic feasibility and potential for increasing power densities in MFCs are the main strengths of carbon-based material. Besides, the use of clay for membrane fabrication is suggested due to the low cost of both, the material itself and its manufacturing process. Also, this membrane produces power density similar to that obtained with the Nafion membranes. The use of a mixed consortium for biofilm formation at the anode, with domestic wastewater as substrate in the anodic chamber and HCl as catholyte, are the most favorable conditions, since they resemble those found in real scenarios. In this way, it is possible to consider wastewater as a potential source of energy and reusable water, rather than a hazardous waste. Therefore, it is strongly suggested that the main findings of this study be taken into account in order to continue adding efforts in the development and effective application of MFCs.",
"introduction": "1 Introduction Water and energy have a close connection. Water is required for all sources of energy production, including the electrical energy production, and energy is necessary for the disposal of water and the treatment of wastewater [1] . Water is an essential natural resource for humans and all life in the earth. It is an important component in ecosystems health, food production, socio-economic progress, and energy production. Water and sanitation systems must work together to ensure the human health and development [2] . However, according to The United Nations and Water Security and Sustainable Management Report 2020 [3] data, 1.8 million people lack of safe managed sanitation services. Moreover, more than 80 percent of wastewater returns to ecosystems without being treated. This wastewater discharged without any treatment generates negative effects on human health, natural environment, and global economics, in both, local population and far-away population from the pollution source. According to the World Health Organization [4] , it is estimated that in middle-and-low development countries, 842,000 of annual deaths are related to wastewater and sanitation. From the environmental point of view, untreated or partially treated wastewater generates contamination of surface water, soil and groundwater. When it is discharged into natural water bodies such as lakes and rivers, this water can infiltrate into aquifers and deteriorate the quality of fresh water. Also, the untreated wastewater that reaches the oceans contributes increasing the number of de-oxygenated dead zones. The marine ecosystem damage is estimated to reach an area of 245,000 k m 2 . This has a direct impact on the economy of the fishing industry, livelihoods, and food chains. Therefore, poor water quality interferes with economic development [5] . On the other hand, access to energy is key to social development. According to the British Petroleum Statistical Review of World Energy data from 2020 [6] , the global energy consumption grew at a rate of 1.3 percent. This growth was less than the growth reported in 2018 with 2.8 percent. Although the energy production was led by natural gas and renewable sources, global carbon dioxide ( C O 2 ) emissions keep growing. While the C O 2 growth rate 0.5 percent in 2019 was less than the annual average of 1.1 percent reported since 2010, it is still imperative to try to curb its growth. Notice that, C O 2 is the most abundant greenhouse gas that is mainly generated by burning fossil fuel and it is directly related to the global temperature and sea-level rise, sudden weather changes, and other adverse effects of unprecedented scale [7] . The world population continues to increase, and fossil fuels are being over-exploited faster than new sources are being discovered. It is important to develop green energies to reduce the negative effects of fossil fuel [8] . Therefore, wastewater treatment and alternative energy production is of main concern worldwide, as a society. Treatment of wastewater not only reduces pollutants from the water, but also enables the reuse of water [9] . A significant fraction of the world's energy demand can be obtained from wastewater, which contains an average chemical energy of 1.9 kWh m −3 stored as organic compounds, as long as it is converted into useful and economic energy [10] , [11] , [12] , [13] . It has been estimated that domestic wastewater approximately contains 9 times the amount of energy that is used to treat it [14] . Additionally, the development of renewable energy sources contributes to the reduction of greenhouse gas emissions and their associated negative effects. All this is reflected in positive effects on health, society, economics, and natural environment. An emerging technology that has aroused great interest among the scientific community due to its great potential to treat wastewater and generate bioenergy, are the microbial fuel cells (MFCs). These devices use bacteria as a catalyst to oxidize organic and inorganic matter and generate electrical current. Bacteria degrade the substrate contained in the wastewater, generate protons, and release electrons and carbon dioxide. Typical configuration of MFCs consists of an anode chamber, a cathode chamber, a membrane between the chambers, and an external electrical circuit. The released electrons flow from the anode (negative terminal) to the cathode (positive terminal), through a conductive material [15] . The electrons, protons, and oxygen present in the cathode (conventional configuration) react to form water [16] . In recent years, there have been important advances in MFCs research. Some studies have reported applications of MFCs at pilot-scales and field-scales (30 L [17] , 200 L [18] , 225 L [19] , 1,000 L [20] ) that allow to visualize the potential of MFCs in more realistic scenarios and the importance of further work on scaling up. The aim of this study is to review and analyze the most recent scientific publications on MFCs technology. Particular emphasis is done on the analysis of the configurations, anode, cathode and membrane materials, as well as the mechanisms for electron transfer and their influence on MFC performance. In addition, the study discusses the strategies implemented to improve those elements, and the effect they have on the ability of MFCs to treat wastewater and generate bioenergy. Furthermore, a cost analysis of the main MFCs materials (anode, cathode and membrane) is conducted. Finally, the areas with the greatest potential to promote the development of MFCs effectively and economically are identified."
} | 2,241 |
21829627 | PMC3150374 | pmc | 1,858 | {
"abstract": "Flocks of birds are highly variable in shape in all contexts (while travelling, avoiding predation, wheeling above the roost). Particularly amazing in this respect are the aerial displays of huge flocks of starlings ( Sturnus vulgaris ) above the sleeping site at dawn. The causes of this variability are hardly known, however. Here we hypothesise that variability of shape increases when there are larger local differences in movement behaviour in the flock. We investigate this hypothesis with the help of a model of the self-organisation of travelling groups, called StarDisplay, since such a model has also increased our understanding of what causes the oblong shape of schools of fish. The flocking patterns in the model prove to resemble those of real birds, in particular of starlings and rock doves. As to shape, we measure the relative proportions of the flock in several ways, which either depend on the direction of movement or do not. We confirm that flock shape is usually more variable when local differences in movement in the flock are larger. This happens when a) flock size is larger, b) interacting partners are fewer, c) the flock turnings are stronger, and d) individuals roll into the turn. In contrast to our expectations, when variability of speed in the flock is higher, flock shape and the positions of members in the flock are more static. We explain this and indicate the adaptive value of low variability of speed and spatial restriction of interaction and develop testable hypotheses.",
"introduction": "Introduction The beautiful coordination in flocks of birds has raised scientific interest since ages in both laymen and scientists [1] , [2] , [3] , [4] , [5] . Flocks of birds have great variation in shape: often different flocks have different shapes and a single flock changes its shape over time [1] , [5] , [6] . Extreme changes in shape and density of flocks occur during the aerial displays of thousands of starlings at dusk. For instance, sometimes during turning the flock may change in relative proportions, density and volume [7] , [8] , whereas at other times the shape of a flock may remain intact while only changing its orientation relative to the movement direction [4] . Further, during turning individuals may reposition their location within a flock in an amazing way [1] , [4] , [5] , [8] . This variability of shape differs markedly from what is described for schools of fish. Schools of fish are usually oblong in the movement direction [9] , [10] , [11] . However, under specific conditions, shapes of schools of fish are variable too, for instance, when a school is very large, and also when it is attacked by a predator. Very large schools have been described to be amorphous and to comprise extensions at the border, so-called pseudopodia, and sparse areas in the interior, called vacuoles, as if they consist of subgroups that move in somewhat different directions [12] . Similarly, in our model of very large schools (comprising up till 10.000 individuals) in which individuals have a limited view because it is blocked by those that are closest around them, shape appears more variable than in other models. This is due to the occurrence of subgroups with different movement directions in the school (Kunz and Hemelrijk, under review). Further, when being under attack of a predator, the shape of schools may become highly diverse. The shapes that emerge are for instance coined as ‘bend’, ‘flash expansion’, ‘herd’, ‘split’, and ‘hour glass’ [13] . Computer models of such attacks show that this diversity arises from the local differences of prey behaviour in the flock [14] , [15] . These depend on the prey's distance to the predator: Individuals close to the predator are avoiding it, while those further away from the predator are coordinating with the other school members. In conclusion it seems that the variability of school shape may arise from local differences in movement behaviour, thus, from reduced synchronisation of the school of fish. Since it is very difficult to study empirically [16] whether local differences in behaviour lead to a greater variation of shapes of flocks of birds, we will study it in a model of self-organised travelling groups, because such models have helped to create a better understanding of travelling groups in many aspects, such as their alignment [17] , [18] , [19] and direction choice [20] , [21] and, most importantly, also their shape. They show, for instance, that shape of a group of fish and birds changes when it is under attack of a predator [14] , [15] , [22] , that shape of fish schools depends on the synchronisation of spawning tendency [23] , and on density and school size [24] , [25] , [26] . Our models of fish schools have shown that the commonly observed oblong shape in the movement direction emerges as a side-effect of coordination and slowing down to avoid collisions [24] , . The elongated shape emerges, for all school sizes, in models in two dimensions or three, when individuals move at slow speed or fast and when a single school comprises individuals of a single body size or of two sizes. Furthermore, in our models of fish schools, schools appear to be more oblong the greater the number of individuals they include. We have confirmed these patterns in our empirical studies of three-dimensional positions of individuals in schools of 10 to 60 mullets: larger schools are both, denser and more oblong [25] . We attribute the fact that larger schools are more oblong to the higher number of adjustments necessary to avoid collisions in larger schools, because in larger schools individuals are closer to their nearest neighbours up till a certain saturation point [24] . Individual fish in larger schools are closer to their nearest neighbours. This emerges, because the attraction to other school members in larger schools is stronger because of the higher number of interaction partners. In our model of bird flocks [26] , however, like in flocks of real birds [4] , the relationship between density and group size is known to be absent. This may be due to the usually much larger group sizes that are investigated in studies of flocks of birds than of schools of fish: The flock sizes studied are already in the range in which density is saturated. In our present study of bird flocks we will use a model, called StarDisplay [26] . StarDisplay combines an adapted version of our former model of travelling schools of fish with characteristics of birds [24] , [26] . Modelled individuals fly following simplified aerodynamics, i.e. they experience lift, drag and the force of gravity [29] and in order to fly along a curve, like real birds, individuals roll into the direction of the turn until they are at a certain angle to the horizontal plane, the so-called banking angle [30] . The model is parameterised so that individuals resemble starlings, as regards body weight, speed, lift-drag coefficient [31] , roll rate [26] , number of interaction partners [4] and the way in which the flocks remain above a sleeping site of size similar to that of Termini in Rome [7] , [32] . Its patterns of flocking have been shown to resemble remarkably those of huge flocks of real starlings when flying above the roost recently studied with the help of stereo-photography above Rome [4] . The resemblance concerns the flat shape of flocks, the relative proportions (aspect ratios) of the flock shapes, their distribution of distances and angles to the nearest neighbours, their orientation, their balanced density between front and back and the way flocks turn [26] . Here, we investigate to what degree flock shape and its variability depends on local differences in behaviour. We assume that greater local differences in behaviour arise from larger flock size, lower number of interaction partners, sharp turning, rolling during turning and greater adjustment (and thus variability) of speed. We confirm that these traits cause larger differences in behaviour among individuals that indeed result in a greater variability of shape, except for one trait, namely variability of speed. We explain how high variability of speed results in low variability of shape of the flock. We derive testable hypotheses for real animals and speculate about the adaptive value of locality of interaction and adjustment of speed.",
"discussion": "Discussion We show that local variability of behaviour in a group generally leads to more variable flock-shape, but not in cases of local variability of speed. Instead, high variability of speed results in an oblong shape that is permanently oriented in the movement direction. Remarkably, a lower variability of speed, thus, a stronger synchronisation in a flock, leads to a variable orientation of the longest dimension of the shape relative to the movement direction. The present study shows that group size has a great impact on the variability of shape. The local variation in larger flocks is greater as is apparent from the greater changes in volume during sharp turns, from the lower global polarisation, and from the scale free correlation of the deviation from the average velocity with flock size in the model. The scale free correlation resembles that in real starlings [40] . The increase in the size of the subgroups with flock size is however larger in our model than in real starlings (gradient of the scale free correlation in the model is 0.44 and in starlings 0.35) indicating that in the model there is less local variation than in reality. This may arise from the greater uniformity of the environment in the model, due to the absence of all kinds of disturbances (such as other birds, including predators, wind, airplanes and very high buildings) [26] . The greater uniformity of environment may also be the cause that the volume of the flock in the model is smaller than in reality [26] . A higher number of interaction partners in our model decreases the variability of flock shape as a consequence of the greater synchronisation of the flock-members (as is apparent from the stronger scale free correlation between subgroup size and flock size, from the stronger global and local polarisation and the smaller changes of volume during turns). Similarly, when in a model of predation on fish schools prey- individuals interact with more neighbours while evading attacks of a predator, the shape of their schools becomes less diverse than when interacting with fewer partners [15] . Turning has a big impact on the variability of shape. Turning in the model resembles descriptions of turning of real flocks, for instance, of rock doves in several aspects [8] . This concerns the temporary changes of volume of the flock and its loss of altitude during a turn, see Fig. 6C of our earlier work [26] , the frequently occurring change in orientation of the flock and the repositioning of individuals as shown by Pomeroy and Heppner for rock doves in their Fig. 4B and 5 \n [8] . Large changes in volume arise only when flocks are large and individuals interact with few neighbours, because in this case individuals sometimes experience different environments (above and outside the roost), which desynchronises behaviour in the flock. Rolling into a turn is essential for creating both the reduction in volume and the loss in altitude in our model. The loss of altitude is a consequence of the reduced lift that individuals experience when banking. Without banking, the shape of the bird flock resembles that of a fish school, since it is very elongated in the movement direction ( Movie S7 ). Together these traits (large flock size, few interaction neighbours and rolling into a turn) cause the great variability of shape. Change of shape during turning and repositioning of individuals are a consequence of low variability of speed. Repositioning has been observed in several species, such as dunlins [5] , pewits [1] , rock doves [8] and starlings [4] . Repositioning of individuals in the flock arises, because all individuals follow an equal path length during a turn, as show for rock doves [8] . Low variability of speed causes the change of orientation of the flock and the repositioning of individuals, as is shown in our fish model, because these traits are absent when variability of speed is high ( Fig. 9 ). Here, when adjustability is high, due to the close proximity in the inner corner of the turn individuals slow down and in the outer corner, due to the large inter-individual distances, they speed up. Consequently, during a turn the shape of the school is maintained and individuals stay at approximately the same location in the school. This can be seen in Fig. 9B in which we gave individuals different grey-shades depending on their location in the group in the initial snapshot: they appear to be faithful to approximately the same location, left, front etcetera during the whole series of snapshots (for colour-version see Supplementary material, S1). The permanency of shape during turning due to high variability of speed extends our former theory about the causation of the oblong shape of fish schools to include turning behaviour [25] . This theory implies that the group shape becomes more oblong due to frequent slowing down by its members in order to avoid collisions [24] , [25] , [27] , [28] . Our finding that in StarDisplay variability of speed is accompanied also by elongation of the flock in the movement direction, suggests that if their speed could deviate from cruise speed more, this mechanism of elongation would work for birds also. Since shape of fish-schools is more oblong than that of bird flocks, we hypothesise that the variability of speed of birds is lower than that of fish. There may be several biological advantages to locality of interaction and low variability of speed. Locality of interaction may result in greater variability of behaviour among individuals in a flock. This may confuse predators and reduce their success at catching of prey. Low variability of speed, may confuse predators also through the accompanying repositioning of flock members during turns, the so-called ‘crossing paths’ [8] , [41] . Further, it may be advantageous by saving of energy through elimination of acceleration and for avoiding collisions by preventing collisions from front to back [26] (Hemelrijk & Hildenbrandt in prep). Collision avoidance may be more important for birds than for fish, because collisions are more dangerous for birds, because their movement is faster and the viscosity of their medium is lower. Despite its usefulness, our model has shortcomings. First, of such complex animals as birds, it concerns merely their movement behaviour in relation to the position and heading of others and of the roost, while using a simple model of flying behaviour, ignoring e.g. flapping flight. It ignores any behaviour related to other motivations, such as nutritional [42] , reproductive [23] or motivations to avoid a predator [15] . It also ignores environmental disturbances, e.g. by physical forces, such as wind. Thus, in nature, there will definitively be additional reasons that cause flock shape to be variable beyond those that we consider in this paper. Indeed, in the model the variability of shape of, for instance, small flocks of 200 birds is below that observed in real flocks in nature. A number of the explanations generated by our model can be used as testable hypotheses for empirical data, not only of birds but also of other animals moving in groups. Testable hypotheses from the present investigation concern effects of locality of interaction and variability of speed ( Table 2 ). Particularly in the light of the great effort and difficulties of collecting empirical data of three dimensional positions of animal groups [25] and particularly of flocks of birds [6] , [16] , such model-based hypotheses are valuable. 10.1371/journal.pone.0022479.t002 Table 2 Hypotheses for empirical testing derived from the model. 1) Greater locality of interaction causes more variable shape in terms of volume and aspect ratios a) in larger groups b) when individuals interact with relatively fewer interacting partners c) in a heterogeneous environment 2) Lower variability of speed causes higher variability of shape a) It induces shape to be less oblong in the movement direction b) It induces an almost random orientation of the oblong shape c) It causes changes in the orientation of the shape relative to the movement direction during turning d) It causes individuals to reposition themselves in the group during turns 3) Higher variability of speed is expected a) in fish rather than in birds b) to result in slowing down in inner corners during turning"
} | 4,191 |
30381948 | null | s2 | 1,859 | {
"abstract": "The survival of all organisms depends on implementation of appropriate phenotypic responses upon perception of relevant environmental stimuli. Sensory inputs are propagated via interconnected biochemical and/or electrical cascades mediated by diverse signaling molecules, including gases, metal cations, lipids, peptides, and nucleotides. These networks often comprise second messenger signaling systems in which a ligand (the primary messenger) binds to an extracellular receptor, thereby altering the intracellular concentration of a second messenger molecule which ultimately modulates gene expression through interaction with various effectors. The identification of intersections of these signaling pathways, such as nucleotide second messengers and quorum sensing, provides new insights into the mechanisms by which bacteria use multiple inputs to regulate cellular metabolism and phenotypes. Further investigations of the overlap between bacterial signaling pathways may yield new targets and methods to control bacterial behavior, such as biofilm formation and virulence."
} | 269 |
35889183 | PMC9319577 | pmc | 1,860 | {
"abstract": "Industrial production of synthetic nitrogen fertilizers and their crop application have caused considerable environmental impacts. Some eco-friendly alternatives try to solve them but raise some restrictions. We tested a novel method to produce a nitrogen bioinoculant by enriching a soil microbial community in bioreactors supplying N 2 by air pumping. The biomass enriched with diazotrophic bacteria was diluted and applied to N-depleted and sterilized soil of tomato plants. We estimated microbial composition and diversity by 16S rRNA metabarcoding from soil and bioreactors at different run times and during plant uprooting. Bioreactors promoted the N-fixing microbial community and revealed a hided diversity. One hundred twenty-four (124) operational taxonomic units (OTUs) were assigned to bacteria with a greater Shannon diversity during the reactor’s steady state. A total of 753 OTUs were found in the rhizospheres with higher biodiversity when the lowest concentration of bacteria was applied. The apparent bacterial abundance in the batch and continuous bioreactors suggested a more specific functional ecological organization. We demonstrate the usefulness of bioreactors to evidence hidden diversity in the soil when it passes through bioreactors. By obtaining the same growth of inoculated plants and the control with chemical synthesis fertilizers, we evidence the potential of the methodology that we have called directed bioprospecting to grow a complex nitrogen-fixing microbial community. The simplicity of the reactor’s operation makes its application promising for developing countries with low technological progress.",
"conclusion": "5. Conclusions During the different experimental stages, the soil microbial community composition and prevalent families reflected the differences in the substrate state (dry-soil, liquid, and biofilm) and mineral and carbon sources. The reactors fed with a modified Hoagland solution with low-cost N supplied only by air and citrate as the carbon source allowed the growth of a complex nitrogen-fixing microbial community. Nevertheless, the process of reaching a steady state lasted around 80 days. NH 4 , NO 2 , and NO 3 concentrations in the effluent were negligible, and TKN corresponded to the microbial biomass. The different communities’ compositions and their metabolic characteristics revealed the bioreactors configuration effect (soil-organic particles, batch-flocs, and packed-biofilm). The acclimatization process in the batch complete mix bioreactor selected a ‘free-swimmer’ community dominated by Beijerinkia and Zooglea . The microbial community in the biofilm from the nitrogen-fixing reactors was more diverse than those from soil or the acclimatization process and was dominated by Xanthobacter and Sphingobium . The microbial composition suggests the participation of the complete nitrogen transformations in the bioreactors. In addition, a complex nitrogen-fixing microbial community was identified in the plant root rhizospheres in each treatment. The COVID-19 pandemic and the war in Ukraine have highlighted the difficulties, especially for emerging countries with an agricultural base to access fertilizers, which, together with the environmental problems caused by their industrial production, urge local solutions to countries. Bioreactors have proved their value for enriching nitrogen-fixing populations capable of replacing ammoniacal nitrogen and studying microbial ecology and soil diversity with successive enrichment steps of the same sample. Microbial biomass from N-fixer reactors at different concentrations allowed the tomato plants to grow and the chemical nitrogen-containing solution. Thus, packed bioreactors and the experimental framework, demonstrated for the first time in the current study, could play a future role in augmenting agricultural practices.",
"introduction": "1. Introduction The intensification of agriculture through methods including intensive fertilization has substantially increased food availability but has imposed severe environmental consequences. Nitrogen-based synthetic fertilizer production through the Haber–Bosch process consumes more than 1% of the world’s total energy generated, using about 2% of the natural gas extracted and emitting more than 300 million metric tons of carbon dioxide [ 1 ]. In addition, applying N-based fertilizers to crops has led to a global nitrogen cycle disturbance, illustrated by the increasing eutrophication of land and water bodies, greenhouse gas emissions, and biodiversity losses, particularly of soil microbial communities [ 2 , 3 ]. In 2050, the world’s population is predicted to increase to around 10 billion, with consequent increases in the demand for food and agricultural expansion [ 4 , 5 ]. Some ecological alternatives involve the intensive use of crops that replace soil nitrogen content, such as legume-based fertilization systems, organic farming, and microbial bio-inoculants, using endophytic microbes to increase the supply of nutrients to crops. However, not all legumes are liable to intercrop with nonleguminous plants; the rotation process can take longer than expected or defy large-scale application [ 6 ]. Using manure, compost, and plant residues may lead to the accumulation of pharmaceuticals and antibiotics [ 7 , 8 ]. Moreover, it is insufficient to supply the demand for crop production [ 9 ]. As a result, abandoning synthetic N-fertilizers could lead to nutrient undersupply, even with increased legume cropping. Regarding bio-inoculants, even though their world market is growing, their application is still marginal compared with chemical fertilizers [ 10 ]. Several factors hinder their use, including identifying and tracking inoculated strains in the field, the poor understanding of relationships between microorganisms and plants, and complex production technology [ 11 ]. Plants and microbial communities co-exist depending on their mutual species interactions, the microenvironment generated by the physicochemical conditions of the soil, and the ability to adapt to changes in each of these conditions quickly. Therefore, the effects of plant species on microbial taxa are often not easy to predict a priori [ 12 ]. Even within a single species, plants can select different subsets of microorganisms at different stages of development, presumably relating to specific functions [ 13 ]. Thus, there is not a universally applicable bioinoculant for all crop types and soils. Manipulating active microbial communities in agriculture and developing new microbiome engineering approaches to address these challenges is a priority. These microbiome engineering approaches allow the manipulation and study of microbial communities in situ, without isolating species or model communities in the laboratory [ 14 ], and promote plant fitness and health [ 15 ]. Specifically, integrating microbiome engineering theory with bioprocess engineering offers a top-down approach. The sample is a blank canvas, and external stimuli are applied to identify patterns and self-assembled construction forms associated with specific conditions [ 16 ]. This approach has been used to develop strategies for enhancing bioremediation, providing additional nitrogen, and identifying adaptation strategies associated with nitrogen fixer communities [ 17 , 18 ]. The present study aimed to engineer a diazotrophic community in reactors using the self-assembly approach. Subsequently, the community was introduced into the rhizosphere of a nonlegume plant, and its capacity as a growth promoter was assessed. There was no difference in growth parameters between plants grown in chemical fertilizer versus inoculation, which suggests that this methodology could be used to replace chemical fertilization.",
"discussion": "4. Discussion 4.1. Bioreactors and Microbial Biodiversity The results indicate that the microbial community in bioreactors was capable of fixing nitrogen when provided with citric acid as the only carbon source. Different bacterial species carried out nitrogen fixation, and the nitrogen-fixing genera evolved with the growth of the plants. The use of bioreactors to detect and study complex microbial communities has been demonstrated for bacteria from the human microbiome and many environmental systems [ 28 , 29 ]. The bioreactor strongly impacted the community structure of the soil microbiome, as shown in Figure 6 of principal components. After seeding each bioreactor with the original soil solution, the bioreactor was isolated from the surrounding environment, and media and air, which were introduced to support the microbial community, were sterilized to prevent the introduction of new micro-organisms. Therefore, all the species detected in the different stages of adaptation were present in the initial soil sample. The changes in community structure reflect the dependency on co-metabolism and nutrient transformations in a complex ecological structure. Thus, one benefit of using bioreactors is that the relative abundance of species that are initially present in low quantities and which may go undetected can be magnified, allowing the further study of their metabolic capabilities and role in the community. We found at least three genera, Candidatus finniella , Azonexus , and Cloacibacterium , which were only identified in the bioreactors but not in the soil. Candidatus finniella has been described as an obligate intracellular parasite, an amoeboflagellate that perforates the cell walls of freshwater green algae and feeds on the algal cell contents by phagocytosis [ 30 ]. The presence of this organism means that the host protozoa are present and benefit from the bioreactors to grow. However, we do not know their role in the consortium, because we did not measure the eucaryotes; it could also mean that other protozoa are present. Protozoa are natural predators of bacteria, the species that prey on nitrogen-fixers in the soil have been studied, and their impact has been evaluated [ 31 , 32 ]. Therefore, it is important to control the growth of protozoa in bioreactors. In other experiments, we have achieved significant results by limiting oxygen. In the present work, a decrease in Candidatus finniella was observed in packed bioreactors (1.12%, 0.41%, 0.91%, and 0.27% in PcB t1 to PcB r4 , respectively). Azonexus , a genus that has been identified as a critical endophyte for nonlegumes, occurs only under microaerobic conditions and in the absence of high concentrations of other nitrogen sources [ 33 ]. The presence of this species refers to how beneficial microorganisms can be enriched in bioreactors with potential application to agricultural practice. The third genus detected only in the bioreactors was Cloacibacterium , a heterotrophic bacteria identified in wastewater treatment that plays an essential role in organic matter fermentation [ 34 ]. Therefore, its presence in the bioreactors is unsurprising as fermentative bacteria are highly favored in bioreactor environments. We analyzed the genus in the bioreactor involved with the nitrogen cycle as described by [ 35 ]. As we did not amplify the gene, we used the cross-matched names of the Uniprot database retrieved list (February 2022) and the assignment obtained from Silva2020. As expected, with no combined nitrogen input to the medium, the nitrogen fixers adapted to the growing conditions in the bioreactors determined the microbial succession. Importantly, different groups of nitrogen fixers were observed at high and low densities, demonstrating their importance in shaping ecology throughout diversity. Zooglea and Beijerinkia , in the batch reactor and the first days of adaptation in the packed reactors, at the end of the growth in the packed bioreactors, were replaced by Sphingobim and Xhantobacter. Other organisms not dependent on inorganic nitrogen also evolve differently in each type of bioreactor. The physical structure of the bioreactor environment significantly impacts the composition and, therefore, functionality of the bacterial community. The zeolite particles in the packed bioreactor provide a greater surface area for biofilms and more opportunities for localized communities. Moreover, the environment provided by the packed bioreactor is closer to that of the natural soil environment, thereby increasing the relevance of in vitro studies. For example, Candidatus obscurbacter was abundant in the batch bioreactor, whose physical and possibly functional space was occupied by an unknown group of the Class Bacteroidea. Our analysis also suggests a cluster organization of fixers and nonfixers (the complete database for the study is provided in the Supplementary Materials File S3 ). The possible flocs in batch reactors and biofilms in the packed reactors are schematized in Figure 7 . Zooglea was favored only in the batch reactor and has stood out for its ability to form flocs in activated sludge systems and flooded soils for rice cultivation [ 36 ]. One study found that aerobiosis favored Zooglea fluctuation because of the high affinity with oxygen [ 37 ]. The genus Beijerinckia is also found in environments rich in oxygen and carbohydrates, a positive environment for its growth [ 38 ]. On the other hand, Pelomonas has been identified in fluids for hemodialysis. Its ability to grow in oligotrophic media has been demonstrated [ 39 ], while Candidatus obscuribacter is a facultative anaerobe that could occupy the interior of the floc [ 40 ]. These genera found in relatively high density only in BB, with diverse metabolic characteristics, could suggest the organization of co-metabolism in the floc structure facilitated by Zooglea . Although it was impossible to measure dissolved oxygen in the packed reactors’ microenvironment, it is assumed that less oxygen reaches the microenvironments between the zeolite particles in the bioreactor. This condition, together with the carbon source change (glucose was gradually replaced by citrate in BB, while only citrate was used in the packed bioreactors), was a determinant of the growth of the genera Sphingomona , Xanthobacter , and Reyranella that exhibit the ability to use in N-deficient environments and organic substrates ranging from acetate to aromatic hydrocarbons [ 41 , 42 ]. Proteobacteria generally prefer environments rich in labile—easily transformed—carbon sources [ 43 ] and are found to fix nitrogen in bioreactors treating wastewater with a high carbon-to-nitrogen ratio [ 44 ]. Some aquatic Verrucomicrobia (Spartobacteria and Opitutae classes with aerobic and heterotrophic metabolism) can fix nitrogen. However, they are the most abundant taxa in some natural wetlands where the organic matter and nutrient contents are significantly higher [ 45 ]. This suggests that, in BB reactors, the phyla abundances were highly affected by the nutrient ratio. 4.2. Rhizosphere Soil Microbiome and Plant Growth Performance We found that plants inoculated with packed bioreactor effluent had the same growth as C+ plants that were fertilized with Hoagland solution. Plants in C− treatment died at seven weeks due to the absence of any nutrient addition, meaning that plant growth in plants in P 0.1 OD , P 0.2 OD , and P 0.3 OD was enhanced by microbial inoculation. The soil used to grow the tomato plant was characterized by scarce nutrients, low humidity, without surface vegetation, and little contribution of organic matter. These characteristics are hostile to many groups of microorganisms and generate two parallel scenarios: one in which microorganisms remain in latent forms through resistance structures such as spores (Firmicutes 89.8%). Other groups remain metabolically active but in low abundance due to environmental and nutritional limitations (Proteobacteria 9.8%) ( Figure 3 a). In a recent study, it was found that the Actinobacteria (0.3%), encode a full set of CAZymes, nitrogenases, and antibiotic synthetases and is related with infertile soils [ 46 ]. A similar microbial community between the initial soil and the soil inoculated with a low inoculum density (P_0.1 OD ) suggests that the plant initially recruits a microbiome in the soil that remained viable after sterilization and colonized the rhizosphere faster compared to the taxa present in the effluent from packed bioreactors at 0.1 of optical density P_0.2 OD . A greater abundance of genera such as Opitutus was observed in the rhizosphere with treatments S OD 0.2 and S OD 0.3. Phenotypic changes were also observed in plants roots; there was an increase in the root width and root volume as the amount of inoculum increased ( Figure 5 B,E); as there is a more significant microbial population, it possibly induces the plant to facilitate a more significant translocation of compounds toward the root to maintain the sizeable microbial community. However, this does not negatively affect the plant growth in stem length, meaning that an enhancement in Rubisco enzyme activity possibly causes a more efficient photosynthetic machinery due to higher CO 2 concentrations emitted by a high rate of soil microbial respiration. Many reports show the root phenotypic changes in response to microbial inoculations, especially under limiting environmental and sparse nutritional conditions [ 47 , 48 , 49 ]. For microbial community assessment, the space is documented as a critical factor. Reduced root diameter means less surface area for bacterial colonization for an individual root. The plant increases the surface area of its roots to recruit more microbes [ 49 ], which results in better long-term metabolic conditions for growth."
} | 4,404 |
36087062 | PMC9826415 | pmc | 1,861 | {
"abstract": "Abstract Marine chemical sedimentary deposits known as Banded Iron Formations (BIFs) archive Archean ocean chemistry and, potentially, signs of ancient microbial life. BIFs contain a diversity of iron‐ and silica‐rich minerals in disequilibrium, and thus many interpretations of these phases suggest they formed secondarily during early diagenetic processes. One such hypothesis posits that the early diagenetic microbial respiration of primary iron(III) oxides in BIFs resulted in the formation of other iron phases, including the iron‐rich silicates, carbonates, and magnetite common in BIF assemblages. Here, we simulated this proposed pathway in laboratory incubations combining a model dissimilatory iron‐reducing (DIR) bacterium, Shewanella putrefaciens CN32, and the ferric oxyhydroxide mineral ferrihydrite under conditions mimicking the predicted Archean seawater geochemistry. We assessed the impact of dissolved silica, calcium, and magnesium on the bioreduced precipitates. After harvesting the solid products from these experiments, we analyzed the reduced mineral phases using Raman spectroscopy, electron microscopy, powder x‐ray diffraction, and spectrophotometric techniques to identify mineral precipitates and track the bulk distributions of Fe(II) and Fe(III). These techniques detected a diverse range of calcium carbonate morphologies and polymorphism in incubations with calcium, as well as secondary ferric oxide phases like goethite in silica‐free experiments. We also identified aggregates of curling, iron‐ and silica‐rich amorphous precipitates in all incubations amended with silica. Although ferric oxides persist even in our electron acceptor‐limited incubations, our observations indicate that microbial iron reduction of ferrihydrite is a viable pathway for the formation of early iron silicate phases. This finding allows us to draw parallels between our experimental proto‐silicates and the recently characterized iron silicate nanoinclusions in BIF chert deposits, suggesting that early iron silicates could possibly be signatures of iron‐reducing metabolisms on early Earth.",
"introduction": "1 INTRODUCTION Precambrian Banded Iron Formations (BIFs) are widespread iron‐ and silica‐rich chemical sedimentary deposits that precipitated from anoxic, ferruginous seawater throughout the Archean Eon and Paleoproterozoic Era (Bekker et al., 2010 ; Farquhar et al., 2010 ; Johnson & Molnar, 2019 ; Cornelis Klein, 2005 ; Poulton & Canfeld, 2011 ). BIFs are often interpreted as illuminating records of Archean marine biogeochemical cycling, where chemical precipitates may have captured the intersection of ancient biological activity and seawater chemistry (Konhauser et al., 2017 ). The accurate identification of the primary BIF precipitates—and any evidence of early diagenetic reworking—is critical to reconstructing early seawater chemistry and deciphering the potential impact of microbial life on BIF formation, particularly in the absence of a robust fossil record (Knoll et al., 2016 ). However, the complex and micron‐scale BIF mineral assemblage often confuses the identification of primary versus secondary BIF mineral component(s). BIFs contain iron oxides (magnetite [Fe 3 O 4 ] and hematite [Fe 2 O 3 ]), iron silicates (greenalite [(Fe) 3 Si 2 O 5 (OH) 4 ], riebeckite, and stilpnomelane), microcrystalline quartz ([SiO 2 ], also known as chert), and carbonates (siderite [FeCO 3 ], dolomite [CaMg(CO 3 ) 2 ], and ankerite [Ca(Fe 2+ ,Mg)(CO 3 ) 2 ] (Bekker et al., 2014 ; Klein, 2005 ; Trendall, 2002 ). While ferric hydroxides [Fe(III)(OH) 3 ]) are widely hypothesized to be the initial BIF precipitate (e.g., Konhauser et al., 2017 ), much of the BIF assemblage likely derived from post‐depositional processes including microbial respiration (Craddock & Dauphas, 2011 ; Heimann et al., 2010 ; Johnson et al., 2013 ; Konhauser et al., 2005 ; Walker, 1984 ), fluid‐mediated reactions (Sun, Konhauser, et al., 2015 ), various grades of metamorphism (Bekker et al., 2014 ; Klein, 2005 ; Krapež et al., 2003 ; Trendall, 2002 ), and oxidizing fluids (Rasmussen et al., 2013 , 2014 ). Despite efforts to distinguish the imprint of post‐depositional processes on BIF mineral formation, the depositional environment, precursor phases, and the role of biology in BIF genesis remain disputed (Johnson et al., 2013 ; Johnson, 2019 ; Konhauser et al., 2017 ; Rasmussen et al., 2021 ; Tosca et al., 2019 ). In an effort to resolve the initial BIF mineral, a series of studies recently utilized high‐resolution petrography of BIF cherts to identify the original BIF phase(s) (Muhling & Rasmussen, 2020 ; Rasmussen et al., 2021 ; Sun, Konhauser, et al., 2015 ). Ubiquitous early‐mineralizing chert in BIFs is often attributed to the estimated high dissolved silica concentrations in the Archean ocean (Siever, 1992 ), which lacked silica‐secreting organisms (Maliva et al., 2005 ). The precompaction mineralization of low porosity chert (Fischer & Knoll, 2009 ; Krapež et al., 2003 ) makes it exemplary for the preservation of microfossils (Knoll et al., 2016 ; Maliva et al., 2005 ) and early minerals, either primary minerals (Rasmussen, Krapež, & Muhling, 2015 ) or early‐forming diagenetic phases. Therefore, recent discoveries of widespread and abundant greenalite nanoinclusions in well‐preserved 2.4 to 3.5 billion‐year‐old South African and Australian BIF cherts suggest that iron silicates were an important iron mineral in nascent BIF sediments (Muhling & Rasmussen, 2020 ; B. Rasmussen et al., 2021 ; Rasmussen, Krapež, & Muhling, 2015 ; Rasmussen, Krapež, Muhling, & Suvorova, 2015 ). Rasmussen et al. ( 2019 ) argue that the relatively uniform distribution and random orientations of these particles indicate that these iron silicate nanoinclusions were deposited as settling seawater precipitates and, therefore, were the primary BIF particles. However, another proposed mechanism for the precipitation of early iron silicate minerals is through the microbial respiration of iron oxides in BIF sediments in the presence of high dissolved silica (Fischer & Knoll, 2009 ; Percak‐Dennet et al., 2011 ; Robbins et al., 2019 ; Walker, 1984 ). Silicate minerals interpreted as early diagenetic precipitates forming in sedimentary porewaters also display random orientations in other Precambrian deposits (Lepot et al., 2017 ; Michalopoulos & Aller, 1995 ; Wacey et al., 2014 ), similar to the poorly defined orientation of iron aluminosilicate clays formed during simulated diagenesis of Amazon delta sediments (Michalopoulos & Aller, 1995 ). Furthermore, the preserved iron silicate inclusions in BIF cherts contain 10%–20% Fe 3+ (Johnson et al., 2018 ), which could represent relict Fe(III) originating from initial iron oxide minerals reduced through post‐depositional iron respiration (Johnson, 2019 ; Johnson et al., 2018 ). Dissimilatory iron reduction (DIR) is an anaerobic respiratory metabolism, performed by a variety of microbes, that couples the oxidation of organic carbon with the reduction of Fe(III) substrates (Lovley and Phillips, 1988 ; Kato et al., 2019 ; Myers & Nealson, 1990 ; Miot & Etique, 2016 ) as in Reaction ( 1 ), using CH 2 O as a simplified formula for organic carbon: (Reaction 1) \n 4 Fe OH 3 + 7 H + + C H 2 O → 4 Fe 2 + + HC O 3 − + 10 H 2 O . \n This paired iron‐ and carbon‐cycling metabolism has been suggested as both the progenitor of iron formation carbonates, where Fe(II)CO 3 precipitated from porewaters saturated in Fe(II) and inorganic carbon, and as an explanation for the lack of organic carbon in Precambrian BIFs (Baur et al., 1985 ; Craddock & Dauphas, 2011 ; Heimann et al., 2010 ; Johnson et al., 2008 ; Konhauser et al., 2005 ; Walker, 1984 ). Likewise, microbial iron respiration in the high‐silica pore waters of Archean sediments could have produced secondary iron silicate biominerals, as proposed by Fischer and Knoll ( 2009 ), through the saturation of Fe(II) and dissolved silica. Extensive DIR research describes the diverse secondary minerals associated with microbial iron respiration (further details in Supporting Information ). While this prior microbial iron reduction work displays a spectrum of DIR‐mediated mineral products, ranging from secondary Fe(III) oxides to Fe(II,III) phases such as magnetite or an Fe(II,III) green rust salt (e.g., Bae & Lee, 2013 ; Etique et al., 2016 ; Han et al., 2020 ; O'Loughlin et al., 2019 ; Salas et al., 2009 ; Zegeye et al., 2010 ), the absence of dissolved silica in these studies decouples their mineral observations from predictions of ancient DIR products in the siliceous geochemistry of Archean seawater. A limited set of experimental and environmental DIR investigations have included dissolved silica or used silicate minerals as Fe(III) substrates (e.g., Dong et al., 2009 ; Kostka et al., 1999 ; O'Reilly, 2005 ; Phoenix et al., 2003 ; Sergent et al., 2011 ); see Supporting Information for more details. Notably, Komlos et al. ( 2007 ) detected putative secondary Fe(II) silicates by Mössbauer spectroscopy but used aluminosilicate clay minerals as a starting iron substrate. While this earlier DIR work explored clay‐microbe interactions and the impact of silicic acid on iron reduction, the results of many silica‐amended DIR experiments are ill‐suited for comparison to the Archean environment due to their incongruent sources of silica, such as sand or clay minerals, and/or the addition of other improbable Archean chemical species, like organic or phosphate buffers. Additionally, ancient seawater composition likely varied from the experimental conditions employed in the vast majority of DIR studies. Recent models suggest the Archean ocean had [Si aq ] = 0.67 to 2 m m (Jones et al., 2015 ; Konhauser et al., 2007 ; Maliva et al., 2005 ; Siever et al., 1992 ), a circumneutral pH of 6.5–7.5 (Halevy & Bachan, 2017 ; Jones et al., 2015 ; Krissansen‐Totton et al., 2018 ) and approximately 10 m m dissolved inorganic carbon (DIC) ca. 2.5 Ga (Halevy & Bachan, 2017 ). Generally, calcium [Ca 2+ ] estimates in Archean seawater are analogous to the modern ocean concentration of 10 m m (Higgins et al., 2009 ; Holland, 1984 ), though other recent estimates suggest higher [Ca 2+ ] concentrations of 35–40 m m (Jones et al., 2015 ) or 50–100 m m (Halevy & Bachan, 2017 ). Research investigating DIR in simulated ancient marine conditions remains limited. Studies replicating Archean environmental conditions mainly focused on distinguishing the BIF Fe isotopic signals during Fe(III) oxide respiration in the presence of dissolved silica (Wu et al., 2009 ) and during respiration of Si–Fe(III) gels (Percak‐Dennet et al., 2011 ), as well as the implications of DIR of Si–Fe(III) gels on Si isotope fractionation (Reddy et al., 2016 ). However, these isotope‐centered studies did not characterize the secondary mineral products from DIR of Fe(III)‐Si gels in artificial Archean seawater, with the exception of reporting a singular, putative smectite mineral (Percak‐Dennet et al., 2011 ). Moreover, this tentatively identified three‐layered silicate phase is distinct from the two‐layer early greenalite minerals hosted in BIF cherts. Additionally, these prior Archean‐like DIR studies were completed at an elevated pH of 8.7 (Wu et al., 2009 ) or high DIC concentrations (≥30 m m ) (Percak‐Dennet et al., 2011 ; Reddy et al., 2016 ). In another set of Archean‐relevant experiments that characterized DIR‐promoted Ca/Fe‐carbonate mineral precipitation (Zeng & Tice, 2014 ), the experimental matrix lacked dissolved silica. Thus, a more extensive biomineral evaluation of DIR products in simulated Archean seawater, with high levels of silica, is essential to understanding whether DIR could be responsible for the genesis of early iron silicate minerals in BIFs. Despite the range of previously explored experimental conditions, no study has heretofore demonstrated the formation of authigenic Fe(II)‐iron silicates as a secondary mineral product of Fe(III) oxide respiration. In this study, we used batch experiments to examine and characterize the secondary phases that form after microbial iron reduction, as performed by a model iron‐reducing bacterium, with synthesized ferrihydrite as the electron acceptor and lactate as the electron donor across a range of Archean‐relevant chemistries. Specifically, we tested the impact of dissolved silica and divalent cations (calcium and magnesium) on the products from microbial reduction of ferrihydrite in conditions representative of Archean marine environments. Our exploration of DIR secondary mineral precipitates reveals how the interplay of iron, dissolved silica, and divalent cations can alter the products of microbial iron respiration, including the formation of potential precursor iron silicates.",
"discussion": "4 DISCUSSION Overall, these experiments testing microbial iron reduction in silica‐rich simulated Archean marine environments produced several secondary mineral products. Depending on whether silica, calcium, and/or magnesium was added to the experimental medium, we observed secondary ferric oxyhydroxides, such as goethite and lepidocrocite, and other precipitates including calcium carbonate polymorphs and proto‐iron silicates (Figure 6 ). We explore these products and hypothesize how they formed below. FIGURE 6 Summary of mineral formation pathways observed in our experimental simulations of Archean marine environments testing the effects of dissolved Si, Ca 2+ , and Mg 2+ on the secondary products of DIR. (a) Pathways of iron mineral formation showed a dependence on the dissolved ions, with the basal AAS media (without added Si, Ca 2+ , and Mg 2+ ) forming lepidocrocite, γ‐FeOOH, upon DIR (top arrow); the addition of Ca 2+ and/or Mg 2+ resulting in the formation of goethite, α‐FeOOH, after DIR (middle arrow); and the addition of Si with or without Ca 2+ and/or Mg 2+ generating proto‐iron silicates upon DIR (bottom arrow). (b) Calcium carbonate precipitation was also observed, with calcite forming in the presence of added Ca 2+ (top arrow) while calcite and aragonite both precipitated in the presence of added Ca 2+ when Mg 2+ was present, or when Fe 2+ was present as a result of DIR, and/or when organics were released by S. putrefaciens . 4.1 Secondary mineral products from DIR \n 4.1.1 Proto‐iron silicates We observed the formation of novel Fe(II)‐rich iron‐silica coprecipitates in our experimental incubations inoculated with S. putrefaciens . In tandem, our Ferrozine and silicomolybdate assay results suggest that silica sequesters ferrous iron into iron‐silica solid phases. In experiments without amended silica, biogenic Fe(II) aq increased in solution, while this Fe(II) predominantly partitioned into solid precipitates under siliceous conditions (Figure 1a,b , Table S3 ). In silica‐containing experiments, the average Fe(II)/FeT solid was approximately 25% higher than those in experiments without added dissolved silica (Table S3 ). The incorporation of biogenic ferrous iron and dissolved silica into newly formed iron‐silica coprecipitates could explain the depletion of Fe(II) aq and Si aq in the silica‐amended solutions. Silica‐amended experiments without added divalent cations or solely with added magnesium contained both negligible (<0.1 m m ) final Fe(II) aq and large decreases in Si aq , with approximately 40%–50% of the starting Si removed from solution (Figure 1c ; Table S2 ). These AAS and Mg‐amended experiments additionally had the highest Fe(II)/FeT solid ratios, as the magnesium‐amended solutions generated solids with 74% Fe(II) and the simple AAS solutions formed solids with 87% Fe(II) (Figure 1a,b , Table S3 ). However, the addition of Ca 2+ distinctly altered the distribution of ferrous iron in silica‐amended experiments, increasing [Fe(II) aq ] to 0.24–0.3 m m and lowering Fe(II)/FeT solid ratios (Figure 1a,b , Table S3 ), which we explore further below. In all experimental conditions without added silica, substantially more (0.1–0.5 m m ) Fe(II) aq accumulated in solution, consistent with the formation of Fe(II)‐rich siliceous phases in silica‐amended experiments. Indeed, in incubations amended with dissolved silica, micro‐ and nano‐scale imaging revealed the formation of novel iron‐silica coprecipitates. The elevated Fe(II)/FeT solid in silica‐amended precipitation products could point to the formation of mixed‐valence Fe‐oxides like magnetite, but we did not detect magnetite under these conditions by SEM‐ or TEM‐based analyses, Raman spectroscopy, XRD, or magnetic reaction. Instead, we observed the formation of abundant, light gray, wispy precipitates over time (Figure S6 ). In our electron microscopy imaging and analyses, extensive Fe–Si clusters of amorphous curling structures were present in all silica‐bearing bioreduction conditions (e.g., Figures 2j–l , 3m–o , 4f–h and 5l–n ) but absent in control experiments. These iron‐ and silica‐rich precipitates lack definitive crystal structure, including any resolvable lattice fringe in HRTEM imaging. Though SAED patterns suggest some iron‐silica precipitates contain a component of 2LFh, the coprecipitates are morphologically distinct from their antecedent ferrihydrite minerals, which are composed of pseudohexagonal crystallites (Figure S3 ). The iron‐silica phase is instead characterized by thin, curling, and folding layers (e.g., Figure 2j ), as well as high Fe, Si, and O (Table S5B–E ). Furthermore, the higher Fe(II)/FeT solid of the silica‐amended solids compared to bioreduced solids without added silica implies that these curled silica‐rich phases are Fe(II)‐rich. Therefore, we suggest that these structures are incipient Fe(II)‐rich proto‐silicates (Figure 6a ). We hypothesize that the precipitation of Fe(II)‐rich phases is promoted by cell‐mineral interactions resulting in biomass‐mineral aggregates that retain and adsorb local ferrous iron. The surface properties of S. putrefaciens have been observed to enhance aggregation of ferrihydrite particles and cells, forming packed structures (Glasauer et al., 2001 ; Zachara et al., 2002 ; Zegeye et al., 2010 ). These cell‐mineral aggregates create microenvironments with elevated pH and higher concentrations of metabolic byproducts, like Fe(II) and CO 3 \n 2− (Pallud et al., 2010 ; Roden & Urrutia, 2002 ; Zachara et al., 2002 ; Zegeye et al., 2010 ). For example, Zegeye et al. ( 2010 ) reported the highest retention of ferrous iron in dense microbe‐mineral aggregates of S. putrefaciens and γ‐FeOOH, and these Fe(II)‐rich microenvironments led to the formation of a more reduced product, green rust, rather than magnetite. Similarly, Pallud et al. ( 2010 ) observed that sand‐ferrihydrite aggregates inoculated with S. putrefaciens produced secondary minerals, including goethite/lepidocrocite and siderite, within the inner fractions of aggregates, while outer portions showed limited secondary iron phases. Indeed, the S. putrefaciens biofilm‐mineral aggregates that developed in our AAS‐only experiments—and more extensively in the silica‐amended condition—appeared to sequester Fe(II), as indicated by our EDS observations of high Fe in AAS biomass, very low Fe(II) aq in the final media solutions, and high Fe(II)/FeT solid in the precipitate‐biomass solid pellet (Figures 1a,b and 2a–c,g , Tables S3 and S5B ). We further posit that ferrihydrite‐adsorbed silica and ferrous iron promoted the precipitation of iron‐silica coprecipitates in these experiments. SAED patterns of the iron‐silica coprecipitates in our biotic experiments showed weak ferrihydrite reflections (Figures 3n and 4f ). Not only are ferric hydroxides highly adsorptive of Si, but ferrihydrite surfaces also act as templates for silica polymerization at pH 6–8.5 (Swedlund & Webster, 1999 ), potentially facilitating the formation of Fe–Si precipitates. Siliceous conditions could also enhance the aggregation of clustered microbial biomass, silica, and FeOOH minerals. Zachara et al. ( 2002 ) noted that, at circumneutral pH, the ferrihydrite surface favors the strong sorption of silicic acid and Fe 2+ , and upon Fe(III) respiration, these adsorbed ions are concentrated in the microenvironments within the aggregated ferrihydrite‐biomass clusters. These ions and reaction products may then react to form aqueous complexes and ultimately new solid phases. Therefore, biomass‐ferrihydrite aggregates in our experiments may have produced microenvironments with increased concentrations of Si and Fe 2+ , as well as templating ferrihydrite surfaces, stimulating the precipitation of iron‐silica phases. The inclusion of calcium in the experimental medium likely resulted in differences in these localized microenvironments. The presence of calcium appeared to produce much higher final Fe(II) aq and lower Fe(II)/FeT solid ratios, potentially associated with tendency of calcium to disaggregate microbe‐mineral aggregates and outcompete Fe 2+ adsorption on ferrihydrite surfaces. We observed the highest concentrations of Fe(II) aq in experiments without added silica but with 10 m m Ca 2+ , a trend that increased with the addition of magnesium (Table S3 ). Even in calcium‐bearing experiments amended with silica, we identified substantially higher Fe(II) aq compared to calcium‐lacking experiments (Figure 1a,b , Table S3 ). A previous study found that multicellular aggregates of Shewanella oneidensis MR1 cultures did not form or disaggregated in the presence of even low Ca levels (0.68 m m Ca 2+ ) under anaerobic growth conditions (McLean et al., 2008 ). If calcium similarly limited cell‐ferrihydrite aggregation in our anaerobic Fe(III) respiration experiments, this would likely allow more bio‐generated Fe(II) to escape into solution while increasing the exposed surface area of ferrihydrite particles to bioreduction and Si adsorption. Additionally, Ca 2+ is used to dislodge weakly adsorbed divalent ions, including Fe 2+ , from ferrihydrite surfaces (Cismasu et al., 2013 ; Kinsela et al., 2016 ). In calcium‐amended experiments, calcium would thus be expected to decrease the retention of Fe(II) in ferrihydrite‐associated complexes, as well as inhibit the formation of microbe‐mineral aggregates. Previous abiotic experiments formed proto‐Fe(II) silicates in Archean‐like conditions when pH was elevated to ≥7.5 (Hinz et al., 2021 ; Tosca et al., 2016 ). Microbial iron respiration increases alkalinity and drives pH to these higher values, exceeding pH 7 (e.g., Bergmann et al., 2013 ). While pH values of ≥7.5 are higher than the predicted pH of the Archean ocean, DIR increased the bulk pH in our experiments (Table S1 ) and potentially created even higher‐pH microenvironments in microbe‐mineral aggregates. Thus, the increasing shift in pH could have made the precipitation of proto‐iron silicates more favorable in our experimental set‐up. In the presence of Ca 2+ , the experiments concluded with lower pH (<7.5), likely due to the formation of calcium carbonate minerals. The lower pH and extensive precipitation of calcium carbonate minerals may have contributed to the heightened Fe(II) aq and Si aq of the combined Ca 2+ ‐ and Mg 2+ ‐amended experiments, potentially limiting–but not preventing–the formation of proto‐iron silicates. Because the iron‐silica curling precipitates were absent in silica‐amended control products, we further propose that Fe(II) was essential to the formation of these phases. Our abiotic control experiments solely contained ferrihydrite and calcium carbonate phases after over the >4.5 month incubations (Figures S3 and S4 ). Conversely, experiments with S. putrefaciens performing Fe(III) respiration and producing Fe(II) in the presence of dissolved silica invariably resulted in proto‐iron silicate phases, regardless of whether Ca 2+ or Mg 2+ were in the solution. Our findings support the Harder ( 1978 ) hypothesis that Fe(II) plays a critical role in the formation of proto‐iron silicates under environmentally relevant conditions (i.e., <2 m m Fe and Si). With iron introduced solely as ferrihydrite, we found that the biogenic production of Fe(II) from DIR was required to produce these precursor iron silicates. 4.1.2 Carbonates The most evident effect of adding calcium was in the production of calcium carbonate polymorphs (Figure 6b ), which formed in both S. putrefaciens incubations and control experiments (see Figure S4 ). Both aragonite and calcite were slightly supersaturated in the starting solutions, with the initial saturation index (SI) of each phase beginning at SI Calcite of 0.16–0.28 and SI Aragonite of 0.017–0.13 in experiments with either Ca 2+ ‐only or both Ca 2+ and Mg 2+ (Table S4 ). Over the course of the experiments, as pH increased either due to the DIR reaction, CO 2 exsolution in the N 2 headspace, and/or the higher‐pH buffering of the sodium metasilicate Si source, these SI values would have increased. Indeed, at the final pH of 7.5–7.6 in the post‐DIR experiments, the SI Calcite increased to ~0.6–0.8 while SI Aragonite increases to ~0.4–0.65, again with either Ca‐bearing chemistry. Our observation of aragonite, despite being thermodynamically unfavorable relative to calcite, is likely due to the favorable precipitation of aragonite in the presence of Fe(II) and Mg(II) (Berner, 1975 ; Mucci & Morse, 1982 ; Zeller & Wray, 1956 ). Fe(II) inhibits normal calcite spiral growth (Di Lorenzo et al., 2017 ) and the incorporation of Mg(II) into the calcite lattice inhibits crystal growth either by increasing surface energy (Sun, Jayaraman, et al., 2015 ) or increasing calcite solubility (Davis et al., 2000 ). However, we observed both aragonite and calcite polymorphs in S. putrefaciens incubations amended with calcium, under conditions both with and without magnesium. This dual finding signals the intricacies of polymorph formation in biological solutions. Although the iron‐reducing environment and supplemented magnesium would suggest aragonite precipitation, the [Mg]/[Ca] ratio in our experiments (initially at 1 where both present) remained lower than the critical limit established for aragonite selection (i.e., [Mg]/[Ca] ≥ 2; (Sun, Jayaraman, et al., 2015 )), potentially enabling the formation of both polymorphs across local heterogeneity. In calcium‐amended conditions with S. putrefaciens , initial calcite precipitation was likely overtaken by aragonite mineralization due to the increasing ferrous iron inhibiting calcite (Dromgoole & Walter, 1990 ; Meyer, 1984 ) and/or the development of an EPS matrix (Lyu et al., 2020 ). In contrast, we identified only calcite and no aragonite in Raman scans of minerals in Ca‐amended abiotic controls (Figure S4 ), confirming the role of cellular organics, biogenic Fe(II), and Mg 2+ in promoting aragonite precipitation under the other experimental conditions. Dissolved silica could also play a role in carbonate crystal growth or polymorphism in the bioreduction incubations with S. putrefaciens . Kellermeier et al. ( 2013 ) observed both the stabilizing effects of surface‐adsorbed silica on metastable CaCO 3 polymorphs and silica‐enhanced calcite growth at low temperatures. Unlike silica‐free experiments that contained large, more singular calcium carbonate crystals, we frequently observed smaller individual needles and blades fusing into larger mineral aggregates in Si‐amended experiments (e.g., Figure 5f,h ); however, the exact role of dissolved silica in carbonate morphology is difficult to quantify in these limited experiments. Iron carbonates, including siderite, were not detected in our experimental conditions, which we attribute to the low concentrations of Fe(II) aq and the large divergence in the growth kinetics between calcite and siderite. Our experiments started with 1.5 m m Fe(III) in ferrihydrite and 10 m m HCO 3 \n − , evolving to a final measured solution chemistry with a ~1:20 ratio of aqueous Fe(II) aq to calcium or magnesium (where [Fe(II) aq ] ≤ ~500 μ m and initial [Ca] and [Mg] was 10 m m ). Siderite saturation was maximized in DIR experiments with combined Ca and Mg but without Si, where Fe(II) aq reached 500 μ m at the final pH of 7.5; under these conditions, SI siderite = 1.7 while SI calcite = 0.7. At environmental conditions of 25°C and 1 atm, when both siderite and calcite are supersaturated, the growth rate of calcite is approximately seven orders of magnitude faster than that of siderite, a behavior likely explained by the smaller ionic radius of Fe 2+ and requisite higher activation energy to form siderite (Jiang & Tosca, 2020 ). More importantly, siderite precipitation proceeds through an amorphous ferrous carbonate [AFC: Fe(CO 3 ) 0.5 OH] precursor, which requires a higher supersaturation threshold than obtained by our experiments. The threshold for AFC precipitation is an SI siderite > ~2.8 (Jiang & Tosca, 2019 ), and our final solution chemistry remained below this minimum. In contrast, the supersaturation required for aragonite and calcite precipitation (Busenberg & Plummer, 1986 ) were surpassed in all our Ca‐bearing solutions. Therefore, based on kinetic considerations, we would expect calcium carbonates to precipitate instead of iron‐rich carbonates in this experimental set‐up. Our experimental chemistry was distinct from a previous set of experiments simulating BIF bioreduction where Shewanella oneidensis MR‐1 cultures reduced 5 m m ferric hydroxides to >3 m m Fe(II) aq in solutions with 0–10 m m Ca 2+ and 0–100 m m Mg 2+ (Zeng & Tice, 2014 ). In contrast to our findings, Zeng and Tice observed calcium‐rich siderite or ankerite precipitation in experiments supplemented with calcium, and they did not report calcium carbonate mineralization. We hypothesize that the higher bioproduced aqueous ferrous iron concentrations enabled AFC (and siderite) to reach sufficient supersaturation for precipitation in the Zeng and Tice experiments, while the lower bioproduced Fe(II) concentrations in our experiments precluded the formation of these iron‐rich carbonate minerals. 4.1.3 Goethite We identified secondary Fe(III) phases including goethite (α‐FeOOH) in silica‐free S. putrefaciens incubations, confirming Ferrozine results that indicated there was only partial reduction of the starting ferrihydrite substrate. This partial reduction occurred despite the excess electron donor (lactate) that was provided for S. putrefaciens . Secondary goethite mineralization arising from the microbial respiration of ferrihydrite is well documented (Bae & Lee, 2013 ; Fredrickson et al., 2001 ; Fredrickson et al., 2003 ; R. Han et al., 2018 ; Hansel et al., 2003 ; Pallud et al., 2010 ). Hansel et al. ( 2003 ) found that this secondary α‐FeOOH mineralizes through dissolution and reprecipitation transformations of ferrihydrite during reactions with Fe(II), where sorbed Fe(II) passivates the ferrihydrite surface at low initial Fe(II) concentrations (<0.3 m m ), inducing the mineralization of goethite instead of magnetite. After our 19 week incubations, the bioreduced solutions without silica had a mean [Fe(II)] of 0.34 m m (Figure 1 ; Table S3 ). Since the initial Fe(II) concentration was probably similar to control experiments (0.03 m m ), increasing over time during bioreduction, goethite plausibly formed through this Fe(II)‐induced reprecipitation pathway. In addition, we found further evidence of remnant Fe(III) with our tentative identifications of 6‐line ferrihydrite and hematite alongside goethite in silica‐free and magnesium‐supplemented experiments (Figure 4c ), suggesting there was an increase in particle size and crystallinity of the starting ferrihydrite substrate during microbial Fe(III) respiration. In contrast, experiments augmented with silica did not form goethite. In fact, silica may inhibit goethite formation as it blocks the sorption of Fe(II) at reactive ferrihydrite surface sites (Jones et al., 2009 ; Schwertmann & Thalmann, 1976 ), subsequently stabilizing the ferrihydrite and hindering its transformation to other Fe‐oxides (Anderson & Benjamin, 1985 ; Lee & Xu, 2019 ). As previously discussed, our results also suggest that released Fe(II) bonded with silica to form iron‐silica coprecipitates. It is possible that the ‘captured’ Fe(II) in these siliceous coprecipitates limits Fe(II) interaction with remnant ferrihydrite. Regardless of the mechanism, the absence of goethite or detection of other secondary iron oxides in our silica‐rich incubations suggests that the dissolved silica component prevented Fe(II)‐induced solid state transformation of ferrihydrite to other Fe oxides (e.g., goethite, magnetite, hematite). Instead, our measurements showed siliceous experiments retained 13%–34% Fe(III) in solid products (Table S3 ), mainly as remnant two‐line ferrihydrite that formed surfaces for iron‐silica coprecipitation (e.g., Figure 4f ). 4.2 Implications for Precambrian BIFs \n Microbial Fe(III) respiration likely influenced BIF deposition and/or the early stages of BIF diagenesis (Konhauser et al., 2005 ; Walker, 1984 ). One proposed mechanism for BIF formation argues for the deposition of primary iron oxides (e.g., ferrihydrite) (Beukes et al., 2008 ; Konhauser et al., 2017 ; LaBerge, 1964 ) that transformed into iron silicates, carbonates, magnetite, or other phases during iron bioreduction hosted in ancient sediments (Fischer & Knoll, 2009 ). We tested this hypothesis using laboratory simulations of S. putrefaciens suspensions in artificial Archean seawater to examine microbial iron reduction as a viable pathway for ancient iron silicate formation. The carbonate minerals produced in our DIR incubations and control solutions, however, were unlike those recorded in early BIF assemblages. We observed pervasive calcium carbonate precipitation, and no evidence of iron carbonates—siderite or ankerite—that are ubiquitous in the BIF record (Klein, 2005 ; Trendall, 2002 ) and often attributed to dissimilatory iron reduction (Becker & Clayton, 1972 ; Heimann et al., 2010 ; Johnson et al., 2013 ; Mozley & Carothers, 1992 ; Walker, 1984 ). This CaCO 3 precipitation likely dominated due to inhibited iron carbonate growth kinetics and a low experimental Fe/Ca ratio with low total [Fe], conditions disparate from ancient sedimentary packages that hosted high concentrations of deposited iron precipitates. Hence, our experimental results are potentially more suitable for examining the mineral products from microbial iron respiration in an environment mimicking a water column containing dilute Fe(III) oxides. Furthermore, our observations of proliferate calcium carbonate forming in simulated Archean seawater conditions may imply that the deeper Archean ocean was actually less saturated with respect to calcium carbonate than our experimental conditions and/or that BIF depositional environments can be constrained to below the carbonate compensation depth (CCD). Finally, our results demonstrate the precipitation of potential Fe(II)‐rich proto‐silicates from the microbial respiration of Fe(III) oxides. Although the experimental proto‐silicates lack the well‐layered structure of the greenalite particles hosted by Archean BIF cherts, the morphology of these phases is remarkably similar to 25°C, pH 7.5, proto‐Fe(II,III) silicates that developed a coherent layered structure reminiscent of greenalite upon hydrothermal aging (Hinz et al., 2021 ). These comparable observations imply that the DIR‐produced proto‐silicate phases would evolve into more crystalline silicate minerals with simulated higher temperature diagenesis. However, even with a surplus of electron donor (lactate) and conditions apt for extensive Fe(III) respiration, we still observed substantial Fe(III) remaining in solids (Figure 1 ; Table S2 ). Given the Fe(III) content of the experimental solids, our findings may suggest that initial iron oxides would be preserved in the BIF record. Indeed, early nanoinclusions of hematite in 2.5 Ga Hamersley Group BIFs reported by Sun, Konhauser, et al. ( 2015 ) and Li et al. ( 2013 ) could be relicts of a primary ferric oxyhydroxide. Notably, the average redox state of iron in BIFs is +2.4 (Klein & Beukes, 1992 ) and our most relevant experiments with calcium, magnesium, and silica produced solids with a final iron redox of 66% Fe(II)/FeT solid or +2.3. The proximity of these experimental values with the BIF average composition may support the hypothesis that the partial reduction of initial ferric iron precipitates transpired during the genesis of BIF assemblages. In contrast, Rasmussen et al. ( 2016 , 2021 ) argue that the nanoscale ferric oxide particles are secondary to primary Fe(II) silicates (i.e., greenalite) based on textural relationships and paleomagnetic evidence that indicate iron oxides in BIFs formed during later stage fluid–rock interactions, metamorphism, and/or oxidative weathering (Muhling & Rasmussen, 2020 ; Rasmussen et al., 2021 ; Rasmussen et al., 2013 , 2019 , Rasmussen et al., 2016 ). We were unable to observe the extent of bioreduction necessary for the sole deposition of iron silicate mud after water column or sedimentary DIR, potentially due to restraints imposed by our nutrient‐depleted medium (e.g. lacking phosphate and possible consumption of other essential trace elements) or the time limitation of our 19‐week bioreduction incubations. However, investigating microbial Fe(III) respiration on more substantial timescales or with continuous chemostat cultures could possibly result in the complete reduction of ferrihydrite and singular production of Fe(II)‐rich proto‐silicates. While it is impossible to truly replicate the Archean ocean or ancient sedimentary porewaters in modern laboratory reconstructions, our experiments demonstrate that incipient iron silicates may have formed from microbial respiration of early ferric substrates, suggesting a possible alternative genesis for early BIF iron silicates."
} | 9,554 |
38711153 | PMC11075230 | pmc | 1,865 | {
"abstract": "Background Lignocellulosic biomass as feedstock has a huge potential for biochemical production. Still, efficient utilization of hydrolysates derived from lignocellulose is challenged by their complex and heterogeneous composition and the presence of inhibitory compounds, such as furan aldehydes. Using microbial consortia where two specialized microbes complement each other could serve as a potential approach to improve the efficiency of lignocellulosic biomass upgrading. Results This study describes the simultaneous inhibitor detoxification and production of lactic acid and wax esters from a synthetic lignocellulosic hydrolysate by a defined coculture of engineered Saccharomyces cerevisiae and Acinetobacter baylyi ADP1. A. baylyi ADP1 showed efficient bioconversion of furan aldehydes present in the hydrolysate, namely furfural and 5-hydroxymethylfurfural, and did not compete for substrates with S. cerevisiae , highlighting its potential as a coculture partner. Furthermore, the remaining carbon sources and byproducts of S. cerevisiae were directed to wax ester production by A. baylyi ADP1. The lactic acid productivity of S. cerevisiae was improved approximately 1.5-fold (to 0.41 ± 0.08 g/L/h) in the coculture with A. baylyi ADP1, compared to a monoculture of S. cerevisiae . Conclusion The coculture of yeast and bacterium was shown to improve the consumption of lignocellulosic substrates and the productivity of lactic acid from a synthetic lignocellulosic hydrolysate. The high detoxification capacity and the ability to produce high-value products by A. baylyi ADP1 demonstrates the strain to be a potential candidate for coculture to increase production efficiency and economics of S. cerevisiae fermentations. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-024-02510-8.",
"conclusion": "Conclusion In this study, our aim was to increase the efficiency and feasibility of lactic acid production by S. cerevisiae using a lignocellulosic feedstock. Toward that end, we established a synthetic microbial consortium employing a strictly aerobic soil bacterium ADP1 with S. cerevisiae . ADP1 exhibited high bioconversion rate on furan derivatives in the lignocellulosic hydrolysate, which supported S. cerevisiae ’s growth and lactic acid production. With the synthetic microbial consortium, the productivity of lactic acid was increased 1.5-fold. No competition for sugars between S. cerevisiae and ADP1 occurred in the coculture, allowing for sustained lactic acid titer and yield. In addition, ADP1 consumed residual carbon in the coculture and directed it to high-value lipid (WE) production. The work demonstrates the potential of cocultures for the production of biochemicals from complex and challenging feedstock, such as lignocellulose hydrolysates.",
"discussion": "Discussion Using lignocellulosic biomass as feedstock in bioprocesses is economically and environmentally justified. However, the hydrolysates derived from biomass contain a mixture of different substrates and toxic compounds that can result in growth inhibition of the host microorganism, lowered productivities and yields [ 52 , 53 ], and sometimes accumulation of unwanted byproducts [ 54 ]. For the lignocellulose-based processes to be successful, comprehensive utilization of carbon and the ability to produce a broad range of relevant biochemicals is crucial. Many of the issues could potentially be addressed using a coculture of optimally selected and engineered strains. In this work, our goal was to establish a coculture for efficient production of lactic acid from a synthetic lignocellulose hydrolysate (SLH) containing xylose, hexoses, aliphatic acids, furan derivatives, and phenolic acids, thus mimicking lignocellulose hydrolysates made from spruce. Lignocellulosic biomass derived from forestry, such as spruce residues, are widely abundant, especially in Northern Europe [ 55 ]. Therefore, valorization of lignocellulosic biomass derived from wood has great potential in countries with extensive forest coverage and well-developed forest industries. To further improve the carbon recovery and process feasibility, we aimed to direct the residual carbon and potential byproducts of S. cerevisiae toward the production of a high-value product, WEs. To allow efficient carbon recovery in the process, we selected the yeast S. cerevisiae and the bacterium A. baylyi ADP1 as the host organisms for the coculture. S. cerevisiae can efficiently produce lactic acid from major lignocellulose components [ 56 , 57 ], while ADP1 can metabolize a wide range of organic acids and phenolic compounds that are known to be harmful for S. cerevisiae [ 52 ], and direct the carbon to secondary products, namely high-value lipids, WEs. While lactic acid can be separated from the cultivation broth using several reported methods for its purification from the fermentation medium [ 58 ], WEs accumulate intracellularly and can be extracted from harvested cells, potentially providing an advantage for product recovery. To engineer ADP1 to be suitable for the coculture, we first characterized wild-type ADP1 for its tolerance to furfural and HMF, as well as its ability to consume lactate. Previously, ADP1 has been shown to grow on acetate in minimal medium with 1 g/L furfural [ 22 ]. We confirmed that wild-type ADP1 can tolerate up to 2 g/L furfural and 1.5 g/L HMF when glucose was provided as a carbon source (Additional file 1 : Fig. S1). We also demonstrated that ADP1 was able to grow well in SLH and exhibited sufficient tolerance to the mixture of inhibitors present. We further confirmed that wild-type ADP1 naturally utilizes l - and d -lactate, which aligns with previous research [ 59 ]. As the starting point for engineering, we selected a previously described strain ASA507 [ 39 ]. The strain ASA507 is derived from a genetically stable, transposon-free ADP1-ISx [ 37 ] and has exceptionally high tolerance against p -coumarate and ferulate [ 39 ], the key aromatic monomers of the lignin fraction of lignocellulose. We first removed the complete lactate operon ( lldPRD , dld ; ACIAD 0106-0109) encoding a lactate permease, a l -lactate dehydrogenase operon regulator, a l -lactate dehydrogenase, and a d -lactate dehydrogenase. Furthermore, to increase the production of WEs, we overexpressed a fatty acyl-CoA reductase acr1 (ACIAD 3383), which has been previously shown to improve WE production in ADP1 [ 45 ]. We also showed that lactate concentrations relevant to what could be expected from the fermentation of the SLH containing approximately 38 g/L of sugars, e.g., up to 18 g/L lactic acid did not inhibit the growth of ASA714 (Additional file 1 : Fig. S5). Lactic acid at a yield of 0.26 g/g was recently reported for S. cerevisiae during fermentation of corn stover and wheat straw hydrolysate [ 60 ]. ASA714 consumed all acetate, ferulate, and p -coumarate in the SLH medium, and as a result of its incomplete sugar metabolism [ 21 ], no xylose, galactose, or mannose was consumed (data not shown). Thus, glucose is the only sugar found in the SLH medium that ADP1 potentially competes for with S. cerevisiae . Unexpectedly, ASA714 consumed only about 0.7 g/L of glucose in the SLH medium; the same was observed for wild-type ADP1 grown in minimal medium supplemented with 10 g/L glucose and 2.6 g/L acetate (data not shown), demonstrating that the inefficiency of glucose utilization was not a result of strain engineering but a phenotypic feature of ADP1. In fact, ADP1 did not start utilizing glucose even in 3 days after acetate depletion. On the contrary, in a previous study, ADP1 was shown to co-utilize acetate and glucose when grown in minimal medium containing less substrates, e.g., 2 g/L of both acetate and glucose [ 61 ]. Moreover, ADP1 has been shown to utilize gluconate concurrently with acetate, even at substrate concentrations as high as 100 mM (5.9 g/L acetate and 19.5 g/L gluconate) [ 28 ]. Gluconate is metabolized by the same modified Entner–Doudoroff pathway as glucose [ 21 ], so it is not clear why the presence of acetate has such an inhibitory effect on glucose but not on gluconate utilization. Although interesting, the determination of the reason behind the reduced ability to consume glucose of ADP1 in the studied conditions remains to be investigated in the future. In the context of the studied coculture, however, the hampered glucose metabolism in the presence of acetate was found to be beneficial, as ADP1 did not compete for sugars with S. cerevisiae . Taken together, the characteristics of ASA714 in terms of tolerance and carbon utilization were found to be well-suited for the coculture. The lactic-acid-producing strain LX3 was constructed based on the rational design of genetic perturbations to increase the activity of the lactic acid metabolic pathway with introduction of multiple copies of the LDH gene and simultaneously deleting CYB2 , ERF2 , and GPD1 in the xylose-fermenting strain, XXX [ 38 ]. Deletion of CYB2 , encoding a lactate cytochrome-c oxidoreductase, has been known to prevent assimilation of lactate at low pH [ 62 ], whereas deletion of ERF2 , which encodes a palmitoyl transferase, enhanced acid tolerance of S. cerevisiae [ 63 ]. GPD1 , encoding a glycerol-3-phosphate dehydrogenase, was deleted to increase the lactic acid metabolic pathway flux by reducing the activity of the glycerol pathway [ 63 ]. The advantage of lactic acid fermentation from xylose was described in a previous study: when xylose was used as substrate, 28-times less ethanol, the major fermentation product in S. cerevisiae , was produced compared to when glucose was used as the substrate [ 64 ]. This was also observed in LX3 fermentation. The LX3 strain produced 5.9-times less ethanol from xylose than from glucose. The lactic acid production of LX3 was 4 times higher from xylose than from glucose. Turner et al. [ 64 ] proposed that the slower rate of xylose utilization, compared to glucose, prevents the accumulation of an excessive amount of intracellular pyruvate, which typically leads to ethanol production. Instead, this slower utilization rate results in the conversion of pyruvate to lactic acid [ 64 ]. These characteristics make the LX3 strain not only an ideal host for carbon utilization in lignocellulosic hydrolysates, but also an efficient producer of lactic acid. The aerobic biotransformation of furfural and HMF to their corresponding alcohols (and to some extent, furfural to furoic acid) by S. cerevisiae has been previously reported [ 65 , 66 ]. Despite the ability of S. cerevisiae to detoxify furan derivatives, the presence of these compounds has been shown to interrupt the performance of S. cerevisiae already in the early stages of fermentations [ 67 ]. In addition, furfural has been shown to negatively affect glycolysis [ 68 ] and both HMF and furfural can cause a prolonged lag phase of S. cerevisiae [ 66 , 69 , 70 ]. Besides inhibiting cell growth, previous research has demonstrated that furfural and HMF can directly negatively impact bioproduction, such as ethanol production [ 71 ]. While acetic acid is toxic to S. cerevisiae on its own [ 72 ], it has also been reported to increase the toxicity of furfural to S. cerevisiae [ 73 ] . Species of Acinetobacter genus have been also described to metabolize furfural and HMF [ 22 – 24 ]. To investigate whether coculturing the strains would accelerate the detoxification and reduce the inhibitory effect of furan-derived compounds on S. cerevisiae , we first studied the bioconversion capacity of both cocultivation partners individually. LX3 converted furfural and HMF in 12 h and 17 h (conversion rate of furfural 0.07 g/L/h and HMF 0.02 g/L/h). Previously, S. cerevisiae was reported to convert furfural and HMF with a higher conversion rate (furfural 0.14 g/L/h and HMF 0.27 g/L/h; estimated from the figures) [ 74 ] compared to what we measure for LX3; the conversion rate is likely to be strain and condition dependent. ASA714 converted all furfural and HMF within 6 h. Previously, ADP1 was shown to convert 1 g/L furfural (in a minimal medium with acetate as the sole carbon source) [ 22 ] slightly faster than in our experiments. This is likely due to the complex nature and combined inhibitors in the SLH. Taken together, the detoxification studies demonstrated that ADP1 has potential to serve as an efficient detoxifier of lignocellulosic hydrolysates and can, thus, potentially support S. cerevisiae growth and production process. For a coculture to strive, not only the choice of microorganisms but also the inoculation ratio needs to be considered [ 75 ]. Therefore, prior conducting the coculture of ASA714 and LX3 with SLH medium, the effect of initial inoculation ratio was investigated. The improved coculture performance seen with increasing the proportion of ADP1 (to 2:1) can be explained by ADP1’s more prominent role in detoxification, given its high furan derivative bioconversion rate. A similar trend was previously demonstrated in a coculture with a filamentous fungus and a yeast: a higher inoculation size of Fusarium striatum that converted furfural and HMF in a lignocellulosic hydrolysate resulted in more efficient production of ethanol by S. cerevisiae in their coculture [ 76 ]. Based on the initial strain characterizations, it is likely that LX3 consumed most of the sugars also in the coculture. Hexoses probably contributed mainly to the biomass accumulation of LX3 while the xylose was primarily converted to lactic acid. Notably, sugar consumption was faster in the coculture compared to the monoculture of LX3 (Fig. 4 ). Ethanol production was observed both in monoculture and cocultures (Fig. 5 B). We observed that ethanol was depleted faster in the coculture than in the monoculture, probably due to the metabolic activity of ASA714; ethanol is one of the most preferred carbon sources for ADP1 [ 61 ]. ADP1 converts ethanol to acetaldehyde, followed by conversion of acetaldehyde to acetate. Consequently, acetate was first generated from ethanol before it was reassimilated by ADP1 (Additional file 1 : Fig. S12). Even though the acetate concentration measured by HPLC did not decrease in the coculture, the acetate produced from ethanol (that was depleted) was probably consumed simultaneously as it was produced by ASA714. While some acetate remained in the coculture, the acetate concentration was lower compared to that of the monoculture. Thus, it can be concluded that both substrates and byproducts (acetate and ethanol) of S. cerevisiae could be more effectively utilized in the coculture. The monoculture and coculture experiments were repeated four times (for the data of individual experiments, see Additional file 1 : Figs. S10 and S11). There were batch to batch variations in the consumption of acetate and phenolic compounds and the production of WEs among the repetitions. In the 1st and 4th experiment, all acetate was depleted by the 28-h timepoint, followed by the consumption of the phenolic compounds. While it has been previously reported that S. cerevisiae has the ability to convert phenolic compounds into less toxic derivatives, consumption of ferulate and p -coumarate during the mono-cultivations of LX3 did not occur. Therefore, we assume that ASA714 was solely responsible for the consumption of the phenolic compounds via the 3-ketoadipate pathway, as was demonstrated earlier [ 19 , 20 ]. In the 2nd and 3rd repletion of the cocultures no phenolic acids were consumed. We reason that this may be due to the higher concentration of acetate in these cultures, which serves as a carbon catabolite repressor for the aromatics degradation pathway [ 20 ]. Furthermore, the differences in carbon consumption (Additional file 1 : Fig. S11) and WE production profiles among the replicate cultures indicate that the population ratios vary in time across the batches. This could be expected as we did not have any means for population control during the cocultures. A small change in the performance of either coculture partner at the beginning of the culture due to, for instance, decreased viability or fitness could affect the population dynamics during the entire cultivation. The interaction in our bacterium-yeast consortium of ADP1 and S. cerevisiae can be classified as facultative mutualism, in which both strains are able to independently grow in the SLH medium. ASA714 was able to successfully direct some of the carbon, likely from acetate and phenolic acids, toward WE production, despite the process conditions in the coculture being far from optimal for lipid accumulation. In contrast, no WEs were detected from LX3 monocultures, which is consistent with the previous studies showing that S. cerevisiae does not naturally produce WEs [ 77 ]. ADP1 natively produces WEs mainly in conditions with high carbon–nitrogen ratio [ 78 ], whereas the process conditions used in this study exhibited a relatively low carbon–nitrogen ratio. The high nitrogen content directs substrates toward biomass production and more importantly, can drive the cells to rapidly consume the accumulated WEs as the carbon source [ 50 ], which likely also explains why some WEs were lost over time (24.4 mg/L at 28 h and 11.9 mg/L at 48 h). In addition, based on the carbon consumption data and visual observation of produced mScarlet fluorescence protein (this dyed the cells of ASA714 to pink), the number of ASA714 cells was likely not very high in the cocultures, emphasizing the difficulty to achieve high WEs titers. We have previously described an autonomous switch for WE accumulation in low carbon conditions [ 26 ], as well as improved WE production in nitrogen-rich conditions [ 44 ] and by partitioning metabolism for strict control of biomass and WE production [ 28 ]. Adapting these approaches in addition to overexpression of the WE synthesis pathway could have further enhanced the WE production metrics, but on the other hand, may have potentially compromised the robustness of the production strain. The greatest benefit of the coculture was observed in the production of lactic acid. The productivity of lactic acid was improved 1.5-fold (to 0.41 ± 0.08 g/L/h) compared to that of the monoculture. This report presents the first instance of using coculture to improve production of lactic acid by S. cerevisiae in a lignocellulosic feedstock. Based on the cell growth and sugar consumption during the first 28 h of the coculture, LX3 grew faster and to a higher biomass concentration in the coculture, which we assume was due to the rapid detoxification of the inhibitors in the early stages of fermentation by ASA714. Despite sharing some of the resources in the culture, such as oxygen and nitrogen source, it is worth noting that the titer and yield of lactic acid were not negatively affected by potential coculture interactions or competition for resources. Taken together, the improved cell growth, carbon recovery, and productivity demonstrate the compatibility of S. cerevisiae and A. baylyi ADP1 and the potential of cocultures for upgrading complex and heterogeneous substrates."
} | 4,814 |
27999590 | PMC5143786 | pmc | 1,866 | {
"abstract": "Ant Colony Optimization (ACO) has been successfully applied to solve a wide range of combinatorial optimization problems such as minimum spanning tree, traveling salesman problem, and quadratic assignment problem. Basic ACO has drawbacks of trapping into local minimum and low convergence rate. Simulated annealing (SA) and mutation operator have the jumping ability and global convergence; and local search has the ability to speed up the convergence. Therefore, this paper proposed a hybrid ACO algorithm integrating the advantages of ACO, SA, mutation operator, and local search procedure to solve the traveling salesman problem. The core of algorithm is based on the ACO. SA and mutation operator were used to increase the ants population diversity from time to time and the local search was used to exploit the current search area efficiently. The comparative experiments, using 24 TSP instances from TSPLIB, show that the proposed algorithm outperformed some well-known algorithms in the literature in terms of solution quality.",
"conclusion": "5. Conclusion In this paper, a new hybridized metaheuristic algorithm, called annealing elitist ant system with mutation operator for traveling salesman problem, has been introduced. Experiments were conducted using 24 data sets obtained from the TSPLIB and the experimental findings of the proposed algorithm were compared with different state-of-the-art algorithms. The results illustrate that the proposed algorithm outperforms other algorithms and has smaller percentage deviations in comparison to Chen and Chien [ 27 ], Wang et al. [ 41 ], Yousefikhoshbakht et al. [ 32 ], and Mahi et al. [ 31 ] algorithms. For future work, the proposed hybrid algorithm can be enhanced by using adaptive parameter on-the-fly or tuning using fuzzy logic. In addition, further evaluation of the performance of the proposed hybrid algorithm can be done using asymmetric TSP. For generalization, the proposed algorithm can be applied for different optimization problems. Another enhancement can be introduced by implementing a parallel version of this algorithm.",
"introduction": "1. Introduction One of the most popular combinatorial optimization problems is the traveling salesman problem (TSP) [ 1 ]. Given a set of cities, a salesman attempts to find the shortest or at least near to the shortest tour by visiting each city only once and turning back to the starting city. TSP is a representative of variety of combinatorial problems. It has been studied for the last 40 years. It has many real world applications such as the movement of people, postal delivery, school bus routes, garbage collection, design of hardware devices and radio electronic systems, machine scheduling, integrated circuits, and computer networks [ 2 – 4 ]. Metaheuristic algorithms are formally defined as algorithms that inspired by nature and biological behaviors. They produce high-quality solutions by applying a robust iterative generation process for exploring and exploiting the search space efficiently and effectively. Recently, metaheuristic algorithms seem to be a hot and promising research areas [ 5 ]. They can be applied to find near-optimal solutions in a reasonable time for different combinatorial optimization problems [ 6 ]. Metaheuristic algorithms such as genetic algorithms (GAs) [ 7 ], particle swarm optimization (PSO) [ 8 ], tabu search (TS) [ 7 ], simulated annealing (SA) [ 9 ], and ant colony optimizations (ACO) [ 10 ] are widely used for solving the TSP. Ant colony optimization proposed by Dorigo et al. in 1996 [ 10 ] simulates the intelligent behavior of real ants seeking for the food in nature. It has been successfully applied to solve many optimization problems such as TSP [ 10 ], quadratic assignment [ 11 ], job-shop scheduling [ 12 ], and load balancing in telecommunications networks [ 13 ]. In applying standalone metaheuristic algorithms, there is possibility of losing the diversity of the population through premature convergence and thus the algorithm gets stuck in local optima. Therefore, maintaining the diversity and making tradeoff between diversification and intensification by combining two or more algorithms to produce high-quality solutions and speed up the execution time is indispensable [ 14 ]. For hybrid ACO, the earliest study was conducted by McKendall and Shang [ 15 ]. They presented a hybrid ant system algorithm to solve dynamic facility layout problem. Another research was a hybrid ant system algorithm for solving TSP in which ant colony, genetic algorithm, and simulated annealing are hybridized [ 16 ]. For the hybrid ant colony system (ACS), many researches were conducted including the work by Huang and Liao [ 17 ], Yoshikawa and Otani [ 18 ], Xing et al. [ 19 ], Liao et al. [ 20 ], Lin et al. [ 21 ], Hajipour et al. [ 22 ], and Min et al. [ 16 ]. The research by Katagiri et al. [ 23 ] is an example for hybrid MAX-MIN Ant System. To solve TSP problems, several hybrid ACO variants with other metaheuristic algorithms such as SA, PSO, ACO, ABC, and ANN were proposed. Bontoux and Feillet [ 24 ] proposed a hybrid ACO algorithm with local search procedures to solve TSP. Tsai et al. [ 25 ] presented a hybrid ACO called ACOMAC algorithm for solving TSP. Beam-ACO algorithm is a hybrid ACO with beam search for solving TSP [ 26 ]. Chen and Chien presented a hybrid algorithm, called the genetic simulated annealing ant colony system with particle swarm optimization techniques, for solving TSP [ 27 ]. Junqiang and Aijia proposed a hybrid ant colony algorithm (HACO), which combined ACO with delete-cross to overcome the shortcoming of slow convergence speed of ACO [ 28 ]. Dong et al. [ 29 ] proposed an algorithm, called cooperative genetic ant system (CGAS) for solving TSP, which hybridized both GA and ACO to improve the performance of ACO. Recently, Gündüz et al. [ 30 ] presented a hybrid ACO with ABC for solving TSP. In addition, Mahi et al. [ 31 ] proposed a new algorithm in which ACO was hybridized with PSO and 3-Opt for solving small TSP instances. The PSO was used to determine the optimum values of the two main parameters of ACO which affected algorithm performance and the 3-Opt was used to escape from the local optima found by ACO algorithm. Furthermore, Yousefikhoshbakht et al. [ 32 ] proposed REACSGA for solving small TSP instances which employed the modified ACS for generating initial diversified solutions and GA for intensification mechanisms. As noted above, previous studies show that ACO still has drawbacks. The performance of these studies was dramatically decreased when dealing with large-scale instances. To the best of my knowledge, no research has been done to hybridize elitist ant system with SA, mutation, and local search. Therefore, in this research a new hybrid elitist ant system with SA, mutation operator, and local search procedure is introduced for solving TSP. Introducing SA can help ACO to escape from the local optima. On the other hand, determining initial solution of SA is almost difficult. Therefore, the use of the ACO is useful in the generation of SA initial solution. While introducing the mutation operation to ACO algorithm will enhance the algorithm performance, expand the diversity of population, and inhibit the premature convergence. Applying either SA or mutation is based on the diversity level of the population. After applying SA or mutation, elitist ant system goes through a local search procedure to speed up the convergence. The rest of the paper is structured as follows. Section 2 presents the TSP formulation. Section 3 describes the hybrid algorithm. The experimental results are presented in Section 4 . Conclusions and future work are given in Section 5 .",
"discussion": "4. Experimental Results and Discussion The experimental results to examine the validity and the performance of the proposed algorithm were introduced in this section. The experiments were conducted using 24 TSP standard benchmark problems, with different length, from TSPLIB [ 39 , 40 ]. The proposed algorithm has been implemented in Java on an Intel Core-i7 PC. The Object Oriented Paradigm (OOP) and different data structures have been used to optimize its code. All experiments were conducted on the symmetric TSP. As in other metaheuristic algorithms, the quality of the solutions created by the proposed algorithm was affected largely by the different values of the parameters. Thus, a number of different alternative values were examined to tune the parameters of the proposed algorithm. Table 1 provides the parameters which show variations in the range of values while default values of other parameters were taken. Finally, the selected parameter values are those that achieved the best computational results with respect to the quality of the solution. All of the parameter values have been determined by the experiments on Eil51, lin318, and fl1400 TSP instances which represent small, medium, and large instances, respectively. In these experiments, the proposed algorithm was stopped when reaching the optimal solution or 1000 iterations. The optimum combinations of the parameters are shown in Table 1 . Afterward, the algorithm was initialized with a population of 25 ants using α = 1 and β = 5 which control the influence of pheromone trail and heuristic information (edge cost) in selection of the next city by transition rule. The parameter q \n 0 was set to 0.05 which specifies the intensification/diversification rate. The initial value of the pheromone trail was set to be τ \n 0 = 0.5 and the maximum number of iteration was 1000 iterations. All of the instances included in TSPLIB have already been examined in the literature and their optimality results can be used to compare algorithms according to the best and the average values. Different instances with different size were selected. These instances can be classified into three groups based on their lengths. The first group is the smallest group which includes 8 instances varying in length between 51 and 100 cities. The second is the medium group which includes 10 instances varying in length between 101 and 318 cities. The third is the large-scale group which includes 5 instances with length between 575 and 1655 cities. Therefore, the results were collected after conducting the experiments 10 times for each instance and took the best results, the average results, and the standard deviation for comparison with previous work. Two different experiments were conducted for the evaluation. Both of them were configured according to the parameters setting as shown in Table 1 . In the first experiment, the proposed algorithm was compared with the basic elitist ant system algorithm (EAS). Six evaluation measures were used to evaluate both algorithms. These measures are best solution, worst solution, average solution, standard deviation, number of iterations, and running time of algorithms. Table 2 shows the superiority of the proposed algorithm over EAS in computational results for all evaluation measures. As can be seen from the tabulated values, the quality of the solutions obtained by the proposed algorithm was significantly better than the solutions obtained by EAS. This superiority of the proposed algorithm may be attributed to the introduction of SA which exploited the detected promising solutions to speed up the learning capability of the algorithm. Meanwhile, the addition of the mutation operator enhanced global search capability of the algorithm, prevented it from being trapped in local optima, and thus improved its performance. Moreover, the introduction of the local search strategy increased the speed of algorithm convergence. \n Figure 2 shows the behavior of the proposed algorithm with three TSP instances. They were chosen to represent short, medium, and large instances, and as such, the findings can be generalized to the other instances. \n Figure 2(a) shows the results from a single run for korB200 instance. In this instance, the proposed algorithm reached the optimal solution that is 29437 in all ten runs. This figure depicts best, average, and worst solution obtained during the run. Strong optimization capability of proposed algorithm could be inferred. Diversified solutions, high convergence speed, and stagnation avoidance can be observed from the convergence behavior of the algorithm. Specifically, the figure shows how quickly the optimal solution, that is 29437, was found after 26 iterations. The trend goes down fast towards the optimum solution in the early iterations until it approaches the optimum solution. It is clear that no stagnation happened during the search process as observed from the figure. \n Figure 2(b) illustrates the results from a single run of the proposed algorithm for the lin318 instance. In nine out of ten runs, the proposed algorithm reached the optimal solution that is 42029. This figure presents the graph which shows the convergence behavior of the algorithm. It is clearly noticed that the proposed algorithm has high convergence speed with diversified solutions. The algorithm succeeded in avoiding the potential stagnation and premature convergence as can be observed from the convergence behavior of the algorithm. According to the figure, there is no stagnation happened during the search process. Additionally, the algorithm converged to the best solution after a maximum of 106 iterations. The search space for this instance is medium although the algorithm has no problem in quickly finding the optimum solution. \n Figure 2(c) demonstrates a single run for rl1323 instance containing 1323 cities. This graph plots the convergence behavior of the proposed algorithm over 1000 iterations. The algorithm reached a solution with cost near to optimal solution at iteration number 293 and never once changed afterwards. As it can be seen from the figure, in the initial stage, the diversity was high due to the variation of the population. However, as fitness function decreased, the diversity also decreased until the suboptimal solution was attained that is 270388. This smooth convergence was due to the good balance between diversification and intensification that proposed algorithm could provide. Although the search space for that instance was large, no stagnation happened during the search process. In the second experiment, the proposed algorithm was compared with four state-of-the-art metaheuristic algorithms: Chen and Chien [ 27 ], Wang et al. [ 41 ], Yousefikhoshbakht et al. [ 32 ], and Mahi et al. [ 31 ]. The authors of these algorithms proved that their algorithms outperformed the other algorithms in the literature. The best found solution, the average one over all runs, the standard deviation, the percentage deviation of the best results, and the percentage deviation of the average results were used as evaluation measures for the comparison. The results are presented in Tables 3 and 4 and Figure 3 . In Tables 3 and 4 , column 1 shows the TSP instances, column 2 shows the best known solutions, column 3 shows the algorithms, column 4 shows the best solutions over all runs, column 5 shows the average solution of all runs, column 6 presents the standard deviation, column 7 reveals the percentage deviation of the best results (PD_Best) compared to those of the best known solution, and column 8 reveals the percentage deviation of the average of the best solution of all runs (PD_Avg) in comparison to the best known solution. PD_Best was calculated by ( 7 ) and PD_Avg was calculated by ( 8 ). (7) P D _ B e s t = b e s t s o l u t i o n − b e s t k n o w n s o l u t i o n b e s t k n o w n s o l u t i o n × 100 , \n (8) P D _ A v g = a v g s o l u t i o n − b e s t k n o w n s o l u t i o n b e s t k n o w n s o l u t i o n × 100 . \n In Table 3 , the proposed algorithm was compared with Chen and Chien [ 27 ] and Wang et al. [ 41 ] on 24 benchmark instances with cities from 51 to 1655. As can be seen in Table 3 , for the 24 TSP instances, the proposed algorithm was much better than both algorithms on all medium and large instances, such as lin318, rat575, rat783, rl1323, fl1400, and d1655, with respect to the five evaluation measures mentioned above. There was no significant difference between the proposed algorithm, Chen and Chien [ 27 ] and Wang et al. [ 41 ] on the small instances with cities less than or equal to 100, with respect to best found solution and PD_Best. In Table 4 , the proposed algorithm was compared with Yousefikhoshbakht et al. [ 32 ] and Mahi et al. [ 31 ] on 15 and 8 benchmark instances, respectively, with cities from 51 to 200 as reported in their studies. As can be seen in Table 4 , the values of columns best and PD_Best show that there was no significant difference between the proposed algorithm and Yousefikhoshbakht et al. [ 32 ] on the small instances with cities less than or equal to 100. For the larger instances, the proposed algorithm gained much better results than Yousefikhoshbakht et al. [ 32 ]. Comparing with Mahi et al. [ 31 ], the proposed algorithm achieved better results in all the 8 instances with respect to best found solution, average solution, PD_Best, and PD_Avg. \n Figure 3 shows a comparison of the proposed algorithm to Chen and Chien [ 27 ] and Wang et al. [ 41 ] based on the percentage deviations of the average solution to the best known solution. It is clear that the proposed algorithm significantly gained smaller percentage deviations than Chen and Chien [ 27 ] and Wang et al. [ 41 ] in the large-scale TSP instances that is lin318, rat575, rat783, rl1323, fl1400, and d1655. In summary, numerical results show that the proposed algorithm was effective. It was able to solve small and large size instances better than the existing algorithms. This is because of the proposed algorithm capability of searching the optimal solution until the last iterations without stagnation or premature convergence, especially for the medium and large TSP instances, compared to the other algorithms. In general, the results indicate that the structure of the proposed algorithm, which depends on the concepts of embedding simulated annealing, mutation operation, and local search procedure, achieved the balance between diversification and intensification and enabled algorithm to escape from local optima and speed up the convergence. These gave the proposed algorithm the superiority over the other algorithms in reaching the suboptimal/optimal solutions for TSP problems."
} | 4,620 |
29312225 | PMC5742216 | pmc | 1,867 | {
"abstract": "Microbial biofilms are highly structured and dynamic communities in which phenotypic diversification allows microorganisms to adapt to different environments under distinct conditions. The environmentally ubiquitous pathogen Cryptococcus neoformans colonizes many niches of the human body and implanted medical devices in the form of biofilms, an important virulence factor. A new approach was used to characterize the underlying geometrical distribution of C. neoformans cells during the adhesion stage of biofilm formation. Geometrical aspects of adhered cells were calculated from the Delaunay triangulation and Voronoi diagram obtained from scanning electron microscopy images (SEM). A correlation between increased biofilm formation and higher ordering of the underlying cell distribution was found. Mature biofilm aggregates were analyzed by applying an adapted protocol developed for ultrastructure visualization of cryptococcal cells by SEM. Flower-like clusters consisting of cells embedded in a dense layer of extracellular matrix were observed as well as distinct levels of spatial organization: adhered cells, clusters of cells and community of clusters. The results add insights into yeast motility during the dispersion stage of biofilm formation. This study highlights the importance of cellular organization for biofilm growth and presents a novel application of the geometrical method of analysis.",
"introduction": "Introduction Microorganisms have been traditionally analyzed using planktonic microbial cells; however, this lifestyle is not necessarily related with the growth of microbes in their most prevalent habitat. Recent approaches in confocal microscopy and molecular biology have provided evidence that biofilm formation represents the most common mode of microbial growth in nature (Costerton et al., 1995 ; Jabra-Rizk et al., 2004 ; Ramage et al., 2009 ; Martinez and Casadevall, 2015 ) and is a response to ecological competition in the environment (Oliveira et al., 2015 ). A wide range of microorganisms are able to switch from a planktonic to a colonial lifestyle in the form of a biofilm, creating aggregated communities that are enclosed by an extracellular matrix (ECM) (Costerton et al., 1995 ). Microbial biofilms are now recognized as highly structured and dynamic communities, in which phenotypic diversification allows microorganisms to adapt to diverse environments under different conditions (Watnick and Kolter, 2000 ; Parsek and Fuqua, 2004 ; Drescher et al., 2016 ; Gulati and Nobile, 2016 ; Sheppard and Howell, 2016 ). Importantly, biofilms can be composed of thousands of cells encased in a matrix and attached to a surface, but they can also contain as few as tens of cells arranged as small clusters or aggregates (Stacy et al., 2016 ). Open channels interspersing the microcolonies allow water and nutrients to reach their interior and contribute to the nutrition and formation of mature biofilms, possibly mimicking a primitive circulatory system. Waste products might also be removed through this system (Flemming and Wingender, 2010 ). Cells growing within biofilms exhibit unique phenotypic features compared to their planktonic counterparts, with the increased resistance to antimicrobial agents provided by biofilms being the more drastic example (Martinez and Casadevall, 2006a ; Clatworthy et al., 2007 ; Lewis, 2008 ; Ramage et al., 2009 ). Biofilm formation in the environment and in the host can be induced by sub-lethal concentrations of antibiotics or secondary metabolites, respectively (Kumar and Ting, 2013 ; Oliveira et al., 2015 ). In this context, biofilm formation is an important feature of Cryptococcus neoformans because it is an environmentally ubiquitous fungal pathogen that causes cryptococcosis, a lethal disease with a worldwide distribution related to bioclimatic conditions as well as to soil characteristics and land use (Cogliati et al., 2017 ). Almost 200,000 deaths per year are estimated to be due to cryptococcal meningitis (Rajasingham et al., 2017 ). The major virulence factor of this fungus is the polysaccharide capsule that surrounds the cell wall and is responsible for fungal attachment to surfaces and subsequent biofilm formation (Martinez et al., 2010 ; de S Araújo et al., 2016 ). The C. neoformans capsule is composed mainly of glucuronoxylomannan (GXM), a polysaccharide generated intracellularly and exported to the extracellular space via vesicle-mediate secretion (Rodrigues et al., 2007 ). GXM is also a constituent of the cryptococcal biofilm ECM (Martinez and Casadevall, 2005 ; Park et al., 2009 ). C. neoforma ns can form biofilms on medical devices, including ventriculoatrial shunt catheters (used to manage intracranial hypertension), peritoneal dialysis fistulae, cardiac valves and prosthetic joints (Walsh et al., 1986 ; Braun et al., 1994 ; Banerjee et al., 1997 ; Johannsson and Callaghan, 2009 ; Shah et al., 2015 ). On biotic surfaces, after traversing the blood brain barrier in meningoencephalitis, C. neoformans has the ability to form biofilm-like structures known as cryptococcomas (Aslanyan et al., 2017 ). Although previous studies using confocal microscopy provided initial insights into cryptococcal biofilm structure, conventional scanning electron microscopy (SEM) techniques do not preserve the mature biofilm ultrastructure (Martinez and Casadevall, 2005 , 2007 ). The highly hydrated matrix is greatly deformed and the cell samples undergo distortion and may present artifacts. Also, C. neoformans capsule is sensitive to dehydration and is easily disrupted during routine sample preparation (Edwards et al., 1967 ; Sakaguchi, 1993 ). As a consequence, considerable effort is currently being spent on the development of new methods and instrumentation for its visualization. By applying an adapted protocol for SEM, we characterized the underlying geometrical structure of cell distribution during biofilm formation. The degree of order was numerically quantified and we revealed a correlation between higher levels of biofilm formation and more ordered underlying structures. Order/disorder are very relevant in physical systems. In crystals, for example, deformations can only occur near defects due to the high energetic cost of their occurrence elsewhere. Besides, some phase transitions are defect mediated. Moreover, in the last decades the interplay among defects, geometry and statistical physics has been highlighted (Nelson, 2002 ). Here we propose the application of parameters designed to measure order in physical systems (Nelson and Halperin, 1979 ; Aeppli and Bruinsma, 1984 ; Okabe et al., 2000 ; Bernard and Krauth, 2011 ; Borba et al., 2013 ) to the microbial populations. We also investigated the details of the ultrastructural organization of cryptococcal biofilms and show that cryptococcal cells aggregate with a specific ordered structure favoring biofilm formation as compared to disorganized conglomerates.",
"discussion": "Discussion The hallmarks of this study were the use of a numerical measure to quantify the geometrical order of the first layer of adhered cells in the process of biofilm formation as well as the detection of well-shaped ultrastructure of C. neoformans biofilms. To verify the relation between cellular order and biofilm production, we analyzed C. neoformans H99 and the mutants grasp and cap67 , and the usual model C. neoformans B3501. Once we showed that there is indeed a correlation between increased order and increased biofilm production, we focused on the standard strain to further study the ultrastructure. This analysis was made possible by the introduction of a modified protocol for SEM visualization of microbial biofilms, due to the fact that standard protocols greatly distort the matrix. To minimize artifacts, a shortened time of fixation and careful dehydration is optimal for ultrastructural SEM analysis (Joubert et al., 2015 ). The ultrastructure preservation was achieved by combining appropriated techniques, a reduced period of incubation during SEM preparation and good grade reagents. Distinct levels of spatial organization were observed: adhered cells, clusters of cells, as well as the community of clusters. The affinity of attachment to different surfaces is strongly related to the presence of the cryptococcal capsule. In fixed C. neoformans cells, the fibers surrounding the cell (capsule filaments) directly stretch and link cells to surface, promoting attachment (de S Araújo et al., 2016 ). Cryptococcal biofilm formation seems to be driven by a communication system via adhesion/matrix protein signaling (Wang et al., 2013 ) and directional proliferation of the original adhered cells. Cfl1, the first prominent ECM secreted protein of C. neoformans , is highly expressed in subpopulations located at the periphery of a mating community and is concentrated in the extracellular matrix boundary. This protein orchestrates yeast-hypha morphotype transition, cell adhesion, and virulence. This suggests that Cfl1 possibly serves as a signal regulating morphotype transition in the cells enclosed or adjacent to the ECM (Wang et al., 2012 , 2013 ; Wang and Lin, 2015 ). We hypothesize that the reversible cell attachment is mediated by capsule interactions and the orderly distribution of cells, as described above. As the capsule is primarily responsible for the high negative zeta potential of C. neoformans cells, variations in the structure of GXM could also influence the Zeta potential. However, zeta potential determinations of C. neoformans H99 and B3501 strains did not reveal major differences (Kozel and Gotschlich, 1982 ; Nosanchuk and Casadevall, 1997 ; Cordero et al., 2011 ). In this way, we assume that biofilm formation capability is a serotype-dependent process and is influenced by either biological or environmental factors. Based on this, the resulting patterns observed for H99 and B3501 strains cannot be explained by charge distribution. The ordering reported is due to an effective repulsion among cells, the nature of which remains a mystery. It may be due to chemical sensing or excluded volume that hinders the free motion of cells (Movie S1 ), in which cells flow following similar paths and tend to adhere at an approximately constant distance from one another. Our data supports that once irreversible attachment occurs, cryptococcal cells may form a narrow ECM layer around the cell body where cells rapidly proliferate, but the surface-attached and peripheral anchored cryptococcal cells may restrict their expansion to the plane. As initial small clusters proliferate, their shape increasingly becomes anisotropic. At this point, the biofilm consists of several layers of cells grouped into clusters resembling extremely organized flower-like patterns. After maturation, cells may detach as microcolonies or as isolated planktonic cells, which auto-organize following an approximately hexagonal distribution. Cells tend to follow similar trajectories and may initiate the process again (Figure 10 and Movie S1 ). Figure 10 Scheme of C. neoformans B3501 biofilm formation. (1,2) Adhesion of planktonic cells follows an approximately hexagonal distribution. (3) Cluster expansion and shaping. (4) Flower-like mature biofilm. (5) Detachment of microcolonies or planktonic cells. (a) SEM of biofilm development stages. Yan et al. ( 2016 ) discovered that the cluster ultrastructure of Vibrio cholerae biofilm results from the combination of expansion and confinement of surface-attached cells that generates an effective anisotropic stress. Such stress overpowers the cell-to-surface adhesion force for cells at the cluster center, causing these cells to realign in the vertical direction and forcing the transition from 2D expansion to 3D growth (Yan et al., 2016 ). Moreover, if selection pressure is high, it has been shown that clusters of Pseudomonas aeruginosa have higher fitness than isolated cells because cells at the top of the clusters have better access to nutrients (Kragh et al., 2016 ). Cluster morphogenesis results from a great number of variables capable of shaping the ultrastructure. Physical and demographic processes are demonstrated to act as key factors in biofilm architectures (Hödl et al., 2014 ). Cryptococcal cells are likely most susceptible to the hydrodynamics constraints due to low motility. In contrast, more motile microorganisms may escape these constraints and develop biofilm morphogenesis related to cellular migration and biofilm coalescence (Hödl et al., 2014 ). Mathematical studies have related that biofilm architecture depends on the availability of nutrients, carbon and oxygen, uptake processes linked to hydrodynamics and diffusion limitation of substrate transport through the biofilm. More generally, metabolic capabilities, genotypic and phenotypic adaptations could result in different behaviors within the biofilm, allowing organisms to choose between a number of strategies (Klapper and Szomolay, 2011 ; Klapper, 2012 ). Interestingly, for V. cholerae the presence of low cell number in cluster biofilm results in increased volume when compared to biofilms with a larger population. The hypothesis is that the significant changes in cell–cell spacing between small and large clusters in biofilms are due to strong temporal variation in ECM composition or production levels per cell (Drescher et al., 2016 ). In agreement, the flower-like clusters of C. neoformans present a high volume of ECM and relatively low cell concentration, as supported by our findings. For instance, the shunting procedures used to treat cryptococcal meningitis hypertension are risk associated and have historically discouraged surgeons due to its complications (Liu et al., 2014 ; Cherian et al., 2016 ) since it can provide a surface for cryptococcal attachment. It is common knowledge that uropathogenic strains of Escherichia coli can successfully adhere to and colonize the kidney, despite the presence of high flow rates. Since kidney tubules are narrow (<50 μm), bacterial attachment patterns at even very small spatial scales can easily block them, increasing the severity of kidney infection (Melican et al., 2011 ). The architectural flower-like cluster organization observed in serotype D B3501 strain might provide the yeast cells with a protected niche against antifungals, host defenses, environmental predators and dehydration. Physical differences in C. neoformans serotypes A and D biofilms may reflect the predilection of some serotype D strains for peripheral tissue (e.g., skin) whereas the structure of serotype A biofilms may select these strains in tissues such as the lungs (Abdulkareem et al., 2015 ). As demonstrated, cryptococcal cells may detach from the biofilm in an organized manner. It is plausible to assume that organization may be needed for the successful dissemination to the host. In fact, researchers showed that Candida albicans detached cells from biofilms are more metabolically active than planktonic cells (Uppuluri et al., 2010 ). Upon this scenario, special treatment of the devices or the use of materials that hinder the initial organization may be used clinically to avoid the development of infection, by disrupting the initial organization. Moreover, the introduction of an objective measure of order (ψ 6 ) obtained from an image may facilitate the analysis of whether a given surface is prone to biofilm formation. Continued studies are required to provide a greater understanding of the importance to investigate the complications of cryptococcal meningoencephalitis associated to the spatial distribution of clusters, as well as new methods of imaging for helping the development of new anti-biofilm targets."
} | 3,954 |
28779290 | PMC5569662 | pmc | 1,868 | {
"abstract": "Paddy fields are a significant source of methane and contribute up to 20% of total methane emissions from wetland ecosystems. These inundated, anoxic soils featuring abundant nitrogen compounds and methane are an ideal niche for nitrate-dependent anaerobic methanotrophs. After 2 years of enrichment with a continuous supply of methane and nitrate as the sole electron donor and acceptor, a stable enrichment dominated by ‘ Candidatus Methanoperedens nitroreducens’ archaea and ‘ Candidatus Methylomirabilis oxyfera’ NC10 phylum bacteria was achieved. In this community, the methanotrophic archaea supplied the NC10 phylum bacteria with the necessary nitrite through nitrate reduction coupled to methane oxidation. The results of qPCR quantification of 16S ribosomal RNA (rRNA) gene copies, analysis of metagenomic 16S rRNA reads, and fluorescence in situ hybridization (FISH) correlated well and showed that after 2 years, ‘ Candidatus Methanoperedens nitroreducens’ had the highest abundance of (2.2 ± 0.4 × 10 8 ) 16S rRNA copies per milliliter and constituted approximately 22% of the total microbial community. Phylogenetic analysis showed that the 16S rRNA genes of the dominant microorganisms clustered with previously described ‘ Candidatus Methanoperedens nitroreducens ANME2D’ (96% identity) and ‘ Candidatus Methylomirabilis oxyfera’ (99% identity) strains. The pooled metagenomic sequences resulted in a high-quality draft genome assembly of ‘ Candidatus Methanoperedens nitroreducens Vercelli’ that contained all key functional genes for the reverse methanogenesis pathway and nitrate reduction. The diagnostic mcrA gene was 96% similar to ‘ Candidatus Methanoperedens nitroreducens ANME2D’ (WP_048089615.1) at the protein level. The ‘ Candidatus Methylomirabilis oxyfera’ draft genome contained the marker genes pmoCAB , mdh , and nirS and putative NO dismutase genes. Whole-reactor anaerobic activity measurements with methane and nitrate revealed an average methane oxidation rate of 0.012 mmol/h/L, with cell-specific methane oxidation rates up to 0.57 fmol/cell/day for ‘ Candidatus Methanoperedens nitroreducens’. In summary, this study describes the first enrichment and draft genome of methanotrophic archaea from paddy field soil, where these organisms can contribute significantly to the mitigation of methane emissions. Electronic supplementary material The online version of this article (doi:10.1007/s00253-017-8416-0) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction The methane concentration in the atmosphere has increased continuously over the last 150 years. Methane is the second most abundant greenhouse gas and exhibits radiative forcing up to 34 times higher than that of CO 2 (Myhre et al. 2013 ). Paddy fields are a significant source of methane and contribute 10–20% to global methane emissions (Bodelier 2011 ; Conrad 2009 ). The cultivated area dedicated to rice agriculture occupies approximately 160 million ha of land worldwide and is predicted to increase by 60% in the coming decades. Without changes in cultivation practices, such increases will result in even higher atmospheric methane emissions. The global biogenic methane budget is directly linked to the activity of methanogenic and methanotrophic microorganisms in the environment. Methanotrophic organisms function as a biofilter, and without their contribution, it is estimated that the atmospheric methane concentration would be 10–60% higher (Conrad 2009 ). Whereas aerobic methanotrophs are well studied, much less is known about methane removal in oxygen-limited nitrogen-loaded freshwater environments. NC10 phylum (‘ Candidatus Methylomirabilis oxyfera’) bacteria and ‘ Candidatus Methanoperedens nitroreducens’ archaea are the only methanotrophic microorganisms known to directly couple the anaerobic oxidation of methane to the nitrogen cycle . NC10 phylum bacteria use nitrite as an electron acceptor (A), and ‘ Candidatus Methanoperedens nitroreducens’ archaea perform nitrate reduction (B) with methane as an electron donor according to the following reactions: \\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}$$ 8{{\\mathrm{N}\\mathrm{O}}_2}^{-}+3{CH}_4+8{\\mathrm{H}}^{+}\\to 4{\\mathrm{N}}_2+3{CO}_2+10{\\mathrm{H}}_2\\mathrm{O}\\ \\left(\\mathrm{A},\\Delta G{0}^{\\prime }=-987\\ \\mathrm{kJ}/\\mathrm{mol}\\ {CH}_4\\right) $$\\end{document} 8 NO 2 − + 3 CH 4 + 8 H + → 4 N 2 + 3 CO 2 + 10 H 2 O A Δ G 0 ′ = − 987 kJ / mol CH 4 \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}$$ 2\\mathrm{x}\\ \\left(4{{\\mathrm{NO}}_3}^{-}+{CH}_4\\to 4{{\\mathrm{NO}}_2}^{-}+{CO}_2+2{\\mathrm{H}}_2\\mathrm{O}\\right)\\ \\left(\\mathrm{B},\\Delta G{0}^{\\prime }=-503\\ \\mathrm{kJ}/\\mathrm{mol}\\right) $$\\end{document} 2 x 4 NO 3 − + CH 4 → 4 NO 2 − + CO 2 + 2 H 2 O B Δ G 0 ′ = − 503 kJ / mol \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}$$ 8{HNO}_3+5{CH}_4\\to 5{CO}_2+4{\\mathrm{N}}_2+14{\\mathrm{H}}_2\\mathrm{O}\\ \\left(\\mathrm{sum}\\right) $$\\end{document} 8 HNO 3 + 5 CH 4 → 5 CO 2 + 4 N 2 + 14 H 2 O sum \n In 2006, an enrichment culture in which nitrate and nitrite reduction were coupled to the anaerobic oxidation of methane was described for the first time (Raghoebarsing et al. 2006 ). In that study, an inoculum from the sediment of a freshwater canal was used to start an anaerobic enrichment. After 16 months, the culture was dominated by a consortium consisting of archaea (10–15% of cells) belonging to the Methanosarcinales family that were only distantly related to ANME2D (86–87% in 16S ribosomal RNA (rRNA identity)) and a bacterium (approximately 80% of cells) of the candidate division NC10 . The enriched co-culture preferred nitrite over nitrate as the substrate, although activity was observed with both substrates (Raghoebarsing et al. 2006 ). The nitrite-dependent anaerobic oxidation of methane (AOM) was later assigned to phylum NC10 bacteria, which are able to carry out this process in the absence of other microorganisms (Ettwig et al. 2008 ). The bacterium uses an intra-aerobic mechanism in which oxygen is produced via a putative nitric oxide dismutase and subsequently used for methane oxidation via the particulate methane monooxygenase complex. Assembly of the genome of the NC10 bacterium revealed a complete methane oxidation pathway that included the pmoCAB operon and an incomplete denitrification pathway. It was hypothesized that the dismutation of nitric oxide to oxygen and nitrogen supplies O 2 for the methane monooxygenase. These NC10 phylum bacteria were named ‘ Candidatus Methylomirabilis oxyfera’ (Ettwig et al. 2010 ). Sequencing of the genome of the AOM archaea and identification of their nitrate reductase indicated that these archaea could couple nitrate reduction to AOM (Arshad et al. 2015 ; Haroon et al. 2013 ). The responsible archaea were named ‘ Candidatus Methanoperedens nitroreducens.’ Recent microbial ecology studies have indicated sufficient presence and activity of methanotrophic archaea in paddy field soils (Lee et al. 2015 ; Vaksmaa et al. 2016 ) to warrant investment in a long-term enrichment procedure to obtain these paddy field AOM archaea and to study their physiology and metabolic potential in more detail. To achieve this goal, we started a sequencing batch bioreactor continuously fed with nitrate and methane and inoculated with soil from an Italian paddy field soil harboring substantial AOM archaeal cell numbers (Vaksmaa et al. 2016 ). After establishing nitrate-dependent methane oxidation, the total DNA of this biomass was sequenced using Ion Torrent technology, and the draft genome was annotated and analyzed. The enriched microbial community was further characterized by microscopy, 13 CH 4 and 15 N activity assays, and qPCR.",
"discussion": "Discussion Nitrate-dependent anaerobic oxidation of methane (N-AOM) was discovered a decade ago, but the characterization of the metabolism has been hindered by the slow growth of the responsible organisms. The N-AOM microorganisms ‘ Candidatus Methanoperedens nitroreducens’ archaea and NC10 phylum bacteria have been detected in various fresh water sediments (Welte et al. 2016 ). In this study, we started an enrichment culture fed solely with methane and nitrate using a paddy field soil harboring significant amounts of ‘ Candidatus Methanoperedens nitroreducens’ (Vaksmaa et al. 2016 ) as the inoculum. Based on qPCR, FISH, and metagenome analyses, the enrichment was dominated by ‘ Candidatus Methanoperedens nitroreducens’ after 2 years of enrichment. Many previous enrichments were fed with nitrite or a mixture of nitrite and nitrate instead of nitrate only; such conditions are presumably advantageous to NC10 phylum bacteria. We intentionally omitted ammonium from the medium as other studies showed that such cultures would yield a mixed culture of ‘ Candidatus Methanoperedens nitroreducens’ and anammox bacteria, which could outcompete NC10 phylum bacteria for nitrite (Shi et al. 2013 ). Our previous field work demonstrated a high abundance of ‘ Candidatus Methanoperedens nitroreducens’ in the paddy field soil (Vaksmaa et al. 2016 ), which was confirmed by qPCR of the inoculum slurry. The inoculum slurry contained approximately 1.9 ± 0.1 × 10 5 copies per milliliter of the 16S rRNA gene of ‘ Candidatus Methanoperedens nitroreducens’ and 1.9 ± 0.3 × 10 3 copies per milliliter of NC10 phylum bacteria. After 2 years of enrichment, these numbers had increased to 2.2 ± 0.4 × 10 8 16S rRNA copies per milliliter of ‘ Candidatus Methanoperedens nitroreducens’, corresponding to 22% of the total detected 16S rRNA copies (bacteria plus archaea). These numbers indicate a doubling time of 1 to 2 months. The abundance of 16S rRNA gene copies of NC10 phylum bacteria was 7.9 ± 0.3 × 10 7 , corresponding to approximately 4% of the total copies. The qPCR copy numbers correlated well with the metagenome sequencing results for ‘ Candidatus Methanoperedens nitroreducens’, with an abundance of 16S rRNA reads of 22% after 2 years. The percentage of reads assigned to NC10 phylum bacteria was 15%, possibly indicating underestimation by qPCR. The results based on the two methods presented here provided insight into the growth dynamics of both methane oxidizers. The growth of ‘ Candidatus Methanoperedens nitroreducens’ was observed after approximately 10 months of acclimatization, whereas for NC10 phylum bacteria, more than a year was necessary before a substantial increase in cell numbers was observed. The initial growth of NC10 phylum bacteria was presumably nitrite-limited. Similar lag phases of the growth of NC10 phylum bacteria in enrichment cultures have been reported previously. Zhu et al. showed that ‘ Candidatus Methanoperedens nitroreducens’ only started to increase in an enrichment obtained from minerotrophic peatland after 9 months, when significant methane oxidation rates (9 nmol/day/g in serum bottles, based on CO 2 production) indicated microbial growth (Zhu et al. 2012 ). In addition to substrate preference and availability, temperature has been implicated as a decisive factor in the outcome of AOM enrichments. In enrichments started from wastewater treatment plant sludge and lake sediments, a co-enrichment of NC10 phylum bacteria and ‘ Candidatus Methanoperedens nitroreducens’ was obtained at 35 °C, whereas at 22 °C, only NC10 phylum bacteria were enriched (Hu et al. 2009 ). In our enrichment culture, the methane oxidation potential increased in accordance with the 16S rRNA copy number. The batch incubations performed with the whole bioreactor revealed average methane oxidation and nitrate reduction after 2 years of 0.055 and 0.012 mmol/h/L, respectively. Based on 13 C-CO 2 production, the cell-specific methane oxidation rates after 2 years were 0.57 fmol/cell/day for ‘ Candidatus Methanoperedens nitroreducens’. This is in the same range as a previously reported nitrate-dependent AOM rates that we measured in paddy field soil in which the estimated cell-specific rates were 1.2 fmol/cell/day of CH 4 (Vaksmaa et al. 2016 ) as well as NC10 phylum bacteria enrichment in which the cell-specific rates were about 0.2 fmol/cell/day of CH 4 (Ettwig et al. 2009b ) and is also comparable to rates reported for sulfate-dependent AOM by ANMEs (0.7 fmol CH 4 /cell/day) (Nauhaus et al. 2005 ). Unfortunately, we did not observe nitrite-dependent methane oxidation in the batch incubation, suggesting that either 1 mM nitrite was greater than the inhibitory concentration for the organism, regardless of their capacity to metabolize nitrite or the biomass requires a longer adaptation time to overcome the previous nitrite limitation. Based on the stoichiometry of the reactions for nitrate- and nitrite-dependent anaerobic oxidation of methane, AOM organisms accounted for approximately 46% of nitrate consumption, whereas presumably other nitrate reducers in the reactor, such as denitrifiers, were responsible for the remaining 54% of nitrate loss. Metagenome analysis revealed that only a few phyla other than ‘ Candidatus Methanoperedens nitroreducens’ and NC10 phylum bacteria were represented in greater than 5% abundance. Anaerolineales (8.5% abundance) belonging to Chloroflexi are obligate anaerobes that have previously been observed in anaerobic methanotrophic (Ettwig et al. 2009a ; Siniscalchi et al. 2015 ) and methanogenic enrichment cultures (Gray et al. 2011 ; Liang et al. 2015 ; Yamada et al. 2005 ). Anaerolineales may be responsible for the degradation of n -alkanes and release formate, acetate, hydrogen, and carbon dioxide. Hug et al. indicated that Anaerolineales may provide organic acids to other microorganisms such as acetoclastic methanogens (DeSantis et al. 2006 ). The physiology of Candidate division OC31 (7% abundance), which was discovered more recently, remains unknown. Phycisphaerae , a class of Planctomycetes (4.4% abundance), has also been shown to degrade heteropolysaccharides (Wang et al. 2015 ) and was previously found to be highly abundant in AOM and other anaerobic enrichment cultures. After 2 years of enrichment, Rhodocyclaceae accounted for 6.1% and Comamonadaceae for 5.6%. Both of these belong to Betaproteobacteria . Members of Comamonadaceae can perform denitrification, which may explain the observed nitrate reduction rate, which was higher than expected based on the methane oxidation rate alone. The 3.5-Mb size of the draft genome of ‘ Candidatus Methanoperedens nitroreducens Vercelli’ is comparable to those of the publicly available genomes of ‘ Candidatus Methanoperedens BLZ1’ (3.7 Mb) and ‘ Candidatus Methanoperedens nitroreducens ANME2D’ (3.2 Mb). The GC content of the ‘ Candidatus Methanoperedens nitroreducens Vercelli’ genome is 44.1% and is more similar to that of ‘ Candidatus Methanoperedens nitroreducens ANME2D’ (GC content 43.2%) than ‘ Candidatus Methanoperedens BLZ1’ (40.8%). Functional gene analysis revealed that the mcrA gene has 96% identity to ‘ Candidatus Methanoperedens nitroreducens ANME2D’ and 89% identity to ‘ Candidatus Methanoperedens BLZ1’ at the protein level. A similar trend was observed for the majority of enzymes in the reverse methanogenesis pathway. Analysis of the denitrification pathway revealed the presence of nitrate reductases as well as nitric and nitrous oxide reductases in the draft genome, whereas no nitrite reductase could be identified. In summary, this is the first enrichment culture from paddy field soil supplied solely with nitrate and methane to enrich ‘ Candidatus Methanoperedens nitroreducens’ and NC10 phylum bacteria. The newly enriched co-culture will be used in future studies to unravel the ecophysiological properties of AOM microbes and investigate their role in mitigating methane emissions from paddy fields."
} | 4,125 |
27035705 | PMC5042319 | pmc | 1,869 | {
"abstract": "The microbial world displays an immense taxonomic diversity. This diversity is manifested also in a multitude of metabolic pathways that can utilise different substrates and produce different products. Here, we propose that these observations directly link to thermodynamic constraints that inherently arise from the metabolic basis of microbial growth. We show that thermodynamic constraints can enable coexistence of microbes that utilise the same substrate but produce different end products. We find that this thermodynamics-driven emergence of diversity is most relevant for metabolic conversions with low free energy as seen for example under anaerobic conditions, where population dynamics is governed by thermodynamic effects rather than kinetic factors such as substrate uptake rates. These findings provide a general understanding of the microbial diversity based on the first principles of thermodynamics. As such they provide a thermodynamics-based framework for explaining the observed microbial diversity in different natural and synthetic environments.",
"introduction": "Introduction There is an immense diversity of microbes in the natural environment ( Curtis et al. , 2002 ). One major challenge for microbial ecology, besides achieving more complete enumeration of the total diversity ( Bunge et al. , 2013 ), is to explain how this diversity is generated and maintained over evolutionary time. In particular, understanding the set of environmental, biochemical and evolutionary conditions that can lead to the generation and maintenance of microbial diversity is a prerequisite to understand and control natural microbial populations ( Gudelj et al. , 2010 ; O'Brien et al. , 2013 ) and engineer synthetic microbial communities ( Großkopf and Soyer, 2014 ). In ecology, a historically dominant idea in the study of diversity is the ‘competitive exclusion principle', which states that at equilibrium no two species can coexist occupying the same niche ( Hardin, 1960 ). This principle is shown theoretically in the context of microbial ecology using mathematical models of the well-mixed, single substrate chemostat environment. In such environments, the theory predicts that coexistence of two species can only be possible for a unique combination of kinetic parameters, and outside this combination only a single species, that has the highest substrate affinity, can survive at steady state ( Hsu et al. , 1977 ). Thus, it is expected that a single organism should monopolise each substrate; and the number of observed species in the environment should not surpass the number of limiting nutrients or substrates. This exclusion theory led to the proposition of the ‘paradox of the plankton' as the problem of how a high biological diversity can be maintained on a relatively limited number of niches ( Hutchins, 1961 ). One way to resolve this paradox is to invoke spatial and temporal variations in substrate levels ( Hsu et al. , 1977 ; Pfeiffer et al. , 2001 ; Kerr et al. , 2002 ; MacLean and Gudelj, 2006 ). Although spatial and temporal variation in a single substrate can certainly contribute to microbial diversity, it cannot explain species harbouring different metabolic pathways that enable conversion of the same substrate into different end products ( Rodríguez et al. , 2008 ). Even a common substrate such as glucose can be converted into a variety of end products by different species or even within a single species ( Gupta and Clark, 1989 ). This catabolic diversity contributes to the observed species diversity in the environment; and it is possible that these two observations of diversity are linked. For example, it is proposed that metabolic byproducts ( Schink, 1997 ) as well as specific production of toxic substances ( Lenski and Hattingh, 1986 ) could differentially inhibit competing species or result in autoinhibition ( De Freitas and Fredrickson, 1978 ). When such inhibition affects species competing for a given substrate in a differential way, it can allow coexistence on that single substrate ( De Freitas and Fredrickson, 1978 ; Lenski and Hattingh, 1986 ). Arguably the most fundamental inhibitory constraint on microbial growth is that arising from thermodynamics of metabolism. As microbial growth depends on energy harvested from substrates converted into end products, it is governed by the thermodynamics of such metabolic conversions. Here, we consider this relation between the thermodynamic constraints placed on growth sustaining metabolic conversions and the resulting population-level dynamics. Using thermodynamic models of microbial growth, we show that the inevitable slowing down of microbial growth ensuing from product built-up can lead to coexistence of different species implementing different metabolic conversions and consuming the same substrate. As each metabolic conversion operates with different thermodynamics, species utilising these are governed by different growth and product-inhibition dynamics that result in their coexistence. We find this ‘thermodynamics inhibition' effect to be strongest for reactions leading to a low change in free energy, where it dominates over kinetic factors such as substrate uptake rates. In line with this fundamental observation, we find that several biologically relevant microbial conversions as well as theoretically possible metabolic conversions of glucose fit in the regime of strong thermodynamic effects and readily lead to coexistence of different species on a single substrate. These findings provide a thermodynamic basis to evaluate observed microbial diversity.",
"discussion": "Discussion We have considered a thermodynamic model for microbial growth and analysed the ensuing population dynamics in the context of competition and coexistence. The model is derived from the fundamental principles of conservation of energy and thermodynamics in metabolic conversions fuelling microbial growth. The key result from this first-principles model is that utilisation of different metabolic conversions by different species can allow for their coexistence on a single substrate under a homogenous environment. We find that the abundance of different species under this circumstance is governed by the change in standard free energy of the reaction that they utilise, and that species utilising reactions that are within ~50 kJ (mol Substrate) −1 of the reaction with highest change in free energy would coexist at significant frequency. An analysis of known and chemically possible metabolic conversion reactions shows that this is a biologically relevant regime, where many biochemical reactions known to sustain microbial growth are found. This indicates that significant amounts of microbial diversity on a single substrate could have initially emerged from, or are being sustained by thermodynamic constraints. As the change in standard free energy of the utilised metabolic conversion reactions increases, however, we find that kinetic effects such as differences in substrate uptake rates overcome the thermodynamics effects on microbial growth dynamics. As a result, it can be expected that any change in kinetic parameters (for example, by evolutionary change) would easily disrupt thermodynamic-driven diversity emerging under metabolic conversions with large change in free energy. The same is not true for metabolic conversions with small change in Gibbs free energy, where we find that species utilising metabolic pathways with of −20 vs −25 kJ (mol Substrate) −1 can still coexist even with a 10-fold difference in their substrate half saturation constants ( Figure 5 ). We emphasise that these predictions on the energy ranges leading to coexistence are conservative estimates, as the thermodynamic model used here considers that all of the free energy available from the metabolic conversion is invested into growth rate ( Hoh and Cord-Ruwisch, 1996 ). In reality, some of this free energy would need to be invested in driving anabolic reactions and other cellular maintenance processes ( Kleerebezem and Stams, 2000 ; Jin and Bethke, 2003 ; Rodríguez et al. , 2008 ). As a result, even metabolic conversions with higher free energy change could enter a regime of thermodynamic inhibition, and offer a window for the emergence of thermodynamics-driven diversity. The presented model considers the thermodynamics of microbial growth, by considering a overall growth-supporting metabolic reaction (for example, glucose to acetate). In reality, cellular metabolism takes place over many reactions that finally reach a metabolic end product. Thus, the reaction Gibbs free energy change from the overall reaction is split among all the individual reactions and some of it needs to be invested to achieve an appropriate flux for these reactions ( Flamholz et al. , 2013 ). The consequence of this is that not all of the Gibbs free energy change from the overall reaction can be invested in growth rate, as we assume here. Therefore, our estimates for the effects of thermodynamic inhibition could act in a larger parameter regime, that is, even for overall reactions with larger Gibbs free energies than studied here. In future work, it could be possible to consider different reaction pathways in different species to get a more accurate model of their thermodynamic growth dynamics. Indeed, studies in this direction are already being employed to compare different pathways ( Flamholz et al. , 2013 ) and assess pathway feasibility under different conditions ( González-Cabaleiro et al. , 2013 ; Cueto-Rojas et al. , 2015 ). Although the thermodynamic constraints highlighted here act similar to inhibition driven by metabolic byproducts ( De Freitas and Fredrickson, 1978 ; Lenski and Hattingh, 1986 ), it is important to note that the former does not inhibit microbial growth per se , but emerge from the drive towards chemical equilibrium in the given metabolic reaction sustaining growth. Therefore, each given species is specifically affected by the build-up of its own products. The inherent thermodynamic mechanism is thus to punish specifically the fastest growing organisms the earliest, and thereby favouring the coexistence of a high number of different metabolic conversions in the environment. The products of these diverse metabolic conversions can then be utilised by the same species or different ones as energy source, resulting in an example of ‘niche creation', and potentially leading to the emergence of further microbial diversity and interactions through adaptation. Indeed, anaerobic microbial communities are frequently characterized by abundance of such interlinked metabolic conversions ( Schink, 1997 , 2002 ). The findings of this study suggest that microbial coexistence can readily arise under metabolic growth-supporting reactions with low free energy change. A direct experimental test for this proposition for this would be to grow two different species, or synthetically tagged variants of the same species, on a single substrate. The choice of the species and substrate should be such that each species can only utilise metabolic growth-supporting reactions with low free energy change. Such an experiment can be run under chemostat conditions, as well as under batch conditions that test mutual invasion from low frequency. Examples would include anaerobic growth on propionate or glycerol, where different metabolic pathways are known to exist. A broader suggestion from this study is that environments that mainly allow for growth-supporting metabolic reactions with low free energy change should harbour more metabolic diversity compared with environments that allow for metabolic reactions with high free energy change. This proposition could potentially be tested through the use of increasingly available metagenomics data from different environments. Examples for former type of environments would include anaerobic digesters, animal guts, wet soils and ocean sediments of highly productive regions. In these environments, high concentrations of substrates along with the lack of strong oxidising agents like oxygen or nitrate lead to accumulation of high concentrations of waste compounds. Thus, we expect microbial growth in these environments to be mainly limited by the lack of free energy available from the specific metabolic conversions a given species utilises. Microbes can overcome this limitation by evolving an ability to produce different waste products from the same substrate molecule. This way they can maintain growth by producing those products that are only present at very low concentrations in the environment, and overcoming any thermodynamic inhibition from accumulated products. This situation is analogous to a river flowing from the mountains to the sea, at its steeper parts the river is maintained in a single basin, while running over the flat surface close to the coast it starts to meander and form many little rivers in a delta."
} | 3,233 |
28210070 | null | s2 | 1,870 | {
"abstract": "A novel high throughput method for synthesis and screening of customized protein-resistant surfaces was developed. This method is an inexpensive, fast, reproducible and scalable approach to synthesize and screen protein-resistant surfaces appropriate for a specific feed. The method is illustrated here by combining a high throughput platform (HTP) approach together with our patented photo-induced graft polymerization (PGP) method developed for facile modification of commercial poly(aryl sulfone) membranes. We demonstrate that the HTP-PGP approach to synthesize and screen fouling-resistant surfaces is general, and thus provides the capability to develop surfaces optimized for specific feeds. Surfaces were prepared "
} | 180 |
29491985 | PMC5804180 | pmc | 1,871 | {
"abstract": "Abstract Relocation is an important event in the lives of several social insects whereby all colony members have to be transferred to a new nest when conditions in the old nest become unfavorable. In the current study, network tools were used to examine the organization of this goal-oriented task in the Indian queenless ant Diacamma indicum which relocate their colonies by means of tandem running. Individual ants were used as nodes and tandem runs as directed edges to construct unweighted networks. Network parameters were characterized in control relocations (CRs) and in relocations where the node with the highest outdegree, that is, the Maximum tandem leader (Max TL) was experimentally removed. These were then compared to 1) randomized networks, 2) simulated networks in which Max TL was removed, and 3) simulated networks with removal of a random leader. Not only was there complete recovery of the task, but the manner in which it was organized when Max TL was removed was comparable to CRs. The results obtained from our empirical study were significantly different from the results predicted by simulations of leader removal. At an individual level, the Max TL had a significantly higher outdegree than expected by chance alone and in her absence the substitute Max TL did comparable work. In addition, the position of the Max TL in the pathway of information flow was conserved in control and experimentally manipulated conditions. Understanding the organization of this critical event as more than the sum of individual interactions using network parameters allows us to appreciate the dynamic response of groups to perturbations.",
"discussion": "Discussion In the current study, we looked at the task organization and information flow in D . indicum colonies during the goal-oriented task of relocation with the help of network tools. Leaders are responsible for relocating their colonies as they tandem run nestmates one at a time to the new shelter. In particular, a single leader designated as the Max TL plays a pivotal role in the organization and execution of colony relocation ( Sumana and Sona 2013 ). We examined the position occupied by the Max TL in relocation networks. We also studied the effects of experimental removal of such a leader on the network structure and how it differed from the effects of the simulated removal of a single leader. The organization of the process of relocation was considerably different from random work distribution within the colony illustrating that results obtained from simulation studies should be interpreted with caution ( Christley et al. 2005 ; Pinter-Wollman et al. 2013 ). Fewer leaders were active in CRs than predicted by chance alone. Work distribution among these leaders was significantly more right skewed in CRs than in the RND networks where the interactions were distributed randomly with very few leaders initiating more than 5% interactions. Although most leaders in CRs led few tandem runs, there were a few who performed up to 45% of the total tandem runs. There are several underlying mechanisms operating within animal societies that make their structure and functioning significantly different from random organization ( Hamede et al. 2009 ; Naug 2009 ). Further studies will have to be conducted to understand the mechanisms that determine which individuals will become leaders during a given relocation and the amount of work they perform in D. indicum colonies. Lower average path length and diameter in CR indicate optimization of the paths of information flow within the colonies ensuring faster transfer of information among colony members. The higher outdegree centralization of CR networks as compared to RND networks indicate that task execution, relocation in this case, is dependent on only a small subset of colony members. Several short paths are maintained between all individuals of the colony during relocation as is signified by the higher outcloseness values of CR networks. These together ensure fast, efficient, and accurate transfer of information within all colony members which in turn will maintain colony cohesion during relocation. It has been speculated that colony size could affect task organization and the consequent pattern of information flow within social insect colonies. In general, individuals in smaller groups are more homogeneously connected to each other and heterogeneity in connectivity tends to increase as group size increases ( Fewell 2003 ; Naug 2008 ; Naug 2009 ). Spatial constraints prevent all individuals of large groups from interacting with each other at random giving rise to non random patterns of interaction within larger insect societies. These studies are based on non specific interactions such as proximity between individuals while we have used tandem running, a behavior that has specific functional significance. We observe the same connectivity patterns as has been reported earlier with most of the network parameters being negatively correlated to colony size in both control and randomized networks. Density decreases with increasing colony size indicating that relatively fewer tandem runs are required to relocate larger colonies. This could be due to the fact that followers are usually led only once to the new nest; hence, the increase in the actual number of ties is low compared to the number of potential ties as colony size increases. Decreasing centralization signifies that more leaders become involved in performing tandem runs as colony size increases. Diameter decreases in larger colonies while average path length is not correlated to colony size which indicates that individuals in larger colonies are equally well connected to each other as individuals in smaller colonies. Relocation becomes more complex as group size increases since more individuals have to be relocated to the new nest while at the same time chances of fragmentation of the colony also increases. In addition, there may be an upper limit to the number of tandem runs that can be performed by each leader. These constraints may result differences in patterns in work organization in colonies of different sizes ( Jeanne 1999 ; Naug 2009 ). Work distribution among leaders was right-skewed with the majority of the leaders performing few tandem runs while there were a few leaders who performed many tandem runs in the post removal period of MLR as well as in the corresponding period of CRs. This confirms previous observations in D. indicum ( Kaur et al. 2012 ; Sumana and Sona 2012 ) and is consistent with patterns of task distribution observed in other social insects ( Robinson 1992 ; Gordon 1996 ; Beshers and Fewell 2001 ). However, frequency distribution of tandem runs among leaders was more right-skewed in MLR than in CR signifying that work distribution became more unequal upon removal of only the Max TL. This is contrary to the response shown when a large number of leaders are removed as substitute leaders distribute the task more evenly among themselves ( Kolay and Annagiri 2015 ). However, removal of Max TL does not have any impact on the structure of relocation networks as all network parameters are comparable between CR and MLR indicating that work organization within the colony remains unaffected. In contrast, density and closeness was observed to increase when many leaders were removed during relocation ( Kolay and Annagiri 2015 ). Thus, the colony seems to respond to different degrees of stress in different ways to accomplish the same task. We further compared these results with the effects of simulated removal of a single leader, either the Max TL or a random leader and find that simulated networks were different from networks based on experimental data. Density and outcloseness are higher in both MLRA and RLRA simulations than MLR while network diameter, average path length, and outdegree centralization were comparable. This indicates that removal of any node with a nonzero outdegree value (leaders) produces a similar effect as removing a node with the highest outdegree value in the simulations. This interesting disparity with experimental removal required further investigation. This also illustrates that the empirical effects of removing an important individual from a social group could be different from that predicted by simulated removals ( Flack et al. 2006 ; Hamede et al. 2009 ).This could be due to the fact that biological systems have intrinsic mechanisms to cope with the loss of individuals and respond in an appropriate manner which are not clearly understood and, therefore, not accounted for while performing simulations. Most studies on task organization in social insect colonies focus on the behavior of groups of individuals performing the task but there is evidence to suggest that there is behavioral variability among individuals within this group rather than equal participation by all ( Kaur et al. 2012 ; Pinter-Wollman et al. 2012 ). Although nearly 20% of colony members become leaders during a single relocation event in D . indicum , all of them do not perform equally with few leaders carrying out most of the workload. This is further exemplified by the disproportionate role played by the Max TL. Not only does she perform significantly more tandem runs than other leaders, she also plays a crucial role in the process of information transfer among nestmates during relocation. This is indicated by the significantly higher outdegree and outcloseness values of Max TL in CRs than predicted by the RND simulations. The role of the Max TL during relocation is conserved as is indicated by the comparable outdegree and outcloseness of Max TL in both CR and MLR. However, the identity of the individual who assumes this position varies from one relocation to the next and persistence of an individual in the position of Max TL is for the duration of one relocation event in most cases. In fact, it is so flexible that even when the prospective Max TL is removed during the course of relocation, another leader can seamlessly take up this role without any discernible perturbations to task organization. This is evident from the fact that the outdegree and outcloseness of the individual who emerges as the Max TL in the post manipulative phase of MLR is comparable to that of the Max TL in the equivalent phase of CR. In fact, the performance of the emergent Max TL in MLR is better than that of the leader who performs highest number of tandem runs after simulated removal of the initial Max TL in MLRA and is comparable with performance of the Max TL after a random leader is removed in RLRA. However, there is a slight reduction in efficiency of relocation with an overall decrease in the rate at which tandem runs are performed and the time taken to complete relocation ( Sumana and Sona 2013 ). Thus, it seems that there are several individuals in the colony who are capable of becoming the Max TL during colony relocation. Further studies need to be carried out to elucidate the factors which determine the identity of the Max TL during a given relocation. The use of simulations and social network analysis in this study allows us to compare the consequences of targeted leader removal during relocations with observed results and contrast our findings with changes occurring by chance alone. The leaders organize the movement of the colony from the old nest to the new one in a manner that is very different from random and they also maintain this organizational structure in the absence of the most hardworking leader (Max TL). This study allows us to get a glimpse of task organization as more than the sum of individual events and its robustness to perturbations within social insect colonies."
} | 2,932 |
36823146 | PMC10236175 | pmc | 1,872 | {
"abstract": "Identification of novel, electricity-producing bacteria has garnered remarkable interest because of the various applications of electricigens in microbial fuel cell and bioelectrochemical systems. Shewanella marisflavi BBL25, an electricity-generating microorganism, uses various carbon sources and shows broader sugar utilization than the better-known S. oneidensis MR-1. To determine the sugar-utilizing genes and electricity production and transfer system in S. marisflavi BBL25, we performed an in-depth analysis using whole-genome sequencing. We identified various genes associated with carbon source utilization and the electron transfer system, similar to those of S. oneidensis MR-1. In addition, we identified genes related to hydrogen production systems in S. marisflavi BBL25, which were different from those in S. oneidensis MR-1. When we cultured S. marisflavi BBL25 under anaerobic conditions, the strain produced 427.58 ± 5.85 μl of biohydrogen from pyruvate and 877.43 ± 28.53 μl from xylose. As S. oneidensis MR-1 could not utilize glucose well, we introduced the glk gene from S. marisflavi BBL25 into S. oneidensis MR-1, resulting in a 117.35% increase in growth and a 17.64% increase in glucose consumption. The results of S. marisflavi BBL25 genome sequencing aided in the understanding of sugar utilization, electron transfer systems, and hydrogen production systems in other Shewanella species.",
"introduction": "Introduction With the global population growing exponentially, energy demand has increased remarkably, placing unprecedented pressure on the global ecosystem. To solve this issue, various renewable energy sources and production methods are currently under research [ 1 - 3 ]. Among them, microbial fuel cells and bioelectrochemical systems, which are recognized as novel biotechnologies for energy generation via sustainable biodegradation of chemical compounds using microorganisms as catalysts [ 4 , 5 ], have been promoted as a solution. Many electricigens, including archaea, Cyanobacteria, Firmicutes, Proteobacteria (α-Proteobacteria, β-Proteobacteria, γ-Proteobacteria, and δ-Proteobacteria), yeast, and eukaryotic algae oxidize organic compounds and transfer electrons to anodes [ 6 - 8 ]. Among these electricigens, Shewanella species are widely studied owing to their metal-reducing activity that reduces solid metal oxides [ 9 ]. The electrogenic facultative anaerobe Shewanella oneidensis MR-1 is a γ-Proteobacterium belonging to the order Alteromonadales [ 10 ]. Shewanella oneidensis MR-1 has multiple extracellular electron transport mechanisms, including direct electron transfer to electrodes via c-type cytochromes, electroactive biofilms, conductive nanowires/pili, and indirect transfer via mediators/shuttles. Based on these characteristics, this strain can reduce insoluble metal oxides and is a promising agent for bioremediating contaminated soils and water [ 11 , 12 ]. Apart from these abilities, Shewanella oneidensis MR-1 can also accumulate formate via lactate or pyruvate oxidation and produce hydrogen [ 13 , 14 ]. Although most microbial biohydrogen production occurs in Clostridium spp., Enterobacter spp., and Bacillus spp. [ 15 - 17 ], S. oneidensis MR-1 is also considered a potential hydrogen producer owing to its ability to oxidize carbon sources. Moreover, S. oneidensis MR-1 has various hydrogenases and could be engineered to produce biohydrogen [ 18 ]. Despite these advantages, some disadvantages hinder the application of S. oneidensis MR-1 [ 19 ]. One major disadvantage is the limited utilization of particular carbon sources, such as N -acetylglucosamine, lactate, and pyruvate [ 20 , 21 ]. Previously, we reported that the newly isolated S. marisflavi BBL25 strain could utilize multiple carbon sources, such as glucose, xylose, galactose, fructose, and lactose [ 22 , 23 ]. In addition, comparable amounts of electricity could be produced from lignocellulosic biomass. Thus, S. marisflavi BBL25 has more practical applications in hydrogen production than S. oneidensis MR-1 based on carbon source utilization. Through whole-genome analysis, we not only identified various carbon source-consuming pathways but also determined the associated electron transfer and hydrogen production systems. Based on these findings, we hypothesized that S. marisflavi BBL25 could produce hydrogen in the presence of exogenous electron acceptors using various carbon sources. Therefore, we aimed in this study to explore hydrogen production and carbon source metabolism in S. marisflavi BBL25 using whole-genome sequencing and demonstrate the hydrogen production system of S. marisflavi BBL25. Hence, we herein present the process from genetic information discovery to biohydrogen production and carbon source utilization by S. marisflavi BBL25.",
"discussion": "Discussion In this study, we previously found that S. marisflavi BBL25 can utilize various carbon sources and produce electricity. However, the mechanism involved remained unclear. Using genome sequencing, we demonstrated the existence of genes related to carbon source uptake, utilization systems, and electron transfer systems. We also identified the pathways for the biohydrogen-producing system in S. marisflavi BBL25. After whole-genome sequencing of this strain, we attempted to identify and confirm its potential as a hydrogen producer. In comparison to hydrogen production in S. oneidensis MR-1, S. marisflavi BBL25 had more formate dehydrogenase, which indicated that BBL25 had greater potential as a hydrogen producer. This was confirmed by hydrogen production experiments under anaerobic conditions. Moreover, by introducing the glk gene from S. marisflavi BBL25 into S. oneidensis MR-1, we observed glucose utilization and growth using glucose in S. oneidensis MR-1. Based on these findings, we verified that the glucose utilization system of S. marisflavi BBL25 worked, and the possibility of improving S. oneidensis MR-1 using a carbon source-utilizing gene from S. marisflavi BBL25 was confirmed. Thus, our study demonstrated various carbon sources and biohydrogen production systems, which offset the shortcomings of the existing representative strains. Furthermore, the whole-genome data generated in this study indicated the possibility of engineering this strain for higher electricity generation, hydrogen production, and carbon source utilization in other Shewanella species."
} | 1,619 |
36010832 | PMC9407070 | pmc | 1,874 | {
"abstract": "This paper addresses the problem of detecting multiple static and mobile targets by an autonomous mobile agent acting under uncertainty. It is assumed that the agent is able to detect targets at different distances and that the detection includes errors of the first and second types. The goal of the agent is to plan and follow a trajectory that results in the detection of the targets in a minimal time. The suggested solution implements the approach of deep Q-learning applied to maximize the cumulative information gain regarding the targets’ locations and minimize the trajectory length on the map with a predefined detection probability. The Q-learning process is based on a neural network that receives the agent location and current probability map and results in the preferred move of the agent. The presented procedure is compared with the previously developed techniques of sequential decision making, and it is demonstrated that the suggested novel algorithm strongly outperforms the existing methods.",
"conclusion": "6. Conclusions This paper considered the detection problem for multiple static and moving targets hidden in a domain, directly extending the classical Koopman search problem. Following previously developed methods, we addressed detection with both false-positive and false-negative detection errors. In the exploration stage, the suggested algorithm implements the deep Q-learning approach and applies neural network techniques for learning the probabilities of the targets’ locations and their motion patterns; then, in the exploitation stage, it chooses actions based on the decisions made by the trained neural network. The research suggested two possible procedures. In the first, called the model-free procedure, the agent detects the targets in the environment and simultaneously, online, learns the environment and trains a neural network that supports the agent’s decision-making processes. In the second procedure, called the model-based procedure, the agent begins detection only after offline learning and requires an exact model of the targets’ activity. The results obtained by maximizing the discounted cumulative expected information gain and by minimizing the expected length of the agent’s path demonstrate that the suggested algorithm outperforms previously developed information-based procedures and provides a nearly optimal solution even in cases in which the existing techniques require an unreasonable computation time. The proposed algorithms were implemented in the Python programming language and can be used both for further development of the methods of probabilistic search and detection and for practical applications in the appropriate fields.",
"introduction": "1. Introduction The detection of hidden stationary or moving targets is the first task of search procedures; this task focuses on recognizing target locations and precedes the chasing of the targets by the search agent [ 1 , 2 ]. Usually, the solution of the detection problem is represented by a certain distribution of the search effort over the considered domain [ 3 , 4 ]; for recent results and an overview of the progress in this field, see, e.g., [ 5 , 6 , 7 , 8 ]. In the simplest scenario of the detection of static targets by a static agent, it is assumed that the agent is equipped with a sensor that can obtain information (complete or incomplete) from all points in the domain. Using such a sensor, the agent screens the environment and accumulates information about the targets’ locations; when the resulting accumulated information becomes sufficiently exact, the agent returns a map of the domain with the marked locations of the targets. In the case of a moving agent, the detection process acts similarly, but it is assumed that the agent is able to move over the domain to clarify the obtained information or to reach a point from which the targets can be better recognized. A decision regarding the agent’s movement is made at each step and leads the agent to follow the shortest trajectory to achieve the detection of all targets. Finally, in the most complex scenario of moving target detection, the agent both moves within the domain to find a better observation position and tracks the targets to obtain exact information about each of their locations. It is clear that in the first scenario, the agent has a passive role, and the problem is focused not on decision making, but on sensing and sensor fusion. However, in the case of a moving agent, the problem focuses on planning the agent’s path. In recent decades, several approaches have been suggested for planning the agent’s motion and specifying decision-making techniques for detection tasks; for an overview of such research, see, e.g., [ 9 , 10 ]. Formally, such research addresses stochastic optimization methods that process offline and result in a complete agent trajectory or involve certain heuristic algorithms that allow the agent’s path to be planned in real time. In this research, we follow the direction of heuristic algorithms for search and detection with false positive and false negative detection errors [ 7 , 11 , 12 , 13 ] and consider the detection of static and moving targets. In addition, we assume that the agent is equipped with an on-board controller that is powerful enough to process deep Q-learning and train neural networks on relatively large data sets. Similar to previously obtained solutions [ 12 , 13 ], a data set is represented by an occupancy grid [ 14 , 15 ], and the decision making for the probability maps follows the Bayesian approach [ 7 , 8 ]. The implemented deep Q-learning scheme follows general deep learning techniques [ 16 , 17 ] applied to search and detection processes [ 18 ] and to navigation of mobile agents [ 19 ]. However, in addition to usual functionality, the suggested method utilizes the knowledge about the targets’ locations in the form of probability map. In the suggested algorithm, it is assumed that the agent starts with an initial probability map of the targets’ locations and makes decisions about its further movements either by maximizing the expected cumulative information gain regarding the targets’ locations or by minimizing the expected length of the agent’s trajectory up to obtaining the desired probability map. For brevity, we refer to the first approach as the Q-max algorithm and the second approach as the Shortest Path Length (SPL) algorithm. The maximization of the expected information gain and minimization of the expected path length are performed with a conventional dynamic programming approach, while the decision regarding the next step of the agent is obtained by the deep Q-learning of the appropriate neural network. As an input, the network receives the agent location and current probability map, and the output is the preferred move of the agent. The a priori training of the network is conducted on the basis of a set of simulated realizations of the considered detection process. Thus, the main contributions of the paper are the following. In contrast to known search algorithms with learning, the suggested algorithm allows search and detection with false positive and false negative detection errors, and, in addition to general deep learning scheme, the suggested algorithm utilizes the current agent’s knowledge about the targets’ locations. Note that both featured of the suggested algorithm can be used for solving the other problems that can be formulated in the terms of autonomous agents and probability maps. The algorithm and the training data set were implemented in the Python programming language with the PyTorch machine learning library. The performance of the algorithm was compared with the performance of previously developed methods. It was found that the novel deep Q-learning algorithm strongly outperforms (in the sense of obtaining the shortest agent path length) the existing algorithms with sequential decision-making and no learning ability. Therefore, it allows the targets to be detected in less time than the known methods.",
"discussion": "5. Discussion This paper presents a novel algorithm for the navigation of mobile agents detecting static and moving hidden targets in the presence of false-positive and false-negative errors. The suggested algorithm continues in the direction of previously developed procedures [ 12 , 13 ] for seeking and detecting hidden targets. However, in contrast to these procedures, which follow an immediate one-step decision making process, the proposed method implements the deep Q-learning approach and neural network techniques. The suggested algorithm is implemented in two versions: a procedure that maximizes the cumulative discounted expected information gain over the domain (Q-max algorithm) and a procedure that minimizes the expected path length of the agent in detecting all the targets (SPL algorithm). Formally, the first procedure is an extension of previously developed techniques based on the expected information gain calculated over the local neighborhood of the agent, while the second is a direct application of Q-learning techniques to the required value of the agent’s path length. The simulations show that after offline training of the neural network using the generated data set, the algorithm provides solutions that outperform the results obtained by the previously developed procedures, both in terms of the cumulative information gain and in terms of the agent’s path length. Moreover, the expected number of actions obtained by the Q-max algorithm by maximizing the cumulative discounted expected information gain is statistically equal to the number of actions obtained by the SPL algorithm by minimizing the expected path length. This equivalence follows directly from the nature of the problem: in terms of information, the detection of the targets means accumulating as much information as possible about the targets’ locations, and in terms of the path length, the detection of the targets means making as few movements as possible in order to specify the exact target locations."
} | 2,508 |
35515857 | PMC9063668 | pmc | 1,875 | {
"abstract": "Oil/water separation (OWS) technology has become an increasingly crucial tool to protect the environment and reduce the economic losses caused by the discharge of oily wastewater and oil spills. Recently, porous materials with superwettability have been applied in effective OWS and have achieved tremendous success. Herein, we review recent advancements of OWS utilizing femtosecond (fs) laser-structured superhydrophobic or underwater superoleophobic porous materials. We will review the enabling materials processing and treatment methods, their surface wettability, the separating methods and processes, and the separation mechanisms. Inspired by lotus leaves and fish scales, superhydrophobic and underwater superoleophobic properties are artificially achieved on substrate surfaces by fs laser processing. By using fs laser-structured superwetting porous materials, various oil/water mixtures (OWMs) are successfully separated through different separation methods. Presently, the research of fs laser-based OWS is still in its infancy. We will also discuss the current challenges and future prospects in this emerging field. It is expected that the advanced features of fs laser microfabrication will lead to exciting applications for OWS.",
"conclusion": "4. Conclusions and perspective OWS is a global challenge in the aspect of protecting ecological environment and reducing the economic loss caused by the oil spill and the discharge of oily wastewater. Recently, porous materials with superhydrophobicity or superoleophobicity have been developed to efficiently separate various OWMs, because OWS is actually an interfacial issue. Due to the features of small heat-affected zone, high spatial resolution, extensive material processing, and non-contact manufacturing, fs laser microfabrication has been successfully applied to design extreme wettability on various material surfaces. Such technology is able to not only process a broad range of substrate materials but also create both large-area micro/nanoscale hierarchical structures and perforating microholes array on a thin film. In this review, we mainly summarize the recent applications of the fs laser-structured superhydrophobic or underwater superoleophobic porous materials in OWS. Inspired by lotus leaf and fish scale, superhydrophobicity and underwater superoleophobicity are respectively achieved on the surfaces of various materials by combining fs laser-designed surface microstructures and hydrophobic/hydrophilic chemistry. The diversity of the surface wettabilities of the fs laser-structured materials endows these materials with the ability to separate OWMs in a variety of ways. The practical oil/water separating application of the fs laser-induced superwetting porous materials is introduced in detail, including the wettability, separating method/operation, and the separation mechanism of each case. The research related to the fs laser-structured oil/water separating materials was first reported in 2016, and is currently still in its infancy. Many problems and challenges should be solved in the coming years. Firstly, the efficiency of fabricating superwetting porous materials needs to be improved because it is important for large-scale real OWS. More advanced laser-ablation manner ( e.g. , processing based on a cylindrical focusing lens, automatic parallel processing) and the laser system with higher power are eagerly awaited. Secondly, poor mechanical/chemical durability usually results in the decline of the surface superwettability as well as the oil/water separating capacity of the separating materials after several cycles of use. Endowing those materials with durable superhydrophobicity and superoleophobicity can extend their service life. Thirdly, the oil contaminants in a real condition are usually very complex rather than the pure oily liquids that are used in the proof-of-concept separation experiments. For example, high-viscosity and high-density oils will potentially foul the separating materials or block the micropores of the substrate, weakening the separation ability of the fabricated materials. There also exist many different kinds of oil/water emulsions. The materials and methods for separating the complex mixtures of water and high-viscosity oils and the emulsions are required to be further developed. Fourthly, more researches need to be focused on understanding the dynamic interaction between the superwetting porous substrate and the OWM. Such interaction mechanism has a positive role in designing and fabricating various separating materials, as well as the use of those materials to achieve efficient OWS. Finally, and perhaps most importantly, researchers should attempt to bring the fs laser-induced superwetting porous materials to market and design practical large instrument by using those oil/water separating materials as the core components in the future, practically preventing the environmental pollution caused by the discharge of oily industrial wastewater and oil spills. It could be expected that the development trend of OWS will have an explosive growth towards the design of materials and the separating methods. We believe that the advanced features of fs laser microfabrication will lead to an exciting future in the field of OWS.",
"introduction": "1. Introduction Oil/water separation (OWS, i.e. , the separation of oil and water) has become an urgent issue and a worldwide challenge because of the ever increasing frequency and amount of oily wastewater discharge and oil spill accidents. 1–6 The oil pollution not only seriously destroys the ecological environment but also causes huge economic loss. 5–9 A typical example is the Gulf of Mexico oil spill, which happened in 2010 and caused about 2 × 10 8 gallons of crude oil to be leaked on the sea surface ( Fig. 1a and b ). 5,10 Such an unprecedented environmental catastrophe has resulted in immeasurable damage to the marine ecosystem as well as public health ( Fig. 1c and d ). To protect the environment and reduce economic loss, there is an urgent need for developing advanced materials and technologies for effective OWS. Conventional materials and methods ( e.g. , absorption, gravity separation, flotation, skimming, and centrifugation) used to solve those oil-pollution problems often suffer from many limitations, such as low separation efficiency, high cost, low selectivity, the need of driving energy, and the generation of secondary pollutants. 1,3 Recently, porous materials with opposite superwettabilities to water and oil are successfully applied in the field of OWS. 11–33 Those materials usually have both superhydrophobicity and superoleophilicity, or have both superoleophobicity and superhydrophilicity. For instance, Feng et al. firstly fabricated a superhydrophobic film ( i.e. , rough Teflon-coated metal mesh) that also had superoleophilicity. 11 The mesh successfully separated the oil/water mixture (OWM, i.e. , the mixture of oil and water). The porous film allowed oil to permeate through but completely intercepted the water phase. Xue et al. prepared an underwater superoleophobic porous film ( i.e. , nanostructured hydrogel-coated metal mesh) and also achieved OWS by using such a mesh. 19 When poured OWM onto the water-wetted mesh, the water phase could wet and penetrate through the mesh because of the superhydrophilicity, while the oil phase always remained above the mesh. Inspired by the above-mentioned research works, a large number of porous materials with superhydrophobicity/superoleophilicity or superoleophobicity/superhydrophilicity have been developed in order to achieve OWS. 12,13,16,21,22,24,25,34–46 Fig. 1 Gulf of Mexico oil spill accident. (a and b) Leaked crude oil covering on the ocean surface. (c and d) Seabirds and sea turtles being killed by the leaked crude oil. Reproduced from ref. 5 with permission from RSC, copyright 2018. With the advent of a femtosecond (fs) laser, the microfabrication based on the fs laser has been rapidly applied in the field of advanced nano/microfabrication and other modern manufacturing. 47–55 Such technology has many definite advantages in designing surface microstructures, including small heat-affected zone, high spatial resolution, non-contact manufacturing, etc. 56–59 In particular, fs laser can process almost all of the known materials and directly create micro/nanoscale structures on the surfaces of various kinds of materials ( e.g. , semiconductors, metals, polymers, glasses, and ceramics) through simple one-step ablation. 56–60 In addition, the fs laser can not only generate uniform large-area rough microstructures on the surface of a substrate but also drill microholes array through a thin film. 28,47,61,62 In recent years, fs laser microfabrication obtains great success in adjusting the surface wettability of different solid materials, because wettability is mainly depended on the surface microstructure and chemical composition of the material surfaces. 47,48,63–72 Various superhydrophobic surfaces and underwater superoleophobic surfaces have been prepared through fs laser processing. 47–55,61,73–77 The fs laser-structured superwetting porous materials also have important applications in OWS. 16,28,62,78–81 In this paper, we will review the recent advancements of femtosecond laser-induced superhydrophobic or underwater superoleophobic porous materials and their applications in OWS. This review starts with the introduction of the significance and urgency of performing highly efficient OWS (Section 1). Then, the theoretical basis of wettability, the features of fs laser microfabrication, and the typical separating ways based on the superwetting porous materials are presented as the related background (Section 2). The next part shows the practical applications of the fs laser-induced superwetting porous materials towards OWS (Section 3). We will focus on discussing the wettability of the separation materials, the separation method and operation, and the separation mechanisms. In the end, the current challenges in the field of OWS using the fs laser-induced superhydrophobic or superoleophobic porous materials will also be discussed (Section 4)."
} | 2,546 |
27252693 | PMC4879347 | pmc | 1,877 | {
"abstract": "Anaerobic digestion (AD) is a microbial process widely used to treat organic wastes. While the microbes involved in digestion of municipal sludge are increasingly well characterized, the taxonomic and functional compositions of AD digesters treating industrial wastewater have been understudied. This study examined metagenomes from a biogas-producing digester treating municipal sludge in Shek Wu Hui (SWH), Hong Kong and an industrial wastewater digester in Guangzhou (GZ), China, and compared their taxonomic composition and reconstructed biochemical pathways. Genes encoding carbohydrate metabolism and protein metabolism functions were overrepresented in GZ, while genes encoding functions related to fatty acids, lipids and isoprenoids were overrepresented in SWH, reflecting the plants’ feedstocks. Mapping of genera to functions in each community indicated that both digesters had a high level of functional redundancy, and a more even distribution of genera in GZ suggested that it was more functionally stable. While fermentation in both samples was dominated by Clostridia , SWH had an overrepresentation of Proteobacteria , including syntrophic acetogens, reflecting its more complex substrate. Considering the growing importance of biogas as an alternative fuel source, a detailed mechanistic understanding of AD is important and this report will be a basis for further study of industrial wastewater AD.",
"introduction": "Introduction Anaerobic digestion (AD) is a biological decomposition process widely used in municipal wastewater treatment plants (WWTPs). Globally, biogas-producing AD processes are gaining attention because they can not only degrade organic waste, which reduces water quality and poses a danger to public health if not properly treated ( Sahlstrom , 2003 ; Bibby and Peccia, 2013 ), but also provide a renewable source of energy in the form of methane (biogas; Angelidaki and Ellegaard, 2003 ; Weiland, 2003 ; Luo et al., 2013 ). Despite its widespread use worldwide, the biological mechanisms of AD are still poorly understood, mostly due to the complexity of the microbial communities involved ( Wagner and Loy, 2002 ; Nelson et al., 2011 ). Thus, detailed studies of the composition of AD microbial consortia and their metabolic functions are required. An improved understanding of AD could enhance the efficiency of carbon recovery from waste streams, contributing to the global goal of turning WWTPs into sustainable systems ( Jetten et al., 1997 ). Identifying the microorganisms in AD systems has traditionally been accomplished by the construction of 16S rRNA gene clone libraries followed by Sanger sequencing ( Riviere et al., 2009 ), which in recent years has been replaced by high-throughput next-generation sequencing of 16S rRNA gene amplicons ( Schluter et al., 2008 ; Werner et al., 2011 ). However, both these methods focus on taxonomic identification, and the metabolic pathways present can be determined only indirectly. Moreover, these methods can introduce biases as a PCR amplification step is required. Shotgun metagenomic sequencing, which directly sequences the extracted DNA, can provide more detailed information on the identity of the microbes and their metabolisms as well as other biological information including novel genes ( Ferrer et al., 2005a , b ). In addition to these advantages, metagenomic sequencing is gaining importance in the study of microbial communities because the decreasing cost and increasing sequencing depth have enabled high-resolution analysis of complex environmental samples ( Qin et al., 2010 ; Fierer et al., 2012 ) such as sea water ( Tang et al., 2013 ), soils ( Fierer et al., 2012 ), human gut ( Qin et al., 2010 ), and freshwater ( Breitbart et al., 2009 ). Metagenomics was first applied to AD in 2008 with the analysis of a German full-scale biogas plant treating farm waste ( Schluter et al., 2008 ). Several further studies have since examined AD metagenomes, with the focus mainly on taxonomy and gene-centric functional analyses ( Wirth et al., 2012 ; Wang et al., 2013 ; Wong et al., 2013 ; Yang et al., 2014 ). Recently, metagenomic analysis has shifted toward reconstructing important metabolic pathways and genomes present in AD systems ( Li et al., 2013 ). Of the AD metagenomes analyzed to date, samples have been obtained from full-scale biogas plants treating farm waste ( Schluter et al., 2008 ), industrial ( Wang et al., 2013 ) and municipal ( Wong et al., 2013 ; Yang et al., 2014 ) sludge digesters, and lab-scale reactors ( Wirth et al., 2012 ; Li et al., 2013 ). However, little analysis has yet been conducted of full-scale AD systems treating high-strength industrial wastewater. While municipal sludge digesters are important, AD systems treating high-strength wastewater should not be neglected because a substantial volume of industrial wastewater is generated every year. In China alone, the discharge of industrial wastewater was 6.9 × 10 10 metric tons in 2012 and estimated to be 7.8 × 10 10 metric tons in 2015, accounting for as high as 35% of total national wastewater discharge ( Feng et al., 2015 ). This study aimed to determine whether and how the major AD processes and the taxonomic groups performing them differ in a high-strength industrial wastewater system compared to better-studied municipal sludge systems. We obtained metagenomes from one system of each type and reconstructed and compared their key AD metabolic pathways. The major taxonomic groups performing these processes were determined and compared between the two systems, and pathways with particular functional significance analyzed. The redundancy of organisms in the major pathways of the multi-step AD process was also examined.",
"discussion": "Discussion Anaerobic digestion has been widely applied in the treatment of municipal sludge ( Riviere et al., 2009 ; Yang et al., 2014 ) and industrial wastewater ( Rajeshwari et al., 2000 ) as this technology decomposes organic waste while simultaneously producing biogas ( Angelidaki and Ellegaard, 2003 ; Weiland, 2003 ; Luo et al., 2013 ). However, previous AD studies have mainly focused on the phylogenetic diversity of municipal digesters treating waste sludge from secondary treatment ( Yang et al., 2014 ) or farm waste ( Schluter et al., 2008 ). The taxonomic and functional composition of AD digesters treating high-strength industrial wastewater have not been extensively studied, especially those taxa and functions involved with the major AD steps ( Wagner and Loy, 2002 ; Nelson et al., 2011 ). Given the vast and increasing volume of high-strength industrial wastewater produced worldwide, the application of AD to wastewater is likely to grow. Therefore, it is important to thoroughly understand the microbiology involved in the anaerobic treatment of wastewater and how it differs from the more common sludge digesters. We analyzed metagenomes (i.e., covering both taxonomy and metabolic functions) from an industrial wastewater AD system and a municipal system treating sludge to reconstruct the major AD biochemical processes and examine how they differ between these two digester types. In this study, we chose not to assemble the reads before annotation, as low-coverage sequences might be excluded by an assembler ( Howe et al., 2014 ) and bias the taxonomic profile against rare taxa. Unlike some studies using unassembled reads ( Fierer et al., 2012 ), here we used merged paired end reads for more reliable annotation results. However, the protein annotation rates (53.4% for GZ and 48.3% for SWH) were similar to reported rates for unmerged reads (42.7–56.1%; Fierer et al., 2012 ; Wirth et al., 2012 ). We also assembled contigs to provide a basis for phylogenetic comparison between the metagenomes, although these were not used to generate taxonomic and functional abundances. The annotation rates in these contigs were < 68%, suggesting incomplete sequencing of genes played some role in the annotation rate but did not account for all unannotated reads. Unannotated reads could also be attributed to the high diversity of the samples ( Wilkins et al., 2015a ) and/or incomplete databases ( Ye et al., 2012 ). The presence of functional genes in an AD system should generally be correlated to the substrate it is treating. For example, a metagenome from anaerobic digestion of tannery wastewater found genes assigned to protein metabolism were the most abundant, making up about 15% of the metagenome ( Wang et al., 2013 ). In the metagenome from the GZ digester, which treats carbohydrate-rich beverage wastewater, genes related to carbohydrate transport and metabolism were the most abundant and strongly overrepresented relative to SWH ( Figure 2 ; Supplementary Figure S3 ; Hu et al., 2010 ). In contrast, functions related to metabolism of lipids and fatty acids were significantly overrepresented in SWH, including the SEED subsystem for fatty acids, lipids, and isoprenoids ( Figure 2 ), COG category for lipid transport and metabolism (Supplementary Figure S3B ) and genes encoding enzymes from the cytochrome P450 family (Supplementary Figure S4 ). Cytochrome P450 enzymes catalyze reactions on a broad range of substrates, particularly lipid metabolites such as steroids, eicosanoids, fatty acids, and lipids ( Coon et al., 1992 ). This functional profile of SWH suggests there was a higher lipid content in the sludge being treated there, as expected for a municipal sewerage waste stream. Similarly, the overrepresentation of SEED subsystems related to the metabolism of aromatic compounds ( Figure 2 ) reflects the probable presence of polycyclic aromatic hydrocarbons (PAHs) in sewerage waste streams, particularly concentrated in digester sludge ( Blanchard et al., 2004 ). Genes encoding the cobalt-zinc-cadmium resistance protein CzcA were abundant in both metagenomes. Cobalt-zinc-cadmium resistance protein CzcA is essential for resistance against certain heavy metals ( Diels et al., 1995 ). Its presence suggests heavy metals such as cobalt, zinc and/or cadmium, could be present in both samples. A previous study has shown that heavy metals are released with the decomposition of organic matter ( Dong et al., 2013 ). The taxonomic composition of the two digesters also reflected the complexity of hydrolytic functions required to process their feedstocks. The phylum Proteobacteria was strongly overrepresented in SWH, with the exception of the class delta-Proteobacteria ( Figure 1B ). This overrepresentation was quite evenly spread over a large number of Proteobacteria genera (data not shown), indicating it was not due to a few exceptional species but rather a systematic difference. Members of the Proteobacteria have a broad range of roles in all AD steps except for methanogenesis. In the pathway reconstruction, Proteobacteria genera were among the top contributors to the production of all major fermentation products except lactate and to acetogenesis in both digesters ( Figure 3 ). Given this broad range of roles in both digesters, the overabundance of Proteobacteria in SWH reflects the more complex range of substrates treated by that system, requiring a more diverse repertoire of functions. Bacteria in both metagenomes were numerically dominated by reads assigned to the genus Clostridium , which can ferment a wide variety of carbon sources and produce VFAs and alcohols that serve as substrates for methanogenesis. A high Clostridium abundance has been reported in a range of anaerobic reactors ( Nelson et al., 2011 ) including those fed with crops and a mixture of animal manure ( Rademacher et al., 2012 ) or excess sludge ( Li et al., 2013 ), and in our previous amplicon sequencing-based studies of the GZ and SWH digesters ( Wilkins et al., 2015b ; Jia et al., 2016 ). In this study, the class Clostridia was slightly overrepresented in the GZ metagenome ( Figure 1 ). Mapping of genera to the formation of AD intermediates found that Clostridium spp. were the main fermenters forming four of the six major acidogenesis products (formate, acetate, butyrate, and lactate; Figure 3 ), consistent with their reported versatility in AD fermentation ( Li et al., 2011 ). Members of the Clostridiales also have roles in initial hydrolysis ( Moon et al., 2011 ; Dassa et al., 2014 ) and syntrophic acetate oxidation (SAO; Müller et al., 2013 ). However, it is notable that in our previous study of enrichment cultures inoculated from both digesters Clostridium abundance increased in tandem with the methanogen genus Methanobacterium ( Jia et al., 2016 ); in this study, the class Methanobacteriales was likewise slightly overrepresented in GZ (Supplementary Table S4 ). SAO bacteria, including Clostridium ultunense (<0.1% in both samples; Müller et al., 2013 ), oxidize acetate to provide H 2 and CO 2 to syntrophic partner methanogens such as Methanobacterium ( Zinder and Koch, 1984 ), and it is possible that at least some methane production from acetate in the digesters proceeds via this route. As SAO bacteria likely oxidize acetate via the reversible Wood–Ljungdahl pathway (reductive acetyl-CoA pathway; Lee and Zinder, 1988 ) also used in other AD processes, e.g., methanogenesis, the presence or absence of this route cannot be determined by the presence or absence of marker genes. As expected for a sewage waste stream, SWH contained a high abundance of human-associated pathogen genera such as Mycobacterium ( Mycobacterium tuberculosis, M. bovis and M. avium accounted for ∼72% of Mycobacterium in both samples) and Burkholderia , although Mycobacterium was also unexpectedly abundant in GZ. This high abundance of Mycobacterium in both digesters is especially noteworthy as they are unlikely to be important in the degradation of organic compounds under anaerobic conditions ( de Oliveira et al., 2010 ). As removal of Mycobacterium at mesophilic temperatures typically requires weeks to months ( Sahlstrom, 2003 ; El-Mashad et al., 2004 ), the high abundance detected in both systems suggests neither is successful in removing these potential pathogens with the retention time of the systems. Methanomicrobiales was the most abundant archaeal order in both metagenomes (Supplementary Table S4 ), and contributed most of the abundant genera mapped to the methanogenesis pathway from both digesters ( Figure 3 ; Supplementary Figure S5 ). The orders Methanosarcinales and Methanobacteriales were also abundant, with both overrepresented in GZ (Supplementary Table S4 ). Our previous study using amplicon sequencing of the archaeal rRNA gene ( Wilkins et al., 2015a ) found that Methanomicrobiales was dominant in GZ but order Methanosarcinales in SWH, while a previous metagenomic study of SWH sludge found an overwhelming (>70%) dominance of Methanomicrobiales . These different results may reflect differences in methods, for example the short read length of shotgun metagenomic sequencing leading to conflation of protein sequences from the closely related orders. They could also reflect functional redundancy in the methanogenesis step of AD, i.e., maintenance of the biochemical functions over time independent of variance in taxonomic composition. Such redundancy has been proposed as a feature of higher AD steps, particularly fermentation ( Werner et al., 2011 ), and our previous study of AD enrichment cultures has shown that similar methane yields can be obtained from systems fed with different substrates and containing diverse methanogen communities ( Wilkins et al., 2015b ). Detailed reconstruction of the methanogenesis pathway (Supplementary Figure S5 ) found that the functional potential for complete acetoclastic and hydrogenotrophic pathways were present in both metagenomes, with genes encoding enzymes involved in the acetoclastic route dominant in both samples. This agrees with our previous amplicon sequencing study ( Wilkins et al., 2015a ) and other previous studies that found the acetoclastic pathway tends to be dominant in methanogenic systems ( Li et al., 2013 ; Yang et al., 2014 ). The time series of SWH metagenomes also suggested greater taxonomic than functional variability, with e.g., the class Methanomicrobia increasing 1.4-fold in relative abundance between September 2011 and March 2012 while functional abundances remained relatively constant (Supplementary Figure S7 ). As different microbes share similar functions ( Hashsham et al., 2000 ), it is quite likely that the presence of functional genes, rather than particular microbial taxa, determines the functional stability of the AD digesters. The presence of the same core functions in each digester contributed by different consortia was also illustrated by the reconstruction of the denitrification pathway, with almost non-overlapping sets of genera responsible for the same biochemical steps in each digester (Supplementary Figure S6 ). The abundance of the genus Methanosaeta was unusually low in both digesters compared to other reported anaerobic environments, where it is often the most abundant methanogen genus ( Ariesyady et al., 2007 ; Nelson et al., 2011 ). Our previous amplicon-sequencing based study found the Methanosaetaceae to outnumber Methanosarcinaceae in both GZ and SWH ( Wilkins et al., 2015a ). Variation in the relative proportions of Methanosaeta and Methanosarcina in anaerobic digesters has been linked to operating parameters such as the frequency of substrate feeding ( Conklin et al., 2006 ) and retention time ( Ma et al., 2013 ), as well as biochemical factors including H 2 partial pressure and the presence of heavy metals ( Yilmaz et al., 2014 ). Pairwise comparison of the contigs assembled from GZ and SWH found that the most similar sequences were mainly methanogens (based on NCBI nr annotation), particularly the strain Methanosaeta concilii GP6 (similarity >99%), indicating the two systems shared overall similar methanogen compositions. In addition to analyzing the distribution of phylogenetic and functional genes as in previous studies ( Schluter et al., 2008 ; Wirth et al., 2012 ; Wang et al., 2013 ) or describing the dominant pathways ( Li et al., 2013 ), functional AD pathways were reconstructed and the relative abundances of genera from both metagenomes that may be performing these functions mapped on to these pathways. Genera associated with acidogenesis were less evenly distributed (Supplementary Table S5 ) than those associated with other steps. This was unexpected, as acidogenesis populations are more reliant on functional redundancy ( Werner et al., 2011 ), while syntrophic bacteria, which are much less abundant than other functional bacteria in AD process, are sensitive to environmental change ( McInerney et al., 2009 ; Werner et al., 2011 ). The more even distribution of genera in GZ (Supplementary Table S5 ) indicates that the community should be more functionally stable, as the presence of more parallel pathways provides resilience against fluctuations in substrate loading ( Hashsham et al., 2000 ). This has been experimentally verified in both lab-scale ( Hashsham et al., 2000 ) and full-scale ( Werner et al., 2011 ) reactors, and it has been further shown that evenness rather than richness is the key factor in preserving the functional stability of an ecosystem ( Wittebolle et al., 2009 ). In studies of anaerobic reactors with differing evenness, Werner et al. (2011) found that communities with greater evenness had higher methanogenic activity, suggesting that GZ may have higher methanogenic potential than SWH. Manipulating community evenness (e.g., by transient disturbances; Cabrol et al., 2016 ) may be a practical strategy for optimizing biogas production. This study compared municipal wastewater and sludge metagenomes. Despite having different taxonomic profiles, GZ and SWH shared mostly similar potential microbial functions, and of the major functional differences most could be related directly to the digester feedstocks. Mapping of taxa to the major metabolic pathways in AD allowed the major functional taxa in each digester to be determined, and the more even distribution of genera performing major AD functions in GZ suggested a stronger adaptive capability than in SWH (functional stability). We found the metagenome of GZ was similar to that of a production-scale biogas plant both at the phylogenetic and functional level, confirming the biogas-producing potential of industrial wastewater AD. While this study of a single high-strength industrial wastewater AD system may not be globally representative, it strongly suggests that there are major functional differences compared to the better-studied municipal sludge systems that can be directly linked to feedstock, and provides a basis for further investigation of industrial wastewater AD. Future studies should also further explore the AD microbial community with metatranscriptomic and metaproteomic analyses to better understand the metabolic functions."
} | 5,265 |
29372184 | PMC5775026 | pmc | 1,878 | {
"abstract": "A process for converting fructose to 2,5-furandicarboxylic acid, a monomer used in the production of a renewable plastics.",
"introduction": "INTRODUCTION Biomass is an abundant renewable carbon source for the sustainable supply of valuable intermediates for the production of fuels, chemicals, and bio-based plastics ( 1 ). Furan-2,5-dicarboxylic acid (FDCA) is an important renewable building block because of its potential as a substitute for a variety of petrochemicals, such as terephthalic acid and adipic acid ( 2 ). Potential applications for FDCA include polyesters, polyurethanes, and polyamides ( 3 ). A bio-based polymer of commercial interest is polyethylene furanoate (PEF), a copolymer of ethylene glycol and FDCA. PEF has improved mechanical properties compared to polyethylene terephthalate, such as higher glass transition temperature and improved tensile modulus ( 4 ). In addition, PEF has better gas barrier properties for oxygen [9× lower ( 5 )], carbon dioxide [11× lower ( 6 )], and water vapor [2× lower ( 7 )]. Avantium has shown that PEF can be used for the production of water bottles, food packaging, sports apparel, and footwear. Accordingly, a strategic consortium of global companies is developing technology for the production of this bioplastic ( 8 ). Economic production of bio-based ethylene glycol from carbohydrates has been recently achieved, and as such, the current limitation for commercial production of 100% bio-based PEF is the economical production of pure FDCA. The general reaction scheme for the production of FDCA is shown in Scheme 1 . Fructose [(2 R ,3 S ,4 S ,5 R )-2,5-bis(hydroxymethyl)oxolane-2,3,4-triol] is selectively dehydrated to produce 5-(hydroxymethyl)furan-2-carbaldehyde (HMF), which is subsequently oxidized to yield FDCA. Unfortunately, as noted by various investigators ( 9 – 11 ), economical production of FDCA must address technological challenges, such as (i) instability of HMF caused by undesirable condensation reactions at moderate temperatures (for example, 373 K), especially in the presence of an acid and/or a base; (ii) low solubility of FDCA in commonly used solvents; and (iii) incomplete oxidation of HMF toward FDCA, producing small quantities of partially oxidized intermediates, that is, 5-formylfuran-2-carboxylic acid (FFCA) and 5-(hydroxymethyl)furan-2-carboxylic acid (HMFCA), that terminate the chain growth during downstream polymerization process, leading to a poor-quality polymer ( 10 ). Avantium has developed a homogeneous process that uses methoxymethylfurfural (obtained from dehydration of fructose in methanol) as the substrate; acetic acid as a solvent; and Co(OAc) 2 , Mn(OAc) 2 , and HBr as the homogeneous catalyst system to overcome the abovementioned challenges to produce FDCA ( 11 ). The homogeneous process of HMF oxidation is effective, yielding more than 96% FDCA; however, a small fraction of FFCA is retained with FDCA in solid state and requires further purification by catalytic hydrogenation of FFCA. This step requires control of hydrogenation conditions to reduce/avoid reduction of FDCA ( 12 ). Moreover, catalyst recovery and recycle require additional unit operations, and a process using a heterogeneous catalyst is desired. Scheme 1 General reaction scheme for the production of FDCA from fructose. Oxidation of HMF has been extensively investigated over various heterogeneous catalysts ( 13 – 21 ). Although high yields of FDCA are reported for HMF oxidation in water, there are potential problems with this approach. First, production of HMF by dehydration of fructose in water is not effective because of excessive HMF degradation, leading to low HMF yield (<20%) ( 22 ). Organic solvents, such as dimethyl sulfoxide ( 22 ) and γ-valerolactone [5-methyldihydrofuran-2(3 H )-one] (GVL) ( 23 ), have been used to achieve high yields of HMF (>70%). However, these solvents have high boiling points, and the high reactivity of HMF at elevated temperatures prohibits the use of traditional separation techniques, such as distillation, for separation and purification of HMF ( 24 ). Another limitation of HMF oxidation in water is that FDCA has low solubility in water, and stoichiometric amounts of base (for example, sodium hydroxide) are required to keep FDCA dissolved in solution and prevent its precipitation onto the catalyst surface. The requirement of a homogeneous base is disadvantageous because mineral salt (for example, NaCl) is produced along with FDCA. Here, we present a strategy for the conversion of fructose to HMF and the subsequent oxidation of HMF to FDCA in an organic solvent mixture using a heterogeneous catalyst. In our proposed process, HMF is obtained at high concentrations by fructose dehydration, and it is subsequently oxidized over a heterogeneous catalyst to produce FDCA at high yields. We show that the heterogeneous catalyst is stable and can be recovered and recycled, a significant advantage over previous processes based on homogeneous catalysis. We demonstrate that the use of a GVL/H 2 O solvent system enables the formation of HMF at high yields and at high concentrations from fructose dehydration, and the high solubility of FDCA in this solvent system at reaction temperatures (for example, 383 K) enables the oxidation of HMF to FDCA at high concentrations. This latter attribute of the GVL/H 2 O solvent system alleviates the need for a homogeneous base, such as NaOH or Na 2 CO 3 , to enhance FDCA solubility, thereby eliminating the production of undesired salts along with the production of FDCA. Our process also replaces corrosive mineral acids, such as hydrochloric acid and sulfuric acid, which are frequently used as catalysts for fructose dehydration, with an organic acid, FDCA, which is also the desired product of the process. Moreover, we demonstrate that FDCA can be separated from the GVL/H 2 O solvent system by cooling the reaction system, leading to crystallization and facile product separation and solvent recycling. Furthermore, we show, using techno-economic analysis (TEA), that the proposed process is a cost-effective route for the synthesis of FDCA.",
"discussion": "RESULTS AND DISCUSSION Carbon-supported platinum catalysts (Pt/C) have been studied in the literature for the oxidation of HMF to FDCA in aqueous medium both in the presence ( 17 ) and absence ( 16 ) of a homogeneous base. Accordingly, we studied HMF oxidation in the GVL/H 2 O solvent system over a Pt/C catalyst without using a homogeneous base. Oxidation of HMF catalyzed by 5% Pt/C at 383 K under an oxygen pressure of 40 bar for 20 hours leads to complete conversion of HMF and 95% yield of FDCA ( Table 1 , entry 1). Table 1 Results for HMF oxidation reactions. Reaction was carried over the 5% Pt/C catalyst (under 40-bar O 2 pressure and 383 K). DFF, furan-2,5-dicarbaldehyde. # HMF concentration Solvent (GVL/H 2 O) HMF/Pt* Time (hours) HMF conversion (%) DFF yield (%) FFCA yield (%) FDCA yield (%) 1 0.5 wt % 80:20 15:1 20 97 – – 95 2 5 wt % 80:20 20:1 20 100 9 31 11 3 7.5 wt % 50:50 30:1 20 100 – – 94 4 7.5 wt % F-D HMF 50:50 30:1 16 100 – – 0 5 7.5 wt % F-D HMF using HCl + ion exchange + humin adsorption 50:50 30:1 16 100 – – 93 6 7.5 wt % F-D HMF using FDCA + humin adsorption 50:50 30:1 16 100 – – 91 *Molar ratio of HMF to platinum. The stability of the Pt/C catalyst was investigated in a continuous tubular fixed bed reactor with 0.5 weight % (wt %) HMF feed in the GVL/H 2 O (80:20) solvent system. Under the specified reaction conditions, complete HMF conversion, 60% yield of FDCA and 38% yield of FFCA, was achieved. Figure 1A shows that the Pt/C catalyst is stable for the oxidation of HMF in the GVL/H 2 O solvent system. Thus, HMF can be selectively oxidized to FDCA in the GVL/H 2 O solvent system where it can be produced at high yields from fructose. Fig. 1 HMF oxidation, FDCA solubility, and fructose dehydration. ( A ) HMF oxidation over 5 wt % Pt/C. 0.5 wt % HMF in GVL/H 2 O (80:20) solution; temperature, 373 K; pressure, 40 bar; 5 wt % Pt/C, 2.0 g; solvent flow rate, 0.05 ml/min; O 2 flow rate, 20 ml/min. Black squares represents FDCA yield. Red circles represents FFCA yield. ( B ) HMF oxidation over 5 wt % Pt/C. 1.0 wt % HMF in GVL/H 2 O (50:50) solution, temperature, 373 K; pressure, 40 bar; 5 wt % Pt/C, 2.0 g; solvent flow rate, 0.02 ml/min; O 2 flow rate, 25 ml/min. Black squares represent FDCA yield. Red circles represent FFCA yield. ( C ) FDCA solubility as a function of GVL concentration. Red circles represent solubility of FDCA at 303 K. Red triangles represent solubility of FDCA at 373 K. Black squares represent heat of mixing of GVL and H 2 O. ( D ) FDCA solubility as a function of temperature. Red circles represent GVL/H 2 O (50:50). Black squares represent H 2 O. ( E and F ) Fructose conversion and HMF yield for fructose dehydration at 453 K. Black squares represent fructose dehydration using 3 mM HCl. Blue triangles represent fructose dehydration using 0.53 wt % FDCA. Red diamonds represent FDCA stability under dehydration reaction. Solid lines are visual guides. For economical production of FDCA from HMF, it is desirable that oxidation is carried out at high HMF concentrations. However, as noted by various investigators ( 16 – 21 ), FDCA has low solubility in commonly used solvents. We have alleviated this limitation by using the GVL/H 2 O solvent system. We show in Fig. 1C that the solubility of FDCA is enhanced in a binary solution of GVL and water. FDCA has low solubility in both GVL and water; however, the solubility reaches a maximum value at 80 wt % GVL. Excess enthalpies of mixing are indicative of the solvation potential of the solvent system, and the solvation of the solute determines its solubility in the solvent system ( 25 ). We observe that the increase in solubility of FDCA with GVL concentration correlates well with the enthalpy of mixing of GVL and water ( Fig. 1C ) ( 26 ). Accordingly, we performed HMF oxidation over a 5% Pt/C catalyst with high HMF concentration in GVL/H 2 O (80:20) ( Table 1 , entry 2). This experiment resulted in lower FDCA yield, in part, due to the low water content, because HMF oxidation is shown to proceed through the formation of geminal diol, and the equilibrium shifts toward the formation of the corresponding carbonyl compound in the solution with low water concentration ( 27 ). In addition, water is shown to be the oxygen source in HMF oxidation ( 17 ). Thus, we performed HMF oxidation with a solvent system containing higher water content while maintaining high FDCA solubility, that is, GVL/H 2 O (50:50). Figure 1D shows a comparison between the solubility of FDCA in water and in GVL/H 2 O (50:50) as a function of temperature. It can be seen that the solubility of FDCA increases exponentially in the GVL/H 2 O (50:50) with temperature, and at the temperature of 383 K, the solubility of FDCA in the GVL/H 2 O (50:50) system is >9 wt %. Oxidation of 7.5 wt % HMF solution in GVL/H 2 O (50:50) over the 5% Pt/C catalyst resulted in 94% FDCA yield ( Table 1 , entry 3). As the liquid product stream cooled, FDCA crystallized out from the solution, leading to facile product separation and solvent recycling. Moreover, we show that the Pt/C catalyst is stable for the oxidation of HMF in the GVL/H 2 O (50:50) solvent system using a three-phase trickle bed reactor (see Fig. 1B and the Supplementary Materials). In addition, we studied the influence of transport resistances in the trickle bed reactor by changing the flow rates while keeping space velocities of the liquid feed and gaseous oxygen constant. We observe that the changing hydrodynamic properties of the flow have minimal effects on the yield of FDCA produced (table S9), indicating that there are negligible external transport resistances present in the reactor. We also probed the effects of oxygen pressure by performing the oxidation under two different oxygen pressures (40 and 20 bar). We observe that the oxygen pressure has no effect on the FDCA yield, indicating that HMF oxidation is zero order with respect to oxygen pressure. Our results are in agreement with the recent findings of Davis et al. ( 17 ). We have previously shown that high yields of HMF are obtained by dehydration of fructose using HCl as catalyst and with GVL/H 2 O solvent containing high GVL content (90 wt % GVL and 10 wt % water) ( 23 ). Here, we show that similar yields (for example, >70%) are obtained using the GVL/H 2 O (50:50) solvent with 15 wt % fructose ( Fig. 1 , E and F), yielding a liquid product stream containing 7.4 wt % HMF. It is desirable to use glucose instead of fructose as the feed material, but we obtained low HMF yield from glucose dehydration (~20%), and as such, oxidation of glucose-derived HMF was not performed. However, we found that direct oxidation of the fructose-derived HMF (F-D HMF) is complicated by the by-products of the dehydration reaction, especially humins, which are produced by the polymerization of HMF and/or fructose ( Table 1 , entry 4). Oxidation of HMF using the homogeneous Co/Mn/Br system has been reported to suffer from deactivation of the catalyst ( 28 ). Oxidation of F-D HMF solution obtained in GVL/H 2 O (50:50) resulted in low FDCA yield due to catalyst deactivation (see the Supplementary Materials). We determined that both Cl − and humins lead to catalyst deactivation (see the Supplementary Materials). Treatment of F-D HMF by ion exchange to remove Cl − ions and by contacting with activated carbon to remove humins produced a liquid stream of GVL/H 2 O containing HMF that could be oxidized over Pt/C to FDCA, resulting in 93% FDCA yield ( Table 1 , entry 5). In addition, FDCA crystallized out as the solvent was cooled, and solid FDCA was obtained with >99% purity (see the Supplementary Materials). Note that complete removal of FDCA from the GVL/H 2 O (50:50) solvent is not achieved during crystallization. Specifically, crystallization at 277 K leads to two phases, a substantially pure crystallized solid FDCA phase that accounts for 94% of the FDCA produced in the oxidation step and can be directly used in downstream processes without further purification and a liquid phase containing ~0.5 wt % FDCA in the GVL/H 2 O (50:50) solvent system. FDCA is stable under the conditions used for fructose dehydration, and thus, the liquid phase obtained after crystallization can be recycled to the HMF production reactor. To avoid the use of corrosive mineral acids, we replaced HCl with FDCA, creating a more sustainable system capable of achieving high yields of HMF ( Fig. 1 , E and F). Replacing the mineral acid, HCl, with an organic acid, which is the product of the process, improves the economics of the process, because the Cl − ion is eliminated and the ion exchange operation is not required. In addition, the capital cost of the process decreases because FDCA is not as corrosive as the mineral acid. Oxidation of HMF obtained by the dehydration of 15 wt % fructose in GVL/H 2 O (50:50) using 0.53 wt % FDCA as catalyst resulted in 91% FDCA yield ( Table 1 , entry 6). Figure 2A shows pictorially each step of the process. Fig. 2 Process and economics for the production of FDCA from fructose. ( A ) Pictorial representation of FDCA production from fructose. (i) 15 wt % fructose in GVL/H 2 O (50:50) containing 0.53 wt % FDCA. (ii) Solution after dehydration at 453 K containing 7.5% HMF and humins. (iii) Humin removal by adsorption over activated carbon (a red colored solution instead of a black solution is obtained). (iv) Solution obtained after oxidation over a Pt/C catalyst. ( B ) Sankey diagram for FDCA production process and ( C ) costs and revenues. LA, levulinic acid; AC, activated carbon; ROI, return on investment. On the basis of our experimental data, we developed a process model and performed TEA of the process. The analysis suggests that our approach can produce FDCA at a minimum selling price (MSP) of $1490/metric ton ( Fig. 2 , B and C). The reduced MSP for the approach described in the present paper is due to the enhanced solubility of FDCA in the GVL/H 2 O solvent system, the ease of product separation, the use of stable heterogeneous catalysts, and the use of FDCA as a dehydration catalyst. A detailed description of the process, the assumptions made, and the model used are available in the Supplementary Materials. Among the process sections, the HMF production section is the highest contributor due to the high feedstock cost ($650/metric ton), whereas the FDCA production section is the second highest contributor due to the long residence time (4.3 hours), requiring a large volume for the HMF oxidation reactor. If a 40% reduction in the residence time can be achieved, via increased catalyst activity, and the feedstock cost is reduced by 10%, then the MSP of FDCA can be reduced by 11.9% to $1310/metric ton. Thus, the proposed approach could become an economically competitive alternative to current terephthalic acid processes based on fossil fuels [that is, a benzoic counterpart of FDCA ($1445/metric ton)]."
} | 4,269 |
22427720 | PMC3289792 | pmc | 1,879 | {
"abstract": "In this study, a metabolic network describing the primary metabolism of Chlamydomonas reinhardtii was constructed. By performing chemostat experiments at different growth rates, energy parameters for maintenance and biomass formation were determined. The chemostats were run at low irradiances resulting in a high biomass yield on light of 1.25 g mol −1 . The ATP requirement for biomass formation from biopolymers ( K \n x ) was determined to be 109 mmol g −1 (18.9 mol mol −1 ) and the maintenance requirement ( m \n ATP ) was determined to be 2.85 mmol g −1 h −1 . With these energy requirements included in the metabolic network, the network accurately describes the primary metabolism of C. reinhardtii and can be used for modeling of C. reinhardtii growth and metabolism. Simulations confirmed that cultivating microalgae at low growth rates is unfavorable because of the high maintenance requirements which result in low biomass yields. At high light supply rates, biomass yields will decrease due to light saturation effects. Thus, to optimize biomass yield on light energy in photobioreactors, an optimum between low and high light supply rates should be found. These simulations show that metabolic flux analysis can be used as a tool to gain insight into the metabolism of algae and ultimately can be used for the maximization of algal biomass and product yield. Electronic supplementary material The online version of this article (doi:10.1007/s10811-011-9674-3) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions A metabolic network describing the primary metabolism of C. reinhardtii was constructed. By performing chemostat experiments, energy parameters for maintenance and biomass formation were obtained. The chemostats were run at low irradiances resulting in a high biomass yield on light of 1.25 g mol −1 . The ATP requirement for biomass formation from biopolymers ( K \n x ) was determined to be 109 mmol g −1 (18.9 mol mol −1 ), and the maintenance requirement ( m \n ATP ) was determined to be 2.85 mmol g −1 h −1 . These values are in the same range as literature values. With these energy requirements included in the metabolic network, the network accurately describes the primary metabolism of C. reinhardtii and can be used for modeling of C. reinhardtii growth and metabolism. Simulations with this metabolic model confirmed that mitochondrial respiration both provides energy for maintenance and additional energy to support growth. The high maintenance requirements at low growth rates result in low biomass yields; thus, cultivating algae at low growth rates is unfavorable. Cultivating algae at high light supply rates is less favorable as well, because the biomass yield will decrease due to light saturation effects. Thus, to optimize biomass yield on light energy in photobioreactors, an optimum between these two situations should be found. The simulations presented in this paper show that the metabolic model can be used as a tool to gain insight into the metabolism of C. reinhardtii and ultimately can be used for the maximization of biomass and product yield.",
"introduction": "Introduction Microalgae are interesting organisms because of their ability to produce a wide range of compounds, such as carotenoids (Ben-Amotz et al. 1982 ; Kleinegris et al. 2010 ; Lamers et al. 2008 ), lipids (Chisti 2007 ; Hu et al. 2008 ; Griffiths and Harrison 2009 ), hydrogen (Ghirardi et al. 2000 ; Melis et al. 2000 ), protein (Boyd 1968 ; Becker 2007 ), and starch (Delrue et al. 1992 ). These algal compounds have numerous applications, varying from fine chemicals to biofuels to food additives. To make commercial bulk production of these compounds economically feasible, maximization of algal biomass production and optimization of biomass composition are necessary. Understanding of how compounds are produced in the algal metabolism will help to fully exploit the potential of microalgae and their products. Metabolic flux analysis (MFA) is a powerful tool to study the fluxes through metabolic pathways of any organism of interest. It provides information on how nutrients and energy are utilized to form biomass and other components. Using information on metabolic fluxes and pathways, a better understanding of metabolism is obtained, and targets for process optimization or genetic modification can be identified. Several metabolic network models have been published for common production organisms, such as Escherichia coli (Carlson and Srienc 2004 ; Kayser et al. 2005 ), Saccharomyces cerevisiae (Forster et al. 2003 ), and Corynebacterium glutamicum (Kieldsen and Nielsen 2009 ). Due to the increasing interest in microalgae as production organisms, metabolic networks for photoautotrophic organisms such as Chlorella vulgaris (Yang et al. 2000 ) and Arthrospira ( Spirulina ) platensis (Cogne et al. 2003 ) have been developed as well. Metabolic flux analysis will improve understanding of algal metabolism and ultimately can be used for the maximization of algal biomass and product yield (Schmidt et al. 2010 ). \n Chlamydomonas reinhardtii has been studied extensively in the past decades. It is regarded as a model organism for green microalgae because of its diverse metabolism and its ability to grow photoautotrophically as well as heterotrophically on acetate (Gfeller and Gibbs 1984 ; Heifetz et al. 2000 ). In addition, C. reinhardtii is able to accumulate starch (Ball et al. 1990 ) and produce hydrogen when grown anaerobically (Melis et al. 2000 ). The Chlamydomonas genome has been sequenced (Merchant et al. 2007 ), and the availability of this genetic information provides a sound basis for the development of metabolic network models. Recently, Boyle and Morgan ( 2009 ) and Manichaikul et al. ( 2009 ) developed two extensive genome-scale models describing the primary metabolism of C. reinhardtii divided over several cellular compartments, and as such, both models have a high degree of compartmentalization. They qualitatively end up to a limited extent quantitatively predict algal metabolism and are well suited to get a better insight in algal metabolism. However, quantitative validation is very limited, and there are some difficulties in developing such detailed models. For a number of reactions, the localization is not known and has to be assumed. Also, information on the exchange of metabolites between compartments is limited. Finally, to reduce the number of fluxes that cannot be calculated, many, often complex, measurements are needed. When developing a metabolic model, the practical applicability should be considered. Extensive, fully compartmentalized models are excellent tools for the qualitative study of cellular reaction networks and their regulation but are generally not suited for quantitative studies because of their underdetermined characteristics. A model for finding specific engineering bottlenecks and solutions would call for simple, easy to handle networks, which require less input parameters and optimization commands but still represent the important characteristics of metabolism (Burgard et al. 2001 ). Therefore, we developed a more condensed metabolic model describing the primary metabolism of C. reinhardtii , in which reactions are less extensively compartmentalized and which is thus less underdetermined. Apart from the abovementioned uncertainties on compartmentalization and transport steps, the stoichiometry of the energy metabolism is not fixed in these models. The amount of energy in the form of adenosine triphosphate (ATP) required for biomass formation and maintenance is difficult to determine and varies between different microorganisms and growth conditions (Pirt 1965 ). These parameters are essential in metabolic modeling because they largely influence the growth dynamics and biomass or product yields calculated by the model. It is known from previous studies that theoretical estimates of the amount of ATP used for the production of biomass based solely on the required energy for the formation of biopolymers is much lower than the experimentally determined value (Baart et al. 2008 ; Kayser et al. 2005 ; Roels 1983 ). Additional energy is required for the assembly of biopolymers into growing biomass. In order to obtain a correct metabolic network model, these parameters have to be determined experimentally. In this paper, we describe the construction of a metabolic model for C. reinhardtii and the subsequent experimental determination of the energy requirements for maintenance and biomass formation. In this model, photosynthesis and the Calvin cycle are the only processes that are compartmentalized in order to separate these processes from the pentose phosphate pathway (PPP) in the cytosol and energy generation in the mitochondria. Furthermore, reactions in linear pathways are lumped. The energy parameters are estimated from a series of chemostats operated at different dilution rates, a commonly used method for heterotrophic microorganisms (Kayser et al. 2005 ; Taymaz-Nikerel et al. 2010 ). To the best of our knowledge, this has not been applied to photoautotrophically grown microalgae, most probably because light is a challenging energy source to measure accurately. This study shows how this method can be applied to determine energy parameters for photoautotrophic organisms. The final model including the determined energy parameters was used to calculate the respiration rate at different specific growth rates, which enabled prediction of optimal growth rates for efficient light use.",
"discussion": "Results and discussion Metabolic network construction A metabolic network describing the primary metabolism of C. reinhardtii was constructed based on literature (Berg et al. 2003 ; Boyle and Morgan 2009 ; Cogne et al. 2003 ; Harris 2009 ; Yang et al. 2000 ) and the KEGG database (Kanehisa and Goto 2000 ; http://www.kegg.com ). We cross-checked the model with the genome of C. reinhardtii (Merchant et al. 2007 ) to ensure the presence of the enzymes catalyzing the modeled reactions. In this model, we described two cell compartments, the chloroplast and the cytosol, to be able to uncouple the Calvin cycle, the PPP, and the production and consumption of NADPH. For the light reaction, only linear electron transport was modeled. A large network with over 300 enzymatic reactions was obtained. This extensive model was reduced to a smaller, more practical network. This was done by lumping linear pathways into one reaction equation. The resulting network contains 160 reactions and 164 compounds and is listed in Appendix A . A simplified overview of this metabolic network is shown in Fig. 2 .\n Fig. 2 Overview of an algal cell in the light, showing the main metabolic processes. Only the chloroplast and the cytosol were modeled as cell compartments; therefore, mitochondrial processes such as respiration through the electron transport chain were placed in the cytosol. Light is fixed in the chloroplast, yielding O 2 , ATP, and NADPH. These are needed for the fixation of carbon dioxide in the Calvin cycle into glyceraldehyde 3-phosphate ( GAP ). GAP can be transported to the cytosol to be converted into building blocks for biomass. Lipids are formed through glycolysis and the tricarboxylic acid (TCA) cycle. Nitrate is taken up by the cell and converted into glutamate which in turn can be converted to protein and chlorophyll. GAP can be converted to glucose 6-phosphate (G6P) from which carbohydrates are formed. G6P can also enter the pentose phosphate pathway (PPP) which yields NADPH, DNA and RNA. Electrons are carried by NADH and FADH 2 to the mitochondrial electron transport chain, yielding ATP by taking up O 2 \n \n We found that several enzymatic steps which were necessary in the network were not annotated in the KEGG database. Therefore, we performed protein basic local alignment search tool (BLAST) (Altschul et al. 1990 ) searches against the C. reinhardtii genome, using amino acid sequences from green (micro)organisms for the “missing” enzymes. Appendix C shows the enzymes, the E.C. numbers, and the corresponding geneIDs that were found in this way. The reactions in the network (partly) catalyzed by these enzymes are also given. In total, 41 enzymes were found not to be annotated in the KEGG database, of which 39 were retrieved by BLAST searches. Of these 39 enzymes, several have a geneID and are annotated but are not taken up in the KEGG database yet. Other enzymes have a draft geneID and still need to be annotated. Only two enzymes could not be found in this way. The first one is ATP phosphoribosyltransferase (E.C. 2.4.2.17). This enzyme is essential in histidine formation and is described in the Chlamydomonas Sourcebook (Harris 2009 ). The second enzyme is homoserine acetyltransferase (E.C. 2.3.1.31), which is necessary in cysteine formation. This enzyme is present in other green microalgae ( Ostreococcus lucimarinus ), cyanobacteria ( Synechococcus elongates , Anabaena variabilis ), and diatoms ( Phaeodactylum tricornutum , Thalassiosira pseudonana ). BLAST searches with the amino acid sequences from these organisms did not give a result. Therefore, we assume this reaction is performed by another but similar enzyme because this step is essential in the formation of cysteine. Because both histidine and cysteine were measured in the amino acid composition and not added to the medium, they had to be formed within the metabolism. Thus, both reactions were taken up in the model. By studying the null space of matrix A \n c (Eq. 2 ), 12 underdetermined parts were revealed in the model. This means that there is no unique solution for Eq. 2 . By measuring and setting constraints to some of the fluxes in the underdetermined parts of the metabolism and using optimization of objective functions, unique solutions could be obtained. Reactions that are irreversible were constrained to one direction, and thermodynamically impossible combinations of reactions were also constrained in the correct direction. In the appendix, the arrows indicate whether a reaction is reversible or irreversible and if so, in which direction it is set. By choosing the objective functions “maximize biomass yield” and “maximize ATP yield” unique values could be calculated for the fluxes in all but two underdetermined parts. The first remaining underdeterminancy involved the anaplerotic routes between phosphoenolpyruvate, pyruvate, and oxaloacetate. This was solved by setting the upper and lower boundaries of the flux through reaction 35 to 0. The second underdetermined part involved the coupling of the PPP, the glycolysis, and the tricarboxylic acid cycle (TCA). The aerobic degradation of sugars can occur through both the PPP as well as the TCA if NADPH and NADH are freely exchangeable through reaction 54, a transhydrogenase reaction. We restricted this by setting this reaction forward so NADPH can only be converted to NADH. In this way, the fluxes through the PPP will only be dictated by the demand of NADPH and PPP intermediates. Chemostat experiments To estimate the energy parameters for maintenance and the formation of biomass, we performed seven chemostat experiments at low irradiances and low dilution rates ranging from 0.018 to 0.064 h −1 . The experiments were performed at irradiances to make sure the photosystems were working at maximum efficiency (Baker et al. 2007 ) and to prevent light damage to the algal cells. When steady state was reached, the biomass density C \n x (g L −1 ) and the photon flux density absorbed by the algae (PFD abs ) were measured for each dilution rate. In Table 2 , the biomass density and the PFD abs at the different growth rates are given as well as the residence time for each steady state. With the biomass density and the photon flux density values, a light supply rate per amount of biomass ( r \n Ex , mmol g −1 h −1 ) could be calculated for each growth rate μ (h −1 ) according to Eq. 5 . The relationship between the specific light supply rate r \n Ex (mmol photons g −1 h −1 ) and the growth rate μ (h −1 ) can be described using the model of Pirt ( 1965 ) as used by Zijffers et al. ( 2010 ) according to Eq. 7 :\n Table 2 Biomass density, absorbed photon flux density, and growth rates determined for each chemostat experiment Growth rate a \n μ (h −1 ) Residence time a (h) Biomass density C \n x \n b (g L −1 ) Photon flux density absorbed PFD abs (μmol photons m −2 s −1 ) Specific light utilization rate r \n Ex, μ (mmol photons g −1 h −1 ) 0.018 ± 0.000 55.1 ± 0.5 0.78 ± 0.04 88 15.8 ± 0.8 0.019 ± 0.000 52.8 ± 1.2 0.84 ± 0.08 87 16.7 ± 1.5 0.031 ± 0.001 32.3 ± 1.0 0.41 ± 0.02 80 25.7 ± 1.4 0.034 ± 0.001 29.7 ± 1.1 0.39 ± 0.02 73 26.8 ± 1.6 0.052 ± 0.001 19.3 ± 0.5 0.21 ± 0.01 51 36.1 ± 1.7 0.061 ± 0.003 16.5 ± 0.7 0.11 ± 0.01 36 45.5 ± 3.7 0.064 ± 0.001 15.6 ± 0.0 0.10 ± 0.01 31 44.8 ± 3.7 The residence time is calculated as 1/ μ for each steady state. The specific light utilization rate is calculated from these data using Eq. 5 and 6 . Mean ± standard deviation \n a \n N = 13, 4, 6, 7, 13, 6, 8, respectively \n b \n N = 8, 3, 6, 5, 5, 4, 4, respectively \n 7 \\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}$$ {r_{\\text{Ex}}} = \\frac{\\mu }{{{Y_{{xE}}}}} + {m_E} $$\\end{document} in which m \n E is the maintenance requirement (mmol photons g −1 h −1 ) and Y \n xE the yield of biomass on light (g biomass mmol photons −1 ). The amount of light used by the algae increases proportionally to the growth rate while a fixed amount of maintenance light energy is necessary to keep the algae in a healthy state. Regression on the specific growth rates μ and specific light supply rates r \n Ex for all chemostat experiments yields a straight line with a R \n 2 of 0.989. According to Eq. 7 , the offset of this line gives a m \n E of 5.98 ± 1.63 mmol g h −1 and the inverse of the slope gives an Y \n xE of 1.25 ± 0.06 g mol −1 . This Y \n xE is high compared to biomass yields found for other green microalgae as can be seen from Table 3 . Cuaresma et al. ( 2009 ), Kliphuis et al. ( 2010b ), and Zijffers et al. ( 2010 ) found yields ranging from 0.5 to 1.0 g mol −1 for several green microalgae. These yields were all obtained at high irradiances of 1,000 μmol photons m −2 s −1 or more. At low irradiances, we expect a higher yield than at high irradiances because at high irradiances, the antenna complexes in the algal photosystems become saturated. The remainder of the absorbed light will be dissipated as heat or fluorescence (Krause and Weis 1991 ; van der Tol et al. 2009 ). In addition, elevated irradiances also induce damage to the algal cells (Kok 1955 ), possibly increasing the energy consumption for maintenance purposes. Therefore, high irradiances will lead to low photosynthetic efficiencies as can also be seen from the yields for C. reinhardtii obtained by Takache et al. ( 2010 ). The difference between the yields at high and low irradiances in these experiments reflects the fact that a large part of the light is “wasted” at higher irradiances. The high biomass yield found in our experiments supports the hypothesis that the efficiency of the algal photosystems was indeed high at a light intensity of 80 μmol photons m −2 s −1 . Moreover, it seems that we were working at maximal efficiency because the relation between specific light utilization and specific growth rate is linear ( R \n 2 = 0.988). This shows that the biomass yield is constant and not influenced by the change in light regime (Table 2 ), indicating we reached the maximal value of this yield parameter. Furthermore, it also shows that photorespiration, which results in waste of ATP, is negligible.\n Table 3 Comparison of biomass yields on light energy and the used irradiances for different microalgae from literature Organism \n Y \n xE or Y \n xE \n obs (g mol −1 ) Light intensity (μmol photons m −2 s −1 ) Reference \n Chlamydomonas reinhardtii \n 1.25 ± 0.06 a,d \n 80 This paper \n Dunaliella tertiolecta \n 0.78 a \n 930 Zijffers et al. ( 2010 ) \n Chlorella sorokiniana \n 0.75 a \n 930 Zijffers et al. ( 2010 ) \n Chlorella sorokiniana \n 0.80 b \n 1500 Kliphuis et al. ( 2010b ) \n Chlorella sorokiniana \n 1.0 b \n 2100 Cuaresma et al. ( 2009 ) \n Chlamydomonas reinhardtii \n 1.11 b,c \n 110 Takache et al. ( 2010 ) \n Chlamydomonas reinhardtii \n 0.73 b,c \n 500 Takache et al. ( 2010 ) \n Chlamydomonas reinhardtii \n 0.51 b,c \n 1000 Takache et al. ( 2010 ) \n a \n Y \n xE = μ / (r Ex + m E ), according to Eq. 7 \n \n b \n Y \n xE \n obs = μ / r \n Ex (the observed yield was not corrected for maintenance requirements) \n c Recalculated from data obtained in flat Torus photobioreactor by Takache et al. ( 2010 ) \n d Calculated by linear regression ( P < 0.05) \n It needs to be stressed that even at these low irradiances, not all absorbed light can be converted in the photosystems. Therefore, the light supply rate r \n Ex was corrected for the maximum efficiency (Φ Pmax = 0.8) of the photosystems to obtain the specific light utilization rate r \n Ex, μ (Eq. 6 ), which is also shown in Table 2 . This specific light utilization rate r \n Ex, μ should be used for energy parameter estimation, since this rate represents the actual amount of photons that enter the algal metabolism. As expected, regression on the specific growth rates μ and specific light utilization rates r \n Ex, μ also yields a straight line according to Eq. 7 . The offset of this line gives m \n E of 4.79 ± 1.31 mmol g −1 h −1 and the inverse of the slope gives Y \n xE of 1.57 ± 0.07 g mol −1 . These values represent the yield and maintenance requirements corrected for the inefficiency of light use, Φ Pmax . Biomass composition The composition of the biomass can have a significant effect on the flux distribution in the model and thus on the estimation of energy parameters (Pramanik and Keasling 1998 ). Therefore, the macromolecular biomass composition (% w / w ) was determined for six of the steady states as shown in Table 4 . In addition, the average biomass composition for all growth rates is given here. These compositions, along with the corresponding growth rates μ and light utilization rates r \n Ex, μ , were used for the model simulation of each steady state. The macromolecular biomass composition was not measured for μ = 0.019 h −1 ; therefore, the composition of μ = 0.018 h −1 was used to perform model simulations for this growth rate.\n Table 4 Measured biomass composition (% w / w ) normalized to 100% and average biomass composition of C. reinhardtii at different specific growth rates (mean ± standard deviation) μ (h −1 ) Average c \n 0.018 0.031 0.034 0.052 0.061 0.064 Protein a \n 42.48 ± 0.55 39.34 ± 0.77 40.17 ± 1.10 41.63 ± 0.67 40.16 ± 0.40 37.61 ± 0.88 40.23 ± 1.71 Carbohydrate a \n 24.09 ± 2.00 21.70 ± 0.73 24.62 ± 2.15 27.29 ± 4.08 26.26 ± 1.59 28.87 ± 1.11 25.47 ± 2.54 Lipids a \n 14.13 ± 0.28 22.16 ± 2.24 17.72 ± 0.88 14.25 ± 0.47 16.78 ± 3.85 18.65 ± 1.84 17.28 ± 3.01 DNA b \n 0.21 ± 0.008 0.20 ± 0.004 0.23 ± 0.006 0.22 ± 0.002 0.23 ± 0.002 0.19 ± 0.002 0.21 ± 0.02 RNA b \n 5.99 ± 0.23 5.57 ± 0.12 6.48 ± 0.16 6.04 ± 0.06 6.57 ± 0.06 5.27 ± 0.05 5.99 ± 0.50 Chlorophyll a \n 7.06 ± 0.07 5.59 ± 0.03 5.13 ± 0.05 4.73 ± 0.07 4.34 ± 0.20 4.01 ± 0.06 5.14 ± 1.09 Ash a \n 6.05 ± 0.03 5.44 ± 0.40 5.66 ± 0.24 5.84 ± 0.11 5.66 ± 0.24 5.41 ± 0.42 5.68 ± 0.24 The sum of the individual biomass components including ash was comparable to the measured dry weights within 10% \n a \n N = 3 for each steady state \n b Based on literature values (Merchant et al. 2007 ; Valle et al. 1981 ) and the measured dry weight and cell numbers. N = 3, 6, 5, 5, 4, 8, respectively \n c Average biomass composition for all growth rates with the standard deviation for all six growth rates \n The biomass composition did not show much variation as a function of growth rate, although the carbohydrate content seemed to increase at increasing growth rates. The pigment content, on the other hand, decreased at increasing growth rate, which can be explained by the fact that the amount of light per cell increased at increasing growth rate, because the culture became more diluted. Such a response of microalgae to decrease the amount of photosynthetic pigments upon an increase in irradiance is known as photoacclimation (Dubinsky and Stambler 2009 ). Protein, nucleic acids, and ash content did not vary significantly. Since the composition of the biomass did not change very much in this range of growth rates, the average biomass composition was used in the final model. In Appendix D , the elemental composition of all macromolecules and of the C. reinhardtii biomass itself is given. The amino acid composition and the average fatty acid composition can be found in Appendixes E and F . Energy requirements for growth and maintenance Using the specific growth rate μ , the specific light utilization rate r \n Ex, μ (Table 2 ) and the biomass composition as determined for each steady state (Table 4 ) as input for the model, q \n ATP can be calculated as described in the Theoretical aspects section. Figure 3 shows the plot of q \n ATP against the specific growth rate, μ . Linear regression through these points yields a straight line ( P < 0.05). According to Eq. 4 , the offset of this line gives the ATP required for maintenance ( m \n ATP ), being 2.85 ± 0.82 mmol g −1 h −1 . The slope represents K \n x , the amount of ATP needed to make biomass from biopolymers and has a value of 109 ± 19 mmol g −1 . From this number, it could be calculated that 18.9 mol ATP is required to transport and assemble 1 mol biomass. This amount was added to the biomass synthesis reaction (mol ATP per mol biomass, reaction 147) of the final model. The ATP hydrolysis flux (reaction 57) was set to the value for m \n ATP to fix the maintenance requirements of the final metabolic model.\n Fig. 3 Plot of the specific calculated overall ATP production rate q \n ATP against the experimentally determined growth rate. q \n ATP was calculated with the model for each growth rate. Regression through these points yields a straight line of which the offset gives the ATP required for maintenance m \n ATP , 2.85 mmol g −1 h −1 . The slope gives the constant which represents the additional amount of ATP needed to make biomass from the biopolymers ( K \n x ), 109 mmol g −1 , according to Eq. 4 . The error bars represent the minimum and maximum values for q \n ATP and the growth rate μ , which follow from the relative errors of biomass measurements \n An overview of values for K \n x and m \n ATP for several microorganisms is presented in Table 5 . The value for K \n x depends on the characteristics of the model, for example the degree of compartmentalization and the ATP stoichiometry for biopolymer formation and thus the value for K \n x differs per model. To make a good comparison of the total ATP use in several species, 1/ Y \n x ATP (mol ATP g −1 ) is also given, which is the total amount of ATP required to make 1 g of biomass. As can be seen from Table 5 , there is considerable variation in the values for 1/ Y \n x ATP , K \n x , and m \n ATP for different microorganisms. This can be due to the type of microorganism, the culture conditions, and the assumptions made in the models.\n Table 5 Comparison of yields and maintenance coefficients of different microorganisms from literature Organism 1/ Y \n x ATP (mol g −1 ) \n K \n x (mol g −1 ) \n m \n ATP (mmol g −1 h −1 ) Reference \n Chlamydomonas reinhardtii \n a \n 0.43 0.11 ± 0.018 c \n 2.85 ± 0.82 c \n This paper \n Petunia hybrida (cell culture) b \n 0.44 0.15 1.41 de Gucht and van der Plas ( 1995 ) \n Saccharomyces cerevisiae \n b \n 0.35 0.07 1.00 Forster et al. ( 2003 ) \n Escherichia coli \n b \n 0.31 0.09 2.81 Kayser et al. ( 2005 ) \n Escherichia coli \n b \n 0.40 0.10 4.70 Carlson and Srienc ( 2004 ) \n Streptomyces coelicolor \n b \n – 0.08 3.69 Borodina et al. ( 2005 ) \n Neisseria meningitides \n b \n 0.40 0.05 1.61 Baart et al. ( 2008 ) \n Corynebacterium glutamicum \n b \n 0.29 0.03 – Kieldsen and Nielsen ( 2009 ) \n a Photoautotrophic growth on light and CO 2 \n \n b Heterotrophic growth on glucose. Glucose yields 32 ATP (Berg et al. 2003 ) \n c Values are given ± their standard error as calculated by linear regression ( P < 0.05) \n Firstly, these parameters depend on the assumed P/O ratio, the relationship between ATP synthesis, and oxygen consumption. We assumed that NADH yields 2.5 ATP and FADH 2 yields 1.5 ATP upon respiration. However, lower values for these ratios would result in lower values for K \n x and m \n ATP . Secondly, we assumed a ratio between NADPH and ATP production during linear electron transport in chloroplast of 3:2 (ATP/NADPH) which exactly fits the requirements of the Calvin cycle. Studies on Spinach chloroplasts, however, show that linear electron transport only can deliver 2.57 ATP per every two NADPH and that cyclic photosynthetic electron transport (i.e., cyclic photophosphorylation) is needed to generate additional ATP (Allen 2003 ). This cyclic pathway was not included in our model because the additional ATP requirement is small, and it would make our model underdetermined. Its necessity, however, could partly explain the fact that the experimentally determined minimal quantum requirement of oxygen evolution is 10 instead of 8 as discussed before (equivalent to Φ Pmax = 0.8) and as such, it is implicitly present in our model. Besides the balance of ATP and NADPH in the chloroplast, the microalgal cells as a whole require substantial ATP as is extensively discussed in this study. According to our model, ATP must be generated by the complete conversion of sugars (GAP) produced in the chloroplast by the combined action of glycolysis, TCA cycle, and oxidative phosphorylation in the cytosol and mitochondria (Fig. 2 ). It is interesting to note that this pathway ultimately yields 1 mol of ATP per 1.5 mol photons (see “ Simulation of oxygen uptake through respiration ” for calculation). Cyclic photophosphorylation, on the other hand, would only yield 1 mol of ATP per two photons (Allen 2003 ). This shows that energy generation through linear photosynthetic electron transport is energetically more favorable than through cyclic electron transport and supports our description of C. reinhardtii metabolism. Thirdly, the parameter K \n x accounts for the requirement of ATP for biomass formation which is not accounted for in the part of the network with a known energy stoichiometry. For more complex models involving more compartmentalization, a larger part of the ATP may be accounted for, and consequently, the value of this parameter would become lower. However, for the two more extensive models described for Chlamydomonas , the energy parameters were not properly estimated. Manichaikul et al. ( 2009 ) used parameters taken from yeast cultivation and not from C. reinhardtii itself. Boyle and Morgan ( 2009 ) estimated the energy parameters for autotrophic, heterotrophic, and mixotrophic growth by fitting their model to one experimentally determined biomass yield, based on experiments performed in shake flasks. The maintenance parameter was taken from literature and not measured. Boyle and Morgan estimated a K \n x of 29.89 mmol g −1 , which is indeed lower than our value, as would be expected for a more detailed model. However, they used shake flask experiments to estimate this parameter. Shake flasks usually have undefined light regimes, which makes it difficult to properly measure the absorbed light by the culture (PFD abs ) and thus the light supply rate ( r \n Ex ). Also, the efficiency of photosynthesis is unknown under these cultivation conditions. Furthermore, only a single growth rate was used, making the estimation highly dependent on the assumed value of the maintenance parameter. Simulation of oxygen uptake through respiration With the average biomass composition (Table 4 ) for all chemostats and the energy requirements for both maintenance and growth associated processes, a working model was obtained as shown in the appendix. With this model, we simulated the mitochondrial respiration rate (mmol O 2 g −1 h −1 ) at different growth rates ( μ , h −1 ). Appendix C shows the flux distribution through the network at a specific growth rate μ of 0 and 0.062 h −1 . The size of the fluxes in mmol g −1 h −1 are shown in boxes, red boxes for μ = 0 h −1 and black boxes for μ = 0.062 h −1 . Light is taken up in the light reaction of photosynthesis, and with the energy that is formed, CO 2 is fixed in the Calvin cycle. The carbon is moved to the cytosol in the form of GAP, where it enters the glycolysis and TCA cycle. Part of the GAP is used to synthesize biopolymers and part is used here to generate energy in the form of ATP, NADH, and FADH 2 . NADH and FADH 2 are in turn respired in the mitochondria to generate additional ATP to fulfill the energy requirements for growth and maintenance. The flux distribution at μ = 0 h −1 shows the maintenance metabolism without biomass formation. It also shows that the minimal light uptake rate is 4.28 mmol photons g −1 h −1 , which is used to provide energy for maintenance (2.85 mmol photons g −1 h −1 ). Using these fluxes, it can be calculated that in this model 1.5 mol photons should be consumed to produce 1 mol ATP. Figure 4 shows the simulated biomass yield on light ( Y \n xE in g mol −1 , closed dots) for several growth rates μ . These simulations give an ideal situation since no light saturation is modeled and consequently the biomass yield Y \n xE increases asymptotically to the maximal value of 1.57 g mol −1 at very high specific growth rates. In reality, light saturation will occur due to the high irradiances necessary to reach maximal growth rates. The photons are absorbed, but the energy is only partly used for growth, causing a decrease in the biomass yield on light energy (Φ P < Φ Pmax , Eq. 6 ). If light saturation (Baker et al. 2007 ; Krause and Weis 1991 ; van der Tol et al. 2009 ) would be taken into account, the yield in Fig. 4 would reach an optimum and decrease as soon as light saturation occurs. The simulated respiration rate (open diamonds), calculated by the sum of the oxygen consumption rates in oxidative phosphorylation (reactions 52 and 53), and the fraction of oxygen used for maintenance (open triangles) are also plotted in Fig. 4 . Respiration is a linear function of the growth rate. This is expected, since most ATP required for growth is generated by oxidation of GAP in the mitochondria. Regression through these points yields a straight line with an intercept of 0.53 mmol O 2 g −1 h −1 at μ = 0 h −1 . This value is the oxygen uptake rate through respiration which is necessary for maintenance purposes. In earlier experimental work (Kliphuis et al. 2010a ), we measured that the respiration rate required for maintenance for Chlorella sorokiniana was 0.3 mmol O 2 g −1 h −1 . This is in the same order as the value we now calculated for C. reinhardtii but almost twofold lower. As mentioned before, this is likely to be species-specific. It can be seen that at low growth rates, a relatively large part of respiration is needed for maintenance purposes. At high growth rates, the largest part of respiration is needed for energy generation for growth purposes. Consequently, cultivating microalgae at low specific growth rates results in a low biomass yield Y \n xE , since a large part of the energy is used for maintenance processes rather than growth. This effect was experimentally confirmed for C. sorokiniana and Dunaliella tertiolecta by Zijffers et al. ( 2010 ).\n Fig. 4 Simulated respiration rates (mmol O 2 g −1 h −1 ) at several growth rates ( μ , h −1 ). Regression through these points yields a straight line with an intercept at μ = 0 of 0.53 mmol O 2 g −1 h −1 . This value is the oxygen uptake rate through respiration which is necessary for maintenance purposes. The maintenance fraction of the respiration rate is also plotted and shows which part of respiration is used for maintenance purposes. This graph also shows the simulated biomass yields on light energy ( Y \n xE , g mol −1 ) for these growth rates. The simulations give an ideal situation since light saturation is not modeled. If light saturation would be taken into account the yield would reach an optimum and decrease as soon as light saturation occurs"
} | 9,219 |
20333304 | PMC2841643 | pmc | 1,882 | {
"abstract": "The Yellowstone caldera contains the most numerous and diverse geothermal systems on Earth, yielding an extensive array of unique high-temperature environments that host a variety of deeply-rooted and understudied Archaea , Bacteria and Eukarya . The combination of extreme temperature and chemical conditions encountered in geothermal environments often results in considerably less microbial diversity than other terrestrial habitats and offers a tremendous opportunity for studying the structure and function of indigenous microbial communities and for establishing linkages between putative metabolisms and element cycling. Metagenome sequence (14–15,000 Sanger reads per site) was obtained for five high-temperature (>65°C) chemotrophic microbial communities sampled from geothermal springs (or pools) in Yellowstone National Park (YNP) that exhibit a wide range in geochemistry including pH, dissolved sulfide, dissolved oxygen and ferrous iron. Metagenome data revealed significant differences in the predominant phyla associated with each of these geochemical environments. Novel members of the Sulfolobales are dominant in low pH environments, while other Crenarchaeota including distantly-related Thermoproteales and Desulfurococcales populations dominate in suboxic sulfidic sediments. Several novel archaeal groups are well represented in an acidic (pH 3) Fe-oxyhydroxide mat, where a higher O 2 influx is accompanied with an increase in archaeal diversity. The presence or absence of genes and pathways important in S oxidation-reduction, H 2 -oxidation, and aerobic respiration (terminal oxidation) provide insight regarding the metabolic strategies of indigenous organisms present in geothermal systems. Multiple-pathway and protein-specific functional analysis of metagenome sequence data corroborated results from phylogenetic analyses and clearly demonstrate major differences in metabolic potential across sites. The distribution of functional genes involved in electron transport is consistent with the hypothesis that geochemical parameters (e.g., pH, sulfide, Fe, O 2 ) control microbial community structure and function in YNP geothermal springs.",
"introduction": "Introduction Metagenome sequencing of microbial community DNA holds tremendous promise for determining the properties of indigenous microbial populations and the composition and structure of microbial communities in natural environments [1] – [6] . Recent studies suggest that metagenome sequencing can be quite effective in characterizing low-complexity sites such as the extremely acidic (pH<1) mine-drainage biofilm at Iron Mountain [7] , [8] or deep subsurface (2.8 km) drainage waters from an African gold mine [9] . The extreme geochemical conditions at the Iron Mountain site (i.e., high H + , Fe II , As III ) limit the microbial community composition, and metagenomic sequencing has been used to successfully assemble near-complete, consensus genomes of indigenous Ferroplasma type II and Leptospirillum group II populations. The metagenomic data also provided necessary tools (e.g., expression arrays) for evaluating key genetic determinants important to the function of these organisms in this geochemical context [7] , [8] . Specifically, oxidation of Fe II , arsenic resistance, and defense against oxidative stress are important genetic attributes of the organisms inhabiting these environments [8] . In more complex microbial communities, significantly greater sequencing is required to obtain adequate depth of coverage for phylogenetic and functional analyses [3] , [5] . Surface soils and marine photic zones are among the most diverse environments on Earth [3] , [5] , [10] – [11] , and the gene diversity observed in the Global Ocean Survey metagenomes [5] precluded extensive assembly of individual sequence reads into larger contigs and scaffolds. However, the metagenomes clearly revealed dominant organisms important in marine systems, as well as immense diversity and identification of numerous new protein families [5] . The relative simplicity of high-temperature environments as indicated from prior 16S rRNA gene surveys [12] – [15] provides a unique opportunity for utilizing metagenome sequencing to elucidate phylogenetic and functional diversity in model environmental systems. The primary goal of this work was to evaluate the phylogeny and ecology of five disparate chemotrophic microbial communities in Yellowstone National Park (YNP) using environmental shotgun sequencing in the context of extensive geochemical characterization. Our specific objectives were to (i) identify predominant indigenous populations of five high-temperature geothermal microbial communities in YNP using multiple phylogenetic analysis approaches of metagenome sequence data, (ii) determine the metabolic potential of these indigenous microorganisms using bioinformatic and functional analysis of metagenome sequence, and (iii) identify candidate protein-coding genes that may have relevance to variable geochemical conditions across these geothermal systems. The phylogeny of specific functional genes provides direct insight towards the possible role of individual population(s) within each community, and provided candidate genes whose distribution may be a function of major geochemical attributes such as pH, dissolved oxygen, Fe and or S species.",
"discussion": "Results and Discussion Geochemical Context of Chemotrophic Geothermal Habitats The extensive geochemical diversity of terrestrial hot springs in YNP provides a natural laboratory for evaluating the role of specific geochemical variables such as pH, dissolved oxygen, ferrous iron and the presence of sulfide (or elemental S) on the distribution and functional adaptations of thermophilic microorganisms. The five chemotrophic environments chosen for this study encompass a representative range of habitat types characteristic of non-phototrophic high temperature environments in YNP. They exhibit major differences in pH (∼3–8), dissolved oxygen, Fe, total dissolved sulfide, as well as predominant solid phases intimately associated with the microbial community ( Figure 1 , Table 1 ). It is hypothesized that geochemical and hydrodynamic attributes of each site control the phylogenetic composition and corresponding functional capabilities of these microbial communities. 10.1371/journal.pone.0009773.g001 Figure 1 Habitat context and geothermal site characteristics. Site photographs and scanning electron micrographs (SEM) of microbial mats and solid phases associated with each geothermal sample used for metagenome sequencing (map of Yellowstone National Park and site locations shown in top left panel). A. Crater Hills (CH, gold); B. Norris Geyser Basin (NGB, red); C. Joseph's Coat Hot Springs (JCHS, blue); D. Mammoth Hot Springs (MHS, green). E. Calcite Springs (CS, violet). 10.1371/journal.pone.0009773.t001 Table 1 Aqueous geochemical parameters 1 and predominant solid phases associated with the five geothermal microbial communities sampled for metagenome sequencing. Location T pH I DIC DS O 2 \n As Fe CH 4 \n H 2 \n Solid Phases 2 \n Site 3 \n Coordinates °C –mM– ------------uM------------ ----nM---- \n Crater Hills (CH) \n 75 2.5 18 1.3 1–2 <3 2 230 300 67 \n S 0 , SiO 2 \n \n Alice Spring CHANN041 44° 39′ 12.108″ N. Lat 110° 28′ 39.6″ W. Lon \n Norris Geyser Basin (NGB) \n 65 3.0 17 0.82 <1 50 27 37 300 15 \n Fe(AsO 4 ) 0.6 (OH) 3 \n \n Beowulf Spring NHSP35 44° 43′ 53.4″ N. Lat 110° 42′ 40.9″ W. Lon \n Joseph's Coat Hot Springs (JCHS) \n 80 6.1 23 0.45 25 <3 130 0.7 900 107 \n S 0 , Sb 2 S 3 , FeS 2 , As 2 S 3 , SiO 2 \n \n Scorodite Spring JCS083 44° 44′ 21.4 N. Lat 110° 19′ 28.2″ W. Lon \n Mammoth Hot Springs (MHS) \n 71 6.6 32 16.5 80 <3 20 0.4 <10 17 \n CaCO 3 (aragonite), S 0 \n \n Narrow Gauge MA041 44° 58′ 9.915″ N. Lat 110° 42′ 35.4″ W. Lon \n Calcite Springs (CS) \n 75 7.8 16 0.8 70 <3 18 3.4 <10 30 \n FeS 2 , S 0 \n \n Scary Spring \n 44° 54′ 17.46″ N. Lat 110° 24′ 14.5″ W. Lon 1 I = ionic strength calculated from aqueous geochemical modeling at sample temperature; DIC = dissolved inorganic C; DS = dissolved sulfide; DO = dissolved oxygen. 2 predominant solid phases determined using scanning electron microscopy (FE-SEM) coupled with energy dispersive analysis of x-rays (EDAX) and x-ray diffraction (XRD). 3 Site = specific spring name and Yellowstone National Park Thermal Inventory Number (when available, www.rcn.montana.edu ). A brief comparison of the geochemical attributes across these five sites is necessary for evaluating potential functional differences among the numerically dominant phyla identified in each microbial community. The low pH (2.6) turbid pool at Crater Hills (CH) contains suspended particulates (∼1–2 g/L) comprised primarily of elemental S and SiO 2 ( Figure 1a ). Dissolved O 2 values are below detection (∼3 µM), and the low concentrations of other dissolved gases including H 2 S, H 2 and CH 4 are characteristic of a steam-dominated, acid-sulfate system [16] ( Table 1 ). In contrast, the acidic (pH 3.1) mat community sampled from a geothermal spring in Norris Geyser Basin (NGB) is from an oxygenated outflow channel (∼65–70°C) dominated by Fe III -oxides [17] . The electron and x-ray amorphous Fe-oxides in NGB form as encrustations and nodules around filamentous organisms ( Figure 1b ) at channel locations where dissolved O 2 values range from 30–100 µM (20–60% of saturation, Table 1 ). The Fe-oxide mats at NGB and the S-rich sediments of CH have both been shown to contain significant numbers of crenarchaea within the order Sulfolobales [13] , [18] – [19] . Consequently, a comparison between these two sites provides an excellent opportunity to study geochemical factors responsible for functional diversity of different members of these acidophilic crenarchaea. Although the three higher pH sites at Joseph's Coat (JCHS), Calcite (CS), and Mammoth Hot Springs (MHS) are all sulfidic and sub-oxic, they are geochemically distinct from one another and yield significantly different microbial communities ( Figure 1 ). The anoxic, submerged sediments sampled at JCHS (80°C) are dominated by reduced phases of sulfur including pyrite (FeS 2 ), stibnite (Sb 2 S 3 ), orpiment (As 2 S 3 ) and elemental S ( Figure 1c ). The aqueous phase of JCHS contains high concentrations of CH 4 , H 2 , NH 4 , arsenite, and thiosulfate ( Table 1 ). Consequently, numerous reduced chemical species could serve as electron donors to support chemolithotrophic metabolism in the absence of O 2 \n [13] . The Calcite and Mammoth Springs ‘streamer’ communities were sampled from high-velocity (i.e., ∼0.1–0.3 m s −1 ), highly-sulfidic outflow channels (>80 µM total dissolved sulfide, Table 1 ), and have been shown to be dominated by microorganisms of the deeply-rooted bacterial Order Aquificales [20] – [22] . Although total soluble Fe is low in the source waters of JCHS and CS, a combination of higher pH (6.1 and 7.8, respectively) and high sulfide results in the deposition of pyrite (FeS 2 ) in both environments. At CS, the deposition of pyrite on filamentous cell walls yields characteristic black ‘streamers’ ( Figure 1e ) that are variably intermixed with thin, but visible coatings of elemental S. In contrast, the predominant mineralization processes at MHS [20] – [21] result in the deposition of CaCO 3 (aragonite) and elemental S to form white, pale yellow ‘streamers’ ( Figure 1d ). A comparison of the geochemistry and functional gene content across these three sulfidic sites provides a unique opportunity for identifying the role of S in the energetics and metabolism of different thermophilic microorganisms [23] – [24] . Analysis of Metagenome Sequence Approximately 14–15,000 sequence reads of an average length of ∼800 bp (∼11–12 Mbp) were obtained for each site ( Table S1 ). The number of individual sequence reads that assembled into contigs and scaffolds (Celera Assembler [5] ) varied considerably across these five sites. Though the total amount of sequence obtained per site was relatively small, there was considerable assembly producing scaffolds of significant length (>174 kb). The length of these scaffolds indicates that these communities are dominated by a small number of relatively homogenous microbial species, which facilitated the phylogenetic and functional annotation of these datasets. Coverage estimates for contigs ranged from ∼6.6 at MHS, 5 at CH, and just over 2 at JCHS, CS and NGB (additional assembly statistics in Table S1 ). Archaeal-Dominated Communities (Crater Hills, Norris Geyser Basin, Joseph's Coat) Analysis of individual sequence fragments (e.g. ∼800 bp) from Crater Hills (CH) using binning and fragment recruitment approaches both reveal that the majority of sequence reads (∼60%) are phylogenetically related to crenarchaea within the order Sulfolobales, and that a smaller number of sequence reads are more closely related to members of the Desulfurococcales (∼12%) and Thermoproteales (∼5%). The partial genome sequence of Acidilobus sulfurireducens str. 18D70, an anaerobic, S-respiring Desulfurococcales isolated from YNP [25] recruits approximately 10% of the sequence reads from CH ( Figure 2a ). However, none of the current genome sequences (e.g., Sulfolobus solfataricus \n [26] , Aeropyrum pernix \n [27] , Hyperthermus butylicus \n [28] and Staphylothermus marinus \n [29] ) are good references for organisms present at this site, as evinced by nucleotide identities generally less than 70%, and inconsistent coverage relative to reference genomes ( Figure 2a ). 16S rRNA gene sequences observed in the metagenome data are consistent with prior 16S rRNA gene analysis at CH using PCR and clone library analysis ( Figure S1 ). One near full-length 16S rRNA gene observed in the assembled sequence data ( Table S2 ) was present in a large scaffold of 165,334 bp, and this novel taxa (91% nucleotide identity to S. solfataricus ) represents an important Sulfolobales population in CH. 10.1371/journal.pone.0009773.g002 Figure 2 Phylogenetic analysis of metagenome sequence data. Binning of metagenome sequence reads (l eft column) from Crater Hills (gold), Norris Geyser Basin (red), Joseph's Coat Springs (blue), Mammoth Hot Springs (green) and Calcite Springs (violet) (with blastn similarity scores (E-values) of <10 −10 ) to closest reference microbial genomes (abbreviations below). Environmental sequence reads were further categorized based on nucleotide identity ranging from 47–100% (shaded from light to dark, legend shown only for MHS). Fragment recruitment (r ight column) of metagenome sequence reads to reference microbial genomes is plotted across each reference genome (x-axis) at a nucleotide identity ranging from 50–100% (y-axis). Reference genomes: MK1 = Metallosphaera sp. str. MK1 (partial genome sequence) [19] ; AS \n = Acidilobus sulfurireducens (partial genome sequence) [25] ; SS \n = Sulfolobus solfataricus \n [26] ; CM \n = Caldivirga maquilingensis \n [34] ; AP \n = Aeropyrum pernix \n [27] ; PA \n = Pyrobaculum arsenaticum \n [38] ; TV \n = Thermoplasma volcanium \n [31] ; NM \n = Nitrosopumilus maritimus \n [32] – [33] ; Hyd \n = Hydrogenobaculum sp. Y04AAS1 [35] ; GK \n = Geobacillus kaustophilus \n [36] ; TP \n = Thermofilum pendens ; CS \n = Caldicellulosiruptor saccharolyticus \n [37] ; SY03 \n = Sulfurihydrogenibium sp. Y03AOP1 [35] , [39] ; SY \n = Sulfurihydrogenibium yellowstonensis \n [35] , [39] ; TA \n = Thermus aquaticus Y5.1 MC23; TL \n = Thermotoga lettingae \n [40] . The acidic Fe III -oxide microbial mat from Norris Geyser Basin (NGB) is the most diverse archaeal community of the five sites included in this study, and contains several novel lineages within the current Phylum Crenarchaeota, Candidate Phylum Thaumarchaeota [30] and Phylum Euryarchaeota ( Figure 2b ). A significant fraction (∼11%) of the metagenome sequence reads exhibit reasonable nucleotide identity (80–99%) to the available genomic sequence (∼2200 ORFs) of Metallosphaera sp. str. MK1 [19] , representing a coverage of ∼0.3x ( Figure 2b ). Additional Sulfolobales-like sequence reads (∼9%) do not recruit well to current reference genomes, and exhibit low nucleotide identity to S. solfataricus P2 [26] ( Figure 2b ). Approximately 8% of the total sequence reads are related to a novel euryarchaeal lineage, and ∼5% are related to organisms within the candidate Phylum Thaumarchaeota [30] . These sequences are contributed by indigenous organisms distantly related to currently cultivated relatives, and exhibit low nucleotide identity (47–70%) to reference genomes within the Thermoplasmatales (i.e., Thermoplasma volcanium , [31] ) and Nitrosopumilales (i.e., Nitrosopumilus maritimus, Nitrosocaldus yellowstoni \n [32] – [33] , respectively). Novel members of the Thermoproteales (∼5%) and Desulfurococcales (∼5%) are also present in the Fe-oxide mats. Currently, the best references for these sequence reads include Caldivirga maquilingensis \n [34] and the partial genome sequence data of Acidilobus sulfurireducens \n [25] , respectively ( Figure 2b ). A smaller subset of environmental sequence reads (∼3%) from NGB show excellent identity to the genome of Hydrogenobaculum sp. Y04AAS1 [35] ( Figure 2b ), and although the estimated coverage of this genome is only 0.2x, the number of high-identity (∼90%) sequence matches, as well as the thorough distribution of fragments across the reference genome, suggests that highly similar Hydrogenobaculum -like organisms are important members of the Fe-mat community. Other minor (<0.5% of sequence reads) bacterial populations present include distant relatives of Geobacillus kaustophilus \n [36] and Caldicellulosiruptor saccharolyticus \n [37] . Metagenome sequence from Joseph's Coat Hot Springs (JCHS) is largely contributed by indigenous members of the Thermoproteales and Desulfurococcales ( Figure 2c ). The largest fraction of sequence reads (∼40%) ‘bin’ nearly equivalently to the Pyrobaculum spp. genomes [38] (i.e., P. aerophilum , P. arsenaticum DSM 13514, P. caldifontis JCM 11548). Recruitment of sequence reads to these reference organisms shows similar nucleotide identities (60–80%) across the genomes of Pyrobaculum spp. and Thermoproteus neutrophilus , ( Figure 2c ) representing significant and fairly uniform coverage (∼1x). A second Thermoproteales population in JCHS (∼10% of sequence reads) is more closely related to Caldivirga maquilingensis \n [34] ( Figure 2c ), while approximately 20% of the metagenome sequence reads show homology to the partial genome sequence (∼2200 ORFs) from A. sulfurireducens str. 18D70 [25] . As found in Crater Hills, these sequences do not show significant nucleotide identity to currently available Desulfurococcales genomes such as A. pernix (not shown). Phylogenetic assignment of 16S rRNA genes observed within the metagenome sequence data from JCHS ( Table S2 ) is consistent with prior identification of 16S rRNA genes using PCR and cloning ( Figure S1 ). Bacterial-Dominated Communities (Mammoth Hot Springs and Calcite Springs) A large majority (∼90%) of metagenome sequence reads from the streamer community at Mammoth Hot Springs (MHS) were highly similar (nucleotide identity >90%) to the genome of Sulfurihydrogenibium sp. Y03AOP1 [35] , isolated from Obsidian Pool, YNP ( Figure 2d ). The fact that this microbial community is dominated by a single bacterial population(s) within the Aquificales suggests that this population is highly adapted to the geochemical and hydrological attributes of this environment. Significant and relatively uniform coverage (∼3.9x) of either of the two reference Sulfurihydrogenibium genomes (strain Y03AOP1 or S. yellowstonensis , [35] ) was obtained from this site with only 12 Mbp of random sequence. Metagenomic sequence from the higher pH (7.8), sulfidic site at Calcite Springs (CS) is also largely bacterial, exhibiting high nucleotide identity to genome sequences of Thermus aquaticus and Sulfurihydrogenibium yellowstonensis SS-5, which was isolated from this site [39] . These two bacterial populations represent a major fraction of the metagenome sequence from this community (∼45% Thermus and ∼36% Sulfurihydrogenibium ), resulting in ∼1.5x and 1.7x coverage relative to the reference genomes of T. aquaticus and S. yellowstonense (or strain Y03AOP1), respectively ( Figure 2e ). A small fraction (<1–2%) of sequence reads from CS are similar to Pyrobaculum -like organisms (the dominant archaea in this system) as well as bacterial population(s) distantly related to genome sequence of Thermotoga lettingae \n [40] ( Figure 2e ). Predominant Sequence Assemblies Principal components analysis (PCA) of nucleotide word frequencies (1–5) derived from the assembled libraries from each site was used to cluster and classify the metagenome assemblies ( Figure 3 ). When contigs >1500 bp from all sites are included together in this analysis, the individual sites separated into distinct and highly-constrained clusters ( Figure 3a ). The distinct sequence signatures across sites reflect differences in the dominant phyla present within these five different geothermal habitats. The PCA plot was then fixed in this orientation, and sequences classified phylogenetically ( Figure 3b ). Distinct, but partially overlapping Sulfolobales sequence assemblies (gold) are evident in the acidic, elemental sulfur site (CH) and the acidic Fe-oxide mats (NGB). The anaerobic, sulfidic sediments of Joseph's Coat Springs (JCHS) are dominated by two major sequence assemblies that are most similar to reference genomes within the crenarchaeal orders Thermoproteales and Desulfurococcales (blue, Figure 3b ). Bacterial sequences within the order Aquificales dominate the carbonate ‘streamer’ communities (70–72°C) at MHS, and are also one of the dominant phyla present in the pyritic ‘streamers’ from CS. These scatter plots are presented in three dimensions thereby providing better separation of different clusters, which can only be fully appreciated by interacting with the data directly (nucleotide word frequency plots can be accessed at http://gos.jcvi.org/openAccess/scatterPlotViewer.html ). 10.1371/journal.pone.0009773.g003 Figure 3 Nucleotide word frequency plots and phylogenetic analysis of metagenome assemblies. Nucleotide word frequency principal component analysis (PCA) of assembled metagenome sequence data (contigs>1500 bp) from five chemotrophic geothermal habitats in YNP: A. Metagenome sequence colored by site (Crater Hills = gold, Norris Geyser Basin = red, Joseph's Coat = blue, Mammoth Hot Springs = green, Calcite Springs = violet). B. Identical PCA orientation of metagenome sequence observed in Panel A , but colors now designate phylogenetic affiliation at the order level (Sulfolobales = gold; Desulfurococcales = light blue; Thermoproteales = dark blue; Aquificales = green; Thermales = violet; Unassigned = black), and C. Identical PCA orientation with phylogenetic classification at the domain-level ( Archaea = gold, Bacteria = violet). Viral Sequences Metagenome data from all five sites was analyzed for sequences exhibiting significant similarity to known viruses. A small subset of sequence reads (∼1%) from CH, JCHS, and NGB recruit to crenarchaeal virus genomes at nucleotide identities ranging from 50–90%. The majority of virus-like sequences at JCHS (∼70%) were related to Pyrobaculum Spherical Virus (PSV), and of 150 viral-like sequence reads, 112 were assembled into larger contigs and scaffolds resulting in an average coverage of 2.5x across the reference PSV genome ( Figure S2 ). The metagenome sequence data suggest that PSV was present in the sediment used to extract DNA, or present in host cells. Recent work also suggests possible assembly of Pyrobaculum -like viruses from thermal springs in YNP [41] . The majority of virus-like sequence reads from CH and NGB were most closely related to the reference genomes of Sulfolobales viruses including Acidianus two-tailed virus [42] , Stygiolobus rod-shaped virus [43] and other miscellaneous Sulfolobus viruses [18] , [44] . However, the viral-related sequence reads from CH and NGB do not assemble into larger contigs or scaffolds and do not result in uniformly random coverage across the viral genomes (data not shown). Functional Analysis of Metagenome Sequence One of the premises of this study is that the metabolic attributes of microbial populations in high-temperature, geothermal environments are influenced by geochemical parameters [45] , [46] , a subset of which can also be influenced by physical processes such as velocity, turbulence and gas exchange. Following, the genes required for specific physiologies of the populations inhabiting these springs should reflect the geochemical constraints defining these microenvironments (e.g., the presence of specific electron donors and or acceptors). To examine this tenet in an unbiased manner, we first performed a thorough and integrated statistical analysis of metagenome sequence data to identify the biochemical pathways and functions that are utilized differentially in the five microbial communities. Metabolic Pathway Reconstruction We explored differences and similarities in metabolic pathways among the five communities using a custom metabolic pathway reconstruction database created based on metagenome sequence from each site as well as reference genomes and databases [47] – [49] (see Methods ). Pathway completeness scores for all MetaCyc [47] pathways found in at least one community (or reference genome) were first subjected to principal component analysis (PCA) to identify pathways that contribute most to variability across the genomes/metagenomes ( Figure S3 ). The completeness data for the key pathways identified by PCA analysis was then clustered using average linkage hierarchical clustering ( Figure 4a ). As expected, the archaeal and bacterial sites are readily separated at the functional level, and the reference genomes group with the relevant metagenomes ( Figure 4a ). Results from functional analysis converge with phylogenetic analysis ( Figure 2 – 3 ) and reveal the importance of distinct phyla in each site (e.g. Sulfolobales in CH; Aquificales in MHS). Although contributions from specific phyla are clearly recognizable, the more diverse communities (especially the acidic Fe-oxide mats) are functionally relatively distant from the major reference species. The major pathways that contribute to differentiating the sites follow the division between bacteria and archaea-specific pathways ( Figure 4a ). For example, bacteria synthesize terpenoid compounds through the MEP pathway whereas archaea use the mevalonate pathway. The pathway responsible for autotrophic metabolism recently characterized in Metallosphaera sedula \n [50] was clearly represented in the sites dominated by archaea (CH, JCHS, NGB). While the pathway completeness-based analysis can highlight major functional differences between metagenomes and reference genomes, it does not account for the abundance of specific genes, and thus does poorly in differentiating relatively similar communities (e.g. the three archaeal communities). 10.1371/journal.pone.0009773.g004 Figure 4 Functional gene analysis. Two-way clustering of biochemical pathways that contributed most to the variability between sites based on PCA analysis. A. Comparison of metagenomes and relevant reference genomes based on pathway completeness. Reference genomes: Sulfolobus solfataricus P2, Metallosphaera sedula DSM 5348, Pyrobaculum arsenaticum , Thermus aquaticus Y5.1 MC23, Sulfurihydrogenibium sp. YO3AOP1. B. Comparison of metagenomes based on the median number of blast hits to enzymes in a pathway on a log scale. Abbreviations: MEP, methylerythritol phosphate; GGPP, geranylgeranyl-diphosphate; THF, tetrahydrofolate; 3-HP/4-HB, 3-hydroxypropionate/4-hydroxybutyrate; BCKA, branched-chain keto-acid. To assess the similarity of the five communities in terms of abundance of specific genes (i.e., pathway activity), we performed the same analysis outlined above using the median number of blast hits to proteins in each pathway for each site. Results from this analysis, which accounts for differences in organism abundance within sites provide examples of the role of specific electron donors or acceptors (e.g., sulfur, tetrathionate, arsenate) across the three archaeal environments ( Figure 4b ). Functional analysis provides clues to the types of metabolic potential represented in metagenomes and genomes, but differentiating putative function within protein families based on analysis of partial protein sequences is challenging. For this reason, we also analyzed the genes related to the use of electron donors and acceptors (including those identified in Figure 4a ) and CO 2 fixation in greater detail using assembled metagenome sequence. Carbon Fixation and Energy Cycling: Linkage to Geochemistry To assess differences in the potential metabolism of organisms found within sites exhibiting disparate geochemistry, we searched the five metagenomic assemblies using query sequences ( Table S3 ) of proteins associated with specific autotrophic (CO 2 fixation) pathways and electron transfer processes involving C, Fe, S, As, N, H 2 or O 2 ( Table 2 ). In contrast to the analysis using generic EC-number/reaction associated sequences from public databases described above, the query sequences used for detailed functional analysis were selected from reference organisms that are phylogenetically related to members of these environments ( Table S3 ). 10.1371/journal.pone.0009773.t002 Table 2 Identification of metagenome sequences associated with CO 2 fixation, electron transfer reactions and detoxification across five high-temperature chemotrophic systems in Yellowstone National Park (YNP). Process Substrate Marker Gene 1 \n Number of Probable Sequence Matches 2 in Site \n Crater Hills \n \n Norris Geyser Basin \n \n Joseph's Coat HS \n \n Mammoth Hot Springs \n \n Calcite Springs \n CO 2 fixation (reductive TCA) CO 2 \n \n aclB \n 0 0 0 \n 1 \n \n 1 \n CO 2 fixation (reductive acetyl-coA) CO 2 \n \n acs, cooS \n 0, 0 0, 0 0, 0 0, 0 \n 2, 1 \n CO 2 fixation (3-HP/4-HB) 3 \n CO 2 \n \n 4hcd, mcm \n \n 2, 3 \n \n 3, 8 \n \n 4, 2 \n 0, 0 0, 0 Thiosulfate oxidation S 2 O 3 \n 2− \n \n tqoAB \n \n 2 \n \n 2 \n \n 1 \n 0 0 Oxidation of reduced S S 2− \n \n sqr \n \n 4 \n \n 10 \n \n 6 \n \n 2 \n \n 3 \n Sulfite oxidation SO 3 \n 2− \n \n sox \n \n 1 \n \n 3 \n \n 3 \n \n 1 \n \n 1 \n Sulfite oxidation SO 3 \n 2− \n \n soxC \n 0 0 0 0 \n 4 \n Hydrogen oxidation (Group 1 Ni-Fe Hyd) H 2 \n \n hynS, hynL-like \n \n 3, 1 \n 0, 0 \n 1, 3 \n 0, 0 0, 1 Arsenite oxidation As III \n \n aroA \n 0 \n 1 \n \n 1 \n 0 0 Terminal oxidation 4 \n O 2 \n \n doxB \n \n 3 \n \n 5 \n \n 1 \n 0 0 Terminal oxidation O 2 \n \n aoxB \n 0 \n 10 \n 0 0 0 Terminal oxidation O 2 \n \n foxA \n 0 \n 4 \n 0 0 0 Terminal oxidation O 2 \n \n other \n 0 \n 1 \n 0 \n 2 \n \n 2 \n Dissimilatory sulfur reduction 5 \n S 0 , S n \n x− \n \n sreA-like \n 3 \n \n 2 \n 0 \n 10 \n \n 2 \n \n 7 \n Dissimilatory sulfate reduction SO 3 \n 2− \n \n dsrA \n 0 0 \n 3 \n 0 \n 1 \n Dissimilatory N oxide reduction NO/N 2 O \n norB, nosZ \n 0, 0 0, 0 \n 2, 0 \n 0, 0 0, 0 Dissimilatory nitrate reduction NO 3 \n − \n \n narG \n 0 \n 2 \n \n 4 \n 0 \n 2 \n Arsenic detoxification As III , As V \n \n arsB, arsC \n \n 3, 0 \n \n 8, 0 \n \n 5, 0 \n \n 1, 1 \n \n 4, 4 \n Mercury detoxification Hg \n merA \n \n 2 \n \n 3 \n \n 1 \n 0 0 1 marker genes code for proteins with high specificity for possible pathway. 2 number of different ‘high-confidence’ sequence matches to marker genes (see Supplementary Table S4 for details on individual sequence matches. 3 3-HP/4-HB = 3-hydroxypropionate/4-hydroxybutyrate pathway; terminal oxidation = reduction of O 2 via heme Cu oxidase. 4 includes Mo-pterin proteins similar to sre A and arr A. 5 no gene sequences with homology to sox B, sox M, nir K, nir S, nap A, mcrA , or amo A genes were noted (gene symbols also described in Table S3 ). An inventory of genes known to be involved in the five major chemotrophic CO 2 fixation pathways reveals major differences across sites ( Table 2 ), consistent with the dominant phyla found in each habitat. Genes coding for the key enzyme required for CO 2 fixation via the reverse tricarboxylic acid (rTCA) cycle (ATP citrate lyase, acl B [51] ) were observed only in MHS and CS; both metagenome sequences show excellent identity to the acl B gene annotated in Sulfurihydrogenibium spp. genomes (NCBI). The high-pH (7.8), pyritic mat at Calcite Springs (CS) is the only community to contain evidence of the reductive acetyl-CoA pathway for CO 2 fixation (marked by genes acs of the acetyl-CoA decarboxylase/synthase complex and coos , the carbon monoxide dehydrogenase catalytic subunit [52] ). These sequences are phylogenetically related to members of sub-dominant populations of Delta-proteobacteria and Firmicutes. Genes specific to the recently reported 3-hydroxyproprionate/4-hydroxybutyrate CO 2 fixation pathway (4-hydroxybutyryl-CoA-dehydratase, methylmalonyl-CoA-mutase) in Metallosphaera sedula \n [50] were found in sites dominated by archaea (CH, JCHS and NGB). The majority of environmental sequence hits to genes in this pathway were related to Sulfolobales reference genomes, consistent with the dominant phyla observed in CH and NGB, as well as a minor Sulfolobales population in JCHS ( Figure 2 ). Possible Chemotrophic Metabolism: Evaluation of Electron Donors and Acceptors Genes responsible for a sulfide-quinone reductase (SQR, glutathione reductase family of flavoproteins [53] – [54] ) were identified in all sites ( Table 2 ), and this is consistent with the presence of dissolved sulfide and other forms of reduced S (e.g. elemental S) in these geothermal environments ( Table 1 , Figure 1 ). The environmental sqr sequences exhibit closest matches to expected phyla for each site including members of the Sulfolobales (sites CH, NGB, JCHS), the Thermoproteales (site JCHS), the Aquificales (sites CS, MHS) and the Thermales (site CS). An additional sulfur oxidation pathway ( sox gene cluster [55] – [56] ; not to be confused with sox -type terminal oxidases, to be discussed below) was observed in CS (one Thermus -like and one Sulfurihydrogenibium -like sequence) as inferred by the presence of the sox C gene. Genes coding for the oxidation of thiosulfate (via the membrane bound tqo AB subunits [55] – [57] were noted in sites containing Sulfolobales (CH, NGB, and to a lesser extent JCHS), but not in sites dominated by Aquificales (MHS, CS). Genes coding for Group 1 membrane-bound Ni-Fe hydrogenases [58] were observed primarily in anoxic sulfidic sites containing Sulfolobales and Thermoproteales (CH and JCHS) ( Table 2 ), with several sequences showing significant amino acid identity to the Acidianus ambivalens Ni-Fe hydrogenase thought to be linked with a membrane-bound, sulfur-reductase (SreA) [59] – [60] . Arsenite oxidase genes ( aro A) were observed in two of the five sites (NGB and JCHS), and arsenite oxidation has been measured in both of these systems [61] – [62] . However, it is yet unclear whether arsenite-oxidizing organisms (including the Aquificales) derive energy from catalyzing this exergonic reaction [63] . No evidence of genes responsible for the synthesis of key enzymes in ammonium oxidation ( amoA ) or methanotrophy/methanogenesis ( mcr A) [64] – [65] were found in the assembled metagenome sequence (not shown in Table 2 ), suggesting that these may not be dominant microbial processes in the habitats studied here. Heme Cu oxidases (subunit 1 of terminal oxidase complexes) catalyze the reduction of O 2 to H 2 O and are an excellent indicator of the potential for aerobic metabolism (or in certain cases, O 2 detoxification) [66] – [67] . Genes coding for a subunit 1 were identified in all sites, but the diversity and number of gene copies was especially extensive in the Fe-oxide mat of NGB ( Figure 5 , Table 2 ). Over 20 different gene sequences matching different types of terminal oxidases were identified in the Fe-oxide mat including aox B, sox M, dox B and foxA - like cytochrome c oxidases [27] , [68] – [71] . The abundance and diversity of terminal oxidase genes is consistent with the observed O 2 influx in this outflow channel habitat [17] , [61] , [72] . Four different copies of the fox A gene observed in the Fe-mats are all related to Metallosphaera spp. sequences, and is consistent with observations suggesting that this terminal oxidase is utilized when Fe 2+ serves as an electron donor [71] , [73] – [74] . The complete absence of this gene in the other four sites is also consistent with the fact that the oxidation of Fe(II)is not a dominant process in the sulfidic habitats. 10.1371/journal.pone.0009773.g005 Figure 5 Diversity of heme copper oxidases present in metagenome sequence data. Phylogenetic tree (deduced protein sequences) of heme Cu oxidases and their relationship to nitric oxide (NO) reductases (NorB). Metagenome sequences observed across the five sites are included (Site_Meta ). All other entries are from annotated genomes found on NCBI. [notations for heme Cu oxidases: AoxB = A. pernix \n [27] ; SoxB, SoxM = S. acidocaldarius \n [68] , [70] ; DoxB = A. ambivalens \n [69] ; FoxA = M. sedula \n [71] , [73] and NorB = nitric oxide reductases. Tree = distance tree created with MEGA using the neighbor-joining method with 100 bootstraps]. The only heme Cu oxidase genes found in CH and JCHS were dox B-like sequences with ∼70% amino acid identity to other Sulfolobales and Thermoproteales reference genomes, respectively ( Figure 5 , Table S4 ). Recent evidence suggests that the terminal oxidase complex containing doxB is utilized when reduced S species (e.g., elemental S) serve as the electron donor [57] , which is consistent with the fact that both CH and JCHS contain dissolved sulfide, elemental S, and are sub-oxic ( Figure 1 , Table 1 ). The sites dominated by bacteria (MHS and CS) each contain two sequences matching the sub-unit I of cytochrome c oxidases annotated in the Sulfurihydrogenibium and Thermus spp. reference genomes, respectively ( Figure 5 , Table S4 ). Other terminal electron acceptors besides O 2 that may be important in these microbial habitats include nitrate, ferric iron, arsenate, thiosulfate, elemental S, sulfate or CO 2 . Gene sequences similar to putative molybdenum (Mo)-pterin subunit I arsenate reductases ( arr A) [75] – [76] were abundant in JCHS ( Figure 4b , Table 2 ), and this correlates with the high As concentrations at this site. However, genes coding for dissimilatory nitrite reductases ( nir K, nir S) [77] , ferric iron reductases ( fer ) [78] – [79] and methyl coenzyme M reductase ( mcr A) [65] were not observed in the metagenome data. Interestingly, genes coding for known dissimilatory nitrate reductases ( nar G) and nitric oxide reductases ( nor B) were found in JCHS ( Table 2 , Figure 5 ) and show excellent identity (E-values<10 −63 ) to those annotated in the Pyrobaculum spp. genomes [79] . Based on current models of dissimilatory nitrate reduction in bacteria [77] , [80] – [81] , a nitrite reductase ( nir K or nir S) would be required to produce NO, which serves as a substrate for nitric oxide reductase ( nor B) to produce N 2 O. Nevertheless, based on currently known gene function, the indigenous Pyrobaculum -like populations exhibit partial metabolic potential for denitrification. The nitrate reductase genes ( nar G) found in NGB were affiliated with sub-dominant phyla within the bacterial order Bacillales, while those in CS were affiliated with Thermus -like organisms. Genes coding for putative sulfur reductases (SreA-like) were observed in all habitats that contain reduced forms of sulfur (CH, JCHS, MHS, CS; Table 2 , Figure 4b ). The exact function of these putative Mo-pterin proteins remains to be elucidated, but work with related proteins in Acidianus ambivalens , Aquifex aeolicus and Pyrococcus furiosus (representing several thermophilic groups) suggests that a membrane-bound SreA (Mo-S binding site) acts to transfer electrons to elemental S [60] , [82] . In several cases, H 2 can serve as the electron donor as has been noted in A. ambivalens and P. furiosus \n [60] , [83] – [84] . Genes known to be involved in dissimilatory sulfate reduction including dsr A (codes for the sulfite reductase subunit [85] ) were only observed in JCHS and CS. The phylogenetic identity of dsr A genes (3) observed at JCHS suggests that the indigenous relatives of Pyrobaculum spp. and or Caldivirga spp. exhibit metabolic potential for sulfate reduction, in addition to the possible reduction of more reduced forms of S (i.e., sre A-like discussed above). The dsr A sequence from CS is contributed from a relative of the less-dominant deltaproteobacterial population(s) present at this site [e.g. Desulfovibrio and or Desulfococcus -related sequence matches; Table S4 ). Trace Element Detoxification All sites contained evidence of ars B genes ( Table 2 ), which code for efflux proteins used to transport arsenite out of the cell under toxic conditions [86] – [87] . The high arsenic concentrations associated with Yellowstone's geothermal ecosystem may necessitate that these microorganisms be capable of efficient arsenite efflux (aqueous As levels ranged from 10 to 130 µM across the sites discussed here; Table 1 ). The ars C gene, which is often found together with ars B on the ars operon [87] and codes for an arsenate reductase associated with detoxification, was only found in the bacterial dominated sites (MHS, CS), and these sequences were affiliated with Thermus and Sulfurihydrogenibium- like organisms. Genes that code for the mercuric reductase ( mer A) used in Hg detoxification [88] – [90] were found in sites dominated by archaea (CH, NGB, JCHS); all mer A sequences (5 total) were most closely related to genome sequences from the Sulfolobales ( Table S4 ). Mercury and arsenic are two of the most important toxic constituents originating from Yellowstone's geothermal features [91] – [93] , and it is noteworthy that the deeply-rooted phyla present in these environments exhibit potential for the detoxification of these elements. Summary Phylogenetic and functional analysis of random shotgun sequence data from five different geothermal environments ranging in pH from 2.5 to 7.8 suggest that these microbial communities are composed of numerically predominant microbial populations whose functional attributes are consistent with geochemical conditions. The two acidic sites (CH, NGB) and the near-neutral sulfidic site (JCHS) were dominated by sequences belonging to members of the Archaea . In contrast, the two microbial ‘streamer’ communities were dominated by sequences belonging to the Bacteria , including organisms within the deeply-rooted bacterial lineages of Aquificales (MHS, CS) and Thermales (CS). High-temperature springs with pH less than ∼6 were dominated by archaea (although Hydrogenobaculum -like organisms are important in NGB), whereas sites with pH values above ∼6 were dominated by bacteria. In addition, the distribution of different archaeal sequence reads from pH 2.5 to 7.8 confirmed the importance of Sulfolobales relatives at low pH (2.5 and 3.0 at CHAS and NGB), compared to Thermoproteales relatives at near-neutral pH and above (6.1 at JCHS). However, all three archaeal-dominated sites contained a significant number of contigs corresponding to novel Desulfurococcales populations. Moreover, a modest number of Sulfolobales sequences were observed in JCHS at a pH of 6.1. Consequently, the metagenome data show that members within the Class Thermoprotei commonly co-occur in the archaeal communities of YNP, and that the relative abundance of specific members of this Class change across sites with major differences in pH and or the presence of dissolved oxygen ( Table 1 ). The sequencing depth of indigenous populations (estimated based on coverage of reference genomes) varied significantly across the five geothermal sites, tracking inversely with microbial diversity. For example, environmental sequence data from the lowest diversity site (MHS) provided ∼4x coverage relative to the Sulfurihydrogenibium sp. Y03AOP1 or S. yellowstonensis genomes. At the other extreme, the 65°C, Fe III -oxide mat (NGB) exhibited considerable diversity including several novel archaeal populations within the Crenarchaeota (e.g. Sulfolobales, Desulfurococcales, Cenarchaeales, other uncharacterized Groups within the Crenarchaeota) and the Euryarchaeota, as well as a dominant bacterial population(s) of Hydrogenobaculum -like organisms, acidophilic members of the Aquificales [22] . Consequently, the sequencing depth of the dominant organisms in NGB (<1x coverage for Metallosphaera sp. str. MK1 and Hydrogenobaculum sp. Y04AAS1) is considerably lower than the other sites. Analysis of this sequence diversity is also compounded by the fact that the archaea present in NGB (as well as CH and JCHS) are only distantly related to organisms whose genomes have been sequenced to date, and in some cases represent order-level (or higher) lineages that do not yet have a cultured representative. With the rapid decline in sequencing costs, and the adoption of new pyrosequencing technologies, the amount of sequence coverage reported here is modest. However, our results indicate that modest metagenome sequencing in high-temperature geothermal environments provides an excellent tool for assessing and characterizing the predominant members of these microbial communities, as well as the possible functional attributes of these indigenous populations. This study was initiated as the first phase of a more extensive project (DOE-Joint Genome Institute Community Sequencing Project) aimed at characterizing the prokaryotic gene diversity found within phototrophic and chemotrophic geothermal sites of YNP. Consequently, further sequencing results will focus on building consensus genomic content of predominant indigenous and novel microorganisms in geothermal chemotrophic environments, and expand the inventory and analysis of functional attributes important for the survival and growth of these extremophiles."
} | 11,639 |
27999158 | PMC5181773 | pmc | 1,884 | {
"abstract": "ABSTRACT Diversity is often associated with the functional stability of ecological communities from microbes to macroorganisms. Understanding how diversity responds to environmental perturbations and the consequences of this relationship for ecosystem function are thus central challenges in microbial ecology. Unimodal diversity-disturbance relationships, in which maximum diversity occurs at intermediate levels of disturbance, have been predicted for ecosystems where life history tradeoffs separate organisms along a disturbance gradient. However, empirical support for such peaked relationships in macrosystems is mixed, and few studies have explored these relationships in microbial systems. Here we use complex microbial microcosm communities to systematically determine diversity-disturbance relationships over a range of disturbance regimes. We observed a reproducible switch between community states, which gave rise to transient diversity maxima when community states were forced to mix. Communities showed reduced compositional stability when diversity was highest. To further explore these dynamics, we formulated a simple model that reveals specific regimes under which diversity maxima are stable. Together, our results show how both unimodal and non-unimodal diversity-disturbance relationships can be observed as a system switches between two distinct microbial community states; this process likely occurs across a wide range of spatially and temporally heterogeneous microbial ecosystems.",
"introduction": "INTRODUCTION Microbial communities are the foundation of all ecosystems on Earth ( 1 ). Microbes live in fluctuating environments, and this heterogeneity influences their ecological structure and diversity ( 2 ). Similar to what has been found in large-scale ecosystems ( 3 ), diversity in microbial systems is often linked to ecological function and stability. For example, recent studies have revealed that higher community evenness is associated with improved functional stability in microcosms containing denitrifying bacteria ( 4 , 5 ). Similarly, the diversity of natural phytoplankton communities has been associated with increased resource use efficiency and ecological stability ( 6 , 7 ). In the human gut, microbial diversity appears to be connected with community stability and host health ( 8 , 9 ). Thus, in order to predict how disturbance might impact ecosystem function, it is important to determine whether there are general rules governing the relationship between microbial community diversity and environmental change. Disturbances introduce spatiotemporal heterogeneity to an environment and push ecosystems outside their normal range of variability, generally resulting in differential mortality and/or growth of community members. To better understand diversity-disturbance relationships (DDRs) in microbial systems, we can turn to the decades of literature surrounding this topic from traditional ecology ( 10 – 14 ). Despite important differences between microbes and macrobes, such as higher passive dispersal rates ( 15 ) and the potential for mixing together of entire microbial ecosystems ( 16 ), ecological systems often behave quite similarly across vastly different scales ( 17 – 20 ). In systems ranging from forests to coral reefs to grasslands, maximal diversity has been observed at intermediate frequencies or intensities of disturbance ( 21 – 26 ). The intermediate disturbance hypothesis (IDH) postulates that these peaks in diversity at intermediate levels of disturbance stem from the coexistence of organisms with different life history traits, such as those defined by competition-colonization tradeoffs, along a successional gradient ( 21 , 22 , 27 , 28 ). However, alternative non-unimodal DDRs, in which diversity peaks at low or high levels of disturbance, are also frequently detected in nature ( 29 – 31 ). Here we sought to systematically characterize the relationship between disturbance rate (i.e., the product of disturbance intensity and frequency) and diversity in a microbial ecosystem ( 32 ). Prior work using a single-strain system ( Pseudomonas fluorescens ) showed that coexistence of colony morphotypes peaked at intermediate levels of disturbance and productivity ( 33 – 35 ), but it is unclear whether these results translate to more complex communities. Recent laboratory work has shown that multispecies microbial community responses to disturbance depend not only on intensity but also on the frequency of disturbance application ( 34 , 36 ), suggesting that a simple universal relationship between diversity and disturbance may not exist. We expanded on this prior work and developed an experimental system comprised of a complex bacterial community enriched from Lake Ontario, wherein we could precisely define the dimensions of disturbance (e.g., disturbance type, range, intensity, and frequency) and quantitatively sample community diversity. We imposed disturbance regimes ranging from no applied disturbance to a near-total collapse of ecosystem biomass, and frequency and intensity were independently modulated to assess potential interactions between these factors ( 34 , 37 ). Further, we employed two qualitatively different disturbances: biomass removal and dilution, which is commonly applied in microcosm studies and is predicted to maintain relative taxon abundances ( 33 – 35 ), and UV radiation, which should induce taxon-specific mortality ( 38 ). We predicted that biomass removal/dilution would act as an indiscriminate mortality event with no marked effect on niche-defining environmental properties and therefore would not alter community diversity; we predicted that UV, on the other hand, would give rise to a unimodal DDR due to the coexistence of resistance phenotypes and tradeoffs between resistance and other traits. We show that our results are consistent with a minimal model of resource-coupled species, and we map the landscape of possible DDRs for the model to set our observations in broader context. Together our experimental and modeling results suggest conditions under which a variety of unimodal and non-unimodal DDRs can be observed, not only in microbial systems, but potentially in macroscale ecosystems as well.",
"discussion": "DISCUSSION We developed a tractable yet realistic experimental model system for assessing diversity-disturbance relationships in microbial systems. Microcosm communities underwent a reproducible succession between distinct ecological states, each characterized by a dominant species (i.e., het1, het2, and het3). We observed different apparent relationships between diversity and disturbance rate over the course of the experiment, which we propose to be due to the combined effects of biologically induced environmental change (i.e., limiting resource drawdown) and our imposed disturbance regimes. Alternatively, it is possible that community succession could be decoupled from environmental conditions. To test this hypothesis, one would need an experimental system in which environmental conditions could be manipulated separately from the community that normally creates them—for example, with a flowthrough system. Initially, we did not expect to find changes in community diversity in the biomass removal/dilution treatments, because indiscriminate mortality events alone are not expected to alter relative taxonomic abundances ( 45 ). Accordingly, the similarity in community structures between undisturbed controls and the highest-rate biomass removal treatments on day 16 ( Fig. 2A and B ) suggests that density-independent mortality alone has no detectable effect on community diversity. However, we did observe changes in diversity in other biomass removal treatments, which we suggest were mediated by changes in resource availability. Complementing our results, prior work in P. fluorescens microcosms found that biomass removal disturbances modified oxygen concentrations, which in turn altered community composition ( 46 ). In our work, we propose that the indirect effect of biomass removal on the abiotic environment shifted over time due to a biological feedback on resource availability, leading to a transient unimodal pattern. Similar dynamics also occurred in our simplified model, but we showed that we could obtain persistent coexistence if the environmental disturbance remained within a small range of rates. If our explanation is correct, such temporal dynamics may partly explain the variety of DDR patterns observed in nature and may be a common feature of both microbial and macrobial systems in fluctuating environments ( 30 ). Recent work has implied that disturbance generally reduces variability in microbial community composition ( 47 , 48 ), somewhat contradicting our findings ( Fig. 2A and C ). We found that community structure was more variable among replicates in the high-diversity intermediate-disturbance treatments, which may be symptomatic of a transition toward more stochastic community assembly ( 49 ). The lack of community convergence to an “intermediate state” of maximum diversity suggests that DDR peaks are transient, where there are no clear winners or losers. We speculate that peak community variance centered at the apex of a unimodal DDR curve might be a general phenomenon and that this would be an interesting subject for future investigation. Indeed, these results mirror prior work in a grassland ecosystem, where species compositional stability was lowest in high-diversity plots, while functional stability was greatest in those same plots ( 3 ). Our very limited model illustrated how DDR structures can shift over different time scales in the presence of consumer-resource feedbacks ( Fig. 4 ). Diversity responded uniformly to the product of frequency and intensity in the model, which matched what we saw in our more complex enrichment communities ( Fig. 5 ). This multiplicative interaction between intensity and frequency was also observed for a special case of a recent vegetation model (i.e., when age to maturity and dispersal capability were similar across competitors) and is consistent with results from experiments involving coexistence of colony morphotypes across a range of disturbance intensities and frequencies ( 34 ). We saw no evidence for nonmultiplicative behaviors in our experiments, e.g., U-shaped DDRs ( 37 ). Our model was constructed with independent contributions from frequency and intensity, which is consistent with experimental results ( Fig. 5B ). However, it is possible that different experimental conditions or a higher-dimensional model might identify these types of complex relationships. Regardless of whether the IDH is valid, our work highlights the importance of understanding the mechanisms underlying disturbance-induced changes in diversity. In our case, we have framed unimodal DDRs as nonequilibrium mixtures of incompatible ecological states (defined by fitness tradeoffs along an environmental gradient) maintained by disturbance-induced environmental heterogeneity. We suggest that temporally stable DDRs (unimodal or otherwise) are rare due to the ubiquity of ecological and environmental feedbacks ( 50 ), which can dampen disturbance-induced environmental heterogeneity. Over longer time scales, evolutionary adaptation can alter the relationships between species traits and the environment, which could in turn alter DDR structure in persistently disturbed ecosystems. Moving forward, it will be important to replicate our results in other controlled systems with higher temporal resolution to assess their generality, investigate how ecosystem function and stability are affected by a community’s position along a DDR curve, and explore the intersection between disturbance-based coexistence and evolution (e.g., the potential for horizontal gene transfer, competition/cooperate tradeoffs, etc.). From the influence of antibiotics on pathogen susceptibility in the gut ( 51 ) to eutrophication in freshwater ecosystems ( 52 ), more reliable models for how microbial diversity responds to disturbance will inform our ability to predict the ecological stability of microbial systems in the presence of perturbations."
} | 3,044 |
29111438 | PMC5776718 | pmc | 1,885 | {
"abstract": "L-lysine and other amino acids are commonly produced through fermentation using strains of heterotrophic bacteria such as Corynebacterium glutamicum . Given the large amount of sugar this process consumes, direct photosynthetic production is intriguing alternative. In this study, we report the development of a cyanobacterium, Synechococcus sp. strain PCC 7002, capable of producing L-lysine with CO 2 as the sole carbon-source. We found that heterologous expression of a lysine transporter was required to excrete lysine and avoid intracellular accumulation that correlated with poor fitness. Simultaneous expression of a feedback inhibition resistant aspartate kinase and lysine transporter were sufficient for high productivities, but this was also met with a decreased chlorophyll content and reduced growth rates. Increasing the reductant supply by using NH 4 + , a more reduced nitrogen source relative to NO 3 − , resulted in a two-fold increase in productivity directing 18% of fixed carbon to lysine. Given this advantage, we demonstrated lysine production from media formulated with a municipal wastewater treatment sidestream as a nutrient source for increased economic and environmental sustainability. Based on our results, we project that Synechococcus sp. strain PCC 7002 could produce lysine at areal productivities approaching that of sugar cane to lysine via fermentation using non-agricultural lands and low-cost feedstocks.",
"introduction": "1. Introduction L-lysine is one of the essential amino acids required for human and animal growth. As demand for meat (e.g. poultry, swine, cattle) has grown, so has demand for essential amino acids, especially lysine ( Eggeling and Bott, 2015 ). Lysine and other amino acids are commonly produced by fermentation using strains of heterotrophic bacteria, such as Escherichia coli and Corynebacterium glutamicum ( Brautaset and Ellingsen, 2011 ). C. glutamicum has been engineered to produce lysine with a yield of 0.31 g lysine/g sugar. At this yield, current lysine demand (~1.85 million tons per year) would consume approximately 3% of world sugar production (~200 million tons per year – USDA-ERS). For this reason, it is important to consider alternative feedstocks and production strategies for essential amino acids. Direct photosynthetic production of L-lysine is an attractive process because it circumvents the need for harvesting, processing and fermentation of plant-derived sugars, by directly coupling lysine biosynthesis to photosynthesis. Cyanobacteria are attractive photosynthetic organisms for L-lysine production due to the availability of genetic tools, rapid growth rates, halotolerance, and ability to be grown on non-productive land with simple nutrient requirements ( Oliver and Atsumi, 2014 ; Pate et al., 2011 ) or wastestreams ( Korosh et al., 2017 ). Here, we report successful metabolic engineering strategies for production of L-lysine from a strain of the fast-growing Synechococcus sp. strain PCC 7002. The structure and regulation of amino acid biosynthesis has been well studied and used to design metabolic engineering strategies in various bacteria. For example, a lysine producing C. glutamicum strain was rationally designed by overexpressing feedback-resistant enzyme variants that increased flux in relevant anaplerotic and biosynthetic reactions ( Becker et al., 2011 ). L-lysine is a part of the aspartate-family of amino acids that also includes L-isoleucine, L-methionine, and L-threonine ( Kirma et al., 2012 ). The precursors and metabolites in this amino acid family are interconnected to several parts of central metabolism and are homeostatically controlled by several transcriptional and/or post-translational mechanisms ( Wittmann and Becker, 2007 ). In cyanobacteria, the anaplerotic reaction of phosphoenolpyruvate carboxylase (PEPC) combines PEP with HCO 3 − to generate oxaloacetate ( Fig. 1 ). Oxaloacetate, a component of the modified cyanobacteria TCA cycle ( Zhang and Bryant, 2011 ), is then converted to L-aspartate by aspartate aminotransferase where it enters the lysine biosynthetic pathway. High levels of L-aspartate negatively affect PEPC activity ( O’Regan et al., 1989 ). L-aspartate is further converted to L-aspartyl-phosphate by aspartate kinase (AK), an enzyme that is subject to feedback inhibition by L-lysine and L-threonine at regulatory subunits. L-aspartyl-phosphate is then converted by aspartate semialdehyde dehydrogenase into L-aspartate semialdehyde. From this branch point, L-aspartate semialdehyde may either be converted into L-homoserine, which is used for biosynthesis of the other members of the aspartate family of amino acids; or into L-2,3 dihydropicolinate by dihydrodipicolinate synthase (DHDPS) in the lysine biosynthetic pathway. Like AK, DHDPS is also subject to feedback regulation by L-lysine and L-threonine. The variant of the lysine biosynthetic pathway in plants and cyanobacteria is distinct in that it uses a L,L-diaminopimelate aminotransferase to convert L-2,3,4,5 tetrahydrodipicolinate to L,L-diaminopimelate as opposed to the dehydrogenase, acetylase, and succinylase pathways found in most other organisms ( Hudson et al., 2006 ). From L,L-diaminopimelate, subsequent epimerization and decarboxylation reactions are used to ultimately generate L-lysine, which then is then excreted by the transporter, lysE , in C. glutamicum ( Kelle et al., 1996 ). A corresponding transporter in Synechococcus sp. strain PCC 7002 has not been identified. Here, we report increased L-lysine excretion in the cyanobacterium Synechococcu s sp. strain PCC 7002 (hereafter PCC 7002) with heterologous expression of the amino acid exporter ybjE from Escherichia coli. We also examined the effects of feedback-resistant AK, DHDPS, and PEPC variants on L-lysine productivity. The best producing strain was cultivated in laboratory photobioreactors to study L-lysine production dynamics during growth to a light-limited stationary phase. We examined the effects of media composition on L-lysine production and cellular growth rates by using nitrate, ammonia, and a N-/P-rich sidestream from a municipal wastewater treatment facility as a nutrient source. Our work demonstrates the potential of engineered photoautotrophic amino acid production in cyanobacteria.",
"discussion": "4. Discussion In this project, we successfully engineered the cyanobacterium, Synechococcus sp. PCC 7002 to express feedback resistant versions of enzymes in the lysine biosynthetic pathway as a model for photosynthetic production of L-lysine. Simultaneous expression of an engineered aspartate kinase and lysine transporter was sufficient for high productivities, but also led to decreased chlorophyll content. Increasing the reductant supply by using NH 4 + , a more reduced nitrogen source relative to NO 3 , gave a twofold increase in productivity. Our highest volumetric productivity for lysine was 0.003 g L −1 h −1 , which is comparable carbon partitioning for PCC 7002 engineered to produce ethanol (0.010 g L −1 h −1 ) ( Dühring et al., 2014 ). However, this pales in comparison to the heterotrophic lysine productivity of 4.0 g L −1 h −1 achieved with C. glutamicum ( Becker et al., 2011 ). Some of the limitations for producing amino acids in photoautotrophic hosts may be due to the relatively small flux of the TCA cycle, which has been demonstrated under numerous growth conditions in 13 C-MFA studies ( Wan et al., 2017 ). The aspartate family of amino acids is intertwined in both catabolic and anabolic processes in the TCA cycle ( Galili, 2011 ). During photoautotrophic growth, the TCA cycle has been shown to primarily operate as a bifurcated pathway to generate necessary precursor metabolites, oxaloacetate and α-ketoglutarate ( Steinhauser et al., 2012 ). Although, several species of cyanobacteria have since been shown to have the enzymatic capacity to operate a closed TCA cycle through bypass reactions ( Xiong et al., 2014 ; Zhang and Bryant, 2011 ). The exact reason for this discrepancy has been highly debated, but it has been proposed to be a result of niche specialization ( Zhang et al., 2016 ). Researchers have recently modified photoautotrophic flux in the TCA cycle by the introduction of a heterologous GABA shunt, which may be used to extend the plasticity of this metabolic node in future studies ( Zhang et al., 2016 ). Tools such as flux balance analysis may also reveal novel metabolic engineering approaches suitable for the constraints of photoautotrophic metabolism ( Rügen et al., 2015 ). For example, introduction of a pyruvate carboxylase may also increase the metabolic flexibility of the PEP-pyruvate-oxaloacetate node, thereby leading to lead to an increase in precursor supply. While most organisms contain a subset of enzymes to ensure optimal carbon and energy flux at this anaplerotic node, C. glutamicum is a noted exception ( Sauer and Eikmanns, 2005 ). In lysine producing strains of C. glutamicum, flux through the energy consuming C3 rather than the C4 carboxlylation reaction is the predominant anaplerotic route, suggesting a need for increased futile cycling in these conditions ( Marx et al., 1999 , 1996 ). The high pyruvate/PEP ratio in PCC 7002 ( Dempo et al., 2014 ) may make this an attractive option for future work. Lysine production in TK.032 was met with decreased chlorophyll content per cell ( Fig. 3c ). Decreased chlorophyll content has previously been seen in tobacco plants engineered to have elevated levels of free lysine ( Azevedo et al., 1997 ). This implies a change in physiology during lysine production. This is contrary to cyanobacteria engineered to export sucrose ( Ducat, 2016 ). This finding may represent the current ceiling in carbon-flux redistribution during photoautotrophic lysine production, due to the limited metabolic flux of the autotrophic TCA cycle, which has been put forth by experimental and computational approaches ( Broddrick et al., 2016 ; Nogales et al., 2012 ; Song et al., 2015 ; Young et al., 2011 ). Given our laboratory strain performance in wastewater medium, we can extrapolate to a conservative aerial lysine productivity of ~100 (Winter) - 250 (Summer) g of Lysine/m 2 /year using estimates of aerial biomass productivities from 9.4 (Winter) - 23.5 (Summer) gDCW/m 2 /day in an open pond system with a marine cyanobacterium ( Moreno et al., 2003 ), and the 3% of carbon flux to lysine in our wastewater production conditions. In comparison, current United States areal productivities are ~775 g of Lysine/m 2 /Year. This figure is based on the mean annual yield of sugar crops in the United States (Michael McConnell, n.d.), assuming a glucose production rate of 6.5 mol/m 2 /Year and the maximum theoretical yield (0.82 mol L-lysine mol −1 glucose) for C. glutamicum ( Wittmann and Becker, 2007 ). These figures highlight the areal productivity of cyanobacteria utilizing non-traditional nutrient sources, with the added benefit of wastewater polishing for discharge. That said, to take advantage of this trait, titers must be increased to keep the cost of purification low ( Hermann, 2003 ) and productivities must be increased to reduce the capital costs needed to produce a target amount of amino acid. It is also important to note that these extrapolated productivities do not take into account operating costs associated with purification and extraction ( Quiroz-Arita et al., 2017 ) or any spatial variation in the availability of wastewater ( Roostaei and Zhang, 2017 ). The integration of the multitude of ‘-omics’ data into genome scale metabolic models, in particular with isotopically nonstationary metabolic flux analysis in over producer strains ( Adebiyi et al., 2015 ), will enable more accurate representations of the flux control of the branch points in non-model organisms, and further development of a photosynthetic chemical platform."
} | 2,979 |
35316950 | PMC8889185 | pmc | 1,886 | {
"abstract": "Coordinated responses in eusocial insect colonies arise from worker interaction networks that enable collective processing of ecologically relevant information. Previous studies have detected a structural motif in these networks known as the feed-forward loop, which functions to process information in other biological regulatory networks (e.g. transcriptional networks). However, the processes that generate feed-forward loops among workers and the consequences for information flow within the colony remain largely unexplored. We constructed an agent-based model to investigate how individual variation in activity and movement shaped the production of feed-forward loops in a simulated insect colony. We hypothesized that individual variation along these axes would generate feed-forward loops by driving variation in interaction frequency among workers. We found that among-individual variation in activity drove over-representation of feed-forward loops in the interaction networks by determining the directionality of interactions. However, despite previous work linking feed-forward loops with efficient information transfer, activity variation did not promote faster or more efficient information flow, thus providing no support for the hypothesis that feed-forward loops reflect selection for enhanced collective functioning. Conversely, individual variation in movement trajectory, despite playing no role in generating feed-forward loops, promoted fast and efficient information flow by linking together otherwise unconnected regions of the nest.",
"conclusion": "5 . Concluding remarks and future directions Further research is required to establish whether and how feed-forward loops impact the collective functioning of social insect colonies. Central to these efforts is quantifying the extent to which feed-forward loops and other network motifs are present within colony interaction networks. A common approach is to compare empirical networks with Erdős-Rényi random networks matched for size and density, yet these null models often lack biological and physical relevance [ 58 , 59 ]. For example, random graphs typically assume that all individuals are equally likely to interact, thus ignoring spatial and temporal constraints on interactions (e.g. two individuals that generally occupy opposite sides of the nest are unlikely to interact). A potential application of our model lies in the generation of spatially explicit null models, tuned to a particular system, that will enable realistic comparison with empirical data. To illustrate this point, we reanalysed previously published data on interaction networks of the ant P. californicus [ 7 ], using our agent-based simulations to generate spatially explicit null models that match the empirical data in network size and density (see electronic supplementary material for details on this analysis). Comparing the empirical networks with random graph models, the original study concluded that feed-forward loops were over-represented, while three-cycles ( figure 1 b ) matched expected frequencies (figure 3 in [ 7 ]). Conversely, our method suggests that both substructures are over-represented in the empirical data relative to the simulated data ( figure 7 ). We stress that our reanalysis does not invalidate the findings of [ 7 ]—indeed, our model is not parametrized appropriately for their data in terms of ant worker activity and movement. Rather, these results emphasize the important role that selecting a null model plays in the interpretation of network analyses. By offering a means to generate spatially explicit null models, we anticipate that our model will prove useful for future investigations into the mechanisms that drive the structure of animal social networks.\n Figure 7 . TSP comparing the relative significance (measured by Z scores) of triangle configurations found in Waters and Fewell's [ 7 ] data compared with null models generated from our simulation using either activity variable or uniform conditions. Normalized Z-scores were averaged across 12 networks with varied sizes and across 100 simulation runs for each condition; bars indicate the standard errors.",
"introduction": "1 . Introduction In many group-living species, social interaction patterns play an important role in shaping fitness outcomes, such as by impacting access to social information, the likelihood of cooperation or exposure to pathogens [ 1 ]. Beyond an individual's direct connections, evolutionary fitness may further be influenced by the patterning of interactions at the group level. For instance, a minority of highly interactive individuals can accelerate the spread of information or disease throughout a population by linking together otherwise unconnected individuals [ 2 ] and modular social structures can contain the spread of information within tightly knit communities [ 3 ]. These group-level properties are probably especially important in eusocial insect colonies in which only one or a few colony members reproduce, such that the fitness of individual workers is tightly linked to colony collective performance [ 4 ]. In the absence of any central control, interactions among workers regulate task allocation through a distributed process, ensuring that the effort devoted to various tasks matches a colony's internal needs and external conditions [ 5 , 6 ]. Meeting the demands of this regulatory role is likely to favour different interaction patterns to those that are observed in social systems where individual success is key. In other words, a social insect worker's position in the network is often less important for its fitness than higher level network properties [ 5 – 7 ]. For instance, among-individual variation in interaction frequency in harvester ants ( Pogonomyrmex barbatus ) generates networks with a few highly interconnected individuals while the majority of workers remain only weakly connected [ 2 ]. Networks structured in this way permit rapid information transfer throughout a population and can thereby facilitate swift collective responses to changing conditions [ 2 ]. Social network analysis has emerged as a key approach for quantifying variation in social connectivity and investigating its ecological and evolutionary consequences [ 1 , 8 , 9 ]. A useful means to gain insight into a network's functionality is to deconstruct it into its constituent subcomponents [ 10 ]. A network can, for instance, be described in terms of the different three-node subgraphs (or triads) from which it is composed. Because subgraphs differ in their functional properties [ 11 – 13 ], over-representation of a given subgraph within a network (relative to its typical representation within an ensemble of appropriately randomized networks) can suggest the processes or functions that have helped to shape that network. For example, food webs display an over-representation of simple chains derived from trophic interactions—e.g. species A consumes species B, which in turn consumes C [ 10 ]—whereas gene transcription networks contain an over-abundance of a triadic configuration known as the ‘feed-forward loop’ ( figure 1 a ; [ 12 ]), whereby a gene A transcriptionally regulates the activity of a second gene B, and both A and B jointly regulate a third gene C. As feed-forward loops are well-suited to carry out signal processing tasks (e.g. amplifying responses to external environmental cues), this structural feature may reflect the regulatory function of these networks [ 11 ].\n Figure 1 . Examples of triangle configurations. ( a ) A transitive triangle, or feed-forward loop, in which one individual, A, has two outgoing edges and another, C, has two incoming edges. ( b ) A cyclic triangle, in which all individuals have one outgoing and incoming edge. ( c ) A triangle with a bidirectional edge connecting B and C. The regulatory role of interaction networks within social insect colonies may likewise be reflected in their constituent subcomponents. In common with other biological regulatory systems, the antennation patterns of harvester ants ( P. californicus ), which play a key role in transmitting task-relevant information between colony members, show an over-representation of feed-forward loops [ 7 ], at least relative to their appearance in size- and density-matched random graphs. Similarly, dominance relationships in the eusocial wasp Ropalidia marginata are made up predominately of feed-forward loops and are involved in regulating worker activity through agonistic interactions [ 14 ]. However, the way in which such network structures develop in social insect colonies is unclear. In contrast with other biological regulatory networks, where the relationships between nodes are relatively stable (e.g. one gene produces a transcription factor that activates or inhibits another gene), the nodes in social insect interaction networks represent individual workers that engage in brief pairwise interactions with one another and often lack stable relationships. The patterning of interactions among workers instead arises from the behaviour of individuals that influence their likelihood of interacting. In some cases, the presence of feed-forward loops might simply reflect the tendency of a particular type of relationship to be transitive—e.g. in the dominance networks of R. marginata [ 14 ], if worker A is dominant over B and B is dominant over C, A is likely to also be dominant over C. Feed-forward loops consequently tend to form in the network. Yet even in this case, the formation of dominance relationships is dependent on other aspects of behaviour that influence individuals' likelihood of interacting, such as their spatial location on the nest. If A and C never interact, a feed-forward loop will never form between A, B and C. Furthermore, the reason for the development of feed-forward loops in interaction networks that lack such hierarchical organization (e.g. the antennation patterns of P. californicus ) is less clear and suggests that more subtle behavioural mechanisms may be responsible for generating this structural feature in these populations. The structure of a feed-forward loop inherently implies among-individual variation in contact patterns, as each node differs in the number of incoming and outgoing connections (or edges). Insect workers express substantial among-individual variation along a number of behavioural axes that may contribute to the generation of such network structures [ 15 ]. For instance, workers vary in the proportion of time that they are actively engaged in tasks: e.g. a minority of workers often carry out the majority of work [ 16 – 20 ], with some workers even appearing to specialize in inactivity [ 21 ]. Workers also vary in their spatial behaviour within the nest. This can partly be determined by activity levels—more active individuals will tend to cover more ground per unit time—but can also result from variation in movement patterns. Some P. barbatus workers, for example, walk very sinuous paths, causing them to occupy relatively restricted regions within the nest, while others walk straighter paths and so roam more extensively [ 2 ]. Such among-individual variation in activity and space-use may play central roles in shaping social contact patterns by influencing the likelihood that particular individuals will contact one another. For example, workers that move in straighter paths will probably contact a greater number of nest-mates than workers that remain restricted to small regions within the nest. That different pairs of individuals vary in their likelihood of interacting further suggests that random graph models, which typically assume an equal probability of interaction between any pair of nodes, may not be the most appropriate null model with which to assess the presence of network motifs in empirical social insect interaction networks. Here, we construct an agent-based model to investigate how among-individual variation in activity and movement patterns in a simulated insect colony contributes to the formation of interaction networks dominated by feed-forward loops. We further consider how this variation drives the speed and efficiency of information flow within the colony. Our model is not designed to reproduce the dynamics of any specific species. Rather, we seek to evaluate structural and functional consequences of patterns of behavioural variation that are commonly observed across eusocial insects [ 15 , 22 ], with a particular focus on how such variation shapes patterns of physical contact between workers (e.g. antennation), which are central in regulating collective behaviour [ 2 , 6 , 7 ]. We first predict that, by determining how frequently individuals contact others and how diverse those contacts are, among-individual variation in activity and movement will drive over-representation of feed-forward loops in the resulting interaction networks. We further predict that, when these sources of variation are treated as a ‘behavioural syndrome’ (i.e. individual activity and patterns of movement covary, such that the most active agents also walk straighter paths), they will have a synergistic effect on the production of feed-forward loops by emphasizing among-individual variation in space-use. Second, due to the tendency of feed-forward loops to move information in a directional manner [ 11 , 12 ], we predict that patterns of behavioural variation that generate feed-forward loops will also lead to faster and more efficient information flow, in the sense that fewer interactions will be needed to drive the spread of information throughout a colony [ 7 ].",
"discussion": "4 . Discussion The superorganismal nature of eusocial insect colonies means that natural selection is increasingly expected to act on colony-level traits [ 4 ], such as the ability to generate robust, yet flexible, colony-level responses to ecological challenges. Collective coordination relies on interactions that transfer information between nest-mates, raising the possibility that natural selection has acted on the behavioural algorithms that determine whether and how workers interact. Using a simple agent-based model, we demonstrate that among-individual variation in the likelihood of sending outgoing (or receiving incoming) links is sufficient to generate an over-abundance of a triadic network substructure known as the ‘feed-forward loop’ ( figure 1 a ). This motif is commonly found in biological regulatory networks where it performs various signal processing tasks, e.g. discriminating persistent signals from short-lived pulses [ 10 , 11 ], and is also over-represented within social insect interaction networks, where similar regulatory roles have been demonstrated [ 7 , 14 ]. Nevertheless, our model found that feed-forward loops alone had little impact on information transmission processes. Rather, among-individual variation in movement patterns (either alone or as part of a behavioural syndrome) promoted faster and more efficient information transfer, despite contributing little to the production of feed-forward loops. Our model thus demonstrates how collective properties that support colony functioning can be tuned by modifying both the behavioural variation present among workers and correlations across traits. Insect workers often vary considerably in their activity levels [ 16 – 19 ], with a minority of individuals generally carrying out most of the work [ 19 , 20 , 24 ]. These individuals can also play a key role in transmitting task-relevant information through interactions with nest-mates [ 15 ]. Honeybee ( Apis mellifera ) foragers, for example, vary dramatically in their likelihood to produce recruitment dances, even when collecting from identical resources [ 41 ]. Similarly, highly active ‘keystone individuals' catalyse worker activity in ant colonies [ 42 ]. We therefore linked activity in our model to the likelihood of directed information transfer between individuals and found that, when this criterion was satisfied, individual variation in activity drove the production of feed-forward loops within the interaction networks. However, given that other effects of activity variation in our model (e.g. total distance moved) were unimportant for the generation of these motifs, it seems likely that any behavioural trait that (i) varies among individuals and (ii) directly influences the directionality of pairwise interactions (e.g. the direction of information transfer) could drive an over-abundance of feed-forward loops. One such trait may be the propensity to interact with nest-mates. For example, honeybees vary in their likelihood to engage in trophallactic food-sharing interactions [ 43 ], with some individuals potentially specializing in offering food [ 44 ]. Dominance interactions are also characterized by clear directional relationships—indeed, transitive relationships are a common feature in dominance hierarchies, in both insects [ 14 ] and other taxa [ 13 ]. In various ant species, for example, trophallaxis is generally directed from subordinate to dominant individuals [ 22 ]. Variation in knowledge or past experience is also likely to promote transitive network structures when it results in directed information transfer among workers. For example, more experienced Temnothorax albipennis ants are more likely to engage in tandem runs, where they directly lead naive followers to a resource [ 45 ]. Similarly, it has been suggested that in the grass-cutting ant ( Acromyrmex heyeri ), workers initially sacrifice foraging efficiency in order to more rapidly provide nest-mates with information about newly discovered foraging resources [ 46 ]. Nevertheless, while transitive network structures are a potentially common feature of social insect colonies, whether they offer any functional benefit remains unclear. Previous analyses of empirical social insect networks have shown that an over-representation of feed-forward loops could reflect selection for more efficient information transfer in insect colonies [ 7 , 14 ]. However, our model found that the speed and efficiency of information transfer was unrelated to the proportion of transitive triangles in the population social network. For example, among-individual variation in activity alone produced comparable triangle transitivity levels compared with when individuals varied in both activity and turning index, but the former was associated with relatively slow and inefficient transmission compared with the latter. This suggests that the effects of feed-forward loops on collective functioning are likely to be context-dependent. It is also possible that feed-forward loops confer regulatory properties beyond those considered here. For instance, in transcriptional networks, feed-forward loops can dampen responses to external signals to ensure that transient signals are ignored [ 11 ]. A similar role may be present in insect colonies by limiting collective responses to weak signals about low-quality resources and thereby promote effective worker allocation. Workers often vary in their response thresholds to task-related stimuli, with some requiring relatively little stimulation to begin work, while others must experience substantially higher intensities of task-related stimuli before acting [ 22 ]. Feed-forward loops may regulate worker activation by limiting responses to weak task-relevant stimuli, while ensuring sufficient stimulation (e.g. multiple signals from active workers) is received by inactive workers when more help is truly needed. It is also possible that, in some cases, the production of feed-forward loops is simply an inadvertent by-product of the behavioural variation present within insect colonies and not itself a target of selection. Previous work has shown, for example, that the frequency and nature of lower level dyadic interactions play a key role in determining the types of triadic configurations that can arise in a network [ 47 ]. It is worth noting that our model assumed that behavioural variation remained constant over time. In reality, an individual's activity and/or propensity to interact with others may shift in response to factors such as worker loss, changes in colony food stores or the discovery of a new resource, and these changes may in turn influence how information is transferred through the colony [ 19 , 29 ]. Nevertheless, while our model represents a simplified transmission scenario, it demonstrates clearly how variation in simple individual-level behaviours can significantly impact colony-level information transfer. It also highlights the challenge in inferring the functionality of dynamic systems from knowledge of the static network structure alone. Within insect colonies, interactions are often brief and stable relationships between particular individuals are generally absent. Under such conditions, very different patterns of interaction can give rise to similar network structures when aggregated over time [ 39 ]. Whereas previous analyses of the function of feed-forward loops have focused on systems with relatively fixed relationships (e.g. gene regulatory networks; [ 11 ]), within insect colonies, the timing and order of interactions is of critical importance. Indeed, when we simulated information flow on the static networks derived from our time-ordered interaction lists, rather than on the time-ordered interactions themselves, we found that in agreement with previous studies [ 11 , 14 ] information spread more efficiently on networks characterized by an over-representation of transitive triangles (electronic supplementary material, tables S6 and S7 and figure S6). In contrast with among-individual variation in activity, individual variation in movement paths often improved both the speed and efficiency of information transfer in our model, despite having limited impact on the generation of network transitivity. Spatial behavioural variation was included in our model in terms of walking path sinuosity, causing some individuals to remain in restricted areas of the nest, while others traversed the entire nest space [ 2 , 28 ] (electronic supplementary material, figure S3). Under certain conditions, such variation in space-use allowed for faster and more efficient information transfer through the colony. In particular, these effects were observed either when individuals varied in path sinuosity alone or when activity levels were negatively correlated with turning indices across the population—that is, agents with sinuous walking paths tended to be inactive while those with straighter walking paths were often active. In many eusocial insect species, similar patterns of space-use variation have been observed. Bumblebees ( Bombus terrestris ), for example, perform irregular ‘excited’ runs throughout the nest after returning from successful foraging trips, which serve to increase foraging activity in other workers by rapidly distributing pheromones, and potentially through physical contacts [ 48 , 49 ]. Similarly, red harvester ants ( P. barbatus ) vary in the sinuosity of their walking trajectories, which influences their interaction frequency. Ants with straighter walking paths contact more nest-mates than those with more tortuous paths [ 2 ]. Our model is consistent with the hypothesis that such variation in connectivity facilitates rapid information flow throughout the population due to workers with straighter walking paths linking isolated clusters of individuals [ 2 ]. It should be noted, however, that the adaptiveness of fast, efficient information transfer is highly context dependent. In response to predation, for example, insect colonies are likely to benefit from rapid alarm propagation that can quickly marshal colony defences [ 50 , 51 ], whereas rapid information transfer may be less valuable in a foraging context. Instead, the regulation of information transmission in response to environmental feedback is key to ensuring worker effort is divided according to the quality of resources [ 52 ], and colonies that show restraint in foraging efforts can often be more successful [ 53 , 54 ]. In addition, behavioural variation that promotes fast and efficient information transfer may also promote faster transmission of pathogens. In this case, we would expect natural selection to favour collective responses to the infiltration of pathogens that limit unnecessary interactions. On exposure to pathogens, for example, some ant species switch from allogrooming to self-grooming—or even isolate themselves from other workers completely—thus reducing potential infection of healthy nest-mates [ 55 , 56 ]. Similarly, nest architecture can influence disease spread throughout a colony, with physically or behaviourally segmented nests tending to dampen the spread of disease [ 56 , 57 ]."
} | 6,198 |
37454222 | PMC10349867 | pmc | 1,887 | {
"abstract": "The role of microbial interactions and the underlying mechanisms that shape complex biofilm communities are poorly understood. Here we employ a microfluidic chip to represent porous subsurface environments and show that cooperative microbial interactions between free-living and biofilm-forming bacteria trigger active spatial segregation to promote their respective dominance in segregated microhabitats. During initial colonization, free-living and biofilm-forming microbes are segregated from the mixed planktonic inoculum to occupy the ambient fluid and grain surface. Contrary to spatial exclusion through competition, the active spatial segregation is induced by cooperative interactions which improves the fitness of both biofilm and planktonic populations. We further show that free-living Arthrobacter induces the surface colonization by scavenging the biofilm inhibitor, D-amino acids and receives benefits from the public goods secreted by the biofilm-forming strains. Collectively, our results reveal how cooperative microbial interactions may contribute to microbial coexistence in segregated microhabitats and drive subsurface biofilm community succession.",
"introduction": "Introduction The terrestrial and oceanic subsurface hosts over 80% of microorganisms on Earth and is thus the major microbial habitat on our planet 1 , 2 . Unlike aqueous environments (e.g. open ocean) where microbes are mostly free living (planktonic), the subsurface provides immensely large surface area for microbial attachment. The surface-attached microbes sequester nutrients from pore water and grow into dense multi-species assemblages, called biofilms 3 , 4 . The close proximity of diverse species in biofilms facilitates various interactions between them, such as quorum sensing and synergistic metabolism, which determine the community traits and functions 5 – 7 . In the past decades, theoretical and experimental research has been performed to dissect the intricate interactions that dictate subsurface biofilm community structure. Cooperative microorganisms such as cross-feeding partners are found to co-aggregate in biofilm communities to allow reciprocal benefits 8 – 10 . By contrast, the mutually antagonistic microorganisms tend to exclude each other from local niches and segregate spatially 7 , 11 . Besides direct functional consequences, the physical structure of subsurface environments can determine ecological stability and functional activities by modulating the spatial distribution of cooperative and competitive genotypes. Compared with well-mixed environments, spatial segregation under structured conditions balances the competitive and cooperative interactions to stabilize the community 12 . For instance, physical separation in porous media enables the coexistence of slow-growing species with fast-growing competitors, as the rapid biofilm formation blocks fluid flow and redirects nutrients to its competitors 13 , 14 . One recent experiment also demonstrated that the spatial segregation of biofilm consortia governs the metabolite cross-feeding and microbial growth via tuning the fidelity of quorum sensing signal transmission 15 . The current understanding of interaction-derived subsurface biofilm communities however, largely rests on dual-species communities. How microbial interactions shape diverse biofilm communities in spatially structured environments is still poorly understood. Here, we investigated the biofilm colonization process in a porous medium where soil bacteria self-assemble into structured microbial communities. Using microfluidics, 16S rRNA gene amplicon sequencing, and fluorescence in situ hybridization (FISH), we observed that during early biofilm development, biofilm-deficient species actively primed the microscale environment for biofilm-forming microbes to colonize surfaces. We further performed exometabolomics, transcriptomics, pairwise interaction analyses and genetic manipulation to uncover the mechanisms of interspecific interactions. We find that the interaction between biofilm-deficient and biofilm-forming species drives microbial community succession through active spatial segregation in the subsurface environment.",
"discussion": "Discussion Subsurface porous environments such as soil and sediment are one of the major microbial habitats on Earth 1 , 2 . In these environments, biofilm formation is a fundamental living strategy for microorganisms. The intricate interactions within local communities give rise to community diversity and stability. The role of microbial interactions in shaping complex natural biofilm community successions, however, is rarely examined. Here, using a microfluidic chip environment, we demonstrate that a positive microbial interaction drives the community spatial structure during biofilm colonization. Biofilm-deficient or more planktonic Arthrobacter triggers biofilm formation via removal of biofilm inhibitors. With the decreased level of biofilm inhibitors, Pseudomonas , the most abundant genus in the total community, reduced its proportion in the ambient fluid and switched to a sessile lifestyle. The spatial segregation was also observed in co-culture containing Arthrobacter and Pseudomonas strains with comparable initial cell densities. The pairwise interaction analyses revealed that Arthrobacter and biofilm-forming strains dominated the plankton and biofilm in the co-culture systems, respectively. The proportion of Pseudomonas that inhabited biofilms in co-culture was significantly higher than that in monoculture ( p < 0.001, one-way ANOVA, Supplementary Table 6 ). Although the growth in ISEM and M9 minimal medium with different DAA levels suggests that Arthrobacter doesn’t gain fitness benefits through DAA hydrolysis (Supplementary Fig. 19 ), it receives return benefits from biofilm-forming species which secrete public goods. The mutually beneficial interaction facilitates the occupation of segregated niches by free-living Arthrobacter and biofilm-forming bacteria and shapes community spatial organization. It indicates that the planktonic microbial minorities can impose strong selection pressure on biofilm community succession, by serving as inhibitor scavengers and, in turn, enhance their own fitness in the pore fluid. Simulated and experimental results have demonstrated that spatial segregation is generally an outcome of competitive interactions in structured environments 5 , 30 , 31 . As competing microorganisms reciprocally exclude one another, spatial segregation contributes to the stable coexistence of competitors and increases community diversity 12 , 32 . In particular, biofilm-forming species gain competitive advantages over free-living or biofilm-deficient species in natural environments. Previous studies found biofilm-forming species outcompete free-living cells via smothering them or cutting off nutrient access 33 , 34 . By contrast, we show that free-living species, which are hitherto recognized as the outcompeted species, can in fact establish mutualistic cooperation with biofilm-forming cells. In the early stage of colonization, the free-living microbes scavenge the universal biofilm inhibitor to induce biofilm formation. With reduced inhibitory effect, biofilm-forming cells depart from ambient fluid to colonize grain surfaces and secrete public goods in return. These early surface settlers can rapidly preempt or modify the microhabitats, which impacts the colonization of late-arriving species and drives subsequent community succession 8 , 35 , 36 . This indicates that the cooperative interaction between free-living and biofilm-forming partners may hold a central role during the colonization of virgin territories. We find that DAAs are the key metabolites that mediate spatial segregation in our system. The changes in DAAs during biofilm colonization reveal a V-shaped pattern. At the late stage, the frequency of Arthrobacter capable of efficiently hydrolyzing DAAs decreased significantly due to its deficiency in biofilm formation. The mean travel time of colloids in the microfluidic chamber decreased from 51.48 ± 2.16 to 30.72 ± 5.16 min after the biofilm development (two-tailed Student’s t -test, n = 3 independent replicates for each condition, p = 0.011, Supplementary Fig. 2 ), indicating an accelerated elution of planktonic Arthrobacter at the late stage. The extinction of Arthrobacter in the porous environments led to the diminished metabolic capabilities of the remaining microbial community towards DAAs, the concentration of which in the effluent gradually increased to the original level in ISEM (Fig. 4d ). As a common group of secondary metabolites and major constituents of bacterial walls, these amino acid enantiomers can be actively excreted and passively released to surrounding environments 37 , 38 . Micromoles of DAA per gram dry weight were detected in sediment and soil 37 , 39 , 40 . As DAAs over this concentration range are sufficient to suppress biofilm formation 26 , 27 , they may therefore play a pervasive role in governing biofilm community structure over a wide range of ecosystems. This study provides a mechanistic understanding of how microbial interactions may govern biofilm community succession in subsurface porous environments. Further efforts are required to address the roles of different biotic interactions, such as competition and mutualism in shaping the ecological stability and functions of complex biofilm communities, and to validate these findings in real soils and microbial communities. This experimental platform can be adapted to enable in situ visualization and integrate multiple omics technologies to elucidate spatiotemporal dynamics of microbial activities and underlying mechanisms, which will advance our understanding of the spatial organization of microbial communities and functional traits in ecosystems."
} | 2,467 |
27775869 | PMC5399005 | pmc | 1,888 | {
"abstract": "Summary Lignin is a major polymer in the secondary plant cell wall and composed of hydrophobic interlinked hydroxyphenylpropanoid units. The presence of lignin hampers conversion of plant biomass into biofuels; plants with modified lignin are therefore being investigated for increased digestibility. The bacterium Sphingomonas paucimobilis produces lignin‐degrading enzymes including LigD, LigF and LigG involved in cleaving the most abundant lignin interunit linkage, the β‐aryl ether bond. In this study, we expressed the LigD , LigF and LigG ( Lig DFG \n ) genes in Arabidopsis thaliana to introduce postlignification modifications into the lignin structure. The three enzymes were targeted to the secretory pathway. Phenolic metabolite profiling and 2D HSQC NMR of the transgenic lines showed an increase in oxidized guaiacyl and syringyl units without concomitant increase in oxidized β‐aryl ether units, showing lignin bond cleavage. Saccharification yield increased significantly in transgenic lines expressing Lig DFG \n , showing the applicability of our approach. Additional new information on substrate specificity of the Lig DFG enzymes is also provided.",
"introduction": "Introduction Lignin is one of the most abundant biopolymers in the world. Together with cellulose, hemicelluloses, pectin and additional minor components that constitute the plant cell wall, lignin offers mechanical strength and affords protection against pathogens. The hydrophobic properties of lignin make it a crucial polymer controlling water conduction (Bonawitz and Chapple, 2010 ). During the polymerization process, lignin fills up spaces between polysaccharides in the secondary plant cell wall and may be covalently cross‐linked to some of these. The properties of lignin impede enzymatic lignocellulose deconstruction and thus constitute a major bottleneck in biofuel production (Pauly and Keegstra, 2010 ; Van Acker et al ., 2013 ; Vanholme et al ., 2013a ). Energy‐demanding and expensive pretreatment is required to facilitate the access of hydrolytic enzymes to the polysaccharides in the cell wall. Modification of the cell wall structure in ways that ease deconstruction without compromising the agronomic performance of the crop plant is therefore an important research topic. However, plants with large reductions in lignin content exhibit reduced growth, fitness and development. Altering lignin composition and structure might therefore be a better strategy to improve the efficiency of biomass processing (Bonawitz and Chapple, 2010 ; Li et al ., 2008 ; Ralph, 2007 ; Sederoff et al ., 1999 ; Vanholme et al ., 2008 , 2012 ). Lignin is a complex aromatic polymer derived mainly from three monolignols: p ‐coumaryl alcohol, coniferyl alcohol and sinapyl alcohol. When incorporated into lignin, these monolignols give rise to the p‐ hydroxyphenyl (H), guaiacyl (G) and syringyl (S) units, respectively. Once monolignols are synthesized in the cytoplasm, they are exported to the apoplast, activated to form radicals by the action of peroxidases or laccases and thereby subjected to radical coupling polymerization reactions. The most abundant interunit linkage type, formed primarily by the coupling of a monolignol (at its β‐position) to the phenolic end (at its 4–O‐position) of a growing polymer chain, is β–O–4 referring to an ether linkage between the aliphatic side‐chain of one monolignol and the phenolic moiety of another. Additional linkage types formed are β–5, 5–5, β–β and 4–O–5 (Boerjan et al ., 2003 ; Ralph et al ., 2004 ). The ability of microorganisms to degrade lignin has been extensively studied. White‐rot and brown‐rot fungi are involved in high‐molecular‐mass lignin degradation whereas bacteria like Sphingomonas paucimobilis SYK‐6 degrade low‐molecular‐mass lignin fragments and small oligomers (Kirk and Farrell, 1987 ; Masai et al ., 2007 ). The metabolism of S. paucimobilis has been evolutionarily adapted to allow the bacterium to grow on lignin‐derived compounds. This involves development of enzymatic machinery to fully degrade different types of units. Reductive cleavage of β‐aryl ether units has been demonstrated in vitro using guaiacylglycerol‐β‐guaiacyl ether (GGE) as a model substrate (Masai et al ., 2003 ) (Figure 1 ). The β–O–4‐linked moieties are reductively cleaved in the course of a three‐step sequence that involves initial dehydrogenation (catalysed by LigD, LigL and LigN enzymes), followed by reductive cleavage of the ether linkage catalysed by a glutathione S‐transferase (LigE or LigF), involving the formation of a glutathione conjugate, and finally reductive cleavage of this conjugate, producing the monomer and oxidized glutathione in a step using additional glutathione, catalysed by a glutathione lyase (LigG). The enzymes catalysing these processes are highly stereospecific. The GGE substrate exists as four enantiomers. The α R ‐GGE enantiomers are dehydrogenated by LigD, whereas the αS‐GGE enantiomers are dehydrogenated by LigL or LigN. The LigF enzyme conjugates the β S enantiomer of α‐(2‐methoxyphenoxy)‐β‐hydroxypropio‐vanillone (MPHPV) with glutathione (GSH) whereas the β R enantiomer is conjugated by LigE (Sato et al ., 2009 ). The substrate specificity assays were mainly performed on guaiacyl dimers (Gall et al ., 2014a ; Masai et al ., 2007 ), but in vitro the Lig enzymes are also capable of degrading synthetic high‐molecular‐mass polymers mimicking the structure of lignin (Sonoki et al ., 2002 ). Figure 1 Pathway for degradation of β–O–4‐linked units by Lig enzymes from S. paucimobilis exemplified with the model compound GGE (Masai et al ., 2003 ). GGE , guaiacylglycerol‐β‐guaiacyl ether; MPHPV , α‐(2‐methoxyphenoxy)‐β‐hydroxypropiovanillone; GS ‐ HPV , α‐glutathionyl‐β‐hydroxypropiovanillone; HPV , β‐hydroxypropiovanillone; GSH , glutathione. A recent paper described an attempt to engineer LigD into the plant with the idea of creating benzylic‐oxidized lignins that might chemically degrade more readily (Tsuji et al ., 2015 ). In this study, the codon‐optimized S. paucimobilis SYK‐6 genes encoding the complete set of lignin‐degrading enzymes LigD, LigF and LigG (henceforth abbreviated LigDFG) were inserted into the Arabidopsis genome by stable transformation using transit peptides targeting each of the heterologous proteins to the secretory pathway.",
"discussion": "Discussion Lignin embedded in the plant cell wall impedes enzymatic lignocellulose deconstruction for biofuel production. In this study, we targeted lignin‐degrading Lig enzymes from the bacterium S. paucimobilis to the apoplast of Arabidopsis with the aim of modifying lignin structure in a manner that would improve the accessibility of polysaccharide‐hydrolysing enzymes without influencing the overall content of lignin. The overall lignin composition, in terms of G and S units and linkage types, was not changed in the transgenic LigDFG Arabidopsis lines, except that the degree of α‐oxidation of G and S units was approximately twofold increased, as deduced from the aromatic regions of the 2D HSQC NMR spectra (Figures 4 and S3). The detected changes in lignin structure were significant and displayed little variation between the individual transgenic plant lines obtained. In Arabidopsis plants expressing only LigD , an increase in oxidized β–O–4‐ether structures (Aox structures in Figure 5 ) was observed consistent with the action of LigD on the Cα‐OH of β–O–4‐linked units (A structures in Figure 5 ) as exemplified in Figure 8 (Tsuji et al ., 2015 ). Upon expression of the entire LigDFG pathway, the LigD product serves as substrate for LigF, and therefore, the oxidized β–O–4‐ether structures (Aox) are further metabolized. In the LigDFG lines, the observed increase in the content of oxidized structures (Gox and Sox units) in the absence of elevated levels of oxidized β–O–4‐ether structures (Aox) (Figure 4 ) demonstrated that the entire LigDFG pathway was functional: the set of LigDFG enzymes reductively cleaved some of the β–O–4‐units resulting in an increase in HPV‐like end‐units. Naturally occurring benzaldehyde and benzoic acid end‐units will contribute to the intensity of Gox and Sox signals (Boerjan et al ., 2003 ; Rahimi et al ., 2013 ). However, LigDFG activity would not influence the content of, nor modify, these groups and therefore the increase in Gox and Sox in the transgenic LigDFG lines was derived from reductive cleavage of α‐oxidized β–O–4‐units. The unaltered Aox/A ratio demonstrated that under the given experimental conditions, the LigD‐step was rate limiting, that is those G units with β–O–4‐bonds that are oxidized to α‐ketones by LigD are cleaved by the action of LigF and LigG. Figure 8 Cleavage of a β–O–4‐bond in lignin by the Lig DFG enzymes depicting their theoretical contribution to the observed increase in Gox units as revealed by the 2D HSQC NMR spectra. The observed reductive cleavage of some of the β–O–4‐linked lignin units in the transgenic LigDFG lines caused a moderate, but significant, increase in the digestibility of the cell wall by polysaccharide‐hydrolysing enzymes. In a recently published study, Tsuji et al . ( 2015 ) did not detect increased saccharification yields in transgenic Arabidopsis plants expressing LigD alone, although it was expected that the increase in oxidized β–O–4‐structures should render the lignin more easily degradable by alkaline pretreatment. Similarly to the present study, only minor changes in lignin structure were observed by 2D HSQC NMR analysis, although this might reflect issues with respect to achieving apoplastic localization of the enzyme. At the current stage, it may thus be concluded that expression of the entire LigDFG pathway with the amy and ppi signal peptides may be more efficient in terms of achieving increased enzymatic cell wall decomposition. The data seem to suggest that the cleavage of the oxidized β–O–4‐structures (Aox) achieved by expressing LigDFG is needed to improve the saccharification, not merely the oxidation of the β–O–4‐structures. Phenolic profiling of the transgenic amyDFG10 Arabidopsis line demonstrated accumulation of dimers oxidized at the Cα position. This is consistent with LigD activity. In agreement with previously reported observations (Tsuji et al ., 2015 ), the detected compounds were mainly oxidized G units linked with other G units or with ferulic acid and their hexose and malate derivatives. Such phenylpropanoid dimers are classified as neolignans or lignan‐like and are thought to be involved in plant defence mechanisms rather than in the lignification process (Davin and Lewis, 2005 ). The cellular localization of this coupling reaction is not known. In lignin formation, oxidative coupling of monolignols takes place in the cell wall by radical reactions catalysed by the action of oxidases (peroxidases or laccases) (Boerjan et al ., 2003 ). The G(β‐O‐4)ferulic acid is unlikely to be formed in the cell wall as it would mean that such units would subsequently be incorporated into lignin, but no or very little cell wall‐bound ferulic acid can be detected in dicots (Vogel, 2008 ). It has recently been demonstrated that radical coupling may also proceed in the cytosol (Dima et al ., 2015 ; Niculaes et al ., 2014 ) and peroxidases are also found in the vacuole (Welinder et al ., 2002 ; Weng and Chapple, 2010 ). However, as glycosylation is a cytoplasmic process and the vacuole is a storage compartment for glycosylated monolignols (Dima et al ., 2015 ; Liu et al ., 2011 ), the radical coupling of monolignols to form neolignan‐like compounds most likely takes place in the cytosol before glycosylation and transport into the vacuole. This is corroborated by the observation that the phenolic alterations observed by Tsuji and co‐workers in both purely cytoplasmic and mixed cytoplasmic/apoplastic localized LigD‐expressing transformants are also neolignan‐like compounds (Tsuji et al ., 2015 ). Malate (trans‐) esterification occurs inside the vacuole (Hause et al ., 2002 ; Strack and Sharma, 1985 ). In this study, the LigDFG proteins were targeted to the secretory pathway using well‐proven signal peptides (Liu et al ., 2004 ; Rogers, 1985 ), although leakage or mis‐targeting cannot be fully excluded. The presence of oxidized neolignan‐like compounds such as Gox(β‐O‐4)FA in the transgenic amyDFG10 Arabidopsis line indicates that the LigD enzyme and G(β‐O‐4)ferulic acid have at some point been present in the same cellular compartment. This could either be caused by mis‐targeting of LigD to the cytoplasm in the amyDFG10 line, or by co‐occurrence of the LigD enzyme and its substrate G(β‐O‐4)ferulic acid as the protein travels through the secretory pathway to the apoplast. The former explanation is the most likely due to similarities in phenolic profiling with the cytoplasmically localized LigD (Tsuji et al ., 2015 ). Quantification of the LigD enzyme in soluble protein extracts from all three LigDFG ‐expressing lines using Western blot analysis (Figure 3 ) demonstrated that the expression level of LigD was strongly variable between replicate plants from the two homozygous transgenic lines ( amyDFG12 and ppiDFG1 ), but high and more stable in the transgenic amyDFG10 line, potentially due to the mixed localization of LigD in amyDFG10 lines. All three transgenic Arabidopsis lines possessed LigDFG activity in the in vitro assay, expressed the LigD enzyme to a level detectable by Western blot and displayed benzylic oxidation in β–O–4‐structures. However, two of the lines ( amyDFG12 and ppiDFG1 ) did not display a significant change in soluble phenolics, supporting the notion that the absolute levels of intracellular enzyme accumulation in these lines were negligible. Furthermore, the required cofactors [nicotinamide adenine dinucleotide (NAD + ) and glutathione (GSH)] are expected to be nonlimiting in the cytosol, meaning that if LigDFG had been present in this compartment in the amyDFG12 and ppiDFG1 lines, we would have detected changes in soluble phenolics. The fact that we did not strongly implies that the detected oxidation of G and S lignin units (Gox and Sox in the NMR) and the proposed cleavage of β–O–4‐bonds had indeed taken place in the apoplast. Two cofactors are required for the ability of LigDFG enzymes from S. paucimobilis to catalyse cleavage of the β‐aryl ether units in lignin. LigD activity is dependent on the presence of NAD + whereas LigF and LigG activity requires the presence of GSH. The concentration of these two cofactors is expected to be low in the apoplast although NAD + is detected in the apoplastic fluid and associated with peroxidase activity (Otter and Polle, 1997 ) and GSH is involved in detoxification processes and small amounts are found in the apoplast (Foyer et al ., 2001 ). Low concentrations of cofactors are therefore considered to be the limiting factor for LigDFG activity in the apoplast rather than the level of the LigDFG proteins. A completely separate factor that may also restrict the level of modifications introduced into lignin by the expression of LigDFG is that LigD and LigF show high substrate specificity with only one of four stereoisomers of G(β‐O‐4)G being accepted as substrate and cleaved. The results obtained from the characterization of the transgenic Arabidopsis plants expressing LigDFG provided additional knowledge on the substrate specificity of LigDFG. The 2D HSQC NMR analysis showed an increase in the amount of Sox units formed. In addition, S(β‐O‐4)G was converted into HPS by the LigDFG enzymes. The phenolic analysis also revealed formation of Gox(β‐O‐4)sinapic acid. This demonstrated that coniferyl alcohol dimers as well as sinapyl alcohol and sinapic acid ether dimers were recognized by the LigDFG enzymes. These in vivo activities match recently reported in vitro activities of LigD on S(β‐O‐4) units (Gall et al ., 2014b ; Tsuji et al ., 2015 ). The ability of the LigDFG enzymes to reductively cleavage not only G units expands the potential of using this approach to modify lignin structure in plants as this allows a greater variety of ether‐linked units to be cleaved. Expression of the individual LigD , LigF and LigG genes in transgenic Arabidopsis lines provided interesting insights into the ability of endogenous plant enzyme activities to metabolize intermediates in the LigD‐LigF‐LigG pathway. When protein extracts from LigF ‐expressing lines were incubated with MPHPV as substrate, HPV was formed in addition to the expected product GS‐HPV (Figure S2). In the bacterial pathway, the conversion of GS‐HPV to HPV is catalysed by LigG. We did not detect LigG activity in enzyme extracts from WT Arabidopsis leaves, even upon prolonged incubation, and based on this it can be suggested that the enzyme responsible for the conversion in the transgenic lines was LigF. However, the phenolic profiling demonstrated that HPV derivatives are indeed present in stems from WT plants, and re‐examination of the results from Tsuji et al . ( 2015 ) demonstrated that the amounts of these compounds were also slightly increased in plants expressing only LigD. These results suggest that Arabidopsis does harbour enzymes with LigF and LigG activities. The discrepancy in enzyme activity between leaves and stems can be explained either by differential expression of endogenous LigF and LigG‐like enzymes in different tissues or at different growth stages, or by the endogenous activities being collated into one enzyme with LigFG activity. LigG belongs to the omega class of GSTs (Meux et al ., 2012 ) that are characterized by catalysing redox reactions and reductive deglutathionylations instead of a classical glutathionylation reactions (Board et al ., 2008 ). A cysteine residue in the active site of the omega class of GSTs is essential for their different mode of action. In plants, GSTs of the lambda class are known to have cysteine in the active site as opposed to glutathionylating plant GSTs that harbour a serine in this position (Lallement et al ., 2014b ). The lambda class GSTs from wheat and poplar are able to reduce glutathionylquercetin to quercetin (Dixon and Edwards, 2010 ; Lallement et al ., 2014a ). A lambda class GST is therefore a good candidate for the LigG activity found in wild‐type Arabidopsis, but no plant enzymes are known which catalyse a LigF‐like reaction. In bacterial degradation of pentachlorophenols, a single enzyme is responsible for a presumably two‐step reaction similar to the combined LigFG‐reaction (Huang et al ., 2008 ). The reaction consumes two GSH‐molecules and, if the enzyme is damaged by oxidation at the catalytic Cys‐residue, it produces a GSH‐conjugate that it is unable to cleave. It is therefore possible that under certain conditions, LigF harbours additional LigG activity, and that Arabidopsis produces an enzyme with both activities. Further analysis is required in order to identify the Arabidopsis enzymes and compare their activity to LigF and LigG to elucidate whether the introduction of these enzymes into the apoplast in concert with LigD is sufficient to introduce changes into lignin structure. As endogenous lambda GSTs in plants are targeted to either the cytosol, chloroplast or peroxisome (Dixon et al ., 2009 ), such genetic engineering will still be required to obtain degradation of lignin. In conclusion, we have shown that expression of LigDFG encoding the β‐aryl ether‐unit‐degrading enzymes from S. paucimobilis, has an impact on lignin structure, thus providing a platform for saccharification yield improvement, a desired feature of plants for biofuel production. Even though there are limitations to the amount of change possible from this approach, we have shown that the changes observed are statistically significant and that it is possible to modify lignin structure, without significant alterations to neolignan‐like compounds, by targeting bacteria‐derived lignin‐degrading enzymes to the secretory pathway. Fine‐tuning of lignin cross‐linking may be combined with other modest modifications of lignin structure offering improved properties with respect to biomass conversion into biofuels without compromising plant robustness."
} | 5,092 |
21334448 | null | s2 | 1,889 | {
"abstract": "Spider silk has been evolutionarily optimized for contextual mechanical performance over the last 400 Ma. Despite precisely balanced mechanical properties, which have yet to be reproduced, the underlying molecular architecture of major ampullate spider silk can be simplified being viewed as a versatile block copolymer. Four primary amino acid motifs: polyalanine, (GA)(n), GPGXX, and GGX (X = G,A,S,Q,L,Y) will be considered in this study. Although synthetic mimetics of many of these amino acid motifs have been produced in several biological systems, the source of spider silk's mechanical integrity remains elusive. Mechanical robustness may be a product not only of the amino acid structure but also of the tertiary structure of the silk. Historically, solid state nuclear magnetic resonance (ssNMR) has been used to reveal the crystalline structure of the polyalanine motif; however, limitations in amino acid labeling techniques have obscured the structures of the GGX and GPGXX motifs thought to be responsible for the structural mobility of spider silk. We describe the use of metabolic pathways to label tyrosine for the first time as well as to improve the labeling efficiency of proline. These improved labeling techniques will allow the previously unknown tertiary structures of major ampullate silk to be probed."
} | 331 |
25586643 | null | s2 | 1,890 | {
"abstract": "Cells acclimate to fluctuating environments by utilizing sensory circuits. One common sensory pathway used by bacteria is two-component signaling (TCS), composed of an environmental sensor [the sensor kinase (SK)] and a cognate, intracellular effector [the response regulator (RR)]. The squid symbiont Vibrio fischeri uses an elaborate TCS phosphorelay containing a hybrid SK, RscS, and two RRs, SypE and SypG, to control biofilm formation and host colonization. Here, we found that another hybrid SK, SypF, was essential for biofilms by functioning downstream of RscS to directly control SypE and SypG. Surprisingly, although wild-type SypF functioned as an SK in vitro, this activity was dispensable for colonization. In fact, only a single non-enzymatic domain within SypF, the HPt domain, was critical in vivo. Remarkably, this domain within SypF interacted with RscS to permit a bypass of RscS's own HPt domain and SypF's enzymatic function. This represents the first in vivo example of a functional SK that exploits the enzymatic activity of another SK, an adaptation that demonstrates the elegant plasticity in the arrangement of TCS regulators."
} | 288 |
25323067 | PMC4200409 | pmc | 1,891 | {
"abstract": "An array of micron-sized setal hairs offers geckos a unique ability to walk on vertical surfaces using van der Waals interactions. Although many studies have focused on the role of surface morphology of the hairs, very little is known about the role of surface chemistry on wetting and adhesion. We expect that both surface chemistry and morphology are important, not only to achieve optimum dry adhesion but also for increased efficiency in self-cleaning of water and adhesion under wet conditions. Here, we used a plasma-based vapor deposition process to coat the hairy patterns on gecko toe pad sheds with polar and non-polar coatings without significantly perturbing the setal morphology. By a comparison of wetting across treatments, we show that the intrinsic surface of gecko setae has a water contact angle between 70–90°. As expected, under wet conditions, adhesion on a hydrophilic surface (glass) was lower than that on a hydrophobic surface (alkyl-silane monolayer on glass). Surprisingly under wet and dry conditions the adhesion was comparable on the hydrophobic surface, independent of the surface chemistry of the setal hairs. This work highlights the need to utilize morphology and surface chemistry in developing successful synthetic adhesives with desirable adhesion and self-cleaning properties.",
"discussion": "Results and Discussion To investigate the effect of surface chemistry on wetting and adhesion we first discuss the characterization of coated sheds using x-ray photoelectron spectroscopy (XPS) and scanning electron microscopy (SEM) measurements. We then discuss differences in wetting as a function of surface chemistry. Finally, we discuss how shear adhesion in air and water varies as a function of setal surface chemistry and substrate wettability. Surface morphology and chemical analysis Figure 1 shows SEM images of a untreated shed (B-S, Figures 1a and 1b ) and the PECVD coated sheds, using two precursors, maleic anhydride (M-S, Figures 1c and 1d ) and 1H, 1H, 2H-perfluoro-1-dodecene (F-S, Figures 1e and 1f ) along with a oxygen plasma treated sample (P-S, Figures 1g and 1h ). The micrometer-scale tetrad pattern common to Tokay geckos ( Gekko gecko ) can be clearly seen, as can the hierarchal setal branches, terminating into hundreds of finer flattened pads (spatulae) on the untreated and PECVD coated sheds ( Figure 1 ). Visual similarity in these SEMs suggest that PECVD did not result in changes in morphology. Interestingly however, in the P-S samples we see a unique clumping behavior at the terminal branches of the setae ( Figure 1h ). We believe this could be due to removal of the lipids and an increase in adhesion energy between setae leading to bunching. The role of adhesion in stabilizing the clumping of fibrillar structure has been studied before and supports our observations of the P-S samples 25 26 . To investigate the effect of PECVD and plasma treatment on the samples we used surface sensitive XPS to measure the elemental composition of the toe pad shed surface (the probe depth of XPS is around 10 nm). The survey spectrum of the B-S sample (untreated shed) shows the presence of C1s, N1s, O1s and S2p peaks, representing the primary elements: carbon, nitrogen, oxygen and sulfur ( Figure 2 ). While this is not surprising given the proposed constituents of the setae ( β -keratin and lipid) 8 10 , this scan is the first report of the surface elements involved in the adhesion of the gecko adhesive system. Conversely, when compared to the B-S scan, the survey scan of the P-S sample (oxygen plasma-treated shed) shows an increased relative atomic composition of N1s, implying a change has occurred in the atomic composition of the surface after plasma treatment. The increased atomic composition of N1s in the plasma treatment (P-S) suggests that a lipid layer was removed, similar to the removal of lipids on the surface of wool 23 24 , except in this case the increased presence of β -keratin at the surface is likely causing the elevated nitrogen signal. It is also interesting to note that the peaks in the C1s spectrum of B-S and P-S samples are at the same positions, however, their relative percentages are significantly different. This provides further evidence that surface lipids are removed upon plasma treatment, at least in fractions if not completely, since we would expect different amounts of carbon in the lipid and keratin components of the setae. It is also important to note that the surfaces of the setae could be partially covered with keratin and the surface structure of keratin may also be affected by oxygen plasma. In the PECVD coated samples, the absence of N1s or S2p (nitrogen and sulfur) peaks found in the B-S sample implies that the coating was successfully deposited on the sample and allows us to assume that the coating thickness is around 10–15 nm, comparable to the analysis depth of XPS ( Figure S1 ). The M-S sample shows the presence of C1s and O1s peaks (carbon and oxygen) whereas, the F-S sample shows C1s and F1s peaks (carbon and fluorine). These elements are consistent with the chemical structure of the PECVD coatings and the atomic compositions of control measurements. The relative atomic compositions calculated using survey scans and also the results of the high resolution XPS scans are provided in the supporting information ( SI Text, section S1, Table S1, Figure S2 and Table S2 ). In summary, using XPS and SEM we confirmed that either a thin layer of PECVD coating was deposited on the structured surface of the gecko toe sheds or the surface was at least partially stripped of lipids, such that all four treatments have distinctly different surface chemistry yet similar morphology. Surface wetting and intrinsic surface chemistry modeling To study the wettability and thus the ability of gecko toes to self-clean water from their adhesive toe pads, we first measured the water contact angles of the B-S, P-S, M-S and F-S shed surfaces ( Table 1 ). Both the B-S and F-S surfaces show characteristics of a typical superhydrophobic surface, such as a contact angle of 150° ( Figure 3 ) and a very low contact angle hysteresis. The water droplet in contact with both of these surfaces corresponds to a Cassie-Baxter state. In contrast, the water droplet on the P-S and M-S surfaces almost instantaneously spreads on the surface, resulting in complete wetting ( Figure 3 ). The color of the setae on these surfaces also changes from a shiny white to grey once penetrated by water, which is a common indication of surface wetting in gecko toes 16 . Because these samples have similar morphology, the differences in surface wettability can be attributed to surface chemistry. Due to the limitations of SEM however, it is possible that there are morphological differences across samples at the nanometer-level scale which cannot be observed visually. We do not believe this has a significant effect on wetting behavior however because the two PECVD coated samples (M-S and F-S) have such extreme differences in wetting behavior despite likely having similar roughness changes related to the coating process. To test if the superhydrophobicity of the B-S surface is metastable we measured the B-S surface along with the M-S, F-S and P-S surfaces under water condensation, which is expected to induce a wetting transition 27 28 . The samples were kept under 100% humidity for 3–4 days, which was then followed by the measurement of water droplet contact angle. We found, as expected, that both the M-S and P-S samples wetted completely under condensation. Similarly, as expected, the F-S surface retained its superhydrophobic characteristics and continued to remain dry, even after prolonged exposure to high humidity. Conversely, we found that in the B-S surface water began to penetrate inside the surface roughness and resulted in the wetting of the surface, unlike water droplet contact angle measurements made in ambient conditions. Thus the B-S surface changed from a Cassie-Baxter wetting state to a Wenzel wetting state. When comparing the B-S surface (147°) to the known inherent chemistries of the M-S (complete wetting) and F-S (149°) surfaces we find that the B-S surface is inherently hydrophobic, enough to form a stable superhydrophobic Cassie-Baxter state, like the F-S surface. However it appears that this state is only a metastable state and a transition to the Wenzel state, similar to the M-S or P-S surface, occurs after water condensation on the surface. It is important to note here that the wetting transition is completely reversible and the B-S samples regain their superhydrophobicity after drying. Interestingly the B-S sample is different than the P-S sample, which does not have a layer of coating deposited but rather has been stripped of surface layers (most likely hydrophobic lipids). We hypothesize that after oxygen plasma treatment a hydrophilic surface has been exposed (likely β -keratin) which results in the complete wetting of the surface, similar to the hydrophilic M-S surfaces. Since superhydrophobicity is a result of both surface roughness and intrinsic surface wettability, we carried out apparent contact angle calculations using a model unit cell shown as an inset in Figure 4b ( SI Text, section 2 ). The hierarchical unit cell consists of four setae (tetrad pattern), each with a square cross section. The square pillar is 60 μ m tall and 4 μ m wide. The top face of each pillar consists of a number of cubes, 0.2 μ m in size. The roughness parameters were calculated for the unit cell and incorporated into the Cassie-Baxter and Wenzel equations to predict the apparent contact angles ( θ CB and θ W ) and to calculate G* CB and G* W , the thermodynamic free energy corresponding to Cassie-Baxter and Wenzel states, respectively 29 30 31 . The results show that the range of 70° ≤ θ Y ≤ 120° resulted in θ CB of about 145°–160° ( Table S3 ), matching the range of values measured experimentally on the superhydrophobic B-S and F-S samples. This suggests that the intrinsic wettability of gecko setae can fall somewhere within this range. Figure 4a shows ΔG* calculated as a function of roughness, R, for different intrinsic wettability θ Y . Figure 4b shows the expanded region near ΔG* = 0. Although the range of R where contact angles could be predicted is very small compared to the actual roughness of the gecko setal surface, it qualitatively predicts wettability as a function of θ Y . For θ Y ≤ 90°, ΔG* is an increasing function of R, implying that a Wenzel state is thermodynamically more favorable. On the other hand, for θ Y ≥ 90°, ΔG* is a decreasing function of R, becoming negative at a critical R, implying that the Cassie-Baxter state is thermodynamically more favorable for a roughness value more than the critical R. We find that the model predictions are consistent with our experimental observations, where M-S surfaces ( θ Y ≈ 48°) completely wet with water (Wenzel wetting state), whereas the F-S surface ( θ Y ≈ 110°) is superhydrophobic (Cassie-Baxter state). Although the θ Y for the B-S surface is not known, our model predictions and experimental observations allow us to narrow down a range for θ Y for the B-S surface to be between 70°–90°. In this range a stable Cassie-Baxter wetting is possible but also is not the most favorable state thermodynamically, which is in agreement with our observations of B-S surface wetting after exposure to water vapor condensation. It is interesting that this range encompasses the θ Y ≈ 90° value which was previously suggested to be the intrinsic surface chemistry of the gecko setae based on the contact angle measurement of the smooth gecko eye scale 14 and that it is neither strongly hydrophobic nor is it strongly hydrophilic. Despite this consistency, it is important to be cautious of our predictions, as this predicted range (70°–90°) does not take into account any changes in surface chemistry which may cause the structure to behave more hydrophobic or hydrophilic when in contact or exposed to water, a behavior mounting evidence suggests may occur (Pesika et al 17 and Hsu et al 10 ). To quantify the transition barrier from the Cassie-Baxter to Wenzel wetting states we used a water column to measure the hydrostatic pressure necessary to induce a wetting transition in all four surface treatments. We found that the P-S and M-S sample wetted as soon as they were introduced into the water column, as expected by static water contact angle measurements. Conversely, both the B-S and F-S samples appeared shiny and silvery, implying the presence of an air plastron layer. Both surfaces continued to retain the plastron layer as the immersion depth increased up to a maximum of 4′. The plastron layer did not disappear in either sample, even after 7–8 hours. Thus the transition barrier between the Cassie-Baxter and Wenzel states in both the B-S and F-S samples can be estimated to be greater than 11.95 kPa, almost three orders of magnitude higher than the M-S and P-S samples. Although we anticipate the barrier for the F-S surface to be higher than the B-S surface based on model predictions and the condensation results, it is surprising how resilient the native sample (B-S) is when protected by an air plastron. This has clear implications for the whole animal, as we found previously that the maintenance of an air plastron likely allows geckos to remain adhesive on hydrophobic surfaces 22 32 . As such we can better appreciate the various scenarios that must occur for natural toe pads to become wet, which seem to be when applied repeatedly on wet surfaces or when agitated on rough, wet surfaces 16 . During these instances it is likely that the setal mat first must be penetrated by water and the plastron broken in order to become wet. Similarly in humidity is it likely small droplets of water on the setae cause the plastron to fail to form and allows water to be pulled into the setal mat and wet. Regardless, it is clear that the different transition barriers and wetting behaviors in each of the surface treatments can be attributed to differences in inherent surface chemistry and thus their ability to self-clean water from the adhesive toe pad. Dry and wet adhesion To study the effect that surface chemistry has on adhesion we tested the shear adhesion of natural, uncoated toe sheds (B-S) and modified sheds (P-S, M-S, and F-S) on a hydrophilic glass surface and a hydrophobic OTS-SAM coated glass surface in air (W dry ) and water (W wet ). We have measured the forces parallel to the surface, similar to friction measurements, and we refer to these forces as shear adhesion because we expect the origin of this force to be due to adhesion rather than friction. The effect of coating (B-S, M-S, F-S, P-S), test substrate (glass or OTS-SAM coated glass) and treatment conditions (air or water) was highly significant (df = 16, F = 22.7982, p ≤ 0.0001*) as were many of the interactions, including the three-way interaction of coating*surface*treatment ( SI Text, section 3, Table S4 ). Subsequent two-way ANCOVAs, one for each surface demonstrates that the three-way interaction is driven by a strong negative effect of water on adhesion for all four coatings on glass, but the absence of such an effect when adhesion is tested on OTS-SAM coated glass. Specifically, there was a significant drop in force when samples were tested underwater on the hydrophilic glass substrate compared to tests performed in air ( Figure 5a ). However, this did not occur on the hydrophobic OTS-SAM coated glass ( Figure 5b ). Overall force values from samples tested on the OTS-SAM substrate in air were lower than those tested on the glass substrate in air. The ratio of Hamaker constants between OTS-SAM coated glass and glass is 0.87 ( SI Text, section 3, Table S5 ) 33 , and we expect the shear adhesion to be lower for OTS-coated glass. However, the measured ratio is much smaller (around 0.14). This ratio is also lower than that measured for adhesion of Tokay geckos on glass and OTS-SAM coated glass 16 . The differences could be due to much cleaner glass used in the shed experiments or due to the differences between the animal and shed experiments. In summary, adhesion appears to be highly sensitive to substrate wettability, either reducing adhesion on the hydrophilic glass substrate or maintaining adhesion on the hydrophobic OTS-SAM coated glass. Surprisingly, the surface chemistry of the setae does not have an affect on adhesion across all substrates and treatments, suggesting that setal surface chemistry is not critical to successful adhesion in air or water. To investigate the experimentally measured results, we calculated the W wet /W dry ratio for B-S, P-S, M-S and F-S shed samples in contact with glass or OTS-SAM coated glass and modeled the predicted W wet /W dry ratios of shed samples in contact with the coated glass surface ( SI Text, section 3.3, Table S6 and Table S7 ) using models derived elsewhere 22 . Model calculations for the glass surface were not possible in this context. Experimental ratios on glass clearly show that dry adhesion is favored (the ratio is below 1), whereas on coated glass adhesion in dry and wet conditions are virtually the same (the ratio is near 1) in all sample treatments. In contrast, model calculations predict that only the B-S and the F-S samples should be lower in air compared to water on the coated surface and the hydrophilic samples (P-S and M-S) should have much weaker adhesion in water than air. This is clearly inconsistent with our experimental results. During tests with the B-S and F-S samples in water, we could clearly see that a layer of air was always present at the shed surface, consistent with our hydrostatic pressure experiments. When testing on a hydrophilic glass surface, water formed a lubricating layer between the air plastron and the glass slide, preventing the hydrophobic sheds from contacting the surface and thus resulting in much lower adhesion (ratios are below 1). This is consistent with whole animal adhesion to wet glass 16 22 . Likewise, when testing on the hydrophobic OTS-SAM coated glass, the air plastron on the B-S and F-S samples allowed for dry contact to occur, similar to whole-animal experiments which show a dry region of the toe is maintained when in contact with a hydrophobic surface 22 . What was intriguing however was the adhesion results of the hydrophilic P-S and M-S samples on the coated surface. As expected, the M-S and P-S samples wetted immediately underwater. However, we found that wetting only affected adhesion when tested on the hydrophilic glass surface, not the hydrophobic OTS-SAM coated glass. Our results suggest that wetting due to inherent surface chemistry does not affect adhesion in wet conditions, and rather substrate wettability is the driving factor in wet conditions. This is clearly contradictory to what we expect from theoretical models for adhesion in wet environment and observations of geckos sliding down surfaces when their toes become wet 16 . Several factors may explain differences between the experimental and theoretical results. Firstly, small differences in mechanical properties, surface roughness and clumping in the P-S samples may have a significant effect on adhesion between uncoated and coated samples. In wool fibers plasma treatment may increase roughness by about 5 nm 34 , and in a PECVD coating similar to that used in this study a Wenzel value of 1.19 (ratio of actual area divided by projected area) was found on flat surfaces after PECVD treatment 31 . Secondly, model predictions are based on normal adhesion between the shed surface and substrate, however experimental measurements use shear geometry. Even though there is a correlation between normal and shear adhesion 35 , in the case of M-S and P-S which wet almost instantaneously in water, it is possible that water drains out of micro and nano channels of the shed to establish temporary dry contact with the surface as it is sheared over the shed 36 37 , a behavior which is not accounted for in the model. Additionally, it is possible that coating the setae with either polymer coating or by plasma treating the setae alters the mechanical properties of the setal array. However, it is unlikely that the large difference in shear adhesion force between the glass and the OTS-SAM coated glass is due to changes in the modulus of the setal array alone. Instead we expect changes in modulus to affect the shear adhesion in wet and dry conditions, where plasma treated setae or setae that are coated with a hydrophilic polymer may become softer in water, similar to results found when measuring setal modulus in highly humid conditions (>80% RH) 38 . Finally, measurements of shear adhesion in shed samples suggest that maximum force (F max ) is reached by significant differences in the force profiles. F max was reached either just before the sliding started, followed by a decrease in the force as the sliding continued, a case we refer to as stiction; or F max was reached after sliding started, in which case the force either plateaued at the maximum value or continued to increase until it reached F max , a case we refer to as friction ( Figure 6 ). When we compared stiction and friction responses in air and water we found that P-S and M-S tested on OTS-SAM coated glass were significantly different, and the B-S samples tested on OTS-SAM coated glass were nearly significantly different. In fact, when comparing instances of friction to stiction it is very clear that friction dominates in all surface chemistries tested on OTS-SAM coated glass in air ( Table S8 ). When tested in water however, there is no clear sample behavior on OTS-SAM coated glass ( Table S8 ). Conversely on glass we only found that adhesion behavior in air and water was significantly different for B-S samples where stiction was the dominant behavior in air. Interestingly it appears friction may be more dominant in water for the B-S samples tested on glass. The association of either stiction or friction with specific treatment groups was very intriguing and clearly suggests an influence of shed surface chemistry on adhesion behavior and performance. Summary Tokay geckos ( Gekko gecko ) are native to wet tropical regions of South East Asia and because of this we expect their adhesive system to remain functional at high humidity and in the presence of surface water. We find that toe pad sheds coated with a nanometer-thick hydrophilic coating wet immediately, creating a super wetting state. In contrast, sheds with a hydrophobic fluorinated coating are superhydrophobic, even after prolonged exposure to humidity. From these results we infer that the intrinsic surface chemistry of the gecko setae is more similar to hydrophobic surfaces with high water contact angles (150°). Nevertheless, after prolonged exposure to humidity, untreated samples wet similarly to hydrophilic-coated samples, although they do recover their superhydrophobic state upon drying. This suggests that the water contact angle for intrinsic surface chemistry of gecko setae is between 70–90°, resulting in a metastable Cassie-Baxter superhydrophobic state that changed to the thermodynamically more stable Wenzel state upon exposure to humid air. However, we cannot exclude the possibility that the wetting behavior results from an intrinsic surface chemistry of greater than 90° that restructures to a more hydrophilic surface after prolonged exposure to water or humid air. We also found that adhesion is significantly impacted by the surface chemistry of the substrate. As expected, adhesion in wet conditions on hydrophilic glass is much lower than dry adhesion. Surprisingly, adhesion in wet conditions for hydrophilic-coated sheds on hydrophobic OTS-SAM coated glass was almost the same as dry adhesion. This suggests that the hydrophilic structured surfaces (toe pad sheds) do not show the same trends as the hydrophilic flat surfaces (substrates where adhesion is made). Our results raise several important questions about the characteristics of the gecko adhesive system. Gecko adhesion is very sensitive to the presence of water on hydrophilic surfaces but not hydrophobic surfaces. If the latter are better reflective of the surface chemistry of natural surfaces, this could help explain the abundance and success of geckos in wet tropical habitats. The hydrophobicity of gecko toe pads arises from roughness (fine hair-like structures) as well as surface chemistry (likely lipids), and the non-wetting (Cassie-Baxter) state of the toe pads should enhance the ability of the toe pads to exclude water at the contact interface on hydrophobic surfaces. Interestingly, our F-S coated toe pad sheds remained non-wettable under conditions that led to the wetting of the native toe pad sheds, raising the question of whether the critical pressure to prevent transition from the Cassie to Wenzel state in the natural toe pad is 1) not attainable by the natural system, 2) trades-off with some other performance characteristic (e.g., self cleaning, resiliency, etc.), or 3) does not translate into an adhesion effect that affects organismal fitness. Until the material components of gecko setae are better understood, we will not know the answers to these and other questions relevant to a deeper understanding of the ecology and evolution of gecko adhesion and the current limitations in synthetic mimics. We hope our findings motivate more research into the role of chemistry and the nature of the materials making up adhesive gecko toe pads."
} | 6,414 |
37713483 | PMC10881039 | pmc | 1,892 | {
"abstract": "As the most promising candidates for the implementation of in-sensor computing, retinomorphic vision sensors can constitute built-in neural networks and directly implement multiply-and-accumulation operations using responsivities as the weights. However, existing retinomorphic vision sensors mainly use a sustained gate bias to maintain the responsivity due to its volatile nature. Here, we propose an ion-induced localized-field strategy to develop retinomorphic vision sensors with nonvolatile tunable responsivity in both positive and negative regimes and construct a broadband and reconfigurable sensory network with locally stored weights to implement in-sensor convolutional processing in spectral range of 400 to 1800 nanometers. In addition to in-sensor computing, this retinomorphic device can implement in-memory computing benefiting from the nonvolatile tunable conductance, and a complete neuromorphic visual system involving front-end in-sensor computing and back-end in-memory computing architectures has been constructed, executing supervised and unsupervised learning tasks as demonstrations. This work paves the way for the development of high-speed and low-power neuromorphic machine vision for time-critical and data-intensive applications.",
"introduction": "INTRODUCTION Machine vision plays a crucial role in time-critical applications, which require real-time object recognition and classification, such as autonomous driving and robotics ( 1 ). With the increase of frame rates and pixel densities of sensors, large quantities of raw and unstructured data are generated in sensory terminals ( 2 , 3 ), and image processing becomes a data-intensive task, which requires high-efficiency and low-power image processing directly at the sensory terminals ( 2 , 4 , 5 ). The emerging in-sensor computing that can perform low-level and high-level image processing tasks in sensory networks has attracted numerous attentions in recent years, and various neuromorphic devices and structures including optoelectronic synapses ( 6 – 11 ) and one-photosensor–one-memristor arrays ( 3 , 12 ) have been developed to implement in-sensor computing. In particular, human retina–inspired retinomorphic vision sensors have demonstrated their great potential in in-sensor computing because they can constitute built-in artificial neural networks (ANNs) and implement multiply-and-accumulation (MAC) operations using tunable responsivities as the weights of ANNs ( 1 , 13 – 18 ). However, these retinomorphic devices still suffer from the volatile nature of their responsivities, which require a sustained gate bias to maintain ( 1 , 13 , 14 , 19 , 20 ). That is, the weights of ANNs need to be stored remotely and supplied to each vision sensor via complex external circuits, which would inevitably consume additional energies. More seriously, extremely complex circuits are required for the individual control of each unit in networks, which is almost impossible and unacceptable for large-scale sensory networks considering the limited power and resources available at the edge. In addition, the limited operating spectra of these retinomorphic devices, mainly in ultraviolet (UV) and visible (vis) range ( 1 , 13 , 15 , 16 , 19 , 20 ), hinder the implementation of in-sensor computing in more important infrared band. It is highly desirable to develop broadband retinomorphic vision sensors with nonvolatile tunable responsivities to constitute sensory networks for neuromorphic machine vision. Here, we propose an ion-induced localized-field modulation strategy to realize the nonvolatile modulation of responsivity and develop a broadband retinomorphic vision sensor with operating spectra covering 400 to 1800 nm based on core-sheath single-walled carbon nanotube@graphdiyne (SWNT@GDY). The responsivity can be linearly modulated in both positive and negative regimes by controlling the localized field induced by the trapped Li + ions in GDY, enabling the implementation of in-sensor MAC operation and convolutional processing. A 3 × 3 × 3 sensor array is fabricated to constitute a reconfigurable convolutional neural network (CNN) involving three kernels (3 × 3), and low-level and high-level processing tasks for multiband images, including edge detection and sharpness of hyperspectral images and classification of colored letters, are performed as demonstrations. The conductance of this retinomorphic device can also be linearly modulated, and, thus, the device array can also implement in-memory computing tasks. A complete hardware emulation of human visual system including retina and brain visual cortex is achieved by networking the front-end in-sensor CNN and back-end in-memory ANN based on the same device structure, and unsupervised learning (autoencoder) is demonstrated. The nonvolatile nature of responsivity induced by the localized field enables the local storage of weights in sensory networks, which is essential for high-efficiency low-power in-sensor computing at edge.",
"discussion": "DISCUSSION Table S1 compares the parameters of the existing retinomorphic vision sensors with our device. The most important advantage of our device is that it can realize the nonvolatile modulation of the responsivity in a quite wide spectral range, which enables the implementation of in-sensor computing for broadband images. Of course, this device still suffers from some limitations for practical applications. (i) A drain bias is still required during the read process, inducing a high dark current and consuming more energy. (ii) The response speed is limited to milliseconds due to the charge trapping/detrapping mechanism, restricting its potential for high-speed image processing. (iii) Although it can operate in a wide spectral range, the discrimination of colors (wavelengths) is still a great challenge, and the wavelength-dependent photoresponse will affect the processing accuracy of the sensory network for color images. (iv) The scale of the sensory network is still limited, which is difficult in meeting the requirements of machine vision for millions of pixels and tens of categories in real scenarios. In the future, we will pay more efforts to solve these challenges. Fortunately, the photovoltaic effect provides a feasible approach to realize high-speed photoresponse in nanosecond timescale ( 1 ), and no drain bias is required for devices based on the photovoltaic effect. The RC delay, however, should also be fully considered for high-speed devices because the parasitic capacitance, together with the large resistance of the device, would induce a nonnegligible RC time constant, which limits the further improvement of the processing speed. Mimicking the various human cone cells (short-wave, middle-wave, and long-wave cones) that respond to specific wavelengths ( 38 ) might provide a possible approach to realize color discrimination. In addition, the compatibility of the device with modern complementary metal-oxide semiconductor (CMOS) technology is crucial for the large-scale fabrication of the sensory network. The polymer electrolyte used in this work is considered to be the main obstacle for CMOS process due to its poor thermodynamical stability and processability. Nevertheless, inorganic solid electrolyte (e.g., Li 3 PO 4 and Li x SiO 2 ) provides a promising candidate to replace the polymer electrolyte ( 39 ), which can improve the CMOS compatibility of our device. High-speed, low-power, and large-scale retinomorphic vision sensors with nonvolatile modulation of responsivity and color discrimination ability should be further developed in the future. In conclusion, we have proposed a strategy that uses localized field induced by captured Li + ions instead of the externally controlled gate bias to realize nonvolatile modulation of responsivity. Moreover, correspondingly, retinomorphic vision sensors based on specially designed core-sheath SWNT@GDY have been developed to constitute built-in neural networks with locally stored weights, which can implement in-sensor convolutional processing in a wide spectral range covering 400 to 1800 nm. Closely analogous to human visual system, neuromorphic machine vision involving front-end in-sensor CNN and back-end in-memory ANN is constructed to implement supervised and unsupervised learning tasks. It is a great breakthrough for the development of in-sensor neural networks with locally stored responsivities, providing a feasible strategy to develop high-speed and low-power neuromorphic machine vision for real-time object recognition and classification."
} | 2,140 |
32219464 | PMC7190589 | pmc | 1,893 | {
"abstract": "Substrates with high sulfate levels pose problems for biogas production as they allow sulfate reducing bacteria to compete with syntrophic and methanogenic members of the community. In addition, the end product of sulfate reduction, hydrogen sulfide, is toxic and corrosive. Here we show how sulfate addition affects physiological processes in a thermophilic methanogenic system by analyzing the carbon flow and the microbial community with quantitative PCR and amplicon sequencing of the 16s rRNA gene. A sulfate addition of 0.5 to 3 g/L caused a decline in methane production by 73–92%, while higher sulfate concentrations had no additional inhibitory effect. Generally, sulfate addition induced a shift in the composition of the microbial community towards a higher dominance of Firmicutes and decreasing abundances of Bacteroidetes and Euryarchaeota . The abundance of methanogens (e.g., Methanoculleus and Methanosarcina ) was reduced, while sulfate reducing bacteria (especially Candidatus Desulforudis and Desulfotomaculum ) increased significantly in presence of sulfate. The sulfate addition had a significant impact on the carbon flow within the system, shifting the end product from methane and carbon dioxide to acetate and carbon dioxide. Interestingly, methane production quickly resumed, when sulfate was no longer present in the system. Despite the strong impact of sulfate addition on the carbon flow and the microbial community structure during thermophilic biogas production, short-term process disturbances caused by unexpected introduction of sulfate may be overcome due to the high resilience of the engaged microorganisms.",
"introduction": "Introduction The anaerobic digestion of organic waste of changing composition and quantity for biogas production involves the risk of introducing undesirable substances to the system, which endanger optimal process performance (Illmer and Gstraunthaler 2009 ; Wagner et al. 2014 ). One of these substances is sulfate (SO 4 2− ), which can be introduced in the reactor when digesting wastes from the food industry, particularly from the production of alcohol, yeast, citric acid and edible oils, and the paper industry (Colleran et al. 1995 ). The input of high levels of SO 4 2− into biogas fermenters results in lower methane (CH 4 ) production rates and the evolution of hydrogen sulfide (H 2 S), which causes odors and corrosion. In the presence of SO 4 2− as a terminal electron acceptor, SO 4 2− reducing bacteria (SRB) compete with syntrophic bacteria and methanogens for their common substrates (lactate, acetate, propionate, butyrate, and H 2 ) and thereby produce cytotoxic H 2 S (Muyzer and Stams 2008 ). Data on inhibiting SO 4 2− levels vary largely as the effect of SO 4 2− in an anaerobic digestion process depends on various factors, such as the type of reactor, the operation temperature, the used substrates, the pH, and the native degrading community (Colleran and Pender 2002 ; Chen et al. 2008 ). Nevertheless, it is generally agreed that the ratio of available carbon to SO 4 2− is more decisive than the SO 4 2− concentration itself and that at a chemical oxygen demand (COD) to SO 4 2− ratio of 10 or larger methanogenesis is not inhibited, while below this value both, successful and unsuccessful anaerobic digestion may occur (Hulshoff Pol et al. 1998 ). SRB affect the anaerobic community at multiple degradation levels, although it is assumed that SRB cannot compete with fast-growing acidogenic, fermenting organisms (Chen et al. 2008 ). From a thermodynamic point of view, SRB should out-compete acetogenic and methanogenic organisms for hydrogen (H 2 ), acetate, propionate, and butyrate (Stams et al. 2005 ). In reality, however, the dominance of SRB is, depending on the substrate, less distinct, as acetogens or methanogens may be advantageous concerning their pH and temperature optimum, sensitivity to H 2 S toxicity, initial abundance, or growth rate (Colleran and Pender 2002 ; Paulo et al. 2015 ; Chen et al. 2008 ). Therefore, experimental data on the competition of SRB with acetogenic and methanogenic organisms are often contradictory (Chen et al. 2008 ). Nevertheless, it was mostly found that SRB outcompete their opponents for H 2 and propionate, while the outcome of the competition for acetate and butyrate is rather unclear (Colleran et al. 1995 ; Paulo et al. 2015 ; Chen et al. 2008 ). Under SO 4 2− limiting conditions, SRB can grow as acetogens, meaning that they may be abundant in biogas reactors even after long periods of SO 4 2− absence (Visser et al. 1993 ). The toxicity of H 2 S is dependent on the pH, as unionized H 2 S can diffuse through the cell membrane and cause damage to proteins, coenzymes and can interfere with the assimilatory sulfur metabolism (Chen et al. 2008 ). The susceptibility to sulfide inhibition differs between the trophic groups and can be sorted as follows: hydrogenotrophic methanogens < hydrogenotrophic SRB < propionate oxidizing SRB < acetotrophic methanogens < syntrophic propionate oxidizing bacteria < acetate oxidizing SRB (Maillacheruvu and Parkin 1996 ). The inhibitory levels for dissolved sulfides in the literature range from 100 to 800 mg/L (Chen et al. 2008 ). Most research on SO 4 2− in anaerobic digestion is descriptively investigating real-life substrates at mesophilic conditions. The present study combines a defined substrate with an undefined mixed methanogenic community at thermophilic conditions to investigate competition-related and inhibitory effects of SO 4 2− addition to a reproducible system. In a first step, we examined the effects of various SO 4 2− levels on the applied methanogenic system regarding CH 4 production, SO 4 2− reduction, microbial community composition, and abundance of the relevant physiological groups. In a second step, we focused on SO 4 2− induced shifts in carbon flow over a period of 4 weeks of anaerobic digestion.",
"discussion": "Discussion The addition of SO 4 2− to the system affected the anaerobic digestion process on multiple levels, including gas production, VFA concentrations, the microbial community structure, and the absolute abundance of SRB and methanogens. The threshold for maximal inhibition of methanogenesis was 3 g SO 4 2− /L, which corresponds to a COD/SO 4 2− ratio of 5. Data on inhibitory SO 4 2− levels at thermophilic conditions are less available than at mesophilic conditions, but existing values are in a similar range as in our study. Siles et al. ( 2010 ) observed a total inhibition of CH 4 production at 1.8 g SO 4 2− /L when fermenting glucose (6 COD g/L), McFarland and Jewell (McFarland and Jewell 1990 ) found that CH 4 production decreased to 30% of the maximum at 0.8 g SO 4 2− /L when fermenting sucrose (10 g COD/L), and the CH 4 content in biogas from the digestion of sugar beet molasses (12 g COD/L) was reduced by half at a SO 4 2− concentration of 3 g/L (Colleran and Pender 2002 ). In general, the inhibitory effect of SO 4 2− addition could be caused by both toxic and competitive effects. The competition for carbon sources and electron donors between microorganisms is determined by Gibbs free energy, substrate affinities, cell numbers and growth rates, and the availability of electron acceptors (Lovley et al. 1982 ; Stams et al. 2005 ; Elferink et al. 1998 ; Visser et al. 1993 ). In the present investigation, SO 4 2− was available in excess, at least in the samples containing 3 and 5 g SO 4 2− /L, in which it was not depleted until the end of the incubation. Initial abundances favored methanogens over SRB, as the inoculum derived from a well-performing biogas reactor with low SO 4 2− levels, which was reflected by dsrA copy numbers below detection limit and a low relative abundance of SRB found during NGS analysis. In the early phase of incubation, the main inhibitory mechanism of SO 4 2− addition was probably the successful competition of SRB for H 2 , explaining the H 2 levels below detection limit. H 2 consumption combined with SO 4 2− reduction yields more free energy than hydrogenotrophic methanogenesis, and SRB have a lower threshold for H 2 uptake than methanogens, being 0.002 and 0.011 mbar, respectively (Lovley et al. 1982 ). Therefore, SRB outcompeted methanogens by decreasing the H 2 partial pressure below the level necessary for successful methanogenesis, making thermodynamics irrelevant for the outcome of the competition as long as SO 4 2− was not limited (Lovley et al. 1982 ). Subsequently, during SO 4 2− reduction, toxic levels of H 2 S were reached (up to 6%), which enhanced the inhibition of CH 4 production until the end of the incubation. In this context, the pH in the present study was 7.5 ± 0.1 at day 28, meaning that approximately 20% of the total dissolved sulfide should have been present as free H 2 S (Koster et al. 1986 ). This two-step inhibition mechanism was also suggested before (Karhadkar et al. 1987 ; McFarland and Jewell 1990 ). The composition of the microbial community on phylum level was similar to results of other investigations of thermophilic methanogenic microbiomes. Considering all samples from day 28 regardless of the SO 4 2− concentration, Firmicutes was the dominant prokaryotic taxon (45–84%), with comparable relative abundances also found by (De Vrieze et al. 2018 ; Liang et al. 2018 ; Lin et al. 2017 ; Maus et al. 2016 ; Tuan et al. 2014 ), ranging from 37 to 96%. Also, Thermotogae , Bacteroidetes , Tenericutes , Synergistes , and Chloroflexi were found, which are common members of a thermophilic CH 4 producing community (De Vrieze et al. 2018 ; Liang et al. 2018 ; Lin et al. 2017 ; Maus et al. 2016 ; Tuan et al. 2014 ), with Bacteroidetes being exceptional abundant in our control samples. Generally, microorganisms can be affected by increased SO 4 2− concentrations on three levels: (i) they are able to directly use SO 4 2− as terminal electron acceptor, (ii) they are positively or negatively influenced by the change in abiotic conditions (including pH and substrate availability) caused by ongoing SO 4 2− reduction, and (iii) they are inhibited by H 2 S or they benefit from the inhibition of competing microorganisms. To our knowledge, no NGS data on SO 4 2− -influenced communities under thermophilic conditions has been available until now. Under mesophilic conditions, the dominant SRB were diverse, with Desulfomicrobium , Desulfobulbus , and Desulfovibrio (all Proteobacteria ) when digesting maize silage and pig manure (Kushkevych et al. 2017 ); Desulfosporosinus ( Firmicutes ) when digesting manure (St-Pierre and Wright 2017 ); and Desulfonauticus ( Proteobacteria ) when digesting synthetic wastewater containing sucrose and ethanol under a high phenol load. In the present investigation, the candidate genus Desulforudis ( Firmicutes ) was most abundant, with up to 2% in samples with 3 or 5 g SO 4 2− /L. The description of this genus is based on the genome analysis of a single uncultivated bacterium found in a South African gold mine, as no cultivated representative is available at present (Chivian et al. 2008 ). Furthermore, with increasing SO 4 2− concentrations, species belonging to the OTU Desulfotomaculum ( Firmicutes ) occurred in increasing abundances up to 1% of total prokaryotes. Until 2018, the genus Desulfotomaculum comprised a heterogeneous and polyphyletic group of thermophilic spore-forming SRB, which were reclassified into five genera by Watanabe et al. ( 2018 ). Since the SILVA database used in this investigation was from December 2017, the OTU “ Desulfotomaculum ” contains the present-day genera Desulfotomaculum , Desulfallas , Desulfofundulus , Desulfofarcimen , and Desulfohalotomaculum (Watanabe et al. 2018 ). Both OTUs, Candidatus Desulforudis and Desulfotomaculum , are thermophilic and are able to oxidize H 2 among other substrates, including various sugars, VFAs, and alcohols (Muyzer and Stams 2008 ; Chivian et al. 2008 ). Besides, it should be considered that further members of the microbial community could be able to reduce SO 4 2− but were not isolated or described as such this far. Especially, the family Peptococcaceae contains versatile physiological groups including SRB, as the two genera mentioned above (Stackebrandt 2014 ). Apart from known SRB, seven bacterial OTUs were positively affected by the SO 4 2− addition and showed more than 1% relative abundance at high SO 4 2− levels. The most abundant of those was Caldicoprobacter (up to 29%), a genus, comprised of only four cultivated species, known to ferment a range of sugars including glucose and cellobiose (both intermediates of CMC mineralization) to lactate, acetate, ethanol, CO 2 , and H 2 under thermophilic conditions (Yokoyama et al. 2010 ; Bouanane-Darenfed et al. 2013 , 2011 , 2015 ). Up to date, no SO 4 2− reducing Caldicoprobacter species was found and the genus was rarely detected during NGS investigations of biogas sludge, e.g., De Vrieze et al. ( 2018 ). The OTU unclassified Clostridiales_vadinBB60_group was less abundant (up to 7%) and is based on a 16s rRNA gene fragment sequenced, when investigating the microbial diversity in an anaerobic digester fermenting vinasses (Godon et al. 1997 ). Further, positively affected OTUs include the following: the genus Proteiniborus , mainly known for fermenting proteins to acetate, ethanol, and H 2 /CO 2 (Hahnke et al. 2018 ; Niu et al. 2008 ); members of the Clostridium sensu stricto 15 group ; and uncultured members of the family Peptococcaceae . The latter is, as mentioned above, an ecologically and physiologically heterogeneous group of obligate anaerobes, including besides SRB also syntrophic, autotrophic as well as heterotrophic bacteria (Stackebrandt 2014 ). Negatively influenced bacterial OTUs (> 1% abundance in control samples) were Lentimicrobiaceae , Izimaplasmatales , and uncultured Clostridiales . At the moment, the family Lentimicrobiaceae , representing the third abundant OTU (up to 21%) in the present study, includes only one described species, Lentimicrobium saccharophilum , which was isolated from a mesophilic UASB reactor, fermenting sugars to acetate, malate, propionate, formate and H 2 (Sun et al. 2016 ). The order Izimaplasmatales belongs to the phylum Tenericutes , whose members are cell wall free bacteria typically found as parasites or commensals of eukaryotic hosts (Skennerton et al. 2016 ). The order is based on the genomes of two free-living bacteria extracted from a deep-sea CH 4 seep sediment (Skennerton et al. 2016 ). In contrast, some relatively abundant OTUs were not affected by the SO 4 2− concentrations. The second most abundant OTU Defluviitoga comprises only one cultivated species, Defluviitoga tunisiensis , isolated from a mesophilic anaerobic reactor, fermenting among others cellobiose and glucose to acetate, H 2 and CO 2 (Hania et al. 2012 ). Members of the genus Defluviitoga have been detected in great abundances in thermophilic reactors (Liang et al. 2018 ; Maus et al. 2016 ). Further on, both OTUs Clostridia_DTU014 and Clostridia MBA03 showed average relative abundance of 5% and are based only on sequences isolated from thermophilic biogas reactors (Campanaro et al. 2016 ). The archaeal community in the control samples consisted mainly of two methanogenic genera, Methanoculleus and Methanosarcina , comprising 10% and 6% of prokaryotic 16 s DNA, respectively. Both were inhibited by SO 4 2− addition, with Methanosarcina being more sensitive and totally inhibited at more than 0.5 g SO 4 2− /L, while Methanoculleus preserved a relative abundance of 4% at 1.5 g SO 4 2− /L. The higher susceptibility towards sulfide inhibition of acetoclastic compared with hydrogenotrophic methanogens was also observed previously by Maillacheruvu and Parkin ( 1996 ). As expected, the absolute abundance of SRB was pushed by the SO 4 2− addition to 10 6 copies/mL but surprisingly did not exceed the abundance of methanogens, even in samples with excessive SO 4 2− . Moestedt et al. ( 2013 ), who investigated the abundance of SRB in 25 industrial biogas fermenters with various SO 4 2− loads, found 10 5 to 10 7 copies/mL. In our investigation, mcrA copy numbers, which were used to quantify methanogen, of more than 10 6 could be found even in samples in which methanogenesis had been totally inhibited for at least 2 weeks, probably deriving from inactive or dead cells (Wagner et al. 2008 ). This result highlights the discrepancy between DNA-based abundance measurements and physiologically determined activity parameters and underlines the importance of biochemical data. To determine if the measured methanogen-specific DNA derived from free DNA, inactive cells or dead cells, a viability test was performed. It could be shown that methanogenesis restarted immediately when a small aliquot of organisms of a 3 g SO 4 2− /L variant was transferred into new SO 4 2− free medium, meaning that inhibition by H 2 S was not lethal but reversible. The same is valid for the SRB, which also seemed inhibited by H 2 S, referring to the low CO 2 production in week three and four of cultivation; transferred into fresh SO 4 2− -containing medium, SRB instantly resumed their metabolic activities. In the present study, the addition of SO 4 2− led to a decrease in propionate and butyrate concentrations and to an accumulation of acetate, although both, methanogens and SRB, should be able to utilize acetate (Muyzer and Stams 2008 ). This phenomenon was also observed by other authors under mesophilic, but also thermophilic conditions (O’Flaherty et al. 1998 ; Colleran and Pender 2002 ). As SRB have a lower threshold for H 2 , these organisms create a lower H 2 partial pressure than methanogens, which improves the thermodynamics for VFA-oxidizing H 2 -producing acetogens (Stams et al. 2005 ). Furthermore, SRB themselves incompletely oxidize propionate and butyrate to acetate, with especially propionate oxidation being thermodynamically extremely favorable (O’Flaherty et al. 1998 ). Under the physiochemical conditions at day 14, the oxidation of propionate and butyrate with SO 4 2− yields a ΔG of − 265 and − 94 kJ/mol, respectively. The preference of SRB for propionate over other organic acids was experimentally shown previously by Visser et al. ( 1993 ). This means, SRB directly and indirectly cause low levels of propionate and high levels of acetate. The higher the SO 4 2− level, the more decisive is the direct effect. In this context, Li et al. ( 2015 ) successfully tested the application of low additions of SO 4 2− (COD/SO 4 2− 200 to 350) to fight propionate accumulation during a thermophilic co-digestion of coffee grounds, milk, and waste activated sludge. The prevalence of acetate in the SO 4 2− samples in the present study might be explained by the inhibition of the acetoclastic methanogenic genus Methanosarcina by H 2 S, as seen in the NGS results. Moreover, the SO 4 2− samples were metabolically not very active in the last 2 weeks, indicating that all organisms including the SRB were at least partially inhibited by the high H 2 S concentrations. Moreover, it is possible that acetate-utilizing SRB were not present in the applied inoculum, although thermophilic acetate-oxidizing SRB have previously been isolated from methanogenic reactors (Min and Zinder 1990 ; Hattori et al. 2000 ; Balk et al. 2007 ). The OTU Desulfotomaculum comprises species able as well as species unable to oxidize acetate, while corresponding information is not available for Candidatus Desulforudis . These possible explanations were also mentioned by Colleran and Pender ( 2002 ), who also observed high effluent levels of acetate during thermophilic digestion of SO 4 2− -rich wastewater. Contrary to that, Visser et al. ( 1992 ) found that SO 4 2− reducers could outcompete acetoclastic methanogens under thermophilic conditions fed on an acetate/propionate/butyrate substrate mixture. In conclusion, our results show that SRB are able to strongly influence their abiotic environment and thus the entire microbial community even in low abundances. Finally, the altered community and chemical properties lead to a distinct shift in the carbon flow of the entire system."
} | 5,101 |
38017526 | PMC10685487 | pmc | 1,895 | {
"abstract": "The biogas produced through anaerobic digestion (AD) of renewable feedstocks is one of the promising alternatives to replace fossil-derived energy. Even though lignocellulosic biomass is the most abundant biomass on earth, only a small fraction is being used towards resources recovery, leaving a great potential unexploited. In this study, the combination of state-of-art genomic techniques and engineered systems were used to further advance the knowledge on biogas production from lignocellulosic-rich residues and the microbiome involved in the anaerobic digestion hereof. A long-term adapted anaerobic microbiome capable of degrading wheat straw as the sole substrate was investigated using protein stable isotope probing (protein-SIP). The results indicated that a diverse microbial community, primarily composed of Firmicutes and Methanogens, played crucial roles in cellulose degradation and methane production. Notably, Defluviitoga tunisiensis , Syntrophothermus lipocalidus , and Pelobacter carbinolicus were identified as direct metabolizers of cellulose, while Dehalobacterium assimilated labelled carbon through cross-feeding. This study provides direct evidence of primary cellulose degraders and sheds light on their genomic composition. By harnessing the potential of lignocellulosic biomass and understanding the microbial communities involved, we can promote sustainable biogas production, contributing to energy security and environmental preservation. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-023-02432-x.",
"conclusion": "Conclusion Protein-SIP analysis revealed 3 organisms, Defluviitoga tunisiensis , Syntrophothermus lipocalidus , and Pelobacter carbinolicus , directly metabolizing the cellulose in a digester fed with wheat straw. However, at the end of the incubation (≥ 36 h) Dehalobacterium had assimilated labelled carbon, indicating that this organism assimilated the labelled carbon through cross-feeding. This was further validated through the isotopic distribution profiles. This study presents, as one of the first, direct evidence of who are the primary degraders of cellulose by tracking the 13 C assimilation, while most other studies are based on sequence homology and therefore are limited to identify the organisms with potential to degrade the substrate. In this study, we elucidated organisms involved in the degradation of cellulose, both primary and secondary degraders, and their genomic composition. The results from this study therefore contribute to a better understanding and treatment of lignocellulosic biomass.",
"discussion": "Discussion The aim of the study was to identify and elucidate microorganisms actively involved as the primary degraders of lignocellulosic biomass. A reactor was operated to obtain a long-term adapted microbiome to degrade wheat straw under anaerobic and thermophilic conditions, followed by a protein-SIP experiment to further investigate the key-players on the cellulose degradation. Utilizing crop residues, like wheat straw, for anaerobic digestion presents a promising opportunity to capture energy through biogas production while recycling valuable nutrients for agricultural land. By harnessing the untapped potential of these crop residues, we can foster sustainable practices that contribute to energy security and soil health, forging a path towards a greener and more efficient agricultural landscape. Reactor performance The semi-continuous reactor performance for biogas production yielded a VMP and MY of 0.205 ± 0.06 L CH 4 L −1 d −1 and 0.201 ± 0.07 L CH 4 gCOD rem −1 , respectively, which corresponds to 57% of the theoretical methane yield. It is suggested that the composition of the wheat straw has influenced the biogas production due to its recalcitrant nature and effect on the hydrolytic stage of the AD. These results are in agreement with previous studies [ 12 ], in which the obtained methane yield ranged from 0.231 to 0.296 L CH 4 gVS −1 , corresponding to 65 and 85% of the theoretical methane yield, in similar experimental set-up treating wheat straw at the same OLR. It is suggested that the highest methane yield (85%) obtained by the authors may be a result of co-digesting wheat straw with sewage sludge. Changes in the bacterial and methanogenic community Regarding methanogenesis, Methanosarcina is known to have the capability to utilize all three known pathways and a minimum of nine different substrates, allowing it to utilize all three products from acetogenesis. Moreover, this genus has been reported to exhibit exceptional tolerance to high ammonium concentrations, high salt concentrations, changes in pH, and temperature [ 13 ]. The presence of high diversity in a community is generally considered a sign of ecosystem stability, as it provides a mechanism for coping with stress and environmental changes. This flexibility allows the sensitive population to adapt robustly and alter the preferred pathway of methanogenesis if needed. In contrast, Lentimicrobium and Acetomicrobium possess chemoorganotrophic metabolism and act as fermenters of a narrow range of carbohydrates [ 14 – 17 ]. As for Proteiniphilum , it has been shown that substrates such as starch and sugars can support anaerobic growth [ 18 ]. In our study, we observed the presence of Proteiniphilum throughout the 14 weeks, with relative abundances reaching up to 8.8% of the total population (Fig. 1 ). A few organisms, including Defluviitoga and Caldicoprobacter , were observed to increase in abundance right after the change in substrate from week 3 to week 4. Bacteria belonging to the genus Defluviitoga has previously been demonstrated to have a fermentative metabolism in similar environments [ 19 , 20 ]. Species of this genus can utilize a large variety of carbohydrates, as well as several complex polysaccharides such as chitin and cellulose [ 21 ]. Members of Caldicoprobacter are chemoorganotrophs, grow under strictly anaerobic conditions, and have a fermentative metabolism with an affinity for sugar substrates [ 22 , 23 ]. It seems that for the organisms which increased in the relative abundance after change in the substrate (from mix of cellulose and wheat straw till only wheat straw) all prefer sugars/sugar rich substrates, whereas the organisms with decreased abundance resemble organisms with ability to grow on less diverse mixtures of mono- and disaccharides. But common for all these organisms, whether they increased or decreased in abundance, is that after 8 weeks of adaptation to the straw substrate the microbial community is almost back at the starting point (Figs. 1 and 3 ). So, the shift in community composition appeared to be temporary and returned to the original after 14 weeks. Protein-SIP analysis Several peptides with a clear isotopical shift were found. Some of these labelled proteins were identified as hypothetical proteins (103 unique accessions), flagellin (18 unique accessions), ABC transporters (5 unique accessions), and elongation factor Tu (5 unique accessions). All of the above-mentioned proteins have common features of high turnover rate, highly abundant among many different organisms, and the detection of multiple of these proteins are in line with the results of previous studies where protein-SIP was used in different settings of AD [ 24 – 27 ]. To ensure a high significance in the detection of labelled proteins, we have in this study applied very conservative restrictions (RIA > 10%) and both automatically and manually curated the data. The isotopic distribution of most of the identified peptides followed a tailed distribution, revealing that the organisms synthesizing the peptides got the labelled carbon through cross-feeding [ 28 , 29 ] and not through direct metabolization of the labelled substrate. This was similarly observed for the two methanogens ( Methanosarcina and Methanoculleus ). When organisms directly metabolize isotopically labelled substrates, the resulting isotope pattern closely resembles a normal distribution [ 28 , 29 ]. However, when cross-feeding occurs, meaning the usage of labelled intermediates from degradation processes that accumulate in the culture, it leads to tailed distributions (negative skewness). Dehalobacterium formicoacenticum was seen to assimilate labelled carbon at the end of the incubation (after minimum 12 h). This organism has previously been reported not to be able to metabolize sugars for energy [ 30 , 31 ]. Therefore, the labelled carbon assimilation must derive from a cross-feeding of labelled intermediates from the degradations process of cellulose, which is also supported by the isotopic distribution. Three organisms have been identified as direct metabolizers of the labelled cellulose: Defluviitoga tunisiensis , Syntrophothermus lipocalidus and Pelobacter carbinolicus . Defluviitoga tunisiensis was the first to assimilate the labelled carbon from the cellulose. This organism has previously been reported as a cellulose degrader with a high H 2 -production ability [ 32 ]. The organism Syntrophothermus lipocalidus has not to our knowledge previously been described as able to degrade/metabolize cellulose. However, it has previously been shown that 3.4% of all genes in this organism were affiliated to carbohydrate transport and metabolism [ 33 ]. It was a bit surprising to see Pelobacter carbinolicus identified as one of the main degraders of the cellulose, considering that an inability to ferment sugars through glycolysis is thought to be a defining characteristic for this species [ 34 ]. However, looking into the genome of Pelobacter carbinolicus it was discovered that it encodes for proteins involved in sugar uptake [ 35 ]. This study presents clear evidence that this organism is one of the primary metabolizers and assimilates the cellulose. Also, from the genomic material of the HQ-MAG (bin.3), classified as a Pelobacter carbinolicus , multiple genes encoding carbohydrate degradation is observed, supporting that this organism can take up and metabolize sugars, such as cellulose. This study offers a significant advancement in the construction of a microbiome for biofuel production from recalcitrant lignocellulosic biomass, such as wheat straw, with broader implications for other challenging feedstocks. By employing a combination of state-of-the-art genomic techniques and long-term adaptation strategies, the research identified key microbial species directly involved in cellulose degradation and methane production. This knowledge is valuable as a tool for monitoring and optimizing microbiomes for biofuel production. Instead of relying solely on metagenomic analysis, which provides limited functional insights, the study directly links microbial identity with their functional roles in cellulose degradation. This direct evidence allows for more precise manipulation and engineering of microbial communities, ultimately leading to improved biofuel production efficiency and sustainability. Moreover, the findings shed light on the potential of underutilized lignocellulosic biomass resources, contributing to a more environmentally friendly and economically viable bioenergy sector."
} | 2,811 |
37610031 | PMC10651141 | pmc | 1,896 | {
"abstract": "Summary Plant‐based co‐production of polyhydroxyalkanoates (PHAs) and seed oil has the potential to create a viable domestic source of feedstocks for renewable fuels and plastics. PHAs, a class of biodegradable polyesters, can replace conventional plastics in many applications while providing full degradation in all biologically active environments. Here we report the production of the PHA poly[( R )‐3‐hydroxybutyrate] (PHB) in the seed cytosol of the emerging bioenergy crop Camelina sativa engineered with a bacterial PHB biosynthetic pathway. Two approaches were used: cytosolic localization of all three enzymes of the PHB pathway in the seed, or localization of the first two enzymes of the pathway in the cytosol and anchoring of the third enzyme required for polymerization to the cytosolic face of the endoplasmic reticulum (ER). The ER‐targeted approach was found to provide more stable polymer production with PHB levels up to 10.2% of the mature seed weight achieved in seeds with good viability. These results mark a significant step forward towards engineering lines for commercial use. Plant‐based PHA production would enable a direct link between low‐cost large‐scale agricultural production of biodegradable polymers and seed oil with the global plastics and renewable fuels markets.",
"introduction": "Introduction One of the most interesting classes of biomaterials are the polyhydroxyalkanoate (PHA) family of microbial biopolymers (Madison and Huisman, 1999 ; Suriyamongkol et al ., 2007 ) that have unique properties depending on their monomer unit composition. PHAs are a class of polyesters produced by a biological polymerization process and are accumulated by some microbes as a carbon storage and energy reserve under nutrient‐limiting conditions. The subsequent intracellular or extracellular consumption of PHAs by microbes upon the return of favourable growth conditions provides for their natural biodegradability in many environments (Madison and Huisman, 1999 ; Snell and Peoples, 2009 ; Suriyamongkol et al ., 2007 ). A wide range of monomers can be incorporated into PHAs (Steinbuchel and Valentin, 1995 ) which most often consist of short‐chain (3–5 carbon units) and/or medium‐chain‐length (6–14 carbon units) monomers. Polymers that have been produced in engineered bacteria by fermentation for commercial purposes include the homopolymers poly[( R )‐3‐hydroxybutyrate] (PHB) and poly(4‐hydroxybutyrate) (P4HB), and copolymers poly[( R )‐3‐hydroxybutyrate‐co‐4‐hydroxybutyrate] (PHB‐4HB), poly[( R )‐3‐hydroxybutyrate‐co‐( R )‐3‐hydroxyvalerate] (PHBV) and poly[( R )‐3‐hydroxybutyrate‐co‐( R )‐3‐hydroxyhexanoate] (Koller and Mukherjee, 2022 ). Varying the monomer composition of PHAs yields materials with a range of properties such that suitable replacements for many petroleum‐derived plastics can be obtained. Besides plastics, PHAs can also be used in medical applications (Chen and Zhang, 2018 ; Koller, 2018 ; Koller and Mukherjee, 2022 ; Lim et al ., 2017 ), converted by thermolysis to small molecule (Kang and Yu, 2015 ; Mamat et al ., 2014 ; Mullen et al ., 2014 ) precursors to commodity chemicals (Schweitzer et al ., 2015 ; Schweitzer and Snell, 2015 ), and used in wastewater treatment applications (Chu and Wang, 2016 ; Hiraishi and Khan, 2003 ; Wang and Chu, 2016 ). In the latter, degradation of PHAs by denitrifying bacteria supplies reducing power for reactions that convert nitrates and other nitrogen containing species to inert nitrogen gas (Hiraishi and Khan, 2003 ). PHAs produced by bacterial fermentation have successfully accessed markets in high‐value medical applications, but the high capital and operating costs of fermentation have so far prohibited their widespread use in commodity applications. Regardless of the end‐of‐life environmental benefits enabled by biodegradability, cost remains the key factor for broad adoption of new products. Engineering plants to produce PHAs enables polymer production from CO 2 at the capital efficiency and scale of agriculture. Extensive work has been performed to produce PHB in leaf chloroplasts (Poirier and Brumbley, 2010 ; Snell et al ., 2015 ; Somleva et al ., 2013 ; Van Beilen and Poirier, 2012 ) by co‐expressing genes encoding plastid‐targeted β‐ketothiolase (PhaA), NADPH‐dependent reductase (PhaB) and PHA synthase (PhaC) enzymes to convert plastidial acetyl‐CoA to polymer (Figure S1 ; Snell et al ., 2015 ). High level production of PHB in chloroplasts has been achieved, with up to ~40% dry cell weight PHB achieved in Arabidopsis thaliana (Bohmert et al ., 2000 ), but this has often been met with stunted plant phenotypes when significant levels (>3% dry cell weight) are produced (Bohmert et al ., 2000 , 2002 , 2004 ; Somleva et al ., 2013 ). Prior work in oilseeds has targeted PHB synthesis to seed plastids (Houmiel et al ., 1999 ; Malik et al ., 2015 ; Valentin et al ., 1999 ) by co‐expressing genes encoding plastid‐targeted PhaA, PhaB, and PhaC enzymes to divert acetyl‐CoA from fatty acid biosynthesis to produce polymer. Up to 7.7% of the seed fresh weight was produced in seed plastids of Brassica napus (Houmiel et al ., 1999 ; Valentin et al ., 1999 ) and up to 15.2% (mature seed weight) was obtained in Camelina sativa (Malik et al ., 2015 ). The high levels achieved in Camelina, however, resulted in chlorotic cotyledons, poor emergence and low‐seedling survival (Malik et al ., 2015 ). Despite the challenges, production of PHAs in oilseeds can be advantageous since it simplifies logistics and storage of the harvested product prior to processing, and would coproduce revenue generating seed oil and protein rich meal in a biorefinery (Snell and Peoples, 2013 ). In this study, we chose to investigate PHB production in the cytosol of seeds to determine if moving PHB production away from the developing photosynthetic machinery in the seed embryo plastid to the cytosol would yield seeds with high levels of PHB and good seedling emergence. PHB was chosen as the target with the goal of enabling production of a cheap source of homopolymer that could be blended with other biopolymers to create materials with improved properties (Koller and Mukherjee, 2022 ; Li et al ., 2016 ). Prior work with PHB production in the cytosol has been limited to leaves and has resulted in the production of low levels of polymer (Poirier et al ., 1992b ) with significantly reduced fresh plant weight (Poirier et al ., 1992a , b ; Snell et al ., 2015 ; Somleva et al ., 2013 ). To our knowledge, cytosolic production of PHB in seeds has not been previously reported. The highest reported level of cytosolic PHB produced in any organ to date is 0.34% dry weight (DW) PHB produced in cotton fibres (John and Keller, 1996 ). C. sativa , an annual plant from the Brassicaceae family, was chosen as the host for our work since its seeds contain high levels of oil and the plant can thrive in marginal growth conditions (Eynck et al ., 2013 ). Camelina also does not cross‐pollinate with B. napus (FitzJohn et al ., 2007 ), an important edible oilseed crop. In addition to expressing the PHB biosynthesis pathway in the seed's cytosol, a key element of our strategy was to modify the last enzyme of the PHB biosynthetic pathway, PHA synthase (PhaC), with a C‐terminal sequence that has been previously shown to tail‐anchor (Abell and Mullen, 2011 ) recombinant proteins to the cytosolic face of the endoplasmic reticulum (ER) (Barbante et al ., 2008 ). Seed PHB levels, the viability and health of seedlings, and the partitioning of carbon between polymer, seed oil and seed protein are described.",
"discussion": "Discussion Interest in renewable sources of biodegradable plastics has increased with the closing of overseas markets for plastic waste and recent public awareness that recycling is not economically viable (Sullivan, 2020 ). PHAs are a leading solution to the world's plastics problem since they are renewable, biodegradable materials with properties that can replace petroleum‐based plastics in many applications. In fact, PHA bioplastics were named one of the 25 ideas that will shape this decade (Fortune, 2019 ) because of their unique biodegradability in all biologically active environments, including soil, rivers, oceans, compost and sewage (Snell and Peoples, 2009 ). The widespread use of PHAs produced by large‐scale bacterial fermentations has been hindered by the high capital and operating costs of production. Today, starches and vegetable oils are produced on an enormous scale at very low cost through agriculture. Similarly, plant‐based production of PHAs could significantly decrease costs of production. The concept of plant‐based PHA production was first described over 30 years ago (Pool, 1989 ), and despite much effort by academic and industrial researchers, primarily targeted towards the production of PHB (Snell et al ., 2015 ; Somleva et al ., 2013 ; Suriyamongkol et al ., 2007 ), industrially relevant success has not yet been achieved. Production of PHAs in plants has often led to negative agronomic effects when the biosynthetic pathways are expressed, including impaired plant growth when pathways are expressed in vegetative tissue (Bohmert et al ., 2000 , 2004 ; Somleva et al ., 2013 ) or impaired seedling survival with seed‐specific expression (Malik et al ., 2015 ). Oilseeds such as Camelina sativa are ideal for biosynthesis of PHB and its copolymers, since acetyl‐CoA is a common substrate for both PHB and seed oil biosynthesis (Figure 1 ). Previous work in our labs was successful in producing high levels of polymer in Camelina seed plastids, reaching up to 15.2% of the mature seed weight, but cotyledons were chlorotic and a significant negative impact on seedling emergence and viability was observed (Malik et al ., 2015 ). We reasoned that production of PHB in an alternate location, away from the developing photosynthetic machinery of the seedling, might increase seedling vigour. In this study, we evaluated the cytosol as an alternative seed‐specific production site for PHB. At the onset of our work, production of PHB in the cytosol of plants was considered problematic as prior attempts in leaves yielded only low levels of PHB (highest level in a plant, 0.34% DW PHB, cotton fibres (John and Keller, 1996 )) with often stunted plant phenotypes (Poirier and Brumbley, 2010 ; Snell et al ., 2015 ; Somleva et al ., 2013 ). To our knowledge, no efforts to produce PHB in the cytosol of seeds have been reported. Since the cytosol of seeds supplies malonyl‐CoA for ER‐associated fatty acid elongation reactions (Figure 1 ), we reasoned that a greater pool of accessible acetyl‐CoA might be available in the cytosol of developing seeds than in leaves for production of PHB. Cytosolic production of PHB using strong seed‐specific promoters (transformation construct pMBXS394) yielded PHB levels of up to 4.5% of the mature seed weight in T 2 seeds, 13 times higher than the previously reported highest amounts of cytosolic PHA produced in plants (0.34% DW PHB, cotton fibres; John and Keller, 1996 ). Surprisingly, PHB levels dropped in subsequent generations suggesting that cytosolic PHB production was not stable. We next pursued a strategy of anchoring PHA synthase to the cytosolic face of the ER to attempt to stabilize the PHA synthase and thus polymer production. ER targeting has previously been used to increase the production of human immunodeficiency virus proteins Nef (Barbante et al ., 2008 ) and p24 (Virgili‐López et al ., 2013 ) in tobacco, possibly by increasing the stability of the proteins and/or making them less susceptible to degradation by proteases (Pedrazzini, 2009 ). Fusion of the gene encoding PHA synthase (PhaC) to a similar ER tail‐anchoring sequence and co‐expression with the other genes of the PHB biosynthetic pathway (transformation construct pMBXS763) yielded lines that produced PHB up to 4.9% of the mature seed weight in T 2 seeds. Importantly, these lines showed stable PHB production through multiple generations with some lines showing increased polymer levels producing up 7.1% PHB in homozygous T 4 seeds in the greenhouse (Table 2 ). While targeting of the PHA synthase (PhaC) to the ER was not experimentally confirmed, our results suggest that PhaC fused to the ER targeting signal in transformation construct pMBXS763 provides more stable PHB synthesis than transformation construct pMBXS394 containing PhaC without an ER targeting signal. Several independent ER targeted events showed more stable PHB production through multiple generations and, in some cases, produced higher polymer levels. Additional work is required to understand the factors contributing to increased stability of product formation in ER‐targeted lines compared to their cytosolic counterparts. Use of a controlled environmental chamber to simulate spring growth conditions produced higher levels of PHB with up to 10.2% PHB observed in T 4 seeds with good emergence and survival of T3 seedlings (78%, Table 2 ). ER‐targeted PHB producing seedlings were green and healthy but possessed cotyledons that were narrower than wild‐type controls (Figure 3 ). The chlorotic phenotype previously observed in seedlings of seed‐specific plastid PHB producers (Malik et al ., 2015 ) was not observed. Past attempts at commercialization of microbial based PHAs have been difficult, primarily due to the high cost of advanced fermentation‐produced copolymers that have restricted entry into many markets. We believe crop‐based production will enable an advantaged cost structure, thereby eliminating the remaining barrier to entry for broad adoption of these materials. The multiple co‐products harvested in a PHB‐producing seed, including oil, polymer and protein‐rich seed meal, increase the seed value, providing additional revenue streams to support a business (Snell and Peoples, 2013 ). We have demonstrated the production of up to 10.2% PHB in seeds of Camelina with good seedling viability. This is a significant step towards producing field‐ready lines for a commercial pipeline. This study is the first report of cytosolic production of PHB in seeds, and is the highest level of cytosolic PHB production in a plant to date at 30 times the previously reported highest amounts of cytosolic PHA produced (John and Keller, 1996 )."
} | 3,610 |
28577167 | null | s2 | 1,898 | {
"abstract": "Plant establishment during phytostabilization of legacy mine tailings in semiarid regions is challenging due to low pH, low organic carbon, low nutrients, and high toxic metal(loid) concentrations. Plant-associated bacterial communities are particularly important under these harsh conditions because of their beneficial services to plants. We hypothesize that bacterial colonization profiles on rhizoplane surfaces reflect deterministic processes that are governed by plant health and the root environment. The aim of this study was to identify associations between bacterial colonization patterns on buffalo grass (Buchloe dactyloides) rhizoplanes and both plant status (leaf chlorophyll and plant cover) and substrate biogeochemistry (pH, electrical conductivity, total organic carbon, total nitrogen, and rhizosphere microbial community). Buffalo grass plants from mesocosm- and field-scale phytostabilization trials conducted with tailings from the Iron King Mine and Humboldt Smelter Superfund Site in Dewey-Humboldt, Arizona, were analyzed. These tailings are extremely acidic and have arsenic and lead concentrations of 2-4 g kg"
} | 284 |
24949257 | PMC4052673 | pmc | 1,900 | {
"abstract": "High power densities have been obtained from MFC reactors having a purple color characteristic of Rhodopseudomonas . We investigated the microbial community structure and population in developed purple MFC medium (DPMM) and MFC effluent (DPME) using 16S rRNA pyrosequencing. In DPMM, dominant bacteria were Comamonas (44.6%), Rhodopseudomonas (19.5%) and Pseudomonas (17.2%). The bacterial community of DPME mainly consisted of bacteria related to Rhodopseudomonas (72.2%). Hydrogen oxidizing bacteria were identified in both purple-colored samples: Hydrogenophaga and Sphaerochaeta in the DPMM, and Arcobacter , unclassified Ignavibacteriaceae , Acinetobacter, Desulfovibrio and Wolinella in the DPME. The methanogenic community of both purple-colored samples was dominated by hydrogenotrophic methanogens including Methanobacterium, Methanobrevibacter and Methanocorpusculum with significantly lower numbers of Methanosarcina . These results suggeste that hydrogen is actively produced by Rhodopseudomonas that leads to the dominance of hydrogen consuming microorganisms in both purple-colored samples. The syntrophic relationship between Rhodopseudomonas and hydrogenotrophic microbes might be important for producing high power density in the acetate-fed MFC under light conditions.",
"introduction": "Introduction Microbial fuel cell (MFC) is a new technology in renewable energy. It generates electrical power while accomplishing waste water treatment by utilizing microorganisms (Pant et al. [ 2012 ]). Although MFCs have been comprehensively investigated, this technology is still at an early stage and under extensive laboratory research. Over the past 10 years, considerable effort has been made to improving power generation efficiency, focusing mainly on different MFC setups and new materials (Kim et al. [ 2008 ]; Lovley [ 2006 ]; Sleutels et al. [ 2012 ]). Microbial ecology studies in MFC systems are important for understanding the mechanism of microbial electricity generation (Rabaey et al. [ 2007 ]; Rabaey and Rozendal [ 2010 ]). In addition, MFC systems also provide insight into the physiological roles of microbes and better understanding of interactions in complex microbial communities within natural environments (Bretschger et al. [ 2010 ]). The bacterial population and predominant species vary depending on operational conditions such as inocula, substrate nature and electrode materials (Sun et al. [ 2012 ]; Logan and Regan [ 2006 ]; Logan [ 2009 ]). In general, MFCs using mixed culture produce more power than ones with pure culture (Watson and Logan [ 2010 ]). However, complex syntrophic interactions existing in MFC systems relating to high power densities have not been well studied. Previous studies have also shown that MFC performances are affected by light, which can cause the solution medium of respective MFCs to enrich in the phototrophic purple nonsulfur (PNS) bacterium, Rhodopseudomonas palustris DX-1. This bacterium has been shown to produce higher power densities in pure culture than in mixed culture and increase power production along with light intensity. In contrast to, the R. palustris ATCC 17001 does not generate power (Xing et al. [ 2008 ]). The genus Rhodopseudomonas liberates hydrogen when illuminated anaerobically in the presence of a carbon source such as acetate and malate (Barbosa et al. [ 2001 ]). The hydrogen production is mediated by a nitrogenase enzyme and dependent on light (Basak and Das [ 2007 ]; Rey et al. [ 2007 ]). Previous MFC microbial ecology studies have used Sanger-based 16S rRNA sequencing for microbial community characterization. Recent advances in next generation sequencing such as pyrosequencing offer a better alternative for comprehensively characterizing microbial communities, especially in terms of less abundant members (Dowd et al. [ 2008 ]; Huse et al. [ 2007 ]; Lee et al. [ 2010 ]). 454 pyrosequencing of 16S rRNA gene has been revealed highly diverse microbial communities in MFCs and MECs (Lee et al. [ 2010 ]; Jia et al. [ 2013 ]; Lu et al. [ 2012a ]). The present study aimed at better understanding the high power density achieved in MFC reactors having purple color characteristic of Rhodopseudomonas. To do this, the microbial communities in the purple-colored samples collected from a media bottle and a MFC reactor respectively were investigated using 16S rRNA amplicon pyrosequencing.",
"discussion": "Discussion Light conditions significantly enhanced power densities (around 8–16%) in the MFCs using mixed and pure cultures of R. palustris strains DX-1 and RE-2 respectively (Xing et al. [ 2009 ]). In order to better understand electricity generation in MFCs containing purple photosynthetic Rhodopseudomonas palustris in mixed culture, microbial communities in DPMM and DPME were investigated using 16S rRNA gene pyrosequencing. As expected, Rhodopseudomonas was found to be dominant in both samples, but it was present in much higher abundance in DPME than DPMM; this might be due to the fact that DPME has developed for a longer period of time (2 years) than DPMM (4 weeks). The Rhodopseudomonas sp. detected in our study does not provide insight on its capability for electricity generation. The Rhodopseudomonas sp. was found to be more closely related with R. faecalis and R. rhenobacensis which are unknown for their ability to generate electricity, and R . palustris ATCC 17001 which does not produce an electric current, but it is less closely related to known electrogenic bacteria R. palustris strains DX-1 (Xing et al. [ 2008 ]). Previous studies have shown that Geobacter , Comamonas and Pseudomonas dominated the anode communities of acetate-fed MFCs under light conditions, revealed by PCR/DGGE and 16S rRNA gene library analyses (Xing et al. [ 2009 ]). The capacity of electricity generation in acetate-fed MFCs by those genera has also been revealed (Xing et al. [ 2010 ]; Pham et al. [ 2008 ]; Xing et al. [ 2008 ]). However, these genera were found with low abundance in the DPME. This contrast in abundance might indicate that the species structure of the bacterial community differ between the MFC anode biofilm and effluent, whilst also being dependent on inoculum source. Chemolithotrophic bacterium Hydrogenophaga , which consumes hydrogen and carbon dioxide as energy and carbon sources respectively (Willems et al. [ 1989 ]). They were found in DPMM but not in DPME. Geobacter sulfurreducens has been found in syntrophic cooperation with the Hydrogenophaga sp. strain AR20 in acetate-fed MFC (Kimura and Okabe [ 2013a ]). In DPMM, Hydrogenophaga might use the hydrogen produced by Rhodopseudomonas but not the acetate. Members of chemolithotrophic bacteria were also found in DPME, including unclassified Ignavibacteriaceae , Acinetobacter , Desulfovibrio and Azospirillum which are known to be capable of using hydrogen as an electron donor (Wong et al. [ 1986 ]; Gross and Simon [ 2003 ]; Yong et al. [ 2002 ]; Iino et al. [ 2010 ]). It has been reported that the growth of G. sulfurreducens was more efficient when co-cultured with Wolinella and Desulfovibrio, which act as hydrogen-consuming partners; with nitrate as the electron acceptor, acetate oxidation was more rapid, resulting in faster growth of Geobacter sulfurreducens under low hydrogen partial pressure (Cord-Ruwisch et al. [ 1998 ]). The removal of hydrogen by the hydrogen-consuming bacteria in the DPME might have stimulated the growth of Rhodopseudomonas . Our results suggested that the growth of Rhodopseudomonas was due to the syntrophic cooperation with hydrogen-consuming bacteria in both purple-colored samples. Anodic hydrogen oxidation by hydrogenotrophic exoelectrogens produces electrical current after complete oxidation of acetate, and a rapid current increase occurs when hydrogen gas is supplied to the reactor (Lee et al. [ 2009 ]). However, a previous study showed that electricity generation by hydrogenotrophic exoelectrogens was excluded due to inhibition of nitrogenase-dependent hydrogen production in Rhodopseudomonas by NH 4 Cl in the medium (Hillmer and Gest [ 1977 ]; Rey et al. [ 2007 ]). The presence of hydrogen-oxidizing bacteria in the purple-colored samples suggested that hydrogen production by Rhodopseudomonas could not be completely inhibited by NH 4 Cl in the medium (Rey et al. [ 2007 ]). Potential hydrogenotrophic exoelectrogens might coexist with Rhodopseudomonas in the anode biofilm due to their involvement in electricity generation via hydrogen oxidation. This potential syntrophic interaction between Rhodopseudomonas and hydrogen-oxidizing bacteria might be one of the explanations for the high power density obtained from illuminated MFCs. Bacterial 16S rRNA gene sequences were also detected by the V3-V5 primer set and inconsistent results in the bacterial diversity and the abundance patterns of bacterial community were observed between V1-V3 and V3-V5 data sets. In our previous study (unpublished data), the V3-V5 primer set was also applied for methane-producing biocathodes in MECs, which accounted for 99% of the total sequences belonged to the archaea domain. The PCR amplification of the V3–V5 region of 16S rRNA can be biased depending on the activity of the archaeal population present in the samples. The differences in portions of archaeal sequences between two purple-colored samples suggest that the DPME could have a more active archaeal population than DPMM. Archaeal community in MFC reactors enriched with Rhodopseudomonas sp. has not been reported yet. The predominant hydrogenotrophic methanogens were found in both purple-colored samples, but not acetoclastic methanogens despite the MFC medium containing high concentration of acetate. Under anaerobic conditions, methanogenic archaea often partner with heterotrophic H 2 -producing bacteria which catalyze oxidation of a variety of organic compounds (fatty acids, alcohols, and aromatic compounds). The methanogens utilize the H 2 produced by these heterotrophic bacteria during methanogenesis while the heterotrophic bacteria benefit from the methanogens, which play a role in the removal of excess hydrogen that would inhibit their growth (McInerney et al. [ 2008 ]; Sakai et al. [ 2009 ]). Hydrogenotrophic methanogenesis by Methanobrevibacter in DPME might be relying on the predominance of Rhodopseudomonas for interspecies H 2 transfer. However, it seems likely that CO 2 derived from Rhodopseudomonas is limited as a major source of carbon for methane generation due to its known role in being recycled into biomass (McKinlay and Harwood [ 2010 ]). Hydrogen generation by Rhodopseudomonas in both purple-colored samples containing acetate could share similar environmental condition with hydrogen-producing MEC reactors fed with acetate, where hydrogenotrophic methanogens have emerged as the most active methanogens (Lu et al. [ 2012a ]). In general, few acetoclastic methanogens are present in acetate fed-MFCs due to their inhibition in growth as a result of air exposure and outcompetition for organic substrates by facultative anaerobes and exoelectrogens (Chae et al. [ 2010 ]; Dar et al. [ 2008 ]). In addition, hydrogen inhibition of growth and acetate metabolism in Methanosarcina species has been reported (Ahring et al. [ 1991 ]; Ferguson and Mah [ 1983 ]). Minor acetoclastic methanogen ( Methanosarcina ) present in the DPMM and the non-detection of Methanosarcina in the DPME might be indication of the robust inhibition by hydrogen produced by Rhodopseudomonas and its acetate competition. In conclusion, although the electric generation ability of Rhodopseudomonas detected in our study is unknown, putative hydrogenotrophic exoelectrogens existing in MFC reactors having purple-colored effluents might contribute to increasing power density. For future studies the above observations are needed to investigate how known-exoelectrogenic phototrophic bacteria can be applied to efficiently generate electricity via syntrophic relationships with the hydrogenotrophic exoelectrogens and methanogens in the acetate-fed MFC under light conditions."
} | 3,043 |
26350329 | null | s2 | 1,903 | {
"abstract": "Bacteria have traditionally been studied as single-cell organisms. In laboratory settings, aerobic bacteria are usually cultured in aerated flasks, where the cells are considered essentially homogenous. However, in many natural environments, bacteria and other microorganisms grow in mixed communities, often associated with surfaces. Biofilms are comprised of surface-associated microorganisms, their extracellular matrix material, and environmental chemicals that have adsorbed to the bacteria or their matrix material. While this definition of a biofilm is fairly simple, biofilms are complex and dynamic. Our understanding of the activities of individual biofilm cells and whole biofilm systems has developed rapidly, due in part to advances in molecular, analytical, and imaging tools and the miniaturization of tools designed to characterize biofilms at the enzyme level, cellular level, and systems level."
} | 228 |
37688153 | PMC10490179 | pmc | 1,904 | {
"abstract": "Despite being primarily categorized as non-autonomous self-healing polymers, we demonstrate the ability of Diels–Alder polymers to heal macroscopic damages at room temperature, resulting in complete restoration of their mechanical properties within a few hours. Moreover, we observe immediate partial recovery, occurring mere minutes after reuniting the fractured surfaces. This fast room-temperature healing is accomplished by employing an off-stoichiometric maleimide-to-furan ratio in the polymer network. Through an extensive investigation of seven Diels–Alder polymers, the influence of crosslink density on self-healing, thermal, and (thermo-)mechanical performance was thoroughly examined. Crosslink density variations were achieved by adjusting the molecular weight of the monomers or utilizing the off-stoichiometric maleimide-to-furan ratio. Quasistatic tensile testing, dynamic mechanical analysis, dynamic rheometry, differential scanning calorimetry, and thermogravimetric analysis were employed to evaluate the individual effects of these parameters on material performance. While lowering the crosslink density in the polymer network via decreasing the off-stoichiometric ratio demonstrated the greatest acceleration of healing, it also led to a slight decrease in (dynamic) mechanical performance. On the other hand, reducing crosslink density using longer monomers resulted in faster healing, albeit to a lesser extent, while maintaining the (dynamic) mechanical performance.",
"conclusion": "5. Conclusions In conclusion, caution must be exercised when categorizing intrinsic self-healing polymers into autonomous and non-autonomous (stimulus-dependent) categories, as the ability to self-heal in ambient conditions relies not only on the reversible chemistry but also on the composition of the polymer network. In the case of Diels–Alder polymers, if the primary objective is rapid healing, the preferred approach is to accelerate the process by reducing the crosslink density through a lower maleimide-to-furan ratio ( r ). However, if the material’s mechanical properties, such as fracture strain and toughness, as well as its dynamic performance, are of utmost importance, it is recommended to expedite healing by decreasing the crosslink density through the utilization of monomers with higher molecular weight. Nevertheless, both techniques demonstrate the remarkable effectiveness of fast room temperature self-healing in Diels–Alder polymers, thereby invalidating the previous classification of these materials as non-autonomous.",
"introduction": "1. Introduction In light of the prevailing sustainability goals, extending the lifespan of products is of paramount importance. Consequently, there is a growing emphasis on the development of self-healing polymers, which possess the capability to fully recover from substantial damage. In the first classification, self-healing polymers are classified as extrinsic [ 1 ] or intrinsic [ 2 ]. In extrinsic self-healing polymers, the healing process relies on the release and solidification of a healing agent that is extrinsically encapsulated within the polymer network (e.g., microcapsules, nanocapsules, or vascular network). However, this approach has drawbacks: the healing capacity diminishes with repeated damage–healing cycles, and the material properties may not be identical to the original in the healed area. In contrast, intrinsic self-healing polymers, such as those utilized in this study, employ inherently reversible chemistries and can theoretically heal an infinite number of times. In a secondary classification, self-healing polymers can be categorized as either autonomous or non-autonomous. The majority of extrinsic self-healing polymers exhibit autonomous healing capabilities, whereas intrinsic self-healing polymers can be further classified into autonomous or non-autonomous categories. Autonomous self-healing polymers can heal under ambient conditions without requiring any external stimulus. On the other hand, non-autonomous self-healing polymers necessitate an external stimulus, typically heat, to activate the healing process. However, it should be noted that the autonomous healing ability is not solely dependent on the healing mechanism but also on the composition of the polymeric network in which it is incorporated. Our paper demonstrates that these factors should not be overlooked. Within the domain of intrinsic self-healing polymers, a fundamental trade-off exists between mechanical strength/stability and the time/temperature required for healing [ 3 , 4 ]. Generally, polymers with higher bond strengths [ 5 ], such as those formed through dissociative covalent interactions like Diels–Alder bonds or associative covalent interactions [ 6 ] like disulfide bonds [ 7 ] and transesterifications [ 8 ], exhibit superior mechanical strength and stability due to the increased energy required to break these bonds. However, this also implies that a greater amount of energy must be supplied by the healing stimulus, resulting in slower healing mechanisms or even infeasibility at ambient temperatures. Conversely, the total bond energy of physical crosslinks [ 9 ], like hydrogen bond crosslinks [ 10 ] and coordination complexes [ 11 , 12 , 13 ], is significantly lower, necessitating less energy for their disruption. As a result, these physico-chemical networks can heal damage at lower or ambient temperatures. Nevertheless, the lower total bond energy has adverse effects on the mechanical properties of the polymers. Physically crosslinked self-healing polymers tend to have reduced mechanical strength, and in certain cases, they may exhibit creep behavior and non-negligible self-adhesion. These factors can lead to permanent plastic deformations and suboptimal strain recovery. It is crucial to acknowledge that the healing characteristics of self-healing materials are not solely determined by the type of reversible bond they possess. The composition of their network also plays a significant role, an aspect that is often overlooked and underestimated in the existing literature. Therefore, the primary focus of this paper is to explore the impact of altering the network composition on enhancing the healing capabilities of these materials. We employ the extensively utilized reversible Diels–Alder reaction to synthesize our self-healing polymers. Compared to other reversible bonds, the Diels–Alder bond possesses a moderate strength, which leads to an outstanding balance between mechanical strength, stability, and healing time on the order of hours at mild temperatures [ 14 ]. Therefore, they have been used to introduce healing capacities in multiple applications, including soft robotics [ 15 , 16 ], flexible electronics [ 17 ], and coatings [ 18 ]. While the self-healing capability of furan and maleimide-based Diels–Alder polymers have been extensively documented, the majority of self-healing procedures described involve non-autonomous healing at temperatures ranging from 80 to 130 °C. However, in most applications, the need for external heat stimulus to activate the healing process complicates the system. Hence, the demand for autonomous healing under application conditions, predominantly at ambient conditions, is significant. In an initial proof-of-concept study, the authors have successfully demonstrated that by modifying the network composition to an off-stoichiometric maleimide-to-furan ratio, Diels–Alder polymers can autonomously undergo self-healing even under ambient conditions [ 19 ]. Building upon this discovery, this research delves into an in-depth investigation of methods to expedite the self-healing process. Alternatively, there have been reports on using the Diels–Alder chemistry in autonomous extrinsic self-healing polymers [ 20 ]. The primary enabling factor for our study is the significant advantage of the tunability of Diels–Alder polymers. Previous studies have demonstrated a remarkable range of mechanical properties, spanning from highly flexible elastomers with Young’s modulus in the 100 kPa range to rigid thermosets with moduli reaching a few GPa [ 21 ]. Achieving this versatility involves manipulating various network design parameters. These parameters include adjusting the molecular weight of the monomer units [ 21 ], controlling the functionality of the monomers [ 22 ] (e.g., the number of reactive maleimides or furan per monomer unit), and tuning the stoichiometric ratio between furan and maleimide [ 23 , 24 , 25 ]. Furthermore, blending monomers of the same type but with different molecular weights can create polymer network blends. Through blending, the resulting self-healing Diels–Alder polymers may exhibit partial phase separation within the network, leading to enhanced toughness and stiffness. In this paper, we demonstrate that these network design parameters, particularly the molecular weight of the monomer and the extent of off-stoichiometry, can also accelerate the healing process under ambient conditions. This acceleration leads to autonomous healing within minutes at room temperature ( Figure 1 ). Diels–Alder polymers possess an additional advantage in terms of their remarkable manufacturability. Unlike classical elastomers and thermosets, which are characterized by irreversible crosslinks in their network structure, Diels–Alder polymers can undergo thermal reprocessing. This unique feature not only presents significant potential for material recycling but also enables a wide array of processing techniques to be employed [ 14 ]. These techniques encompass formative manufacturing methods such as casting [ 26 ], injection molding [ 27 ], and compression molding [ 28 ], as well as additive manufacturing approaches including fused filament fabrication [ 29 , 30 ], direct ink writing [ 31 ], and selective laser sintering [ 32 ]. Moreover, assembly and binding techniques can also be utilized with these polymers [ 14 , 15 ]. To optimize the healing process, it is crucial to understand the specific factors that contribute to its enhancement when modifying the network composition: (i) Firstly, it is essential to have a significant quantity of reactive furan and maleimide groups present on the surfaces of fractures to facilitate the formation of Diels–Alder bonds through reaction. When damage occurs, the reversible Diels–Alder bonds break along the fracture line since, although relatively strong, they are the weakest covalent bond within the network. By increasing the concentration of Diels–Alder bonds in the network (e.g., the crosslink density), a greater number of bonds will break during damage, resulting in increased availability of reactive furan and maleimide groups at the fracture surfaces. (ii) Secondly, network mobility plays a vital role in the healing process. On a microscopic level, increased mobility leads to improved contact between the fracture surfaces. Furthermore, at a molecular level, this increased mobility facilitates the encounter between reactive furan and maleimide groups, making it easier for them to find suitable binding partners and form Diels–Alder bonds. The network mobility is increased upon decreasing the Diels–Alder crosslink density. The inverse relationship between these two contributions is evident, highlighting the necessity for a comprehensive study to determine the relative importance of each factor in the context of autonomous self-healing. Two methods can be employed to modify the Diels–Alder crosslink density in Diels–Alder polymers. The first method involves adjusting the molecular weight of the monomers within the network, while the second method involves varying the off-stoichiometric ratio of maleimide to furan in the network. To investigate the impact of these network design parameters on thermal characteristics and (thermo-)mechanical performance, a series of tests were conducted on two sets of materials comprising four and three Diels–Alder polymers. The tests included quasistatic tensile testing, dynamic mechanical analysis (DMA), dynamic rheometry, differential scanning calorimetry (DSC), and thermogravimetric analysis (TGA). Furthermore, the influence of these parameters on the self-healing capabilities was analyzed. The series of materials were subjected to quasistatic tensile testing until fracture, both before and after healing at room temperature (25 °C) for various durations ranging from 5 min to 24 h. This allowed for the observation of the evolution of Young’s modulus, stress at fracture, strain at fracture, and toughness as a function of healing time, as well as the resulting healing efficiencies. Notably, these tests involved healing significant damages, which involved cutting the tensile samples in half using a scalpel blade.",
"discussion": "4. Discussion Although classified as non-autonomous in the literature [ 39 ], Diels–Alder-based reversible polymers exhibit autonomous self-healing of macroscopic damages at ambient conditions, such as room temperature. This paper showed five Diels–Alder networks that could heal within 24 h from significant damage (e.g., being cut in half) and at ambient conditions without the addition of any external heat. However, achieving this requires mobility within the network, which can be attained by reducing the crosslink density. This reduction can be accomplished through two strategies investigated in this paper: increasing the molecular weight of the monomers or decreasing the maleimide-to-furan ( r ) ratio. In general, reducing the crosslink density, whether through an increase in molecular weight or reduction in the r -ratio, results in a faster increase in healing efficiency, e.g., a shorter healing time. However, care should be taken when evaluating the healing performance based on the healing efficiency alone. The healing efficiency is defined as the recovery of the mechanical properties prior to damage, including Young’s modulus, fracture stress, fracture strain, and toughness. The lower these initial mechanical properties are, the faster they can be recovered at room temperature. Materials with a higher crosslink density have a higher Young’s modulus, fracture stress, fracture strain, and toughness; therefore, it takes longer to recover these initial properties after damage. Nevertheless, although the healing efficiency increases lower for these higher crosslinked materials, they can withstand higher stresses at a faster rate. This is attributed to the larger number of Diels–Alder bonds that are formed across the fracture surfaces, resulting from the more densely crosslinked network structure. Depending on the application, one will prefer faster relative (e.g., healing efficiency) or absolute recovery (e.g., fracture stress or toughness) of the mechanical properties. The reduction in crosslink density also affects the thermomechanical behavior of the material at higher temperatures. Whether the crosslink density is decreased through adjustments in molecular weight or the r-ratio, the degelation temperature is lowered. Consequently, the material can be reprocessed at lower temperatures, which can be advantageous in terms of energy consumption. However, this decrease in crosslink density also results in reduced temperature resistance, as the material experiences a loss of mechanical and structural stability at lower temperatures. Nevertheless, we present in this paper room temperature healing Diels–Alder polymers that can operate in applications extending up to temperatures of 60 °C or even higher. All Diels–Alder networks shown in this paper also exhibit almost instantaneous partial healing minutes after being damaged. The immediate healing occurs due to the presence of reactive maleimide and furan groups at the fracture surfaces, which are formed upon mechanical breaking of the Diels–Alder bonds (cutting). When the fracture surfaces are brought back into contact, these reactive groups exhibit a strong propensity to react. Furthermore, additional adhesive forces and secondary interactions contribute to the initial healing process. The influence of these non-covalent interactions is more pronounced in the elastomer with lower crosslink density. This immediate healing property holds significant value in various future applications as it facilitates the establishment of excellent contact and ensures the cohesion of fracture surfaces throughout the subsequent healing process. In numerous applications, such as self-healing soft robots, as previously published [ 19 ], damages are self-sealed through the elastic recovery of the polymeric network (e.g., snap back). Combining this elastic recovery with instantaneous healing enables autonomous self-sealing followed by self-healing, eliminating the need for external or human intervention. Moreover, the rapid autonomous sealing and healing process significantly minimizes the risk of incorporating dirt and dust particles into the healed area, thereby reducing any adverse impact on healing performance. One major drawback of the presented Diels–Alder polymers, as well as autonomous intrinsic self-healing polymers in general, is that the fracture surfaces must be promptly recombined after damage occurs. Failure to do so results in the separate recombination of the mechanically formed reactive groups on the two fracture surfaces, leading to a loss of autonomous self-healing capability. Nevertheless, if this happens, long gaping damages in the presented Diels–Alder polymers can still be healed using a heat cool cycle, as demonstrated in [ 40 ]. This is made possible by the thermal dissociation of Diels–Alder bonds during the heating phase, leading to an increased presence of reactive maleimide and furan groups both within the polymer network and on the fracture surfaces. Upon subsequent cooling, the Diels–Alder bonds reform throughout the network and across the fracture surfaces, ultimately restoring the initial properties. The most effective acceleration of healing occurs when the crosslink density is reduced by decreasing the r -ratio. Our findings demonstrate that extremely low r -values of 0.3 enable Diels–Alder polymers to rapidly heal significant damage and fully restore their initial mechanical performance within a mere five hours at room temperature. The reason for this enhanced healing is an increase in mobility that goes together with an increase in excess of reactive furan. This excess of furan extensively increases the rate of reaction. However, an excessive increase in the amount of furan has adverse effects on other material parameters, which can be crucial depending on the specific application. A higher concentration of unbound furan results in a greater number of dangling chains within the network. These dangling chains have a negative impact on the elastomers’ dynamics, as they contribute to increased viscous portion in their viscoelastic behavior. In applications such as soft robotics, rapid actuation and efficient energy storage are of great importance. The presence of a significant viscous contribution in the viscoelastic behavior introduces time-dependent responses that limit the system’s dynamics, such as actuation speed, and lead to higher energy consumption due to viscous losses. Furthermore, when reducing the crosslink density by adjusting the r -value, the spacing between the crosslinks (e.g., the coiled propylene oxide chains) is not increased. As a result, the increase in fracture strain is limited, which in turn, restricts the toughness of the material. By decreasing the crosslink density through an increase in the molecular weight of the monomers, the healing process is also accelerated, but to a lesser extent. Nonetheless, this approach enables the development of Diels–Alder polymers that can fully restore their performance after significant damage within a single day under normal ambient conditions. However, this method does not compromise other mechanical properties. It does not result in an increased presence of dangling chains, thereby minimizing the occurrence of viscous losses. By utilizing longer monomers, the spacer length between the crosslinks is increased. Consequently, the polymer chains between the crosslinks exhibit more extensive coiling, allowing the material to undergo stretching to longer strains. This augmentation in fracture strain also results in a higher toughness of the material. While reducing the crosslink density through the r -ratio has minimal impact on the glass transition temperature ( T g ), decreasing it via the molecular weight of the monomers leads to a decrease in T g . This can also be attributed to the increase in spacer length between the crosslinks."
} | 5,161 |
27805065 | PMC5090987 | pmc | 1,906 | {
"abstract": "This paper studied and realized a flexible nanogenerator based on P(VDF-TrFE) nanofibers and PDMS/MWCNT thin composite membrane, which worked under triboelectric and piezoelectric hybrid mechanisms. The P(VDF-TrFE) nanofibers as a piezoelectric functional layer and a triboelectric friction layer are formed by electrospinning process. In order to improve the performance of triboelectric nanogenerator, the multiwall carbon nanotubes (MWCNT) is doped into PDMS patterned films as the other flexible friction layer to increase the initial capacitance. The flexible nanogenerator is fabricated by low cost MEMS processes. Its output performance is characterized in detail and structural optimization is performed. The device’s output peak-peak voltage, power and power density under triboelectric mechanism are 25 V, 98.56 μW and 1.98 mW/cm 3 under the pressure force of 5 N, respectively. The output peak-peak voltage, power and power density under piezoelectric working principle are 2.5 V, 9.74 μW, and 0.689 mW/cm 3 under the same condition, respectively. We believe that the proposed flexible, biocompatible, lightweight, low cost nanogenerator will supply effective power energy sustainably for wearable devices in practical applications.",
"discussion": "Discussion As mentioned before that the roughness of triboelectric friction layer strongly affected output performance of TENG. It is described that the TENG would achieve ideal and optimum output performance when the friction layers had micro or nano scale structures. Here, we prepared different friction layer structures to investigate the outputs of TPENG. The spin-coated P(VDF-TrFE) film with vapored electrodes, electrospinning film with vapored electrodes and electrospinning film with screen-printed electrodes are fabricated and their AFM surface images are shown in Fig. 7a (1–3), respectively. Spin-coated P(VDF-TrFE) film has relatively lower roughness of 30 nm. While electrospinning film with vapored electrodes has surface roughness of 2000 nm and screen-printed one has 800 nm. Meantime, five types of the other friction layers are produced, including spin-coated PDMS film, PDMS film with micro structures, spin-coated PDMS/MWCNT film and PDMS/MWCNT film with micro structures. Their combinations are shown in Fig. 7b (i–v) and their triboelectric output voltages of NG under the pressure of 5 N and the frequency of 5 Hz are shown in Fig. 7b (vi). Their corresponding voltages are 8.92 V, 14.22 V, 16.64 V, 20.08 V and 30.06 V, respectively. By comparison, it can conclude that PDMS/MWCNT film has better output performance than PDMS film due to faster electron flowing of MWCNTs doping, and micro structures also contributes to increase the output voltages. Meantime, the optimum property happened to the combination of PDMS/MWCNT film with micro structures and electrospinning P(VDF-TrFE) film with screen-printed electrodes. Due to the flexibility of fabricated TPENG, the full-contact state can be easily achieved by fingertip pressing. Thus we utilized it to monitor fingertip motion. The proposed device is shown in Fig. 8(a) and it can light up a commercial LED bulb when the force is applied by finger (snapshot of LED shown in Fig. 8(a) ). When the finger forces are varied from 2 N to 4 N, as shown in Fig. 8(b) , the corresponding output voltages of TPENG are increased from 7 V to 11 V, as shown in Fig. 8(c) . It can also be used as a finger-motion sensor. The proposed hybrid NG has some advantages: a) it produces relatively high output power in small scale matched with human fingertip force, which is capable of lighting LED; b) it can be easily fabricated with lower cost and achieved with flexibility and biocompatibility; c) it utilizes PDMS/MWCNT composite membrane not only to tune the internal resistance of device but also to combine with P(VDF-TrFE) nanofiber to achieve a higher output performance for matching the loading resistance. Compared with the other NGs mentioned above, the achieved hybrid NG is superior in terms of output voltage. In summary, a flexible triboelectric and piezoelectric hybrid NG is proposed and fabricated, which is composed of PDMS/MWCNT membrane, P(VDF-TrFE) nanofibers by electrospinning method, and electrodes by screen-printing process. It is observed that the output performance of TPENG will be improved by MWCNTs doped PDMS membrane as well as patterned pillars micro-structure. Meantime, as an optimum proposal, electrospinning P(VDF-TrFE) nanofibers film with screen-printed electrodes will achieve higher outpower. In the open-circuit, the triboelectric and piezoelectric output peak voltages of TPENG are 25 V and 2.5 V under the pressure of 5 N, respectively. While connecting with vaired external resistances, the triboelectric output power and power density are 98.56 μW and 1.98 mW/cm 3 under the matched resistance of 5 MΩ, and the piezoelectric output power and power density are 9.747 μW and 0.689 mW/cm 3 under the matched resistance of 30 MΩ. This proposed TPENG will be a small-scale, flexible and apposite power promising source for wearable and implantable devices and portable electronic devices."
} | 1,287 |
37735458 | PMC10515101 | pmc | 1,907 | {
"abstract": "Climate change globally endangers certain marine species, but at the same time, such changes may promote species that can tolerate and adapt to varying environmental conditions. Such acclimatization can be accompanied or possibly even be enabled by a host’s microbiome; however, few studies have so far directly addressed this process. Here we show that acute, individual rises in seawater temperature and salinity to sub-lethal levels diminished host fitness of the benthic Aurelia aurita polyp, demonstrated by up to 34% reduced survival rate, shrinking of the animals, and almost halted asexual reproduction. Changes in the fitness of the polyps to environmental stressors coincided with microbiome changes, mainly within the phyla Proteobacteria and Bacteroidota. The absence of bacteria amplified these effects, pointing to the benefit of a balanced microbiota to cope with a changing environment. In a future ocean scenario, mimicked by a combined but milder rise of temperature and salinity, the fitness of polyps was severely less impaired, together with condition-specific changes in the microbiome composition. Our results show that the effects on host fitness correlate with the strength of environmental stress, while salt-conveyed thermotolerance might be involved. Further, a specific, balanced microbiome of A. aurita polyps supports the host’s acclimatization. Microbiomes may provide a means for acclimatization, and microbiome flexibility can be a fundamental strategy for marine animals to adapt to future ocean scenarios and maintain biodiversity and ecosystem functioning. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-023-00266-4.",
"conclusion": "Conclusions The role of metaorganism’s microbiomes in host fitness and ecological interactions is increasingly evident. A. aurita is one of the main contributors to jellyfish blooms that cause enormous ecological and socioeconomic damage, and this study identifies the response of its microbiome to environmental challenges, coinciding with changes in the fitness of the polyps. A microbiome’s presence is beneficial for these animals’ stress tolerance, and microbial community changes correlate with impaired host fitness of A. aurita when the temperature or salinity is increased to sub-lethal levels. In a future ocean scenario, mimicked here by a combined but milder rise of temperature and salinity, the fitness of polyps was less severely impaired, together with condition-specific changes in the microbiome composition. Our results show that the effects on host fitness correlate with the strength of environmental stress, while salt-conveyed thermotolerance might be involved. Microbiome-mediated acclimatization and adaptation may provide a mechanism for hosts besides phenotypic plasticity. Thus, microbiome flexibility can be a fundamental strategy for marine animals to adapt to future ocean scenarios to maintain biodiversity and ecosystem functioning.",
"introduction": "Introduction It is widely recognized that marine ecosystems are under threat [ 1 ]. Ocean acidification, the global increase in sea surface temperature, and changes in salinity, along with overfishing, eutrophication, sedimentation, and pollution, endanger marine species globally [ 2 – 4 ]. At the same time, marine organisms adjust and cope with changing environments [ 5 – 8 ]. Eco-physiological studies typically focus on solitary macroorganisms or interactions among them [ 9 , 10 ], such as competition (e.g., acidification influencing turf algae-kelp interactions) and predation (e.g., warming leading to kelp grazing by range-expanding herbivorous fishes) [ 11 – 13 ]. In contrast, less is known about the impact of ocean climate change on interactions between macro- and microorganisms [ 14 – 16 ]. All multicellular organisms live in an intimate and interdependent association with their microbiome, which includes bacteria, archaea, viruses, fungi, and protists [ 17 – 20 ]. Consequently, animals and plants represent functional biological entities comprising a host and its microbiome, so-called metaorganisms [ 17 ]. Members of a host-associated microbiota have various functions within a metaorganism and display fundamental roles in host health by contributing, for instance, to host development [ 21 ], organ morphogenesis [ 22 ], metabolism [ 23 , 24 ], aging [ 25 ], behavior [ 26 ], and reproduction [ 27 – 29 ]. Microbes may further be essential for macroorganisms living in extreme environmental conditions [ 30 ] and for acclimating and adapting to environmental changes [ 31 – 35 ]. Although the host can also respond to environmental perturbations through phenotypic plasticity [ 7 , 36 , 37 ], microbial-mediated acclimatization has received particular attention. Microbes can play a critical role in controlling host responses to environmental stress through various mechanisms [ 38 ], including metabolites and signaling molecules production [ 39 ], host stress response stimulation [ 40 ], modulation of the host immune response [ 41 ], metabolic cooperation [ 42 ], biofilm formation [ 43 ], and detoxification [ 44 ]. Furthermore, microorganisms have shorter generation times, respond more rapidly and are therefore better suited to persist through these stressors [ 45 ]. In nature, metaorganisms face a diversity of biotic and abiotic stressors that may require an associated microbial community that responds adequately by changing the composition and/or producing protective molecules or modulating host responses [ 46 ]. Consequently, the metaorganism’s fitness may be optimized by altering the composition of its associated microbiota in terms of abundance and/or diversity [ 30 , 34 , 45 , 47 , 48 ]. Such dynamic restructuring of a host’s community through environmental change is known as microbiome flexibility. For instance, microbiome flexibility has been proposed to play a role in the rapid acclimatization of Fungia granulosa after long-term exposure to high-salinity levels [ 49 ], acclimatization of Acropora hyacinthus to increased thermal stress [ 32 ], and the ability of the coral and sponge holobiont to cope with environmental change [ 34 , 50 , 51 ]. [ 30 , 52 ]However, only a few studies have directly addressed how a microbiome enables acclimatization to short-term changes in a local environment or enables host adaptation (e.g., [ 53 – 55 , 2019 #4567]). To provide insights into these processes, our research is focused on the microbiome of the moon jellyfish Aurelia aurita (Linnaeus, 1758) and its involvement in the eco-physiological responses of that host. The scyphozoan A. aurita is a cosmopolitan species documented worldwide in various coastal and shelf sea environments [ 56 ] and is also one of the most frequent blooming jellyfish species [ 56 , 57 ]. Jellyfish blooms, which are significant and sudden increases in jellyfish populations, have been receiving increased attention in the context of climate change [ 58 ]. These blooms can significantly impact marine ecosystems, disrupting the balance of marine food webs and posing threats to biodiversity [ 59 ], since jellies predominantly feed on plankton, including fish eggs and larvae. This leads to competition with other planktivorous organisms and potentially reduces food availability for fish and other marine species [ 60 ]. The predation pressure from jellies can have cascading effects on the abundance and distribution of various marine organisms, affecting ecosystem stability and biodiversity [ 61 ]. With rising sea temperatures and altered ocean currents, climate change is often believed to influence jellyfish population dynamics, facilitating their proliferation. However, the relationship between rising sea temperatures and jellyfish blooms is complex and under scientific discourse. Perceived recent increase in global jellyfish abundance, often seen as a sign of deteriorating oceans, is not conclusively supported by formal analysis of long-term data [ 62 ]. While there has been a slight linear increase in jellyfish populations since the 1970s, this trend is not robust and may be part of a larger cyclical pattern. The strongest observed trend indicates that jellyfish populations undergo significant worldwide oscillations with approximately a 20-year periodicity [ 62 ]. Nevertheless, the implications of jellyfish blooms extend beyond marine ecosystems, affecting human industries [ 56 ]. Commercial fishing, aquaculture, and tourism industries can suffer from jellyfish outbreaks, as these gelatinous creatures can damage fishing gear, clog fishnets, and deter tourists [ 58 ]. A. aurita blooms are a significant concern in marine ecosystems of the Mediterranean Sea [ 63 ], the East Sea [ 56 ], the Gulf of Mexico [ 64 ], and the Atlantic Ocean [ 65 ]. The dynamics of jellyfish blooms, including those of A. aurita , are complex and influenced by various factors, including climate change, nutrient inputs, and predator-prey interactions [ 58 ]. Understanding the drivers and consequences of these blooms is essential for effective management and conservation strategies. By monitoring jellyfish populations and studying the influencing factors that cause such blooms, scientists and policymakers can develop measures to mitigate the negative impacts of jellyfish blooms and promote the health and resilience of marine environments in the face of climate change [ 66 ]. We hypothesize that the microbiome is one of those influencing factors. The life cycle of Aurelia is biphasic and alternates between free-living pelagic medusae and sessile benthic polyps. While only the medusa can sexually reproduce to form planula larvae, the polyps can undergo asexual reproduction through both budding (clonal polyp generation) and strobilation (production of precursor medusa, i.e., ephyra) (Fig. 1 A) [ 67 ]. Environmental factors such as temperature, salinity, or food supply influence both the asexual reproduction of the polyps and medusa ecology, such as somatic growth and sexual maturation [ 68 – 70 ]. A. aurita is highly flexible and can adapt to a wide range of environmental conditions and survive and reproduce between 4 and 28 °C and 15–38 PSU salinity [ 71 – 74 ]. Temperature plays a crucial role in the reproduction of polyps (e.g., [ 70 , 75 – 77 ]. At higher temperatures (20–28 °C), polyps tend to reproduce daughter polyps by budding, while below a certain threshold (< 16 °C), strobilation is triggered to reproduce planktonic ephyrae [ 70 ]. Salinity is also expected to determine the settlement of planulae and subsequent development of polyps [ 56 , 78 ] and may also affect the distribution of polyps in coastal waters (e.g., [ 75 , 79 , 80 ], and the mortality of polyps [ 80 , 81 ]. Understanding the effect of abiotic factors on the survival and reproduction of A. aurita is essential for accurate predictions on the species’ future under climate change and its potential to bloom [ 56 , 82 ]. The composition and structure of the microbial communities associated with A. aurita are well characterized and was shown to be crucial for A. aurita’s fitness (survival, feeding, and growth) and particularly for the generation of offspring [ 29 ]. Bacterial colonizers belong to various phyla, including Proteobacteria, Firmicutes, Bacteroidetes, and Actinobacteria. Some common bacterial genera in the complex and highly diverse A. aurita microbiome include Vibrio , Pseudomonas , Pseudoalteromonas , Alteromonas , Roseobacter , and Ruegeria . The composition of the bacterial communities associated with the moon jellyfish changes with compartment, life stage, and population [ 29 , 83 , 84 ]. Here, we tested how Aurelia ’s microbiome is changing in composition due to acute temperature and salinity rises, thus affecting host fitness. Ultimately, the microbiome of this metaorganism might mediate the acclimatization of A. aurita to climate change, and as a first step in this process, short-term changes were investigated here, which may even help mitigate jellyfish blooms in the future. \n Fig. 1 Study design of the host-fitness experiment. ( A ) The life cycle of Aurelia aurita alternates between pelagic medusae and benthic polyps. The host-fitness experiments were conducted with polyps exposed to increased temperature and salinity. ( B ) Each treatment comprised 96 native (n) or sterile (s) polyps (the latter were kept under sterile conditions throughout the experiment). Control conditions included a salinity of 30 PSU and an ambient temperature of 20 °C. Salinity was increased to 40 PSU (salt) or 37 PSU (fo); temperature was raised to 30 °C (temp) or 25 PSU (fo)",
"discussion": "Discussion Over the past 100 years, the sea surface temperature has increased on average by 0.6 °C [ 98 ]. Moreover, more frequent climatic extremes, like marine heatwaves, result in animal performance declines, mitigation, and local mortality [ 99 – 101 ]. In addition to temperature changes, historical records show that the ocean salinity increased by 4% between 1950 and 2000 [ 102 ]. Many marine species are stenohaline and cannot tolerate a wide fluctuation in the salinity of water; thus, their narrow range of salt tolerance limits their survival, reproduction, and germination [ 103 ]. Salinity can act synergistically or antagonistically with other environmental stressors; for instance, salt stress was reported to cross-protect thermal stress [ 104 – 106 ]. The capability of marine animals to adapt to future ocean scenarios is crucial for maintaining biodiversity and ecosystem functions [ 35 ]. Host-associated microbial communities represent a major factor regulating the host’s response to their external environment [ 31 – 35 ]. Change in the composition of a host’s microbiome (both loss of taxa, shifts in relative abundance, or appearance of novel taxa) has been linked to adapted host fitness as a function of environmental change [ 34 , 107 ]. Correlative observational studies were reported for salt stress in algae [ 108 ], thermal tolerance in sea anemones [ 109 , 110 ], and heat stress in corals and sponges [ 32 , 111 – 113 ]. Consequently, a shift in the microbiome toward a microbial community that supports host fitness could reinforce rapid host acclimatization [ 114 ]. High microbiome flexibility may promotes metaorganismal acclimatization, at the risk of losing putatively essential associates and possibly allowing pathogen invasion [ 34 ]. Low microbiome flexibility in Pocillopora coral was linked to coral disease outbreaks, whereas high microbiome flexibility in Acropora corals was linked to rapid adaptation to escape the disease [ 51 ]. The microbiome of A. aurita benefits the host in adjusting to changes in the environment, such as temperature and salinity, and plays a supportive role for host acclimatization. A diverse and flexible microbiome might assist in maintaining host fitness in a climate-changed ocean. This assumption is supported by our host-fitness experiments conducted with sterile animals, which led to losses in survival, growth, and progeny output under standard (though sterile) conditions, exacerbated under environmental stressors (Figs. 2 and 3 ). We had already demonstrated in a previous study that bacteria function as a protective shield, and their absence impaired host fitness and affected life cycle decisions, resulting in the halt of offspring generation [ 29 ]. These results were verified with sterile polyps under standard conditions, which were also almost completely impaired in asexual reproduction, especially in the release of the ephyrae. Here, we additionally demonstrate that the associated microbiota of A. aurita is changing in composition due to acute, sublethal temperature and salinity increases. This consequently affects host survival, and for those polyps that do survive, growth and asexual reproduction are impaired (Fig. 2 ). Note that energy-intensive fitness parameters related to reproduction were more affected than mere survival. Raising the salinity and temperature to sublethal levels impaired all analyzed fitness traits, leading to 66% reduced survival rates and halting offspring generation. Several studies observed that environmental stressors diminish invertebrate reproduction (i.a., [ 115 , 116 ]); however, only a few studies link microbiome shifts of these hosts to those effects [ 117 ]. We assume that changes in the microbial composition support acclimatization by the host, but drastic changes are associated with loss of microbial function, causing fitness deficits. In a natural setting, exposure to changed salinity or temperature may be more gradual than the abrupt changes applied here, possibly allowing for a slow but steady adaptation of the microbiome and its host. Nevertheless, during heatwaves [ 99 – 101 ], which are expected to increase in severity and frequency due to climate changes, local temperature and salinity changes can be relatively rapid, especially in shallow waters [ 99 – 101 ]. When more moderate increases of salinity and temperature were combined in a future ocean scenario, this resulted in a less impaired fitness than for the more severe, single stressors (Fig. 6 A). This showed that the effects on host fitness correlate with the strength of the environmental stress, while salt-conveyed thermotolerance may also be involved. To our knowledge, salinity-conveyed thermotolerance in marine macroorganisms has only been described in corals [ 23 , 118 ] and data on A. aurita are lacking. Currently, it is unknown whether the bacterial community patterns and the response of the corals to different salinities are causally linked or whether they represent parallel responses of the host and its associated bacteria [ 110 ]. Recent studies propose that osmolytes like floridoside might play a role in adjusting osmotic pressure by counteracting oxidative stress due to combined salinity and heat stress, thereby contributing to stress resilience [ 23 , 118 ]. Similar studies would need to be conducted with A. aurita to gain deeper insights into the salinity-driven thermotolerance of this host. Following analysis of the polyp’s microbiomes, we observed major changes in relative abundance that occurred on phylum, genus, and OTU levels (Figs. 4 , 5 and 6 ). Highly-abundant genera like Alteromonas , Pseudoalteromonas , and Pseudomonas (all Proteobacteria) declined under all environmental stress conditions, while various unclassified genera assigned to Proteobacteria and Bacteroidota increased. Notably, some Vibrio OTUs increased, whereas others decreased (depending on the condition), indicating that reporting findings on genus level only can be imprecise. We demonstrate that approximately a quarter of the detected community members (26%) maintain their relative abundance irrespective of environmental change. Other members may be interchangeable and act as microbiome regulators that maintain a constant microbiome functionality, irrespective of individual members during environmental change. Alternatively, those bacterial members that change in abundance due to environmental conditions may represent microbiome conformers that adapt to their surrounding environment and change the functionality of the complete microbiome [ 51 ]. We noted that the 31 most abundant OTUs all changed their abundance as a result of environmental stress (Fig. 5 B), and the intensity of the environmental stressor drives the degree of community change. Thermal tolerance of animals is assumed to be associated with an increase in Alpha- and Gamma-Proteobacteria [ 35 , 119 , 120 ]. That was not observed in our experiments, as the Proteobacteria phylum decreased under high salt or high temperature conditions. Alpha-Proteobacteria produce protecting antioxidants within the coral holobiont [ 121 ], and Gamma-Proteobacteria representatives inhibited the growth of coral pathogens and provided additional nutrients for the host [ 122 , 123 ]. Clearly, such observations cannot be generalized and extended to different hosts, such as jellyfish. Impaired fitness of A. aurita polyps correlates with complex abundance shifts on the OTU level. Our results suggest that microbial communities play a critical role in affecting the response of animals to ambient temperature and salinity. Recent studies have suggested that the microbiome might be crucial in mediating the resilience of marine organisms, including jellyfish, to climate change stressors [ 58 , 124 ]. A healthy and diverse microbiome could enhance the host’s ability to withstand environmental challenges and promote overall ecosystem stability [ 125 , 126 ]. Consequently, the metaorganism concept should be considered for predicting species’ responses to global climate change. Climate change producing warmer ocean temperatures and increased salinity may enhance jellyfish reproduction and growth rates, leading to population booms. This study’s simulated future ocean scenario demonstrated that jellyfish bloom-causing A. aurita can adapt and survive under changing environmental conditions. The relationship between the host’s microbiome, stress tolerance, and climate change concerning jellyfish blooms is complex and likely involves numerous interacting factors. Understanding these intricate connections is essential for predicting and managing jellyfish blooms in the context of ongoing climate change. By gaining a deeper understanding of these processes, thus implementing the metaorganism concept, researchers can develop more effective strategies for managing and mitigating the impacts of jellyfish blooms in the context of a changing climate."
} | 5,425 |
24265707 | PMC3827042 | pmc | 1,910 | {
"abstract": "Understanding strategies used by animals to explore their landscape is essential to predict how they exploit patchy resources, and consequently how they are likely to respond to changes in resource distribution. Social bees provide a good model for this and, whilst there are published descriptions of their behaviour on initial learning flights close to the colony, it is still unclear how bees find floral resources over hundreds of metres and how these flights become directed foraging trips. We investigated the spatial ecology of exploration by radar tracking bumblebees, and comparing the flight trajectories of bees with differing experience. The bees left the colony within a day or two of eclosion and flew in complex loops of ever-increasing size around the colony, exhibiting Lévy-flight characteristics constituting an optimal searching strategy. This mathematical pattern can be used to predict how animals exploring individually might exploit a patchy landscape. The bees’ groundspeed, maximum displacement from the nest and total distance travelled on a trip increased significantly with experience. More experienced bees flew direct paths, predominantly flying upwind on their outward trips although forage was available in all directions. The flights differed from those of naïve honeybees: they occurred at an earlier age, showed more complex looping, and resulted in earlier returns of pollen to the colony. In summary bumblebees learn to find home and food rapidly, though phases of orientation, learning and searching were not easily separable, suggesting some multi-tasking.",
"introduction": "Introduction Constantly changing temporal and spatial distributions of resources provide complex challenges to animals. Understanding how they explore the landscape can give insight into how they find and selectively exploit these resources efficiently. The impressive abilities of bumblebees and honeybees to exploit a landscape for nectar and pollen for their colony have been investigated in terms of their ability to learn and memorize visually complex routes in pursuit of these rewards, their sophisticated spatial navigational abilities, and their energetic efficiency at reward collection [1] – [5] . However, most deductions have been made without researchers being able to map the complete flight paths taken by bees in real landscapes whilst they learn, search and forage. Instead, researchers have analyzed detailed sections of flight such as flying near the colony entrance [6] – [9] or at flower patches [5] , [10] , [11] , or designed elegant experiments to measure flight characteristics in a simulated foraging environment [12] – [15] . The objective of this study was, for the first time, to map and characterize the flights of bumblebee workers, starting with naïve bees on their first exploratory flights. We examined whether the shapes of these flights indicate an optimal strategy for searching or learning, and analyzed the changes in flight trajectories with experience as they developed into successful foraging flights. Learning About the Colony Entrance When a bumblebee first leaves the colony, she makes short flights which have been described as ‘learning flights’ or ‘orientation flights’ [6] , [8] , [16] . Learning flights in social and solitary Hymenoptera start with circumscribed movements close to the nest, backing away in a series of zigzags or arcs of constant angular velocity, but increasing radius, roughly centered on the entrance hole [16] . During these arcing maneuvers a bee gathers visio-spatial information (and possibly olfactory information) relating to the colony entrance and nearby landmarks to enable a successful return at the end of a trip (reviewed in [8] , [9] ). The description of these ‘learning flights’ has previously focused on the portion visible to an observer or video at the colony [8] , [9] , [17] , and indeed sometimes the flights only cover this short range. However, the bee may fly beyond view and there are no published data on what the bees do next. During the ‘unseen’ portions of these preliminary flights, away from the colony entrance, not only is the bee likely to be learning the landscape, but it is also the bee’s first opportunity to search for flowers and to manipulate flowers to gather nectar and pollen. Since the flights studied here are likely to include learning, orientation, searching and possibly some foraging; then we refer to them as ‘preliminary flights’ rather than ‘orientation flights’ to avoid confusion with previous literature. Exploring the Landscape and Searching for Forage How do bumblebees explore and choose where to forage in a heterogeneous environment? They show constancy to plant species and to forage area over several days [1] , [4] , [18] , but how do they make these choices in the first place? As [19] note with respect to honeybees “little is known about the actual process of searching, because of the difficulty of following individual bees in the field” . Does a bumblebee, leaving the nest for the first time, fly in one direction until suitable forage is reached and then start feeding? Or does the bee make several flights to learn about the vicinity before starting to forage? In exploring, they may use an optimal strategy in terms of the energy and time utilized to find patches of flowers, such as a random walk, or spiral pattern or random looping pattern [20] – [22] . Honeybees fly in distinctive looped search patterns when attempting to locate their hive, after their hive-centred navigation mechanisms have been disrupted [23] , and when attempting to relocate a food source [24] and the tendency for loop sizes to increase over time results in ‘scale-free’ (Lévy flight) characteristics. This strategy is considered optimal in these circumstances because (a) it ensures that the area where the target is expected to lie is searched most intensively [25] , and (b) the bee has a low chance of getting lost by centring the search on a known location. We looked for evidence of these characteristics in bumblebees on their preliminary flights. In contrast to honeybees, bumblebees cannot rely on nest-mates to tell them the location of suitable forage, although they do communicate olfactory information about forage in the area [26] . Characterization of these early bumblebee flights will provide some insight into how they compare with the preliminary flights of honeybees [17] , [27] that can acquire spatially explicit information on forage sources from dancing nest-mates in the colony [28] . We used harmonic radar [29] , [30] to plot the flight trajectories of individual bumblebee workers ( Bombus terrestris L.) with increasing experience, from naïve bees to experienced foragers. We analysed the flight patterns for evidence of learning, searching, and the tracks were superimposed on the landscape for evidence of the start of foraging.",
"discussion": "Discussion We have illustrated the ontogeny of bumblebee flights from naïve explorer to effective forager in a series of animated flight tracks ( Figure S2 ), and shown the bees’ expanding use of space around the colony as they become more experienced ( Figure 2A–C ). The data were collected from bees in one colony in one location and, although this is often the case for such intricate behavioural studies of free flight [14] , [17] , [23] , [27] , inference should still be made with caution. Despite this limitation, it is the only dataset available to our knowledge to show the transition in flight between naïve exploration and foraging in bumblebees. The results indicate that bumblebees are impressively fast at learning the location of home, the location of food and memorizing efficient straight routes between these goals [2] , [14] , [39] . Analyses of simple parameters showed the bees few faster, further, straighter and covered less angular range around the colony as they became more experienced ( Figure 3 ). It is striking that the bumblebees in this study made their first flights within a few days of emerging from pupae (average: 2 days), and started collecting pollen within a day or two of their first flight (also in [40] ). One bumblebee in our study (not tracked) even brought back pollen on her first flight. This speed of development from naïve bee to forager is faster than documented for honeybees, who spend several days in a colony as nurse bees [38] before flying out of the colony, and perform many orientation flights before foraging effectively [27] . In future studies it would be interesting to compare whether the level of stores and demand in the colony alter the speed of progression to foraging. Learning, Exploring and Searching How much time or effort is spent on learning and how much is spent on searching during these preliminary flights? Wei & Dyer [17] hypothesized for honeybees that duration of the ‘learning flight’ portion of a trip (meaning the small scale arcs in front of the colony) relates to investment in learning and stated that the bees fly off in one direction after the ‘learning phase’ suggesting they are then involved in another activity. Biesmeijer & Seeley [41] separate honeybee ‘orientation flights’ from the ‘search trips’ of scouts by using the duration of the excursion and the return of nectar and/or pollen to distinguish between them, which was pragmatic given that they did not have access to trajectory data. However the bumblebees’ complex and gradually expanding flight patterns illustrated here suggest that phases of orientation, learning and searching may not be easily separable, at least for bumblebees. As expected, bumblebees demonstrated arcing behaviour at the beginning and end of their preliminary flights (1 st to 3 rd trips), turning to face the colony at distances of 0–2 m, to gain visio-spatial and olfactory information about the location of the colony and landmarks in the vicinity [6] , [9] , [16] . The radar tracks also show evidence of either arcing or fine scale looping around the colony entrance at distances of up to 10 m ( Table 1 ; Table S1 , Figure S2 ). Whilst the data are not at a high enough resolution to confirm their similarity to the very fine scale arcs seen by an observer, it is indicative of a learning phase at a larger scale than previously reported [9] , [17] . The loops made by a bumblebee on her first flight gradually increase in size and are directed in different azimuthal directions. These first flights have Lévy characteristics that are consistent with the execution of optimal random looping searching strategy [25] , adding to a growing body of evidence showing that several groups of animals use looping search patterns (e.g. honeybees, butterflies, moths and ants [20] – [23] , [42] , [43] . During these looping flights, the bees may not only be searching for the colony, but they may also be memorizing landmarks or scenes [44] to aid relocation of the colony; and they may also be searching for a food source. Our evidence, although limited to 38 tracks, suggests a high degree of multi-tasking within one or two flights (learning where home is, finding food and sampling it within one or two flights). How these neurologically complex tasks are prioritized or combined is an avenue of future study. An anecdotal example of this is the use of the small patch of hogweed, located 30 m from the colony by at least two bumblebees during preliminary flights. This patch of plants was not visible to the human eye from a long distance, although it was next to a wooden structure (the moth trap) so whilst the bees appeared to use the plants as a ‘service station’ to top up with nectar, the structure may also have been acting as a local landmark of known distance and bearing from the colony. Our data suggest that bumblebees gather information on the landscape and forage sources from one or two complex flights covering all four quadrants around the colony, whereas Capaldi et al [27] showed that honeybees made a series simple single looped orientation flights, confined to a narrow sector around the hive. The two datasets are not directly comparable because they were collected in different years, with differing foraging availability, but the contrasts described raise a question of whether the simpler honeybee orientation flights could be shaped by their ability to use shared information on the location of forage resources [26] , [45] . Have naïve honeybees already received information on where to find forage from dancing scouts in the colony? We find little evidence in the literature either for or against this hypothesis. Biesmeijer & Seeley [41] showed that 60% of ‘novice foragers’ engaged in their first few flights rely at least in part on acquiring information from following dances. But, the notion that differences in searching flight patterns may relate to differences in recruitment behaviour is supported by the fact that honeybees adopt optimal Lévy looping searching flights when they are triggered to search without the benefit of shared information [23] , [24] after their hive centered navigational systems have been disrupted or rendered ineffective. Foraging Bumblebees on their third flights showed lengthened and straightened flight paths which become, with further experience, straight vector flights to and from a forage location (similar to those illustrated in [29] , [46] . Of particular interest in this study was that all of the experienced foragers were flying to the south west of the colony, where field beans were flowering. The landscape had patchy forage available in all directions from the colony ( Figure S1 ). This strong bias in direction, with the bees flying in a predominantly upwind direction on their outward flights leads us to hypothesize that the floral olfactory cues may be providing strong directional information guiding their choice of forage over hundreds of metres. Bumblebees are known to share olfactory information in the colony when stimulating other bees to forage, even if they do not communicate location of that forage [26] , [47] , but in honeybees it is generally considered a short range cue if used directly [48] , [49] . Examining the cues used by bumblebees to find forage sources was beyond the scope of this study, but further research to discover the scale over which bumblebees use visual and olfactory cues to make foraging decisions would help predict resource use. Seeley [50] suggests that if bees can utilize floral scent over hundreds of metres, then individual exploration is a very effective way of finding food in the landscape, without recruitment. In summary, tracking bumblebees from their first flights to experienced foragers has shown their capability to quickly learn the location of home and the location of forage resources in a complex landscape, using a series of arcing and multiple looping flights (predicted by [45] ), followed by vector flights. The general characteristics of these looping flights can be used to start to predict how bumblebees will find resources in complex landscapes. Further progress in elucidating how bees learn to utilize their landscape would be made by tracking individuals over sequential flights and simultaneously monitoring behaviour of individuals in the colony (sic [41] ), together with repeated studies on different colonies with differing recruitment behaviour. These are required to tease apart the mechanisms of the apparent multi-tasking as the bees learn and explore the landscape."
} | 3,884 |
36161026 | PMC9495442 | pmc | 1,911 | {
"abstract": "Microbes, especially abundant microbes in bulk soils, form multiple ecosystem functions, which is relatively well studied. However, the role of rhizosphere microbes, especially rhizosphere rare taxa vs. rhizosphere abundant taxa in regulating the element circling, multifunctionality, aboveground net primary productivity (ANPP) and the trade-offs of multiple functions remains largely unknown. Here, we compared the multiple ecosystem functions, the structure and function of rhizosphere soil bacterial and fungal subcommunities (locally rare, locally abundant, regionally rare, regionally abundant, and entire), and the role of subcommunities in the Zea mays and Sophora davidii sole and Z. mays/S. davidii intercropping ecosystems in subtropical China. Results showed that intercropping altered multiple ecosystem functions individually and simultaneously. Intercropped Z. mays significantly decreased the trade-off intensity compared to sole Z. mays , the trade-off intensity under intercropped S. davidii was significantly higher than under intercropped Z. mays . The beta diversities of bacterial and fungal communities, and fungal functions in each subcommunity significantly differed among groups. Network analysis showed intercropping increased the complexity and positive links of rare bacteria in Z. mays rhizosphere, but decreased the complexity and positive links of rare bacteria in S. davidii rhizosphere and the complexity and positive links of fungi in both intercropped plants rhizosphere. Mantel test showed significant changes in species of locally rare bacteria were most strongly related to nitrogen-cycling multifunctionality, ANPP and trade-offs intensity, significant changes in species of locally rare fungus were most strongly related to carbon-cycling multifunctionality, phosphorus-cycling multifunctionality, and average ecosystem multifunctionality. This research highlights the potential and role of rare rhizosphere microorganisms in predicting and regulating system functions, productivity, and trade-offs.",
"conclusion": "Conclusion This study investigated the rhizosphere abundant and rare bacteria and fungi, rhizosphere C/N/P-cycling multifunctionality, ecosystem multifunctionality, aboveground net primary productivity and trade-offs in the Z. mays and S. davidii sole and Z. mays/S. davidii intercropping ecosystems. Results demonstrated that intercropping altered multiple ecosystem functions individually and simultaneously. Intercropped Z. mays significantly decreased the trade-off intensity compared to sole Z. mays , the trade-off intensity under intercropped S. davidii was significantly higher than under intercropped Z. mays . Moreover, both rhizosphere abundant and rare could predict and might affect rhizosphere elements circling, multifunctionality, aboveground productivity and trade-offs, whereas, the significant changes in species of locally rare microbes were the best predictor of rhizosphere elements circling, multifunctionality, aboveground productivity and trade-offs. We thus ascribe great importance to the rare species. Indeed, our results may help in driving a high functionality by directing future efforts on collection, conservation and manipulation of rhizosphere rare species. Further research with more ecosystems and operation of rare microbe combinations will facilitate to maintain a higher ecosystem function and a better understanding of cause-and-effect mechanisms.",
"introduction": "Introduction Losses in taxonomic and functional diversities are pervasively at both global and local scales ( Radchuk et al., 2016 ; Baldrighi et al., 2017 ; Huang et al., 2019 ). This trend is predicted to continue over this century ( Huang et al., 2019 ), and raises increasing concerns on the influence of biodiversity on ecosystem functions ( Karolína et al., 2014 ; Radchuk et al., 2016 ; Baldrighi et al., 2017 ). Unlike flora ( Niklaus et al., 2017 ; Wang et al., 2020b ) or fauna ( Gagic et al., 2015 ; Tonin et al., 2018 ) both which have been well studied in the studies of the relationship between biodiversity and ecosystem function, soil microorganisms represent the richest and highest diverse life ( Delgado-Baquerizo et al., 2017 ; Luo et al., 2018 ; Chen et al., 2020a ), nevertheless, the relationships between soil microorganisms and ecosystem functions are not fully understood ( Delgado-Baquerizo et al., 2017 ). Previous studies have found that soil microbes contribute to driving multiple ecosystem functions simultaneously (MF) ( Fanin et al., 2017 ; Jiao et al., 2018 ; Chen et al., 2020a ), such as carbon and nutrient cycling ( Delgado-Baquerizo et al., 2017 ; Jiao et al., 2018 ) and productivity ( George et al., 2019 ; Zheng et al., 2019 ). However, most studies have centered on the temperate communities ( Guerra et al., 2020 ) and the bulk soil communities ( Luo et al., 2018 ; Wen et al., 2020 ). In contrast, subtropical data on soil microbial diversity and ecosystem functions are particularly scant ( Guerra et al., 2020 ). Furthermore, the rhizosphere is a true hot point of plant-microbial-soil interactions, and the rhizosphere microbial communities are clearly distinctive from the surrounding bulk soil’s ( Fan et al., 2017 ; Jiao et al., 2018 ), because plants filter microbes for special structures and function ( Del Galdo Jos et al., 2003 ), to benefit their growth, nutrition ( Mendes et al., 2014 ), and function ( Lu et al., 2018 ). Notwithstanding, how rhizosphere bacterial and fungal communities participate in the nutrient cycling of rhizosphere soil, and drive the rhizosphere multiple functions, aboveground primary productivity and trade-offs among functions remains less explicitly acknowledged. Although bacteria and fungi are the most frequently studied communities in soil biodiversity and ecosystem function research ( Guerra et al., 2020 ), their diversity effects on ecosystem functions have not yet been fully explored for the rhizosphere soils. Besides, the majority of studies were based on abundant taxa of microbiome ( Chen et al., 2020a ; Liang et al., 2020 ). A large number of low abundance taxa (rare microbial taxa) were deleted before analyzing data of microbiome ( Lynch and Neufeld, 2015 ; Jousset et al., 2017 ; Chen et al., 2020a ). However, there is a theoretical contradiction for this deletion. On the one hand, the influential mass ratio hypothesis states ( Grime, 1998 ) that the influences of species/functions on an ecosystem function/process is in proportion to their biomass/relative abundance ( Karolína et al., 2014 ; Bagousse-Pingueta et al., 2019 ), in this sense, it may be reasonable to focus only on the role of abundant taxa. On the other hand, the most of taxa in almost all ecosystems are low abundance (rare) ( David et al., 2013 ; Jain et al., 2014 ; Lynch and Neufeld, 2015 ), and dominant species account for most of the total abundance ( Dee et al., 2019 ), resulting in high diversity in the rare subcommunity ( Jain et al., 2014 ; Lynch and Neufeld, 2015 ) and low diversity in the dominant subcommunity. Generally, the selection effect [selection/occurrence of some particular species (or functions)/identity effects] and complementary effect (niche partitioning/different resource utilization/facilitation) are recognized mechanisms to interpret the role of biodiversity in shaping ecosystem functions ( Loreau, 2000 ; Mensah et al., 2020 ; Ding and Wang, 2021 ). Since high species diversity should increase the likelihood of the selection effect ( Mensah et al., 2020 ) and/or the complementary effect ( Mensah et al., 2018 ), rare subcommunities or their functions is inferred to contribute in greater proportion to a given ecosystem function than the dominant. This contradicts the mass ratio hypothesis. Besides, the deletion changed substantially the profile of rare taxa ( Chen et al., 2020a ). Unfortunately, rare species are often more sensitive ( Jain et al., 2014 ; Guo et al., 2020 ; Zhou et al., 2020b ) and vulnerable to vanish firstly ( David et al., 2013 ). Therefore, if the above inference is true, conservation and use of rare species will be more imperative than those of the abundance species to maintain ecosystem functions. However, the ecological role of rare species is poorly known ( Säterberg et al., 2019 ). Although the complementarity and selection effects contribute to MF, their relative contributions remains controversial ( Mensah et al., 2020 ). Moreover, the relative importance of these two effects of rare rhizosphere microbes in explaining rhizosphere functions, plant productivity and trade-offs are understudied. In this study, our hypothesis was that rare rhizosphere taxa contributed to multiple functions and trade-offs in larger proportion than the abundant taxa did. To examine this hypothesis, we characterized the rhizosphere abundant and rare bacteria and fungi, rhizosphere C/N/P-cycling multifunctionality, ecosystem multifunctionality, aboveground net primary productivity and trade-offs, as well as explored how bacteria and fungi with different abundance differently link to the multiple rhizosphere functions, aboveground net primary productivity and trade-offs in subtropical sole and intercropping systems in China.",
"discussion": "Discussion The selection effect was stronger than the complementary effect on multiple functions In agreement with findings from macroecology ( Karolína et al., 2014 ; Baldrighi et al., 2017 ; Fanin et al., 2017 ; Bagousse-Pingueta et al., 2019 ) and bulk soils ( Bastida et al., 2016 ; Delgado-Baquerizo et al., 2016 ; Mori et al., 2016 ), the diversities of the microbes in the rhizosphere played an essential role in driving multiple functions ( Supplementary Table 3 ). This indicated that maintaining functioning needs protection and use of rhizosphere microbial diversity. Furthermore, the positive effects dominated CCMF, but the negative effects dominated AEMF. The positive and negative effects on NCMF and ANPP were detected. The negative effects (negative correlation, 15/17) were more frequent than the positive effects (positive correlation, 2/17) in driving NCMF ( Supplementary Table 3 ), indicating that the negative effects dominated the NCMF. However, the positive effects (positive correlation, 8/9) were more frequent than the negative effects (negative correlation, 1/9) in driving ANPP ( Supplementary Table 3 ), suggested that the positive effects dominated ANPP. Here, we showed for the first time that ( Figure 5 ) the significant changes in rhizosphere species and functions had a stronger relationship with CCMF, NCMF, PCMF, AEMF, ANPP, and trade-offs than the significantly diversity indices, indicating that the selection effect played a chief role in driving multiple functions. This finding was different from the findings of studies in the macroecology ( Tonin et al., 2018 ; Woodcock et al., 2019 ; Chun et al., 2020 ) and bulk soils ( Wen et al., 2020 ), which indicated that the complementarity effect played a chief role in driving multiple functions whereas the selection effect had a limited role ( Emily et al., 2018 ; Li et al., 2020a ). This inconsistency indicated that the selection effect in the rhizosphere could not be ignored and might be a novel way to regulate productivity. Our result could be explained by the following potential causes: (1) The rhizosphere community was under strong selective pressure for special microbes ( Poole et al., 2018 ) based on crucial functions associated with the metabolism of N and P, which are associated with plant growth promotion and nutrition ( Mendes et al., 2014 ), therefore, plant productivity relied on the microbiome for the uptake of nutrients ( Chen et al., 2020b ). (2) The rhizosphere microbes mediated the root exudation ( Korenblum et al., 2020 ), changed the nutrient supply and absorption of plants and rhizosphere NCMF. In turn, rhizosphere NCMF negatively modulated plant productivity (Spearman r = −0.70, p < 0.001, Supplementary Figure 2 ). Collectively, the strong selectivity of plants to rhizosphere microorganisms ( Philippot et al., 2013 ) determined which microorganisms or functions appear; therefore, the multiple functions of the rhizosphere and ANPP were more dependent on the selection effect than the complementary effect, indicating that community assembly processes might determine the diversity and composition of community, and it was the result of these processes that determined how selection and complementary effects occur ( Leibold et al., 2017 ). Therefore, our study provided new insight into the rhizosphere that differ from bulk soils. Rare rhizosphere taxa might contribute over proportionately to multiple functions Consistent with our previous findings from bulk soils ( Ding and Wang, 2021 ), distinct microbes (bacteria or fungi) dominated the given functions. In this study, fungi likely dominated the CCMF and PCMF, bacteria likely dominated the NCMF and ANPP ( Figure 6 ). This might be due to bacteria and fungus had different metabolic niches ( Ding and Wang, 2021 ). For instance, fungus rather than most bacteria could secrete lignin-degrading enzymes ( Ding et al., 2020a ), enhanced rhizodeposition, and suppressed the organic matter degradation ( Zhou et al., 2020a ). Fungi dominated P uptake in symbiotic plants ( Wang et al., 2020a ; Jiang et al., 2021 ). The denitrification, mineralization, and assimilation of N were mainly driven by bacteria ( Starke et al., 2016 ; Li et al., 2019 ; Ding and Wang, 2021 ). N is the major nutrient limiting plant growth ( Moreau et al., 2019 ; Wang et al., 2020a ), thus, if bacteria dominated the NCMF, bacteria were expected to dominate ANPP ( Figure 6 ). Furthermore, this was the first study to report on locally rare bacteria and fungi were overlooked keystone taxa shaping ecosystem functions and trade-offs. These findings supported established results ( Radchuk et al., 2016 ; Jousset et al., 2017 ; Dee et al., 2019 ; Zhang et al., 2019 ; Chen et al., 2020a ). Rare taxa could act as keystone species through several mechanisms. We grouped the potential mechanisms accounting for the greater contribution of the rare species to ecosystem functions into three pathways that may operate simultaneously: (1) Rare species occupied the majority of species in ecosystems ( Jain et al., 2014 ; Lynch and Neufeld, 2015 ; Jousset et al., 2017 ; Zhang et al., 2019 ; Guo et al., 2020 ) and had extremely high diversity ( Mo et al., 2018 ). Rare taxa had more stronger impact on the multifaceted diversity of the community than abundant taxa ( Zhou et al., 2019a ). In this study, the species richness of locally rare bacteria was 433- to 677-fold that of locally abundant bacteria, the species richness of regionally rare bacteria was six–seven fold that of regionally abundant bacteria, the species richness of locally rare fungus was 20- to 52-fold that of locally abundant fungus, the species richness of regionally rare bacteria was 1.2- to 1.8- fold that of regionally abundant fungus ( Supplementary Table 2 ). On the one hand, rare species provided insurance effects ( Jousset et al., 2017 ; Chen et al., 2020a ) as implied by the insurance hypothesis ( Jiao et al., 2017 ), and recruitment from the persistent rare microbial seed bank provided a broad reservoir of ecological function ( Lynch and Neufeld, 2015 ). On the other hand, most distinct traits combinations were supported predominantly by rare species ( David et al., 2013 ). Similar observations were detected in our study. Both locally and regionally, the function richness of rare bacteria and fungus occupied 57–100% of that of whole bacteria and fungus ( Supplementary Table 2 ). According to the hypothesis of complementary and selection hypotheses ( Mensah et al., 2020 ), high diversity elevated the chance of rare taxa to contributing to ecosystem functionality; in this sense, rare species had the potential to change an ecosystem’s multifunctionality via the enhancement of biodiversity ( Angelini et al., 2015 ; Jousset et al., 2017 ). (2) Many studies had supported that “being different” was crucial for the influence of rare taxa/. In plant communities, rare species had a higher effect on ecosystem functioning because the of the rare taxa individual mass is higher than that of the abundant taxa ( Radchuk et al., 2016 ). Similar effects had been observed in decomposition systems ( Guo et al., 2020 ) and microbial system ( Wei et al., 2019 ). Essentially, microbes changed system functions through their metabolic functions ( Ding and Wang, 2021 ). In this study, there were obvious differences in functions ( Figures 5C,D ) and functional diversity ( Supplementary Table 2 ) of the rare versus abundant species, suggested that rare species had the potential to act significant roles in ecosystem functioning via providing different functions. (3) Affecting the interactions. The effect of some taxa on ecosystem functions were not independent from their interactions with other taxa ( Wagg et al., 2019 ). The disappearance of rare caused an obvious or bad alteration in the composition or function of the community ( Nannipieri et al., 2020 ). In contrast, the occurrence of rare could have a good effect on the community ( Xue et al., 2020 ). They could create circumstances that supported the co-occurrence of high densities of different functional organisms, thereby, enhancing MF ( Angelini et al., 2015 ). Rare microbes could also heighten the role of abundant microbes ( Jousset et al., 2017 ). Rare dissimilated low content compounds into materials needed by other microbes or synthesized effective bioactive compounds ( Harrison et al., 2021 ). The rare could reshape rhizosphere community, thereby promote crop growth ( Li et al., 2020b ). Rare taxa could affect species interactions ( Xue et al., 2020 ). In this study, rare contributed more to the microbe’s positive interaction than abundant ( Table 1 and Supplementary Figures 4 , 5 ). Higher frequent facilitations than competitions possibly yielded the complementarity effects ( Ding and Wang, 2021 ). Therefore, rare contributed to ecosystem functions through species interactions ( Jousset et al., 2017 ; Dee et al., 2019 ). Since rare species are most vulnerable to be lost ( Dee et al., 2019 ) and are largely unexplored, we ascribed great importance to the rare species and suggested to optimize their taxa for maintaining a high ecosystem functionality. Rare rhizosphere taxa might contribute to trade-offs of multiple functions Monoculture had higher ANPP than intercropping, this implied that monoculture brought higher benefits than intercropping, which was consistent with the recent findings in tropics ( Grass et al., 2020 ). However, intercropping also shifted the ANPP-CCMF relationship from being none to being positive for Z. mays ( Figures 2B,C ), shifted the ANPP-NCMF relationship from being negative to being positive for S. davidii ( Figures 2D,E ). Intercropping decreased the trade-offs intensity compared to sole for Z. mays ( Figure 3E ). MF and yields were not always synergistic ( Supplementary Figure 2 ), confirming recent findings ( Garland et al., 2021 ). Our study also suggested, for the first time, that the locally rare bacteria species were most strongly related to the trade-offs of multiple functions, and indicated that the trade-offs would likely be reduced by optimizing the taxa of locally rare bacteria."
} | 4,869 |
25309916 | PMC4189764 | pmc | 1,912 | {
"abstract": "The increased demand and consumption of fossil fuels have raised interest in finding renewable energy sources throughout the globe. Much focus has been placed on optimizing microorganisms and primarily microalgae, to efficiently produce compounds that can substitute for fossil fuels. However, the path to achieving economic feasibility is likely to require strain optimization through using available tools and technologies in the fields of systems and synthetic biology. Such approaches invoke a deep understanding of the metabolic networks of the organisms and their genomic and proteomic profiles. The advent of next generation sequencing and other high throughput methods has led to a major increase in availability of biological data. Integration of such disparate data can help define the emergent metabolic system properties, which is of crucial importance in addressing biofuel production optimization. Herein, we review major computational tools and approaches developed and used in order to potentially identify target genes, pathways, and reactions of particular interest to biofuel production in algae. As the use of these tools and approaches has not been fully implemented in algal biofuel research, the aim of this review is to highlight the potential utility of these resources toward their future implementation in algal research.",
"conclusion": "9. Conclusion The above reviewed computational tools and approaches in conjunction with the high interests of the scientific community in synthetic biology offer a new perspective in accelerating biofuel production and microalgal optimization research. The pressing economical and environmental challenges of the use of fossil fuels will furthermore lead to a positive selective pressure towards the use of these strategies aiming at the optimization of biofuel producing strains. A large set of biofuel types can serve as alternative energy sources which currently include ethanol, n-butanol, iso-butanol, short chain alcohols, short chain alkanes, biodiesel (FAMEs), and fatty alcohols. These tools and applications are promising yet much more optimizations need to be achieved in order for biofuel production to compete with available fossil fuels. With the “green revolution” and the more environmentally conscious population, we expect this field to expand significantly in the coming years, building on the available resources for systems and synthetic biology and achieving the generation of strains optimized for biofuel production.",
"introduction": "1. Introduction Biofuel production from microalgae has been receiving attention as an alternative energy source due to its high biomass productivity and minimal land resource requirement. However, there is still a need to improve algal productivity in order to make algal-based bioproducts economically viable. Metabolic network reconstructions of algae can offer insight into genetic modification strategies that can be used to improve microalgal strains. A large number of computational tools have been developed, allowing a range of analyses and predictions, based on genetic and thermodynamic constraints embedded in in the network, to identify bioengineering strategies that can result in enhanced biofuel production of the engineered algal strain. Although a fair number of algal genomes have been fully sequenced, only a few metabolic network models have been reconstructed for these species, hampering algal bioengineering progress [ 1 ]. The utilities of metabolic network models span over several types of applications. On one hand, these models help contextualizing high throughput experimental data, for example, integrating gene expression data with metabolic pathways under different growth conditions [ 2 ]. Metabolic models can also unveil targets for metabolic engineering approaches, which can lead to increased production of target metabolites [ 3 ] or preferentially increase respiration rates [ 4 ]. On the other hand, with the availability of large and diverse biological data sets, metabolic network models can provide a framework to integrate such omics data and allow the formulation and testing of downstream hypotheses. Last, cross-species metabolic comparison represents one more utility of such reconstructions through which identification of differentially activated metabolic pathways can be achieved among other comparative analyses [ 5 ]. Herein we review the reconstruction of metabolic network models and major computational tools and pipelines that hold the potential to contribute to the optimization of algal strains for biofuel production. We describe a number of tools that remain mostly unused by the algal research community. This is reflected from the observation that only 7 algal-based PGDBs (Pathway/Genome Database) are available in Pathway Tools [ 6 ], while approximately 3,500 PGDBs are available for nonalgal species (please see below for more information). The use of some of the herein discussed tools, already applied to the multitude nonalgal organisms, ranging from human to E. coli , provides strategies for algal biofuels optimization with major enhancement potential."
} | 1,278 |
29761099 | PMC5936788 | pmc | 1,913 | {
"abstract": "Colorful super anti-wetting coatings are receiving growing attention, but are challenging to invent. Here, we report a general method for preparing mechanically robust and thermally stable colorful superamphiphobic coatings. A composite of palygorskite (PAL) nanorods and iron oxide red (IOR) was prepared by solid-state grinding or hydrothermal reaction, which was then modified by hydrolytic condensation of silanes to form a suspension. Superamphiphobic coatings were prepared by spray-coating the suspension onto substrates. The superamphiphobicity depends upon the surface microstructure and chemical composition, which are controllable by the PAL/IOR concentration and the solid-state grinding time. The colorful coatings show excellent superamphiphobicity with high contact angles and low sliding angles for water and various organic liquids of low surface tension, e.g., toluene and n -decane. The coatings also feature high mechanical, chemical and thermal stability, which is superior to all the reported colorful super anti-wetting coatings. Moreover, superamphiphobic coatings of different colors can be prepared via the same procedure using the other metal oxides instead of IOR. We believe the colorful superamphiphobic coatings may find applications in many fields like anti-climbing of oils and restoration of cultural relics, as the coatings are applicable onto various substrates.",
"conclusion": "Conclusion In summary, a general method has been developed for the preparation of mechanically robust and thermally stable colorful superamphiphobic coatings. The superamphiphobicity is closely related to the surface microstructure and chemical composition of the coatings, which can be regulated by the concentrations of PAL/IOR, PFDTES, and TEOS and the solid-state grinding time. The colorful superamphiphobic coatings feature high contact angles and low sliding angles for water and various organic liquids with surface tension as low as 23.8 mN m −1 . In comparison with the existing colorful super anti-wetting coatings, the coatings also feature high mechanical, chemical and thermal stability up to 400°C. Moreover, superamphiphobic coatings in different colors can be prepared via the same procedure, and are applicable onto various substrates. We believe that the findings here will shed light on the design of novel colorful super anti-wetting coatings. The colorful superamphiphobic coatings may find applications in many fields, such as anti-climbing of oils and restoration of cultural relics.",
"introduction": "Introduction Inspired by the unique water repellency of lotus leaves and legs of water striders (Neinhuis and Barthlott, 1997 ; Gao and Jiang, 2004 ), superhydrophobic coatings have received great attention. Superhydrophobic coatings have potential applications in many fields such as self-cleaning, anti-corrosion and oil/water separation, etc. (Liu et al., 2010 , 2013 ; Bhushan and Jung, 2011 ; Chu et al., 2015 ; Li et al., 2016 ). However, superhydrophobic coatings can be easily wetted by liquids of low surface tension, e.g., most of organic liquids and surfactant solutions, which largely limit their applications. Different from superhydrophobic coatings, superamphiphobic coatings repel both water and liquids of low surface tension (Lu et al., 2012 ; Schlaich et al., 2016 ). However, preparation of superamphiphobic coatings is much more difficult theoretically and technically than that of the superhydrophobic ones because of much lower surface tension of organic liquids than water (Steele et al., 2008 ; Bellanger et al., 2014 ; Chu and Seeger, 2014 ). For example, the surface tension of water is 72.8 mN m −1 , whereas the surface tensions of toluene and n -dodecane are 28.4 and 25.4 mN m −1 , respectively. Great efforts have been made for preparing superamphiphobic coatings (Zimmermann et al., 2008 ; Liu et al., 2010 ; Cho et al., 2016 ; Papadopoulos et al., 2016 ; Qu et al., 2016 ). However, only a few studies reported superamphiphobic coatings from which the liquids of low surface tension (< 27.5 mN m −1 ) could roll off. For the design of superamphiphobic coatings with low sliding angles (SAs), some special micro-/nanostructures (e.g., reentrant structures, candle soots and silicone nanofilaments) (Tuteja et al., 2007 ; Zhang and Seeger, 2011b ; Deng et al., 2012 ; Liu and Kim, 2014 ; Wong et al., 2017 ) and materials of very low surface energy (e.g., fluoroPOSS; Liu and Kim, 2014 ) have been invented. Cohen et al. designed superamphiphobic coatings by the combination of reentrant structures and fluoroPOSS (Liu and Kim, 2014 ), which have promoted development of superamphiphobic coatings. Zhang and Seeger invented transparent superamphiphobic coatings based on silicone nanofilaments, which showed low SA even for n -decane (Zhang and Seeger, 2011b ). On the other hand, the mechanical stability of most of superhydrophobic and superamphiphobic coatings is low, which makes them far from practical applications (Tian et al., 2016 ). High mechanical stability is a hot issue in the field, especially for superamphiphobic coatings. We have prepared durable and self-healing superamphiphobic coatings repellent even to hot liquids by the combination of silanes and palygorskite (PAL) (Li and Zhang, 2016 ). Furthermore, the methods for preparing superamphiphobic coatings are often complicated, expensive and confined to specific substrates. A simple, cost-effective and general method remains to be explored. For practical applications, colorful super anti-wetting coatings are receiving growing attention. Colorful superhydrophobic coatings have been prepared by using structural color (Sato et al., 2008 ) or coloring species like pigments and dyes (Ogihara et al., 2011 ; Soler et al., 2011 ; Li et al., 2015 ). Gu et al. invented colorful superhydrophobic inverse opal films by the combination of structural color and lotus effect (Gu et al., 2003 ). We fabricated colorful superhydrophobic coatings by modification of Maya blue-like pigments with silanes (Reinen and Lindner, 1999 ; Zhang et al., 2016b ). Maya blue, composed of PAL and indigo, has remarkable stability to acidic and alkaline corrosion but poor thermal stability (Doménech et al., 2007 ; Zhang et al., 2016a ). However, colorful superamphiphobic coatings are rare because it is difficult to give consideration to the color of the coatings in designing the micro-/nanostructures and surface chemical composition of superamphiphobic coatings. Dong et al. reported the first colorful superamphiphobic coatings using Maya blue-like pigments and fluorinated polysiloxane (fluoroPOS) (Dong et al., 2017 ). Although encouraging results have been obtained in the field of colorful super anti-wetting coatings, their thermal stability is often low owing to the fact that organic coloring species are used. For example, the colorful superamphiphobic coatings from Maya blue-like pigments and fluoroPOS faded after being kept at temperature over 150°C for 1 h (Dong et al., 2017 ). Replacing the organic coloring species with metal oxides should be a possible way to enhance the thermal stability. On the other hand, the mechanical stability of colorful super anti-wetting coatings remains to be improved. Here, we report a general method for preparing mechanically robust and thermally stable colorful superamphiphobic coatings by the combination of PAL nanorods, iron oxide red (IOR, or the other iron oxides and metal oxides) and silanes. Iron oxides are environmental friendly, low-cost, and have much higher thermal stability than organic coloring species (Tian et al., 2017 ). First, a composite was prepared by solid-state grinding of PAL and IOR or by hydrothermal reaction of PAL with Fe(III) salts. Then, the PAL/IOR@fluoroPOS suspension was prepared by hydrolytic condensation of 1H,1H,2H,2H -perfluorodecyltriethoxysilane (PFDTES) and tetraethoxysilane (TEOS) in the presence of the PAL/IOR composite. Finally, the PAL/IOR@fluoroPOS coatings were prepared by spray-coating the suspension onto substrates. The colorful coatings show excellent superamphiphobicity with high contact angles (CAs) and low SAs for water and various organic liquids of low surface tension. The colorful coatings also feature high mechanical, chemical and thermal stability up to 400°C."
} | 2,079 |
27084981 | null | s2 | 1,914 | {
"abstract": "Advances in synthetic biology to build microbes with defined and controllable properties are enabling new approaches to design and program multispecies communities. This emerging field of synthetic ecology will be important for many areas of biotechnology, bioenergy and bioremediation. This endeavor draws upon knowledge from synthetic biology, systems biology, microbial ecology and evolution. Fully realizing the potential of this discipline requires the development of new strategies to control the intercellular interactions, spatiotemporal coordination, robustness, stability and biocontainment of synthetic microbial communities. Here, we review recent experimental, analytical and computational advances to study and build multi-species microbial communities with defined functions and behavior for various applications. We also highlight outstanding challenges and future directions to advance this field."
} | 228 |
21949755 | PMC3176771 | pmc | 1,915 | {
"abstract": "Climate change scenarios suggest an increase in tropical ocean temperature by 1–3°C by 2099, potentially killing many coral reefs. But Arabian/Persian Gulf corals already exist in this future thermal environment predicted for most tropical reefs and survived severe bleaching in 2010, one of the hottest years on record. Exposure to 33–35°C was on average twice as long as in non-bleaching years. Gulf corals bleached after exposure to temperatures above 34°C for a total of 8 weeks of which 3 weeks were above 35°C. This is more heat than any other corals can survive, providing an insight into the present limits of holobiont adaptation. We show that average temperatures as well as heat-waves in the Gulf have been increasing, that coral population levels will fluctuate strongly, and reef-building capability will be compromised. This, in combination with ocean acidification and significant local threats posed by rampant coastal development puts even these most heat-adapted corals at risk. WWF considers the Gulf ecoregion as “critically endangered”. We argue here that Gulf corals should be considered for assisted migration to the tropical Indo-Pacific. This would have the double benefit of avoiding local extinction of the world's most heat-adapted holobionts while at the same time introducing their genetic information to populations naïve to such extremes, potentially assisting their survival. Thus, the heat-adaptation acquired by Gulf corals over 6 k, could benefit tropical Indo-Pacific corals who have <100 y until they will experience a similarly harsh climate. Population models suggest that the heat-adapted corals could become dominant on tropical reefs within ∼20 years.",
"introduction": "Introduction The all-important symbiosis between corals and symbiotic dinoflagellates is destabilized by environmental stress, particularly elevated sea temperatures [1] , [2] . Corals turn white, i.e. bleach, as they lose their symbionts and become vulnerable to increased rates of disease and mortality. In combination with local stresses, increasing global stress is driving a global decline in coral populations [3] , [4] . Outbreaks of mass coral bleaching and mortality are finely tuned (adapted) to the local climatology and have been used with climate projections to explore how coral communities are likely to change as the world warms [1] . The thermal threshold for mass coral bleaching and mortality across the majority of Indo-Pacific and Caribbean habitats lies around 30–32°C [1] , [5] . Some populations of corals, however, experience extreme temperatures and are valuable in terms of understanding the outer envelope of genetic responses to climate change. Here, we present data on corals within the hottest sea with abundant coral on the planet (Arabian/Persian Gulf) that may provide thermally tolerant species and varieties for other tropical habitats as they warm. These corals survive levels of thermal stress that would probably kill even the most tolerant corals growing elsewhere. Here we analyze models of population dynamics to test whether under extreme disturbance scenarios, as are presently experienced by Gulf corals and expected on tropical coral reefs by 2099 [6] , coral populations will increasingly fluctuate and will consequently be forced to abandon the present “reef-building” mode. We also investigate whether, if moved into the tropical Indo-Pacific by currents or assisted migration, today's Gulf corals would have the potential to be dominant and how long replacement of presently resident genotypes would take, given future disturbance scenarios.",
"discussion": "Discussion The corals of the Arabian/Persian Gulf reveal the highest documented temperature tolerance of the coral holobiont. Tropical sea temperatures are projected to increase by at least 1–3°C above today's temperatures by the end of the century, resulting in temperatures throughout the tropics that are comparable to the Gulf today [6] . While Gulf corals were subjected to this climate for ∼6 ky [13] , other corals will have <100 yrs to adapt. Current evidence suggests that adaptation within coral populations at these rates is unlikely to occur naturally, notwithstanding whether heat-resistance is due to adaptation of the symbionts or the coral animal. Gulf corals already have heat-tolerant symbiont populations (primarily clades C and D, [14] , [15] ). Human intervention through assisted colonization may therefore represent an effective management tool [16] . Gulf corals appear likely to withstand the next hundred years of climate change in the tropical Indo-Pacific and are increasingly endangered in their current range over both short [17] and long terms [11] . Rampant coastal development has already physically removed many reefs in the region, lethal red tides have become increasingly common and have caused large-scale reef mortality [17] , and ocean acidification will further stress coral reefs, compromise resilience, and in the long-term even reduce available hardgrounds [11] . Added to this, temperatures in the Gulf are rising [18] ( Fig. 2 ) and extreme heat events are increasingly frequent [19] , suggesting that bleaching events will also be more frequent and more severe. The WWF considers the Gulf ecoregion as ‘critically endangered’. In situ , the threat level for Gulf and tropical Indo-Pacific corals is comparable. Little can be done to ameliorate the situation for Gulf corals and important genetic adaptation that could help corals survive a changed climate might thus be lost. Assisted migration would therefore amount to ex-situ conservation of valuable heat-adapted Gulf stock (both coral animal and zooxanthellae) that, by potentially speeding-up thermal adaptation, would at the same time be beneficial for in-situ conservation of the recipient Indo-Pacific populations. Assisted migration of Gulf corals to the tropical Indo-Pacific is suggested by existing decision frameworks for assessing possible translocations [16] ( Fig. 5 ). Gulf species represent ∼10% of the common Indo-Pacific coral fauna and consequently represent an important heat-resistant core of ∼10% of species. In particular populations of Acropora have already been lost from the tropical Pacific due to lack of heat-adaptation and thus the re-introduction of heat-adapted genotypes may hold little genetic risk [16] . Latitudinal and altitudinal clines in thermal adaptations are common in other taxa [16] . While an anathema to some, considering these uniquely heat-adapted corals for assisted colonization with appropriate precautions may represent at least one option for preserving part of the biodiversity of coral reefs. 10.1371/journal.pone.0024802.g005 Figure 5 Decision framework for assisted migration modified and appended from [ 16 ] which clearly indicates that Gulf corals are prime candidates."
} | 1,711 |
30332416 | PMC6192577 | pmc | 1,916 | {
"abstract": "Natural spider silk is one of the world’s toughest proteinaceous materials, yet a truly biomimetic spider silk is elusive even after several decades of intense focus. In this study, Next-Generation Sequencing was utilised to produce transcriptomes of the major ampullate gland of two Australian golden orb-weavers, Nephila plumipes and Nephila pilipes , in order to identify highly expressed predicted proteins that may co-factor in the construction of the final polymer. Furthermore, proteomics was performed by liquid chromatography tandem-mass spectroscopy to analyse the natural solid silk fibre of each species to confirm highly expressed predicted proteins within the silk gland are present in the final silk product. We assembled the silk gland transcriptomes of N . plumipes and N . pilipes into 69,812 and 70,123 contigs, respectively. Gene expression analysis revealed that silk gene sequences were among the most highly expressed and we were able to procure silk sequences from both species in excess of 1,300 amino acids. However, some of the genes with the highest expression values were not able to be identified from our proteomic analysis. Proteome analysis of “reeled” silk fibres of N . plumipes and N . pilipes revealed 29 and 18 proteins, respectively, most of which were identified as silk fibre proteins. This study is the first silk gland specific transcriptome and proteome analysis for these species and will assist in the future development of a biomimetic spider silk.",
"introduction": "Introduction Spider silk is an outstanding proteinaceous fibre that outperforms other natural and synthetic fibres in tensile strength analyses. In addition to fibre strength, spider silks can also be tough, lightweight, highly extensible/flexible, biodegradable and stable across a broad temperature range [ 1 ]. Studies have also demonstrated the biocompatibility of spider silk: spider silk can be implanted into living tissue without eliciting an immune response [ 2 , 3 ]. The potential of all these qualities present in one fibre type is unique and has led to intensive research interest in the molecular structure of the various silks spiders produce. Orb-weaving spiders have up to seven different glands associated with producing silks and glues used for web architecture [major ampullate gland (MA), minor ampullate (Mi), flagelliform (Flag), aggregate (Ag), and pyriform (Py) glands], protection of eggs [tubuliform glands (Tu)], and prey wrapping [aciniform glands (Ac)] [ 4 ]. The molecular structure underpinning each of these silks and glues is the key to understanding the differing inherent mechanical properties of each type [ 5 ]. The major constituents in each silk gland are spider silk fibroins, called spidroins (Sp). While spidroins were named according to the gland from which they were primarily identified (for example the major ampullate spidroin (MaSp) from the MA gland), spidroins are not exclusively produced in their namesake gland, and proteomes have revealed silk to contain hundreds of proteins other than those derived from the spidroin gene family [ 6 – 11 ]. In the silk glands, silk proteins are secreted and then stored in the sac-like section of the gland in a water-soluble “molten fibril” or “aqua-melt” state, before processing occurs between the gland, duct and spinneret to produce the insoluble fibre we recognise within webs [ 12 , 13 ]. Spidroin sequences are very large (approx. >10 kb) and code for a sequence that includes many repetitions of several repeat motifs, and is flanked by non-repetitive, highly conserved N- and C- termini [ 5 , 14 – 16 ]. Typical repeat motifs in MA protein residues include stretches of poly-alanine (poly-A) and repeat glycine (G) motifs, such as GGX (where X could be A, glutamine, or tyrosine) [ 14 ]. An alternate MA spidroin contains stretches of poly-A and contains GPGGX motifs, where proline (P) is believed to contribute to the greater extensibility of this spidroin [ 10 , 17 ]. Beyond spidroins, much remains to be understood about the tertiary and quaternary interactions of the various proteins and glycoproteins from storage in the soluble phase within the gland, to assembly via processing steps within the duct, and finally, the solid phase fibre as it leaves the spinnerets [ 18 – 20 ]. Synthetic silks have not obtained the level of mechanical performance of their natural counterparts [ 21 – 23 ]. The factors inhibiting successful biomimicry include the inability to mimic the natural production process in current expression systems, including the inability to express full-length recombinant spidroins, and, possibly, the lack of concurrent expression of a set of spidroins and other molecular components found within a silk thread. The non-repetitive N- and C- termini of ampullate silks have been found to play a role in fibre storage and assembly [ 24 ]. Further, the N- and C-termini are also thought to aid assembly of the secondary structures of the repeat regions within the processing duct. Indeed, thus far, the best example of a biomimetic spidroin has included these important highly conserved regions flanking a short repeat region [ 25 ]. The total primary MaSp sequence has been identified within 3 species; Latrodectus hesperus , also known as the Western Black Widow spider, the golden orb-weaving banana spider, Nephila clavipes , and Argiope bruennichi , the Wasp Spider [ 14 , 26 , 27 ]. Of more than 44,000 spider species, entire spider genomes have only been described for 6 species: L . hesperus (GenBank: JJRX00000000.1); the brown recluse, Loxosceles reclusa (JJRW000000000.1); the common house spider, Parasteatoda tepidariorum (AOMJ00000000.2); the African social velvet spider, Stegodyphus mimosarum (AZAQ00000000.1); the Brazilian white-knee tarantula, Acanthoscurria geniculata (GCA_000661875.1); and, most recently, N . clavipes (MWRG00000000.1)[ 27 – 29 ]. In combination with their genome assembly for N . clavipes , Babb et al. characterised spidroin expression within the distinct silk glands [ 27 ]. Significant transcriptome coverage has also been achieved by Prosdocimi and colleagues in their infraorder comparative study of expressed RNAs of the spinning glands [ 30 ]; by Clark et al., based on multi-tissue de novo transcriptome assemblies of closely related cob-weavers [ 31 – 33 ]; and Correa-Garhwal et al. examined silk expression in male spiders of the family Theridiidae [ 34 ]. In Queensland, Australia, Nephila plumipes (Latreille, 1804) and Nephila pilipes (Fabricius, 1793) are commonly encountered golden orb-weaving species [ 35 ]. Studies on N . plumipes have reported on the mechanical properties of their silk, the relationship between protein secondary structure and primary amino acid sequence, populations in the urban environment, and copulation behaviour, however, no transcriptome data is available on the MA gland of this species, and only a handful of silk sequences are available for N . pilipes in online databases [ 36 – 42 ]. In this study, we report on a silk-gland specific transcriptome analysis for these golden orb-weaving species, N . plumipes and N . pilipes . Furthermore, proteomic analysis of silk fibres from these species was undertaken to compare predominant proteins within the silk fibre to predominant proteins expressed within the MA gland transcriptome. This study found that the silk gland transcriptome of N . plumipes and N . pilipes could be assembled into contiguous sequences, and proteome analysis of “reeled” silk fibres could confirm, and be used to mine for, spidroins within the transcriptomes. Novel proteins, which may be important constituents in the structure of spider silk, were also discovered in the silk proteome.",
"discussion": "Results and discussion Nephila plumipes and N . pilipes ( Fig 1A ) were collected and the MA glands were removed ( Fig 1B ) for RNA isolation and sequencing. MA gland reference transcriptomes were constructed for each species by combining next-generation sequence (NGS) data from sequencing runs produced in 2013, with data produced in 2016 (GenBank Accession: SRR6747912, SRR6747911). The combined MA gland transcriptomes for N . plumipes and N . pilipes produced 42,351,802 and 46,060,170 total paired reads, respectively. Paired reads were assembled into 69,812 contiguous sequences (contigs) for N . plumipes with an average length of 685 nucleotides, and into 70,123 contigs for N . pilipes with an average nucleotide length of 672. The data returned from ORF prediction were 67,862 and 67,942 sequences, and Blast2Go annotation allowed for the annotation of approximately 25% and 29% of all transcripts plus identification of 48 and 35 spidroin contigs, for N . plumipes and N . pilipes , respectively. The total spidroin count for each species was increased to 73 and 60 spidroins for N . plumipes and N . pilipes (Table 1 and 2 ), upon mining for spidroin sequences identified within each corresponding transcriptome-derived silk proteome ( S1 and S2 Tables), by analysing the six translated nucleotide reading frames for spidroin-like repeat motifs, and by BLAST searching unique silk sequences found in the related N . clavipes genome reported by Babb et al. [ 27 ]. 10.1371/journal.pone.0204243.g001 Fig 1 The golden orb-weaving spiders and a representative MA gland. (A) Nephila plumipes (photo by Alessandra Whaite) and Nephila pilipes (photo by Amos T Fairchild). (B) The major ampullate gland of Nephila pilipes , indicating the secretory section and the sac that stores silk proteins. 10.1371/journal.pone.0204243.t001 Table 1 Total spidroins in each Nephila plumipes data set (2013 and 2016). ID Predicted Spidroin Length Domain/ Superfamily RPKM 2013 RPKM 2016 TPM 2013 TPM 2016 U_9113 AcSp1 97 0.00 0.31 0.00 0.22 U_25766 AcSp1 156 0.00 1.11 0.00 0.82 U_397 AgSp1 319 59.00 149.92 48.18 110.17 U_1689 AgSp1 338 0.00 39.15 0.00 28.77 U_1891 AgSp1 183 2.01 853.75 1.64 627.39 U_25614 AgSp2 48 26.25 3.78 21.43 2.78 U_7370 AgSp-a 523 Spidroin_N superfamily 0.00 8.89 0.00 6.54 U_1394 AgSp-c 1391 0.16 11.22 0.13 8.25 U_85 AgSp-c 761 30.92 413.08 25.25 303.56 U_6753 AgSp-c 127 0.00 143.05 0.00 105.12 U_6110 AgSp-c 182 Spidroin_N superfamily 0.00 7.40 0.00 5.44 U_2628 ECP-1 † 198 0.00 4088.26 0.00 3004.33 U_4754 Flag1 226 43.48 2.19 35.51 1.61 U_20913 Flag2 107 0.00 0.48 0.00 0.35 U_18953 MaSp 89 8697.28 85599.66 7102.42 62904.26 U_10053 MaSp 88 41232.70 10475.45 33671.66 7698.05 U_11966 MaSp 99 2879.08 16061.76 2351.13 11803.24 U_17526 MaSp 71 1395.00 11596.29 1139.19 8521.72 U_1088 MaSp 103 4472.41 4888.34 3652.28 3592.28 U_47 MaSp1 57 Spidroin_MaSp 16143.09 6830.23 13182.85 5019.30 U_28059 MaSp1 112 Spidroin_MaSp 106.66 10.81 87.10 7.94 U_56 MaSp1 217 Spidroin_N 6718.57 27671.27 5486.55 20334.67 U_82 * MaSp1 189 Spidroin_N superfamily 6.10 6223.69 4.98 4573.57 U_172 MaSp1 60 Spidroin_MaSp 22158.38 581.88 18095.09 427.61 U_2048 MaSp1 60 Spidroin_N superfamily 0.00 32.84 0.00 24.13 U_4815 MaSp1 41 16.48 81.39 13.45 59.81 U_15285 MaSp1 248 Spidroin_MaSp 169.59 934.33 138.49 686.61 U_49434 MaSp1 130 Spidroin_N superfamily 0.00 0.25 0.00 0.18 U_999 * MaSp1 126 5864.84 50336.66 4789.37 36990.68 U_6045 * MaSp1 96 5861.11 7092.78 4786.33 5212.24 U_9403 * MaSp1 92 52042.69 55197.89 42499.36 40563.04 U_10063 * MaSp1 88 5068.79 30452.03 4139.30 22378.15 U_27155 * MaSp1 73 1217.71 1221.18 994.41 897.41 U_8956 * MaSp1 48 202.72 4187.95 165.55 3077.58 U_25731 * MaSp1 119 39.68 328.95 32.40 241.74 U_33 MaSp1 85 Spidroin_MaSp 39225.86 7270.57 32032.82 5342.90 U_18230 MaSp2 154 Spidroin_N 0.89 6.39 0.73 4.70 U_22058 * MaSp-c 68 Spidroin_N superfamily 15.46 1191.73 12.62 875.76 U_14382 * MaSp-d 125 190.96 9074.23 155.94 6668.34 U_6066 * MaSp-d 114 536.54 46701.72 438.15 34319.50 U_46611 MaSp-f 140 0.00 0.59 0.00 0.43 U_10737 MaSp-f 239 Spidroin_MaSp 13.74 50.18 11.22 36.88 U_1060 * MaSp-g 228 317.41 560.02 259.21 411.54 U_17514 * MaSp-g 68 3181.41 1064.47 2598.02 782.25 U_2249 * MaSp-g 135 2005.01 1405.81 1637.35 1033.08 U_679 * MaSp-g 220 406.52 428.27 331.97 314.72 U_24021 MaSp-h 82 Spidroin_N superfamily 31.96 525.52 26.10 386.19 U_17901 MiSp1 153 Spidroin_MaSp 0.00 2.99 0.00 2.20 U_11571 MiSp1 75 Spidroin_MaSp 224.25 100.05 183.13 73.52 U_12772 MiSp1 253 RP1-2 18.62 22.26 15.21 16.36 U_1636 * MiSp1 336 RP1-2 5.21 58.80 4.26 43.21 U_27015 Sp-5803 111 4.31 0.50 3.52 0.37 U_25915 Sp-74867 118 Spidroin_MaSp 3.79 0.66 3.09 0.48 U_17052 Sp-74867 93 Spidroin_MaSp 9185.22 7487.22 7500.88 5502.10 U_37 Sp-8175 920 1.40 47.08 1.15 34.59 U_6378 Sp-907 277 Spidroin_N superfamily 0.33 32.09 0.27 23.58 U_4542 * Sp-907 69 90.31 588.96 73.75 432.80 U_368 * Sp-907 259 29.90 787.06 24.42 578.38 U_17587 * Sp-907 108 1107.42 2025.01 904.35 1488.11 U_1699 * Sp-907 110 55.68 259.41 45.47 190.63 U_2859 * Sp-907 108 293.37 1928.52 239.57 1417.20 U_15734 * Sp-907 124 65.56 2816.43 53.53 2069.70 U_6377 * Sp-907 102 0.63 25.35 0.52 18.63 U_3 * Sp-907 117 422.86 3432.85 345.32 2522.68 U_319 Sp-907 164 1330.27 12939.36 1086.33 9508.69 U_21571 TuSp 179 Spidroin_MaSp 0.00 8676.94 0.00 6376.38 U_586 TuSp1 241 Spidroin_N 0.00 1486.26 0.00 1092.20 U_897 TuSp1 152 RP1-2 0.00 74973.89 0.00 55095.74 U_3907 TuSp1 183 0.00 43.55 0.00 32.01 U_7486 TuSp1 116 RP1-2 0.00 3.58 0.00 2.63 U_23917 TuSp1 149 Spidroin_N superfamily 0.00 1.32 0.00 0.97 U_60428 TuSp1 119 Spidroin_N superfamily 0.00 0.84 0.00 0.62 U_86 TuSp1 89 0.00 40804.39 0.00 29985.74 RPKM, Reads Per Kilobase of transcript per Million mapped reads TPM, Transcripts Per Kilobase Million † ECP-1, egg case protein-1 * Also identified in the corresponding proteome 10.1371/journal.pone.0204243.t002 Table 2 Total spidroins in each Nephila pilipes data set (2013 and 2016). ID Predicted Spidroin Length Domain/ Superfamily RPKM 2013 RPKM 2016 TPM 2013 TPM 2016 I_65000 AcSp 106 0.00 0.16 0.00 0.13 I_55 AcSp1 240 Spidroin_N superfamily 0.00 9455.94 0.00 7450.90 I_29651 AgSp1 143 0.82 1.24 0.55 0.98 I_69410 AgSp1 131 1.04 0.91 0.69 0.72 I_17302 AgSp1 114 5.10 5.89 3.41 4.64 I_4597 AgSp-a 281 0.46 4.64 0.31 3.65 I_17659 AgSp-a 359 Spidroin_N superfamily 0.00 2.94 0.00 2.32 I_1837 AgSp-c 1609 Spidroin_MaSp 18.38 41.31 12.31 32.55 I_31481 AgSp-c 214 0.00 0.69 0.00 0.54 I_13893 AgSp-c 148 2.71 5.22 1.82 4.11 I_1450 AgSp-c 104 10.90 37.31 7.30 29.40 I_67816 AgSp-c 112 Spidroin_N superfamily 1.24 0.47 0.83 0.37 I_18284 Flag 277 23.42 0.72 15.68 0.57 I_14698 Flag1 203 Spidroin_N superfamily 0.00 4.80 0.00 3.78 I_4468 Flag2 245 6.51 93.38 4.36 73.58 I_12595 MaSp 116 17160.99 2766.09 11489.93 2179.57 I_33885 MaSp 110 44532.47 979.27 29816.18 771.62 I_7068 MaSp 79 Spidroin_MaSp 242027.26 87.87 162046.41 69.24 I_228 MaSp1 221 Spidroin_MaSp 689.40 4201.47 461.58 3310.59 I_304 * MaSp1 277 7.87 786.60 5.27 619.81 I_820 MaSp1 173 Spidroin_MaSp 25576.95 2260.04 17124.73 1780.82 I_27560 * MaSp1 105 44.86 3339.47 30.03 2631.37 I_141 MaSp1 306 Spidroin_N superfamily 10.72 3181.16 7.18 2506.63 I_478 * MaSp1 498 Spidroin_N 133.27 362.93 89.23 285.98 I_11111 MaSp1 169 Spidroin_N superfamily 25.36 14.10 16.98 11.11 I_1327 * MaSp1 89 2.63 446.18 1.76 351.57 I_262 * MaSp1 173 2058.66 1616.88 1378.35 1274.04 I_2904 * MaSp1 92 9607.91 2971.56 6432.86 2341.47 I_17778 * MaSp1 79 17.66 1186.59 11.82 934.99 I_2074 * MaSp2 120 138074.73 62.50 92446.26 49.25 I_272 * MaSp2 96 Spidroin_MaSp 238666.86 1906.18 159796.50 1502.00 I_16679 MaSp2 93 60136.69 377.02 40263.79 297.07 I_1340 * MaSp-f 328 Spidroin_MaSp 56.92 344.75 38.11 271.65 I_9111 * MaSp-h 143 53644.90 35.16 35917.29 27.70 I_63355 MaSp-h 104 Spidroin_MaSp 8.03 0.34 5.38 0.26 I_359 * MaSp-h 392 32815.12 2140.32 21970.96 1686.49 I_550 MaSp-h 166 18520.57 5149.23 12400.23 4057.39 I_4156 MiSp 133 10236.26 252.33 6853.56 198.83 I_12574 MiSp 84 0.00 323.96 0.00 255.27 I_10698 MiSp 129 60.56 15064.25 40.55 11870.03 I_20587 MiSp 153 7.17 9866.14 4.80 7774.13 I_4316 MiSp 83 6.03 8829.59 4.04 6957.37 I_26363 MiSp 113 66.04 8262.29 44.22 6510.35 I_3787 MiSp1 66 67.69 13233.55 45.32 10427.51 I_21222 MiSp1 119 Spidroin_MaSp 110.32 7302.47 73.86 5754.06 I_9 MiSp1 136 Spidroin_MaSp 207.34 13762.70 138.82 10844.46 I_7810 MiSp1 95 87.22 9654.94 58.40 7607.71 I_7754 MiSp-a 164 52.36 11827.42 35.05 9319.54 I_8908 * MiSp-d 125 1939.86 211.64 1298.81 166.76 I_61631 Sp-1339 134 0.00 0.50 0.00 0.40 I_52113 Sp-5803 111 19.22 0.84 12.87 0.67 I_20950 Sp-907 147 Spidroin_N superfamily 30.53 19.48 20.44 15.35 I_12762 TuSp1 183 RP1-2 0.00 3.78 0.00 2.98 I_22088 TuSp1 78 Spidroin_MaSp 0.00 14056.52 0.00 11075.98 I_23940 TuSp1 187 0.00 8958.45 0.00 7058.90 I_48 TuSp1 97 0.00 978.98 0.00 771.40 I_1415 TuSp1 112 RP1-2 0.00 6772.23 0.00 5336.25 I_4212 TuSp1 118 0.00 221762.64 0.00 174740.14 I_9940 TuSp1 109 RP1-2 superfamily 0.00 20210.87 0.00 15925.36 I_8400 TuSp1 88 0.00 77285.64 0.00 60898.01 RPKM, Reads Per Kilobase of transcript per Million mapped reads TPM, Transcripts Per Kilobase Million * Also identified in the corresponding proteome A recent paper reporting on the genome and tissue transcriptomes of the golden-orb spider, N . clavipes , identified 28 spidroins [ 27 ]. Our transcriptomes included partial sequences with homology to several of the unique spidroins found in this closely related species, including AgSp-a, AgSp-c, Sp-5803, Sp8175, Sp74867 and, MaSp-c, -d -g and -h (Table 1 and 2 ). In N . plumipes , our study uncovered numerous matches to the N . clavipes spidroin Sp-907, potentially non-overlapping contigs aligning to different regions on the same gene. In N . pilipes , our study found sequences with homology to MiSp-a and -d, and Sp-1339. Our longest assembled spidroin contigs in both species matched to the N . clavipes the glue-like aggregate spidroin, AgSp-c. Aggregate spidroins, which form evenly-spaced droplets along flagelliform prey capture threads, have also been characterised in three other species from the family Araneidae, and three species from the family Theridiidae [ 45 , 46 ], These spidroins vary greatly in length among species and it appears we have recovered a full-length aggregate spidroin (1609 aa) from N . pilipes . However, aggregate spidroins were not evident in the proteome of either species. The N . pilipes and N . plumipes silk proteins were obtained after several predominant bands were excised from the Coomassie stained SDS-PAGE gel ( Fig 2 ), followed by trypsin digestion and LC-MS/MS analysis. This analysis identified 29 and 18 proteins for N . plumipes and N . pilipes (Table 3 and 4 ), respectively (≥20.00 -10lgP, ≥ 2 peptide matches). For the silk proteins of N . plumipes mapped to the transcriptome, 24 were spidroins, mostly MaSp1-like spidroins, MiSp-like and novel N . clavipes spidroin, Sp-907, confirming the abundance of this spidroin in the transcriptome. Besides spidroins, a cuticle protein and a coiled-coil domain-containing protein were identified in the proteome along with two proteins which could not be identified. Cuticle proteins have been described previously in the MA gland [ 7 , 47 ]. The coiled-coil domain protein found within the transcriptome and proteome of N . plumipes is potentially interesting as coiled-coil structures are also found to form in the silk of the Japanese yellow hornet, Vespa simillima [ 48 ]. N . pilipes silk proteins accounted for 13 of 18 proteins mapped back to the corresponding transcriptome. Again, most of these spidroins were MaSp-like, one MiSp and the remaining 5 proteins were not able to be annotated at this stage. 10.1371/journal.pone.0204243.g002 Fig 2 SDS-PAGE and Coomassie stain indicating prominent protein bands that were excised for LC-MS/MS analysis. (A-F) Nephila plumipes , (G-K) Nephila pilipes . MWM, molecular weight marker. 10.1371/journal.pone.0204243.t003 Table 3 Nephila plumipes proteins within the silk matching to the transcriptome. ID -10lgP Cover-age (%) No. of Peptides Unique Peptides PTM Average Mass Example Peptides BLAST Similarity U_999* 484.05 63 27 15 Amidation; Deamidation (NQ); Pyro-glu from Q 12348 R.GAGAAAAAAGGAGQGGYGGLGSQGAGR.G, G.QGAAAAAAGGAGQGGYGGLGGQGAGR.G Major ampullate spidroin 1 [Nephila clavipes] U_1060 403.82 66 17 16 Amidation; Deamidation (NQ); Pyro-glu from Q 19573 R.GPGGYGPGQQGPAQQGPGQQGPGGAGAAAAAGR.G, R.GPGSYGPGQQGPGQQGPR.Q Major ampullate spidroin protein MaSp-g [Nephila clavipes] U_434 282.78 23 15 15 Acetylation (K); Deamidation (NQ); Oxidation (M) 80294 R.GGGGGFNVPSGGGGLNIPSGGGR.G , R.DISSSATSASSASAGDAGGIGQGR.N N/A U_368 237.58 39 8 7 21576 R.GGDSGAAAAAAAADGGR.G, R.GGDTGAAAAAAAADGGR.G Spidroin protein Sp-907 [Nephila clavipes] U_3 172.59 16 7 6 38371 R.GGDSGAAAAAAAADSGR.G, R.GGYGGLGR.G Spidroin protein Sp-907 [Nephila clavipes] U_679 141.03 24 4 3 Deamidation (NQ) 18685 R.YGPSGPGSAAAAAAAAGAGSR.G, G.GYGPGQQGPGQQ(+.98)GPGQQG Major ampullate spidroin protein MaSp-g [Nephila clavipes] U_1081 102.56 18 4 4 Acetylation (K); Deamidation (NQ); Pyro-glu from Q 16054 K.GYDNDFVR.F, R.Q(-17.03)FDHPYK.R Hypothetical Protein NCL1 41264 [Nephila clavipes] U_4542 132.07 29 3 3 5670 L.GGDSGAAAAAAAAADGGR.G, R.GLGGDSGAAAAAAAAADGGR.G Spidroin protein Sp-907 [Nephila clavipes] U_82* 90.27 18 3 3 Acetylation (K); Deamidation (NQ); Pyro-glu from Q 18384 K.AFYQTTGTEDSR.F, G.Q(-17.03)VTPWSNAK.L Major ampullate spidroin 1 variant 1, partial [Nephila clavipes] U_6066* 159.66 46 2 2 Deamidation (NQ) 10109 R.FGSGGPGGDSAAAAAASGGNGGR.F, R.FGSGGPGGDSAAGAAASGGNGGNGGN(+.98)GGR.F Major ampullate spidroin protein MaSp-d [Nephila clavipes] U_6045* 121.31 41 2 2 Deamidation (NQ) 10812 R.AAAAAAGGAGQGGYGSLGSQGAGR.G, G.AGGAAAAAGGAGQ(+.98)GG Spidroin 1 [Nephila clavipes] U_22058 116.5 39 2 1 Oxidation (M) 7566 R.TGAFTADQLDDMSTIGDTLK.T, R.TGAFTADQLDDM(+15.99)STIGDTLK.T Major ampullate spidroin protein MaSp-c [Nephila clavipes] U_14382* 133.02 35 2 2 7519 R.FGSEGPGGDSAAAAAASGGDGGR.F, R.FGTGGPESDSAAASGGNGGNR.Q Major ampullate spidroin protein MaSp-d [Nephila clavipes] U_17587 73.53 23 2 1 8766 G.SSGAAAAAAAADGGIGR.G, S.GGYGGIGR.G Spidroin protein Sp-907 [Nephila clavipes] U_1699 55.99 16 2 1 9285 R.GGYGGLGR.G, A.AAAAAAEGGRGGYGGLGR.G Spidroin protein Sp-907 [Nephila clavipes] U_1669 27.36 3 2 2 85262 L.DVINSNESR.L, K.QKLSELEVQK.Q Coiled-coil domain-containing protein [Nephila clavipes] U_9403 236.31 100 43 36 Acetylation (Protein N-term); Deamidation (NQ); Ethylation; Methyl ester; Acetylation (N-term); 7 more 7117 Q.GAGAAAAAAGGAGQGGYGGLGSQGAGR.G, G.AAAAAAGGAGQGGYGGLGSQGAGR.G Dragline silk fibroin [Nephila clavipes] U_2859 151.32 99 30 29 Acetylation (Protein N-term); Deamidation (NQ); Ethylation; Methyl ester; Octanoyl; 6 more 8958 GLGGDSAAAAAAAADGGR.G, G.GDSAAAAAAAADGGR.G Spidroin protein Sp-907 [Nephila clavipes] U_8956* 99.15 56 19 19 Carbamidomethylation; Deamidation (NQ); Ethylation; Methyl ester; Octanoyl; 9 more 10187 R.GAGAAAAAAAGGAGQGGYRS(-18.01)EE(+14.02).H, R.GAGAAAAAAAGGAGQGGYRS(-18.01)E(+14.02)E.H Dragline silk fibroin [Nephila clavipes] U_2249 57.34 77 14 7 Deamidation (NQ); Ethylation; Methyl ester; Hydroxylation; Dihydroxy; Dehydration 10986 G.GPGGYGPGQQGPGQQGPGQ(+.98)QG.P, P.GGAAAAAAAAGGPGGYGPGQQGP.G Major ampullate spidroin protein MaSp-g [Nephila clavipes] U_10063* 98.84 92 12 7 Acetylation (Protein N-term); Deamidation (NQ); Ethylation; Octanoyl 6547 Q.GGYGGQGAGAAAAAGGAGQ.G, Q.GGYGGQGAGAAAAAGGAGQGGQ.G Major ampullate spidroin-like protein, partial [Nephilengys cruentata] U_27155 116.09 89 11 8 Acetylation (Protein N-term); Deamidation (NQ) 6231 G.AGAGAAAAAAGGAGQGGYGGLGGQ(+.98)GAGQG.G, G.AGAGAAAAAAGGAGQGGYGGLGGQGAGQ(+.98)G.G Dragline silk spidroin 1 [Nephila pilipes] U_17514 68.91 88 11 3 Acetylation (Protein N-term); Deamidation (NQ); Dehydration 5594 Q.Q(+42.01)(+.98)GPSGPGGAAAAAAAA.G, A.GPGGYGPGQQGPGQQGPGQ(+.98)QG.P Major ampullate spidroin protein MaSp-g [Nephila clavipes] U_15734 128.72 36 4 3 10272 R.GGDSGAAAAAAAADGGR.G, G.DSGAAAAAAAADGGR.G Spidroin protein Sp-907 [Nephila clavipes] U_29163 22.58 12 3 3 39198 G.GAGGGRGGGAGGNYPPQPYN.F, Q.VSIVVAALV.G Cuticle protein 10.9 [Nephila clavipes] U_1636 22.56 24 3 3 Deamidation (NQ) 29130 A.AAGGAAGYGRGAGAGAGAAAG.A, S.GAGGGAVAGAGAAAGAV.S Chain A, 3D structure of RP domain of MiSp U_6377 46.76 30 2 1 9218 R.GGYGGLGR.G, R.IGYGPGGVSGAAAVAAAADSGKG.S Spidroin protein Sp-907 [Nephila clavipes] U_97 23.18 1 2 2 Deamidation (NQ) 361672 V.DASVPGGRHK.S, C.RDISLQ(+.98)NVQK.M N/A—short sequence U_25731 20.94 19 2 1 9468 H.GGLGGQGAAAAAAGGAGQGGLGG.L, G.QGAAAAAAGGAGQGGLGG.L Dragline silk fibroin [Nephila clavipes] Proteins from the corresponding transcriptome with 2 or more peptide matches were BLAST annotated (E-value cut-off 10 −3 ). Example matching peptides are shown (full list, see S3 Table ). PTM, posttranslational modifications. 10.1371/journal.pone.0204243.t004 Table 4 Nephila pilipes proteins within the silk matching to the transcriptome. ID -10lgP Cover-age (%) No. of Peptides Unique Peptides PTM Average Mass Example Peptides BLAST Similarity I_478 288.93 25 8 8 Deamidation (NQ); Oxidation (M); Pyro-glu from Q 42362 R.QGGQGAGAATAAASGAGQGGYGR.Q, R.QGGQGAGAAAAGAGGAGR.G Major ampullate spidroin 1 variant 3 [Nephila clavipes] I_304 261.74 21 2 2 22373 R.NAAVAAAAAGGLGGYGLGGQGSGQR.S, R.PSGAGGQGAQAPGGYGTGSGSTIVITAGGQR.G Spidroin 1 [Nephila clavipes] I_33* 233.73 35 8 8 Acetylation (K); Deamidation (NQ); Oxidation (M) 24343 K.DAGGVM(+15.99)QGALGDFKDDLR.E, K.DAGGVMQGALGDFKDDLR.E N/A I_67* 179.9 46 5 5 Deamidation (NQ); Oxidation (M) 11923 R.AISESMANTGGGGLGGSR.A, R.AISESM(+15.99)ANTGGGGLGGSR.A N/A I_2074* 135.34 31 4 3 Amidation 9600 A.SYGPGPQASAAASR.L, Y.GPGPQASAAASR.L Major ampullate spidroin 2 [Nephila senegalensis] I_272* 122.08 33 4 3 Amidation 9371 Y.AAASQSAQVVSR.S, N.YAAASQSAQVVSR.S Major ampullate spidroin 2 variant 1 [Nephila clavipes] I_359* 188.07 51 28 28 Acetylation (Protein N-term); Deamidation (NQ); Ethylation; Dehydration; Dihydroxy 34638 A.GGLGGYGPGQQGPGQGGR.G, A.AAGGLGGYGP(+15.99)GQQGPGQQGPGQR.G Major ampullate spidroin protein MaSp-h [Nephila clavipes] I_1340 178.26 24 18 18 Carbamidomethylation; Deamidation (NQ); Ethylation; Dehydration; Dihydroxy; 4 more 49394 R.LSAPEAGTR.V, I.LSGP(+31.99)GR(+15.99)QASAAASR.L Major ampullate spidroin protein MaSp-f isoform 1 [Nephila clavipes] I_1327 147.26 40 12 12 Acetylation (Protein N-term); Deamidation (NQ); Ethylation; Dehydration; Hydroxylation; O-Diethylphosphorylation 14116 R.GQGGQGPSGQLAQAPSGYGQGSGAAAASGGLGGYGGQGGQR.S, G(+42.01)S(-18.01)GT(-18.01)AIAITAGGQR.G Major ampullate spidroin 1 [Argiope trifasciata] I_262 89.5 39 7 6 Acetylation (Protein N-term); Deamidation (NQ); Ethylation; Hydroxylation; O-Diethylphosphorylation 12276 G.Q(+28.03)GSGAAAAGAGQGGY(+15.99)GR.Q, G.AGAAAAAAGGAGQGGYGGLG.G Dragline silk spidroin 1 [Nephila pilipes] I_27560 129.55 71 5 2 Deamidation (NQ) 7995 R.SLGANSGEADAAGDR.G, G.AGAAAAAAGGAGQGGYGGLG.G Dragline silk spidroin 1 [Nephila pilipes] I_8908 111.49 30 3 3 9045 R.GYGPGSGAGAAAAGGAGEGGR.G, A.AAAAGGAGGEGGR.G Minor ampullate spidroin protein MiSp-d [Nephila clavipes] I_1660 34.34 1 3 3 Acetylation (Protein N-term) 438757 K.IALHLEQ.L, I.VATPDIAGV.H N/A I_9111* 33.72 44 3 3 Acetylation (Protein N-term); Deamidation (NQ) 11976 P(+42.01)GGYGPGQQGPGGYGPGQQ(+.98)GPGGAGAAAAAAAAGG.S, P(+42.01)GGYGPGQQGPGGYGPGQ(+.98)QGPGGAGAAAAAAAAGG.S Major ampullate spidroin protein MaSp-h [Nephila clavipes] I_3749 26.1 1 3 3 Deamidation (NQ) 269920 R.GGKRGN(+.98)THTKK.I, E.TLLSMN(+.98)PTR.G N/A I_23726 45.13 4 2 2 53637 L.LAADDFR.L, K.QNVKVRVASSSK.N N/A I_2904* 30.94 40 2 1 7155 G.GQGAGAAGAAAAAGGAGQGGYGGLG.G, G.QGGYGGLGGQGT.E Dragline silk spidroin 1 [Nephila pilipes] I_17778 28.19 31 2 1 Deamidation (NQ) 8397 G.YGGLGGQ(+.98)GTGAGGAAAA.G, S.ASLGGYGGLG.G Major ampullate spidroin 1A precursor [Nephila clavipes] Proteins from the corresponding transcriptome with 2 or more peptide matches were BLAST annotated (E-value cut-off 10 −3 ). Example supporting peptides are shown (full list, see S4 Table ). PTM, posttranslational modifications. Beyond the matches made to the transcriptome, from the silk of N . plumipes and N . pilipes a further 2,420 and 2,658 de novo only peptides, respectively, were identified with high confidence (average local confidence above 70) that did not match any sequences within the transcriptome. This de novo dataset of unmatched potential proteins was BLASTed against NCBI protein databases, however the focus in this study is on those proteins relevant to their corresponding transcriptomes. Unlike a genome, a transcriptome can only provide us with genes transcribed at the time of RNA isolation and this may partly explain the discrepancy between matched and de novo peptides. A further explanation is that the hand-reeled silk also contains silk from different silk glands. Each silk gland ends at its own spigot on the surface of a spinneret. It is possible that as the silk thread passes past other spigots during collection, it also collects fibres from other silk glands. However, the spigots closest to the major ampullate spigot on the anterior lateral spinnerets produce pyriform spidroins and there was no evidence of these spidroins in the proteome [ 4 ]. Further, our gel-based extraction method might have missed proteins with relatively low molecular weight or low abundance, such as the cysteine-rich proteins identified by Pham et al. which were found to co-localise with spidroins in the MA silk of Latrodectus hesperus [ 8 ]. Quantitative analyses were undertaken and based on reads per kilobase of transcript per million mapped reads (RPKM) values by mapping the 2013 and 2016 data back to a combined de novo reference transcriptome. The 50 most highly expressed sequences of N . pilipes and N . plumipes were manually selected for further annotation (Tables 5 and 6 ). These abundant sequences were matched to sequences found in the NCBI or Uniprot public protein databases (accessed Oct-Dec 2017). Spidroins were, as expected, among the most highly expressed sequences of both datasets, numbering 23 and 26 spidroins for N . plumipes and N . pilipes , respectively. In both species, major and minor ampullate, and tubuliform spidroins were highly expressed in the MA gland. Interestingly, the N . plumipes sequence with the highest RPKM value could not be characterised based on BLAST protein prediction. Uncharacterised highly expressed sequences will be selected for functional annotation in future works. 10.1371/journal.pone.0204243.t005 Table 5 Nephila plumipes silk gland most abundantly expressed genes. ID RPKM TPM Match Description E-value U_3789 346459.71 282927.68 Uncharacterised protein [ Latrodectus hesperus ] 9.40E-05 U_18953 85599.66 62904.26 Major ampullate spidroin-like protein [ Nephilengys cruentata ] 1.50E-31 U_897 74973.89 55095.74 Cylindrical silk protein 1 [ Nephila clavata ] 4.40E-54 U_9403 * 55197.89 40563.04 Dragline silk fibroin [ Nephila clavipes ] 4.80E-29 U_999 † * 50336.66 36990.68 Major ampullate spidroin 1 [ Nephila clavipes ] 1.00E-33 U_20 49180.92 36141.37 N/A U_6066 * 46701.72 34319.50 Major ampullate spidroin protein MaSp-d [ Nephila clavipes ] 1.00E-37 U_10053 41232.70 33671.66 Major ampullate spidroin-like protein [ Nephilengys cruentata ] 4.70E-16 U_86 40804.39 29985.74 Tubuliform spidroin protein TuSp [ Nephila clavipes ] 4.70E-25 U_33 39225.86 32032.82 Major ampullate spidroin 1 [ Nephila clavipes ] 4.90E-49 U_440 32127.85 23609.65 Hypothetical protein NCL1_22245 [ Nephila clavipes ] 7.40E-38 U_10063 * 30452.03 22378.15 Major ampullate spidroin-like protein, partial [ Nephilengys cruentata ] 5.40E-32 U_56 † 27671.27 20334.67 Major ampullate spidroin 1 variant 2 [ Nephila clavipes ] 9.70E-102 U_137 † 24578.13 18061.63 Hypothetical protein NCL1_28494 [ Nephila clavipes ] 1.30E-69 U_172 22158.38 18095.09 Major ampullate spidroin 1A precursor, partial [ Nephila clavipes ] 9.30E-33 U_19886 22132.89 18074.27 Hypothetical protein NCL1_19751 [ Nephila clavipes ] 3.40E-08 U_63 22118.45 18062.48 N/A U_30462 † 19965.96 14672.30 N/A U_4890 18864.55 13862.91 N/A U_47 16143.09 13182.85 Dragline silk spidroin 1 [ Cyrtophora moluccensis ] 4.50E-18 U_11966 16061.76 11803.24 Major ampullate gland dragline silk protein-2, partial [ Araneus ventricosus ] 4.20E-17 U_22856 † 14927.17 12189.90 Uncharacterized protein LOC107452916 isoform X1 [ Parasteatoda tepidariorum ] 1.10E-12 U_339 13236.17 9726.81 N/A U_319 12939.36 9508.69 Spidroin protein Sp-907 [ Nephila clavipes ] 1.20E-56 U_17526 11596.29 8521.72 Major ampullate spidroin-like protein [ Nephilengys cruentata ] 4.10E-11 U_594 9427.11 7698.42 Tubulin alpha chain [ Stegodyphus mimosarum ] 0.00E+00 U_17052 9185.22 7500.88 Spidroin protein Sp-74867 [ Nephila clavipes ] 7.40E-43 U_14382 * 9074.23 6668.34 Major ampullate spidroin protein MaSp-d [ Nephila clavipes ] 2.40E-45 U_32 † 8925.67 7288.93 Hypothetical protein NCL1_19751 [ Nephila clavipes ] 1.30E-43 U_21571 8676.94 6376.38 Tubuliform spidroin-like protein [ Nephilengys cruentata ] 7.30E-42 U_16121 8344.15 6131.83 N/A U_6045 * 7092.78 5212.24 Spidroin 1 [ Nephila clavipes ] 3.00E-38 U_24501 6484.79 5295.65 Hypothetical protein NCL1_37350 [ Nephila clavipes ] 4.90E-13 U_460 6471.84 5285.06 Ferritin [ Stegodyphus mimosarum ] 2.60E-114 U_570 6287.70 4620.62 N/A U_82 † * 6223.69 4573.57 Major ampullate spidroin 1 variant 1, partial [ Nephila clavipes ] 9.20E-85 U_152 6087.77 4971.43 Elongation factor 1-alpha [ Stegodyphus mimosarum ] 0.00E+00 U_17648 5967.76 4385.50 N/A U_134 † 5059.94 4132.07 Cathepsin B [ Araneus ventricosus ] 0.00E+00 U_1088 4888.34 3592.28 Dragline silk protein spidroin 2 [ Nephila clavata ] 3.30E-45 U_346 † 4649.24 3416.57 N/A U_94 4571.75 3733.40 Nucleoside diphosphate kinase [ Latrodectus hesperus ] 3.50E-93 U_197 † 4458.19 3276.17 Putative fasciclin [ Latrodectus hesperus ] 9.40E-42 U_18 4188.49 3077.98 N/A U_8956 † \n * 4187.95 3077.58 Dragline silk fibroin [ Nephila clavipes ] 9.20E-07 U_30 4175.22 3068.23 Cytochrome c oxidase subunit I [ Cyclosa argenteoalba ] 0.00E+00 U_12055 4175.03 3068.08 N/A U_2628 4088.26 3004.33 Egg case protein variant 1 [ Argiope argentata ] 1.50E-38 U_318 † 4079.46 2997.85 N/A U_924 † 4009.24 2946.25 N/A E-value cut-off 10 −3 † Genes found to exhibit a signal sequence * Highly expressed genes also found within the silk proteome 10.1371/journal.pone.0204243.t006 Table 6 Nephila pilipes silk gland most abundantly expressed genes. ID RPKM TPM Match Description E-value I_7068 242027.26 162046.41 Major ampullate spidroin 2 variant 1 [ Nephila clavipes ] 7.70E-39 I_272 * 238666.86 159796.50 Major ampullate spidroin 2 variant 1 [ Nephila clavipes ] 7.70E-39 I_4212 † 221762.64 174740.14 Tubuliform spidroin 1 [ Araneus ventricosus ] 5.20E-11 I_2074 * 138074.73 92446.26 Major ampullate spidroin 2 [ Nephila senegalensis ] 5.00E-13 I_8400 77285.64 60898.01 Tubuliform spidroin 1 [ Agelenopsis aperta ] 5.90E-18 I_16679 60136.69 40263.79 Major ampullate spidroin 2 [ Nephila clavipes ] 3.40E-35 I_9111 * 53644.90 35917.29 Major ampullate spidroin protein MaSp-h [ Nephila clavipes ] 1.30E-59 I_33885 † 44532.47 29816.18 Dragline silk fibroin [ Araneus ventricosus ] 1.40E-20 I_33 * 44168.19 34802.78 NA I_52 34336.68 22989.71 Spider venom protein NPTX_B154 [ Nephila pilipes ] 3.20E-10 I_5 † 33908.87 26718.84 NA I_359 * 32815.12 21970.96 Major ampullate spidroin protein MaSp-h [ Nephila clavipes ] 1.60E-136 I_67 * 26171.69 20622.25 NA I_820 25576.95 17124.73 Dragline silk spidroin 1 [ Nephila pilipes ] 1.80E-94 I_1332 † 23250.68 15567.21 Spider venom protein NPTX_C786 [ Nephila pilipes ] 1.50E-12 I_3479 † 21411.41 14335.75 Hypothetical protein NCL1_21799 [ Nephila clavipes ] 3.00E-04 I_9940 20210.87 15925.36 Cylindrical silk protein 1 [ Nephila clavata ] 4.90E-12 I_2337 † 20187.99 13516.63 NA I_77 19281.52 12909.71 ART2 [ Enterospora canceri ] 4.40E-57 I_550 18520.57 12400.23 Major ampullate spidroin protein MaSp-h [ Nephila clavipes ] 1.20E-58 I_12595 17160.99 11489.93 Dragline silk fibroin, partial [ Araneus ventricosus ] 1.80E-42 I_79 15849.11 12488.47 Hypothetical protein NCL1_39416 [ Nephila clavipes ] 2.10E-14 I_178 * 15167.42 11951.32 NA I_16079 15154.62 10146.59 NA I_10698 15064.25 11870.03 Minor ampullate spidroin-like protein [ Nephilengys cruentata ] 4.10E-63 I_5861 † 14879.35 9962.29 Hypothetical protein NCL1_19751 [ Nephila clavipes ] 3.40E-39 I_22088 14056.52 11075.98 Tubuliform spidroin 1 variant 1 [ Araneus diadematus ] 6.70E-07 I_9 13762.70 10844.46 Minor ampullate fibroin 1 [ Nephila antipodiana ] 1.90E-45 I_12551 † 13639.82 10747.64 Hypothetical protein NCL1_19751 [ Nephila clavipes ] 3.60E-31 I_3787 13233.55 10427.51 Minor ampullate silk protein MiSp1 [ Nephila clavipes ] 8.00E-27 I_576 13206.17 10405.94 Bm3878 [ Brugia malayi ] 4.50E-40 I_37 † 12356.47 9736.41 Venom allergen 5 [ Stegodyphus mimosarum ] 1.10E-41 I_7754 11827.42 9319.54 Minor ampullate spidroin protein MiSp-a [ Nephila clavipes ] 6.00E-16 I_60663 11758.95 9265.58 NA I_2057 † 10260.75 6869.96 Hypothetical protein NCL1_37703 [ Nephila clavipes ] 8.10E-15 I_4156 * 10236.26 6853.56 Minor ampullate spidroin [ Argiope argentata ] 7.10E-27 I_2304 10148.69 6794.93 Hypothetical protein THAOC_21441 [ Thalassiosira oceanica ] 2.50E-22 I_20587 9866.14 7774.13 Minor ampullate spidroin-like protein [ Nephilengys cruentata ] 2.00E-56 I_8429 9848.94 7760.57 NA I_7810 * 9654.94 7607.71 UniProt BLAST Minor ampullate silk protein MiSp1 [ Nephila clavipes ] 4.50E-39 I_2904 * 9607.91 6432.86 Dragline silk spidroin 1 [ Nephila pilipes ] 6.00E-33 I_31392 9573.32 7543.40 CRISP/Allergen/PR-1 [ Parasteatoda tepidariorum ] 6.40E-07 I_55 † 9455.94 7450.90 Cylindrical silk protein 1 [ Nephila clavata ] 6.40E-60 I_55059 9033.46 6048.25 NA I_23940 8958.45 7058.90 Cylindrical silk protein 1 [ Nephila clavata ] 1.80E-66 I_251 8938.21 5984.47 Putative tumor differentially expressed protein [ Latrodectus hesperus ] 2.10E-21 I_4316 8829.59 6957.37 Minor ampullate spidroin-like protein [ Nephilengys cruentata ] 1.70E-30 I_710 8760.58 5865.54 NA I_7 † 8494.88 6693.63 Hypothetical protein X975_26006 [ Stegodyphus mimosarum ] 1.90E-05 I_26363 8262.29 6510.35 Minor ampullate spidroin-like protein [ Nephilengys cruentata ] 1.40E-57 E-value cut-off 10–3 † Genes found to exhibit a signal sequence * Highly expressed genes also found within the silk proteome This study found the MA gland alone produces six of the seven classes of silk products: MA, minor ampullate, flagelliform, tubuliform (also at times referred to as cylindriform silk), aciniform and aggregate silk products. Several other studies have also found multiple spidroin types expressed in a single gland [ 6 , 7 , 49 , 50 ]. The only silk product not found to be produced by the MA gland of both N . pilipes and N . plumipes was pyriform adhesive silk, which is used to attach threads to objects and to each other [ 51 ]. The processing duct of the pyriform gland is shorter than most other ducts suggesting other silks require more extensive processing, which may explain why this silk is absent from the MA gland transcriptome. However, pyriform products are the least intensively studied of the spider silk repertoire, and the lack of pyriform annotation in our MA databases may be a reflection of poor representation in the public databases at the present time [ 51 , 52 ]. Interestingly, in both N . plumipes and N . pilipes , tubuliform spidroins were found to be more highly expressed in the 2016 MA gland transcriptomes yet not expressed in the 2013 transcriptomes (see Table 1 and 2 ). Tubuliform silk is produced during reproduction for the formation of egg sacs [ 4 , 53 ]. While no spiders were gravid at the time of dissection, it is possible they were collected and dissected just after the production of an egg sac, or just prior to vitellogenesis, and the 2016 transcriptomes reflect this in their relatively high expression of tubuliform silk transcripts. Expression of tubuliform spidroins in the MA gland has been previously noted in transcriptomic studies [ 50 ]. Vasanthavada et al. suggest that spiders can downregulate the production of various silks to maintain MA spidroin synthesis as an energetic trade-off, and Larracas et al. suggest that female spiders may shift synthesis of MA gland spidroins to tubuliform spidroins during the reproductive stage [ 6 , 54 , 55 ]. Our study did not find tubuliform spidroins in the silk proteome, however, the silk was collected and digested at the same time as the 2013 transcriptomes. It would be interesting to see if tubuliform spidroins could be found within the dragline silk of spiders prior to, during, or just post egg sac production. This study is the first silk gland-specific transcriptome and proteome analysis in these Australian golden orb-weaving species. Major ampullate transcriptome analysis procured sequences for all silk types thus far known for golden orb spiders with the exception of pyriform adhesive silk. We found differential expression of tubuliform silk in the MA gland, suggesting a greater role for this gland producing tubuliform silks during spider reproduction. The silk proteome analysis resulted in 29 and 18 proteins for N . plumipes and N . pilipes that match to their corresponding MA gland transcriptomes."
} | 10,328 |
35343097 | PMC9130921 | pmc | 1,919 | {
"abstract": "Abstract The rise of emerging technologies such as Big Data, the Internet of Things, and artificial intelligence, which requires efficient power schemes, is driving brainstorming in data computing and storage technologies. In this study, merely relying on the fundamental structure of two memristors and a resistor, arbitrary Boolean logic can be reconfigured and calculated in two steps, while no additional voltage sources are needed beyond “± V \n P ” and 0, and all state reversals are based on memristor set switching. Utilizing the proposed logic scheme in an elegant form of unity structure and minimum cost, the implementation of a 1‐bit adder is demonstrated economically, and a promising circuit scheme for the N ‐bit adder is exhibited. Some critical issues including the crosstalk problem, energy consumption, and peripheral circuits are further simulated and discussed. Compared with existing works on memristive logic, such methods support building a memristor‐based digital in‐memory calculation system with high functional reconfigurability, simple voltage sources, and low power and area consumption.",
"conclusion": "3 Conclusion In brief, this study proposes and experimentally demonstrates that two memristors can achieve the calculation of 16 Boolean logic functions in two steps by adopting V / R – R setting. A prototype MALU is verified. Owing to the unity of memory and computation, the proposed MALU can exhibit high energy‐efficiency cause of the near‐zero static power consumption feature and reduction in data access. The fundamental circuit is based on a memristive device with a Ti/HfO \n x \n /TiN structure exploiting its sufficiently large switching ratio and extremely fast resistance‐state switching speed to improve the calculation accuracy and overall energy consumption. A 1‐bit full adder based on the logic circuit is constructed, highlighting the advantages of easy cascading, the superiority of the logic scheme, and providing inspiration for subsequent research on in‐memory computing. The inevitable crosstalk problem on the memristor crossbar array is also analyzed in detail. Finally, a complete design idea of the N ‐bit serial carry adder using block parallelism is given. We believe this work could be a meaningful step forward to building a practical memristor‐based processor for in‐memory computing.",
"introduction": "1 Introduction Computation and storage are the two basic functions of an integrated circuit chip. The transfer of data between memory and computing unit often consumes a great deal of power, and many acceleration chips do not resolve well this problem yet. [ \n \n 1 \n , \n 2 \n \n ] Accordingly, with computing shifting toward data‐centric, if a novel nonvolatile memory‐based computing structure can be found, the functions of arithmetic and logic unit (ALU) and memory will be fused within a unit (MALU). Thus, the concept of in‐memory computing and potential solutions are proposed, which include logic in‐memory computing (LIM) based on binary devices and neuromorphic computing using multivalue devices. In this solution, data is not required to be saved by a separate memory unit and then calculated by a special ALU. Instead, the storage and calculation operations are conducted directly in this MALU. [ \n \n 3 \n , \n 4 \n \n ] From the perspective of computing, computing‐in‐memory can effectively avoid the traditional widespread “memory wall” problem of von Neuman structure, while reconfigurable technology can effectively circumvent the “compile wall” problem as the reconfigurable operators can reconstitute the circuit into a form that most closely resembles the algorithm. The memory fusion technology investigated in this study exploits possibility of the nonvolatile memory and its conditional resistance characteristics to integrate logical operation and information storage, thereby providing fine‐grained support for implementing this architecture‐level memory‐computing fusion effectively. [ \n \n 5 \n , \n 6 \n , \n 7 \n , \n 8 \n , \n 9 \n , \n 10 \n , \n 11 \n , \n 12 \n \n ] \n Memristor turns out to become a strong candidate device for in‐memory computing for its natural nonvolatility, high speed, high density, low power consumption, and full compatibility with CMOS process at low cost. [ \n \n 13 \n , \n 14 \n , \n 15 \n , \n 16 \n \n ] Having been proposed and proven already, the LIM of memristors can fall into two categories in principle, that is, stateful logic and nonstateful logic. [ \n \n 17 \n , \n 18 \n , \n 19 \n , \n 20 \n \n ] The input and output in the stateful logic are expressed as the resistance state of the memristor. Since 2010, HP Labs has implemented material implication logic functions in a circuit comprising two memristors and a ground resistor. [14 \n ] Such a scheme facilitates logic cascade and parallel computing. However, with the increase in the logic computing complexity, the number of used devices and the operation complexity increase. Nonstateful logic consists of voltage input and resistance output ( V – R ), voltage input and voltage output ( V – V ), resistance input and voltage output ( R – V ), separately. Specific to V – V logic and R – V logic calculation, the output is volatile, and additional storage units are required to store the logic output, which violates the purpose of LIM. Hence, the prospects of these schemes are limited. [ \n \n 21 \n , \n 22 \n , \n 23 \n \n ] \n \n R – R logic and V – R logic are the most promising design schemes at present. R – R logic calculation is composed of the identical physical form of input and output, resistance, which helps build a simple and pure memristor in‐memory computing system. However, it exhibits some defects (e.g., single logic function, considerable memristor devices, large circuit area, and operation time cost). In traditional V – R logic calculation, by assigning logical meaning to the operating voltage directly, the process of writing the logical input signal in the memristor device is eliminated. Although this feature can shorten the operation time of logical calculation, in the process of logical iterative calculation, besides calculation, the resistance output of each calculation must be converted into the voltage input of the next calculation, thus increasing the operation time and power consumption, namely the “cascading problem.” [ \n \n 24 \n , \n 25 \n , \n 26 \n , \n 27 \n , \n 28 \n \n ] \n In this study, a V / R – R scheme is proposed to weigh and balance the performance indicators mentioned above. Combining R – R and V – R advantages, the V / R – R type solves the cascading problem efficiently (compared with the V – R type) and significantly reduces the number of devices and operating steps when implementing the complex logic functions (e.g., XOR and XNOR) (compared with the R – R type). 16 complete Boolean logic functions are altogether implemented in two steps based on two memristors. In this scheme, the input of the binary logic is respectively presented in the form of voltage signal and the resistance state of the memristor. When multiple logic operations are cascaded, the output of the last operation can be directly stored in the memristor, which is used as the input of the next logic operation in the form of resistance. A more detailed comparison of major properties with serial stateful logic is shown in Table S1 , Supporting Information. Meanwhile, when the intermediate results need to be transmitted, signal conversion from resistance to voltage can be achieved synchronously within the time step of logical operation. The corresponding reading circuit has been simulated. Last, a 1‐bit full adder based on the LIM structure is implemented, which uses less devices and operation steps than a full adder based on CMOS and most of the existing works based on memristors. Combined with data manipulation issues on memristor crossbars, a promising N ‐bit parallel full adder circuit scheme is proposed.",
"discussion": "2 Results and Discussion 2.1 Experimental Realization of All 16 Boolean Logic Functions The memristor device structure used to illustrated the logic scheme is presented in Figure \n 1 a . The hierarchical structure of the device can be characterized under the transmission electron microscope (TEM) with image shown in Figure 1b . The TEM‐EDS mapping images of N, Ti, and Hf for the device are given (Figure 1c ). After the 3 V forming voltage initializes the device, 100 consecutive DC cycles are performed (Figure 1d ). Thereafter, the set voltage ( V \n set ) and the reset voltage ( V \n reset ) pulses are configured with magnitudes of nearly 0.6 and −1.1 V, respectively. The device switches from a high‐resistance state (HRS) to a low‐resistance state (LRS) when a voltage larger than the set voltage is applied; vice versa, when a voltage magnitude larger than V \n reset voltage is applied on the device, the device changes from LRS to HRS. Exponential test sampling results of 10 000 reversible switches under set (0.6 V, 30 ns) and reset (−1 V, 30 ns) pulses with an HRS‐to‐LRS ratio of 100 are illustrated in Figure 1e . Figure 1f demonstrates that HRS and LRS are nonvolatile maintained for 10 4 s at 85 °C with a reading bias of 0.1 V. The result of transient switching between the two resistance states is presented in Figure 1g . The device remains at an HRS over 200 kΩ until changed to a ≈400 Ω LRS when pulse width increases to ≈30 ns. According to the test result, for the device at HRS, the resistance drops sharply at a pulse of 0.6 V at 30 ns. Correspondingly, for the device at LRS, the resistance value rises to more than 100 kΩ under a pulse of −1 V in 30 ns. However, only the set operation is employed in the proposed logic scheme. Figure 1 a) The structural diagram of TiN/Ti/HfO \n x \n /TiN memristor. b) High‐resolution TEM image of the cross‐sectional cut of memristor. c) TEM‐EDS mapping images of N, Ti, and Hf in the device. d) Measured resistive switching behavior in dc I – V sweeping mode. e) More than 10 4 cycling endurance under voltage pulses (marked with the mean value and sigma/mean for HRS and LRS). f) The retention test result which is up to 10 4 s at 85 °C (marked with the mean value and sigma/mean for HRS and LRS). g) The switching speed of the memristor. Having the advantages of both R – R and V – R types, a novel scheme with extremely uniform kernel of two memristors (M 1 and M 2 ) and one resistor R, is presented for LIM, and it can efficiently implement the complete 16 Boolean logics by configuring the particular biases on terminals. The two inputs of a given logic caculation are either encoded with logic configuration into the corresponding voltage applied upon the first memristor or the resistance state of the first memristor. A 16 × 16 array composed of the above memristors is used to test the 16 Boolean logic schemes and 1‐bit fuller adder ( Figure \n 2 a ). A Boolean logic circuit, illustrated in Figure 2b , comprises the logic kernel, a switch, and a comparator. The circuit has two working modes. When the switch is off, it works in cascaded mode. That is, the output is a resistance state ( R \n out ). The next logic operation can be carried out directly without signal transformation. When the switch is on, the comparator is connected to the circuit, and the logic output result is achieved by comparing the voltage level of node A with the fixed reference value. The R acts not only as a voltage divider but also as a sampling resistor of the reading circuit coincidentally. A proper load resistor serially connected with memristor has been reported to improve the switching uniformity and endurance of the memristor. [ \n \n 29 \n , \n 30 \n , \n 31 \n \n ] The value of R is set to R = R H · R L which depends on the memristors parameters. The resistance state ( R \n out ) of M 2 and the voltage value of the output of the comparator ( R \n out ) both stand for the result of logic calculation in reading mode. In our work, the HRS and V \n s+ correspond to the logical value “0,” and the LRS and V \n s− represents the logical value “1.” The control terminals T 1 , T 2 , and T 3 are, respectively, connected to the positive pole of M 1 , the positive pole of M 2 , and one end of R. The negative poles of M 1 , M 2 , and the other end of R are linked on the identical word line. The logic implementation depends on the voltage distribution among the devices under applied voltages. T 1 , T 2 , and T 3 only need to configure three types of voltages (± V \n p , 0) to achieve 16 Boolean logic. V \n p denotes a constant that should satisfy V \n set /2 < V \n p < V \n set , so the M 2 initially in HRS cannot switch to the LRS when the voltage drop does not reach V \n set . Moreover, the resistance state of the M 2 should be reversed when the voltage across the memristor exceeds 2 V \n p . In addition, V \n p < 2| V \n reset | is required. Thus, the memristor originally in LRS will maintain its current state during the operation. V \n p is set at 0.4 V as pulse amplitude, and pulse period is set at 50 ns in subsequent simulations and tests. Figure 2 a) Metallographic electron microscopy of the 16 × 16 memristor array. b) The schematic of the memristor‐based LIM circuit. c) The instance diagram of XOR logic gate showing the port configuration. d) The truth table of XOR logic, including the detailed voltage correspondence. Take exclusive OR (XOR) logic as an example to illustrate the logical calculation process. A block diagram (Figure 2c ) shows the ports configuration of the proposed three‐input XOR gate structure. Figure 2d represents XOR logic's truth table, indicating the mapping relationship between logical input/output and physical quantities of voltage and resistance. V represents a variable determined by the logic input p . Logic value “0” represents V = 0; logic value “1” corresponds to V = V \n p . Therefore, the V \n in1 and V \n in3 configured to terminals T 1 and T 3 are determined by the logical input “ p .” Another input q is mapped as the resistance state of M 1 ( R \n in ). It is noteworthy that the original data is undamaged and still stored in M 1 after the calculation, which helps maintain the integrity of the original data. The XOR gate is implemented in one step of writing, one step of logic operation:\n Step 1: The default initial state of memristors is HRS. The second memristor (M 2 ) keep HRS, and the resistance state corresponding to input q is written to the first memristor (M 1 ); Step 2: Apply corresponding voltage pulses to the control terminals T 1 , T 2 , and T 3 , respectively (T 2 is set as a constant value of V \n p regardless of the input). The switch is set to ON when the calculation result needs to be read out, that is, the output is transmitted in the form of voltage. \n According to the truth table of XOR logic, Figure \n 3 a−d presents the resistance transitions measured in the actual circuit under four input conditions. Each condition has been tested ten times. The feasibility of the read mode is verified by involving the read circuit at the time step of a logical operation (Step 2). The rise, fall time, and the width of input pulse for T 1 , T 2 , and T 3 in simulation both are 10, 10, and 30 ns. In addition to the V \n out , the node V A , sampled for comparison with a fixed reference voltage, is also shown in Figure 3e,f . When V \n A > V \n Ref , V \n out = V \n s+ , the output of logic operation is “0.” Vice versa, when V \n A < V \n Ref , V \n out = V \n s− , the output is “1.” Figure 3 a−d) Experimental results for the four possible input−output combination of XOR logic. e–h) Simulation results for the four possible input–output combination of XOR logic in read mode. Based on the theoretical calculation from Ohm's law and Kirchhoff's law, we analyze that by arranging and combining variables at T 1 and T 3 terminals with the resistance state of M 1 , port configuration schemes of 16 Boolean logic can be achieved. There are four scenarios of configurable voltages at T 1 and T 3 with each terminal selected from {−Vp, 0} set, and two possiblilities of M 1 resistance input variable q , thus there are eight input conditions. The circuit diagrams in Figure \n 4 \n enumerate the possible voltage combinations at the control end. The upper and lower bars of the histograms show two scenarios when the input resistance of M 1 is HRS and LRS, respectively. Eight possible scenarios are demonstrated and verified. Consistent with the truth table of a certain logic function, four corresponding actual calculation scenarios can be selected from the mentioned eight scenarios, and then the corresponding port configuration is given from the set of 0, − V , − V \n p , and V − V \n p . XOR logic is taken as an example again. When the input p = 0, q = 0, the output state of R \n out = 0, M 2 should maintain a high impedance state. Meantime, the input resistance state of M 1 is HRS (>> R) after Step 1, thereby concluding that the voltage of the control terminal T 3 in the operation step should be 0. When the input p = 0, q = 1, and the output state of R \n out = 1, M 2 should be transformed to LRS. The state of M 1 is LRS, less than the resistor R, which reveals that the voltage of the control terminal T 1 in the operation step should be − V \n p . When the input p = 1, q = 0, and the output state of R \n out = 1, M 2 should switch to LRS. In this case, the resistance state of M 1 is HRS, suggesting that the voltage of the control terminal T 3 in the operation step should be − V \n p to ensure enough voltage drop. According to the truth table, when the input p = 1, q = 1, and the output state of R \n out = 0, M 2 should maintain HRS, the state of M 1 is LRS after Step 1, and the control terminal T 1 in Step 2 should be 0. Combining the above four conditions, it can be concluded that: when the input p = 0, T 1 = − V \n p and T 3 = 0; when the input p = 1, T 1 = 0 and T 3 = − V \n p . Accordingly, the port signal of the control terminals T 1 and T 3 are V − V \n p , −V, respectively. Depending on the mentioned infer processing, the port configurations of control terminals corresponding to 16 logic functions are deduced with effect ( Table \n 1 \n ). Figure 4 a–d) Experimental results for the eight possible input–output combination of logic scheme. Table 1 16 Boolean logics’ port configurations for control terminal T 1 and T 3 \n Logic function T 1 \n T 3 \n TURE − V \n p \n − V \n p \n FALSE 0 0 COPY P \n − V \n − V \n COPY Q \n − V \n p \n 0 NOT P \n \n V − V \n p \n \n V − V \n p \n NOT Q \n 0 − V \n p \n AND − V \n 0 NAND \n V − V \n p \n − V \n p \n OR − V \n p \n − V \n NOR 0 \n V − V \n p \n IMP − V \n p \n \n V − V \n p \n RIMP − V \n − V \n p \n NIMP 0 − V \n RNIMP \n V − V \n p \n 0 XOR \n V − V \n p \n − V \n NXOR − V \n \n V − V \n p \n John Wiley & Sons, Ltd. 2.2 Implementation of 1‐bit Full Adder Besides Boolean logic computing, arithmetic computing also serves an essential role in ALU. In the CMOS configuration, arithmetic functions are constructed with many Boolean logic gates. This can be implemented in the MALU, where the combinational functions are realized through sequential logic functions. Full adder is the most common and practical form of combinatorial logic. [ \n \n 32 \n , \n 33 \n , \n 34 \n , \n 35 \n , \n 36 \n \n ] A 1‐bit full adder can be realized based on six devices in six steps through the abovementioned logic circuit. For a 1‐bit binary full adder, there are three inputs (i.e., addend a \n i , summand b i \n , and carry‐in c i \n ) and two outputs (i.e., summary s i \n and carry‐out c \n ( \n \n i \n \n +1) ). Furthermore, the output results can be written as:\n \n (1) \n s i = a i ⊕ b i ⊕ c i \n \n \n (2) \n c i + 1 = a i · b i + a i ⊕ b i · c i \n \n \n Figure \n 5 a illustrates the circuit of a 1‐bit full adder, in which six adjacent cells are selected on a row or column in the crossbar. Cascade computing is demonstrated on the memristors. The calculation results are stored in M 5 and M 6 . The calculation steps in Table \n 2 \n clarified the voltage configuration corresponding to each terminal in each step. Practical issues arose regarding data manipulation in a large‐scale array. Considering the crosstalk problem, the biasing voltage setting of the memristor not involved in the calculation must be considered. First, the biasing voltage is selected from the typical V /2, V /3, gnd‐floating, gnd‐gnd, and floating‐gnd schemes. [ \n \n 32 \n , \n 33 \n \n ] Gnd‐floating is referred to the scheme with grounded unselected word lines and floating unselected bit lines and so forth. According to the superposition theorem, when the two ends of the noncalculated memristors in HRS are clamped at V \n p /2 or V \n p /3, the potential of the bottom electrode of the calculated line (WL) will be raised. With the increase of array size, it will approach V \n p /2 ( V \n p /3) which leads to insufficient voltage drop when M 2 needs overturned, resulting in calculation errors. In addition, the calculated memristors along the direction of the cascade can save “0” and “1;” when the storage information is “1” the calculated memristors are at LRS. It is not difficult to infer that the clamp voltage of this type of memristors can be directly loaded to the bottom electrode of the calculation line (WL) through the low resistance memristor. After analysis, to ensure the accuracy of the calculation, the biasing voltage scheme has been determined to use gnd‐floating (Figure 5b ). The above analysis has been verified by simulation combining the conclusions of existing work. [ \n \n 33 \n \n ] It is also found that properly trimming the voltage value at T 2 control end can improve the calculation accuracy inside ( V \n set /2, V \n set ). Unlike the bottom electrode connected to the resistance network and affected by other cells, the top electrode is independent without voltage drop loss. Appropriate adjustment can make up for the voltage loss of the bottom electrode of M 2 . Figure 5 a) The circuit diagram of 1‐bit full adder. b) Crossbar schematic marked with clamp voltage used for one‐full adder. c) Experimental results for eight logic input conditions (including intermediate results). Table 2 Detailed operation steps and voltage configuration in the addition function STEP Operation BL 1 \n BL 2 \n BL 3 \n BL 4 \n BL 5 \n BL 6 \n WL i \n 1 Write in b i \n \n \n Gnd ( b i \n = 0); \n \n V \n p ( b i \n = 1) \n Gnd Gnd Gnd Gnd Gnd \n Gnd ( b i \n = 0); \n − V \n p ( b i \n = 1) \n 2 \n a i \n · b i \n \n \n 0 ( a i \n = 0); \n ‐ V \n p ( a i \n = 1) \n \n V \n P \n Gnd Gnd Gnd Gnd Gnd 3 \n a i \n ⊕ b i \n \n \n − V \n p ( a i \n = 0); \n 0 ( a i \n = 1) \n floating \n V \n p \n Gnd Gnd Gnd \n 0 ( a i \n = 0); \n ‐ V \n p ( a i \n = 1) \n 4 \n c i \n · ( a i \n ⊕ b i \n ) floating floating \n 0 ( c \n i = 0); \n ‐ V \n p ( c \n \n i \n = 1) \n \n V \n p \n Gnd Gnd Gnd 5 \n c i \n ⊕( a i \n ⊕ b i \n ) (S \n i \n ) floating floating \n ‐Vp( c i \n = 0) \n 0 ( c i \n = 1) \n floating \n V \n p \n Gnd \n 0 ( c i \n = 0) \n ‐Vp ( c i \n = 1) \n 6 \n a i \n · b i \n + ( a i \n ⊕ b i \n ) · c i \n ( c \n ( \n \n i \n \n +1) ) floating floating floating floating − V \n p \n \n V \n p \n \n 0 ( a i \n · b i \n = 0) \n ‐Vp ( a i \n · b i \n = 1) \n John Wiley & Sons, Ltd. Figure 5c presents the experimental test results of eight possible inputs of a 1‐bit full adder. In the experimental test, the results which are consistent with the truth table are successfully obtained. Three columns on the left in Figure 5c are three inputs, while the six columns on the right represent the six devices in the 1‐bit full adder circuit. After the calculation, the resistance value in Figure 5c is the state of each memristor. It is noteworthy that the logic implementation proposed in this study is nondestructive, which reveals that the resistance state of the memristor as the input remains unchanged after the logic operation is completed. These features will reduce a deal of backup process and lay great convenience for extending the logic circuit to other applications. In the application of a 1‐bit full adder, a i \n ⊕b \n i \n and a i \n · b \n i \n can be calculated by multiplexing the b \n i \n written in memristor M 1 , thereby saving more hardware consumption for the overall calculation. When the logic circuit proposed in this study is applied to a larger system with a multistep logic calculation, it has been demonstrated to outperform other methods to a certain extent. If the logic calculation of the current step should use the result of the previous step, the characteristic of the circuit that does not damage the written data in the calculation makes the logic cascading easy to realize. 2.3 A Feasible Circuit Architecture for N ‐bit Adder Improving parallelism can maximize the use of computing resources. To increase the computational parallelism, the data manipulation of the N ‐bit adder on the memristor array relies on the bitwise parallelism and blockwise parallelism. [ \n \n 34 \n , \n 37 \n , \n 38 \n \n ] For bitwise parallel computing, the operands must be aligned before the operation, making the data manipulation more important. As can be inferred from the previous section, each line corresponds to a 1‐bit adder, so the N ‐bit adder needs an n‐line parallel operation. However, it is found that the data mapping based on memristor crossbar array inevitably mis‐operate other cells. Figure \n 6 a shows that the cells of noncalculated lines can be possibly mis‐operated when there are more than two LRS in the written data or calculated cells on the same BL. Figure 6b explains that the M 3 ’s voltage drops beyond V \n set simultaneously for the reason of two LRS memristors on the same BL 1 in front when the voltage drop between M 2 ’s control ends exceed V \n set . Increasing the device LRS parameter can effectively mitigate this problem. However, this issue cannot be wiped out from root based on the crossbar array. Figure 6 a) The mis‐operation issue caused by the bitwise parallel operation of memristor crossbar array, the mis‐operated M 3 is marked cross. b) Simulation results of mis‐operation conditions. c) A feasible circuit architecture for N ‐bit adder. Therefore, we focus on block parallelism to design an N ‐bit full adder scheme based on the memristor crossbar array. In the proposed circuit architecture, arrays of memristive switches are dedicated to performing 1‐bit or specific functions. These arrays are defined as function blocks. The BL of each block can be selected uniformly for parallel operation in one clock cycle or separately for serial manner. Specific to the implementation of serial carry N ‐bit adder, the steps marked in blue in Table 2 need to be executed serially from the algorithm, for those steps involving carry must wait for the last bit to start the calculation. Besides, the implementation of computing in‐memory needs the assistance of external circuits, including control unit (CU), that are responsible for the instruction decoding and determining the state trend of the whole system. The CU selects the corresponding block according to whether the command indicates the parallel mode or the serial mode. Each block selects the corresponding two BLs following the address control. It is worthy of mentioning that the number of memristors executing arbitrary logic in the recommended scheme is fixed and can be cascaded, that is, the address control is simple. For the addressing circuit, either direct or counter address can continuously increases one to find the next memristor because of the seamless cascading. After selecting the cells to be calculated, the configurator chooses the voltage source and loads it on the block on the basis of the Table 1 . When it is necessary to perform the signal conversion such as transmitting carry signal, the read circuit can be involved with the on‐state switch. The output V \n out is temporarily stored in the register. The main function of the register is to store the data represented by voltage signals for the proposed scheme involving both voltage and resistance signals. Using AND, OR, and XOR gates of our scheme, this N ‐bit adder based on crossbar array required 3N + 3 clocks and 6N memristors. In contrast with LIM based on memristors, it has been calculated that the energy consumption required to calculate XOR logic is no more than 1 pJ, which is competitive. [ \n \n 39 \n \n ] Compared with CMOS circuits, the area of the crossbar array can be significantly reduced, and the consumption of static power and data handling can be eliminated (Table S2 , Supporting Information). It is suitable for the application scenario of edge processors, due to the fact that power consumption is a very important consideration in the edge computing and design. [ \n \n 40 \n , \n 41 \n , \n 42 \n \n ]"
} | 7,214 |
37727686 | PMC10506583 | pmc | 1,921 | {
"abstract": "Coral reefs face an uncertain future punctuated by recurring climate-induced disturbances. Understanding how reefs can recover from and reassemble after mass bleaching events is therefore important to predict their responses and persistence in a rapidly changing ocean. On naturally extreme reefs characterized by strong daily temperature variability, coral heat tolerance can vary significantly over small spatial gradients but it remains poorly understood how this impacts bleaching resilience and recovery dynamics, despite their importance as resilience hotspots and potential refugia. In the macrotidal Kimberley region in NW Australia, the 2016 global mass bleaching event had a strong habitat-specific impact on intertidal and subtidal coral communities at our study site: corals in the thermally variable intertidal bleached less severely and recovered within six months, while 68% of corals in the moderately variable subtidal died. We therefore conducted benthic surveys 3.5 years after the bleaching event to determine potential changes in benthic cover and coral community composition. In the subtidal, we documented substantial increases in algal cover and live coral cover had not fully recovered to pre-bleaching levels. Furthermore, the subtidal coral community shifted from being dominated by branching Acropora corals with a competitive life history strategy to opportunistic, weedy Pocillopora corals which likely has implications for the functioning and stress resilience of this novel coral community. In contrast, no shifts in algal and live coral cover or coral community composition occurred in the intertidal. These findings demonstrate that differences in coral heat tolerance across small spatial scales can have large consequences for bleaching resilience and that spatial patchiness in recovery trajectories and community reassembly after bleaching might be a common feature on thermally variable reefs. Our findings further confirm that reefs adapted to high daily temperature variability play a key role as resilience hotspots under current climate conditions, but their ability to do so may be limited under intensifying ocean warming.",
"conclusion": "Conclusions As climate-induced disturbances such as marine heatwaves increasingly impact marine ecosystems globally, understanding how coral reefs respond to, and recover or reassemble after, these disturbances is critical ( Graham, Nash & Kool, 2011 ; Hughes et al., 2018a ; Tebbett, Morais & Bellwood, 2022 ). Although the factors driving recovery trajectories of coral reefs are becoming better understood ( Graham, Nash & Kool, 2011 ; Graham et al., 2015 ; Arif et al., 2022 ), it remains unclear whether lessons from more typical coral reefs apply to naturally extreme reef environments, despite their importance as potential refugia and resilience hotspots ( Camp et al., 2018 ; Burt et al., 2020 ). The findings from this study demonstrate that spatial patchiness in recovery and coral community reassembly after bleaching might be a common feature on thermally extreme reefs as temperature variability, and therefore coral heat tolerance, can vary dramatically over small spatial scales (hundreds of meters) on such reefs ( Oliver & Palumbi, 2011 ; Schoepf et al., 2015 ; Thomas et al., 2022 ). This has important implications for marine spatial planning, coral reef conservation and management, particularly in the macrotidal Kimberley region where the extreme seascape imposes significant barriers to larval dispersal ( Underwood et al., 2020 ). Thermally extreme reef environments, such as those found on shallow reefs in the Kimberley, could play an important role as resilience and adaptation hotspots that may fare better under high-emission climate scenarios than other reef areas ( Schoepf et al., 2015 ; Schoepf et al., 2020 ; Adam et al., 2022 ). However, when intensifying climate change is super-imposed on already extreme thermal regimes, the heat tolerance of even naturally heat-resistant coral populations may be overwhelmed in the long-term ( Schoepf et al., 2019 ; Klepac & Barshis, 2020 ).",
"introduction": "Introduction Coral reefs are among the most productive and biodiverse ecosystems on the planet and provide income, food and resources to millions of people ( Moberg & Folke, 1999 ; Brander, Van Beukering & Cesar, 2007 ; Fisher et al., 2015 ). However, they are threatened by a wide range of anthropogenic impacts ranging from climate change, ocean acidification and eutrophication to overfishing and invasive species ( Hoegh-Guldberg et al., 2007 ; Mumby & Steneck, 2008 ). Rising global sea surface temperature (SST) in combination with marine heatwaves have led to widespread coral mass bleaching events that will increase in frequency and intensity as climate change intensifies ( Hughes et al., 2017 ; Hughes et al., 2018a ; Frölicher, Fischer & Gruber, 2018 ). Coral bleaching is a process where heat and/or light stress leads to the breakdown of the symbiosis between the coral and its endosymbiotic algae (family Symbiodiniaceae), resulting in a significant loss of algae from the coral tissue. As the coral host meets most of its energetic requirements from carbon and nutrients acquired autotrophically by the algal symbiont ( Muscatine, McCloskey & Marian, 1981 ), bleaching leads to significant resource limitation as well as cytotoxic stress, and can result in the death of the coral colony if stress is severe or lasts for prolonged periods of time ( Oakley & Davy, 2018 ). Coral mass bleaching events can therefore lead to coral mortality on regional to global scales and are one of the key threats to coral reefs today ( Eakin et al., 2010 ; Hughes et al., 2018b ; Eakin, Sweatman & Brainard, 2019 ). As climate-induced disturbances increasingly impact coral reefs globally, understanding how reefs can recover from and reassemble after bleaching events is important to predict their responses and persistence in a rapidly changing ocean. Since mass bleaching can lead to widespread loss of live coral cover, it is particularly important to understand which factors and mechanisms allow reefs to recover versus those that drive regime shifts towards non-coral dominated states ( Graham, Nash & Kool, 2011 ; Graham et al., 2015 ; Arif et al., 2022 ). For example, over the last decades, many coral reefs have shifted from coral to algal- or animal-dominated states ( e.g. , soft corals, sponges or ascidians) due to eutrophication, loss of grazers, disease outbreak, cyclone damage or mass bleaching ( McManus & Polsenberg, 2004 ; Dudgeon et al., 2010 ; Graham et al., 2015 ; Bell, Micaroni & Strano, 2021 ). These new communities often represent alternative stable states that are difficult to reverse ( Bellwood et al., 2004 ), thus being of major concern for coral reef conservation and management. However, coral reefs have the potential to recover from catastrophic disturbances, though full recovery of coral assemblages generally takes from 10 to 15 years for the fastest growing species and far longer for the full complement of life histories and morphologies of older assemblages ( Gilmour et al., 2013 ; McClanahan & Muthiga, 2014 ; Glynn et al., 2015 ; Hughes et al., 2018a ). In recent years, coral populations with naturally elevated heat resistance have been documented in environmentally extreme reef environments characterized by high temperature variability ( Palumbi et al., 2014 ; Schoepf et al., 2015 ; Safaie et al., 2018 ). Since these locations represent resilience hotspots and potential climate change refuges, they play a critical role in facilitating future coral reef survival under rapid climate change as well as for marine spatial planning and conservation ( Camp et al., 2018 ; Burt et al., 2020 ). Thermally variable reefs have been increasingly studied to investigate the mechanisms underlying the enhanced heat tolerance of resident coral populations ( Barshis et al., 2013 ; Palumbi et al., 2014 ; Jung et al., 2021 ; Thomas et al., 2022 ). However, it remains poorly understood how they respond to, and recover from, climate-induced disturbances ( Morikawa & Palumbi, 2019 ; Schoepf et al., 2020 ; Klepac & Barshis, 2020 ), particularly over longer time scales such as years. Despite the enhanced heat tolerance of their resident coral populations, bleaching events have been documented on some thermally variable and extreme reefs as marine heatwaves increase in frequency and intensity across the globe ( Le Nohaïc et al., 2017 ; Morikawa & Palumbi, 2019 ; Klepac & Barshis, 2020 ). Yet, we currently do not know whether these reefs recover from bleaching events in the same way as reefs in more typical, environmentally more benign environments which have informed most of our knowledge on coral reef recovery dynamics ( Graham, Nash & Kool, 2011 ; Graham et al., 2015 ; Arif et al., 2022 ). One such environmentally extreme reef location is the macrotidal Kimberley region in NW Australia which is characterized by the world’s largest tropical tides yet nevertheless has abundant and highly diverse coral reefs ( Richards et al., 2015 ). This extreme tidal regime exposes resident corals to strong currents and high turbidity, with shallow corals experiencing large daily temperature fluctuations (up to 8 °C) and regular aerial exposure at low tide that can last for several hours ( Dandan et al., 2015 ; Schoepf et al., 2015 ) ( Figs. 1E – 1F ). Thus, strong environmental gradients exist over small spatial scales that have resulted in enhanced heat tolerance of corals in the highly variable intertidal compared to conspecifics in the thermally less variable subtidal ( Schoepf et al., 2015 ). Nevertheless, despite their ability to tolerate extreme environmental conditions, many coral reefs in the Kimberley region bleached extensively during the third documented global mass bleaching event in 2016 ( Le Nohaïc et al., 2017 ; Richards et al., 2019 ; Gilmour et al., 2019 ). Interestingly, intertidal corals bleached less severely and recovered rapidly whereas the subtidal coral community suffered extensive loss of live coral cover (68%) six months after the bleaching event ( Schoepf et al., 2020 ; Jung et al., 2021 ). However, it is currently unknown if the subtidal coral community was able to recover over the following years or whether shifts in coral community composition or toward non-coral dominated states have occurred. Furthermore, it is unknown whether the intertidal coral community was able to maintain its rapid return to pre-bleaching configuration in the long-term. 10.7717/peerj.15987/fig-1 Figure 1 Overview of study site. (A) Dampier Peninsula and Cape Leveque in the southern Kimberley region, NW Australia, with black square showing the study site area near Cygnet Bay. Modified from Wikipedia (Cape Leveque Road —User: Summerdrought). (B) Close-up of the study area showing the intertidal (IT) and subtidal (ST) zone at Shell Island where surveys were conducted. Source: Google ©2023 Maxar Technologies, TerraMetrics. (C) Study area at high tide. (D) Study area during outgoing tide, showing emerging intertidal pool and subtidal zone. Corals exposed at low tide in the (E) intertidal pool and (F) subtidal zone. Photo credit (C–F): V. Schoepf. Here, we investigated how the intertidal and subtidal coral community at a well-studied reef in the inshore Kimberley region had recovered from the 2016 mass bleaching event 3.5 years later and compared these data to previous benthic surveys conducted at the same reef before and during the mass bleaching event ( Le Nohaïc et al., 2017 ) and six months after bleaching ( Schoepf et al., 2020 ; Jung et al., 2021 ). Given that widespread coral mortality had occurred in the subtidal but not intertidal, the aim of this study was to investigate habitat-specific recovery trajectories. Specifically, we asked the following research questions: (1) Did live coral cover in the subtidal recover to pre-bleaching levels over 3.5 years? (2) If not, did this result in changes in benthic cover indicative of a regime shift to non-coral dominated states? (3) Did either the subtidal or intertidal coral community experience a shift in coral community composition in response to the bleaching event? (4) If yes, did this reshuffling also result in a functional community shift due to changes in dominant life history strategies, coral morphologies or community bleaching resistance? Our study provides an important contribution to the understanding of the short-term recovery dynamics (several years) of coral reefs (see also Tebbett, Morais & Bellwood, 2022 ) which is increasingly important as recovery intervals between consecutive bleaching events are becoming shorter under rapid climate change ( Hughes et al., 2018a ).",
"discussion": "Discussion As the frequency and intensity of marine heatwaves increasingly threatens coral reefs ( Frölicher, Fischer & Gruber, 2018 ), it has become ever more important to identify naturally heat-resistant coral populations that are capable of coping with intensifying heat stress. Such populations exist, for example, in thermally extreme reef environments, such as reefs with high-frequency temperature variability, and these populations have been shown to bleach less severely during marine heatwaves and exhibit high bleaching resilience ( e.g. , Safaie et al., 2018 ; Schoepf et al., 2020 ). However, it remains poorly understood how corals from thermally variable reef habitats recover from bleaching over longer time scales (∼years). Here, we show that recovery in the moderately variable, subtidal environment was incomplete 3.5 years after mass bleaching and extensive coral mortality, whereas the coral community in the highly variable intertidal had already returned to their pre-bleaching configuration within six months. Furthermore, the subtidal coral community showed a shift in community composition that did not occur in the intertidal over the same time period, indicating habitat-specific divergence in recovery trajectories. Increases in algal cover and incomplete coral recovery in the subtidal zone Following extensive coral mortality after the 2016 mass bleaching (68% loss of live coral cover in the subtidal) ( Schoepf et al., 2020 ), significant changes in benthic cover had occurred in the subtidal after 3.5 years, with the coral community showing only partial or incomplete recovery ( Fig. 4 ). When mass bleaching leads to severe reductions in live coral cover and recovery is impaired, this can lead to ‘regime’ or ‘phase’ shift towards a system dominated by taxa other than scleractinian corals, such as macro-algae or soft coral ( McManus & Polsenberg, 2004 ; Bell, Micaroni & Strano, 2021 ). In the subtidal, algal cover increased from 2% prior to the bleaching event to 15% 3.5 years later whereas it remained low (4–8% cover) and stable in the intertidal throughout this time period ( Fig. 3 ). Nevertheless, this substantial increase in subtidal macroalgal cover does not represent a phase shift towards an algae-dominated state because they are defined as algae having 50% absolute cover or algal cover exceeding live coral cover—neither of which was the case here ( Norström et al., 2009 ; Bruno et al., 2009 ). It remains to be seen whether algal cover will further increase or decrease over the coming years as recovery continues. It is technically possible that the increase in macroalgal cover in the subtidal represents heterogeneity in benthic cover rather than an actual increase over time since we did not use permanent transects. However, we consider this unlikely given that random transects were used at all time points and in both reef zones and covered the same approximate area, yet macroalgal cover did not fluctuate by more than 1–4% in either zone across the other time points. It is further unlikely that the observed changes in benthic cover occurred due to other heat-related environmental changes. The temperature logger data show that no further prolonged heat stress event occurred in the 3.5 years after the 2016 bleaching events, although some heat stress (maximum w > MMM values of up 2.6) also occurred in late summer/autumn 2017–2019 ( Fig. 2 ). We are also not aware of any reports of major bleaching during this time period. Furthermore, weekly average temperatures were highly similar in both reef zones. Therefore, changes in benthic cover and the increase in subtidal macroalgal cover, in particular, highlight the differential impact that the 2016 bleaching event had on the two different reef zones. Recovery trajectories differed markedly across the two reef zones. Recovery of the subtidal coral community was incomplete 3.5 years after the mass bleaching event because live coral cover had not fully recovered to pre-bleaching levels and a significant shift in community composition had occurred ( Figs. 4 and 5 ). This was in stark contrast to the intertidal coral community which showed full recovery and a return to its pre-bleaching configuration already within six months after the bleaching event ( Schoepf et al., 2020 ). Such habitat-specific divergence in recovery trajectories across small spatial scales (a few hundred meters) is remarkable but has also been documented in other locations and highlights the context-dependent nature of coral recovery ( Golbuu et al., 2007 ; Tebbett, Morais & Bellwood, 2022 ; Thomas et al., 2022 ). High-frequency temperature variability, in particular, has been shown to enhance coral heat tolerance and reduce bleaching risk ( Palumbi et al., 2014 ; Schoepf et al., 2015 ; Safaie et al., 2018 ) because frequent exposure to stressfully high temperatures can promote acclimatization and adaptation ( e.g. , Rivest, Comeau & Cornwall, 2017 ). While weekly average temperatures were very similar in both reef zones, the daily temperature range (DTR) was significantly higher in the intertidal than subtidal (up to 8.5 °C vs up to 4.5 °C, respectively), with maximum short-term temperatures reaching 38.09 °C and 33.84 °C, respectively ( Fig. 2 ). Positively skewed temperature distributions, such as observed here for the intertidal DTR, have been associated with stress-tolerant corals (and lower coral cover) in Western Australia ( Zinke et al., 2018 ), while reef areas with low kurtosis (intertidal < subtidal) have been linked to higher heat tolerance ( e.g. , Ateweberhan & McClanahan, 2010 ). Therefore, strong differences in daily temperature range across small spatial scales have not only resulted in differential heat tolerance of corals at our study site ( Schoepf et al., 2015 ; Le Nohaïc et al., 2017 ) but also remarkable differences in bleaching resilience and recovery potential ( Le Nohaïc et al., 2017 ; Schoepf et al., 2020 ). Interestingly, there was also a large increase in live coral cover (38% cover) in the intertidal 3.5 years after the bleaching event compared to pre-bleaching (25% cover). Although the reasons for this trend are unknown, significant increases in live coral cover despite repeated disturbances have also been documented for other reefs, including Moorea ( Holbrook et al., 2018 ) and another remote reef system in Western Australia, the Rowley Shoals ( Gilmour et al., 2019 ). Since we are not aware of any other major disturbances in 2015 or shortly before the bleaching events in April 2016, it is likely that this increase in coral cover reflects natural year-to-year variation in coral cover and/or the fact that transects were not permanent. In the subtidal, extensive mortality due to the bleaching event resulted in only 13% live coral cover six months after bleaching, compared to 36% live coral cover prior to the bleaching event. Therefore, the 25% live coral cover measured after 3.5 years of recovery indicates substantial, yet at present incomplete recovery. This is not surprising given that full recovery of coral assemblages post bleaching tends to take at least 10 to 15 years ( Gilmour et al., 2013 ; Hughes et al., 2018a ; Gouezo et al., 2019 ). However, this can also depend on whether there is a shift between species, and some coral communities have shown rapid recovery after the 2016 mass bleaching event (within 4 years) ( e.g. , Nakamura et al., 2022 ). The extreme seascape of the macrotidal Kimberley region represents significant barriers to larval dispersal ( Underwood et al., 2020 ). However, we nevertheless consider it likely that further recovery of live coral cover will take place over the coming years for several reasons, at least in the absence of further disturbances. First, many shallow coral reefs in the Kimberley are dominated by broad-cast spawning corals such as Acropora which show greater connectivity than brooders ( Underwood et al., 2020 ). Second, the completed recovery and close proximity (200–300 m) of the intertidal coral community will likely enhance local recruitment and potentially supply more heat-resistant genotypes. Finally, the Kimberley region is a remote, near-pristine marine environment ( Richards et al., 2019 ), thus the absence of local stressors in combination with low initial macroalgal cover and high branching coral cover is likely to promote coral reef recovery ( Graham et al., 2015 ; Donovan et al., 2021 ; Arif et al., 2022 ). Shifts in coral community composition in the subtidal The incomplete recovery of live coral cover in the subtidal was driven by high recruitment of Pocillopora corals which resulted in a strong shift in community composition as Pocillopora cover increased from 2% prior to the bleaching event to 44% cover 3.5 years post bleaching (mostly small colonies) ( Figs. 4B and 5 ). As a consequence, the coral community became dominated by Pocillopora (followed closely by Acropora with 40% cover, Fig. S2 ), even though Acropora made up 81% of live coral cover prior to bleaching. As brooders, Pocillopora corals can have significant advantages over broadcast spawners such as Acropora when coral spawning is disrupted after coral bleaching events ( e.g. , Szmant & Gassman, 1990 ). This is due to their prolific larval production and monthly reproductive cycle whereas spawning typically only occurs once or twice a year. Similar post-disturbance shifts from Acropora to Pocillopora dominance have also been documented on several other reefs, including Moorea ( Lenihan et al., 2011 ; Pratchett et al., 2011 ; Adjeroud et al., 2018 ) and offshore reefs on the Great Barrier Reef ( AIMS, 2021 ). This highlights that live coral cover can show positive recovery trajectories yet the coral community structure may have transitioned to a novel configuration, raising the question of whether this should be considered a ‘full’ or ‘complete’ recovery and which metrics should be used in general to assess coral reef recovery ( e.g. , Berumen & Pratchett, 2006 ). It is currently unclear over what time scales such post-disturbance shifts from Acropora to Pocillopora dominance persist and whether they represent a transitional phase indicative of either continuing degradation or recovery ( Aronson et al., 2004 ; Berumen & Pratchett, 2006 ). In Moorea, for example, Pratchett et al. (2011) argued that the community was unlikely to return to an Acropora dominated state because the low number of juveniles present indicated limited recruitment potential. However, a later modelling study predicted substantial recovery of Acropora and general reassembly to pre-disturbed levels of coral abundance, composition, and size ( Kayal et al., 2018 ). In contrast to Moorea, the percentage of Acropora in the subtidal zone at our study site is still high at 40% cover (compared to <1.0% cover in Moorea) ( Pratchett et al., 2011 ), thus a return to an Acropora dominated state seems possible. Alternatively, Pocillopora dominance may even increase further at the expense of Acropora corals, or the current co-dominance of Pocillopora and Acropora may represent an entirely new, stable, and resilient community structure that will endure unless local conditions change or further disturbances occur ( Pratchett et al., 2011 ). We caution, however, that our study site represents an environmentally extreme reef environment where lessons from more typical, less extreme reef settings may not necessarily apply. Following recovery, deficits in functional trait diversity are common on coral reefs as new coral assemblages often have altered species composition that may be deficient in key functional traits, leading to a loss of reef functionality ( Alvarez-Filip et al., 2013 ; McWilliam et al., 2020 ). Thus, the post-bleaching shift from Acropora to Pocillopora dominance in the subtidal likely has implications for the functioning and resilience of this novel coral community. For example, the subtidal coral community also shifted from being dominated by species with a competitive life history strategy (LHS) ( e.g. , Acropora spp.) to being dominated by corals with a weedy LHS ( Fig. 6 ), especially Pocillopora acuta which we identified as the most common Pocillopora species at our study site ( Fig. S2 , see Methods). In contrast, the intertidal coral community retained dominance by competitive Acropora corals throughout the four survey time points. Such shifts from competitive to weedy and/or stress-tolerant coral species are often observed after disturbances or coral mortality events since weedy corals often survive better and can opportunistically colonize recently disturbed habitats ( Darling, McClanahan & Côté, 2013 ; Kayal et al., 2015 ). The weedy LHS is further characterized by relatively fast reproduction and a brooding reproductive mode ( Darling et al., 2012 ), and some species within the P. damicornis species complex, such as P. acuta , have the ability to produce larvae via parthenogenesis ( Stoddart, 1983 ; Schmidt-Roach et al., 2014 ). Especially the ability to use parthenogenesis could explain why this genus went from being highly uncommon prior to bleaching (2% cover) to being the dominant coral genus 3.5 years later (44% cover) since this strategy would be particularly advantageous for recruitment at low colony densities in disturbed habitats ( Aronson et al., 2004 ; Darling et al., 2012 ; Carlot et al., 2022 ). High photosynthesis rates and resource allocation favoring investment in gamete or larval development over calcification, as observed in P. verrucosa from Moorea, may underlie the success of this life history strategy ( Carlot et al., 2022 ). The shift towards Pocillopora dominance in the subtidal may also have consequences regarding the resistance to bleaching, storms or outbreaks of coral predators ( e.g. , Madin et al., 2014 ). For example, this shift was accompanied by a shift from corals with a high corallite integration index score (CIIS) prior to the bleaching event (76% cover) to corals with low CIIS 3.5 years later (55% cover) ( Fig. S4 ). This was primarily due to the shift from dominance by Acropora (CIIS of 3.5) to dominance by Pocillopora (CIIS of 1.75) ( Swain et al., 2018 ). A high CIIS has been linked to a significantly reduced bleaching response because coral colonial integration and coordination improves responses to injury, predation, disease, and stress ( Swain et al., 2018 ). This shift to low CIIS dominance could therefore indicate that the subtidal may now be more vulnerable to future bleaching events. However, Acropora corals are also highly sensitive to bleaching ( Marshall & Baird, 2000 ; Loya et al., 2001 ), thus communities dominated by either Acropora or Pocillopora may have similar bleaching resistance. Other implications could be a higher resistance of the new subtidal coral community to storms and outbreaks of coral predators such as the crown-of-thorn starfish Acanthaster planci because Acropora corals have a lower resistance to storm damage than Pocillopora ( Berumen & Pratchett, 2006 ; Madin et al., 2014 ) and are also the preferred prey of A. planci ( Pratchett et al., 2017 )."
} | 7,014 |
37712147 | PMC10602577 | pmc | 1,922 | {
"abstract": "Abstract Neuromorphic hardware with a spiking neural network (SNN) can significantly enhance the energy efficiency for artificial intelligence (AI) functions owing to its event‐driven and spatiotemporally sparse operations. However, an artificial neuron and synapse based on complex complementary metal‐oxide‐semiconductor (CMOS) circuits limit the scalability and energy efficiency of neuromorphic hardware. In this work, a neuromorphic module is demonstrated composed of synapses over neurons realized by monolithic vertical integration. The synapse at top is a single thin‐film transistor (1TFT‐synapse) made of poly‐crystalline silicon film and the neuron at bottom is another single transistor (1T‐neuron) made of single‐crystalline silicon. Excimer laser annealing (ELA) is applied to activate dopants for the 1TFT‐synapse at the top and rapid thermal annealing (RTA) is applied to do so for the 1T‐neuron at the bottom. Internal electro‐thermal annealing (ETA) via the generation of Joule heat is also used to enhance the endurance of the 1TFT‐synapse without transferring heat to the 1T‐neuron at the bottom. As neuromorphic vision sensing, classification of American Sign Language (ASL) is conducted with the fabricated neuromorphic module. Its classification accuracy on ASL is ≈92.3% even after 204 800 update pulses.",
"conclusion": "3 Conclusion In summary, we demonstrated 3D‐integrated neuromorphic hardware with a two‐level neuromorphic “synapse over neuron” structure. A superjacent poly‐crystalline Si thin‐film transistor for the synapse (1TFT‐synspse) was vertically positioned over an underlying single crystalline‐Si transistor for the neuron (1T‐neuron) by 3D monolithic integration. Thermal interference between the top‐level synapses and the bottom‐level neurons was avoided with the use of optical annealing with an excimer laser, which was used for dopant activation in the 1TFT‐synapses. After an examination of the individual characteristics of the 1T‐neuron and the 1TFT‐synapse, their cooperation characteristics were investigated to demonstrate spatio‐temporal neural computation capabilities. In addition, ASL classification was conducted to demonstrate the applicability of the 3D‐integrated neuromorphic hardware for neuromorphic vision sensing based on a dynamic vision sensor. Localized electro‐thermal annealing (ETA) using the Joule heat arising from the device itself was selectively applied to the superjacent 1TFT‐synapse to improve the endurance characteristics without affecting the underlying 1T‐neurons. As a result, American Sign Language (ASL) was successfully classified with the aid of ETA even after 204 800 updates pulses. The two most innovative aspects of this work, which are the 3D integration of 1TFT‐synapse over 1T‐neuron and local ETA with Joule heat, will pave the way toward the realization of highly scalable and durable neuromorphic hardware systems in the future.",
"introduction": "1 Introduction Software‐based artificial neural networks (ANNs) have been widely used to conduct various intelligent tasks, such as pattern classification, image classification, and to design circuits. [ \n \n 1 \n , \n 2 \n , \n 3 \n \n ] However, increased energy consumption, ascribed to the massive amount of data transmission between the processors and memory components in traditional von Neumann computing, has been a chronic problem. [ \n \n 4 \n \n ] As an alternative to von Neumann computing, neuromorphic computing, which mimics the functioning of the brain, has attracted attention. [ \n \n 5 \n , \n 6 \n , \n 7 \n \n ] With the mimicry of a biological neural network with a spiking neural network (SNN) that transmits information with spikes, energy‐efficient processing is possible due to its event‐driven and spatiotemporally sparse operations. Neuromorphic hardware is composed of two elementary components: neurons and synapses, akin to a biological neural network, as illustrated in Figure \n 1 a . The neuron operation produces a spike signal as the output voltage when the input current exceeding a certain threshold is transmitted from the artificial synapses. The synapse operation memorizes and modulates the weight between two linked neurons. This feature is referred to as synaptic plasticity. Noting that there are ≈10 11 neurons and 10 15 synapses in the brain, it is difficult directly to mimic a brain system with neuromorphic hardware. One of technical challenges is to realize a high packing density of the neurons and synapses with low energy consumption. This stringent demand becomes more critical when applying neuromorphic hardware to edge computing, such as on a mobile device and on an Internet of Things (IoT) system. A conventional neuron and synapse based on a complex circuit composed of complementary metal‐oxide‐semiconductor (CMOS) transistors and capacitors have limited scalability and relatively low energy efficiency. [ \n \n 8 \n , \n 9 \n \n ] \n Figure 1 3D‐integrated neuromorphic hardware composed of single transistor‐neuron (1T‐neuron) at the top and single thin‐film transistor‐synapse (1TFT‐synapses) at the bottom. a) Schematic illustration of biological neurons and synapses in the human brain, which is composed 10 11 neurons and 10 15 synapses. b) Schematic illustration of 3D‐integrated neuromorphic hardware with superjacent 1TFT‐synapses over underlying 1T‐neurons with other CMOS interface circuits. c) Transmission electron microscope (TEM) images of the 3D‐integrated 1T‐neuron and 1TFT‐synapse. The 1T‐neuron has a single‐crystalline Si channel and SiO 2 gate dielectric, while the 1TFT‐synapse has a poly‐crystalline Si channel and SiO 2 /Si 3 N 4 /SiO 2 gate dielectrics. d) Scanning electron microscope (SEM) images of the 3D‐integrated 1T‐neuron and 1TFT‐synapse. To overcome the limitations of circuit‐based neurons and synapses, device‐level neurons and synapses based on a two‐terminal memristor or a three‐terminal transistor have been developed with their own advantages. The memristor‐based neuron has the advantages of a smaller size and lower operation voltage. [ \n \n 10 \n , \n 11 \n \n ] In contrast, the transistor‐based neuron harnesses a few merits, including better CMOS compatibility with other indispensable control circuits and greater functionality, such as inhibition and homeostasis, due to its three‐terminal operation. [ \n \n 12 \n , \n 13 \n \n ] Like the memristor‐based neuron, a memristor‐based synapse with two terminals is attractive given its smaller size and lower operation voltage. [ \n \n 14 \n , \n 15 \n \n ] In contrast, a transistor‐based synapse with three terminals is favorable for its better stability, higher selective weight update without a selector, and wider conductance range to allow multiple states. [ \n \n 16 \n , \n 17 \n , \n 18 \n \n ] \n Typically, neurons and synapses are studied individually. However, the realization of neuromorphic hardware strictly requires a hybrid form of integration. Recently, the integration for full neuromorphic hardware has been assessed. For example, a fully memristive neural network that integrates memristor‐based artificial neurons and synapses with a single crossbar array on a coplanar surface in a 2D setup was reported. [ \n \n 19 \n , \n 20 \n , \n 21 \n \n ] In addition, our group demonstrated the co‐integration of single‐transistor‐based artificial neurons and synapses on a 2D plane created via 100% CMOS fabrication by laterally positioning the neurons and synapses. [ \n \n 22 \n \n ] Co‐integration was possible in this case on a coplanar surface because the transistor neurons and synapses used are homologs to each other, i.e., they have the same structure but work differently. In terms of the packing density, vertical integration with a structure of one over the other is preferred. Because synapses greatly outnumber neurons in the brain, many layered synapses need to be integrated over neurons, in addition to the fact that complex routing is necessary to form a neural network. [ \n \n 23 \n , \n 24 \n , \n 25 \n , \n 26 \n \n ] Commercial monolithic 3D stacking technology with a trillion transistors, currently being applied to a 3D vertical NAND flash architecture (3D VNAND) in which vertically stacked memory cells are positioned over periphery circuits, can feasibly support the idea of 3D neuromorphic hardware with the structure of “synapses over neurons” for a brain‐level neuromorphic system. On the one hand, there have been a few attempts to integrate memristor‐based synapses over CMOS control circuits. In terms of manufacturing, these efforts have both strengths and weaknesses. One advantage is that the underlying CMOS circuits are scarcely influenced by the post‐processing temperature when creating the superjacent synapses due to the inherent low‐temperature fabrication process. However, a disadvantage is that the manufacturing of the synapse to create cross‐lined top and bottom electrodes can impose an additional burden on the metal interconnections by wiring each synapse at the back‐end‐of‐line (BEOL) stage. From an operational point of view, none of these methods can integrate leaky integrate‐and‐fire (LIF) neuron devices for a SNN. [ \n \n 27 \n , \n 28 \n , \n 29 \n \n ] On the other hand, another approach can be used to integrate CMOS‐based synapses over CMOS‐based neurons. Although vertically stacked neuromorphic CMOS devices are advantageous given their high packing density and CMOS compatibility, they are vulnerable to high‐temperature processes. [ \n \n 30 \n , \n 31 \n \n ] When the thermal annealing temperature exceeding 1000 °C is applied to activate the dopants in the synapse at the top, the neuron at the bottom can also be affected hence an undesirable outcome is made. In this work, we demonstrate 3D neuromorphic hardware by vertically positioning synaptic transistors over neuron transistors by means of a step‐by‐step monolithic integration process, as shown in Figure 1b . Each synapse at the top is a single thin‐film transistor (1TFT‐synapse) consisting of poly‐crystalline silicon (poly‐Si) and charge trap nitride for a tunable synaptic weight. Every neuron at the bottom is another single transistor (1T‐neuron) composed of single‐crystalline silicon (sc‐Si) with a floating body (FB) on a silicon‐on‐insulator (SOI) wafer and thermally grown oxide. Excimer laser annealing (ELA) is adopted to activate the dopants selectively in the superjacent 1TFT‐synapse in the absence of any heat transfer to the underlying 1T‐neuron. The 1T‐neuron must be located at the bottom because the sc‐Si is necessary to enable a single transistor latch for the operation of the neuron. It also must be co‐integrated with other CMOS control circuits on the same plane at the same time to ensure fabrication simplicity. The preferred location for the 1TFT‐synapse is at the top because a poly‐Si channel is not problematic for synaptic operation. Another reason is that layer‐by‐layer integration is possible for multiplying stacked synapses with a poly‐Si channel, as in the abovementioned 3D VNAND. Moreover, good endurance characteristics of the synapse, which is immune to iterative operational stresses, are essential for online training, which typically requires repeated weight updates. [ \n \n 9 \n \n ] For a memristor‐based synapse, breakdown by the set‐stuck effect arising from repetitive cycles is a concern. [ \n \n 32 \n , \n 33 \n \n ] For a transistor‐based synapse, fatigue by iterative charge trapping and de‐trapping stemming from repeated operations is troublesome owing to the accumulated stress. [ \n \n 34 \n , \n 35 \n \n ] At worst, the device can be permanently destroyed. Thermal annealing is known to be effective to cure stress‐induced damage. To repair damaged devices, recovery with wafer‐scale ex situ annealing using a furnace or a process chamber can be done. [ \n \n 36 \n \n ] As an alternative, chip‐scale in situ annealing with an embedded heater can be applied. [ \n \n 37 \n \n ] These approaches are akin to global annealing because heat is propagated to all devices. In contrast to global annealing, local annealing is necessary for the selective curing of a damaged device. This is realized upon the use of Joule heat, which arises from the current flow via the inherent resistor of a 1TFT‐synapse. Herein, we demonstrate a local electro‐thermal annealing (ETA) method with Joule heat to recover a 1TFT‐synapse selectively without thermal deterioration of the 1T‐neurons at the bottom. With this local ETA applied to the top 1TFT‐synapse, it works even after 204 800 update pulses. Finally, we utilize the 3D neuromorphic hardware to classify the patterns of American Sign Language (ASL) to demonstrate the applicability to a dynamic vision sensor (DVS) system.",
"discussion": "2 Results and Discussion 2.1 3D Integration of 1T‐Neurons and 1TFT‐Synapses Figure 1b shows an illustration of the 3D neuromorphic hardware with 1TFT‐synapses over 1T‐neurons. Because the 1T‐neurons in this case have a conventional MOSFET structure using a sc‐Si channel, they can be co‐integrated with other CMOS circuitry for peripheral interfaces, clocking, and for input/output circuits on the same plane at the bottom. [ \n \n 22 \n \n ] The detailed process flow is shown in Figure S1 (Supporting Information). A p‐type 8‐inch SOI wafer with a top sc‐Si layer of 55 nm was used as the starting material. For the fabrication of the 1T‐neurons at the bottom, sc‐Si was patterned for the channel, SiO 2 was thermally grown for the gate dielectric, and n + in situ‐doped poly‐Si was deposited for the gate electrode. Next, the gate was patterned and n + heavy doping with ion implantation was applied for the source and drain (S/D) electrode. To activate the dopant, rapid thermal annealing (RTA) was applied for 3 s at 1000 °C. Due to the floating body (FB) effect in the 1T‐neuron, a single‐transistor latch (STL) is allowed and used for neuronal firing with a spike shape. After the first inter‐layer dielectric (ILD) deposition over the 1T‐neurons, a thin film of undoped poly‐Si was deposited over the first ILD layer, after which patterning was done. As gate dielectrics for the 1TFT‐synapse, tunneling SiO 2 (O), charge trap Si 3 N 4 (N), and blocking SiO 2 (O) were sequentially formed using low‐pressure chemical vapor deposition (LPCVD). Afterwards, n + in situ doped poly‐Si was deposited for the gate electrode and was patterned. Later, n + heavy doping with another ion implantation was applied for the S/D electrode. Because the n + poly‐Si (S) gate covers the overlying ONO gate dielectrics on the undoped poly‐Si (S) channel, this structure has a SONOS configuration. Moreover, the ONO layer creates a quantum well with a finite depth to hold and emit electrons due to the intercalated charge trap nitride between the top blocking oxide and the bottom tunneling oxide. Hence, the 1TFT‐synapse can control the synaptic weight by trapping or de‐trapping electrons. For dopant activation in the S/D of the 1TFT‐synapse, ELA for optical annealing was applied instead of a thermal annealing method such as RTA. Therefore, the underlying 1T‐neurons and other transistors at the bottom are scarcely affected by any thermal stimulus due to the good thermal insulation of the first ILD. This type of light‐induced annealing to confine heat in a irradiated surface is preferred to minimize thermal interference in inter‐devices between the top and bottom level. Compared to the poly‐Si channel thickness of 50 nm and poly‐Si gate thickness of 100 nm and the first ILD layer with a thickness of 300 nm, the penetration depth of ≈6 nm stemming from its short wavelength is shallow. [ \n \n 38 \n \n ] Thus, the heat induced by ELA cannot be transferred to other areas apart from the exposed S/D of the 1TFT‐synapse. After the second ILD deposition step over the 1TFT‐synapses, metallization was applied to connect the underlying 1T‐neurons and the superjacent 1TFT‐synapse. Finally, 3D neuromorphic hardware with 1TFT‐synapses over 1T‐neurons was realized by monolithic integration, as shown in the transmission electron microscope (TEM) image in Figure 1c . The detailed fabrication processes are explained in the Experimental Section. As shown in Figure 1d , the nominal gate width ( W \n G ) and gate length ( L \n G ) are 200 and 400 nm, respectively, for the 1T‐neuron. For the 1TFT‐synapse, the nominal W \n G and L \n G values are 300 and 500 nm, respectively. The detailed device parameters are summarized in Table S1 (Supporting Information). By changing the exposure energy of the excimer laser from 150 to 500 mJ cm −2 but with a fixed laser pulse width of 25 ns, the ELA condition was optimized with two strategies. The first is to find the proper condition under which to generate heat, which is a level similar to that of RTA. The second is to exploit the condition in which the underlying 1T‐neuron is thermally insulated from the heat created at the top. ELA was applied to the experimental group, whereas RTA of 2 s at 1000 °C was used for the control group. Each group has 15 samples. Compared to the annealing effects by RTA, those by ELA were verified by one electrical instance and two physical instances. As shown in Figure S2 (Supporting Information), three representative device parameters, the threshold voltage ( V \n T ), the subthreshold swing ( SS ), and the on‐state current ( I \n ON ), were compared between the two groups. In terms of the electrical evidence, they were comparable to each other. As shown in the TEM images in Figure S3 (Supporting Information) as direct physical evidence, both poly‐Si films were recrystallized from an amorphous phase, which was caused by a heavy dose of ion implantation for the S/D of the 1TFT‐synapse. As indirect physical evidence of recrystallization, optical images were also compared. Their colors look very similar, as shown in Figure S4 (Supporting Information). Based on these observations, ELA activated dopants as RTA did. Despite the fact that heat induced by the ELA is high enough to recrystallize the amorphized S/D silicon film as well as to activate the dopants, it should not be transferred from the top level to the bottom level through the first ILD. To verify this, thermal simulations were conducted with the aid of a 3D thermal simulator (COMSOL). As shown in Figure S5a (Supporting Information), the heat distribution profile across the 3D neuromorphic hardware along the vertical direction was plotted by reflecting each film thickness and the corresponding temperature‐dependent parameters. [ \n \n 39 \n \n ] Referring to Figure S5b (Supporting Information), the top poly‐Si film was heated to 1172 °C with an exposure energy of 200 mJ cm −2 . This was high enough to recrystallize the poly‐Si film but low enough to avoid the melting of the Si. In contrast, the bottom poly‐Si film was sustained at a low temperature below 50 °C, which cannot influence the device characteristics of the underlying 1T‐neurons. On the other hand, ELA at 300 mJ cm −2 heated the top poly‐Si film to 1563 °C, which is higher than the melting temperature of Si. At a higher activation rate, more energy is preferred unless the heat can destroy the device. Therefore, we adopted 200 mJ cm −2 for the optimal condition. As indirect electrical proof of the ELA, the transfer characteristics of the drain current versus gate voltage ( I \n D – V \n G ) and the output characteristics of the drain current versus the drain voltage ( I \n D – V \n D ) of the fabricated 1T‐neuron were measured. These values were then compared before and after ELA for the superjacent 1TFT‐synapses. They were found to be nearly identical, as shown in Figure S6 (Supporting Information). It should also be noted that the 1T‐neurons at the bottom were already annealed by the pre‐applied RTA prior to the first ILD deposition. The instantaneous ELA and the thermal insulation of the first ILD inhibit heat transfer from the top level to the bottom level. 2.2 Decoupled Characteristics of Superjacent 1TFT‐Synapses and Underlying 1T‐Neurons The I \n D ‐ V \n D characteristics of the fabricated underlying 1T‐neuron ( Figure \n 2 a ) are plotted in Figure 2b . When V \n D is lower than the latch‐up voltage ( V \n latch ), I \n D does not flow because the 1T‐neuron is in a high‐resistance state (HRS). However, when V \n D exceeds V \n latch , I \n D flows abruptly given its low‐resistance state (LRS). V \n latch is defined as the critical voltage to trigger the abovementioned STL, causing I \n D to increase abruptly, i . e ., an immediate change from a HRS to a LRS. [ \n \n 40 \n \n ] This abrupt threshold switching is analogous to the firing in a biological neuron. In addition, V \n latch corresponds to the firing threshold voltage ( V \n T,firing ). The charging and discharging process is repeated as long as a constant input current ( I \n in,neu ) is applied to the drain electrode of the 1T‐neuron. Figure 2c shows the neuronal spiking characteristics, represented by the oscillating output voltage ( V \n out,neu ) versus the time. The charging process equivalent to “integrating” is continued, gradually filling the parasitic drain capacitor of the 1T‐neuron; however, a discharging process identical to “firing” suddenly occurs via the abrupt threshold switching. The top peak is V \n T,firing and the bottom peak is the resting voltage ( V \n resting ), as shown in Figure 2c . As shown in Figure 2d , a spiking frequency ( f ) increases as I \n in,neu increases, representing a typical property of a leaky integrate‐and‐fire (LIF) neuron. Given this property, more spiking is generated when more signals are transmitted from previously activated synapses. Energy consumption is crucial for low‐power neuromorphic computing. The estimated amount of neuron energy consumed per spike ( E \n neuron /spike), as extracted from E neuron / spike = I in , neu ∫ 0 1 f V out , neu d t , is 19.3 pJ per spike at I \n in,neu = 10 nA. [ \n \n 22 \n \n ] As summarized in Table S2 (Supporting Information), this value is sufficiently low compared to other artificial neuron devices. Figure 2 Measured characteristics of the decoupled 1T‐neuron (a,b,c,d) and 1TFT‐synapse (e,f,g,h). a) Cross‐sectional schematic of the 1T‐neuron with the corresponding symbolic representation to show the current input ( I \n in,neu ) to the drain and the voltage output ( V \n out,neu ) from the same drain. b) Neuronal output characteristic ( I \n D – V \n D ) of the fabricated 1T‐neuron. Abrupt threshold switching occurs at the latch‐up voltage ( V \n latch ) for the firing operation. c) Spiking characteristics ( V \n out,neu ‐ t ) of the fabricated 1T‐neuron. d) Spiking frequency ( f ) as a function of I \n in,neu with the typical property of a LIF neuron. e) Cross‐sectional schematic of the 1TFT‐synapse with the corresponding symbolic expression to show the voltage input ( V \n in,syn ) to the gate and the current output ( I \n out,syn ) from the source. f) Neuronal transfer characteristic ( I \n D ‐ V \n G ) of the fabricated 1TFT‐synapse with hysteresis caused by charge trapping and detrapping in Si 3 N 4 . g) Potentiation and depression (P/D) characteristics of the fabricated 1TFT‐synapse with 128 conductance states in each case. h) Retention characteristics of the fabricated 1TFT‐synapse. The stable operation of the 1T‐neuron is important because the damaged neuronal characteristics can degrade the system performance. To verify the endurance characteristics of the 1T‐neuron, spiking operations were repeated for 10 h while applying I \n in of 10 nA. Considering that f of the 1T‐neuron is ≈1 kHz at I \n in of 10 nA, it generates 3.6 × 10 7 spikes during the endurance test. Figure S7a shows the transfer characteristics of the 1T‐neuron before and after 10 h of spiking operations. There is no notable degradation in subthreshold swing, threshold voltage, on‐state current, and off‐state current. It implies that an 1T‐neuron is scarcely damaged even after 3.6 × 10 7 spiking. Figure S7b (Supporting Information) compares the spiking characteristics ( V \n out ‐ t ) of the before and after 10 h of spiking operations. Figure S7c,d (Supporting Information) exhibit that V \n T,firing and f were negligibly changed. These data support that the system performance would be maintained even after repetitive operations. There are five representative features in a synaptic device. The first of these is the ability to allow multiple states. Figure 2e–h show the electrical properties of the superjacent 1TFT‐synapse. The synaptic weight can be controlled by changing the trapped charge density in the nitride of the SONOS structure. As shown by I \n D – V \n G in Figure 2f , negative biasing of V \n G causes a leftward shift of V \n T , akin to erasing during memory operation. This corresponds to potentiation during synaptic operation because the conductance ( G ) is increased due to the reduction of V \n T . This negative shift is attributed to the de‐trapping of charges in the nitride of the SONOS. Positive biasing of V \n G induces a rightward shift of V \n T , analogous to programming during memory operation. This corresponds to depression during synaptic operation because G is decreased due to the increase in V \n T . These opposite shifts provoke hysteresis, which tends to be wider due to the larger | V \n G |. This potentiation and depression (P/D) characteristic, as represented by a change of G according to the number of applied pulses, is plotted in Figure 2g . Each of the 128 distinctive G outcomes in the P/D was produced with an identical pulse scheme to use a constant voltage amplitude ( V \n G ) and a fixed pulse width ( t \n pulse ), unlike other works that rely on a variable pulse scheme to utilize an incremental V \n G or a wider t \n pulse . V \n G of −12 V with t \n pulse equal to 100 µs was used for potentiation and V \n G of 9.3 V with t \n pulse set to 10 µs was used for depression. V \n G and V \n D were set to 2.5 V and 1 V, respectively for the reading operation. The second feature makes energy consumption per event ( E \n synapse /event) as low as possible for a synaptic operation. It should be noted that V \n D and the source voltage ( V \n S ) are grounded such that I \n D cannot flow through the channel during the P/D. Therefore, E \n synapse is dominated by the gate leakage current ( I \n G ) and can be estimated by E \n synapse /event = I \n G V \n G t \n pulse . Because I \n G was in a range of a few pA, as shown in Figure S8 (Supporting Information), E \n synapse /event is 2.33 fJ/event for potentiation and 0.03 fJ per event for depression. These values are lower than the biological synapse energy of 10 fJ per event. [ \n \n 41 \n \n ] \n Third is the retention characteristic. As shown in Figure 2h , distinctive states were sustained for longer than 10 5 s due to the non‐volatile charge trapping in the nitride. [ \n \n 42 \n \n ] The fourth feature is endurance, which is important for repetitive online learning. Fifth is recoverability from stress‐induced damage. A synapse will experience more frequent operation‐induced stress due to repeated synaptic updates and will undergo harsher voltage‐induced stress due to the high applied voltage. It should also be noted that the trapping and de‐trapping for synapse operation require a higher | V \n G | above 9 V to allow the tunneling of I \n G intentionally via the ONO dielectrics, whereas the firing for the neuron operation requires a V \n D value below 2 V. If possible, selective curing for a superjacent 1TFT‐syanspe can be done to improve endurance. Curing should selectively mitigate the fatigue or aging of a superjacent 1TFT‐synapse, which arises from the iterative electrical stresses, though it should not have any side effects on any underlying 1T‐neurons. The stress‐induced damage was cured by ETA, which utilized the Joule heat generated in the device. This approach can selectively confine the Joule heat inside the target device. If a wafer‐scale global annealer such as furnace or RTA chamber is used, the well‐optimized thermal budget for an undamaged synapse or underlying 1T‐neurons can become overloaded, in turn causing the corresponding pre‐designed electrical behaviors to be deviated from the target values. Moreover, there is an upper limit of the allowable maximum temperature without the melting of the metal silicide or metal interconnections used. To generate Joule heat inside the 1TFT‐synapse, punchthrough current ( I \n punch ), which flows through n + S via the undoped poly‐Si channel to n + D, was utilized. [ \n \n 43 \n , \n 44 \n \n ] Once more, 3D thermal simulations with the aid of the commercial simulator (COMSOL) were conducted to analyze the temperature ( T ) distribution profile and to optimize the condition of the Joule heat. When I \n punch of less than 105 µA was applied, the induced T near the interface between the ONO gate dielectrics and the undoped poly‐Si channel was only 512 °C, which is not high enough to assist the thermal curing process, as shown in Figure S9a (Supporting Information). Otherwise, when I \n punch exceeds 165 µA, the induced T reached 1415 °C, which is above the melting point of poly‐Si. Thus, the optimal T for curing by ETA is between these two values. When I \n punch is 135 µA, T reached 833 °C, which is high enough to cure the stress‐induced damage on the gate dielectrics, as shown in Figure \n 3 a . [ \n \n 37 \n , \n 44 \n \n ] Although a high T is induced in the poly‐Si at the top, current induced heat is scarcely transferred to the sc‐Si at the bottom due to the good thermal insulation of the first ILD layer, as shown in Figure 3b and Figure S9b (Supporting Information). This facilitates the repair of the damage of the superjacent 1TFT‐synapse selectively and improves the endurance characteristics without affecting the underlying 1T‐neurons. Based on these results, it is inferred that a two‐level neuromorphic module of “synapses over neurons” is attractive not only for the resulting good packing density but also for long‐term endurance with selective curing while preventing thermal disturbances. Figure 3 Improved endurance by electro‐thermal annealing with Joule heat. a) Heat distribution profile in the 1TFT‐synapse obtained from 3D thermal simulation. Joule heat is generated by the internal punchthrough current ( I \n punch ). b) The heat distribution profile along the vertical direction shows that the generated heat is not transferred to the bottom layer. Localized ETA to cure the fatigue of a 1TFT‐synapse does not affect the underlying 1T‐neuron. c) Degraded P/D characteristics caused by iterative operations of the 1TFT‐synapse. d) Cured P/D characteristics of the 1TFT‐synapse by ETA after fatigue arising from iterative operations. As shown in Figure 3c , the P/D characteristics were gradually degraded, i . e ., the range of the G values was reduced owing to stress‐induced damage such as trap generation in the gate dielectric. [ \n \n 35 \n \n ] Note that Fowler‐Nordheim (F‐N) stress induced by repetitive programming and erasing in SONOS flash memory degrades the reading current, which is related to G during the synaptic operation. [ \n \n 44 \n \n ] As shown in Figure 3d , the endurance characteristics were appreciably improved with periodic ETA in each case after 2580 weight updates. Figure S10a (Supporting Information) compares the P/D curves before and after ETA. As expected, the range of G is broadened after ETA. Figure S10b (Supporting Information) compares the I \n D – V \n G plots before and after ETA. The subthreshold swing ( SS ) of the superjacent 1TFT‐synapse was improved by 32.0% and the corresponding on‐state current ( I \n ON ) was enhanced by 50.7%. As shown in Figure S11 (Supporting Information), however, the spiking characteristics of the underlying 1T‐neuron were not changed after ETA was applied to the superjacent 1TFT‐synapse. This outcome confirms that ETA cured the synaptic fatigue selectively but did not change the neuronal characteristics. By the way, it is necessary to evaluate energy consumption ( E \n ETA ) for ETA. The applied I \n punch and voltage ( V \n ETA ) for ETA were 165 µA and 18 V, respectively, while a pulse width ( t \n ETA ) of the V \n ETA was 1 ms. Therefore, the E \n ETA can be calculated as 2.97 µJ with I \n punch V \n ETA t \n ETA . This E \n ETA is much larger than spiking energy of 19.3 pJ per spike by an 1T‐neuron and weight update energy of 0.03 fJ per event in an 1TFT‐synapse. However, it may not be a concern because ETA to enhance the endurance of the 1TFT‐synapse is not frequently applied. Furthermore, down‐scaling of a gate length ( L \n G ) and gate width ( W \n G ) of an 1TFT‐synapse can help lower the E \n ETA because the demanded I \n punch , V \n ETA , and t \n ETA are accordingly reduced. [ \n \n 44 \n , \n 45 \n \n ] \n Table S3 (Supporting Information) compares the synaptic characteristics among previously reported synaptic devices with 3D stacking and those in this work. [ \n \n 26 \n , \n 28 \n , \n 29 \n \n ] There are two notable features in this work. The first is that it represents the first implementation of SNN by stacking synaptic devices onto LIF neurons. The second feature is the selectively curable endurance realized by ETA. Table S4 (Supporting Information) presents the device‐to‐device variation variability of 1T‐neurons and 1TFT‐synapses, which includes the average and standard deviation values of various device parameters. The electrical characteristics of 10 distinct 1T‐neurons and 1TFT‐synapses were measured. Notably, no significant variations were observed, thus facilitating large‐scale integration. 2.3 Coupled Characteristics of Superjacent 1TFT‐Synapses and Underlying 1T‐Neurons To realize an artificial neural network, the cooperative properties of the 1T‐neurons and the 1TFT‐synapses should be investigated in addition to their individual properties. Figure \n 4 a shows a circuit diagram for the cooperation between a pre‐synaptic 1T‐neuron and a 1TFT‐synapse. As long as a constant I \n in,neu is fed to the drain of the pre‐synaptic 1T‐neuron, an oscillating V \n out,neu was emitted from it. Then, this V \n out,neu was applied to the gate of the 1TFT‐synapse as an input signal. Finally, the output current from the 1TFT‐synapse ( I \n out,syn ) reflecting the synaptic weight, which would be transmitted to a post‐synaptic 1T‐neuron, flowed through the 1TFT‐synapse. As shown in Figure 4b , the spiking frequency ( f ) of I \n out,syn increases as I \n in,neu increases. Figure 4 Coupled neuromorphic operation of 1TFT‐synapses at the top and 1T‐neurons at the bottom. a) Vertical hybrid of a pre‐synaptic 1T‐neuron and 1TFT‐synapse. b) Measured I \n out,syn from a 1TFT‐synapse versus time for various values of V \n out,neu from the 1T‐neuron. Here, f increases with a larger I \n in,neu . c) Measured I \n out,syn according to the synaptic weight. I \n out,syn increases with a greater synaptic weight. d) Array structure composed of the pre‐synaptic 1T‐neurons and 1TFT‐synapses. e) Extracted f from each 1TFT‐synapse cell. f) Extracted maximum value of I \n out,syn from each 1TFT‐synapse cell. g) Post‐synaptic 1T‐neuron collecting and receiving I \n out,syn outputs from two different 1TFT‐synapses. h) V \n out,neu from the post‐synaptic 1T‐neuron depending on the two I \n out,syn signals. f is higher when two signals are applied coincidently, thus demonstrating the function of spatio‐temporal coincidence detection. In addition, the maximum I \n out,syn becomes larger as the synaptic weight ( w \n i ) increases, as shown in Figure 4c . This feature is a desired characteristic for the connection between the pre‐synaptic neuron and the synapse. Here, I \n out,syn was measured from the array depicted in Figure 4d . In this case, a different I \n in,neu was applied to each 1T‐neuron in a row and a different w \n i was emitted from each 1TFT‐synapse in a column. Figure 4e,f show the measured f and maximum I \n out,syn from each 1TFT‐synapse, respectively. It could be verified that the higher value of f is produced by the larger I \n in,neu from the pre‐synaptic 1T‐neuron and that the larger I \n out,syn is generated by the stronger w \n i from the 1TFT‐synapse. Figure 4g,h shows the operation of a post‐synaptic 1T‐neuron when I \n out,syn is transferred from the 1TFT‐synapses. As illustrated in Figure 4g , I \n out,syn from the 1TFT‐synapses is applied to the drain electrode of the post‐synaptic 1T‐neuron. Considering that more than a few hundred synapses are connected to each neuron in the brain, the collected current from many synapses with various values of w \n i , assuming that all are to be transmitted to a neuron, is too large to be adapted to a proper value of I \n in,neu . Hence, I \n out,syn must be reduced. A poly‐Si channel with low mobility and a narrow channel width like our 1TFT‐syanpse is preferred to satisfy this demand. More aggressively, an extra circuit component may be needed when numerous synapses, i.e., more than ten thousand, are linked to a neuron. In such a case, the current mirror depicted in Figure S12 (Supporting Information) \n , \n which is co‐integrated with the 1T‐neuron prior to the first ILD deposition, is necessary to reduce the level of I \n out,syn further. [ \n \n 22 \n \n ] As an experimental example of cooperation between the post‐synaptic 1T‐neuron and the 1TFT‐synapses, a coincidence detection process was utilized to depict the spatio‐temporal neural computation. In biology, coincidence detection, referring to the encoding of information by identifying the occurrences of input signals that are spatially distributed but close in time, is used as an important neural computation process in visual and auditory systems. [ \n \n 46 \n , \n 47 \n , \n 48 \n \n ] As shown in Figure 4h , when the two inputs of I \n out,syn1 and I \n out,syn2 were applied together, f was increased because the summed I \n in,neu applied to the post‐synaptic 1T‐neuron was increased. In light of this, it is feasible to ascertain whether two inputs are synchronized. By the way, the number of synapses significantly exceeds the number of neurons in the real neuromorphic system. This indicates that the area of the synapse array is much larger than that of the neurons with supportive circuits, causing challenges in layout. To enhance the layout efficiency of the synapses, a 3D vertical NAND (3D VNAND) structure can be applied. [ \n \n 49 \n , \n 50 \n \n ] It should be noted that the peri‐under‐cell (PUC) or cell‐on‐peri (COP) structure used in 3D VNAND, where their cells are fabricated on the peripheral circuits using monolithic 3D integration, is similar to our 3D neuromorphic hardware. [ \n \n 51 \n \n ] \n 2.4 Neuromorphic Vision Sensing with a Spiking Neural Network In addition to face and pattern recognition abilities, neuromorphic vision sensing (NVS) is promising for gesture classification with reduced power consumption, which is attributed to fewer redundant data transfers and information processing with the SNN. [ \n \n 52 \n , \n 53 \n \n ] To confirm that the 3D‐integrated neuromorphic hardware can be applied to NVS, we performed two types of simulations. The first was a hardware‐based circuit simulation with the aid of SPICE for simple 3 × 3 kernel operation to recognize a letter pattern using a single‐layer SNN. The second was a software‐based neural network simulation with the aid of PyTorch for complex American Sign Language (ASL) classification using a spiking convolutional neural network (spiking‐CNN). Note that the kernel operation is a basic element of the spiking‐CNN. Prior to the circuit simulation with SPICE, the 1T‐neuron must be modeled with sub‐elements. Thus, it was modeled with parallel connection consisting of a threshold switch and a parasitic capacitor. [ \n \n 22 \n \n ] The measured V \n T,firing and V \n resting from Figure 2c for the threshold switch and the extracted C \n par for the parasitic capacitor were reflected in the circuit simulations with SPICE. The 1TFT‐synapse was modeled with a three‐terminal MOSFET and the corresponding V \n T was adjusted to control w \n i . The circuit for the 3 × 3 kernel operation was composed of nine input synapses corresponding to each of nine pixels and one output neuron, with a current mirror located between the synapses and the neuron to reduce I \n out,syn . The circuit of the 3×3 kernel is identical to the circuit schematic in Figure S12 (Supporting Information) except for the number of synapses. V \n in,syn is applied to the gate of the 1TFT‐synapse to modulate w \n i , which is correlated with the pixel intensity. It became elevated at a higher pixel intensity level, as illustrated in Figure S13a (Supporting Information). As a simple example of letter pattern recognition, three different images including noisy signals were prepared. These three letter patterns were passed through three different kernels (“n”‐ kernel, “v”‐kernel, and “z”‐kernel). When the kernel matched the input image, a high f is produced at the corresponding output neuron. As shown in Figure S13b (Supporting Information), the “n”‐kernel generated the highest f for the noisy “n” pattern compared to f for the noisy “v” and “z” patterns. This outcome indicates that the “n”‐kernel successfully recognized the “n” image. In the same way, the “v”‐ and “z”‐kernels generated the highest f for the noisy “v” and “z” patterns, respectively. Prior to the network simulation, a pre‐image processing step is necessary. Most of the image classification process is usually performed using a general image dataset collected with a standard frame‐based camera. [ \n \n 13 \n , \n 17 \n \n ] However, when using a general image dataset, image pixel values must be extracted in every frame and encoded in a spike form. Few difficulties arise when using that approach, e . g ., the limited frame rate, high redundancy between frames, and high power consumption. Therefore, it is recommended to perform NVS using an image dataset obtained directly from a dynamic vision sensor (DVS). In particular, the ASL data were obtained by filming real sign language with a DVS. [ \n \n 54 \n \n ] Because the ASL dataset has 24 classes (A‐Y, excluding J) and each letter consists of 4200 samples, the total number of samples is 100 800. It is important to show whether the proposed 3D‐integrated neuromorphic hardware can classify complicated ASL. Among the total of 100 800 samples, 84 000 samples were randomly selected for the training and the others were used for the testing. \n Figure \n 5 a shows the spiking‐CNN configuration. A few exemplary gestures in ASL are shown in Figure 5b as examples. Each hand gesture has an image size of 180 × 240 pixels. The pixel values stored in ASL data were applied as inputs to the 1T‐neurons and 1TFT‐synapses. Then, LIF operations were performed by 1T‐neurons, and multiply‐and‐accumulate (MAC) calculations were conducted by 1TFT‐synapses, respectively. The spiking‐CNN consists of two spiking convolutional layers, three average pooling layers, and two fully connected layers, as summarized in Table S5 (Supporting Information). It should be noted that the measured characteristics of the underlying 1T‐neuron and the superjacent 1TFT‐synapse were reflected for the spiking‐CNN simulation to classify various hand gestures. For the neuron, the measured the spiking characteristics in Figure 2c were reflected, as mentioned above. For the synapse, the measured P/D characteristics in Figure 2g were reflected to model the conductance modulation using the following equation:\n \n (1) \n G = G max α − G min α + G min α 1 / α if α ≠ 0 G min α ( G max / G min ) if α = 0 \n where α is a parameter that controls the potentiation ( α \n pot ) or depression ( α \n dep ) and η is an internal variable ranging from 0 to 1. [ \n \n 55 \n \n ] Extracted parameters of P/D characteristics for three cases (fresh, after 204 800 update pulses without ETA, after 204 800 update pulses with ETA) were summarized in Table S6 (Supporting Information). To train the network, backpropagation by means of spike layer error reassignment in time (SLAYER) was utilized. [ \n \n 56 \n \n ] When using this method, the probability distribution function (PDF) of the state change was used for differentiating the spike function. Such a PDF change from not‐fire to fire or from fire to not‐fire decays exponentially according to the absolute value of the difference between the membrane potential and V \n T,firing . More details pertaining to the simulation are given in Note S1 (Supporting Information). Figure 5 Neuromorphic vision sensing (NVS) to classify American Sign Language (ASL). a) Configuration of the spiking convolutional neural network (spiking‐CNN) for ASL classification. b) Examples of the ASL dataset obtained by filming actual sign language with the aid of a dynamic vision sensor (DVS). c) Classification accuracy according to the number of training epochs. Three measured synaptic P/D characteristics in three states: fresh, fatigued after 204 800 weight updates, and cured with ETA after 204 800 weight updates. The classification accuracy after curing is comparable to that in the fresh state. The classification accuracy of 93.0% was obtained with a fresh 1TFT‐synapse without cyclic degradation, as shown in Figure 5c . This outcome is close to the upper limit of 93.7%, which is achievable with perfectly linear P/D characteristics ( α \n pot = α \n dep = 1) in the simulation. Successful classification is also confirmed by the Video S1 (Supporting Information), which visualizes the number of output spikes generated from each output neuron according to the time elapse. The corresponding output neuron matches the input image that generates the largest number of spikes. It should be emphasized that the classification accuracy falls to 31.4% after 204800 updates unless ETA is applied for recovery. In contrast, it only slightly decreases to 92.3% even after 204 800 updates with ETA. Finally, we demonstrated the long‐term durable NVS operation with the proposed 3D‐integrated neuromorphic devices with ETA."
} | 11,572 |
39881186 | PMC11779922 | pmc | 1,924 | {
"abstract": "Analog In-memory Computing (IMC) has demonstrated energy-efficient and low latency implementation of convolution and fully-connected layers in deep neural networks (DNN) by using physics for computing in parallel resistive memory arrays. However, recurrent neural networks (RNN) that are widely used for speech-recognition and natural language processing have tasted limited success with this approach. This can be attributed to the significant time and energy penalties incurred in implementing nonlinear activation functions that are abundant in such models. In this work, we experimentally demonstrate the implementation of a non-linear activation function integrated with a ramp analog-to-digital conversion (ADC) at the periphery of the memory to improve in-memory implementation of RNNs. Our approach uses an extra column of memristors to produce an appropriately pre-distorted ramp voltage such that the comparator output directly approximates the desired nonlinear function. We experimentally demonstrate programming different nonlinear functions using a memristive array and simulate its incorporation in RNNs to solve keyword spotting and language modelling tasks. Compared to other approaches, we demonstrate manifold increase in area-efficiency, energy-efficiency and throughput due to the in-memory, programmable ramp generator that removes digital processing overhead.",
"introduction": "Introduction Artificial Intelligence algorithms, spurred by the growth of deep neural networks (DNN), have produced the state-of-the-art solutions in several domains ranging from computer vision 1 , speech recognition 2 , game playing 3 to scientific discovery 4 , natural language processing (NLP) 5 , and more. The general trend in all these applications has been increasing the model size by increasing the number of layers and the number of weights in each layer. This trend has, however, caused growing concern in terms of energy efficiency for both edge applications and servers for training; power is scarce due to battery limits in edge devices while the total energy required for training large models in the cloud raises environmental concerns. Edge devices have a further challenge posed by strong latency requirements in applications such as keyword spotting to turn on mobile devices, augmented reality and virtual reality platforms, anti-collision systems in driverless vehicles etc. The bottleneck for implementing DNNs on current hardware arises due to the frequent memory access necessitated by the von Neumann architecture and the high memory access energy for storing the parameters of a large model 6 . As a solution to this problem, a new architecture of In-memory Computing (IMC) has become increasingly popular. Instead of reading and writing data from memory in every cycle, IMC allows neuronal weights to remain stationary in memory with inputs being applied to it in parallel and the final output prior to neuronal activation being directly read from memory. Among the IMC techniques explored, analog/mixed-signal IMC using non-volatile memory devices such as memristive 7 ones have shown great promise in improving latency, energy and area efficiencies of DNN training 8 – 10 and inference 11 – 13 , combinatorial optimization 14 , 15 , hardware security 11 , 16 , content addressable memory 17 , signal processing 18 – 20 etc. It should be noted that analog IMC does not refer to the input and output signals being analog; rather, it refers to the storage of multi-bit or analog weights in each memory cell (as opposed to using memristor for 1-bit storage 21 , 22 ) and using analog computing techniques (such as Ohm’s law, Kirchoff’s law etc.) for processing inputs. Analog weight storage 23 enables higher density of weights as well as higher parallelism (by enabling multiple rows simultaneously) compared to digital counterparts. Comparing the energy efficiency and throughput of recently reported DNN accelerators (Fig. 1 a) shows the improvements provided by IMC approaches over digital architectures. However, taking a closer look based on DNN architecture exposes an interesting phenomenon–while analog IMC has improved energy efficiency of convolutional and fully connected (FC) layers in DNNs, the same cannot be said for recurrent neural network (RNN) implementations such as long short term memory (LSTM) 24 – 28 . Resistive memories store the layer weights in their resistance values, inputs are typically provided as pulse widths or voltage levels, multiplications between input and weight happen in place by Ohm’s law, and summation of the resulting current occurs naturally by Kirchoff’s current law. This enables an efficient implementation of linear operations in vector spaces such as dot products between inputs and weight vectors. Early implementations of LSTM using memristors have focussed on achieving acceptable accuracy in network ouptut in the presence of programming errors. A 128 × 64 1T1R array 29 was shown to be able to solve real-life regression and classification problems while 2.5 M phase-change memory devices 30 have been programmed to implement LSTM for language modeling. While these were impressive demonstrations, energy efficiency improvements were limited, since RNNs such as LSTMs have a large fraction of nonlinear (NL) operations such as sigmoid and hyperbolic tangent being applied as neuronal activations (Fig. 1 b). With the dot products being very efficiently implemented in the analog IMC, the conventional digital implementation of the NL operations now serves as a critical bottleneck. Fig. 1 Limitation of current In-memory computing (IMC) for Recurrent Neural Networks and our proposed solution. a A survey of DNN accelerators show the improvement in energy efficiency offered by IMC over digital architectures. However, the improvement does not extend to recurrent neural networks (RNN) such as LSTM and there exists a gap in energy efficiency between RNNs and feedforward architectures. Details of the surveyed papers available here 66 . b Architecture of a LSTM cell showing a large number of nonlinear (NL) activations such as sigmoid and hyperbolic tangent which are absent in feedforward architectures that mostly use simple nonlinearities like rectified linear unit (ReLU). c Digital implementation of the NL operations causes a bottleneck in latency and energy efficiency since the linear operations are highly efficient in time and energy usage due to inherent parallelism of IMC. For a LSTM layer with 512 hidden unit and with k = 32 parallel digital processors for the NL operations, the NL operations still take 2–5X longer time for execution due to the need of multiple clock cycles ( N c y c ) per NL activation. d Our proposed solution creates an In-memory analog to digital converter (ADC) that combines NL activation with digitization of the dot product between input and weight vectors. As an example, an RNN transducer was implemented on a 34-tile IMC system with 35 million phase-change memory (PCM) devices and efficient inter-tile communication 31 . While the system integration and scale of this effort 31 is very impressive, the NL operations are performed off-chip using low energy efficiency digital processing reducing the overall system energy efficiency. Another pioneering research 23 integrated 64 cores of PCM arrays for IMC operations with on-chip digital processing units for NL operations. However, the serial nature of the digital processor, which is shared across the neurons in 8 cores, reduced both the energy efficiency and throughput of the overall system. This work used look-up tables (LUT), similar to other works 32 – 34 ; alternate techniques using cordic 35 or piece wise linear approximations 24 , 36 – 38 or quadratic polynomial approximation 39 have also been proposed to reduce overhead and latency ( N c y c ) of computing one function. However, it is the big difference in parallelism of crossbars versus serial digital processors which causes this inherent bottleneck. Even in a hypothetical situation with an increased number of parallel digital computing engines for the NL activations (Fig. 1 c, Supplementary Note S6 ), albeit at a large area penalty, the latency of the NL operations still dominates the overall latency due to the extremely fast implementation of vector-matrix-multiplication (VMM) in memristive crossbars. In this work, we introduce an in-memory analog to digital conversion (ADC) technique for analog IMC systems that can combine nonlinear activation function computations in the data conversion process (Fig. 1 d), and experimentally demonstrate its benefits in an analog memristor array. Utilizing the sense amplifiers (SA) as a comparator in ramp ADC, and creating a ramp voltage by integrating the current from an independent column of memristors which are activated row by row in separate clock cycles, an area-efficient in-memory ADC for memristive crossbars is demonstrated. However, instead of generating a linear ramp voltage as in conventional ADCs, we generated a nonlinear ramp voltage by appropriately choosing different values of memristive conductances such that the shape of the ramp waveform matches that of the inverse of the desired NL activation function. Using this method, we demonstrate energy-efficient 5-bit implementations of commonly used NL functions such as sigmoid, hyperbolic tangent, softsign, softplus, elu, selu etc. A one-point calibration scheme is shown to reduce the integral nonlinearity (INL) from 0.948 to 0.886 LSB for various NL functions. Usage of the same IMC cells for ADC and dot-product also gives added robustness to read voltage variations, reducing INL to ≈0.04 LSB compared to ≈5.0 LSB for conventional methods. Using this approach combined with hardware aware training 40 , we experimentally demonstrate a keyword spotting (KWS) task on the Google speech commands dataset (GSCD) 41 . With a 32 hidden neuron LSTM layer (having 128 nonlinear gating functions) that uses 9216 memristors from the 3 × 64 × 64 memristor array on our chip, we achieve 88.5% accuracy using a 5-bit NL-ADC with a ≈9.9 X and ≈4.5 X improvement of area and energy efficiencies respectively at the system level for the LSTM layer over previous reports. Moreover, compared to a conventional approach using the exact same configuration (input and output bit-widths) as ours, the estimated area and energy efficiency advantages are still retained at c ≈6.2 X and ≈1.46 X respectively for system level of evaluation. Finally, we demonstrate the scalability of our system by performing a character prediction task on the Penn Treebank dataset 42 using a LSTM model ≈100X bigger than the one for KWS using experimentally validated nonideality models and achieving software equivalent accuracy. The improvements in area efficiency are estimated to be 6.6 X over a conventional approach baseline and 125 X over earlier work 31 at the system level, with the drastic increase in performance due to the much higher number of nonlinear functions in the larger model.",
"discussion": "Discussion In conclusion, we proposed and experimentally demonstrated a novel paradigm of nonlinear function approximation through a memristive in-memory ramp ADC. By predistorting the ramp waveform to follow the inverse of the desired nonlinear activation, our NL-ADC removes the need for any digital processor to implement nonlinear activations. The analog conductance states of the memristor enable the creation of different programmable voltage steps using a single device, resulting in great area savings over a similar SRAM-based implementation. Moreover, the in-memory ADC is shown to be more robust to voltage fluctuations compared to a conventional ADC with memristor crossbar based MAC. Using this approach, we implemented a LSTM network using 9216 weights programmed in the 72 × 128 memristor chip to solve a 12-class keyword spotting problem using the GSCD. The results for the 5-bit ADC show better accuracy of 88.5% than previous hardware implementations 31 , 49 with significant advantages in terms of normalized area efficiency (≈9.9 X ) and energy efficiency (≈4.5 X ) compared to previous LSTM circuits. We further tested the scalability of our system by simulating a much larger network (6,112,512 weights) for NLP using the experimentally validated models. Our network with 5-bit NL-ADC again achieves better performance in terms of BPC than recent reports 23 of IMC based LSTM ICs while delivering ≈86 X and ≈24.4 X better area and energy efficiencies at the system level. Our work paves the way for very energy efficient in-memory nonlinear operations that can be used in a wide variety of applications."
} | 3,172 |
37636979 | PMC10448690 | pmc | 1,927 | {
"abstract": "In recent years,\nwith the increasing scarcity of fossil\nresources\nand the worsening environmental pollution, the effective utilization\nof wood and plastic waste has become a critical issue. In this paper,\npropylene glycol (PG) was used as an alcoholysis agent to degrade\nwaste poly(ethylene terephthalate) (PET), and unsaturated polyester\n(UPR) was synthesized by the polycondensation reaction. The Chinese\nfir was modified by chemical impregnation to obtain a new type of\nwaste PET-based wood–plastic composites. It exhibits a compressive\nstrength of about 107 MPa and a water absorption of less than 20%.\nThese results highlight the outstanding modification effect on fir,\ndemonstrating excellent mechanical properties and corrosion resistance.\nThis study presents a green and efficient method for the preparation\nof wood–plastic composites and the recycling of waste PET,\nproviding promising solutions for sustainable resource utilization\nand environmental protection.",
"conclusion": "4 Conclusions In this paper, a p -benzene-type structure was\nintroduced into the impregnation system by the method of depolymerizing\nwaste PET with glycol. The chemical structure lays a foundation for\nthe improvement of heat resistance, corrosion resistance, and mechanical\nproperties of modified wood. The curing characteristics of PET-based\nUPR were comprehensively evaluated with gel time, curing time, exothermic\npeak temperature, and gel content as indicators to achieve the purpose\nof optimizing the UPR impregnation system and the wood modification\nprocess. The FTIR analysis results of PET raw materials and\nproducts of\neach stage show that during the PET dihydric alcoholysis and UPR synthesis\nprocess, the characteristic groups such as alkyl groups, ester groups,\nand benzene ring para-position double substitution structures in the\nPET raw material structure were successfully added to the products\nof subsequent stages, and the UPR oligomer with the p -phenylene structure was successfully synthesized. The participation\nof unsaturated anhydride and the active diluent system provides the\npreconditions for the cross-linking and curing of UPR. The XPS\nanalysis results of waste PET-based UPR-impregnated modified\nwood stated that the relative content of UPR in the modified wood\ndecreased gradually from the outer chord section to the inner chord\nsection, which showed that the impregnation method of UPR is mainly\ntransverse infiltration with a decreasing trend of permeability from\noutside to inside. It shows that the impregnation method of UPR is\nmainly transverse infiltration with a decreasing trend of permeability\nfrom outside to inside. The changes in the micromorphology observed\nby SEM indicated that the microstructure of the modified wood was\nnot damaged. The UPR impregnation system fills and adheres to the\nvessel molecules, wood fibers, tracheids, wood rays, and other cells\nof the wood structural tissue and undergoes cross-linking and solidification.\nThe pit openings on the cell wall are blocked, which greatly improves\nthe water absorption resistance and mechanical strength of fast-growing\nwood. The results of thermogravimetric analysis show that the thermal\ndecomposition initiation temperature of the waste PET-based UPR-impregnated\nmodified wood is about 30 °C higher than that of the log and\ncan maintain relatively high thermal stability. This article\nsolves the cyclic utilization of waste PET and the\nenhancement and modification of fast-growing wood. The modification\neffect is comparable to that of commercially available 191-type resin\nderived from fossil resources, which contains corrosion resistance,\nhigh mechanical strength, and thermal stability. This achievement\nembodies the scientific concept of green chemistry, aligning with\nthe principles of sustainability and environmental responsibility\nin the realm of academic research.",
"introduction": "1 Introduction Wood is a natural and\nrenewable porous polymer material. As a green\nand environmentally friendly material, it has always been favored\nby people and is widely used in construction, industrial component\nconstruction, furniture production, indoor and outdoor decoration,\nand paper industries. However, due to the over-exploitation of natural\nforests in the past and the implementation of logging ban protection\nmeasures in recent years, the supply of high-quality wood is far from\nenough to meet the market demand. 1 Plantation\ntrees not only have a short growth cycle but also have natural defects\nsuch as loose material, low mechanical strength, poor dimensional\nstability, easy mildew, and corrosion, which greatly limit their application\nrange. 2 , 3 PET is a high-performance engineering\npolymer with excellent mechanical\nand barrier properties; thus, it is used in large amounts nowadays,\nand it is widely used in plastic packaging bottles, films, and synthetic\nfibers. 4 − 6 Excellent comprehensive properties make PET the third\nmost widely used polymer in the packaging industry, almost monopolizing\nthe entire beverage packaging bottle market. 7 PET products such as plastic packaging bottles and polyester fibers\nhave become an important part of living. The production of PET worldwide\nis also increasing year by year, and the annual output has increased\nfrom 42 million tons in 2014 to 74 million tons in 2020. 8 − 10 The design life of plastic is usually 1 to 50 years, but most plastic\npackaging materials lose their value after a short period of use. 11 Most plastics are difficult to decompose quickly\nin the ecosystem and easily accumulate in the environment after use,\nwhich causes serious environmental damage, known as “plastic\npollution”. 12 In the chemical\ncyclic utilization process of PET, the polymer\nis depolymerized into oligomers or monomers, which are then separated,\npurified, and used as raw materials for the production of chemical\nproducts. Chemical recycling methods include hydrolysis, methanolysis,\nglycolysis, amine hydrolysis, ammonia hydrolysis, and other methods. 4 , 13 Hydrolysis is the only method that directly depolymerizes PET into\nTPA (terephthalic acid) and EG (ethylene glycol), and the reaction\ncan be carried out in water media with different acid–base\nconditions. Since the technology for synthesizing PET directly from\nTPA and EG through esterification is well-established, and the production\ncost of TPA from para-xylene (PX) oxidation is relatively high, the\nrecovery of TPA from waste PET materials through hydrolysis is of\ngreat significance. Hydrolysis can be classified into alkaline hydrolysis, 14 acid hydrolysis, 15 and neutral hydrolysis. 16 Based on the\ndifferent alcoholysis agents, alcoholysis methods are generally divided\ninto glycolysis 17 and methanolysis, 18 which are the two most commercially mature depolymerization\nmethods. The amine hydrolysis reaction of PET mainly produces TPA\ndiamide and EG, and the most commonly used amine hydrolysis agents\nare methylamine, ethylamine, ethanolamine, and other primary amines,\nas well as ethylenediamine and other diamines, which are generally\nconducted within the temperature range of 20–100 °C. 16 , 19 In the related research of chemical recycling of waste PET, the\nmain objectives are to improve the conversion rate and monomer yield\nof PET, shorten the reaction time, and conduct the reaction under\nmilder conditions as much as possible. In recent years, numerous efforts\nhave focused on the application of emerging technologies such as microwave\nheating, 20 supercritical fluids, 18 , 21 ionic liquids, 22 − 24 deep eutectic solvents, 25 , 26 etc. These approaches aim to address issues such as low thermal\nconductivity of PET, inadequate reaction contact, improvement of heat\ndistribution in reactors, increase of depolymerization energy, enhancement\nof the effective contact area between PET and solvents/catalysts,\nand achievement of higher cyclic utilization efficiency. Efficient\ncyclic utilization of discarded PET products to achieve\nrenewable resource utilization is a major direction for the development\nof the green chemical industry. Duque-Ingunza et al. 27 catalyzed the ethanolysis of PET with Na 2 CO 3 , and the purified product was polycondensed with maleic anhydride\nto produce a linear unsaturated polyester. The resin was then cross-linked\nand cured using peroxide acetone and cobalt 2-ethylhexanoate, achieving\nthe performance standards of commercial unsaturated polyester resins.\nHowever, the application of the polyester in some biomaterials has\nnot been studied. Abdullah and Ahmad 28 synthesized\nUPR from ethanolysis of discarded PET and blended it with 0.3% coconut\nfibers to prepare coconut fiber-reinforced composite materials. In\nthis study, the proportion of coconut fibers was relatively low, leaving\nroom for further utilization of biomaterials. Wang et al. 29 reacted bifunctional epoxy oligomers with poly(ethylene\nglycol) to produce a waterborne epoxy resin, which was then impregnated\ninto wood to study its related properties. For the impregnation modification\nof fir wood, Wang et al. 29 employed phenol–formaldehyde\n(PF) resin for chemical impregnation, while Liu et al. 30 used furfuryl alcohol (FA) resin for the modification\nof fir and poplar wood. However, the organic compounds used in these\nstudies not only pose harm to human health but also do not adhere\nto the concept of green chemistry. To the best of our knowledge, there\nhave been few studies on the application of UPR prepared from discarded\nPET in wood, achieving the combination of sustainable and green resource\nutilization, broadening the application range of fast-growing wood,\nand benefiting resource conservation and environmental protection. We will establish a connection between wood and waste PET products,\nusing propylene glycol as an alcoholysis agent and alcoholysis under\nhigh-temperature conditions; the PET ester bond is broken and replaced\nby hydroxyl end groups, and then, the addition of maleic anhydride\n(MA) and phthalic anhydride (PA) introduces double bonds and benzene\nrings into the molecular chain to obtain unsaturated polyester oligomers.\nThe schematic diagram of the reaction is shown in Figure 1 . Maleic anhydride provides\nactive functional groups for the subsequent cross-linking of UPR oligomers;\nphthalic anhydride is used to adjust the double-bond content in the\nmolecular chain of UPR oligomers, reduce the structural regularity,\nand increase the compatibility of UPR oligomers with cross-linking\nmonomers while controlling its cross-linking density. Then, diluent\nstyrene (St), cross-linking agent divinylbenzene (DVB), and initiator\nbenzoyl peroxide (BPO) were added to the oligomer to prepare an impregnating\nliquid, which was modified by chemical impregnation. 31 The impregnation liquid was injected into the wood voids\nand cured. The result is a wood–plastic composite material\nwith significantly improved mechanical properties, water absorption\nperformance, and thermal stability. This further enhances the service\nlife of the wood, improves its corrosion resistance and insect resistance,\nand provides a new solution for the recycling and utilization of waste\nPET plastics. It should be noted that the curing process of UPR is\nextremely complicated. The C=C bond on the linear unsaturated\npolyester molecular chain undergoes a free radical copolymerization\nreaction with styrene, which can form four possible cross-linking\nstructures: (i) intermolecular cross-linking with or without the styrene\nmonomer; (ii) intramolecular cross-linking with or without the styrene\nmonomer; (iii) branching on polyester molecules by styrene; and (iv)\na free styrene homopolymer. 32 Furthermore,\nthe degree of polymerization of polystyrene between the cross-linked\nchains is different, and the existence of the steric hindrance effect\nalso leads to some unreacted unsaturated double bonds in the cured\nproduct of the unsaturated polyester, which makes the resin-cured\ncross-linked network highly chaotic and difficult to express clearly\nwith chemical reaction formulas. Figure 1 Schematic diagram of the reaction of PG\nalcoholysis of PET and\nthe synthesis of UPR oligomers.",
"discussion": "3 Results\nand Discussion 3.1 FTIR Analysis of PET Raw\nMaterials and Products\nat Each Stage Figure 4 shows the infrared spectra of industrial PET pellets, waste\nPET recycled materials, and their corresponding propylene glycol alcoholysis\nproducts and UPR oligomers from top to bottom. It can be seen that\nthe characteristic absorption peaks of industrial PET pellets and\nwaste PET recycled materials are completely consistent. The double\nabsorption peaks of C–H symmetric and antisymmetric stretching\nvibration near 2950 cm –1 indicate the existence\nof alkyl chains. 1720 cm –1 is the characteristic\npeak of the carbonyl group, and C–O–C symmetric and\nantisymmetric stretching vibration doublets appear at 1245 and 1110\ncm –1 , respectively, indicating the existence of\nester groups. The C–H stretching vibration of the para-disubstituted\nbenzene ring appears near 880 and 730 cm –1 , indicating\nthat there is a para-substituted benzene ring in the structure. The\ncomparison results of characteristic peaks show that there is no difference\nin the chemical composition structure of industrial PET pellets and\nwaste PET recycled materials, and they all completely accorded with\nthe characteristic band properties of PET. Figure 4 Infrared spectra of PET\nraw materials and corresponding products.\nIn the figure, W-PET represents waste PET, W-P represents the alcoholysis\nproducts of waste PET, P-S/D represents the UPR oligomer synthesized\nfrom industrial PET, and W-P-S/D represents the UPR oligomer synthesized\nfrom waste PET. Figure 4 shows the\nC–H stretching vibration absorption peak of the alkyl group\nin the PET structure and the C=O stretching vibration absorption\npeak of the ester group. The C–H out-of-plane bending vibration\nabsorption peaks of the para-disubstituted benzene ring appear at\nthe same frequency position in the spectra of PET alcoholysis products\nand UPR oligomers, indicating that the PET glycolysis and UPR synthesis\nprocess successfully introduced some characteristic groups in the\nPET raw material structure into the subsequent stage products. In\nthis experiment, UPR oligomers with a p -benzene structure\nwere successfully synthesized. The introduction of unsaturated acid\nanhydrides and a reactive dilution system provided conditions for\nthe cross-linking and curing of UPR. 3.2 Research\non the UPR Impregnation System 3.2.1 Analysis\nof Factors Influencing the Viscosity\nof the UPR Oligomer Mixture The viscosity of the UPR oligomer\nmixture to a certain extent reflects the impregnation ability of the\nresin system, directly impacting the performance of modified wood\nas a subsequent impregnation solution. In the synthesis process of\nUPR, both the dilution system and the PET raw material can affect\nthe viscosity of the UPR oligomer mixture. This study referred to\nthe curing method of commercial resin type 191 and selected styrene\nas the diluent 40 and DVB as the cross-linking\nagent. The selection of DVB was based on its unique double-bond structure,\nwhich allows for copolymerization with styrene, facilitating the introduction\nof functional groups while maintaining good strength. This has been\nwidely applied in ion exchange resins. 41 , 42 The dilution\nsystem used in the UPR synthesis experimental scheme consisted of\ntwo active diluents, St and MMA, and one cross-linking agent, DVB,\nin the designed mass ratio of the UPR oligomer, as detailed in Table 2 . Table 2 Combination of Experimental Conditions\nfor Determination of UPR Curing Reactivity UPR oligomer\nmixture code dilution system type of initiator P-S 30%St BPO P-S/D 30%St + 2%DVB BPO P-2S/M 20%St + 10%MMA BPO The\nviscosity of the UPR oligomer mixture is highest\nwhen solely\nusing St as the dilution system, accounting for 30% of the total UPR\noligomer content, as shown in Figure 5 a. When an additional 2% of the cross-linking agent\nDVB is added to the dilution system, the viscosity of the UPR oligomer\nmixture noticeably decreases. Furthermore, when the dilution system\nconsists of 20% St and 10% MMA, the viscosity of the UPR oligomer\nmixture decreases significantly. From Figure 5 b, it can be observed that the variation\nin PET raw materials has a relatively small impact on the viscosity\nof the UPR oligomer mixture. The UPR oligomer mixture synthesized\nfrom recycled PET waste through alcoholysis exhibits higher viscosity\nthan industrial PET granules, which may be attributed to the use of\nrecycled materials. The experimental substitution of MMA for a portion\nof styrene as a diluent in the comprehensive dilution conditions aims\nto investigate whether MMA can replace styrene as a new diluent. The\nresults of this study will be discussed in the following sections. Figure 5 (a) Influence\nof dilution system on the viscosity of the UPR oligomer\nmixture. (b) Effect of the PET raw material on the viscosity of the\nUPR oligomer mixture. 3.2.2 Influence\nof the Dilution System on UPR\nCuring Reactivity The curing reactivity of UPR was comprehensively\nevaluated by gel time, curing time, and exothermic peak temperature.\nThe gel time and curing time are relatively short, and the curing\nreactivity of the UPR oligomer system with a higher exothermic peak\ntemperature is relatively higher. In this experiment, St and MMA were\nused as active diluents for UPR oligomers, DVB was used as a cross-linking\nagent, and BPO was used as an initiator. The effect of the dilution\nsystem on the curing reactivity of UPR was investigated. Table 2 summarizes the combination\nof experimental conditions for the determination of UPR curing reactivity. Table 3 lists the\ncorresponding evaluation indicator data. When BPO is used as the initiator,\nthe exothermic peak temperature of P-S/D is the highest, but its gel\ntime and curing time are slightly longer than those for the other\ntwo systems ( Figure 6 ). The addition of MMA as a diluent reduces the viscosity of the\nsystem, as shown in Figure 5 a, but its performance in the curing reactivity experiment\nis not rational. The incorporation of the cross-linking agent DVB\nslightly extends the gel time and curing time, while it not only reduces\nthe viscosity of the system ( Figure 5 a) but also increases the exothermic peak temperature.\nIn a comprehensive comparison, the curing reactivity of P-S/D is better.\nComprehensive analysis shows that the curing reactivity of P-S/D is\nbetter, and adding 2% of the bifunctional cross-linking agent DVB\nto the dilution system can significantly improve the curing reactivity\nof UPR. Figure 6 Comparison of the UPR curing activity of the PG dilution system.\nIn the figure, P-S represents the complete dilution using styrene\nas the diluent, P-2S/M is the dilution with partial substitution of\nMMA for styrene, and P-S/D is the addition of the cross-linking agent\nDVB on the basis of styrene dilution. The specific amounts of additives\nare shown in Table 2 . Table 3 UPR Curing Reaction\nActivity Data\nof Dilution System curing UPR exothermic peak temperature (°C) gel time(s) curing\ntime(s) P-S 141.8 814 904 P-S/D 147.3 849 979 P-2S/M 112.0 781 892 3.2.3 Influence of PET Raw Materials on UPR Gel\nContent According to the best plan of UPR curing, the UPR\nsystem synthesized from industrial PET pellets and waste PET recycled\nmaterials was cured at 80 °C for 3 h, showing degrees of cure\nof 86.5 and 81.8%, respectively. The reason for the slight decrease\nin the gel content of the UPR product may be the aging or oxidation\nof the recycled material due to light, heat, abrasion, and other factors\nduring the use process. ( Table 4 ) Even so, the gel content of UPR synthesized from PET recycled\nmaterials still reaches more than 80%, and it still has a high curing\nability, which can be used for the research of impregnation and modification\nof fast-growing wood. Table 4 Comparison of the\nGel Content of UPR\nSynthesized from the PET Industrial Material and the Recycled Material a cured UPR designation curing temperature (°C) curing time (h) gel content\n(%) standard deviation of gel content P-S/D 80 ± 2 3 86.5 0.17 W-P-S/D 80 ± 2 3 81.8 0.25 a In the table, P-S/D represents industrial\nPET pellets and W-P-S/D represents waste PET recycled materials. 3.3 Properties\nof UPR-Impregnated Modified Wood 3.3.1 Effect\nof the UPR Impregnation System on\nthe Strengthening Properties of Modified Wood Table 5 shows the impregnation rate,\ncuring weight gain rate, and water absorption resistance data of different\nUPR impregnation systems in Chinese fir modification. It is found\nthat the impregnation rate of PET-based UPR-impregnated modified fir\nis about 190%, the curing weight gain rate is 160–180%, and\nthe water absorption resistance rate is 88–90%. In comparison,\nthe waste PET-based UPR impregnation system and the industrial PET-based\nUPR impregnation system have significant effects on the impregnation\nweight gain and water absorption resistance of Chinese fir, which\nare almost the same. In the case of modified wood impregnated with\n191-type commercially available UPR resin, it has been observed that\nthe impregnation rate, curing weight gain rate, and water absorption\nresistance are marginally superior. Upon analysis, it has been deduced\nthat the primary cause for this disparity lies in the substantially\nlower viscosity of the commercial UPR compared to the PET-based UPR\nsystem. Consequently, this results in relatively enhanced fluidity\nand wetting properties. Table 5 Comparison of Properties\nof Modified\nChinese Fir with Different UPR Impregnation Systems UPR systems wood type viscosity\n(Pa·s) P log (g/cm 3 ) P curing (g/cm 3 ) IY(%) WPG (%) WRE (%) P-S/D S 4.0 0.36 0.98 193.1 182.7 88.6 W-P-S/D S 4.7 0.34 0.92 192.5 168.8 90 191-UPR S 0.6 0.33 1.11 270.7 233.4 92.8 Figure 7 shows the\nwater absorption and compressive strength curves of waste PET-based\nUPR-modified wood and logs. In Figure 7 a, the water absorption of Chinese fir in 8 days is\n167.3%, which is extremely unfavorable for the waterproof and anticorrosion\nrequirements in the subsequent processing and application of wood.\nAfter impregnation and modification of waste PET-based UPR, the water\nabsorption rate of Chinese fir in 8 days dropped to less than 20%,\nand the water absorption resistance of fast-growing wood was greatly\nimproved. The compressive strength curves of Chinese fir before and\nafter modification are shown in Figure 7 b: the compressive strength of unmodified Chinese fir\nis 35.4 MPa, and the compressive strength of modified Chinese fir\nis 107.1 MPa, significantly improved. Figure 7 (a) Water absorption rate–time\ncurves of modified woods\nand (b) compressive strength–time curves of modified woods.\nIn the figure, S represents the unmodified spruce wood sample, and\nthe codes and synthesis conditions for other wood–plastic composite\nmaterials can be found in Table 1 . 3.3.2 Thermal\nStability Analysis of UPR-Impregnated\nWood The thermogravimetric curve reflects the weight loss\nprocess of the sample as the temperature increases. Figure 8 shows the comparison of the\nthermogravimetric curves of the modified wood sample and the log sample,\nand the modified wood sample and the cured UPR sample. Figure 8 TGA and DTG curves of\nmodified wood, cured UPR, and logs. In the\nfigure, S represents the unmodified spruce wood, W-S-P refers to the\nwood–plastic composite material of modified spruce wood using\nW-P-S/D, and W-P-S/D represents the cured sample of waste PET-based\nunsaturated polyester resin. It can be found from Figure 8 that the samples all have a relatively obvious\nweight loss\nprocess, and the weight loss process of the modified wood is slightly\nshifted to the high-temperature direction compared with the log. This\nweight loss process of curing UPR is slightly shifted toward higher\ntemperatures compared to modified wood. The thermal decomposition\nof cured unsaturated polyester resin (W-P-S/D) can be explained through\nthree main stages. The initial decomposition is observed before 250\n°C, which is due to the evaporation of moisture and uncured materials\non the surface of the resin, with a weight loss of around 8 wt %. 43 , 44 Subsequently, between 270 and 480 °C, a steep weight loss is\nexhibited, with the maximum DTG peak at approx. 394 °C. This\nis primarily due to the high temperature, causing rapid decomposition\nand volatilization of the cured UPR until the TGA and DTG curves return\nto their original baseline. Above 500 °C, a gradual and slow\ndecrease in mass is observed, which is attributed to the slow oxidation\nof thermally stable carbon produced during the decomposition process. 45 According to the change of weight loss and weight\nloss rate of the sample with the change of temperature, the thermogravimetric\nanalysis results of the log and modified wood are listed in Table 6 . Table 6 Thermogravimetric Analysis Results\nof Logs and Modified Woods stage\n1 stage\n2 stage\n3 stage\n4 sample\nname temperature range (°C) weight loss rate temperature range (°C) weight loss\nrate (%) temperature range (°C) weight loss rate (%) temperature range (°C) weight loss\nrate (%) S 21–125 4.14 125–220 0.5 220–420 65.31 420–800 10.16 W-S-P 21–116 1.22 116–251 7.76 251–456 74.49 456–800 4.99 It can be found from Table 6 that the weight loss change process of fir\nlog and waste\nPET-based UPR-modified wood can be divided into four stages. The initial\nstage of weight loss from room temperature to 100 °C is mainly\ncaused by the evaporation of residual moisture in the wood cell walls.\nThe first stage is the dehydration stage. The weight loss rate of\nthe modified wood at this stage is significantly lower than that of\nthe log, which shows that under the same sample treatment conditions\nand storage environment, the modified wood can absorb less water with\na lower saturation point, which refers to better water absorption\nresistance. There is a large difference between the weight loss\nrate in the\nsecond stage. The quality of logs remains basically constant in the\ntemperature range of 125–220 °C, which is a stage of slight\nweight loss. The weight loss rate of modified wood is 7.76% in the\ntemperature range of 116–251 °C, mainly because a small\npart of the uncured UPR oligomers in the impregnation system have\npoor thermal stability, and weight loss occurs first. 46 The third stage is the thermal decomposition stage,\nwhich is the\nmain stage of log and modified wood weight loss. The weight loss rate\nof logs in the temperature range of 220–420 °C is 65.3%,\nmainly due to the pyrolysis of cellulose and hemicellulose. 47 The weight loss rate of modified wood is 74.49%\nin the temperature range of 251–456 °C, which is a mixed\npyrolysis process of cellulose and hemicellulose of the wood itself\nand the UPR impregnation system after curing. The fourth stage\nis the carbonization stage. The residual substances\nin the log and modified wood continue to be slowly pyrolyzed until\ncarbonization, and the weight loss rate of the sample is low. Comparing the thermal decomposition temperature range between the\nuntreated spruce wood (S) and the waste PET-modified wood (W-S-P),\nit was observed that the thermal decomposition temperature of W-S-P\nwas approx. 30 °C higher than that of S. This indicates that\nthe modified wood exhibits relatively higher thermal stability. The\nthermogravimetric analysis further confirms that the impregnation\nmodification of waste PET-based UPR resin has improved the thermal\nstability of fast-growing wood. 3.3.3 XPS\nAnalysis of UPR-Impregnated Modified\nWood Chinese fir logs and Chinese fir-modified samples were\nselected for XPS characterization. Figures 9 and 10 are the XPS\nfull scan spectrum and the C 1s high-resolution spectrum of each sample,\nrespectively. It can be seen from the figure that the binding energy\nis 283–290 eV, and there are strong absorption peaks near 532\neV, indicating that the surface of Chinese fir logs and modified Chinese\nfir samples contains a large amount of C and O elements. The bonding\nstate of C atoms in wood with other atoms or atomic groups can be\ndivided into four forms. 48 , 49 In Figure 10 , the C atoms of the C1 component\ncorrespond to aliphatic and aromatic carbon chains. C atoms are only\ncombined with C or H atoms, mainly derived from lignin with the phenylpropane\nstructure in logs, wood extracts, and the main chain structure in\nUPR. Its electron binding energy is about 284.8 eV. Figure 9 Survey XPS spectrum of\nChinese fir log and waste PET-based UPR-impregnated\nmodified Chinese fir samples. Figure 10 High-resolution\nC 1s XPS spectrum of Chinese fir log and\nwaste\nPET-based UPR-impregnated modified Chinese fir samples. The C2 component corresponds to the combination\nof C–O,\nand both cellulose and hemicellulose molecules in wood have a large\nnumber of C atoms connected to hydroxyl groups. Therefore, this combined\nstate represents the chemical structural characteristics of cellulose\nand hemicellulose in wood. The electron binding energy is about 286.4\neV. The C atom of the C3 component is related to the connection\nof\ntwo non-carbonyl-like O atoms or one carbonyl-like O atom, mainly\nderived from cellulose and hemicellulose. 50 The oxidation state of C is higher in the O–C–O structure.\nThe electron binding energy is about 288 eV, which will give a significant\nchemical shift. The C atom of the C4 component is connected\nto a non-carbonyl-like\nO atom and a carbonyl-like O atom and has a high oxidation state with\na binding energy of about 289 eV, which can produce a large chemical\nshift. Its main sources are fatty acids, acetic acid, and other substances\nin wood extracts and ester groups in unsaturated polyester resins. In Table 7 , from\nthe chord section of the log sample to the inner, middle, and outer\nchord sections of the modified wood sample, the changes in the relative\ncontent of carbon atoms in different binding modes were observed.\nIt can be seen that the carbon content ratio of C–C and O–C=O\nrelative to the total carbon gradually increases, while the relative\ncarbon content of C–O and O–C–O gradually decreases.\nAt the same time, the total oxygen content and O/C of the samples\nshowed a decreasing trend in the gradient direction from the log chord\nsection to the inner, middle, and outer chord sections of the modified\nwood. The analysis showed that the relative content of cellulose and\nhemicellulose on the surface of modified Chinese fir decreased gradually\nfrom the inner chord section to the outer chord section compared with\nthe fir log. After analysis, the reason for this changing trend is\nthat the UPR with C–C and O–C=C as the characteristic\nlink mode shows an increasing trend in this gradient direction. Moreover,\nthe relative concentration of UPR on the outer chord section is slightly\nhigher than that in the middle chord section, and that on the middle\nchord section is slightly higher than that in the inner chord section.\nThe impregnation method of the resin is mainly transverse infiltration\nwith a decreasing trend of permeability from outside to inside. Table 7 Contents of Chemical Elements on the\nSurface of Chinese Fir Log and Modified Chinese Fir Samples sample serial number C(%) C–C(%) C–O(%) O–C–O(%) O–C=O(%) O(%) O/C(%) fir log chord section 68.86 32.49 29.92 4.71 1.74 29.37 42.65 modified fir inner chord section 67.61 32.09 27.67 4.37 3.48 30.65 45.33 modified fir middle chord section 69.20 38.09 23.24 3.83 4.04 28.64 41.39 modified fir outer chord section 71.38 50.93 10.44 3.96 6.05 24.40 34.18 3.3.4 SEM Analysis of UPR-Impregnated\nModified\nWood Figure 11 shows the scanning electron microscopy photos of the internal chord\nsection samples of Chinese fir logs and modified Chinese fir W-S-P-B6.\nIt can be seen from the figure that there are no obvious attachments\nin the duct molecules, wood fibers, and wood ray cells on the inner\nchord section of the log, and the pit openings and pit cavities are\nopen, which provides a channel and place for the waste PET-based UPR\nsystem to penetrate into the cell cavity, intercellular space, pit,\nand other voids. In contrast, on the inner chord section of the modified\nwood, the vessel molecules, wood fibers, and wood ray cells were partially\nfilled and adhered with translucent resin. The pit orifices were almost\ncompletely blocked by the resin, which indicates that the UPR system\nwas successfully impregnated into the wood without destroying the\nmicrostructure of the wood. Figure 11 SEM pictures of Chinese fir log and waste PET-based\nUPR-impregnated\nmodified Chinese fir samples. (a–c) SEM images of the internal\nchord section samples of Chinese fir logs. (d–f) SEM images\nof the modified Chinese fir internal chord section samples. (a) and\n(d) are 1000 times magnification, and (b), (c), (e), and (f) are 5000\ntimes magnification. The pits are the channels\nfor water and other liquids\nto flow between\nthe cells of the microscopic tissues of the wood. The UPR impregnation\nsystem blocks most of the pits in the wood, which reduces the permeability\nof the modified wood and greatly improves the water absorption resistance.\nWood fibers (hardwood) and tracheids (softwood) in the wood structure\nplay the main role of mechanical support. The UPR system fills, adheres\nto the cell walls of these cells, and further cross-links and solidifies,\nwhich in turn increases the compressive strength of the modified wood."
} | 8,351 |
27019707 | PMC4786949 | pmc | 1,928 | {
"abstract": "Using a surface-based mimic of a magnetosome interior, the biomineralisation protein Mms6 was found to be a more effective nucleator than binder of magnetite nanoparticles, and performs better than its C-terminal region alone.",
"introduction": "1. Introduction Iron is an essential element in many organisms. It plays a vital role in critical biological processes. 1 – 3 A host of proteins have evolved to coordinate, transport, and harness its useful chemical properties. Examples include: haemoglobin for oxygen transport in erythrocytes, iron storage proteins such as ferritin, and in enzymes which use the change in oxidation state of iron as the basis of electron transport and redox reactions. 1 , 2 Although an essentially useful transition metal, the presence of iron within cells is strictly controlled. High levels of iron can result in the production of damaging oxygen radicals 3 or biogenic iron oxide particles associated with neurodegenerative disorders such as Alzheimer's disease. 4 , 5 However, some organisms have developed methods to exploit the inherent magnetic characteristics of certain iron oxides by forming magnetic nanoparticles. A convenient model system to study iron accumulation and subsequent biomineralisation is the controlled formation of magnetite nanoparticles in magnetotactic bacteria (MTB). 6 – 8 \n These specialised bacteria contain internal vesicle structures termed magnetosomes 9 , 10 which act as nanoreactors for the synthesis of precise nanoparticles of the iron oxide magnetite (Fe 3 O 4 ). 8 , 11 Crystallisation of magnetite is closely regulated by dedicated biomineralisation proteins located within the lipid membrane of the magnetosome. 12 , 13 These proteins control many aspects of the forming crystal, from its specific nucleation to the size and shape of the resulting particle. Within a single MTB strain a highly uniform population of nanoparticles is produced with homogeneous size, shape, and chemical composition. However, between strains these properties can differ significantly. There is intense ongoing research to identify and understand the role of biomineralisation Mms (magnetosome membrane specific) proteins and generate detailed mechanisms for iron oxide biomineralisation. Several key proteins have been discovered tightly bound to the magnetite particles of the MTB Magnetospirillum magneticum AMB-1. 13 One of these, Mms6, is a 6 kDa protein comprising a hydrophobic N-terminal region and a hydrophilic C-terminal region (KSRDIESAQSDEEVELRDALA) rich in acidic residues. 13 – 15 When the Mms6 gene is deleted from MTB, the resulting nanoparticles are smaller and the shape is less well controlled. 16 Importantly, this protein has been used as an additive in synthetic chemical precipitations of magnetite nanoparticles, where it also appears to affect their size, formation and mineral type. 13 , 17 – 19 This has led to the assignment of Mms6 as a key size and morphology controlling magnetite binding protein. Mms6 has been found to self-assemble in solution to form micelle-like structures with a high number of protein subunits. 14 This is likely to be due to the amphiphilic nature of the protein sequence. These protein micelles are able to bind and accumulate iron ions in solution. This is thought to trigger iron oxide nucleation, which in turn aids magnetite crystal growth. 13 , 14 , 20 , 21 The acidic C-terminal region of Mms6 has also been investigated, with the peptide displaying some similar properties to the full length protein in terms of both iron binding and some ability to control iron oxide crystal growth. 14 , 22 \n We have previously demonstrated a novel approach for the formation of magnetite nanoparticles (MNPs) on surfaces, through the patterning of the Mms6 protein and subsequent biomineralisation of magnetite. 23 , 24 This generated consistent microscale MNP arrays when patterned onto functionalised gold surfaces using micro-contact printing (μCP). 23 , 24 Recently we published a variation of this approach to pattern a version of Mms6 engineered to contain an N-terminal cysteine, binding directly to gold and biotemplating MNP arrays of magnetite and magnetically harder cobalt-doped magnetite. 25 In this case, a protein resistant polyethylene glycol (PEG) self-assembled monolayer (SAM) was patterned onto gold surfaces via μCP, with the remaining space backfilled with cysteine-tagged Mms6. 25 The regions of the surface with a locally high Mms6 concentration were surrounded by a dense monolayer of PEG molecules. In this context, our system can be considered as a mimic of the arrangement of Mms6 thought to exist on the interior surface of the magnetosome membrane. Clusters of Mms6 are anchored in the magnetosome membrane through their hydrophobic membrane interacting region, and the C-terminal acid-rich region is exposed within the magnetosome lumen. The N-terminal cysteine–gold attachment allows control over the orientation of the protein on the surface, ensuring that the active C-terminal region is displayed to the reaction solution. This surface based biomineralisation experiment therefore offers a unique in vitro method of studying Mms6 in an environment similar to the native state, anchored in the magnetosome membrane. This is in contrast to the previously performed solution based experiments, 14 , 20 , 21 where the protein is thought to self-assemble into micelle structures, which have the opposite curvature to that found on the interior of the magnetosome membrane. We used this biomimetic system to investigate the differences between the Mms6 C-terminal peptide and the intact Mms6 protein in MNP synthesis to determine if the peptide can be effectively substituted for the intact protein ( Fig. 1 ). Being able to utilise a synthetic peptide offers advantages over the full protein, as peptides are cheaper and easier to produce, which would make the biotemplating properties of Mms6 more industrially amenable. This comparison also uncovers interesting differences between the activity of the protein and the peptide, which provides insight into the function of Mms6 in vivo . Previous studies have found Mms6 tightly associated with the isolated MNPs of magnetotactic bacteria, 13 suggesting that the protein has a strong affinity for MNPs. To test this property in vitro we probed whether Mms6, or its C-terminal peptide region, was able to bind magnetite when supplied with pre-formed MNPs suspended in solution and successfully anchor them to the surface ( Fig. 1 ). These experiments allowed the magnetite templating and magnetite binding activities of Mms6 to be analysed separately. Fig. 1 Experimental scheme. (a) Stamping PEG-thiols onto a gold surface using micro-contact printing. (b) Formation of the protein resistant self-assembled monolayer. (c) Backfilling with the cysteine tagged Mms6 (cys-Mms6) or cysteine tagged C-terminal Mms6 peptide (cys-pep). The surface is then subjected to either a magnetite precipitation reaction (Scheme I) or supplied with preformed magnetite nanoparticles (Scheme II).",
"discussion": "3. Results and discussion We used Quartz Crystal Microbalance with dissipation (QCM-D) to determine if the cysteine labelled Mms6 peptide (cys-pep) was able to interact with a gold surface as we have previously shown for the full length form of Mms6 (cys-Mms6). 25 QCM-D systems are used extensively to study interactions of biomolecules with surfaces. 31 In this experiment, a thin piezoelectric quartz crystal coated with a layer of gold oscillates at its resonant frequency. When material is applied and deposited on the surface, there is a corresponding shift in the resonant frequency. This shift, described by the Sauerbrey equation, is negative if the mass on the surface increases. The dissipated energy loss from the surface can also be measured, which gives an indication of the changes to the visco-elastic properties of the surface adsorbed material. The QCM-D analysis is presented in Fig. 2 for the cysteine labelled Mms6 peptide and, for comparison, the Mms6 protein. These both show the characteristic decrease in frequency associated with adsorption. In phase A, the system is washed with ultra-pure water. During phase B, the peptide or protein is allowed to flow into the system in PBS, before the surface is washed again with ultra-pure water to remove unbound material. The peptide appears to rapidly reach an equilibrium state, with the protein taking longer to plateau in phase B. This is probably because the molar concentration of the protein is less than for the peptide, as the Mms6 protein is larger than the peptide. As both solutions contained 10 μg mL –1 of the respective biomolecule, the molar concentration of cysteines for attachment is higher for the peptide than the protein, leading to more rapid adsorption of the peptide. Using the Sauerbrey equation 32 and estimates of the dimensions of both of the peptide and the protein (obtained from models generated by the Quark protein prediction server 33 ) the coverage of the gold surface by the biomolecule could be calculated ( Table 1 ). This indicates that both molecules produce an almost monolayer coverage of the surface. The Voight model (ESI Table 1 † ) provides an estimated thickness of this layer of 8 nm. This suggests both the protein and the peptide are packed orthogonally to the surface. Furthermore, the return to baseline dissipation in the case of the peptide is consistent with available examples of rigid biopolymer deposition. 31 To confirm that the peptide itself does not interact with the PEG-thiol SAM we conducted further QCM-D experiments (ESI Fig. 1 and ESI Table 2 † ). These clearly show that there is no detectable interaction between the PEG passivated surface and the cysteine tagged biomolecules occurring. Our previous study reveals that Mms6 also does not directly interact with the PEG layer. 25 \n Fig. 2 Frequency (Δ f , solid lines) and dissipation (Δ D , dotted lines) changes of the 7th overtone recorded with QCM-D during adsorption of cys-Mms6 (green) or cys-pep (orange) onto clean gold coated quartz crystals. Grey regions show when Milli-Q water was applied (A and C), and the white region (B) shows when a PBS buffer containing cys-Mms6 or cys-pep at a concentration of 10 μg mL –1 was applied (flow rate 50 μL min –1 ). Table 1 The mass coverage measured with QCM-D of cys-Mms6 and cys-pep adsorbed onto clean gold crystals \n a \n \n Sauerbrey values Cys-Mms6 Cys-pep Mass (ng cm –2 ) 258 182 Coverage (pmol cm –2 ) 23 70 Complete monolayer (pmol cm –2 ) ≈24 ≈83 Coverage (%) ≈96 ≈84 \n a All modelling was performed with QTools 2 Qsense software. Sauerbrey values were calculated from the 7th overtone. The cysteine tagged Mms6 or Mms6 peptide was used to backfill a PEG-thiol patterned gold surface using conditions we have already developed and optimised. 25 These surfaces were subject to a partial oxidation of ferrous hydroxide reaction with potassium hydroxide (POFHK) ( Fig. 1 , Scheme I), which precipitated MNPs. Once the reaction was complete and following cleaning, the surfaces were characterised by scanning electron microscopy (SEM), and the particles formed on both the surfaces and in the bulk solution were probed by X-ray diffraction (XRD) ( Fig. 3 ). The interplanar distances of the particles produced are in close agreement to those corresponding to magnetite ( Table 2 ), rather than the iron oxide maghemite that has a similar crystal structure. The (400) plane in particular, which can be used to distinguish between magnetite and maghemite, confirms the majority of the material is most likely to be magnetite. 34 Diffraction peaks corresponding to gold are also present in surface biomineralisation data, obscuring the (222) magnetite peak. The bulk particles were visualised by transmission electron microscopy (TEM) and their dimensions measured using ImageJ 35 software ( Fig. 4 ). Grainsize analysis was also conducted for particles formed on both the protein and peptide patterned surfaces. Fig. 3 XRD spectra of POFHK (bulk) nanoparticles (black), and of a Mms6 (POFHK) surface. Each spectrum is offset for clarity and peak positions for magnetite (red) and gold (gold) are labelled. Table 2 Interplanar distances from the XRD spectrum of the MNP samples ( Fig. 4 ). Interplanar distances for magnetite and maghemite (all measured in Å) \n a \n \n Peak Magnetite Maghemite POFHK (bulk) \n Mms6 (POFHK) \n (220) 2.966 2.950 2.966 2.962 (311) 2.53 2.520 2.534 2.527 (222) 2.419 2.410 2.423 — \n b \n \n (400) 2.096 2.080 2.097 2.097 (422) 1.712 1.700 1.718 1.711 (511) 1.614 1.610 1.615 1.614 (440) 1.483 1.480 1.483 1.483 (533) 1.279 1.270 1.280 1.276 \n a Based on spectra from DIFFRAC Plus software. \n b Obscured by the Au (111) peak. Fig. 4 TEM image (left) and grainsize analysis of POFHK (bulk) nanoparticles. Scale bar is 100 nm. SEM of the Mms6 surface revealed clear, defined, dense patterns of magnetite nanoparticles in stripes corresponding to the areas covered by the protein. These patterns are consistent with our previous Mms6 surface biomineralisation experiments. 25 When the peptide was used in place of the protein we observed a very different result. As shown in Fig. 5 , the peptide patterns are not as clear or well defined as for the Mms6 protein surface templated particles. The peptide surface templated particles appear to be sparsely distributed on the surface. The extremely low amount of material observed is insufficient for diffraction analysis; although based on the XRD results from the Mms6 surface in the same reaction conditions we infer that the material is magnetite. The grainsize analysis of these particles also shows an interesting difference, ( Fig. 4 & 5 ). The solution phase MNPs gives rise to a mean size of 60 nm, very close to the 65 nm size we observe on the peptide patterned surface, both with a similar broad distribution. By comparison, the Mms6 protein mediated particles are approximately 50% larger, with a mean size of 87 nm, and yet feature a much tighter size distribution. These data indicate that Mms6 is able to enhance both the size and homogeneity of the forming nanoparticles and also successfully anchor these particles to the surface. The peptide on the other hand appears to display particles with similar properties to the MNPs formed in the bulk solution, with no improvement in homogeneity, and with much less dense anchoring of the particles to the surface. It is possible that the peptide may be more susceptible to the destabilising conditions of the POFHK reaction, resulting in loss of function. If this is the case, it indicates that structure and assembly are necessary to the function of Mms6 rather than acidic C-terminal region sequence alone. Fig. 5 SEM analysis of the different surfaces at increasing magnification. Scale bars are 100 μm (left), 20 μm (centre), and 100 nm (right). MNP sizing histograms are shown with Gaussian fitting (GraphPad Prism). MNP coverage from 5 areas of biomolecule patterned regions is shown for each sample with standard deviation. To ascertain if the Mms6 protein or peptide was able to bind pre-made MNPs to the patterned surface, we modified the system. Rather than use the surfaces in an in situ biomineralisation reaction to produce particles, we simply took MNPs in water (prepared from a POFHK reaction), and applied them directly to a surface already patterned with either the Mms6 protein or peptide ( Fig. 1 , Scheme II). The resulting surfaces were washed, and analysed with SEM, and subsequent grainsizing was performed as before ( Fig. 5 ). The main difference observed is between the two Mms6 surfaces (biomineralised and MNP binding), which revealed a less dense MNP pattern had been produced in the pre-formed MNP binding when compared to the in situ POFHK experiments; indicating that fewer MNP had been adsorbed. The protein also showed no selectivity towards binding larger MNPs, as the mean particle size from the grainsize analysis matched those of the applied bulk MNPs. SEM of the Mms6 peptide surface reveals no significant difference to that obtained from the control POFHK reaction, with sizes which again match those of the applied bulk MNPs. Taken together, these results help to build a picture of the differences between the protein and the peptide, and the mode of action of Mms6. Intriguingly, if we compare the density of the MNPs on the Mms6 protein patterns resulting from the in situ biomineralisation to those formed from the addition of pre-made nanoparticles, we see it much reduced in the binding experiment when compared to the biomineralisation one. This is despite the biomineralisation surface being subject to much more extreme conditions of pH and heat, which suggests that the binding of MNPs by Mms6 may be enhanced when the MNPs are formed in the presence of the protein. We hypothesise that by binding iron ions, nucleating and stabilising the formation of the MNP, the C-terminal residues of Mms6 may mediate more contacts with the growing particle than if the MNP is supplied preformed. This may suggest that the strong attachment of Mms6 to magnetite is a by-product of its nucleating activity. In this study, the Mms6 peptide appears to offer no effect on controlling the size or shape of the MNPs produced, and also sequesters nanoparticles with much lower density than the full Mms6 protein. Previous studies of an Mms6 C-terminal peptide in solution phase POFHK magnetite formation show modest particle size effects. 22 A peptide with the additional glycine leucine repeat section displays greater activity. 22 We considered that the shorter length of the peptide (when compared to the Mms6 protein) may mean it is not as accessible on the SAM patterned surface, which may limit its ability to function as fully as when free in solution. To test this, we prepared cys-pep surfaces with no SAM, thereby providing maximum accessibility to the peptides for both our process schemes. The surfaces, visualised by SEM (ESI Fig. 2 & 3 † ), revealed the same type of sparse particle deposition as before, showing that the peptides low activity is not due to masking by the SAM. Therefore, it may be that the Mms6 peptide is crowding itself, by packing more closely than is possible in the full length sequence. Alternatively, the shorter peptide may be more prone to destabilisation by heating than the full length Mms6 protein is. The Mms6 peptide yielded sparse MNP coverage of the patterns under both experimental systems, and the apparent lack of any effect upon size or homogeneity of the surface bound nanoparticles suggests that this molecule exerts no apparent control over magnetite formation in this experimental system. The similarity between the pattern densities of the peptide resulting from Schemes I and II is suggestive of the peptide binding some particles weakly in both cases. It is possible that during the Scheme I POFHK reaction, the peptide may simply be binding to the particles produced in the bulk solution in a similar manner to the MNP binding by cys-pep in Scheme II. This mode of action would explain why both cys-pep patterned surfaces (biomineralised and MNP binding) look very similar. Purified Mms6 forms spontaneous micelle-like structures; indicating that this protein has a natural propensity to aggregate. 14 , 15 , 20 The Mms6 C-terminal peptide contains an abundance of acidic residues (which are considered an essential feature of Mms6 magnetite biomineralisation 14 ) and previous analysis demonstrates some aggregation into oligomeric species of the range dimers to octamers. 15 However, even with the same acidic residues and the locally high concentration brought about through the surface attachment, as well as any natural oligomerisation, the peptide appears to be unable to replicate the activity of the full length protein in our experiments. One important feature which is absent from the peptide is the distinctive glycine–leucine repeat motif (ESI Fig. 5–7 † ) which has been shown to be important in the oligomerisation and activity of Mms6 in previous studies. 15 , 22 This type of low complexity repeating sequence is commonly associated with self-assembling proteins such as silk fibroins. 36 We believe this motif could play a crucial role in the assembly of the complex; bringing about the correct packing and orientation of the proteins to facilitate iron ion coordination, binding, and nucleation of the magnetite nanoparticle. Molecular modelling of this sequence (ESI Fig. 5–7 and ESI † methods) suggests that the glycine and leucine residues in an α-helical conformation could produce regularly spaced interlocking knobs and holes along the length of the repeat motif. A parallel assembly of such helices would allow precise packing of multiple Mms6 molecules to generate a C-terminal surface of iron ion binding residues (aspartate and glutamate). This packing may give rise to an arrangement of these acidic residues that is able to support iron binding and crystallisation of magnetite, as opposed to the potentially uncontrolled surface packing of the peptide form of Mms6 used in our experiments (ESI Fig. 6 † ). Using our biomimetic surface system as a mimic of the magnetosome membrane, we find that Mms6 is able to form nanoparticles which are different (in size and homogeneity) from the particles formed in a bulk solution, which is consistent with our previous studies. It should be noted that the ≈87 nm MNPs formed on our biomimetic surface are approximately twice the size of the 50 nm natural magnetosomes crystals. In previous studies where Mms6 was used in solution in a similar POFHK reaction, the particles were found to be approximately 50% smaller than control particles formed without protein. 17 This is in direct contrast to the 50% size increase we see in our surface based experiment. One key consideration is the effect of the curvature present on the surface of Mms6 soluble micelles when free in solution, when compared to the immobilisation of Mms6 on a flat surface ( Fig. 6 ). A planar arrangement of Mms6 may provide a greater expanse of the active acidic region, giving rise to increased nucleation and growth of larger crystals ( Fig. 6b ). The smaller convex surface present on the Mms6 micelles may provide a smaller nucleation surface and hence form smaller crystals ( Fig. 6a ). Neither the micelle form nor our surface experiment perfectly matches the concave assembly of Mms6 likely to be present on the interior face of the magnetosome ( Fig. 6c ). Further experiments could include enhancing our biomimetic system to better represent the curvature of the magnetosome interior. In addition, this biomimetic system provides a clear marker (larger MNP) of Mms6 activity in vitro . This could be exploited in future experiments to probe the effect of changes to the Mms6 sequence on MNP formation. This may help to further unlock the mode of action of Mms6 at the individual residue level. Fig. 6 The assembly of Mms6 under different conditions (the N-terminal region of Mms6 is represented by a green rectangle and the iron binding C-terminal region by two green cylinders). (a) Mms6 in solution arranged into a micelle, with a convex surface interacting with a magnetite nanoparticle. (b) Surface immobilised Mms6, with a planar interaction with a magnetite crystal. (c) Mms6 within a magnetosome, presenting a concave surface that interacts with a magnetite particle. In summary, the results presented here suggest that Mms6 is a magnetite nucleation protein, where the assembled protein surface binds iron ions specifically to nucleate the formation of magnetite. Furthermore, in our biomimetic system it is not the C-terminal section alone, but the full length protein, which is required to provide the complete function of Mms6."
} | 5,966 |
28617603 | null | s2 | 1,930 | {
"abstract": "In this study, we have fabricated robust patterned surfaces that contain biocompatible and antifouling stripes, which cause microorganisms to consolidate into bare silicon spaces. Copolymers of methacryloyloxyethyl phosphorylcholine (MPC) and a methacrylate-substituted dihydrolipoic acid (DHLA) were spin-coated onto silicon substrates. The MPC units contributed biocompatibility and antifouling properties, and the DHLA units enabled cross-linking and the formation of robust thin films. Photolithography enabled the formation of 200-μm-wide poly(MPC-DHLA) stripped patterns that were characterized using atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS), and rhodamine 6G staining. Regardless of the spacing between poly(MPC-DHLA) stripes (10, 50, or 100 μm), Escherichia coli rapidly adhered to the bare silicon gaps that lacked the copolymer, confirming the antifouling nature of MPC. Overall, this work provides a surface modification strategy for generating alternating biofouling and nonfouling surface structures that are potentially applicable for researchers studying cell biology, drug screening, and biosensor technology."
} | 288 |
35745302 | PMC9228757 | pmc | 1,931 | {
"abstract": "Smart fire-warning sensors based on graphene oxide (GO) nanomaterials, via monitoring their temperature-responsive resistance transition, have attracted considerable interest for several years. However, an important question remains as to whether or not different oxidation degrees of the GO network can produce different impacts on fire-warning responses. In this study, we synthesized three types of GO nanoribbons (GONRs) with different oxidation degrees and morphologies, and thus prepared flame retardant polyethylene glycol (PEG)/GONR/montmorillonite (MMT) nanocomposite papers via a facile, solvent free, and low-temperature evaporation-induced assembly approach. The results showed that the presence of the GONRs in the PEG/MMT promoted the formation of an interconnected nacre-like layered structure, and that appropriate oxidation of the GONRs provided better reinforcing efficiency and lower creep deformation. Furthermore, the different oxidation degrees of the GONRs produced a tunable flame-detection response, and an ideal fire-warning signal in pre-combustion (e.g., 3, 18, and 33 s at 300 °C for the three PEG/GONR/MMT nanocomposite papers), superior to the previous GONR-based fire-warning materials. Clearly, this work provides a novel strategy for the design and development of smart fire-warning sensors.",
"conclusion": "4. Conclusions In summary, we prepared the PEG/MMT/GONR nanocomposite papers with different oxidation degrees and morphologies via a facile, solvent free, and low-temperature evaporation-induced assembly approach. Three different oxidation degrees and morphologies of GONRs were synthesized via adjusting the concertation of the starting oxidant and the oxidation time. The combined use of GONR and MMT, along with PEG molecules, formed an interconnected nacre-like layered structure, and the GONR-2 produced a better-oriented MMT structure. As a result, the mechanical properties showed that the appropriate oxidation of GONR-2 produced an effective reinforcing efficiency. Typically, in comparison with the G 1 and G 3 papers with tensile strength of about 27 MPa and 26 MPa, the tensile strength of the G 2 nanocomposite paper was about 42 MPa. Such artificial nacre of the EG/MMT/GONR nanocomposite paper exhibited an excellent flame-retardant property and showed tunable flame-detection response and sensitive fire early-warning response before being ignited (e.g., 6, 175, and 281 s at 200 °C for the G 1 , G 2 and G 3 papers). Considering the balance between the mechanical property and fire-warning response, the appropriate oxidation degree of GONR is promising for fabricating a fire-warning sensor for potential application. The response time of the fire-warning sensor can be effectively tuned by using the different oxidation degrees of GONRs for monitoring critical fire risk, which would provide a new way to construct a sensitive fire-warning sensor.",
"introduction": "1. Introduction Fire is a “double-edged sword” since it has been of great significance in the history of human civilization, but is also hazardous [ 1 , 2 ], especially since the development of synthesized polymer. As is well known, London’s Grenfell Tower fire in 2017, Brazil’s National Museum fire in 2018, and Paris’ Notre Dame fire in 2019 have caused massive casualties, irreparable property damage, and the loss of priceless artefacts [ 2 , 3 , 4 ]. These severe outdoor fire disasters have been attributed to the high fire-risk of various combustible materials that have low ignition temperatures of 300–500 °C and rapid flame-spread speeds (e.g., about 8 m for <80 s in the large-scale UL experiment [ 5 ]). Some traditional fire alarm strategies, including smoke alarms and heat detectors, have proven to be effective at significantly reducing or even avoiding the risk of indoor fires; however, they have limitations, including a relatively long fire alarm time of >100 s, no fire early-warning signal below ignition temperature, and restricted use in certain outdoor environments [ 6 ]. Therefore, understanding how to monitor the critical outdoor fire risk of combustible materials is imperative, but remains a challenge. To address the above issues, novel and complementary fire-warning materials and sensors have been developed over several years. Typically, the temperature-responsive resistance transition of various nano-fillers, e.g., graphene oxide (GO) [ 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 ], carbon nanotube (CNT) [ 15 ], and MXene [ 16 ], has been widely used to construct sensitive fire-warning sensors. Among them, GO-based fire-warning materials with sensitive flame detection and fire early-warning response have attractive considerable research interest. Typically, pristine GO networks are electrically insulated due to oxygen-containing functional groups in their structure [ 17 , 18 , 19 ]. On encountering a flame, the reduction of oxygen groups in the insulating GO network can chang it into an electrically conductive rGO one [ 20 , 21 , 22 ], thus providing a rapid flame-detection response time (normally less than 10 s [ 11 , 14 , 23 , 24 , 25 , 26 , 27 ]). On the other hand, GO-based fire-warning materials also offer an ideal fire early-warning signal that can be activated below the ignition temperatures of most flammable materials, e.g., 232 s at 200 °C and 35 s at 300 °C for 3 -mercaptopropyltrimethoxysilane modified-GO nanocomposite paper [ 13 ]. A sensitive fire early-warning response is strongly dependent on the formation of the interconnected network and the oxidation degrees of the prepared GO sheets. Moreover, the response of such sensors should also depend on the environmental atmosphere, and this will be further investigated. Among GO derivatives, GO nanoribbon (GONR) shows promise for achieving rapid and sensitive fire-warning response signals during the precombustion process, owing to the effective formation of an interconnected network. Our previous work demonstrated that GONRs produced more rapid resistance-response behaviors than the corresponding GO sheets and GO wide ribbons (GOWR) [ 28 , 29 ]. At an unchanged temperature of ∼300 °C, these fire-warning materials and sensors showed different early-warning response times (to activate the alarm light) of 24.0, 33.5, and 39.0 s for the GO nanoribbon, the GO wide ribbon, and the GO sheet at the same concentration, respectively. This was attributable to the formation of a highly interconnected GONR network, and to GONR thermal reduction behavior that was more effective than that of the other two GO derivatives (i.e., GOWR and GO sheet). Thus, based on the thermal reduction feature of GONR networks [ 30 ], different oxidation degrees are crucial to determining the final fire-warning performance of the fire-warning sensor, and it is therefore necessary to understand the effect of oxidation degrees on the two important parameters (i.e., flame detection response time and fire early-warning response time). Whether or not the different oxidation degrees of GONRs can result in different impacts on temperature-induced resistance transition behaviors remains an important question. In the present work, we synthesized three types of GONR with different oxidation degrees and morphologies, and thus prepared GONR-based nanocomposite papers via a facile and low-temperature evaporation-induced assembly approach. The oxidation degrees and morphologies of the GONRs were analyzed and discussed by various characterizations. The appropriate oxidation degree of the GONRs, along with polyethylene glycol (PEG) molecules, produced a better-oriented montmorillonite (MMT) structure than the other two GONRS, which produced different impacts on the mechanical properties and fire-warning performance. The structure, mechanical properties, and fire-warning response of the three GONR-based nanocomposite papers were investigated and compared."
} | 1,963 |
27386573 | PMC4928988 | pmc | 1,932 | {
"abstract": "High biodiversity promotes the evolution of more biodiversity by selecting for new species exploiting resources more thoroughly.",
"introduction": "INTRODUCTION Biodiversity has always fluctuated during Earth’s history, with phases of massive extinction followed by adaptive radiation ( 1 ). In the light of the current massive global species loss at the hands of recent human development, there is growing interest in predicting how species loss affects the structure and functioning of ecosystems on an evolutionary scale. Evolution of species and the emergence of new taxa can restore ecosystem processes that have been lost during extinction waves ( 2 ), and understanding how changes in diversity patterns feedback to evolutionary rates may provide new tools to predict the shape of tomorrow’s ecosystems and design conservation strategies. Biotic interactions such as competition and predation are major drivers of evolutionary processes ( 3 ), suggesting possible feed back loops between biodiversity and evolutionary diversification. However, the net effect of competition on diversification processes is highly variable and different studies have reported contrasting outcomes. On one hand, a range of studies suggests that evolution will slow down with increasing numbers of competing taxa because available niches will be exploited more completely, thereby preventing the establishment of a novel phenotype ( 4 – 6 ). On the other hand, competition may also stimulate adaptive radiation ( 3 , 7 – 9 ). In natural ecosystems, biodiversity hotspots function as cradles for new species ( 10 ), suggesting that speciation rates may increase with species richness (that is, that biodiversity favors further diversification) ( 11 ). We propose that contrasting observations on the role of competition in evolutionary dynamics can be unified by investigating biodiversity-evolution relationships over a gradient of species richness and niche differentiation. Here, we explored the evolutionary dynamics of a focal bacterial species growing in communities of different biodiversity levels, from one species (focal species competing with itself) to eight species of the genus Pseudomonas . This model system allowed us to explore different ecological scenarios at high functional and taxonomical resolution while drawing on a large pool of life history strategies ( 12 ). We manipulated the functional diversity (FD) of the background community using an index integrating both the number of competitors and niche partitioning ( 13 ). FD is commonly used as predictor of community functioning, and we expected that it may also predict evolutionary processes ( 14 ). For instance, FD is highest in species-rich communities where each species uses different subsets of the available niche space, which likely alters evolutionary diversification processes. We specifically addressed whether higher biodiversity promotes or restricts evolutionary diversification in the focal species. We focused on the first step of evolutionary processes, the generation of mutants harboring altered phenotypes. We expected that biodiversity affects evolutionary dynamics by altering the establishment and maintenance of evolved phenotypes. High biodiversity is often associated with increased resource competition ( 15 ), yet, at the same time, many species may compete for few shared niches, leaving alternative niches underexploited ( 16 ). We therefore expected that novel phenotypes are more likely to be established at high biodiversity because of the novel phenotypes escaping competition by using underexploited resources more efficiently ( 17 ). We used the XerD site-specific recombinase as a model molecular mechanism generating mutations ( 18 ). Site-specific recombinases generate important mutation steps by reshuffling the genome. They rapidly generate new phenotypes and play an essential role for survival in new or competitive environments ( 19 – 21 ). Although this enzyme is not the only mechanism generating mutations, its deactivation slows down evolutionary rates at the tested time scale ( 19 ). We therefore used a functional mutant lacking XerD as a reference to estimate the fitness gain conferred by the evolution of new phenotypes under various conditions. First, we addressed whether high biodiversity accelerates diversification and characterized the phenotypic changes occurring during the diversification process. Second, we assessed whether and when the novel phenotypes take over the ancestral phenotype, considering the spread of novel phenotypes as an indicator for the pressure on the focal species to evolve new features. Finally, we characterized the ancestral strain and evolved phenotypes and tested whether evolved phenotypes occupy a different niche and consume different resources than the ancestral strain.",
"discussion": "DISCUSSION The relationship between biodiversity and ecosystem functioning has been intensively investigated at ecological time scales ( 22 ), yet little is known on how biodiversity change affects ecosystems at evolutionary time scales. Biodiversity has been linked to both increases and decreases in diversification rates ( 3 – 5 , 7 , 11 ). The present study aimed to explore the mechanisms responsible for these variable outcomes by examining experimental communities of varying diversity in respect to adaptive radiation and changes in resource use. This reductionist approach allowed us to disentangle underlying mechanisms that are difficult to track in complex natural ecosystems. We manipulated resource competition by establishing a gradient in FD, including competition with clone mates as reference. We assessed the net impact of biodiversity on diversification as the fitness advantage conferred by the recombinase, allowing rapid manipulation of the genetic material and potential replacement of the ancestral focal species by novel phenotypes. Finally, we inspected changes in the resource use spectrum as a potential mechanism underlying biodiversity-diversification relationships. High biodiversity stimulated the replacement of the ancestral phenotype of the focal species by novel ones, suggesting that novel phenotypes more efficiently exploited the given resource spectrum. Multispecies communities tend to exploit complex resources more intensively, leaving few resources unused ( 23 ). This resource preemption can reduce the fitness of individual species as biodiversity increases ( 15 , 24 ). This matches well with our observations and suggests that the evolution of novel phenotypes allows for a better exploitation of additional resources, thereby escaping a tangled bank situation on the resources ( 25 ). However, high diversity may also result in a more complete use of resources, which may restrict diversification ( 4 , 26 ). Replacement of the ancestral phenotypes by newly evolved phenotypes in diverse communities suggests that the benefit of evolving to more efficiently use underexploited resources outweighs the disadvantages of increased completion in highly diverse communities. Notably, evolved phenotypes essentially gained the ability to better use resources consumed at a low rate by the ancestral phenotype rather than acquiring the ability to use novel resources. This ability may prove useful in escaping competition in many ecosystems, in which species compete for a common set of resources and niches ( 16 ). The decline of the focal species in species-rich communities was likely associated with increased resource competition. At the same time, high FD was associated with a fitness advantage of the wild type over the mutant impaired in XerD recombinase. This suggests that the fitness advantage is based on the ability to rapidly generate new phenotypes at high biodiversity because this allows the enhanced use of resources marginally used by the ancestral strain. Diversification was limited by fitness trade-offs, providing an explanation for the low frequency of evolved phenotypes in the absence of competition. Fitness trade-offs are typical for radiation processes and contribute to species coexistence ( 27 , 28 ). In the present experiment, the more the evolved phenotypes differed from the ancestral phenotype, the lower their overall growth rate was. In the absence of competition, the evolved phenotypes could coexist with the ancestral phenotype; however, their lower growth rate kept them at low frequency. As diversity increases, competition reduces the growth of the ancestral strain, allowing the novel phenotypes to be established in the community using resources underexploited by the competing species. In the present experiment, we focused on site-specific recombinases as a source of mutations. Their ability to excise, move, and invert large pieces of DNA within the genome ( 29 ) rapidly generates new phenotypes ( 19 , 30 ). Site-specific recombinases form the core of rapid adaptive processes in the rhizosphere ( 19 , 21 ) and in gut bacteria ( 20 ), as well as in pathogens ( 31 ). In the experimental system used, site-specific recombinases played an important role in short-term evolution; no novel phenotypes occurred in the mutant lacking XerD. However, this does not exclude the possibility that other mechanisms were involved in the observed pattern. Point mutations certainly also took place, but as they occur at a lower frequency, they are likely more important in long-term adaptive processes ( 32 ). Further, long-term experimental evolution experiments with varying biodiversity levels may help disentangle the relative contribution of different mutation mechanisms for bacterial evolution. Evolutionary and ecosystem processes have long been treated as distinct, but recent studies on grassland communities ( 33 ) and the present study suggest that it is time for unification. The choice of simplified bacterial communities as a model allowed detailed insight into diversification processes and their underlying mechanisms. However, the simplified system and settings also have limitations. We focused on resource competition as the major driver of diversification processes, thereby ignoring other interactions that also drive evolutionary processes, such as predation and mutualism. We targeted trait variation in competing strains because this variation is essential to allow further evolutionary processes to take place, but the evolutionary trajectory will depend on the type of selective regime. Future studies that consider more complex interactions are needed to shed light on the relative importance of different types of interactions for diversification in microbial systems ( 34 ) and extrapolate findings to natural settings in the field. Further, we did not assess the ability of the evolved phenotypes to invade the communities when rare. This might be justified as we sampled evolved phenotypes after they had established themselves in the resident community, which necessarily involved invasion. However, more detailed studies are necessary to disentangle the frequency-dependent fitness of novel phenotypes in communities of varying diversity. Resource use of bacterial strains was based on the Biolog system, which is well established and provides a good proxy for predicting competitive interactions ( 35 ). However, this approach only represents a snapshot of the catabolic potential of bacteria ex situ ( 28 ), and its use in explaining growth kinetics in complex bacterial communities is limited ( 36 ). Considering these limitations, we used the measurements to characterize bacterial phenotypes rather than their resource use in situ. Our results suggest that competition for resources is a good predictor of diversification processes in bacterial communities. However, this does not preclude that other mechanisms were also involved in the observed patterns. For instance, changes in life history strategies may also have driven diversification processes ( 31 ). For instance, the lower growth rate of the evolved phenotypes may have been caused by a shift in the balance between growth and stress resistance ( 37 ). However, lower population growth of more generalist phenotypes indicated that the trade-off between population growth and the ability to use a wide spectrum of resources was involved in the observed diversification processes ( 28 ). Further, cross-feeding and niche construction may have contributed to the observed diversification processes ( 38 , 39 ). Extending the current study and including a wider range of traits involved in trophic and nontrophic interactions are promising perspectives in developing a more general eco-evolutionary theory ( 40 ). Overall, our results suggest that biodiversity can stimulate the evolution of novel phenotypes, provided that (i) part of the resources get increasingly scarce at high diversity, reducing the fitness of the ancestral strain, and (ii) that evolved phenotypes gain access to underutilized resources. Here, we took single resources as different niche dimensions ( 41 ). We stressed that diversification processes increase at high species diversity; however, the relationship is most likely nonlinear. Presumably, at some point, the effect of diversity on diversification will saturate or even reverse as all resources are being consumed, resulting in a hump-shaped biodiversity-diversification function. The changes in colony morphology and resource use patterns were detected after growing the bacteria for at least 10 generations after the end of the experiment, indicating that the changes were inherited and simply not acclimation effects. Considering the exceptional diversity of natural bacterial communities, our experimental approach only explored the low diversity end of this relationship. However, because of the nested structure and compartmentalization of many food webs, where subsets of species are linked with subsets of resources only, we expect the positive biodiversity-diversification relationship to be important even at high species richness. This study provides the first mechanistic explanation of biodiversity-diversification relationships in simplified communities, allowing a resolution that is very difficult to achieve in natural systems. Although caution should be exercised when transferring results from microcosm studies to real-world ecosystems, our results provide evidence that reduced biodiversity may compromise the ability of communities to respond in an evolutionary way to environmental changes. Thus, species loss not only may impair ecosystem functioning through short-term losses of functions but also can result in an “evolutionary debt” by slowing down the evolution of trait variation in the remaining organisms. Thus, species loss may prevent adaptations to changing environmental conditions that are needed to maintain ecosystem functioning in the long term."
} | 3,705 |
26067975 | PMC4463539 | pmc | 1,933 | {
"abstract": "Microbacterium profundi strain Shh49 T was isolated from deep-sea sediment from a polymetallic nodule area located in the East Pacific Ocean. Strain Shh49 T contains genes related to the reduction/oxidation of metals. It has potential application in the bioremediation of heavy metal-contaminated environments."
} | 78 |
34685796 | PMC8538330 | pmc | 1,934 | {
"abstract": "Phytoremediation, a method of phytomanagement using the plant holobiont to clean up polluted soils, is particularly effective for degrading organic pollutants. However, the respective contributions of host plants and their associated microbiota within the holobiont to the efficiency of phytoremediation is poorly understood. The identification of plant-associated bacteria capable of efficiently utilizing these compounds as a carbon source while stimulating plant-growth is a keystone for phytomanagement engineering. In this study, we sampled the rhizosphere and the surrounding bulk soil of Salix \n purpurea and Eleocharis obusta from the site of a former petrochemical plant in Varennes, QC, Canada. Our objectives were to: (i) isolate and identify indigenous bacteria inhabiting these biotopes; (ii) assess the ability of isolated bacteria to utilize alkanes and polycyclic aromatic hydrocarbons (PAHS) as the sole carbon source, and (iii) determine the plant growth-promoting (PGP) potential of the isolates using five key traits. A total of 438 morphologically different bacterial isolates were obtained, purified, preserved and identified through PCR and 16S rRNA gene sequencing. Identified isolates represent 62 genera. Approximately, 32% of bacterial isolates were able to utilize all five different hydrocarbons compounds. Additionally, 5% of tested isolates belonging to genera Pseudomonas , Acinetobacter , Serratia , Klebsiella , Microbacterium, Bacillus and Stenotrophomonas possessed all five of the tested PGP functional traits. This culture collection of diverse, petroleum-hydrocarbon degrading bacteria, with multiple PGP traits, represents a valuable resource for future use in environmental bio- and phyto-technology applications.",
"conclusion": "5. Conclusions S. purpurea and E. obusta are widespread, native plants in North America, distributed in various habitats and ecosystems, and are able to tolerate chronic levels of PHC pollution. Thus, they are ideal candidates for phytoremediation of PHC-contaminated soils. This culture collection holds 438 bacterial isolates with multiple degradative and PGP features, originating from unique soil environments characterized by high levels of PHC contamination. The functional potential of bacterial isolates reported here represents a rich reservoir of metabolically versatile PGPR-PHC degraders that could be used in holistic, bacterial-aided phytomanagement and remediation of PHC contamination in future research.",
"introduction": "1. Introduction Industrial activities such as mining for minerals, oil and gas extraction, inorganic fertilizer-based agriculture, and industrial waste disposal, are all associated with environmental contamination risks which represent a global challenge [ 1 ]. Among pollutants, petroleum hydrocarbons (PHCs) are of great concern and can pose a high risk in oil spills, and environmental contamination of aquatic and terrestrial ecosystems. PHCs, like crude oil, are heterogeneous organic mixtures composed of carbon and hydrogen atoms and are broadly classified into two major fractions: (1) aliphatic hydrocarbons, like alkenes, alkynes, or alkanes, and (2) aromatic hydrocarbons, including mono-aromatic (i.e., benzene, toluene, phenol, etc.), and polycyclic aromatic hydrocarbons (PAHs) [ 2 , 3 ]. The main sources of PHCs contamination in the environment are mostly anthropogenic, and include accidental release (i.e., diesel, solvent), and industrial activities (i.e., production of electricity, petrochemical activities) [ 4 ]. Environmental contamination with PHCs products has adversely affected various ecosystems, including soils, causing damage to natural habitats with serious economic consequences [ 5 ]. Concerns regarding soil pollution with PHCs have initiated the development of several remediation technologies, including biological, chemical and physical methods [ 2 , 6 ]. A promising biological technology for the removal of PHCs from soil is phytoremediation: an eco-friendly, green, solar-driven, and low carbon footprint approach that utilizes plants and their root-associated rhizospheric and endophytic microbiomes to clean-up PHC-contaminated soils [ 7 , 8 ]. Phytoremediation has proven its ability in remediating moderately polluted soils. However, phytoremediation has unreliable effectiveness at high levels of contaminants because of the reduced growth of introduced plants in these conditions. This reduction in plant growth may be partially due to variation in the association between plants and their resident microbiomes [ 9 , 10 , 11 , 12 , 13 ]. Over the last decade, most of the research efforts aimed at enhancing the efficiency of phytoremediation of PHCs focused on using plant species that can tolerate high levels of PHCs, such as Salix spp. [ 2 , 8 , 10 , 11 , 14 ]. Salix spp. (willows), which have been shown to be effective in decontaminating soils polluted with organic compounds, such as PHCs, and trace metals. Willows have several characteristics that may facilitate phytoremediation, including their ease of propagation, fast and perennial growth patterns, high-biomass production, high-contaminants tolerance, and massive deep-root systems [ 2 , 15 , 16 , 17 ]. Additionally, several recent studies have shown that Salix spp. can recruit certain microbial taxa that could help the plant to cope with PHCs contamination stress and accelerate the biodegradation process [ 10 , 18 , 19 , 20 ]. More recently, a promising strategy that includes the screening and identification of native plants growing spontaneously on PHCs-contaminated soils has been adopted [ 21 , 22 , 23 , 24 ]. This is the reason we chose Eleocharis obtusa (Willd), which dominated the vegetation at the site of study [ 21 ]. Eleocharis spp. are ubiquitous plants distributed across Canada and United States, where they grow in wetlands. These plants are not used in phytoremediation. Pérez-Jaramillo et al. [ 25 ] proposed a “back to the roots” frame that involves the survey of native plants, and their associated microbiomes, in their native habitats, with the goal of restoring plant-microbial associations that may have been diluted during plants domestication [ 25 ]. Native plants are more genetically diverse and more adapted to wide-ranging climatic conditions compared to other plant species currently chosen for the phytoremediation of PHCs [ 26 ]. Additionally, native plants have been shown to develop more close relationships with local rhizosphere microbiota than introduced plants [ 11 , 27 ], thus making native plants ideal models to study how microbiomes respond to environmental pollutions and explore their future use in the phytoremediation of PHCs. The rhizomicrobiome, a subset of the plant holobiont, refers to the soil microbiomes associated with a plant’s roots. The rhizomicrobiome contributes to the functioning of plants including through the removal and degradation of PHCs compounds in contaminated soils [ 28 , 29 ]. Plants growth under stress such as PHCs contamination is expected to be lower than it would be under optimal conditions [ 30 ]. However, exploiting the potential of plant growth-promoting rhizobacteria (PGPR) in phytoremediation of PHC-contaminated soils holds great promise as it has recently been demonstrated [ 31 , 32 , 33 ]. PGPR are soil microbes within the rhizomicrobiome with phenotypes that benefit plant growth [ 28 ]. Therefore, plant growth may be positively stimulated by the presence of rhizobacteria with plant growth-promoting (PGP) traits, which alleviate stresses in plants via several mechanisms including: reducing soil nutrient deficiencies (fixing nitrogen, solubilizing phosphorus and enhancing iron uptake), synthesizing plant growth hormones, reduction in ethylene production via 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity, as well as [ 30 , 34 ] degrading a broad range of PHCs compounds [ 2 ]. It is well documented that some rhizospheric bacteria have beneficial effects on their host in natural and anthropized terrestrial ecosystems. However, the role of rhizobacteria in association with plants that spontaneously grow in heavily PHC-polluted areas is not widely explored. However, some reports documented the influence of pollutants on microbial community structures [ 35 , 36 ]. The aim of this study was to isolate and characterize the PGPR and hydrocarbon-degraders associated with Salix purpurea and Eleocharis obtusa plants growing in a long-term petroleum hydrocarbon-polluted petrochemical site. We hypothesized that the rhizosphere of S. purpurea and E. obtusa plants growing in soils chronically contaminated with PHCs would harbor diverse bacterial communities with multiple key species having hydrocarbon degrading potential and PGP traits. To address our hypothesis, a structurally and functionally diverse collection of PGPR and degradative bacteria were isolated from the rhizosphere of Salix and Eleocharis plants collected in the contaminated site. The cultured bacteria were all assessed for their abilities to grow in the presence of alkanes and polycyclic aromatic hydrocarbons as the sole carbon source, as well as for their PGP traits.",
"discussion": "3. Discussion The study of rhizosphere microbial communities associated with plants growing in long-term PHC-contaminated soil represents an opportunity for phytoremediation research. Several reports described the microbial community structures, diversities and functions in the rhizosphere of planted Salix trees as well as in ruderal plants growing spontaneously in soils highly contaminated with PHCs, using different sequencing techniques including cloning [ 37 , 38 ], next generation-targeted amplicon sequencing [ 10 , 11 , 20 ] and metatranscriptomics [ 14 , 39 ]. This study used a conventional microbiological approach to isolate, identify and characterize bacteria with multiple petroleum hydrocarbon-degrading capacities and plant growth-promoting capabilities to generate a bacterial culture collection for future use as a source of bacterial inoculants to enhance phytoremediation of PHCs-contaminated soils. High concentrations of PHCs cause phytotoxic effects on plants growing in contaminated soils [ 23 , 40 ]. For example, the growth rates of corn and red bean plants were reduced at 10000 mg/kg of crude oil [ 40 ]. Similarly, Chaîneau et al. [ 41 ] reported a stunted plant growth and inhibitory effects on the seed germination of several plants such as Helianthus annuus , Zea mays, Lactuca sativa, Phaseolus vulgaris, Triticum sp. and Trifolium sp. when exposed to high concentrations of fuel oil ranging from 3000 to 12,000 mg/kg. However, despite the devastating effects of PHCs on plant growth, recent studies have reported that several spontaneously growing herbaceous plants were found to flourish in highly contaminated soils near abandoned oil wells where the concentrations of PHCs could reach up to 45,000 mg/kg [ 22 , 42 ]. Similarly, Desjardins et al. [ 21 ] reported three indigenous plant species ( Alisma triviale , Eleocharis obtuse and Panicum capillare ) that grow spontaneously in highly petroleum-contaminated decantation basins of a former petrochemical plant in Varennes (southern Québec, Canada). These plants were tolerant of high levels of PHCs where the concentrations could reach up to 26300 mg/kg [ 21 ]. Moreover, not only spontaneously growing indigenous plants could tolerate high concentrations of PHCs, but also introduced pioneering phytoremediator plants such as Salix sp. were able to tolerate such a high level of PHCs [ 10 , 14 , 20 ]. The results of this study support our hypothesis that plants growing in soil chronically contaminated with PHCs would select for rhizospheric bacteria with multiple petroleum hydrocarbon-degrading potential and plant growth-promoting capabilities. In this study, 438 bacterial strains were isolated from bulk soil, and the rhizosphere soil of S. purpurea and E. obtuse rhizosphere soil using three different isolation strategies to enhance the diversity of bacterial isolates with multiple petroleum-hydrocarbon degradation potentials and plant growth-promoting traits. Our isolation strategies resulted in a culture collection of bacterial strains belonging to Actinobacteria, Alpha- Beta- and Gammaproteobacteria, Bacteroidetes and Firmicutes ( Figure 1 ), encompassing a fairly diverse collection of bacterial genera (62 genera) ( Table 1 ), including Acinetobacter, Arthrobacter, Bacillus, Chitinimonas, Enterobacter, Gordonia, Klebsiella, Microbacterium, Mycobacterium, Nocardia, Nocardioides, Pseudomonas, Pseudoxanthomonas, Rhodococcus, Serratia, Sphingomonas, Stenotrophomonas, Streptomyces and Variovorax ( Table 1 ). Several of these genera have previously been shown to hold promising petroleum-hydrocarbons degradation potential and plant growth-promoting activities [ 30 , 43 , 44 , 45 , 46 ]. The selected media used in our study did not result in the cultivation of new phyla; however, expandable bacterial culture collections could be established using additional novel cultivation strategies, as previously demonstrated for Arabidopsis thaliana At -SPHERE culture collection [ 47 ]. Our study revealed that culturable rhizospheric bacteria associated with S. purpurea rhizosphere mainly belonged to Actinobacteria and Gammaproteobacteria ( Figure 1 ). In contrast to our results, Bell et al. [ 10 ] studied the bacterial community structure and composition in the rhizosphere of several willows cultivar growing in PHCs-contaminated soils using 454-pyrosequencing and found that Betaproteobacteria was the predominant phyla. One possible explanation for this result is that a selective medium was used in this study (Bushnell-Haas medium amended with 1% diesel) to isolate PHC-degrading bacteria, while in Bell et al. [ 10 ], all bacteria were potentially amplified and sequenced. In agreement with our explanation, Ferrera-Rodríguez et al. [ 48 ] reported that culturable rhizospheric bacteria from five Arctic native plant species growing in PHC-contaminated soils were similarly dominated by Actinobacteria and Gammaproteobacteria when a selective medium was used to isolate PHC-degrading bacteria. The predominant family within the willow rhizosphere was Enterobacteriaceae ( Figure 2 ). Recent studies have reported that genera belonging to the family Enterobacteriaceae were predominant in the root endosphere of plants growing in Athabasca oil sands reclamation sites [ 49 ] and herbaceous plants growing near natural oil seep fields [ 50 ]. Endophytic bacteria are thought to be a subset of the larger rhizosphere microbiota [ 34 ] and further studies looking at the composition of culturable endophytic bacteria of Salix plants growing in PHC-contaminated soils will be required in order to elucidate the role of bacterial endophytes to improve PHC-phytoremediation. Other predominant families included Micrococcaceae, Nocardioidaceae and Nocardiaceae ( Figure 2 ), which have been shown to possess strong petroleum hydrocarbon degradation capabilities [ 45 , 46 ]. To our knowledge, there are no other reports concerning the isolation and identification of rhizospheric bacteria from E. obusta . Our study revealed that culturable rhizospheric bacteria associated with the E. obusta rhizosphere were mainly affiliated to Gammaproteobacteria, Actinobacteria and Betaproteobacteria phyla ( Figure 1 ). The Dominant families were Comamonadaceae, Xanthomonadaceae and Microbacteriaceae ( Figure 2 ). Comamonad bacteria (phylum Betaproteobacteria), for instance, are known to contain genera such as Comamonas , Delftia and Variovorax , which exhibit an extraordinary capability of degrading wide spectra of PHCs [ 45 , 46 ]. Genera belonging to the phyla Gammaproteobacteria, Actinobacteria are also known to contain bacterial species with efficient petroleum hydrocarbon degradation potentials [ 45 , 46 ] such as Pseudomonas , Streptomyces and Rhodococcus [ 51 , 52 ]. Soil and rhizospheric bacteria can increase the phytoremediation of PHCs by decreasing the level of PHCs in the contaminated soils via their enzymatic machinery mostly under aerobic conditions [ 53 ]. The results obtained in our study indicate that many bacterial isolates originating from the contaminated soil and rhizosphere samples have the potential to degrade a wide range of PHC compounds. More than 32% of our bacterial isolates were able to degrade all PHC being tested ( Figure 4 ). Petroleum hydrocarbon-degrading bacteria isolated in this study belonged mainly to Actinobacteria (mostly Streptomyces , Arthrobacter , Rhodococcus and Nocardia ), Proteobacteria (mostly Pseudomonas , Enterobacter , Stenotrophomonas , Acinetobacter and Variovorax ) and Firmicutes (mostly Bacillus ). Previous reports have shown that many bacterial genera belonging to these phyla were able to degrade a wide range of PHC compounds [ 36 , 46 , 54 ]. For example, the genus Rhodococcus has demonstrated high efficiency in degrading and transforming a wide range of organic substances, including aliphatic and aromatic hydrocarbons, pesticides and petroleum [ 55 , 56 ]. Therefore, there are immense interests in utilizing Rhodococcus in bioremediation of polluted soils due to their safe and ease of culturing and maintenance, and high catabolic versatility [ 52 , 55 , 56 ]. Bacterial isolates with PGP traits provide critical functions for their host plants growing in stressful environments, such as soil contaminated with PHCs. Isolating bacteria from PHC-contaminated environments that have both PGP traits and PHC-degrading activities has been of great interest in a new paradigm of environmental cleanup biotechnology which exploits PGPR. Selecting plants suitable for phytoremediation depends on many criteria, the most important of which is root morphology [ 15 ]. PGPR with the capacity to produce the phytohormones IAA, which plays a role in inducing the formation of lateral roots [ 43 ], would further stimulate plant growth in PHC-contaminated soils. In this study, 43% of bacterial isolates synthesized IAA ( Figure S1 ), which were mostly affiliated to the genera Pseudomonas, Streptomyces, Enterobacter, Arthrobacter and Microbacterium ( Figure S2 ). Previous studies confirmed that IAA-producing genera reported in this work were also found to produce IAA by endophytic and rhizospheric bacteria isolated from various plants [ 24 , 57 , 58 ]. Another mechanism by which PGPR have the potential to improve plant growth under adverse environmental conditions, including PHC contamination, is by producing the enzyme ACC deaminase [ 30 , 43 ]. Stressed plants induce the production of the phytohormone ethylene to bolster their defense. However, ethylene also inhibits plant growth [ 59 ]. Certain PGPR can inhibit ethylene biosynthesis via the production of ACC deaminase which cleaves the ethylene precursor ACC into alpha-ketobutyrate and ammonia [ 30 , 59 ]. In this study, 60% of bacterial isolates were found to produce ACC deaminase ( Figure S1 ). Most isolates that could catabolize ACC reported in this work belonged to genera such as Pseudomonas , Klebsiella, Enterobacter, Stenotrophomonas and Microbacterium ( Figure S2 ). The high percentage of ACC deaminase-producing bacteria among our isolates corroborate previous studies reporting the widespread nature of this trait in various soil bacteria [ 60 , 61 , 62 ]. N fixation, phosphate solubilization, and siderophore production are some of the direct PGP mechanisms making nutrients available to plants. These traits were found among the bacteria isolated of this study ( Figure S1 ). Nitrogen fixation by diazotrophic bacteria is an important trait of PGPR that benefits the plant, especially when growing in nutrients-deficient soils [ 63 ]. Diazotrophic bacteria isolated in this study belonged mainly to genera such as Pseudomonas , Klebsiella, Bacillus , Enterobacter , Acinetobacter and Variovorax ( Figure S2 ). Low levels of soluble P in soils can restrict the growth and development of plants [ 43 ]. Some PGPR solubilize inorganic forms of P and convert it to plant-available forms, thereby facilitating plant growth [ 64 , 65 ]. Our study found that the majority of isolates are able to solubilize inorganic P belonged to the genera Pseudomonas , Acinetobacter , Bacillus and Serratia ( Figure S2 ). Another essential nutrient for plant growth is iron, even if it is present in soils in the highly insoluble form Fe 3+ [ 66 ]. Some PGPR produce low molecular-weight organic compounds, siderophores, that chelate Fe 3+ ions and render them available for reduction to the soluble Fe 2+ form preferred by plants [ 66 ]. The majority of isolates reported in this study that were able to produce siderophores belonged to the genera Pseudomonas , Acinetobacter, Microbacterium, Rhodococcus and Stenotrophomonas ( Figure S2 ). The widespread ability of our isolates to hold PGP traits related to increasing the concentration and availability of nutrients to plants is of great importance to the plant nutrition balance. This study highlights the functional potential of this culture collection in which many bacterial isolates, from the genera Acinetobacter, Arthrobacter, Nocardia, Rhodococcus, Streptomyces and Variovorax , possessed petroleum hydrocarbon degradation capabilities. However, only a small proportion of bacterial isolates (5%) had multiple PGP traits. These strains were isolated from the genera Acinetobacter , Enterobacter , Klebsiella , Pseudomonas and Serratia . Interestingly, in our study, only three bacterial isolates were capable of degrading all five PHCs, and had all five PGP traits ( Figures S3–S5 ): Pseudomonas putida strain EB3, Streptomyces sp. strain WT8 and Bacillus sp. strain WT32. These findings corroborate earlier studies which reported that many isolates from these genera can degrade PHCs and promote plant growth [ 51 , 67 , 68 ]. These bacterial taxa are candidates to look for in follow-up experiments."
} | 5,527 |
34529729 | PMC8445481 | pmc | 1,935 | {
"abstract": "To solve the problem of one-sided pursuit of the shortest distance but ignoring the tourist experience in the process of tourism route planning, an improved ant colony optimization algorithm is proposed for tourism route planning. Contextual information of scenic spots significantly effect people’s choice of tourism destination, so the pheromone update strategy is combined with the contextual information such as weather and comfort degree of the scenic spot in the process of searching the global optimal route, so that the pheromone update tends to the path suitable for tourists. At the same time, in order to avoid falling into local optimization, the sub-path support degree is introduced. The experimental results show that the optimized tourism route has greatly improved the tourist experience, the route distance is shortened by 20.5% and the convergence speed is increased by 21.2% compared with the basic algorithm, which proves that the improved algorithm is notably effective.",
"conclusion": "Conclusion and future work In order to optimize the tourism route planning problem, in this paper, an improved ant colony optimization(ACOCA) algorithm based on context-aware mechanism, pheromone updating strategy is proposed to balance route distance and user comfort degree. In the proposed ACOCA algorithm, we introduce the weather and comfort degree of scenic spots to decide the pheromone updating strategy, and avoid to fall into the local optimum value, the concept of sub-path support degree was set. In order to verify the optimization performance of the ACOCA algorithm, We select three groups of scenic spots with different regional ranges, each group includes 15 scenic spots, and make route planning for each group of scenic spots; At the same time, we also carried out simulation experiments on TSPLIB dataset. The experimental results show that the performance of ACOCA algorithm is compared with basic ACO algorithm, genetic algorithm and Hybrid-ACO algorithm on the two datasets, and its operation results are optimized in route distance, convergence time and user comfort ratio. In particular, compared with Hybrid-ACO algorithm, the travel route distance is reduced by 20.5% and the convergence time is reduced by 21.2%, At the same time, user travel comfort has also been greatly improved. ACOCA algorithm has better optimization ability and user comfort than the baseline algorithms. However, the efficiency of ACOCA algorithm needs to be further improved when the scale of scenic spots increases greatly. At the same time, we plan to include another contextual factors such as tourism cost into the algorthm. In the future work, the ACOCA algorithm will be studied deeply.",
"introduction": "Introduction With the development of mobile Internet technology, the user travel information is becoming more and more diversified. Travel notes, strategies, short videos and so on have become the main basis for the user tourism route planning. While tourism route planning is a kind of loop problem that tourists start from the starting point, pass through the selected scenic spots only once, and finally return to the starting point. Similar to (Travelling Salesman Problem)TSP, tourism route planning is a NP-Hard problem, but it is more complex and affected by many constraints. When tourists arrive in a city to visit the scenic spots in its territory, they are often affected by time, transportation conditions, economic cost and other factors. In practice, people often spend a lot of time on the selection of scenic spots and route planning. ACO algorithm is proposed by Dorigo et al. [ 1 ] according to the intelligent behavior of ant colony in the process of foraging. The algorithm has some advantages such as heuristic, positive feedback and distributed. ACO algorithm has been widely used to solve (Traveling Salesman Problem)TSP. Stutzle et al. [ 2 ] proposed Max-Min ant system (MMAS) to solve the problem but the improved algrithm is easy to fall into local optimum; Yang et al. [ 3 ] proposed an improved ACO algorithm based on game theory, which introduced entropy weighted learning strategy to optimize the accuracy of the optimal solution of (Traveling Salesman Problem)TSP problem on the basis of ACO and MMAS. Deng et al. [ 4 ] divided the optimization problem into several sub problems in order to improve the convergence rate of ACO algorithm and the pheromone update strategy was used to improve the optimization ability, then coevolution mechanism was used to exchange information among different sub populations, so as to avoid the ant colony falling into local optimum. Qian et al. [ 5 ] took the traffic cost as the calculation object, optimized the automatic adjustment mechanism of ant colony pheromone, and improved the performance of ACO algorithm. In addition, some researchers combined ACO algorithm with other algorithms. For example, Liang et al. [ 6 ] combined genetic algorithm with ACO algorithm to further improve the ability to solve the defects of local optimization, obtain the best path and save cost. Che et al. [ 7 ] combined particle swarm optimization with ACO algorithm to find the optimal path and improved the quality of path planning by improving pheromone update rules and heuristic function based on particle swarm optimization. Scholars have put forward a large number of improvement methods and strategies to solve various problems in different boundaries, different disciplines and different fields. Aiming at the problems of local optimization and slow convergence of ACO algorithm, Wang et al. [ 8 ] embedded genetic algorithm and cloud model into ACO algorithm to obtain the optimal solution. Han et al. [ 9 ] proposed an an algorithm called SOS-ACO that combines symbiotic organisms search (SOS) and ACO algorighm to calculate the optimal or near-optimal assembly sequence, which find the optimal or near optimal assembly sequence in fewer iterations. ACO algorithm is also widely used in feature selection. Zhou et al. [ 10 ] proposed a two-stage hybrid ACO for high-dimensional feature se-lection (TSHFS-ACO), which uses the interval strategy to determine the size of OFS for the following OFS search and helps to reduce the complexity of the algorithm and alleviate the algorithm from getting into a local optimum. In solving constraint satisfaction problem (CSP), in order to overcome the shortcomings of low solution quality and slow convergence speed based on ACO algorithm, Guan et al. [ 11 ] proposed an improved ant colony optimization with an automatic updating mechanism (AU-ACO). In order to take advantage of both epsilon greedy and Levy flight, Liu et al. [ 12 ] proposed a greedy-Levy ACO incorporating these two approaches to solve complicated combinatorial optimization problems, which outperforms max-min ACO and other latest solvers. It is useful to implement collective intelligence(CI) evolution using ant colony optimization (ACO), Gan et al. [ 13 ] analyzed the performance of ACO evolution algorithm and verified the feasibility of applying the collective intelligence (CI) evolution theory to a specific application. Besides ACO algorithm, traditional route planning methods include tabu search algorithm [ 14 ], genetic algorithm [ 15 , 16 ], particle swarm optimization algorithm [ 17 ], simulated annealing algorithm [ 18 ] etc. ACO algorithm is a parallel algorithm, the search process of each ant is independent, and ants communicate through pheromone. Therefore, ACO algorithm can be regarded as a distributed multi-agent system, which starts to search for independent solutions at multiple points in the problem space at the same time, which not only increases the reliability of the algorithm, but also makes the algorithm have strong global search ability. In this paper, a solution method of tourism route planning problem based on context feedback and ACO algorithm is proposed. In the above studies, the shortest distance is used as the goal of route planning. However, in the tourism route planning, in addition to the main factors that affect the route planning, the waiting time (comfort degree of the scenic spot) and the weather also have an important impact on the route planning. In ACO algorithm, ants communicate with each other through pheromones, and use learning mechanism to adjust the selection probability of the optimal path. However, the ACO algorithm also has some defects, such as the lack of pheromone in the initial stage of path construction, the evolution speed is very slow; with the ACO algorithm using positive feedback principle to strengthen the optimal solution, after a certain number of iterations, pheromone is mainly concentrated on a few routes, resulting in premature convergence. At the same time, from the perspective of adjusting the load of scenic spots, the routes to individual scenic spots also lead to a sharp increase in the number of tourists, resulting in a sharp decline in tourist experience and a longer waiting time. We propose a tourism route planning method based on contextual information feedback mechanism and ant colony optimization algorithm in this paper. The major works involved in this paper are as follows: Take distance, scenic spots comfort ratio, as comprehensive evaluation indicators to improve tourists experience while ensuring economic costs. The tourism route should pass through all the scenic spots selected by tourists, only once, and finally return to the starting point. Introduce the comfort degree of the scenic spot, weather and other contextual information into the improved algorithm, and dynamically adjusts the tourism route planning through self-learning. From the perspective of tourists, the tourism route planning should be people-oriented, reduce the economic cost of tourists and time, enhance the travel experience of tourist. This paper proceeds in the following order. In section II, introduce ACO algorithm and its lastest researches. In section III, the proposed ACOCA algorithm is described in detail, including contextual information modelling, and ACOCA algorithm description. In section IV, a series of performance tests are simulated and the results of the experiments are analyzed. In section V, the conclusions and our further research directions are given."
} | 2,560 |
39417072 | PMC11480031 | pmc | 1,936 | {
"abstract": "Introduction Despite extensive studies on soil microbial community structure and functions, the significance of plant-associated microorganisms, especially endophytes, has been overlooked. To comprehensively anticipate future changes in forest ecosystem function under future climate change scenarios, it is imperative to gain a thorough understanding of the community structure, diversity, and function of both plant-associated microorganisms and soil microorganisms. Methods In our study, we aimed to elucidate the structure, diversity, and function of leaf endophytes, root endophytes, rhizosphere, and soil microbial communities in boreal forest. The microbial structure and composition were determined by high-throughput sequencing. FAPROTAX and FUNGuild were used to analyze the microbial functional groups. Results Our findings revealed significant differences in the community structure and diversity of fungi and bacteria across leaves, roots, rhizosphere, and soil. Notably, we observed that the endophytic fungal or bacterial communities associated with plants comprised many species distinct from those found in the soil microbial communities, challenging the assumption that most of endophytic fungal or bacterial species in plants originate from the soil. Furthermore, our results indicated noteworthy differences in the composition functional groups of bacteria or fungi in leaf endophytes, root endophytes, rhizosphere, and soil, suggesting distinct roles played by microbial communities in plants and soil. Discussion These findings underscore the importance of recognizing the diverse functions performed by microbial communities in both plant and soil environments. In conclusion, our study emphasizes the necessity of a comprehensive understanding of the structure and function microbial communities in both plants and soil for assessing the functions of boreal forest ecosystems.",
"conclusion": "Conclusion In conclusion, our study elucidates the intricate relationships between microbial diversity, community structure, and functional roles in a boreal forest. The results underscore the significance of considering both bacterial and fungal communities in understanding the dynamic interplay between plants and their associated microbiota. Chemoheterotrophic bacteria emerged as the predominant functional group, exhibiting diverse metabolic capabilities across sampled positions, while fungal functional groups exhibited distinct distribution patterns that reflected their ecological roles. Notably, the leaf endosphere harbored unique microbial communities with specialized functional attributes, suggesting niche-specific adaptations and potential contributions to plant health and fitness. Furthermore, differences in specific bacterial functional groups, such as nitrogen-fixing bacteria and ammonia-oxidizing bacteria, across the sampling positions emphasize the significance of microbially mediated nutrient cycling and plant-microbe interactions. Future research should be focused on unraveling the functional significance of microbial communities in mediating plant-microbe interactions and ecosystem processes, ultimately enhancing our ability to harness the beneficial contributions of microbes for sustainable agriculture and environmental stewardship.",
"introduction": "Introduction Many studies have demonstrated the crucial role of the soil microbiome in maintaining the health of terrestrial ecosystems ( Chaparro et al., 2012 ; Jiao et al., 2018 ; Dubey et al., 2019 ; Banerjee and van der Heijden, 2023 ). This diverse community of bacteria and fungi contributes to various essential functions. One of its primary roles is nutrient cycles ( Bahram et al., 2018 ; Jiao et al., 2018 ; Sokol et al., 2022 ), where microorganisms decompose organic matter, releasing essential nutrients such as nitrogen, phosphorus, and potassium for plant uptake. Additionally, specific fungi can form symbiotic relationships with plant roots, aiding in nutrient absorption and enhancing plant growth ( Bonfante and Genre, 2010 ; Begum et al., 2019 ; De Mandal et al., 2021 ). Furthermore, the soil microbiome acts as a key player in disease suppression, with certain microorganisms inhibiting the growth of harmful pathogens ( De Corato, 2020 ; Pascale et al., 2020 ). Soil microorganisms also contribute to soil structure and stability, playing a crucial role in preventing erosion and promoting water retention ( Hartmann and Six, 2023 ). In essence, the soil microbiome is indispensable for sustaining healthy ecosystems and mitigating environmental challenges. In forest ecosystems, plants provide a diverse array of niches for the growth and proliferation of microorganisms. The plant microbiota, consisting of bacteria, fungi, and so on, colonizes all accessible plant tissues ( Trivedi et al., 2020 ). These microorganisms establish complex associations with plants, playing crucial roles in enhancing plant productivity and maintain health in natural environments ( Laforest-Lapointe et al., 2017 ; Trivedi et al., 2020 ). Previous studies have highlighted the significance of root microbiota in promoting plant growth and resilience to various stresses ( Philippot et al., 2013 ; Trivedi et al., 2020 ). Additionally, leaf-associated microorganisms influence host fitness, growth, resilience to abiotic stresses, and resistance to pathogens ( Vorholt, 2012 ; Laforest-Lapointe et al., 2017 ; Perreault and Laforest-Lapointe, 2022 ). Positive correlations between the diversity of tree-associated microbiota and ecosystem productivity have been observed, while decreases in diversity are linked to disease states and propagation ( Laforest-Lapointe et al., 2017 ; Perreault and Laforest-Lapointe, 2022 ). Moreover, a growing body of evidence suggests that diverse microbial communities associated with roots, leaves, and soil collectively contribute to enhancing plant fitness under environmental changes ( Giauque et al., 2019 ; Hawkes et al., 2020 ; Angulo et al., 2022 ). For example, the rhizosphere microbiome can also form a biological protective barrier, reducing the invasion of pathogenic microorganisms, which is especially important when plants face environmental stresses such as drought or disease outbreaks ( Mendes et al., 2013 ; Singh et al., 2023 ). Despite recognizing the vital functions of soil microorganisms and plant endophytes, previous studies predominantly focused on soil microorganisms and neglected the interconnected dynamics of soil, root, and leaf microorganisms, particularly in boreal forests ( Hassani et al., 2018 ; Hawkes et al., 2020 ; Angulo et al., 2022 ). Importantly, there is a gap in our understanding of microbial endophyte community structure, diversity, and functions in boreal forests. Addressing this gap is crucial, as it could limit our comprehensive understanding of the functions of soil microorganisms and plant endophytes across ecosystems. To comprehensively boreal forest ecosystem function, exploring the structure, diversity, and functional groups of microbial communities within leaves, roots, rhizosphere, and soil is essential. The microbial communities (including their structure, diversity, and functional groups) and their effects on plant systems can influence plant growth, health, and stress resistance, making it a critical aspect of understanding ecosystem dynamics. The living environment acts as a major selective force, shaping the composition of both soil microorganisms and plant endophytes, as highlighted by Trivedi et al. (2020) . Thus, we hypothesize significant differences in microbial community structure and diversity across various sampling positions (including bulk soil (non-rhizosphere soil), rhizosphere soil, leaf and root). Soil microorganisms, with their primary function in promoting soil nutrient cycling, are complemented by rhizosphere microorganisms, which enhance the efficiency of plant nutrient absorption. Simultaneously, plant endophytic microorganisms contribute to the adaptability of plants to their environment. Consequently, we further hypothesize significant differences in the functional groups of microbial communities in different positions, meeting diverse plant needs and ecosystem functions. To investigate these hypotheses, Larix gmelinii , the dominant tree species in the boreal forest, was chosen. We collected samples, including leaves, roots, rhizosphere, and soil, from 36 Larix gmelinii trees. Employing high-throughput sequencing, we determined the microbial community structure in different positions, with a specific focus on endophytic microbial communities identified in leaves and roots.",
"discussion": "Discussion In this study, the diversity of bacteria and fungi in soil was significantly higher than that of endophytic bacteria and fungi, indicating that host could act as a filter, selecting specific populations ( Trivedi et al., 2020 ). The higher bacterial diversity in bulk soil may be attributed to the diverse nutrient sources and environmental conditions available in this habitat ( Bach et al., 2018 ; Xia et al., 2020 ). On the other hand, the lower bacterial diversity in the leaf endosphere could be influenced by factors such as limited nutrient availability or selective pressures exerted by the plant host ( Bulgarelli et al., 2013 ; Trivedi et al., 2020 ). Similarly, the variation in fungal diversity across sampling positions underscores the importance of niche-specific interactions and ecological dynamics in the rhizosphere and endosphere compartments ( Qian et al., 2019 ; Yang et al., 2021 ). The greater fungal diversity in the rhizosphere compared to bulk soil may be linked to the rhizosphere effect, where root exudates influence the structure of the microbial community ( Haichar et al., 2008 ; Peiffer et al., 2013 ; Gong et al., 2023 ). Conversely, the lower fungal diversity in the root endosphere suggests potential selective processes or competition among fungal taxa for colonization within this niche ( Qian et al., 2019 ; Wang et al., 2023 ). Overall, these results shed light on the complex interplay between plants and their associated microbial communities. The significant variations in microbial diversity are accompanied by marked differences in bacterial and fungal community structures across the leaf endosphere, root endosphere, rhizosphere, and bulk soil. This divergence in community structure likely arises from difference in species composition. In the leaf endosphere, bacterial communities are predominantly represented by phyla Proteobacteria, Actinomycetes, Firmicutes, Acidobacteria, and Chloroflexi, with Proteobacteria comprising approximately 50% of the community composition across all sampled environments, which is consistent with findings from previous studies ( Hardoim et al., 2008 ; Compant et al., 2021 ). Moreover, plant endophytic communities exhibit enrichment in Actinobacteria, while displaying depletion in Firmicutes, Acidobacteria, and Chloroflexi compared to the rhizosphere and bulk soil. The observation is consistent with previous study ( Liu et al., 2017 ). The fungal communities inhabiting both leaf and root tissues demonstrate extensive diversity, primarily composed of phyla Ascomycota and Basidiomycota ( Zuo et al., 2021 ). Furthermore, Mortierellomycota phyla were enriched in the rhizosphere and bulk soil, while was depleted in plant endophytic communities. The differences in fungal community composition between plant tissues and bulk soil suggest selective colonization processes within the plant ecosystem ( Trivedi et al., 2020 ; Guo et al., 2021 ). Additionally, the analysis of specific species reveals that bacterial and fungal species in the leaf endosphere exhibit the highest endemism compared to other tissues, with relatively fewer specific species in the rhizosphere. This observation may be attributed to the distinct sources of endophytic microorganisms in leaves and roots, highlighting the influence of plant host specificity and environmental factors on microbial community assembly ( Vandenkoornhuyse et al., 2015 ; Xiong et al., 2021 ). These results underscore the intricate interplay between plant-associated microbial communities and their respective habitats, emphasizing the importance of considering both microbial diversity and community structure in understanding the ecological dynamics within the plant microbiome. This study provides insights that enhance our understanding of the complex relationships between plants and their associated microbial communities, highlighting the need for further investigation into the mechanisms driving microbial community assembly and function within different plant organs ( Bulgarelli et al., 2013 ). The difference of microbial community structure and diversity at different positions may lead to the difference in microbial community function. Our study highlights the diverse functional roles played by microbial communities across various ecological niches within the plant ecosystem. Chemoheterotrophic bacteria were the dominant functional group, with a relative abundance exceeding 50% across all sampled positions. Interestingly, significantly higher relative abundance of hydrocarbon-degrading bacteria (includes species like Pseudomonas aeruginosa , Rhodococcus erythropolis , Mycobacterium vanbaalenii , Sphingomonas paucimobilis , Pseudomonas putida , and so on) was found in leaf endosphere compared to in other positions, suggesting a potential role in detoxification or metabolism of organic compounds in leaf endosphere ( Gupta et al., 2017 ; Feng et al., 2017 ). Moreover, distinct patterns in specific bacterial functional groups were also observed across different sampling positions. The root endosphere and bulk soil showed a higher relative abundance of nitrogen-fixing bacteria compared to the leaf endosphere and rhizosphere, highlighting their importance in nitrogen cycling and plant nutrition ( Beltran-Garcia et al., 2021 ; Zhang et al., 2022 ). Conversely, the rhizosphere displayed a higher relative abundance of fermentation bacteria and nitrate-reducing bacteria, while the bulk soil harbored a greater prevalence of ammonia-oxidizing bacteria and nitrification bacteria, indicating niche-specific microbial metabolic activities ( Singh et al., 2022 ). In terms of fungal functional groups, ectomycorrhizal fungi reached their peak relative abundance in the root endosphere, rhizosphere, and bulk soil, underscoring their symbiotic association with plant roots and potential roles in nutrient uptake ( Philippot et al., 2013 ). Conversely, plant pathogenic fungi exhibited their highest relative abundance in leaves, suggesting a potential threat to plant health within this microenvironment ( Eberl et al., 2020 ; Trivedi et al., 2020 ). Additionally, the root endosphere, rhizosphere, and bulk soil displayed a significantly higher relative abundance of saprophytic fungi compared to the leaf endosphere, highlighting their role in organic matter decomposition and nutrient cycling in soil-associated habitats ( Schappe et al., 2020 ). Notably, a majority of leaf endophytes could not be assigned to familiar functional groups, suggesting unique functional roles that may significantly differ from those observed in other positions ( Trivedi et al., 2020 ). This underscores the need for further exploration to elucidate the specific functions and ecological significance of these enigmatic leaf endophytes within the plant microbiome. Overall, these findings provide valuable insights into the functional diversity and ecological roles of microbial communities in boreal forests, contributing to our understanding of microbial-mediated processes essential for plant health and ecosystem functioning ( Berendsen et al., 2012 )."
} | 3,941 |
26778946 | PMC4701906 | pmc | 1,939 | {
"abstract": "The exponential increase in data over the last decade presents a significant challenge to analytics efforts that seek to process and interpret such data for various applications. Neural-inspired computing approaches are being developed in order to leverage the computational properties of the analog, low-power data processing observed in biological systems. Analog resistive memory crossbars can perform a parallel read or a vector-matrix multiplication as well as a parallel write or a rank-1 update with high computational efficiency. For an N × N crossbar, these two kernels can be O ( N ) more energy efficient than a conventional digital memory-based architecture. If the read operation is noise limited, the energy to read a column can be independent of the crossbar size ( O (1)). These two kernels form the basis of many neuromorphic algorithms such as image, text, and speech recognition. For instance, these kernels can be applied to a neural sparse coding algorithm to give an O ( N ) reduction in energy for the entire algorithm when run with finite precision. Sparse coding is a rich problem with a host of applications including computer vision, object tracking, and more generally unsupervised learning.",
"introduction": "Introduction As transistors start to approach fundamental physical limits and Moore's law slows down, new devices and architectures are needed to enable continued computing performance gains (Theis and Solomon, 2010 ). The computational ability of current microprocessors is limited by the power they consume. For data intensive applications, the computational energy is dominated by moving data between the processor, SRAM (static random access memory), and DRAM (dynamic random access memory). New approaches based on memristor or resistive memory (Chua, 1971 ; Waser and Aono, 2007 ; Strukov et al., 2008 ; Kim et al., 2012 ) crossbars can enable the processing of large amounts of data by significantly reducing data movement. One of the most promising applications for resistive memory crossbars is brain-inspired or neuromorphic computing (Jo et al., 2010 ; Ting et al., 2013 ; Hasan and Taha, 2014 ; Chen et al., 2015 ; Kim et al., 2015 ). The brain is perhaps the most energy-efficient computational system known, requiring only 1–100 femtoJoules per synaptic event (Merkle, 1989 ; Laughlin et al., 1998 ), efficiently solving complex problems such as pattern recognition on which conventional computers struggle. Consequently, there has been great interest in making neuromorphic hardware (Cruz-Albrecht et al., 2013 ; Merolla et al., 2014 ). Resistive memories can effectively model some properties of neural synapses and the crossbar structure allows for high-density interconnectivity as found in the brain. For example, individual neurons in the cerebral cortex can receive roughly 10,000 input synapses from other neurons (Schüz and Palm, 1989 ). Resistive memories are essentially programmable two terminal resistors. If a higher write voltage is applied to the device, the resistance will increase or decrease based on the sign of the voltage, allowing the resistance to be programmed. Consequently, it can be used to model a synapse. Its resistance acts like a weight that modulates the voltage applied to it. This has resulted in a large interest in developing neuromorphic systems based on it (Jo et al., 2010 ; Ting et al., 2013 ; Hasan and Taha, 2014 ; Kim et al., 2015 ). Each cell also has a very small area and the memory can be stacked in 3d when arranged in a crossbar structure. Therefore, industry is developing resistive memories to use as a digital replacement for flash memory (Jo et al., 2009 ; Chen, 2013 ; Chen et al., 2014 ; Cong et al., 2015 ). A pressing question is whether neural-inspired computing systems are able to offer any resource advantage over more conventional digital computing systems. Neural-inspired systems are likely to take the form of a massively parallel collection of neuromorphic computing elements or cores that are each much simpler than conventional CPUs (Merolla et al., 2014 ). Conventionally, each neuromorphic core is based on a local SRAM memory array. This allows for data to be locally stored where it is used, eliminating the need to move large amounts of data. Simply organizing the computing system in this manner can provide 4–5 orders of magnitude reduction in computing energy (Cassidy et al., 2014 ). To get further benefits, the neuromorphic core should be based on an analog resistive memory crossbar array. Both digital and analog neuromorphic cores will have an execution-time advantage as parallelism is easier to leverage in a neuromorphic computational model where communication latency is drastically decreased. Nevertheless, in this work we avoid focusing on a new parallel architecture and instead focus on demonstrating a more fundamental advantage in energy. We will show that performing certain computations on an analog resistive memory crossbar provides fundamental energy scaling advantages over a digital memory based implementation for finite precision computations. This is true for any architecture that uses a conventional digital memory array, even a digital resistive memory crossbar. In addition we give a concrete neural-inspired application, sparse coding, which can be implemented entirely in analog and reap the aforementioned energy advantage. A rich neural-inspired problem is sparse coding (Olshausen, 1996 ; Lee et al., 2008 ; Arora et al., 2015 ), where one seeks to use an overcomplete basis set to represent data with a sparse code. It is used in many applications including computer vision, object tracking, and more generally unsupervised learning. We will show that analog neural-inspired architectures are ideally suited for algorithms like sparse coding, and outline an implementation of a specific sparse coding algorithm. Specifically, there are two key computational kernels that are more efficient on a crossbar. First, the crossbar can perform a parallel read or a vector-matrix multiplication as illustrated in Figure 1 . Second, the crossbar can perform a parallel write or a rank-1 update where every weight is programmed based on the outer product of the row and column inputs. These two kernels form the basis of many neuromorphic algorithms. Figure 1 Analog resistive memories can be used to reduce the energy of a vector-matrix multiply . The conductance of each resistive memory represents a weight. Analog input values are represented by the input voltages or input pulse lengths, and outputs are represented by current values. This allows all the read operations, multiplication operations, and sum operations to occur in a single step. A conventional architecture must perform these operations sequentially for each weight resulting in a higher energy and delay. In this paper we analyze the energy required to perform a parallel read and show that for a fixed finite precision, the noise limited energy to compute a vector dot product can be independent of the size of the vector, O (1), giving the analog resistive memory based dot product a significant scaling advantage over a digital approach. In the more likely situation of a capacitance limited energy, an N × N crossbar still has a factor of N scaling advantage over a digital memory. Similarly, writing a rank-1 update to a crossbar will also have a factor of N scaling advantage over a digital memory. We also analyze the energy cost of precision, energy scaling for communications, energy for accessing one row of the crossbar at a time and energy for accessing one element. Next, we show that these computational kernels can be used with a sparse coding algorithm to make executing the algorithm O ( N ) times more energy efficient.",
"discussion": "Discussion In this analysis we have deliberately avoided specifying constant factors as they can vary by orders of magnitude depending on the technology and design tradeoffs. Particular multiplicative constants apply only to today's hardware, but the big O remains whether new devices change these constants. For instance, the energy to write a resistive memory can be as low as 6 fJ (Cheng et al., 2010 ) or higher than 100 nJ (Mahalanabis et al., 2014 ). The energy for analog driving circuitry around a crossbar can also vary by orders of magnitude depending on the speed and circuit area tradeoffs. Depending on the algorithm, new semiconductor devices such as a spin based neuron (Sharad et al., 2014 ) could also drastically change the energy tradeoff. Nevertheless, it is still useful to consider some specific numbers to understand what is plausible. In running an algorithm on a resistive memory array there are three key components to the energy, the parallel read energy, the parallel write energy and the energy for the driving circuitry. To find the capacitance limited read or write energy we need the capacitance per resistive memory element. The capacitance per element (wire + resistive memory) in an array for a 14 nm process as specified by ITRS will be around 50 aF. If we need to charge the wires to 1 V, that corresponds to 50 aJ per element. For an N × N array the total capacitance limited read or write energy would be 50 × N 2 aJ. As discussed at the end of the Parallel Write Energy section, the current limited write energy could plausibly be on the same order of magnitude. The energy of the driving circuitry depends greatly on what computations are performed, but we can get an order of magnitude estimate by considering one of the most expensive analog operations, an analog to digital converter (ADC). ADCs that require only 0.85 fJ/level (or conversion step) have been demonstrated at 200 kHz (Tai et al., 2014 ). This means that for a 1000 × 1000 crossbar, the energy to run a six bit ADC is roughly the same as the energy to read/write a column of the crossbar. For higher precision ADCs, the ADC will dominate the energy, while for lower precision ADCs the crossbar will dominate the energy. In general, we see that the potential constant factors are on the same order of magnitude and consequently will be very technology dependent. In order to understand the theoretical benefits of a crossbar, we have assumed ideal linear resistive memories. In practice there are many effects that can limit the performance of a resistive memory crossbar in a real algorithm. Access devices are required to be able to individually write a given resistive memory. This limits how low of a voltage can be used. Non-linearities in the resistive memories as well as those introduced by the access device mean that the amount a resistive memory writes will be dependent on its current state. Read and write noise limit the accuracy with which the resistive memories can be read or written. Parasitic voltage drops mean that devices far away from the drivers see a smaller voltage. Despite all of these effects, recent studies are indicating that iterative learning algorithms can tolerate and learn around moderate non-idealities (Burr et al., 2015 ; Chen et al., 2015 ; Cong et al., 2015 ). Given the potential energy scaling benefits of resistive memory crossbars, more work is need to design devices with fewer non-idealities and to better understand how various algorithms can perform given the non-idealities. Overall, we have shown that the energy to perform a parallel read or parallel rank-1 write on an analog N × N resistive memory crossbar typically scales as O ( N 2 ) while a digital implementation scales as O ( N 3 ). Consequently, the analog crossbar has a scaling advantage of O ( N ) in energy. The communications energy between neighboring crossbars scales as O ( N 2 ). Thus, communications are not as important for digital approaches, but once we take advantage of an analog approach the communication energy and computation energy are equally important. For algorithms that operate on only one row of a matrix at a time, both the digital and analog energy scales as O ( N 2 ) per row. Therefore, the better approach will depend on the specifics of a given system. Algorithms such as sparse coding can directly take advantage of the parallel write and parallel read to get an O ( N ) energy savings. Thus, we have shown that performing certain computations on an analog resistive memory crossbar provides fundamental energy scaling advantages over a conventional digital memory based implementation for low precision computations. This is true for any architecture that uses a conventional digital memory array, even a digital resistive memory crossbar. Fundamentally, a digital memory array must be accessed sequentially, one row at a time, while an entire analog memory crossbar can be accessed in parallel. Analog crossbars perform a multiply and accumulate at each crosspoint, while digital memories need to move the data to the edge of the array before it can be processed. In principle, a digital system could be organized to process data at every cell, but the area cost would become prohibitive. Alternatively, optimized digital neural systems will have a processing in memory (PIM; Gokhale et al., 1995 ) type architecture where simple operations are performed near a moderately sized memory array (Merolla et al., 2014 ). While this will give orders of magnitude reduction in energy compared to a CPU (Cassidy et al., 2014 ), the fundamental scaling advantages of an analog crossbar array can further reduce the energy by a few orders of magnitude."
} | 3,378 |
32708476 | PMC7404020 | pmc | 1,940 | {
"abstract": "There are a large number of fouling organisms in the ocean, which easily attach to the surface of ships, oil platforms and breeding facilities, corrode the surface of equipment, accelerate the aging of equipment, affect the stability and safety of marine facilities and cause serious economic losses. Antifouling coating is an effective method to prevent marine biological fouling. Traditional organic tin and copper oxide coatings are toxic and will contaminate seawater and destroy marine ecology and have been banned or restricted. Environmentally friendly antifouling coatings have become a research hotspot. Among them, the use of natural biological products with antifouling activity as antifouling agents is an important research direction. In addition, some fouling release coatings without antifoulants, biomimetic coatings, photocatalytic coatings and other novel antifouling coatings have also developed rapidly. On the basis of revealing the mechanism of marine biofouling, this paper reviews the latest research strategies to develop environmentally friendly marine antifouling coatings. The composition, antifouling characteristics, antifouling mechanism and effects of various coatings were analyzed emphatically. Finally, the development prospects and future development directions of marine antifouling coatings are forecasted.",
"conclusion": "4. Concluding Remarks Every year, marine fouling organisms cause huge economic losses, and antifouling coating is currently the most direct and effective means of prevention. It is important to develop stable, durable and environmentally friendly antifouling coatings. The current antifouling coatings are difficult to cover in terms of stability, toxicity and cost of use, which seriously hinders its popularization and use. In the future, the research and use of marine antifouling coating will surely develop in the direction of being highly efficient, non-toxic, non-polluting and degradable. Antifouling coating containing antifoulant mostly outperform coatings without antifoulant. It is imperative to develop new and highly effective antifoulant to replace cuprous oxide, which is currently widely used. Obtaining natural antifouling substances from nature, especially extracting antifouling active substances from marine organisms, is the most ideal source of antifoulants. Through in-depth research on the antifouling mechanism of natural antifouling active substances, the development of highly effective and non-polluting antifouling agents is an important direction of the development of antifoulants. In addition, releasable coatings are environmentally friendly and have potential development prospects. The fouling release coating is mostly suitable for ships with high speed and fast travel, and it is less effective for ships that are moored for a long time or traveling at low speed. Coatings that rely on a single antifouling mechanism cannot achieve the desired antifouling effect, and future trend will need to integrate multiple methods. Composite coatings that integrate multiple antifouling mechanisms will surely become a research hotspot, especially some nano-composite and modified coatings. With the use of nanotechnology, marine antifouling technology has been greatly improved. When the material reaches the nanometer level, its performance will change qualitatively, and the surface performance of the coating or the release efficiency of the antifoulant will be significantly improved. In addition, combining antifoulant with fouling release coatings is also an important development direction.",
"introduction": "1. Introduction Marine fouling organisms refer to the general term for all kinds of marine organisms attached to the surface of marine facilities and causing damage to human marine economy [ 1 ]. Marine fouling organisms not only endanger the development of the marine economy, but also hinder human exploitation of the ocean. At present, more than 4000 marine fouling organisms have been discovered [ 2 ], such as microorganisms mainly including bacteria, diatoms and Ulva spores, etc.; large fouling organisms mainly include barnacles, bryozoans, mussels and algae. Various fouling organisms easily attach to ships, oil platforms and breeding facilities in the ocean [ 3 ]. Barnacles and other invertebrates are firmly attached to the hull surface by secreting a biological adhesive. As the barnacle grows, its edges will destroy the anticorrosion layer on the hull surface, thereby accelerating the hull corrosion [ 4 ]. As shown in Figure 1 [ 5 ], when the hull surface is attached, fouling organisms will corrode the hull surface and increase the roughness of the hull surface, thereby reducing the speed of the ship [ 6 , 7 , 8 ], increasing fuel consumption and greenhouse gas emissions [ 9 , 10 ]. In recent years, with the rapid development of offshore industries such as submarine oil and marine power generation, the damage of marine fouling organisms to man-made facilities has become more severe. For example, when an offshore oil platform is attached, it will increase the weight of the facility and weaken its ability to resist tsunami and storm risks. In addition, fouling organisms can block drainage pipes, jeopardizing the safety and service life of marine facilities. Some marine fouling organisms enter the non-native land by attaching to the bottom of the ship, thereby causing biological invasion and causing great harm to the global marine ecology [ 11 ]. Every year, a lot of money is invested in ship surface cleaning and maintenance of marine facilities. With the continuous development of the human marine economy, the economic losses caused by marine fouling organisms are getting larger and larger, thus, effective and economic prevention methods are getting more and more attention. The key to managing biological attachment is to block biological attachment from the source. Applying antifouling coating is currently the easiest and most widely used antifouling method. Traditional antifouling coating often use poisonous drugs to poison fouling organisms, the main components of which are organic tin [ 12 ] and cuprous oxide. However, the accumulation of metal elements in metal compounds in fish and shellfish will cause biological variation and death and pollute seawater, and thus harm the marine ecological environment. Taking into account the greater harm of organic tin to the marine environment [ 13 ], organic tin coatings have been banned worldwide. Although cuprous oxide is less toxic, as it continues to accumulate, its impact on marine ecology will become greater and greater. Therefore, the development of efficient and environmentally friendly marine antifouling coatings has become a research hotspot [ 14 ]. There are two main research directions for new antifouling coatings. One direction is to develop non-toxic and environmentally friendly antifoulants that replace metal compounds, that is, to find antifouling active substances from the ocean. An antifoulant developed by marine natural products can effectively reduce the harm to the environment, which is also one of the hot issues in current research [ 2 , 15 ]. In addition, some terrestrial biological products have also attracted attention. The purification or modification of natural products through chemical methods can obtain antifoulants with extremely strong antifouling capabilities. The second direction focuses on fouling release coatings, which will modify the surface characteristics of the material, making it difficult for fouling organisms to adhere or make them easier to remove after attachment. Due to the complex diversity of the marine environment, it is becoming more and more difficult for a single mechanism of antifouling coating to meet the requirements of use, and some composite antifouling coatings that combine multiple antifouling principles have also appeared. With the development and application of nanotechnology and polymer materials, the antifouling ability of antifouling coatings has been further improved [ 16 ]. This article first reveals the fouling mechanism of marine fouling organisms. On this basis, it mainly summarizes the development status of environmentally friendly antifouling coatings in three aspects from antifoulant type antifouling coatings, fouling release antifouling coatings and other important antifouling coatings. Our analysis reveals the antifouling mechanism of various antifouling coatings, and reviews the advantages and disadvantages of the new coating. Finally, the development trend of marine antifouling coatings is forecasted."
} | 2,145 |
38517961 | PMC10959417 | pmc | 1,942 | {
"abstract": "The limited capacity of typical materials to resist stress loading, which affects their mechanical performance, is one of the most formidable challenges in materials science. Here, we propose a bone-inspired stress-gaining concept of converting typically destructive stress into a favorable factor to substantially enhance the mechanical properties of elastomers. The concept was realized by a molecular design of dynamic poly(oxime-urethanes) network with mesophase domains. During external loading, the mesophase domains in the condensed state were aligned into more ordered domains, and the dynamic oxime-urethane bonds served as the dynamic molecular locks disassociating and reorganizing to facilitate and fix the mesophase domains. Consequently, the tensile modulus and strength were enhanced by 1744 and 49.3 times after four cycles of mechanical training, respectively. This study creates a molecular concept with stress-gaining properties induced by repeated mechanical stress loading and will inspire a series of innovative materials for diverse applications.",
"introduction": "INTRODUCTION Materials dominate all aspects of our life to such an extent that it is difficult to imagine life without them ( 1 – 4 ). Mechanical loading environments and internal stress concentration cause great damage to the typical materials ( 5 , 6 ) and remain a critical concern in the field of materials science. The internal stress gained during processing affects the mechanical stability of materials and dramatically shortens their durability and reliability. Furthermore, external stresses are repeatedly loaded on materials during service life, which accelerates aging and may eventually lead to their failure. The stress loading effect (SLE; strength changing after stress loading and unloading divided by the original tensile strength) is proposed to quantify the change in tensile strength. The SLEs of traditional polymers are negative. As a result of the disentanglements and slippages of molecular chains, high-density polyethylene ( 7 ) and thermoplastic polyurethane ( 8 ), both of which are thermoplastic (linear) polymers, exhibited an SLE of −0.0905 and −0.2441, respectively. In the case of thermosetting (cross-linked) polymers, the vulcanized natural rubber ( 9 ) also exhibited a negative SLE of −0.5506. Although their slippage during stretching is substantially retarded by the cross-linked network, which makes macromolecules difficult to disentangle, the restrained slippage hinders the internal stress to relax completely, and the accumulated internal stress causes the fracture of molecular chains until the catastrophic mechanical failure. In contrast to the negative SLEs observed in traditional polymers, several polymers with positive SLEs have been reported recently ( 10 – 15 ). These strategies are classified into two types. The first strategy is a physical rearrangement. Lin et al. ( 10 ) propose a mechanical training strategy to form aligned nanofibrils for the development of fatigue-resistant hydrogels. Inspired by the dynamic sacrificial bonds in biomaterials and the self-strengthening mechanism of skeletal muscles, Tu et al. ( 11 ) fabricated programmable artificial muscle material. Agrawal et al. ( 12 ) developed dynamic self-stiffening in liquid crystal elastomers under repeated low-strain compressive deformation. Liu et al. ( 15 ) used a strain-induced crystallization strategy to prepare a nondestructively reinforced hydrogel. However, in the absence of a chemical cross-linking network to fix the physical rearrangement, it is difficult to ensure its stability upon heating and during use. The second strategy is covalent network growth. By stabilizing the input of monomers in chemical reactions, Matsuda et al. ( 13 ) developed self-growing hydrogels induced by mechanophores. However, the requirement of continuous monomer supplements is usually infeasible in practice. Seshimo et al. ( 14 ) prepared self-strengthening polyurethane elastomers by force-induced cross-linking reactions. The cross-linking reaction of the pre-reserved reactive groups was used to enhance the storage modulus; however, it embrittled the polymer. Yu et al. ( 16 ) used a water-assisted cross-linking reaction mechanism to prepare a self-strengthening polymer, which, however, required aquatic environment. In addition, both SLE and modulus changes after stress loading for all the developed polymers were limited. Bone is regarded as a natural stress-gaining material ( 17 – 19 ). The bone structure is capable of remolding to strengthen itself upon cyclic training with appropriate stress. External tension induces dissociation (unlocking) of the covalent cross-links between collagens. The consequent unlocked mineralized collagen fibrils are aligned into highly ordered structures and covalently cross-linked (locked) again upon unloading to substantially improve the stiffness and strength of the bone ( Fig. 1A ) ( 20 , 21 ). Here, we propose a bone-inspired “dynamic molecular locking” strategy to use the loading stress for polymer strengthening. A covalently cross-linked polyurethane with mesophase domains (MCPU) was designed using the reversible oxime-urethane bonds as dynamic molecular locks, the dissociation of which allowed a gradual release of stress. Meanwhile, the stress induced the rigid segments to align into highly ordered domains in the condensed state, which were locked in the reorganized polymeric covalent network based on dynamic oxime-urethane bonds (see “Design, characterization of polymers and the optimization of training process” in the Supplementary Materials). The cyclic alignment and locking of mesophase domains continuously increased the molecular order to substantially strengthen and toughen the resultant MCPU polymers ( Fig. 1, A and B ). Fig. 1. Design and outcome of stress-gaining polymers. ( A ) Schematics of the stress-gaining process of native bone (a) and covalently cross-linked polymers via dynamic molecular locking (b). ( B ) Strength and modulus changes after stress loading of traditional, state-of-the-art self-strengthening polymers and our polymer [( 9 ): traditional materials; ( 15 ): slide-ring hydrogels; ( 12 ): polydomain nematic liquid crystal elastomers; ( 10 ): freeze-thawed polyvinyl alcohol hydrogel; ( 14 ): segmented polyurethane elastomers; ( 11 ): poly(ethylene-propylene-diene monomer) elastomer; ( 13 ): double-network hydrogels; ( 16 ): water-strengthening polyurethane].",
"discussion": "DISCUSSION Here, we proposed a bone-inspired material concept of stress gaining. We proved that a dynamic poly(oxime-urethanes) with mesophase domains converted typical destructive stress into a favorable factor to substantially strengthen itself, distinguishing it from traditional stress-weakening materials subjected to applied stresses. The stress induced the rigid segment to be aligned into ordered mesophase domains in the condensed state, which was facilitated and locked by the reorganization of the covalent polymeric network via the reversible lock and unlock of the oxime-urethane bonds. The responsive stress-gaining effect on the mechanical properties via dynamic molecular locking strategy in our study is unprecedented, with an increase in the elastic modulus and tensile strength by three and one order of magnitude, respectively. The stress-gaining concept represents a paradigm in material design. This will inspire innovative molecular strategies to improve the performance of materials under dynamic mechanical loads."
} | 1,879 |
33801700 | PMC8065543 | pmc | 1,944 | {
"abstract": "Lignocellulose is a promising feedstock for biofuel production as a renewable, carbohydrate-rich and globally abundant source of biomass. However, challenges faced include environmental and/or financial costs associated with typical lignocellulose pretreatments needed to overcome the natural recalcitrance of the material before conversion to biofuel. Anaerobic fungi are a group of underexplored microorganisms belonging to the early diverging phylum Neocallimastigomycota and are native to the intricately evolved digestive system of mammalian herbivores. Anaerobic fungi have promising potential for application in biofuel production processes due to the combination of their highly effective ability to hydrolyse lignocellulose and capability to convert this substrate to H 2 and ethanol. Furthermore, they can produce volatile fatty acid precursors for subsequent biological conversion to H 2 or CH 4 by other microorganisms. The complex biological characteristics of their natural habitat are described, and these features are contextualised towards the development of suitable industrial systems for in vitro growth. Moreover, progress towards achieving that goal is reviewed in terms of process and genetic engineering. In addition, emerging opportunities are presented for the use of anaerobic fungi for lignocellulose pretreatment; dark fermentation; bioethanol production; and the potential for integration with methanogenesis, microbial electrolysis cells and photofermentation.",
"conclusion": "5. Conclusions Despite their biotechnological relevance and their prevalence as a critical component of ruminant biology for recovering resources from plants, anaerobic fungi remain relatively unexplored as platform organisms for lignocellulosic breakdown and biofuel production. Mimicking natural rumen and hind-gut environments in a scalable bioreactor remains a formidable challenge. The development of robust, low energy, heterologous, scalable processes that are able to make use of complex lignocellulose substrates is critical for effective process-scale production of biofuels. Except for AD reactors, most bioprocessing strategies have been well developed and optimised for aerobic microorganisms with very different growth requirements. Nevertheless, boundless opportunities arise to exploit the enzyme systems and metabolism of anaerobic fungi, whether by producing genes in heterologous platforms, by developing the means to genetically engineer the fungi directly, or by repurposing existing or developing new bioreactor designs. Such work provides an opportunity to potentially produce not only biofuels but platform molecules from one of the most abundant and renewable feedstocks on the planet.",
"introduction": "1. Introduction New approaches are needed to reduce the use of fossil fuels and harness the global abundance of lignocellulosic biomass for biofuel production. Lignocellulosic feedstocks can be converted to biofuels but they require costly and environmentally damaging pre-treatments in order to overcome their inherent recalcitrance to degradation [ 1 , 2 ]. Anaerobic fungi, belonging to the phylum Neocallimastigomycota, may provide a green solution because they have an unprecedented ability to deconstruct crude lignocellulose [ 3 , 4 ] and are able to convert polymeric plant cell wall components to H 2 [ 5 ] and ethanol [ 6 ]. This group of organisms can also produce volatile fatty acids (e.g., acetic and formic acid [ 5 ]) which are suitable substrates for additional downstream biofuel production. Anaerobic fungi are commonly found in the digestive tracts of large mammalian herbivores, including many important livestock and companion animal species such as cattle, sheep, goats and horses. Prior to their correct affiliation [ 7 , 8 , 9 ] zoospores of anaerobic fungi were mistakenly classified as protozoa. Callimastix frontalis and C. equi zoospores, found in horse intestines, were both described as polyflagellated protozoans [ 10 , 11 ] and placed in the same genus as C. cyclopsis , a parasite of freshwater copepods [ 12 , 13 ]. Monoflagellated zoospores, found in ruminants, were also recognised as protozoa and assigned to the protozoan genera Piromonas and Sphaeromonas [ 10 , 14 ]. The discovery that C. cyclopsis was a fungal pathogen (belonging to the Blastocladiomycota ) led to the transfer of C. equi and C. frontalis to a new protozoan genus, Neocallimastix , with N. frontalis as the type species [ 15 ]. Following the seminal work of Orpin [ 7 , 8 , 9 ] and their correct assignment as fungi, based upon the ultrastructure of their motile zoospores, anaerobic fungi were initially placed in the order Spizellomycetales but later transferred to their own order, the Neocallimastigales, in the phylum Chytridiomycota [ 16 ]. Genetic analysis identifies that Neocallimastigomycota is a distinct basal clade of the chytrids [ 17 ]. The order, which now houses 18 different genera of anaerobic fungi, was therefore raised to the level of a phylum, the Neocallimastigomycota, in 2007 [ 17 , 18 ]. The Neocallimastigomycota, Blastocladiomycota and Chytridiomycota are all closely related. Although fungi in the latter two phyla are aerobic and mostly found in fresh water and wet soils (some are parasitic), fungi in all three phyla display similarities in their adaptations to an aquatic lifestyle. These adaptations include a dependence on zoospore release in the aquatic environment for dispersal (via asexual reproduction) and similarities in vegetative thallus morphology, including the ability to grow on and within surfaces and substrates. Additionally, a dormant phase has been observed in all three phyla, where the fungi can survive relatively adverse conditions, sometimes for many months [ 19 , 20 , 21 ]. However, the Neocallimastigomycota differ from the blastoclades and chytrids in their anaerobic lifestyle and flagella apparatus. They also possess hydrogenosomes and a complete absence of mitochondria [ 19 , 22 , 23 ]. From an evolutionary perspective, these three phyla are basal fungal clades with species and genera that are the direct decedents of some of the earliest diverging fungal lineages. It has been proposed that that the Neocallimastigomycota diverged from other primitive aquatic fungi during the late Cretaceous period when grasses and grazing mammalian herbivores first appeared [ 24 ]. Due to their highly effective ability to convert lignocellulose into biofuels and biofuel precursors, anaerobic fungi are biotechnologically interesting. In this review, the challenges and opportunities associated with exploiting anaerobic fungi for the purpose of industrial biofuel production are discussed. Prior to discussing the challenges and opportunities for their exploitation in the biofuel industry, a review of the niche anaerobic fungi occupy in the mammalian digestive tract is presented, drawing in particular upon the substantial amount of literature involving ruminant livestock. This is necessary in the context of this review, as an in-depth appreciation of their natural niche will assist in developing appropriate methodologies for their exploitation in an industrial context."
} | 1,791 |
36133608 | PMC9417570 | pmc | 1,945 | {
"abstract": "A method to electrically induce memristor performance from inkjet-printed silver (Ag) nanoparticles is presented, which is effective on a specifically designed hourglass-shaped Ag metal device. Joule heating-induced oxidation in the bottleneck region, when applying a high current to the device, results in a metal-electrolyte-metal structure produced from just a single metal ink for the memristor operation. This electrically induced memristor shows a nonuniform dispersion of the Ag nanoparticles within the oxide electrolyte layer, depending on the bias polarity adopted during the initial metal rupture process. A versatile and useful range of controllable memristor behaviors, from volatile threshold switching to nonvolatile unipolar as well as bipolar resistive switching, are observed based on the reversible rejuvenation and rupture of the Ag nanofilaments according to the Ag cation migration within the oxide electrolyte. The interplay between the electric field induced redox reaction and thermal diffusion of the Ag nanoparticles constitutes the primary reason for the different switching behaviors, further supported by thermo-field simulation results. The bipolar switching memristor demonstrates reliable endurance even under harsh DC switching conditions with low power consumption compared with its unipolar switching operation. The observed range of controllable switching behavior can be exploited for future low power flexible memory, as a selector in crossbar memory architecture, synaptic learning, and others.",
"conclusion": "Conclusions An hourglass-shaped inkjet-printed Ag nanocluster pattern has been shown to enable memristor operation after an initial rupture of its pristine metallic state via high current flow through the device. The inkjet printing method is very efficient for flexible electronics, as it does not require additional lithography or etching steps to accomplish the desired geometrical deformation of the device. In addition, the metal-oxide electrolyte-metal system is generated here with only a single metal ink and creating the active layer in situ through Joule heating and rupturing in the bottleneck region of the hourglass shape device. The chemical analysis in this region reveals that the Joule heating leads to severe oxidation of the Ag nanoparticles into Ag 2 O insulator. Within the Ag 2 O switching matrix, the generation and rupture of Ag filaments according to the size and polarity of the applied electric field lead to the resistive switching behavior. The initial Ag nanoparticle distribution that is determined by the bias polarity applied during the initial metal rupture stage supports the subsequent asymmetry about the switching polarities. The corresponding switching mechanisms for various switching characteristics ranging from unipolar/bipolar non-volatile switching to threshold switching, all in a controllable manner, were postulated considering both the Joule heating and electric field induced filament dissolution. The electrical method of initiating memristor behavior from the simple printed electrode pattern simplifies the fabrication efforts for printing all the functioning MIM layers and the demanding ink formulation of each individual layer material. Such convenience for attaining the versatile inkjet printed memristors will further pave the way toward their application in the future low-power applications on flexible substrates.",
"introduction": "Introduction Since the first suggestion of the memristor as the basic circuit element correlating charge and flux in 1971 ( ref. 1 ) and experimental demonstration in 2008, 2 memristor research has become very active. The device features a non-linear switching behavior in its resistance values according to the change in the charge flow across the device, which is controlled by the applied voltage/current pulse. A variety of materials showing voltage- or current-controlled resistive switching behavior have been reported to be the flux- or charge-controlled memristors. 3 A wide range of applications is possible based on the highly scalable and low power consuming switching that stems from the simple two-terminal metal-insulator-metal (MIM) structure. 4,5 The nonvolatile resistance switching random access memory (RRAM) has been aiming to replace the incumbent NAND flash technology. 6–8 These devices are also highly promising as the main functional units for the next generation computing paradigms such as in-memory computing 9–11 and neuromorphic computing. 12–14 Most of the memristor work to date has been performed using thin film processing techniques compatible with high-density memory fabrication. However, this may not be ideal for the flexible electronics field. Inkjet printing is highly promising to fabricate memristor devices on flexible substrates because it allows deposition of very small volumes of ink (picolitres) rapidly, achieving high pattern precision and resolution with greater reproducibility compared to other techniques such as screen-printing. 15 In addition to the compact and inexpensive manufacturing capability compared with conventional semiconductor processing including deposition, lithography, etching, and cleaning, 16 inkjet printing is one of the most versatile methods that combines deposition and patterning in a single process without the adoption of a mask. 17,18 Nevertheless, research on inkjet printed memristors has not been as active as those of other electronic devices with rather larger dimensions such as organic transistors, biosensors, and energy storage devices. 19–21 This may presumably be due to the difficulties in securing an appropriate combination of metal electrodes and dielectrics, within the range of available inkjet printable materials, to construct the memristor. First is the choice of defect species that can readily migrate within the dielectric layer and switch the electrical conduction behavior of the device. Second is the need to obtain the materials in the form of jettable ink: the size of the nanoparticles comprising the ink should be smaller than 1/50 of the nozzle orifice to avoid nozzle clogging. The ink viscosity and the surface tension should also be controlled appropriately for high jettability. 22 The inkjet printing method is well suited for organic inks with low viscosity to fabricate the organic-dielectric-based resistive switching devices. 15 However, they inherently have critical reliability issues 23,24 and the switching mechanisms are yet to be established compared to their inorganic counterparts. 25–28 In contrast, transition metal oxides are the most widely studied dielectrics for memristors. 29,30 But, transforming them into printable inks is challenging as it involves many additives and formulation procedures to form stable suspension. 31–34 HfO 2 ink for RRAM fabrication is commercially available, 35 but very little option exists for other materials. Another alternative to obtain inkjet printed metal oxide dielectric layers is to post-anneal a printed metal layer in an oxidizing environment, which requires an additional etching step to reveal the metal contact underneath the oxidized layer. 36 If the device is supposed to be formed on flexible and thermally fragile substrates, the annealing temperature is severely limited, making the stable oxide growth difficult. All these attempts undermine the most critical merit of the inkjet-printed fabrication method, which is simplicity. Here we address this issue by using just single metal ink and generating the active oxide medium in situ through Joule heating in a specially optimized hourglass-shaped device pattern comprising of the electrodes and the active switching layer. Such novel approach of electrically generating an oxide printed device is the simplest yet to print MIM structures for memory operation. Furthermore, this simple method is able to achieve various modes of switching in a single device based on bias polarity and current compliance control. Silver (Ag) nanoparticle is one of the most feasible electrode material in the inkjet printing system. 37–40 It is also one of the most common electrode material for the cation migration-based memristor, which is called electrochemical metallization cell (ECM). 29,30 In an ECM structure, metals with high oxidation potentials such as Ag or Cu easily oxidize into cations within the insulator (electrolyte) layer where they migrate toward the counter electrode, become reduced and generate a conducting filament comprised of their nanoclusters across the dielectric layer. Recent progress in this cation-based memristor system reveals various filament growth and rupture behavior depending on related parameters of the dielectrics. 41,42 Here, we describe a simple and novel electrical method to initiate memristor operation based on the Ag migration of inkjet printed Ag nanoparticles. The geometrically engineered metal (Ag) device printed on the polyimide substrate enables locally enhanced Joule heating, through which an oxide electrolyte can be generated in situ at the center of the device. Based on the resulting Ag/Ag:Ag 2 O/Ag stack, a distinctive feature of this electrically generated memristor is elucidated. The relatively low ion mobility of Ag in Ag 2 O leads to disparate filament growth behavior compared with the ECM devices adopting conventional electrolytes such as chalcogenides.",
"discussion": "Results and discussion The I – V characteristics in Fig. 1a show that the printed metal nanoparticles have been completely sintered during the post-drying and annealing process, showing uniformity in the device-to-device distribution on the polyimide flexible substrate. The inset of Fig. 1a is an optical image of a planar 2p-ink A device. 1p-inkB was fabricated in exactly the same shape, with no visible difference to the device shown in the Fig. 1a inset based on the optical microscope observation. Despite the single-pass printing, ink B shows slightly smaller resistance (∼5 Ω) compared to ink A with 2-pass printing (∼8 Ω). The resistances of single-pass printed ink A (1p-inkA) were slightly high and not as uniform as the others (ESI Fig. S1a † ) implying that they are somewhat incompletely and stochastically sintered, possibly due to the low nanoparticle composition. Intriguingly, as shown in ESI Fig. S1, † several devices commonly encountered abrupt rupture of their metallic properties, and once ruptured, the devices showed a reversible resistive switching upon additional voltage application. This phenomenon could originate from reversible fuse and anti-fuse of the silver nanoparticles at a certain region of the device where the device suffers from locally enhanced Joule heating. Such local heating could be possibly caused by the incomplete sintering and hence the non-uniform resistance across the device. Based on these findings, resistive switching was also attempted with metallic 1p-inkB and 2p-inkA through a novel design of the device geometry. An hourglass-shaped pattern was fabricated so that the resistance could be locally enhanced in the bottleneck region with a narrow gap (50–70 μm) between two 200 μm-long planar ends. The resulting bottleneck region had various widths ranging from 20 to 50 μm, due to the inherent stochasticity of the printer. The bottleneck parts are indicated with red arrows in the inset of Fig. 1c . Both in the planar and hourglass shaped devices, the left-electrode was biased while the right electrode was grounded. This geometry allows spatially confined resistive switching, which is necessary for conducting related analysis as well as for implementing the devices into a memory array. As a result, the rupture of pristine metallic behaviour, which was seen only in the high resistance planar devices (1p-inkA) above, occurred even in these hourglass shape devices made from the 2p-inkA and 1p-inkB conditions. However, as shown in Fig. 1b , the 1p-inkB with its smaller resistances in the planar device form showed the initial rupture process at a high current level over 50 mA. The device stayed in its insulating state afterward even with the high voltage application up to ±10 V, meaning that no resistive switching is possible under this condition. The fabricated pristine device in this work is a metallic conductor, whereas the oxide-based RRAM or ECM usually has a pristine insulation state, so the initial rupture process corresponds to the electroforming process. In contrast, Fig. 1c shows the metallic behaviour of 2p-inkA device ruptured by a relatively low current between 10 mA and 30 mA, and a reliable successive resistive switching behaviour is seen in Fig. 1d . According to the multiple current–voltage ( I – V ) curves of the initial metal rupture in the hourglass shape 2p-inkA devices, the slope of the I – V curves commonly decreases at the verge of the rupture event as seen in the inset of Fig. 1c . This provides evidence for the Joule heating effect along with the high current flow, which is accompanied by the gradual increase in its metallic resistance, immediately before the rupture. The resistance of a metal linearly increases with temperature, being proportional to the characteristic thermal coefficient of the metal. Since the amount of Joule heating (= I 2 R ) is proportional to the resistance when a constant current flows through the devices, the bottleneck part of the hourglass shape is likely to become the favorable part for such metal rupture to take place. Fig. 1d shows typical resistive switching I – V curves of the hourglass 2p-inkA devices after an initial metal rupture event. The red and green curves represent switching with positive and negative bias polarities respectively in the devices where the initiation, i.e. , the initial metal rupture, is performed under the positive bias condition. A compliance current value of 500 mA was applied during set operations to avoid the permanent breakdown of its insulating state and to resume the reversible resistive switching between the low and high resistance states. Due to the much higher metallic resistance after a set process relative to the pristine state, the following set processes are attributed to only a partial recovery from its metallic nature. This is most likely due to the rejuvenation of nanofilaments according to the migration of silver nanoparticles at the initially ruptured region of the device. The detailed switching mechanisms are discussed with the aid of Fig. 3 and 4 later. The present approach based on inkjet printing is useful in effectively enhancing the local Joule heating. Fig. 2a through Fig. 2c provide scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images revealing the geometry of an hourglass shape memristor device (2p-inkA) with the reduced lateral dimension of the bottleneck region (down to 20 μm) and the thickness of the Ag electrode (500 nm). For an accurate detection of the Ag thickness of the bottleneck region, a blanket 100 nm Pt layer was coated on top of the device to avoid severe charge accumulation on the polyimide substrate during focused ion beam (FIB) milling. The resultant cross-sectional TEM image in Fig. 2c reveals a 40–60 nm-thick polycrystalline Ag layer, for which a protrusion on the surface becomes the most preferred nucleation and growth site for the Ag cluster-filaments. The fabrication method of leaving a small gap between the planar Ag electrodes here has effectively reduced both the width and thickness at the bottleneck region. Based on the measured width and thickness of the electrode and bottleneck region, Fig. 2d shows the simulated temperature distribution across the entire device during the initial Ag metal rupture using COMSOL simulation (details given in ESI † ). The electrical conductivity (2 × 10 6 S m −1 ) and the average current level (20 mA) at the verge of metal rupture were extracted from Fig. 1a and c . The substrate is not shown in the device model image of Fig. 2d for simplicity. The hourglass device with only the narrower width (20 μm) in its bottleneck region (upper right corner of Fig. 2d ) shows a small increase (about 5 °C) in its temperature above the ambient of 20 °C. However, the device with both narrower (20 μm) and thinner (50 nm) dimension of Ag metal in its bottleneck region shows more than the 100 °C increase within the region. The width and thickness of other parts are 80 μm and 500 nm, respectively. Such a dramatic thickness modulation across a single device is possible through a one-step printing described here, which otherwise would have required complex lithography and etching steps. The experimental result from a thermal emission measurement shown in Fig. 2e provides a consistent spatial temperature distribution on the device when applying a short AC pulse. Here, a 20 V high pulse height and 30 ns pulse duration were adopted to produce sufficient heat generation within the detectable range. Although the instrument does not provide the exactly calibrated temperature, the hottest spot is obviously placed within the bottleneck region, midway between the contact probes. Once the temperature on the hot spot exceeds a certain level that Ag can endure, the spot is supposed to turn into an insulator (Ag 2 O). Fig. 2 (a) Top-view SEM image of the bottleneck part, (b) cross-sectional SEM image of the planar electrode part, (c) cross-sectional TEM-EDX image of the bottleneck part of an hourglass shape memristor device. (d) The simulated temperature increases in the hourglass shape device based on the geometry observed by the microscopy images, (a) through (c). Inset is the simulated result assuming a geometry that has a uniform thickness across the entire device, i.e. , both the electrode and bottleneck part. The local temperature increase is mostly attributed to much thinner Ag of the bottleneck part compared to that of the electrode part. (e) Experimental evidence of the local temperature increase confined in the bottleneck region observed by a thermal emission detection measurement. (f) Optical microscope image of a pristine hourglass shape memristor (top), the device after going through tens of switching cycles (middle), and a device that showed abnormally high metal rupture current and no following switching behavior (bottom). (g) EDS analysis at position 1 (bottleneck) and 2 (electrode). The upper and middle panels of Fig. 2f show an optical microscopy image of the memristor device prior to and after tens of repetitive switching cycles, respectively. Specifically, the bottleneck region after the switching shows an obvious configurational change from its pristine state. This further confirms that the resistive switching in the hourglass shape device takes place at that part of the device, as intended. Fig. 2g shows EDX (energy-dispersive X-ray spectroscopy) energy spectrum results for the bottleneck region after resistive switching and the other intact metal region. Each position on the device is indicated by a dark and light green mark in the inset device image. The bottleneck region has higher (lower) oxygen (Ag) concentration compared with the metal electrode position. Therefore, it can be postulated that the initial metal rupture generates an oxide layer, most likely a silver oxide, and the switching occurs according to the interplay between the electrode as cation source and the oxide insulator that simultaneously serves as an ionic conductor for the cation migration. Ag 2 O is the most common form of silver oxide, generated easily when silver is exposed to oxygen in the ambient temperature. It is known that the thickness of the oxide increases with temperature up to a certain value, after which the oxide starts to decompose into silver and oxygen. 43,44 The oxide decomposition temperature ranges from 500–700 K in air, depending on the film thickness and fabrication method. A conductive silver device is generated during fabrication owing to the annealing temperature, low enough to avoid any silver oxide generation. Upon biasing, however, Joule heating will increase the temperature, generating the insulating silver oxide at the weakest part of the hourglass device. Once enough silver oxide is available to block the current flow, it will prevent further Joule heating so that a stable insulating state is retained. This is reflected in the I – V curves of the initiation process ( Fig. 1c ), where the current stays at lower value once the rupture of the conducting state happens. In contrast, the lowest panel of Fig. 2f shows one peculiar case, where the device has gone through the initial metal rupture with a very high current almost up to 100 mA, mainly due to an abnormally thick bottleneck part. No resistive switching followed in such high current induced rupture case, whose insulating state featured an order of magnitude lower current level than those from the other cases (ESI Fig. S1c † ). As shown in the figure, the resulting change at the bottleneck is much more severe. Therefore, it can be concluded that just like in the above-mentioned hourglass 1p-inkB case ( Fig. 1b ), some of the 2p-inkA devices with anomalously large bottleneck cross-section cannot initiate subsequent resistive switching despite the obvious metal rupture event, due to the excessive energy applied during the rupture. Such huge energy consumption would generate too thick an insulator (rupture region) as shown in the lowest panel of Fig. 2f so that even a high voltage up to ±10 V cannot trigger the Ag nanoparticle migration within the insulator layer. There could also be a possibility that the high energy consumption might have resulted in a too stoichiometric insulator, which does not contain sufficient defect sites for the cations to migrate through. \n Fig. 3a shows reliable unipolar resistive switching (URS) at both bias polarities for the 2p-inkA device. An on/off ratio over 10 8 at a reading voltage of 0.02 V is attained irrespective of the switching polarities. This kind of unipolar resistive switching is often alternatively referred as non-polar resistive switching, implying that the switching is hardly influenced by the polarity of the applied electric field. Instead, the mechanism relies on the size of the power consumed during the switching, originated from Joule heating. Typically phase change mechanism (PCM) and thermochemical mechanism (TCM) correspond to this type of behaviour. The former involves a transition between the amorphous and crystalline phases of certain chalcogenide materials, widely adopted for phase change random access memory (PRAM), which has little relevance to the oxide switching matrix here. Instead, it is likely that the present non-polar resistive switching phenomenon occurs based on generation and Joule heating-induced rupture of the Ag filaments within the Ag 2 O switching matrix, that is, close to TCM. This is normally accompanied by an abrupt set and reset operation, which is consistent with Fig. 3a . A set voltage is the voltage required to incur electric field-induced Ag cation migration within the switching matrix and as they become reduced and accumulated within the insulator, Ag metallic filament is generated bridging the two metal electrode portions. On the other hand, reset is attributed to the thermally activated diffusion of the Ag metal clusters, thus normally accompanied with the high current flow, which is even higher than the maximum current limit or compliance current assigned during its preceding set process (see Fig. 3a and b ). Fig. 3 (a) I – V curves of unipolar resistive switching (URS) operation at positive bias polarity (left) and negative bias polarity (right) of the memristor device. (b) Proposed Ag mobile ion distribution after a ‘positively’ biased initial metal rupture (left). Conically shaped nanofilament generation by a negatively biased set and hourglass shaped nanofilament generated by positively biased set (right). (c) Simulated temperature and electric field distributions for the hourglass and conical shape nanofilaments. The temperature increase overrides the electric field effect in an hourglass filament, whereas the electric field dominates in the conical filament. (d) Schematic diagram of the devices during a positive bias and negative bias URS. Both URS operations rely on thermal rupture of the filaments, presumably due to the isotropic diffusion or dissolution of Ag particles into the oxide electrolyte. However, due to its strong electric field, the conical shape is likely to go through an evolution toward an hourglass shape before it becomes ruptured. There is, however, a characteristic difference in the switching behaviours depending on the bias polarity, as shown in Fig. 3a . When the device was initially ruptured and subsequently switched with a positive bias, the set state showed mostly simple ohmic behaviour (reset curves of left-hand panel of Fig. 3a ). In contrast, when the device was initially ruptured with a positive bias but subsequently operated with a negative bias, the set state showed additional change as indicated by the “secondary set” in the right-hand panel of the same figure. The on-state resistance decreases at ∼0.15 V in the negative bias set case, meaning that an additional set switching occurs at this voltage during each reset cycle. This can be comprehended in conjunction with the adopted bias polarity during the initiation, which should have led to asymmetry in the initial distribution of mobile Ag nanoparticles within the switching matrix. Since these devices are positively biased during the initial metal rupture stage, most of the Ag cations are likely to move toward the other side (right electrode, RE) of the positively biased electrode (left electrode, LE) during the rupture, as described in the schematic diagram of Fig. 3b . Given this initial asymmetric Ag distribution within the insulator, the negative bias can generate an Ag filament once a little portion of the mobile ions segregated at the RE side reaches the LE side. That is, a conical shape filament having the smallest radius in the vicinity of LE is generated. On the other hand, since the additional Ag cations should be introduced through the oxidation of the LE during the positive bias set, an hourglass shape Ag filament that has the smallest radius in the middle of the switching matrix is likely to be generated. Such a model based on the different filament geometry originating from the asymmetric initial nanoparticle distribution is further supported by a simulated thermo-field distribution shown in Fig. 3c . The conical filament (right-hand panel) was assumed to be comprised of a few atoms for its minimum radius (3 nm), and a slightly larger minimum cross-section (5 nm) was assumed for the hourglass filament (left-hand panel). This is because a larger amount of Ag nanoparticles should be involved to form the hourglass filament due to the additional incorporation of Ag nanoparticles into the oxide layer. This is further supported by the larger positive bias set voltages than the negative set bias voltages, which implies that the positive bias set kinetics involves the additional dissolution of relatively stable Ag electrode particles into the oxide. Although the geometric parameters on the active switching region are unknown, a direct comparison between the two cases can be made by considering the plausible parametric range based on previous reports. The detailed simulation environment is given in ESI Section 2. † The comparison reveals that the thermal and electric field effects are predominant in the hourglass and conical filament, respectively. As shown in the schematic diagram of Fig. 3d , an hourglass shape filament generated by the positive set bias is immediately ruptured once it reaches enough temperature for its isotropic dissolution back into the oxide switching matrix, or rupture. However, the conical shape filament incurred by a negative set suffers from the strong electric field that leverages the additional redox reaction of the Ag filament. Tracing the direction of the electric field represented by the black arrows in the conical shape (bottom right panel of Fig. 3c ), the additional Ag migration would eventually cause the evolution of the conical filament toward an hourglass shape before it reaches enough temperature required for its thermal rupture. This is reflected by the secondary set process in its I – V behaviour before its reset process takes place (right panel of Fig. 3a ). The proximity of the narrowest region of the conical filament to the bulk electrode region incurs heat loss, so the thermal effect has less significance than the other case. Surprisingly enough, despite the symmetric device structure and silver electrode at both sides, bipolar operation is also available in this system, likely due to the asymmetric nanoparticle distribution introduced at the electrical switching initiation process. In Fig. 4a , the set and reset are accomplished at the negative and positive bias polarities, respectively. This operation mode is beneficial in decreasing the switching current (and power too). As shown in Fig. 4b , the switching is done with lower reset current ( I max,reset ) than the compliance current assigned during the preceding set operations, which was not the case in the URS operations of Fig. 3 . This is because reset is mainly governed by the electric field induced oxidation, whereas it is due to thermal rupture in URS, and thus, is comparable to the conventional ECM. The schematic diagrams describing each switching step are included in the insets of Fig. 4a . Based on its much smaller operation current ( Fig. 4b ) and write voltage below 0.5 V (the inset of Fig. 4c ), the present memristor in its bipolar scheme is the most feasible candidate for the printed non-volatile memory application. There is also a clear distinction between the set and reset voltages, which was not the case for URS in Fig. 3 . The cycle endurance of the memory operation lasted over 100 cycles even in the harsh DC measurement mode, implying it could last much longer if applied with much shorter AC pulse switching. Despite the DC bias limited endurance, it is still competitive to its metal oxide counterparts on flexible substrates. 35,36 The non-volatile characteristics of both memory states lasted over 3 × 10 4 s at room temperature. The seemingly unstable high resistance state, in fact, comes from the measurement limit of the instrument, which cannot give accurate data below 10 −13 A. Fig. 4 (a) I – V curves of bipolar memristor operation and the corresponding schematics for each switching step (insets). (b) Maximum current during the reset process ( I max,reset ) of each switching mode. The bipolar operation features reset current smaller than the current compliance ( I cc,set ), which supports the typical redox reaction based switching mechanism, i.e. , electrochemical metallization. (c) Endurance and (d) retention characteristics of the bipolar switching mode. Both set and reset voltage size lower than 0.5 V implies for the promising power efficient printed memristor operation (inset of (c)). (e) Volatile threshold switching behavior of the memristor. As opposed to the conventional electrochemical metallization (ECM) theory, where two electrodes are different as an active (Ag, Cu) and an inert (Pt, Au) metal to enable bipolar switching, the present device performs a robust bipolar operation despite its symmetric electrode adoption with active metals at both sides. And this can be successfully understood in conjunction with the asymmetric nanoparticle distribution secured during the initiation process and the resulting uneven electric field distribution across the filament. The proposed model assumes the Ag nanofilament growth beginning from the cathode to anode, generating a conical filament that has richer nanoparticle composition near the anode side. In fact, the filament growth beginning from the cathode side has been a norm in the conventional ECM theory, as the cations can immediately migrate toward the cathode in the electrolyte and the filament growth kinetics is mainly governed by the redox reaction of the cathode/electrolyte interface. Unlike in the conventional electrolytes such as Ag 2 S and chalcogenide materials, however, the filament growth within the Ag 2 O oxide electrolyte is postulated to be limited by the low ion mobility of the electrolyte. Such a conical shape metal filament with richer Ag nanoparticles in its anode side has been directly observed previously using in situ TEM, 41 further supporting the present switching model described above. Due to the presence of trapped electrons or the electrons injected into the oxide electrolyte, the Ag cations have higher chance of being reduced before reaching the cathode side. 42 Besides non-volatile resistive switching, a volatile resistive switching, namely threshold switching, is also performed by adjusting a compliance current below 50 μA during the set operation, as in Fig. 4e . Despite the obvious set transition or abrupt current jump in the I – V curve, the device remains in the off state in the following voltage sweep at voltage level < holding voltage ( V hold ) under the compliance current fixed at 10 μA. Under such a small compliance current, a relatively weak Ag filament comprised of a few Ag nanoclusters is generated and is readily ruptured by the thermal agitation once the electric field is removed. The filament relaxation back to the Ag nano-particles has recently been verified through in situ TEM observation, and it was ascribed to the Ostwald ripening of the Ag clusters mainly to minimize the filament surface energy in the oxide switching matrix. This has little relevance to the previously described redox based mechanism of Ag ions for the non-volatile switching. Such Gibbs Thomson effect-induced threshold switching in the Ag electrode system is extensively studied for its bio-mimicking capability and sensor applications. 45–48 From a double-sweep result, where the voltage is gradually increased and then sequentially decreased, the minimum voltage at which the device remains on state, V hold , is evaluated. It is slightly above 0 V, and such a small V hold is highly favourable for selector application of the crossbar array architecture. Two-terminal selectors, as well as non-volatile memory components, are essential components for the ultra-high density crossbar memory. Typical two-terminal selectors include threshold switch and diode with high I – V non-linearity or rectifying property, by which the sneak (leakage) currents within the crossbar architecture can be impeded. Moreover, this type of volatile switching element is highly promising for the emulation of synaptic learning, when combined with an appropriate non-volatile switching element that operates in an analog manner. 12"
} | 8,746 |
38586583 | PMC10997342 | pmc | 1,947 | {
"abstract": "Growing interest in renewable energy continues to motivate new work on microbial biohydrogen production and in particular utilizing Escherichia coli a well-studied, facultative anaerobe. Here we characterize, for the first time the H 2 production rate and capacity, of E coli isolates from the 50 000th generation of the Long-Term Evolution Experiment. Under these reaction conditions, peak production rates near or above 5 mL per hour for 100 mL of lysogeny broth (LB media) was established for the ancestral strains and batch efficiencies between 0.15 and 0.22 mL H 2 produced per 1 mL LB media were achieved. All 11 isolates studied, which had been aerobically cultured in minimal media since 1988, exhibited a decreased H 2 production rate or capacity with many strains unable to grow under anaerobic conditions at all. The genomes of these strains have been sequenced and a preliminary analysis of the correlations between genotype and phenotype shows that mutations in gene ydjO are exclusively observed in the two isolates which produce H 2 , potentially suggesting a role for this gene in the maintenance of wild type metabolic pathways in the context of diverse mutational backgrounds. These results provide hints towards uncovering new genetic targets for the pursuit of bacterial strains with increased capacity for H 2 production as well as a case study in speciation and the control of phenotypic switching.",
"conclusion": "6 | CONCLUSIONS We have demonstrated, through the use of LTEE isolates graciously provided by the Lenski lab, that prolonged culture in an aerobic environment is capable of altering the growth strategy and fitness of E coli when reintroduced to an anaerobic environment. We expect these results to extend to other facultative anaerobes evolved under similar conditions. Such populations are of great value not only to the evolutionary biology community but also to the biohydrogen community as their genomes contain mutations in regions important for the regulation of anaerobic metabolism. The identification of these mutations, as negative examples, may reveal strategies for the design of novel microbes with improved hydrogen production rate and yield. To this end, we believe future work focused on the anaerobic culture of the LTEE strains on defined media and the subsequent analysis of spent media for metabolic byproducts and intermediates for pathway identification is motivated.",
"introduction": "1 | INTRODUCTION Growing interest in renewable energy and sustainable manufacturing continues to put a spotlight on hydrogen production. While a range of renewable resources are in a mature state of development including wind and solar, hydrogen may be utilized not only as a fuel source but also as an energy carrier. 1 Biological means of hydrogen production may be less energy intensive than chemical methods, through the capitalization of organic waste products and reaction conditions at less demanding temperatures and pressures. 2 These economic considerations motivate the mapping of cellular metabolic pathways and the identification of efficient routes towards products of interest. The near-ubiquitous capacity for anaerobic fermentation among microbes utilizing a broad array of substrates, 3 has afforded the identification of a host of bacterial and, more recently algal, 4 , 5 candidates. Owing to its easy cultivation, fast growth rate, and mature position as a classic model organism, many studies have focused, and continue to focus, on the use of Escherichia coli : 2 a facultative anaerobe commonly found in the lower intestine of warm-blooded animals, a largely anaerobic environment. 6 Many strategies have been employed to improve hydrogen production rate and yield. In addition to optimizing reaction conditions for a given microbe, it is possible to genetically engineer novel strains for improved production in a given environment. A range of target genes have been identified and successful mutants have been generated. 7 , 8 On the other hand, finding mutants with decreased or cessated hydrogen production 9 is also valuable. Knowledge of these mutants expands the list of target genes known to be important or essential for anaerobic metabolism. Extended culture in an aerobic environment may be expected to select for adaptations in a strain of E coli which could adversely affect its fitness or otherwise alter its growth when returned to an anaerobic environment. The identification of such adaptations and the genomes of such mutants would not only shed light on basic questions regarding evolutionary phenotypic selection and metabolic regulation, but potentially establish a set of genes as targets for engineering strains with higher H 2 yield or better efficiency. Motivated by this hypothesis we set out to characterize, to our knowledge for the first time, the H 2 production of strains produced during the Long-Term Evolution Experiment (LTEE) from the lab of Richard Lenski. 10 The genomes of the strains utilized have been sequenced 11 which enabled us to perform an analysis correlating genotype to H 2 output phenotype. We observed a subset of evolved strains to grow slowly, in comparison to their ancestral strain, or not at all when cultured anaerobically in LB media despite observing robust growth during aerobic culture for all strains. We further observed another subset to grow robustly when anaerobically cultured but fail to produce H 2 . We hope these observations highlight the potential interest of these strains and their genomes to the bioH 2 community as a negative example towards the goal of engineering more productive or more efficient strains. These results further serve as a case study in speciation where adaptation to a novel environment over a reasonably short period of time, on an evolutionary timescale, brought about the complete loss of access to the native metabolic phenotype for a subset of the population. We hope this work brings to light yet another facet of the LTEE which may prove useful towards understanding the control of phenotypic switching and the evolutionary balancing act of remaining adaptable while yet adapting to a specific environment.",
"discussion": "5 | DISCUSSION Our observation that all 11 of the LTEE isolates tested presented with a reduced H 2 production rate or capacity when cultured in LB highlights the potential interest of these strains and their genomes 11 to the bioH 2 community. These results may be utilized similarly to those of previous studies demonstrating genetic variants which ceased hydrogen production. 9 While these negative examples lack the potential for immediate application that mutants with improved yields possess, 7 , 8 they provide insight into the larger signaling framework responsible for maintaining access to anaerobic growth strategies. In fact, no single gene or simple mutational signature was found to be suggestive of the loss of H 2 production or the capacity for anaerobic growth (see Figure 3C ) in these strains indicating these strains likely leverage different pathways to reach the same phenotype. On the other hand, the only two strains which maintain H 2 production display mutations in gene ydjO , one mutation for each strain, both SNPs, 25 nucleotides apart. It is possible that these variations in ydjO are somehow stabilizing access to the wild type metabolic pathways in the context of diverse mutational backgrounds. While believed to be protein-coding, the function of this gene is not well understood 15 , 16 and we hope this work will motivate further investigation into the potential role of ydjO in anaerobic respiration. Multiple growth strategies for anaerobic respiration have been established for E coli . Modulating the carbon source, 17 and the removal of metabolic intermediates 18 prove effective methods for optimizing the yield of desired byproducts. Even doping with oxygen, 19 usually a “poison” for dark fermentation, has been shown capable of improving H 2 yield showcasing the multiple roles every chemical actor may play in the regulation of anaerobic metabolism. With the design of increasingly sophisticated reaction conditions, including the introduction of novel nanomaterials 20 we edge closer to a complete picture of the possible suite of responses to external stimulus. Inclusion of even more exotic nanostructures could serve to enhance the desired output or the incorporation of additional functionality into the reactor including sensors or fuel cells that feed off the product stream of the microbial reactor. 21 , 22 It is tempting to speculate given the data presented in this study how growth strategies differ among these isolates. The striking balance between the reduction in peak H 2 concentration and increased duration of production in the first phenotype resulting in a well conserved range of batch efficiencies and comparable cell numbers suggests that strains exhibiting this phenotype follow the same or similar pathways as their ancestor but with reduced rate constants. Strains exhibiting the second phenotype appear to access different pathways than their ancestors which do not produce H 2 as a byproduct and strains exhibiting the third phenotype appear to have lost either the ability to access any pathways affording anaerobic respiration or the necessary environmental sensing and phenotypic switching to recognize and meet the need. These results further motivate a discussion of the relative fitness of these strains when anaerobically cultured in LB media. It is a far reaching result that the LTEE strains continue to exhibit fitness gains relative to their ancestral strains when measured under culture conditions similar to those in which they are evolving. 23 It may be expected that due to environmental specialization over time, LTEE isolates may exhibit reduced fitness relative to their ancestors when placed in a very different environment even when the ancestral strains are well adapted to that environment. This is the case for strains exhibiting phenotype 3 and while direct competition assays were not completed, it may be expected to be the case for strains exhibiting phenotype 1. On the other hand, strains exhibiting phenotype 2 grew robustly when cultured anaerobically and without investigating direct competition, it is impossible to predict fitness gains or losses. The mutational spectra of a descendant of strain 606 evolving under anaerobic conditions was observed to differ dramatically from that obtained evolving under aerobic conditions 24 , 25 indicating that adaptations beneficial in an anaerobic environment are likely to bear little resemblance, taken at face value, to those acquired during periods of aerobic evolution. It is still possible, however, that adaptations acquired in one environment, even one both constant and tightly regulated, confer a fitness advantage in unrelated environments. Features which regulate the adaptability of an organism and not a specific adaptation, likely related to the control of nongenetic heterogeneity within the population, would generalize well to many environments eliciting a stress response. 26 , 27 It would be interesting to see how the convergence of the LTEE strains to the three distinct phenotypes under the culture conditions described here may relate to the strains’ aerobic phenotypes, 28 the discrete nature of the metabolic pathways occupied, and the underlying control of phenotypic switching."
} | 2,861 |
30449766 | PMC6307996 | pmc | 1,949 | {
"abstract": "The surface crust that caps highly weathered banded iron formations (BIFs) supports a unique ecosystem that is a post-mining restoration priority in iron ore areas. Geochemical evidence indicates that biological processes drive the dissolution of iron oxide minerals and contribute to the ongoing evolution of this duricrust. However, limited information is available on present-day biogeochemical processes in these systems, particularly those that contribute to the precipitation of iron oxides and, thus, the cementation and stabilization of duricrusts. Freshly formed iron precipitates in water bodies perched on cangas in Karijini National Park, Western Australia, were sampled for microscopic and molecular analyses to understand currently active microbial contributions to iron precipitation in these areas. Microscopy revealed sheaths and stalks associated with iron-oxidizing bacteria. The iron-oxidizing lineages Sphaerotilus , Sideroxydans , and Pedomicrobium were identified in various samples and Leptothrix was common in four out of five samples. The iron-reducing bacteria Anaeromyxobacter dehalogens and Geobacter lovleyi were identified in the same four samples, with various heterotrophs and diverse cyanobacteria. Given this arid, deeply weathered environment, the driver of contemporary iron cycling in Karijini National Park appears to be iron-reducing bacteria, which may exist in anaerobic niches through associations with aerobic heterotrophs. Overall oxidizing conditions and Leptothrix iron-oxidizers contribute to net iron oxide precipitation in our sampes, rather than a closed biogeochemical cycle, which would result in net iron oxide dissolution as has been suggested for canga caves in Brazil. Enhancements in microbial iron oxide dissolution and subsequent reprecipitation have potential as a surface-crust-ecosystem remediation strategy at mine sites.",
"discussion": "Discussion The microscopic and molecular evidence presented here supports a role for microorganisms in the present-day precipitation of iron oxides within Karijini National Park. Iron precipitation on cell membranes results from passive metal adsorption and precipitation on functional groups (for an overview on the underlying mechanisms, see [ 27 , 39 ]). In these Karijini samples, we noted different mineral assemblages around cells even within the same sample ( e.g. , fine and dense accumulation, Fig. 4B ; cell surface precipitation vs precipitation in a matrix of extrapolymeric substances, Fig. 4C ), suggesting that differences in cell envelope properties or perhaps microbial activity contribute to different mineral structures. Oxygenic microbes potentially promoted denser mineral assemblage due to oxygen production at their cell surface, resulting in rapid iron oxide precipitation. The detection of iron-oxidizing microbial lineages in our samples also suggests a contribution from direct enzymatic iron oxidation. Features interpreted as stalks produced by some neutrophilic iron oxidizers (compare Fig. 4A and D to Fig. 1B in Krepski et al. [ 30 ]; Fig. 1C in Suzuki et al. [ 43 ]; Fig. 2E and F in Emerson et al. [ 12 ]) were evident. Known stalk-producing iron-oxidizing bacteria ( e.g. , some members of Gallionellaceae and Zetaproteobacteria ) were not detected using molecular methods, and, thus, potentially originate from novel iron oxidizers e.g. , see Krepski et al. ( 30 ). Structures reminiscent of sheaths produced by members of the Leptothrix-Sphaerotilus group ( 40 ) were also observed, although Leptothrix was not a major OTU (70 th most abundant) in one of the samples (Dales Gorge pond) in which they were readily identifiable using LM. Many of the sheaths in that sample appeared to be empty, which is common for Leptothrix ( 11 ), and may explain the lower number of 16S rRNA gene sequences detected than the multitude of sheath structures. All known neutrophilic oxygen-dependent lithotrophic iron oxidizers from freshwater systems are Betaproteobacteria ( 22 ). The samples examined in the present study had a lower proportion of Betaproteobacteria at 4–13% of all sequences than other freshwater iron mats, wetlands, or groundwater seep communities, at between 26 and 75% of all clones as Betaproteobacteria ( 2 , 3 , 17 , 46 ). Within Betaproteobacteria , Gallionellaceae represented ≤2.3% of sequences in Karijini samples, in contrast to up to 25% of all clones in other studies ( 2 ), while Burkholderiaceae (into which Lepthothrix-Sphaerotilus is classified) represented up to 6.0% of all clones; however, Leptothrix is sometimes not detected by molecular methods in other natural iron-rich systems ( 3 , 46 ). Similar to other iron-rich natural environments ( 2 , 3 , 46 ), iron-reducing bacteria were also present in four of the five samples—none were detected in the iron precipitates coating plants on the wall at Dales Gorge. The two OTUs (OTU 2 and 12) from potential iron-reducing bacteria were shared between the four samples, and A. dehalogens (OTU 2) was a major OTU in each of these samples (2 nd , 3 rd , 10 th , and 11 th most abundant in Dales pool, Knox mat, Knox spot, and Hancock Gorge samples, respectively). The highest concentrations of soluble iron were measured in the Dales Gorge pool water sample in which A. dehalogens was the predominant OTU, while the lowest soluble iron concentration (7–28-fold lower than other samples) was measured in the water dripping out of the wall at Dales Gorge in which iron oxides were precipitating on plants, but known iron-reducing bacteria were not detected. Based on aqueous pH values (pH 5–6), the soluble iron measured in these samples was interpreted as ferrous iron ( 41 ). Iron-reducing and iron-oxidizing bacteria have the ability to exist in close proximity to each other in the aerobic zone and contribute to the ongoing cycling of iron even though iron reduction is an anaerobic process ( 17 , 35 ). In environmental studies, the co-existence of iron-oxidizing and -reducing bacteria is often indicative of the coupling of microbial iron oxidation and reduction processes ( 2 , 3 , 46 ). Laboratory studies have demonstrated that it is the activity of iron-reducing bacteria at a microscale anoxic-oxic interface, rather than the activity of iron-oxidizing bacteria, that drives iron oxide precipitation at circumneutral pH in an overall oxidizing environment ( 35 ). The presence of abundant and shared iron-reducing bacteria in four of the five Karijini iron precipitates sampled in the present study is consistent with the hypothesis proposed by Roden et al. ( 35 ) from observations based on microcosm experiments. Iron oxide accumulated in the natural environment sampled for the present study, rather than remaining trapped in a ferrous-ferric cycle between iron-oxidizing and-reducing bacteria ( 35 ). Therefore, iron-reducing bacteria may exist in close association with aerobic heterotrophs or other oxygen-consuming bacteria in their immediate vicinity to maintain micro-anoxic niches for iron reduction, while iron-oxidizing bacteria exist further away at a higher oxygen concentration in the same sample. The iron-oxidizing bacterium that was also detected in four of the five Karijini samples was a Leptothrix OTU, and Leptothrix has been shown to prefer higher oxygen concentrations ( 10 ) than the microaerophilic neutrophilic iron oxidizers that may keep iron trapped in a ferrous-ferric cycle in association with iron reducers ( 35 ). It is important to note that our results differ from recent findings in Brazil. Parker et al. ( 34 ) linked microbial iron reduction in the highly weathered BIF systems in Brazil to cave formation, i.e. , the overall broad scale dissolution rather than net formation of iron oxides. This requires further investigation and is potentially related to oxygen gradients and the presence of organic matter to drive long-term anaerobic conditions for net mineral dissolution that may lead to cave formation, whereas the niches sampled in Karijini National Park for the present study were aerobic overall and appeared to be exposed to ongoing active oxygen generation based on the abundance of the phototrophic oxygenic microbial lineages detected. This may be a critical factor in the overall net precipitation of iron oxides that leads to the ongoing formation and stabilization of iron duricrusts in these ecosystems. The iron precipitate on plants growing on the wall at Dales Gorge was broadly the most different microbiologically of the five samples, with only one major OTU potentially linked to iron cycling (OTU 101) and no phototrophic microbial lineages. The low soluble Fe in this sample ( Table 1 ) suggests that the active biogeochemical cycling of iron (particularly iron reduction) may be limited despite the abundant amounts of visible iron precipitates. This was the only sample that was not recovered from a sitting water body, which may be a large driver for the difference in the microbial community structure from those in other samples. The mineralized stalks evident in this sample ( Fig. 4D ) may originate from novel iron oxidizers. Stalk formation is an important trait for neutrophilic iron oxidizers and is one of the identifying features of this metabolism in nature ( 30 ). However, without supporting functional gene or cultivation data, it is impossible to speculate which of the OTUs identified in this sample may be linked to stalk formation and iron oxidation. Seven of the top 10 OTUs in this sample originate from species for which there is presently no close cultured relative; thus, apart from an indication of microbial ammonia oxidation, there is little else that may be inferred about microbial functionality in this sample at this stage. The abundance of microbial sequences from uncultivated or deep branching lineages in all samples in the present study is of interest ( Fig. 6 and 1C–G ). Many of these sequences were only classified to the class or order level and the absence of close cultured relatives indicates that it is impossible to infer anything about the role of these organisms in the environment based on phylogeny alone. We examined correlation coefficients between OTUs and the geochemical data collected at each site and found correlations with some elements ( Fig. 6 ). These OTUs from uncultivated lineages that correlated positively with Al, Fe, and/or Mn must not be overlooked as potential contributors to metal cycling in Karijini and included OTUs 33, 34, 19, 47, 44, and 62 ( Fig. 6 ). In the future, with cultivation and/or genomic information on relatives to these uncultivated and often phylogenetically deep-branching microorganisms, their possible function in nature will become clearer. In terms of the ecological or biotechnological significance of the present results, a replication of the natural iron oxide formation processes observed in the field may be used as a strategy to promote the formation of fresh iron oxides that contribute to the cementation of iron-rich duricrusts e.g. , in post-mining restoration attempts to reform surface cangas. Overall processes were consistent in several samples collected from across the park. Iron reduction by A. dehalogens and G. lovleyi was fundamental in driving the reductive dissolution of iron oxide minerals, and these iron reducers may exist in close partnership with general aerobic heterotrophs unique to each sample. Phototrophy was critical for the ongoing generation of oxygen to drive aerobic conditions, which is where this system may differ from cave formation processes (and net iron oxide dissolution) in Brazilian canga areas. Heterotrophic iron-depositing Leptothrix species, which grow at sufficiently high concentrations of oxygen to be distant from iron reducers, may have contributed to overall net iron oxide precipitation. Based on these commonalities, the provision of nutrients (N, P) and an appropriate electron donor for iron reducers ( e.g. , lactate or acetate) and a carbon source for Leptothrix and heterotrophs ( e.g. , glucose or peptone) under mining remediation conditions may be sufficient to drive the reduction and subsequent re-oxidation of abundant iron oxides in, for example, crushed BIF or waste iron oxides from mining under water-saturated conditions. Under natural sunlight and with abundant water and nutrients, naturally present phototrophic oxygenic organisms thrive, providing the conditions that favor fresh iron oxide precipitation at the oxic-anoxic interface. With dehydration (natural evaporation under the arid conditions around Karijini), freshly formed iron oxides transform to more crystalline phases ( 6 ). Thus, the stimulation of basic microbial processes and subsequent dehydration of the system may be a platform for the reformation of iron-cemented duricrusts from crushed BIF or promotion of the stabilization of iron oxide waste piles."
} | 3,240 |
33952515 | PMC8099194 | pmc | 1,951 | {
"abstract": "Electron acceptor availability affects the net isotopic fractionation of a global biogeochemical reaction in marine sediments.",
"introduction": "INTRODUCTION The anaerobic oxidation of methane (AOM) with sulfate mitigates the flux of methane (CH 4 ) from the seafloor into the water column ( 1 , 2 ). AOM is mediated by anaerobic methanotrophic (ANME) archaea that reverse the methanogenesis pathway to oxidize methane completely to carbon dioxide (CO 2 ). As ANME archaea cannot reduce sulfate, which is the dominant terminal electron acceptor in marine AOM, they form consortia with sulfate-reducing bacteria ( 3 – 5 ). ANME-1 archaea prevail in deep, diffusion-driven systems not only in anoxic microbial mats but also in hydrothermally heated sediments ( 6 – 8 ). The cultured thermophilic ANME-1 form consortia with the partner bacterium Candidatus Desulfofervidus auxilii ( 9 ). In these consortia, as well as in other AOM consortia, reducing equivalents are exchanged by direct interspecies electron transfer, which is enabled by a high abundance of cytochromes and pili-based, likely conductive, nanowire networks ( 5 , 10 , 11 ). The isotopic composition of methane is commonly used to trace its origin and fate in the environment. For carbon and hydrogen, the isotopic compositions, expressed as ratios of the heavy to the light isotope ( 13 R = 13 C/ 12 C and 2 R = D/ 1 H, with D = 2 H), are usually reported in δ notation relative to standards (std): δ = ( r R sample − r R std )/ r R std − 1 [per mil (‰)], where “ r ” denotes the rare isotope and the standards are Vienna Pee Dee Belemnite (VPDB) for carbon and Vienna Standard Mean Ocean Water (VSMOW) for hydrogen. The isotopic fractionation between a substrate (S) and the product (P) of a reaction is described by α r S / P = R r S / R r P (1) Close to isotopic equilibrium, methane is depleted in 13 C and D relative to CO 2 and H 2 O, respectively ( 12 , 13 ). Farther from equilibrium, kinetic effects contribute to the net isotopic fractionation, and the usually faster reaction of the substrate molecules containing the light isotopes of carbon and hydrogen results in reaction products that are depleted in the heavy isotopes relative to the substrates. Methanogenesis and methane oxidation under far-from-equilibrium conditions should, therefore, show opposing isotopic effects on methane, respectively depleting and enriching the methane in the heavy isotopes (relative to CO 2 and H 2 O). Microbiologically produced methane typically displays strongly negative δ 13 C and δD values ( 14 ). Moreover, experiments with psychrophilic and mesophilic AOM enrichments incubated at seawater sulfate levels (28 mM) yielded methane that was enriched in 13 C and D (relative to CO 2 and H 2 O, respectively), with carbon isotope fractionation ( 13 α) values between 1.012 and 1.039 and hydrogen isotope fractionation ( 2 α) values between 1.109 and 1.315 ( 15 ). Evaluations of sediment porewater profiles of methane concentration and isotopic composition yielded similar estimates for 13 α (values between 1.009 and 1.024) during AOM in natural habitats ( 16 – 18 ). A single study from an anoxic water column showed 2 α values consistent with the AOM enrichment experiments (1.120 and 1.157) ( 19 ). Within sediments, AOM concentrates in sulfate methane interfaces (SMIs), where sulfate diffusing from above meets methane rising from below. In these SMIs, methane isotopic compositions often do not match experimental predictions of preferential oxidation of the isotopically light methane isotopomers. Instead, when methane diffusing from below enters the SMIs, its δ 13 C values relative to that in the deeper methanogenic zone become frequently lighter by up to 30‰ ( 20 – 22 ). This unexpected observation has been interpreted to reflect either concurrent activity of ANME species and methanogens or that ANME-1 may be facultative methanogens ( 23 – 25 ). However, Yoshinaga and co-workers ( 22 ) showed that at low–sulfate concentration AOM cultures can form methane slightly depleted in 13 C. Although concurrent AOM and methanogenesis by the community members in those experiments was not ruled out, the depletion has been alternatively explained by isotopic equilibration between the methane substrate and the dissolved inorganic carbon (DIC) produced through AOM. The authors hypothesized that the gross backward flux from DIC to CH 4 ( J C − ) should be 50 to 75% of the gross forward flux ( J C + ) to generate the observed decrease in δ 13 C of methane in SMIs ( 22 ). This can be recast in terms of a useful concept, the reversibility of the carbon network in AOM, which is defined as the ratio of backward to forward fluxes of AOM ( J C − / J C + ). The value of J C − / J C + can vary from 0 for an irreversible transformation (i.e., unidirectional oxidation of methane) to 1 for equilibrium between reactant and product (i.e., complete reversibility without net methane oxidation). Within this framework, the results of Yoshinaga and co-workers ( 22 ) are explained by J C − / J C + of 0.50 to 0.75. To reach more pronounced declines of methane δ 13 C values in SMIs, J C − / J C + would need to increase to even higher values of 0.9 to 0.99 ( 26 ). However, the AOM reaction is only 2 to 5% reversible at seawater sulfate concentrations (i.e., J C − / J C + is only 0.02 to 0.05) ( 27 , 28 ). In addition, radiotracer measurements of the rate of DIC reduction to methane and methane oxidation to DIC rarely yield J C − / J C + values greater than 0.2 in marine sediments and microbial mats in which AOM occurs ( 29 – 32 ). Last, the Gibbs free energy (Δ G r ) of AOM in most environments is in the range of −20 to −40 kJ mol −1 ( 33 ), more negative than the estimates of the minimal threshold for AOM activity at ~−10 kJ mol −1 ( 34 ), casting further doubt that CH 4 and DIC isotopes nearly equilibrate during AOM in SMIs. Here, using a thermophilic AOM culture devoid of methanogens, we studied the isotope effects of AOM at different concentrations of its terminal electron acceptor, sulfate. We tracked the isotopic composition and the magnitude of forward and backward fluxes in AOM at an unprecedented level of detail. On this basis, we constructed a multistep biochemical-isotopic model, which resolved shifts in the reversibilities of intracellular reactions as the major cause for the highly variable methane isotope patterns during high-sulfate and low-sulfate AOM.",
"discussion": "DISCUSSION The formation of 13 C-depleted methane under net AOM during incubations has been observed before. Incubations of Black Sea AOM-active mats dominated by ANME-1 showed a methane Δδ 13 C of −15‰ at ambient sulfate concentrations of 16 mM, which was explained by a co-occurrence of methanogenesis and AOM ( 38 , 39 ). Later incubations of cold-adapted AOM cultures dominated by either ANME-1 or ANME-2 resulted in a methane Δδ 13 C of −5‰ at sulfate concentrations below 0.5 mM, which was explained by carbon isotopic equilibration between the reactant CH 4 and the AOM product DIC ( 22 ). In our experiments, a co-occurrence of AOM and energy-conserving methanogenesis can be excluded, as the AOM50 culture is devoid of gene sequences of known methanogens and methane production could not be stimulated with typical substrates (i.e., dihydrogen, methanol, methylamine, and acetate) ( 28 ). Hence, all methane isotope effects observed in this culture result from the activity of ANME-1 archaea in association with their partner bacteria. Below, we explore explanations for our experimental results that involve the effects of the evolving ambient conditions on the thermodynamics and kinetics of reactions in the AOM pathway, with implications for other incubation studies and environmental processes. An influence of sulfate concentration on the reversibility of AOM We applied the measured concentrations, rates, and J C − / J C + values of our experiments to the AOM two-box carbon isotope model of Yoshinaga and co-workers ( 22 ). For the measured J C − / J C + range from 0.04 to 0.15, methane δ 13 C and δD values increased under both high-sulfate and low-sulfate conditions because of dominant expression of normal kinetic carbon and hydrogen isotope effects (fig. S2, A and B). This calculated result is inconsistent with the substantial 13 C depletion of the residual CH 4 that we observed at the highest reversibilities reached in our experiments, at low sulfate concentration. For such 13 C depletion in methane to be reached by the two-box model, J C − / J C + would need to be assigned a value of 0.75 ( 22 ). Consequently, it appears that near-equilibrium carbon isotope exchange between the pathway end-members (CH 4 and CO 2 ) as an explanation for the decrease in the methane δ 13 C values during sulfate-limited AOM must be abandoned, especially because this explanation ignores the multistep nature of AOM, in which successive enzymatic steps may exhibit their own reversibilities and isotope effects. The J H + / J C + ratios derived from our radiotracer measurements indicate that carbon-equivalent gross rates of hydrogen transfer from CH 4 to H 2 O were, in all cases, higher than the gross rates of carbon transfer from CH 4 to CO 2 (fig. S1). Moreover, the observed increase in J H + / J C + from ~2 to ~8 with decreasing sulfate concentration indicates even faster relative rates of gross hydrogen transfer, relative to gross carbon transfer at low sulfate concentrations. We reiterate that actual J H + / J C + ratios in our experiments were likely higher than the reported values, as the transfer of protium and deuterium from methane to water is likely faster than the transfer of tritium. Marlow and co-workers ( 40 ) used a similar approach to constrain AOM rates, by measuring D exchange between mono-deuterated methane (CH 3 D) and H 2 O. They similarly found that the CH 3 D-based AOM rates were higher than the 14 C-CH 4 rates by up to a factor of 2, whereas aerobic methanotrophic bacteria did not show such an effect. We suggest that the limited sulfate availability lowers the capacity of the partner bacteria to perform sulfate reduction and to accept reducing equivalents released during AOM. The resulting shortage of electron sinks affects both the thermodynamic drive and the reversibility of AOM, and the kinetics of reactions that depend on the concentrations of the intracellular metabolites whose cycling depends on the existence of electron sinks. While the net rate of AOM is strongly affected by low sulfate, the stability of the gross forward hydrogen transfer rate ( J H + ) indicates that parts of the pathway are less affected by the shortage of electron sinks. We do not have direct estimates of the reversibility of CH 4 -H 2 O hydrogen isotope exchange, as J H – was not measured. However, since the net carbon and hydrogen reaction rates are linked by the stoichiometry of AOM, stability of J H + coupled to a decrease in the gross forward carbon transfer rate ( J C + ) suggests increasing reversibility of CH 4 -H 2 O hydrogen isotope exchange with decreasing substrate concentrations. As our estimates of J H + are lower limits, the increase in the reversibility of CH 4 -H 2 O hydrogen isotope exchange is likely even more pronounced. We hypothesize that this differential reversibility of individual reactions along the metabolic pathway is the driver of the opposing carbon and hydrogen isotope effects that we observed under the different experimental conditions. Although the overall J H + values determined in the radiotracer experiments do not reveal the specific reactions that sustain hydrogen isotope exchange, the methyl-coenzyme M reductase (Mcr)- and heterodisulfide reductase (Hdr)–catalyzed reactions are natural candidates to explain this observation. As long as methane is nonlimiting, Mcr may still activate methane molecules to methyl-coenzyme M (CH 3 -SCoM) and coenzyme B (HS-CoB), even if sulfate is highly limiting. The deuterium from HS-CoB can then be exchanged with H 2 O during HS-CoB cycling by Hdr. In ANME-1, both the Mcr- and Hdr-catalyzed reactions are not coupled to generation of an electrochemical potential across the membrane. Thus, Mcr and Hdr should, in principle, continue to catalyze CH 4 -H 2 O hydrogen isotope exchange, although perhaps at a slower rate. The rates of AOM and methanogenesis are often measured by 14 C tracer experiments similar to those described here [e.g., ( 29 , 41 )]. Because of the difference between the gross forward rate of AOM ( J C + ), as measured in radiotracer experiments, and the net AOM rate, J C net ≡ J C + − J C − , the use of J C + as a proxy for J C net results in an overestimation of J C net , the magnitude of which depends on the magnitude of J C − . A 4 to 15% difference between gross and net fluxes, as observed in our experiments, may be acceptable in some conditions, as long as it is accounted for. We note that the difference may be larger under closer-to-equilibrium conditions (e.g., lower methane or sulfate concentrations). The relatively high reversibility of the hydrogen isotope exchange reactions in the pathway, discussed above, leads to marked overestimation of net AOM rates by a factor of at least 2 to 4 when using 2 H(D)- or 3 H(T)-labeled methane to constrain J H + , even under the relatively far-from-equilibrium conditions in our experiments. Hence, these methods are obviously not ideal to track net AOM rates. The methane δ 13 C values appear to level out in both the high-sulfate and the low-sulfate experiments ( Fig. 1, B and F ). In neither case do the methane δ 13 C values approach the temperature-dependent CH 4 -CO 2 carbon isotope equilibrium, in agreement with the low J C − / J C + values derived from our 14 C measurements (maximal J C − / J C + of 0.15 at 50 μM sulfate). Thus, equilibrium carbon isotope exchange is not the cause of the apparent cessation of change in methane δ 13 C values. Instead, near-zero net isotope fractionation may arise because of a fortuitous combination of equilibrium and kinetic carbon isotope effects with opposing values that cancel each other out. In this case, methane δ 13 C values will remain invariant upon progressive Rayleigh distillation of the methane pool. Alternatively, invariant methane δ 13 C values may arise from a decrease in the gross rates of AOM, which are accompanied by an increase in the turnover time of methane. In this case, the change in methane δ 13 C values may become immeasurably slow over the time frame of the experiment, irrespective of the net carbon isotope fractionation. A multistep isotope model for AOM simulates observed isotope effects To understand the observed isotopic evolution at a more mechanistic level and to test the hypotheses provided above for the differential sulfate dependence of carbon and hydrogen isotope fractionation during AOM, we developed a metabolic-isotopic model. The model accounts for the effect of step-specific reversibility in the AOM pathway on the net fractionation of carbon and hydrogen isotopes (for model architecture and parameterization, see Materials and Methods and text S1). As in previous studies [e.g., ( 22 )], the model is bound by the concentrations and starting isotopic compositions of the end-members methane, DIC, and water (tables S1 and S2). Unlike previous work, our model includes a subset of the intracellular intermediates in AOM, specifically, CH 3 -SCoM, HS-CoB, and formyl methanofuran (CHO-MFR) and the reactions between these metabolites ( Fig. 4 and Table 1 ). We consider these pools as critical because reactions upstream and downstream of these compounds have the most positive and negative standard Gibbs free energy ( Δ G r ′ 0 ) values in the AOM pathway ( 42 ). Upstream of CH 3 -SCoM, the positive Δ G r ′ 0 is likely to lead to high reversibility of the Mcr-catalyzed activation of methane (reaction 1 in Fig. 4 ). Downstream of CHO-MFR, the large negative Δ G r ′ 0 is expected to allow low reversibility of the CHO-MFR dehydrogenase (Fmd)–catalyzed reaction (reaction 3 in Fig. 4 ). Between CH 3 -SCoM and CHO-MFR is a chain of reactions catalyzed by five enzymes, which are not explicitly included in the model. Instead, these reactions are merged and referred to as composite reaction 2 ( Fig. 4 ). Fig. 4 AOM model architecture. Conceptual model of the ANME partner in the ANME-sulfate reducer consortium in our culture experiments. Each ellipse represents a metabolite pool. In the model, the metabolite pools are related by mass and isotope fluxes of carbon and hydrogen. The carbon and hydrogen atoms that we track in the isotope mass balance are depicted in bold. We assume that the diffusion rate of CO 2 , CH 4 , and H 2 O in and out of the cell is nonlimiting for both the isotope and the mass fluxes. The names of the enzymes that catalyze the reactions included in the model are in the ellipses, numbered as referred to in the text (1 to 4). Note that composite reaction 2 is actually a chain of five reactions in the AOM pathway (see text). The hydrogen in the substrates of reactions 1 and 2 ends up in two distinct products (i.e., CH 4 is oxidized to CH 3 -SCoM and HS-CoB, and CH 3 -SCoM is further oxidized to CHO-MFR and F 420 H 2 ). Table 1 Reactions that are included in the biochemical-isotopic model. The reactions are numbered with respect to Table 1 . CoM-SS-CoB, coenzyme M–HTP heterodisulfide; CH 3 -SCoM, methyl-coenzyme M; HS-CoB, methyl-coenzyme B; HS-CoM, coenzyme M; F 420 , oxidized coenzyme F 420 ; F 420 H 2 , reduced coenzyme F 420 ; Fd ox , oxidized ferredoxin; Fd red 2− , reduced ferredoxin. 1. CH 4 + CoM-SS-CoB ⇄ CH 3 -SCoM + HS-CoB 2. CH 3 -SCoM +2F 420 + MFR + H 2 O ⇄ CHO-MFR + 2F 420 H 2 + HS-CoM + H + 3. CHO-MFR + Fd ox + H 2 O ⇄ CO 2 + MFR + Fd red 2− + 2H + 4. HS-CoB + HS-CoM + Fd red 2− + 2F 420 + 2H + ⇄ CoM-SS-CoB + Fd ox + 2F 420 H 2 Model AOM gross and net rates, and the net reversibility of AOM, are constrained by the experimental radiotracer results and imposed rather than calculated, thereby removing a layer of complexity and uncertainty from the model. We do not use the tritium-based forward rates of CH 4 -H 2 O hydrogen isotope exchange to constrain the model but note that the model independently reproduces the high reversibility of hydrogen isotope exchange inferred above from the J C + and J H + measurements. With the experimental methane- and sulfate-dependent rates of AOM (fig. S3), the available energy is randomly split among the four individual reactions in the AOM pathway ( Δ G r i ) and values are randomly assigned to the six unknown kinetic fractionation factors (KFFs). Conceptually, the allocation of Δ G r i values reflects how the ANME distribute the available thermodynamic drive along the metabolic pathway to maximize growth yield and energy conservation, the latter of which may be of greater importance in our experimental conditions, where growth was minimal. For any (random) allocation of Δ G r i values, we use the mismatch between the model results and the isotopic measurements to constrain the parameter values ( Fig. 5 and figs. S4 and S5) and the resulting isotopic evolution of CH 4 and DIC ( Fig. 6 ) over the course of an experiment (details in Materials and Methods and text S1). The ultimate model output is a set of best-fit values for the reversibilities (from the Δ G r i values) of the four reactions and the six unknown KFFs in the AOM pathway, with an estimate of uncertainty on those values. We note that the modeled isotopic evolution in the different experiments is insensitive to four of the six unknown KFFs (fig. S6). Hence, the model results and inferences drawn from them are robust to the values of these parameters, and in essence, the model may be considered to have six free parameters (four reversibilities and two KFFs). Fig. 5 Individual reaction reversibility ( J − /J + ) best fits to the experimental results. The envelopes contain 68% of results based on 10 5 simulations with parameter values drawn from uniform prior distributions and weighted by the inverse of the sum of squared model-measurement mismatch (1/SSE). The schematic chemical reactions in the panels show only the metabolites that participate in isotope exchange, and the full reactions are listed in Fig. 1 . Fig. 6 Modeling of carbon and hydrogen isotopic compositions for high-sulfate and low-sulfate AOM. Temporal evolution of ( A ) methane δ 13 C values, ( B ) methane δD values, and ( C ) DIC δ 13 C values. ( D ) Comparison of the evolution of methane δD and δ 13 C values. Black and red lines are the model results for low-sulfate and high-sulfate conditions, respectively. The circles show the experimental data from Fig. 1 . The error bars are the SDs (1σ) and are smaller than the symbols where not visible. The gray and pink envelopes contain 68% of 10 5 model simulations. High-sulfate experiment During the high-sulfate experiment, the total Gibbs free energy available to the AOM reaction ( Δ G r net ) increased from −34 to −19 kJ mol −1 , reflecting a weakening of the thermodynamic drive with methane consumption. Over the same time interval, radiocarbon-constrained values of J C − / J C + evolved from 0.04 to 0.15, corresponding to a net thermodynamic drive of the carbon pathway in AOM (Δ G C ) in the range −9 to −5 kJ mol −1 . The best-fit distribution of this Δ G C range among the specific pathway steps suggests that the reactions at the beginning and end of the pathway, catalyzed by Mcr and Fmd, respectively (reactions 1 and 3 in Fig. 4 ), were close to equilibrium ( J 1 – / J 1 + of 0.73 to 0.87 and J 3 – / J 3 + of 0.63 to 0.83; Fig. 5, A and C ). In contrast, the composite reaction 2 was far from equilibrium ( J 2 – / J 2 + of 0.07 to 0.33; Fig. 5B ). The hydrogen isotope exchange between HS-CoB and H 2 O is catalyzed by the Hdr/coenzyme F 420 hydrogenase complex (reaction 4 in Fig. 4 ). This reaction recycles the reduced electron carriers HS-CoB and ferredoxin (Fd), and at the high sulfate concentration, it is out of equilibrium ( J 4 – / J 4 + of 0.01 to 0.32; Fig. 5D ). Thus, the increasing methane δ 13 C and δD values in the high-sulfate experiments ( Fig. 6 ), which indicate the preferential reaction of isotopically light carbon and hydrogen during AOM to form DIC, reflect a combination of a small equilibrium isotope effect (EIE) between CH 4 and CH 3 -SCoM (Mcr-catalyzed reaction) and the partial expression of kinetic isotope effects (KIEs) of composite reaction 2 and the Hdr-catalyzed reactions. The downstream carbon EIE between CHO-MFR and DIC is masked by the irreversibility of composite reaction 2. The net apparent 13 α value evolves over the course of the high-sulfate simulation from 0.984 to 1.042, and the 2 α value evolves from 0.906 to 1.303. Our suggested mechanism is in agreement with a recent suggestion that the enrichment of the residual methane in the heavy isotopes and the increase in the relative abundance of the “clumped” isotopologue 13 CH 3 D is due to a combination of expression of KIEs and EIEs with a reversibility of the Mcr-catalyzed reaction ( J 1 – / J 1 + ) up to 0.6 ( 43 ). The slightly lower reversibility predicted in that study relative to the range of reversibility that we predict (0.6 versus 0.73 to 0.87) may arise from the different experimental conditions. Experiments in both studies had excess sulfate (>10 mM), but the initial concentration of methane was much higher in the experiments described by Ono and co-workers ( 43 ) than in our experiments (20 mM versus 1.45 mM). Overall, this resulted in continuously higher Δ G r net and J C net , which may have driven the Mcr-catalyzed reaction farther from equilibrium (i.e., toward lower reversibility) ( 42 , 43 ) Although sulfate was nonlimiting throughout the experiment (>9 mM), we observed a halt in the increase of methane δ 13 C values when methane reached ~0.01 mM after ~30 days. The model captures this trend and identifies a small net fractionation of carbon isotopes as the cause. In the beginning of the experiment, composite reaction 2 is out of equilibrium, and the net isotopic fractionation expressed during this reaction is close to its KIE (with a best-fit carbon KIE value of 20‰). As sulfate is used over the course of the experiment, the reversibility of this reaction increases up to 0.33, resulting in combined expression of the KIE and EIE (the EIE is −47‰ at 50°C) ( 44 ). After 30 days, the shift from the kinetic end-member toward greater expression of the equilibrium fractionation end-member yields a net carbon isotope fractionation of zero, which translates to a net zero rate of change in methane δ 13 C values ( Fig. 6A ). In the high-sulfate experiment, methane concentrations were too low for δD determinations after 20 days, but the model predicts that methane δD values should continue to increase during the experiment and only plateau after about 40 days ( Fig. 6B ). With further decline of J C net and the concurrent increase in J C − / J C + , we predict that both methane δ 13 C and δD values will ultimately decrease toward CH 4 -CO 2 and CH 4 -H 2 O equilibrium values, respectively. Low-sulfate experiment In the low-sulfate experiment, Δ G r net started at −28 kJ mol −1 and reached −12 kJ mol −1 at the end of the experiment, and we calculated Δ G C ranging from −8 to −2 kJ mol −1 , only slightly more positive than in the high-sulfate experiment. The best fits to the measured isotopic compositions were obtained when the Mcr-catalyzed reaction (reaction 1 in Fig. 4 ) is near equilibrium ( J 1 – / J 1 + of 0.86 to 0.94; Fig. 5A ); the composite reaction 2 departs somewhat from equilibrium ( J 2 – / J 2 + of 0.73 to 0.88), and the Fmd- and Hdr-catalyzed reactions (reactions 3 and 4, respectively, in Fig. 4 ) are far from equilibrium at the beginning of the experiment and closer to equilibrium at the end ( J 3 – / J 3 + increasing from 0.07 to 0.37 and J 4 – / J 4 + increasing from 0.02 to 0.43 during the experiment; Fig. 5, C and D ). The best-fit J i – / J i + values show that all reactions become more reversible (i.e., closer to equilibrium) relative to the high-sulfate conditions, as expected at the lower sulfate concentrations, with the exception of the Fmd-catalyzed reaction (reaction 3), which we discuss below. The model reveals that the large negative Δδ 13 C observed in the experiments is not due to near-equilibrium carbon isotope fractionation between DIC and methane during AOM, which should not be expected with the reversibility observed in the low-sulfate experiment ( J C − / J C + = 0.15). The CHO-MFR/CH 3 -SCoM equilibrium carbon isotope fractionation (composite reaction 2 in Fig. 4 ) is large and negative (−47‰ at 50°C; ( 44 )), and the net fractionation associated with this reaction is close to this value because of the reaction’s high reversibility ( J 2 – / J 2 + of 0.73 to 0.88; Fig. 5B ). Together with the negligible EIE of the Mcr-catalyzed reaction (−0.8‰ at 50°C) ( 44 ) and the KIE of Fmd (with a best-fit value of ~5‰), net AOM generates methane Δδ 13 C of up to −26‰, even when J C − / J C + is as low as 0.04 to 0.15. The resulting net apparent 13 α value evolves over the course of the low-sulfate simulation from 0.976 to 0.952. As in the high-sulfate experiment, methane δ 13 C stabilizes after ~30 days when 30% of the initial methane still remained (~0.4 mM methane) because of near-complete depletion of the sulfate. At this point, the slow gross rates of AOM, coupled with the relatively large residual pool of methane yield long methane residence times of 70 days after 30 days of experiment and ~400 days at the end of the experiment. The resulting small relative rate of carbon isotope exchange manifests as near-invariant methane δ 13 C values at the time scale of the experiment. Over the same simulated time period, the apparent 2 α value evolves from 0.905 to 0.950. The hydrogen isotope exchange network with H 2 O in AOM is nonlinear and is mediated by the electron carriers HS-CoB and F 420 . In our model, we simplify this system by assuming that F 420 cycling reactions are at chemical and isotopic equilibrium. In the first days of the low-sulfate experiment, the low J − / J + values of the Fmd-catalyzed reaction (reaction 3 in Fig. 4 ) result in the expression of hydrogen KIEs and a slight overall increase in methane δD values (ΔδD > 0), in contrast to methane δ 13 C values, which decrease throughout the experiment (i.e., display Δδ 13 C < 0). As the J − / J + values of the reactions downstream of CH 3 -SCoM (reactions 2 and 3 in Fig. 4 ) increase as substrates are consumed and Δ G C becomes less negative, the EIEs of these reactions are increasingly expressed, pulling methane δD to more negative values and resulting in ΔδD < 0. However, methane δD values by the end of the experiment (−115‰) are almost 95‰ heavier than expected at hydrogen isotope equilibrium between CH 4 and H 2 O at the experimental temperature (for the measured H 2 O δD of −54‰, methane δD at hydrogen isotope equilibrium would be −208‰), suggesting that CH 4 remains far from equilibrium with the H 2 O, even when J C net is almost zero. The change in the modeled reversibility of the Fmd-catalyzed reaction between the high-sulfate conditions ( J 3 – / J 3 + > 0.62) and low-sulfate conditions ( J 3 – / J 3 + of 0.07 in the beginning, increasing to 0.37) appears counterintuitive. A decrease in sulfate concentrations, through interspecies electron transfer, is expected to decrease the oxidation state of the AOM partner. This may lead to an increase in the ratio of reduced to oxidized electron carriers, which, in turn, leads to a lower thermodynamic drive of the individual reactions in AOM (i.e., an increase in reversibility). However, the adjustment of intracellular metabolite concentrations, other than those of electron carriers, to the lower oxidation state means that while most reactions become less favorable (i.e., more reversible), some may become more favorable (i.e., less reversible). A full metabolic model, which is beyond the scope of the current study, is required to resolve such cases. With the simplified model developed here, we suggest that such a decrease in the reversibility of the Fmd-catalyzed reaction is required to reach agreement between the model and the experimental observations. Such a trend is conserved even when we relax the constraint of consistency between the Gibbs free energy allocated to the different reactions and the J C − / J C + determined by radiocarbon measurements (fig. S7). High-sulfate and low-sulfate AOM experiments reproduce conditions in SMIs Our stable isotope experimental setups were designed to determine methane carbon and hydrogen isotope effects of AOM under high-sulfate and low-sulfate conditions on laboratory time scales. Nonetheless, these experiments show notable similarity to the conditions that prevail in the environment in terms of concentrations of sulfate, methane, DIC, and sulfide. The main difference is the thermophilic cultivation and the higher methane oxidation rate in the culture [150 μmol (L medium) −1 day −1 ] than those in SMIs [<5 μmol (L sediment) −1 day −1 ]. The higher rates allowed us to perform experiments within a reasonable time but kept energetic conditions that mimic environmental settings. This is in contrast to many similar methanogen culture experiments, which are usually grown at H 2 concentrations that are orders of magnitudes higher than those encountered in natural environments and which often are not able to reproduce natural isotope effects [e.g., ( 45 , 46 )]. The two contrasting scenarios investigated here reflect conditions in SMIs ( 1 ), which often appear only centimeters apart from each other. The upper section of SMIs bears relatively high sulfate concentrations (>1 mM) and is similar to our high-sulfate experiments. Under these conditions, KIEs prevail as displayed by increasing methane δ 13 C and δD values upon progressive methane consumption ( Figs. 1B and 6B ). Fractionation factors retrieved from our experiments ( 13 α of 1.016 to 1.017 and 2 α of 1.151 to 1.159) closely match those inferred from natural studies, where values of 13 α of 1.002 to 1.024 and 2 α of 1.120 to 1.157 have been estimated ( 16 – 19 ). In the lower section of SMIs, AOM proceeds under very low sulfate concentrations (<<1 mM) and an excess of methane. Here, methane δ 13 C values often drop instead of becoming increasingly positive ( 22 , 23 , 47 ). It has been suggested that this decline is caused by high rates of hydrogenotrophic methanogenesis (CO 2 reduction with H 2 ) that co-occur with active AOM ( 23 , 47 , 48 ). Since ANME-1 archaea are often the only Mcr-containing microorganisms to be found in this part of the SMIs, some have postulated that ANME-1 are facultative methanogens ( 23 , 24 , 49 ). The characterization of ANME-1 as obligate methane oxidizers with no indications of methanogenesis ( 28 ) weakens this suggestion. Moreover, for methane δ 13 C values to decrease, the apparent rates of CO 2 reduction would need to almost equal AOM rates ( 22 , 38 , 48 ). Radiotracer-based concurrent measurements of methanogenesis and AOM rarely find similar rates of methane production and consumption ( 24 ), and in most of these studies, ratios of methane production to consumption in SMIs are lower than 0.2 ( 29 – 32 ). Such low relative ratios of true methanogenesis are unlikely to be responsible for the low methane δ 13 C values observed globally in SMIs. Our observations of intrinsic formation of 13 C-depleted methane during AOM under substrate limitation resolve this apparent disagreement by obviating the need to invoke methanogenesis by the ANME or other archaea in SMIs. These results notwithstanding, we note that δ 13 C values alone are probably insufficient to distinguish between methanogenesis and AOM, in line with previous suggestions about δ 13 C values of biomass in sediments from methane seeps and vent sites ( 50 ). As an alternative to concurrent methanogenesis and AOM, Yoshinaga and colleagues ( 22 ) suggested that the methane δ 13 C decrease may result from carbon isotopic equilibration between methane and DIC, catalyzed by the enzymatic reaction chain of ANME archaea. However, to reach the steep drop in methane δ 13 C values observed in deeper zones of SMIs, the reversibility of the net AOM reaction would need to be higher than 0.90 ( 26 , 48 ). Such high reversibilities contradict the apparent energy yields of AOM under natural conditions of about −20 to −40 kJ mol −1 . Instead, our study provides evidence that under conditions of low sulfate availability, intracellular reaction reversibilities along the enzymatic AOM pathway give rise to net fractionations that drive the formation of 13 C- and D-depleted methane even with energy yields of −20 kJ mol −1 or more. Strongly negative methane δ 13 C excursions are specifically found in deep SMIs at sulfate concentrations below approximately 1 mM ( 21 , 22 , 26 , 51 ). According to our results, these excursions do not require carbon isotope equilibration between methane and DIC, as proposed by Yoshinaga and co-workers ( 22 ). Instead, the δ 13 C excursions may be attributed to a combination of EIEs and KIEs associated with relatively high upstream reversibilities (Mcr-catalyzed reaction and the following composite reaction 2) and low downstream reversibility of the final reaction in the AOM pathway, catalyzed by Fmd. In other words, at low sulfate concentrations and on the basis of the experimental results, the multistep model predicts a combination of high reversibility between the CH 4 substrate and the CHO-MFR intracellular metabolite and an almost irreversible reaction from CHO-MFR to form DIC ( Fig. 5 ). This downstream irreversibility decouples the isotopic composition of methane from variations in the isotopic composition of DIC, which are only partially communicated upstream in the AOM pathway to the methane. Other SMIs, particularly shallow, advection-influenced sites, do not show negative isotope excursions in δ 13 C values ( 52 , 53 ). In these sites, high fluxes of methane into zones of high sulfate availability may prohibit the expression of the above-described effects of low sulfate concentrations on AOM isotope fractionation. The development of methane δD values within SMIs has been rarely studied ( 18 , 53 ), but the few available observations are well reproduced by our model. Biogenic methane formed below the SMI is expected to have δD values that reflect CH 4 -H 2 O near equilibrium. Our experimental and modeling results suggest that in the lower SMI, methane δD values will remain rather constant or will increase slightly, as was observed in our low-sulfate experiment. With increasing sulfate availability in the upper SMI, methane δD values are expected to increase, as the net fractionation is dominated by the normal KIE of composite reaction 2 (i.e., faster reaction of CH 4 with the light hydrogen isotope). The different behavior of methane carbon and hydrogen isotope patterns derives from the effects of a departure from reversibility in a linear versus a branched reaction network. Specifically, high reversibility of the most upstream, Mcr- and Hdr-catalyzed steps of AOM is, in principle, sufficient to result in hydrogen isotope near equilibrium between CH 4 and H 2 O but insufficient to result in carbon isotope near equilibrium between CH 4 and DIC. We further expect a difference in the dependence of methane δD values on those of the water in different parts of the SMI. Under low-sulfate conditions similar to the deeper zone of the SMI, the δD value of the H 2 O affects the final methane δD values (fig. S8), as the high reversibility of the Mcr-catalyzed reaction results in partial methane-water hydrogen isotopic equilibration. In high-sulfate conditions that resemble the upper part of the SMI, the δD value of the H 2 O has only a small effect on methane δD values (fig. S8), as the lower reversibility of hydrogen isotope exchange reactions results in pronounced expression of KIEs. Under these conditions, Rayleigh distillation of hydrogen isotopes is the main determinant of the methane δD values upon progressive consumption. We predict that the Mcr-catalyzed reaction will remain close to reversibility for both low-sulfate and high-sulfate conditions, which can explain the trend of equilibration of the clumped isotopologs 12 CH 2 D 2 and 13 CH 3 D with time in the environment ( 54 , 55 ). Ash and co-workers ( 54 ), specifically, observed that Δ 12 CH 2 D 2 and Δ 13 CH 3 D values increase toward the expected temperature-dependent equilibrium value in the deep zone of SMI in brackish sediments from the Baltic Sea. These findings support a pronounced reversibility of parts of the AOM pathway, and our findings identify reversibility in the Mcr-catalyzed reaction as a likely candidate to explain the apparent hydrogen and clumped carbon isotope equilibration in the lower parts of SMIs. Our combination of AOM culture experiments and biochemical-isotopic modeling suggest that complete CH 4 -DIC equilibration is not required for depletion in 13 C and D of methane in SMIs. Theoretically, one cannot rule out the possibility that J C − / J C + can increase to values higher than 0.15 in natural environments where sulfate concentrations are even lower than in our tracer experiments ( 56 , 57 ), and well-adapted wild organisms may still respire, thus potentially driving J C − / J C + closer to unity (i.e., full reversibility). However, both the radiotracer rate measurements of gross methane oxidation and DIC reduction and the relatively high Δ G r net of AOM in most marine environments (−20 to −40 kJ mol −1 ) support the hypothesis that AOM does not operate close to thermodynamic equilibrium. In the absence of near-equilibrium conditions in the entire AOM pathway, our results provide indications that the expected reversibility landscape of individual pathway steps can explain natural and experimental observations and, specifically, 13 C- and D-depleted methane in SMIs at low sulfate concentrations. More broadly, our results highlight the strength of combining carefully tailored culture experiments and metabolic-isotopic models of appropriate complexity to understand the controls on the isotopic compositions of natural biogenic materials."
} | 10,166 |
28242282 | PMC5441187 | pmc | 1,952 | {
"abstract": "Enrichment of reef environments with dissolved inorganic nutrients is considered a major threat to the survival of corals living in symbiosis with dinoflagellates ( Symbiodinium sp.). We argue, however, that the direct negative effects on the symbiosis are not necessarily caused by the nutrient enrichment itself but by the phosphorus starvation of the algal symbionts that can be caused by skewed nitrogen (N) to phosphorus (P) ratios. We exposed corals to imbalanced N:P ratios in long-term experiments and found that the undersupply of phosphate severely disturbed the symbiosis, indicated by the loss of coral biomass, malfunctioning of algal photosynthesis and bleaching of the corals. In contrast, the corals tolerated an undersupply with nitrogen at high phosphate concentrations without negative effects on symbiont photosynthesis, suggesting a better adaptation to nitrogen limitation. Transmission electron microscopy analysis revealed that the signatures of ultrastructural biomarkers represent versatile tools for the classification of nutrient stress in symbiotic algae. Notably, high N:P ratios in the water were clearly identified by the accumulation of uric acid crystals.",
"introduction": "1 Introduction The success of coral reefs in oligotrophic environments is owed to the symbiotic association of the habitat-forming scleractinian corals with photosymbionts from the genus Symbiodinium (zooxanthellae). These algal symbionts enable the coral host to access the pool of dissolved inorganic nitrogen and phosphorus in the water column in addition to the nutrient uptake by heterotrophic feeding ( Crossland and Barnes, 1977 , D'Elia and Webb, 1977 , Muscatine and D'Elia, 1978 , Grover et al., 2003 , Titlyanov et al., 2006 , Downs et al., 2009 , Godinot et al., 2009 , Pernice et al., 2012 ). Moreover, the zooxanthellae recycle ammonium excreted as metabolic waste product by the host, thereby efficiently retaining nitrogen within the holobiont ( Muscatine and D'Elia, 1978 , Rahav et al., 1989 , Wang and Douglas, 1998 ). The nutrient limitation experienced by the zooxanthellae in hospite in oligotrophic conditions results in a skewed chemical balance of the cellular nitrogen and phosphorus content relative to the available carbon. As a result, photosynthetic carbon fixation can be uncoupled from cellular growth, facilitating the translocation of a large proportion of photosynthates to the coral host ( Muscatine, 1965 , Muscatine et al., 1989 , Falkowski et al., 1984 , Dubinsky and Jokiel, 1994 ). Reefs and the provision of their valuable ecosystem services are globally threatened by climate change and a range of anthropogenic pressures ( Goreau and Hayes, 1994 , Moberg and Folke, 1999 , Sheppard, 2003 , Hoegh-Guldberg et al., 2007 , Hughes et al., 2007 , Baker et al., 2008 , van Hooidonk et al., 2013 , D'Angelo and Wiedenmann, 2014 , Logan et al., 2014 ). In this context, it has become increasingly clear that the nutrient environment plays a defining role in determining coral reef resilience ( D'Angelo and Wiedenmann, 2014 , Fabricius, 2005 , Szmant, 2002 , Brodie et al., 2012 , Furnas et al., 2005 , Brodie, 1995 ). The ratio of dissolved inorganic nitrogen to phosphorus in the marine environment can be interpreted as an indicator of whether photosynthetic primary production is limited by the availability of nitrogen or phosphorus. In coral reef waters, N:P ratios were found in an approximate range from 4.3:1 to 7.2:1 ( Smith et al., 1981 , Crossland et al., 1984 , Furnas et al., 1995 ) which is lower than the canonical Redfield ratio of 16:1, considered optimal to sustain phytoplankton growth ( Redfield, 1958 ). Consequently, many processes in coral reefs tend to be nitrogen limited ( Furnas et al., 2005 ). Natural nutrient levels in coral reef ecosystems are impacted by the rising anthropogenic nutrient input into the oceans, especially into coastal waters, via the atmospheric deposition of combustion products, agricultural activities, erosion and sewage discharge ( Fabricius, 2005 , Brodie et al., 2012 , D'Angelo and Wiedenmann, 2014 ). Since a number of these sources of nutrient enrichment can be influenced at the local scale ( Brodie et al., 2010 , Kroon et al., 2014 , Aswani et al., 2015 ), the management of nutrification is a promising tool for coral reef protection which also holds potential to mitigate some of the negative effects of rising sea water temperatures on these ecosystems ( D'Angelo and Wiedenmann, 2014 ). It has been conceptualised that some direct negative effects of eutrophication on the Symbiodinium stress tolerance may be caused, paradoxically, by an associated deprivation of nutrients vital for the physiological functioning of the coral symbionts ( Wiedenmann et al., 2013 , D'Angelo and Wiedenmann, 2014 ). The resulting nutrient starvation can occur for example when the availability of one type of essential nutrient (e.g. phosphate) decreases relative to the cellular demand, resulting in imbalanced and unacclimated growth ( Parkhill et al., 2001 ). High nitrate concentrations in combination with low phosphate availability have previously been shown to result in phosphate starvation of the algal symbiont and increased susceptibility of corals to heat- and light-stress-induced bleaching ( Wiedenmann et al., 2013 ). In principle, this condition could not only result from an increased cellular demand due to nutrient (nitrogen) – accelerated cell proliferation rates but also from a selective decrease of one specific nutrient type ( Parkhill et al., 2001 ). Relevant shifts of the nutrient balance in the natural reef environments were reported, for example, for the reefs of Discovery Bay in Jamaica where enrichment with groundwater-borne nitrate resulted in a dissolved inorganic nitrogen to phosphorus ratio of 72:1, coral decline and phase shifts to macroalgal dominance ( Lapointe, 1997 ). However, the functioning of the coral- Symbiodinium association can be severely impaired not only by the imbalanced availability of nutrients, but also by a combined deprivation of both, nitrogen and phosphorus ( Rosset et al., 2015 ). In this light, the expected nutrient impoverishment of oceanic waters that could result from global warming or the rapid uptake of dissolved inorganic nutrients by ephemeral phytoplankton blooms could possibly act in combination with increased heat stress levels to accelerate reef decline ( D'Angelo and Wiedenmann, 2014 , Riegl et al., 2015 ). Due to the fast uptake of dissolved inorganic nutrients by benthic communities it is often difficult to measure the level of nutrient exposure in coral reefs ( Furnas et al., 2005 ). Consequently, biomarkers are required that inform about the nature of the nutrient stress which corals and their symbionts experience under certain conditions ( Cooper and Fabricius, 2012 , D'Angelo and Wiedenmann, 2014 ). Recently, we have demonstrated that bleaching and reduced growth of corals resulting from the deprivation of dissolved inorganic nitrogen and phosphorus is reflected by the ultrastructure of zooxanthellae ( Rosset et al., 2015 ). The undersupply with nutrients manifests in a larger symbiont cell size, increased accumulation of lipid bodies, higher numbers of starch granules and a striking fragmentation of their accumulation bodies. We have exploited the potential of these biomarkers to detect nutrient stress imposed on the coral- Symbiodinium association and explored the response of the algal ultrastructure to skewed dissolved inorganic nitrogen to phosphorus ratios.",
"discussion": "4 Discussion We used ultrastructural biomarkers of zooxanthellae to gain novel insights into the response of the coral – Symbiodinium symbiosis to imbalanced nutrient environments and to analyse the role of nitrogen and phosphorus for the functioning of this association and potential implications for coral reef management. We used the reef coral Euphyllia paradivisa harbouring Symbiodinium sp. (clade C1) as a model system, exposed the corals to HN/LP and to LN/HP conditions and compared them to specimens from nutrient replete (HN/HP) and low nutrient (LN/LP) treatments ( Rosset et al., 2015 ). 4.1 Effect of high nitrate/low phosphate conditions Recently, we demonstrated that corals exposed to HN/LP conditions were more susceptible to bleaching when exposed to heat stress and/or elevated light levels ( Wiedenmann et al., 2013 ). The detrimental effects were linked to the relative undersupply with phosphorus that can result from the higher demand of the proliferating algal populations rather than to the high nitrogen levels. Phosphate starvation in Symbiodinium sp. resulted in a drop of photosynthetic efficiency associated with changes in the ratio of phospho- and sulfo-lipids ( Wiedenmann et al., 2013 ). In other photosynthetic organisms, similar responses to phosphate stress could be attributed to critical changes in the properties of photosynthetic membranes ( Frentzen, 2004 ). Hence, our findings provided a potential mechanistic link between nutrient stress, the malfunctioning of the photosynthetic machinery and the observed bleaching response. With their low zooxanthellae numbers, bleached appearance and small polyp size, the corals from the HN/LP treatment under elevated light levels resembled the low-nutrient phenotype (LN/LP) previously described ( Rosset et al., 2015 ). These two treatments also had similar effects on the ultrastructure of zooxanthellae, with cell size and the accumulation of carbon-rich storage bodies (lipid bodies and starch granules) being increased in comparison to zooxanthellae from nutrient replete conditions. Similar structural changes were found to be indicative of nutrient limitation in zooxanthellae and free-living microalgae ( Hoegh-Guldberg, 1996 , Muller-Parker et al., 1996 , Hu et al., 2008 , Msanne et al., 2012 , Weng et al., 2014 , Rosset et al., 2015 ). Those characteristics have been interpreted as indicators of an uncoupling of carbon fixation from cellular growth. In this state, the nutrient-limited cells sustain a high photosynthetic production while their energy demand is reduced due to slower proliferation rates ( Rosset et al., 2015 , Hu et al., 2008 , Vítová et al., 2015 , Vaulot et al., 1987 ). Since corals from the HN/LP conditions were supplied with excess nitrogen, the nutrient limitation phenotype of corals and symbionts can be clearly attributed to the undersupply with phosphate. Importantly, under both, nutrient replete and low nutrient conditions, the photosynthetic efficiency measured as Fv/Fm was in the healthy range (> 0.5). In contrast, Fv/Fm was strongly reduced in the imbalanced HN/LP treatment, indicative of failing photosynthesis due to phosphate starvation ( Wiedenmann et al., 2013 , D'Angelo and Wiedenmann, 2014 ). At ultrastructural level, the phosphate starvation phenotype resulting from nitrogen enrichment in combination with low phosphate supply can be clearly distinguished from the low-nutrient phenotype by the pronounced accumulation of uric acid crystals. This finding is in line with previous studies that observed comparable deposits in zooxanthellae in response to nitrate enrichment, forming a transitory storage of assimilated nitrogen ( Clode et al., 2009 , Kopp et al., 2013 ). Finally, the phosphate-starved zooxanthellae lack the intriguing fragmentation pattern of the accumulation body, characteristic of strongly nutrient-limited zooxanthellae ( Rosset et al., 2015 ). 4.2 Effect of low nitrate/high phosphate conditions Despite the relative undersupply of nitrogen in the low nitrate/high phosphate treatment, the polyp size and zooxanthellae density of these corals were comparable to those from the replete nutrient treatment. However, the ultrastructural biomarkers revealed signs of nutrient limitation such as elevated levels of lipid bodies and starch granules in symbiont cells from corals under LN/HP. In the light of previous findings, the effects of the low supply with nitrogen could be interpreted to cause an uncoupling of carbon-fixation and cellular growth that manifests in the increased accumulation of carbon-rich storage products. However, as indicated by the smaller cell size and the high number of symbiont cells within the coral tissue, comparable to those from corals experiencing high nutrient levels ( Rosset et al., 2015 , Hu et al., 2008 , Vítová et al., 2015 , Vaulot et al., 1987 ), cell proliferation rates are still high enough to sustain these zooxanthellae densities. These results, together with the high Fv/Fm values of zooxanthellae from LN/HP corals suggest that the N-limitation sustains a slower but chemically balanced growth while maintaining a functional photosynthesis. 4.3 Differential effects of N and P undersupply and critical thresholds Our results suggest that symbiotic corals can tolerate an undersupply with nitrogen much better than an undersupply with phosphorus. These findings likely reflect an adaptation of the algal symbionts to the nutrient environment of coral reefs where processes are mostly nitrogen limited ( Crossland et al., 1984 , Furnas et al., 1995 , Smith et al., 1981 , Furnas et al., 2005 ). In agreement with this assumption, previous studies found a trend that nitrogen enrichment stimulates zooxanthellae growth and results in higher zooxanthellae densities, often without obvious negative effects on the corals ( Fabricius, 2005 ). It cannot be ruled out, however, that nitrogen-fixation by coral-associated microbes in the presence of high phosphate concentration might potentially relieve some of the nitrogen-undersupply of the corals ( Rädecker et al., 2015 ). The present study clearly shows that phosphate deficiency, alone or in combination with a low supply of nitrate, results in a severe disturbance of the symbiotic partnership as indicated by the loss of coral tissue and zooxanthellae. Phosphate starvation of zooxanthellae induced by nitrogen enrichment and resulting high N:P ratios has previously been shown to disturb the photosynthetic capacity of zooxanthellae and increase the vulnerability of corals to light- and heat stress-mediated bleaching ( Wiedenmann et al., 2013 ). The fact that normal photosynthetic efficiency is retained by zooxanthellae in corals from the LN/LP treatment suggests that an undersupply with phosphate has less severe consequences when the algae become limited by nitrogen. This can be explained by the reduced P-demand of the non-/slow-growing algal population ( D'Angelo and Wiedenmann, 2014 ). The concentrations of dissolved inorganic nutrients in our LN/LP treatment (~ 0.7 μM/~0.006 μM) suggest that at measured nitrate concentrations < 0.7 μM the impact of skewed N:P ratio becomes less pronounced. In our experiments, a phosphate concentration of ~ 0.3 μM at a N:P ratio of 22:1 yielded an overall healthy phenotype. Accordingly, it is likely that the absolute N:P ratio becomes also less critical for the proper functioning of the symbionts when phosphate concentrations exceed a vital supply threshold (> 0.3 μM), even when the symbionts are rapidly proliferating. In contrast, a phosphate concentration of ~ 0.18 μM at a ~ 10-fold higher N:P ratio (211:1) yielded a bleached phenotype with the remaining symbionts showing signs of stress (Fv/Fm < 0.4). Therefore, the P-threshold at which corals can become stressed in the presence of high N concentrations can be as high as 0.18 μM. Effects of P deficiency can be expected to become worse if supply from other sources such as particulate food or internal reserves, is low. 4.4 Implications for environmental monitoring and coral reef management Our study suggests that phosphate can become critically limiting even at concentrations ≤ 0.18 μM if the N:P ratios well exceed 22:1. This appears surprising since phosphate concentrations in this range are commonly considered ambient or high in natural reef environments. However, Lapointe (1997) reports phosphate concentrations of 0.1–0.18 μM at N:P ratios in the range of 33–72 to be associated with phosphate limitation of macroalgae in the declining reefs of Discovery Bay (Jamaica). These data suggest that the critical threshold values determined by our laboratory study can indeed be found in reef environments impacted by eutrophication. However, it is important to note that nutrient values measured in the water column of natural or experimental mesocosm settings represent a steady-state equilibrium that depends on their production and uptake by organisms. Since these fluxes vary spatially and temporarily among reef regions, the measured nutrient concentrations have to be considered in the context of the respective environment. Consequently, there is an urgent need to refine these thresholds and quantify the absolute amounts of nutrients and the associated fluxes that are responsible for the observed biological effects. These values are required to provide reliable and effective target values for management purposes. Of particular interest in the context of the present work is also the role of phytoplankton blooms. Stimulated by nutrient enrichment in the first place, coastal blooms can limit primary production by depleting essential nutrients or shifting their ratio over time and space ( D'Angelo and Wiedenmann, 2014 ). Critically, the depletion of dissolved inorganic phosphorus has been reported in the aftermath of phytoplankton blooms that were initially set off by elevated nitrogen levels ( Del Amo et al., 1997 , Fujiki et al., 2004 , Haese et al., 2007 ). Such a lack of phosphate may render benthic corals more susceptible to stress, bleaching and associated mortality ( Wiedenmann et al., 2013 ). Indeed, previous studies have observed a correlation between elevated nitrogen concentrations, increased phytoplankton densities and coral bleaching ( Wagner et al., 2010 , Wooldridge, 2009 , D'Angelo and Wiedenmann, 2014 ). Preventing the enrichment of coral reef waters with excess nitrogen should consequently be a management priority. However, it is important to note that also other forms of nutrient enrichment can have a plethora of direct and indirect negative effect on corals and their symbionts (reviewed by D'Angelo and Wiedenmann, 2014 ). Therefore, the reduction of nutrient enrichment has to be generally high on the agenda of coral reef management ( Riegl et al., 2015 ). The extended set of cumulative, ultrastructural biomarkers provided here ( Table 1 ) can be used to identify different forms of nutrient stress in Euphyllia sp. associated with Symbiodinium (C1). These biomarkers hold promise to indicate nutrient stress also in other symbiotic coral species and in various reef settings. Importantly, they have potential to become part of the toolkit that is required for an in-depth understanding of the nutrient environment in coral reefs by bridging knowledge gaps left by traditional measurements of nutrient levels in the water column. Our findings highlight the key role of phosphorus in sustaining zooxanthellae numbers and coral biomass and for the proper functioning of symbiont photosynthesis, thereby contributing to the critical understanding of the importance of phosphorus for the functioning of symbiotic corals ( Ferrier-Pagès et al., 2016 )."
} | 4,831 |
23610627 | PMC3631397 | pmc | 1,953 | {
"abstract": "Symbiotic dinoflagellates are unicellular photosynthetic algae that live in mutualistic symbioses with many marine organisms. Within the transcriptome of coral endosymbionts Symbiodinium sp. (type C3), we discovered the sequences of two novel and highly polymorphic hemoglobin-like genes and proposed their 3D protein structures. At the protein level, four isoforms shared between 87 and 97% sequence identity for Hb-1 and 78–99% for Hb-2, whereas between Hb-1 and Hb-2 proteins, only 15–21% sequence homology has been preserved. Phylogenetic analyses of the dinoflagellate encoding Hb sequences have revealed a separate evolutionary origin of the discovered globin genes and indicated the possibility of horizontal gene transfer. Transcriptional regulation of the Hb -like genes was studied in the reef-building coral Acropora aspera exposed to elevated temperatures (6–7°C above average sea temperature) over a 24-h period and a 72-h period, as well as to nutrient stress. Exposure to elevated temperatures resulted in an increased Hb- 1 gene expression of 31% after 72 h only, whereas transcript abundance of the Hb -2 gene was enhanced by up to 59% by both 1-day and 3-day thermal stress conditions. Nutrient stress also increased gene expression of Hb -2 gene by 70%. Our findings describe the differential expression patterns of two novel Hb genes from symbiotic dinoflagellates and their polymorphic nature. Furthermore, the inducible nature of Hb -2 gene by both thermal and nutrient stressors indicates a prospective role of this form of hemoglobin in the initial coral–algal responses to changes in environmental conditions. This novel hemoglobin has potential use as a stress biomarker.",
"introduction": "Introduction Globin proteins are a diverse group of proteins, organized in a number of families and represented in all kingdoms of life (Vinogradov et al. 2006 ). Hemoglobin (Hb) proteins are a member of the globin family that contain a prosthetic group (heme) with iron (Fe +2 ) coordinated with the absolutely conserved proximal histidine (Vuletich and Lecomte 2006 ). The average size of Hb is 140–180 aa (Mr 15–18 kDa) and is characterized by a low homology between distant relatives (Suzuki and Imai 1998 ). A variety of hemo-proteins that exist in living organisms share similar tertiary structure (globin fold) and evolutionary origin, while displaying a large sequence diversity in their primary structure (Royer et al. 2005 ). In vertebrates, there are four types of globin proteins including hemoglobin, myoglobin, neuroglobin, and cytoglobin (Pesce et al. 2002 ). In plants, hemoglobins are divided into symbiotic and non-symbiotic hemoglobins, as well as truncated hemoglobins (Shimoda et al. 2005 ). Three groups of hemoglobins have been characterized in microorganisms: truncated hemoglobins containing 110–140 aa, flavohemoglobins containing hemoglobin and a flavin-containing reductase domain and myoglobin-like proteins (Egawa and Yeh 2005 ), with a number of microbial Hbs lacking a completely conserved goblin fold (Bonamore and Boffi 2008 ). Truncated hemoglobins are also found in bacteria, unicellular eukaryotes, and higher plants (Milani et al. 2005 ). Furthermore, hemoglobin proteins show a high diversity of their structural and also functional properties. In vertebrates, globin proteins are involved in the capture, transport, and storage of O 2 and CO 2 , whereas in invertebrates, they have preserved the function of O 2 binding (Lecomte et al. 2005 ). From an evolutionary point of view, the oxygen transport function is proposed to be related to the appearance of multicellular animals (Vinogradov et al. 2006 ). Hbs are also involved in scavenging nitric oxide (NO) and the protection of cells from NO damage (Egawa and Yeh 2005 ; Lecomte et al. 2005 ). During severe hypoxia stress in Arabidopsis , alfalfa, and maize, over-expressed non-symbiotic plant class 1 hemoglobin has been involved in reducing NO level and increasing overall the plants’ survival rate (Dordas et al. 2003a , 2004 ; Perazzolli et al. 2004 ). The importance of Hb in symbiosis has been suggested as high mRNA expression levels of non-symbiotic and truncated Hbs are observed in root nodules of Lotus japonicus compared with other plant tissues (Bustos-Sanmamed et al. 2011 ). Leghemoglobins, symbiotic plant Hbs, which are also found in root nodules of legumes are involved in O 2 transport to the nitrogen fixing bacteria and are as well required for symbiosis (Ott et al. 2005 ). Highly polymorphic and diverse Hb sequences indicate their capacity for a potential molecular mechanism of adaptation, and therefore they have been proposed to present a unique system for studying the effect of environmental changes on molecular evolution (Andersen et al. 2009 ). Symbiodinium are unicellular photosynthetic dinoflagellates, involved in a mutualistic symbiosis with a number of marine organisms such as scleractinian corals (Fig. 1 ; Muscatine et al. 1975 ; Trench 1979 ). Symbiotic dinoflagellates are phylogentically separated into nine clades (A–I) and then additionally into multiple subclades (Santos et al. 2002 ; Coffroth and Santos 2005 ; Pochon et al. 2006 ; Pochon and Gates 2010 ). It has been shown that different Symbiodinium clades and subclades can influence the physiological tolerance of the coral–dinoflagellate symbiosis to environmental stress (Rowan 2004 ; Berkelmans and van Oppen 2006 ; Robison and Warner 2006 ; Loram et al. 2007 ; Reynolds et al. 2008 ; Sampayo et al. 2008 ; DeSalvo et al. 2010 ; Fisher et al. 2012 ). Recent studies of gene expression levels in symbiotic dinoflagellates following the exposure of the coral–algal symbiosis to elevated temperatures have revealed differential regulation of molecular chaperones ( Hsp 70 and Hsp 90) and cytochrome P450 genes (Rosic et al. 2010 , 2011a , b ). Differential responses of Hsp 70 and Hsp 90 orthologs from both partners in symbiosis have been also observed under thermal stress conditions (Leggat et al. 2011 ). Figure 1 Coral Acropora aspera on the Great Barrier Reef, Australia (A). Light micrograph of Symbiodinium maintained in culture at constant temperature and light conditions (B). Temperature variation has been shown to affect the demand and the supply of oxygen, suggesting an adaptive role of different hemoglobin isoforms in marine fishes, in optimizing oxygen transport and the levels of oxygen and carbon dioxide under various thermal conditions (Sartoris et al. 2003 ; Pörtner and Knust 2007 ). Hbs are also involved in the scavenging of NO, which is a free radical and a membrane-permeable molecule involved in the immune responses, and in establishing and maintaining coral–algal symbiosis (Gardner et al. 1998 ; Trapido-Rosenthal et al. 2001 ). The connection between NO and thermal stress has been suggested as elevated temperatures resulted in an increase in NO production in the sea anemone Aiptasia followed by a breakdown of the symbiosis (Trapido-Rosenthal et al. 2001 ; Perez and Weis 2006 ). In the present study, we characterize the sequence polymorphisms of two putative hemoglobin genes identified within expressed sequence tags (ESTs) of Symbiodinium (clade C3) (Leggat et al. 2007 ), provide their phylogenetic analyses, and propose the 3D protein structures. In addition, we apply transcriptional analyses to coral dinoflagellates both in symbiosis and in vitro cultures to determine the changes in the gene expression patterns of hemoglobin-like proteins when exposed to different thermal and nutrient stress conditions. Finally, we discuss the potential importance of Hb genetic polymorphisms as a tool of evolutionary adaptation.",
"discussion": "Discussion and Conclusions An important strategy to increase stress tolerance in plants and therefore the survival rate includes the contribution of hemoglobin proteins (Dordas 2009 ). These heme-containing proteins represent an ancient class of ubiquitous oxygen-binding proteins (Vinogradov et al. 2006 ) that after the extensive evolutionary pressure acquired a number of new features enabling them to adapt to extreme conditions and to preserve their functionality (Perutz 1983 ). Consequently, these proteins have been used for monitoring the adaptive changes in organisms exposed to variable external conditions (Andersen et al. 2009 ). Recent advances in sequencing technologies have resulted in the discovery of Hb -like sequences in many prokaryotic and eukaryotic microorganisms including bacteria, yeasts, algae, protozoa, and fungi and the presence of microbial globins such as truncated hemoglobins (trHb), globin-coupled sensors (GCSs), and flavohemoglobins (flavoHbs) (Bonamore and Boffi 2008 ). In the present study, we report the presence of two globin proteins in coral dinoflagellates, which are represented with several isoforms and a highly conserved hemoglobin residue, the proximal histidine (F8) (Fig. 2 ). The predicated 3D protein structure confirmed the globin fold for these dinoflagellate proteins and their preserved tertiary structure (Fig. 3 ). Likewise, many different hemo-proteins found in nature have similar tertiary structure (globin fold), as well as evolutionary origin, although showing a huge variability in their amino acid sequences (Royer et al. 2005 ). Our phylogenetic studies revealed the existence of two separate groups of hemoglobin proteins (Fig. 4 ). The BLAST search confirmed the Hb-like origin of these proteins and additionally using PSI-BLAST, their evolutionary link with microbial flavohemoproteins. Molecular phylogeny suggested a close evolutionary relationship between the Symbiodinium Hb-2 and its metazoan cytoglobin counterparts that act as a NO scavenger and play a role in oxidative stress response (Trent and Hargrove 2002 ) indicating a possibility of contamination or horizontal gene transfer. However, as Hb sequences were recovered also in cultures of different Symbidinium types (Fig. 2 B), a possibility of contamination has been excluded. Furthermore, microbial origin of eukaryotic globins has been proposed as a result of horizontal gene transfer, which has occurred in the past, during endosymbotic events and lead to the first establishment of mitochondria and plastids such as chloroplasts in the eukaryotic cells (Hoogewijs et al. 2012 ). The complex evolutionary origin of dinoflagellate genes involved in the biosynthesis of mycosporine-like amino acids was also recently reported (Rosic and Dove 2011 ; Rosic 2012 ), as well as the occurrence of these microbial genes within the coral genome (Shinzato et al. 2011 ). The Hb expression patterns can be affected by a number of factors including hypoxia, organogenesis, pathogen infection, and ontogenesis (see review by Kosmachevskaya and Topunov 2009 ). Our results revealed differential gene regulation of two algal Hbs proteins when the coral–dinoflagellate symbiosis was exposed to thermal and nutrient stress. The increased transcript abundance of Hb- 1 after a 3-day period of thermal stress and even more inducible Hb- 2 expression to both thermal and nutrient stress conditions (Fig. 5 ) may be due to a Hb role in the cell protection and in the process of scavenging NO (Dordas 2009 ; Kosmachevskaya and Topunov 2009 ). Consequently, it could be expected that Symbiodinium Hbs may also be involved in the metabolism of NO during thermal stress. The NO molecule is known as a very potent signaling molecule involved in a number of biological processes such as initiating host immunity response against pathogen invasion (Wang and Ruby 2011 ), as well as in signaling within coral–algal symbioses (Safavi-Hemami et al. 2010 ). A high level of NO production in the sea anemone exposed to elevated sea temperature can lead to the collapse of cnidarian–dinoflagellate symbiosis (Trapido-Rosenthal et al. 2001 ; Perez and Weis 2006 ). Additionally, nitrate, nitrite, and NO were found to induce the synthesis of non-symbiotic-Hb (Nsgb) in plants (Wang et al. 2000 ; Ohwaki et al. 2005 ), while Nsgl functioned as a NO dioxygenase, modulating NO metabolisms, and NO detoxification (Dordas et al. 2003b ; Hebelstrup et al. 2007 ). Here, we also report a further increase in Hb -2 expression by 70% when coral–dinoflagellate symbiosis was exposed to ammonium-enriched seawater for a 3-day period. Nutrient over-enrichment is considered as one of the leading factors leading to coral decline (Szmant 2002 ). As algal endosymbionts and their invertebrate host exchange nutrients and metabolic products (Venn et al. 2008 ; Yellowlees et al. 2008 ), they also show the capacity to quickly fix nitrogen from the seawater enriched with ammonium, with much higher intake reported for symbiotic dinoflagellates compared with the host (Pernice et al. 2012 ). Nitrogen assimilation has been stimulated by over-expression of plant Hbs (class 1) that removes NO acting as an inhibitor of nitrogenase (Shimoda et al. 2009 ). Consequently, elevated transcript levels of Hb -2 mRNA reported here may indicate a potential role of this Hb form in NO detoxification and also enhancing the process of nitrogen assimilation in coral endosymbionts. Future studies are needed to determine the molecular mechanism of nitrogen absorption and NO detoxification and the role of Hb proteins during these processes. Low Hb transcript levels observed for Symbiodinium cultures and the abundant and inducible expression of Hb genes in hospite may suggest that algal Hb gene expression requires the symbiotic condition or alternatively the Hb importance for symbiosis as seen in some plants (Bustos-Sanmamed et al. 2011 ). A lack or low transcript abundance for catalase, an antioxidant enzyme, has been also reported for Symbiodinium cultures (Bayer et al. 2012 ), whereas a high level of gene expression was detected in the cnidarian–dinoflagellate symbiosis (Sunagawa et al. 2009 ). Despite the large ecological and socio-economic importance of coral reefs worldwide, our understanding of their ability to adjust to changing environmental conditions is poorly developed. A number of mechanisms have been proposed to potentially increase the coral–algal stress tolerance including inducible HSPs, production of oxidative enzymes, and fluorescent coral pigments (Coles and Brown 2003 ; Baird et al. 2009 ). The differences in stress tolerance found in corals could potentially be driven by genetic adaptation and/or phenotypic acclimatization (Weis 2010 ). Phenotypic plasticity, in response to thermal stress, has been reported for both partners in symbiosis (see review Weis 2010 ). Photo-acclimation of symbiotic dinoflagellates to high light levels can lead to higher thermal tolerance (Robison and Warner 2006 ). Previous exposure to temperature fluctuations in the environment can positively influence coral thermal tolerance (Oliver and Palumbi 2011 ). Hemo-proteins such as hemoglobins have been implicated as an important indicator of an organism's capacity to respond to environmental change due to their involvement in oxygen transport and related metabolic processes (Andersen et al. 2009 ). Our research suggests that Symbiodinium Hb genes, in particular Hb -2, play a role in the mechanisms of the early stress response during exposure of coral–dinoflagellate symbiosis to thermal and nutrient stress. Additional research is needed to elucidate the exact mechanisms of algal Hb transcriptional regulation upon exposure to stress and functional significance of hemoglobin polymorphism in Symbiodinum . In conclusion, this research provides new insights into the molecular changes occurring in symbiotic dinoflagellates under stress. Differential gene expression patterns of these highly polymorphic hemoglobin-like proteins of coral dinoflagellates indicate that these universal globin proteins may play an important role in the coral–algal stress response potentially through physiological acclimatization and/or evolutional adaptation to climate change."
} | 4,019 |
34512246 | PMC8427800 | pmc | 1,954 | {
"abstract": "Computing paradigm based on von Neuman architectures cannot keep up with the ever-increasing data growth (also called “data deluge gap”). This has resulted in investigating novel computing paradigms and design approaches at all levels from materials to system-level implementations and applications. An alternative computing approach based on artificial neural networks uses oscillators to compute or Oscillatory Neural Networks (ONNs). ONNs can perform computations efficiently and can be used to build a more extensive neuromorphic system. Here, we address a fundamental problem: can we efficiently perform artificial intelligence applications with ONNs? We present a digital ONN implementation to show a proof-of-concept of the ONN approach of “computing-in-phase” for pattern recognition applications. To the best of our knowledge, this is the first attempt to implement an FPGA-based fully-digital ONN. We report ONN accuracy, training, inference, memory capacity, operating frequency, hardware resources based on simulations and implementations of 5 × 3 and 10 × 6 ONNs. We present the digital ONN implementation on FPGA for pattern recognition applications such as performing digits recognition from a camera stream. We discuss practical challenges and future directions in implementing digital ONN.",
"conclusion": "5. Conclusion In this paper, we carried out the questions—can we use ONN for image recognition, and—what are the advantages and limitations of ONN for AI at-the-edge applications. To do so, we presented a proof of concept of the ONN neuromorphic computing paradigm with a fully digital design. We validated the computing capability of a 5 × 3 ONN and a 10 × 6 ONN performing pattern recognition both in simulation and FPGA implementation. We used Hebbian or Storkey learning rules to train our ONN. For both learning rules, with three stored patterns, the 5 × 3 ONN retrieved 14 test patterns out of a test set of 15. For the 10 × 6 ONN with five stored patterns, results differ from Hebbian and Storkey learning rules. ONN with weights computed with Storkey can retrieve 25 test patterns out of 25, but using Hebbian ONN can only retrieve 20 test patterns out of 25. Further experiments confirmed that Storkey is more accurate than Hebbian for an equal number of stored patterns. We performed additional experiments to characterize our digital ONN. First, we estimated by simulation a maximum operating frequency for the 5 × 3 ONN to 83,33 MHz. Then, we showed that for a specific application, the ONN implemented on FPGA could go up to 125 MHz, without any changes on the ONN operation. Besides, we performed a timing analysis on digital ONNs. We measured the initialization time needed to apply the input pattern to the ONN, and the computation time, needed by the ONN to stabilize to a stored pattern. From measurements, we were able to calculate the maximum FPS (Frames per second). For the 10 × 6 ONN, we obtained a maximum FPS around 76000, at 31.25 MHz with a training configuration resulting in all test patterns successfully retrieved. Then, we embedded the 10 × 6 ONN into a complete image recognition application performing digit recognition from a camera stream. It respected real-time constraints, and we demonstrated that the ONN paradigm can fit with AI at-the-edge image recognition application. Thus, despite the size limitation (about a 100 neurons) of our digital design due to high FPGA resource consumption, the huge potentiality of ONN is undeniable. ONNs are still in their infancy for comparison with standard benchmarks and it is the focus of future works. The potential of the proposed ONN propels further investigation to explore its capabilities on diverse applications for AI at-the-edge.",
"introduction": "1. Introduction In recent years, we have witnessed a proliferation of smart edge devices adopted by all industry sectors such as smart homes, smart city cameras, smart autonomous driving cars, smart healthcare, smart manufacturing, etc. Most edge devices are getting smaller and compact. Using Artificial Neural Networks (ANNs), specifically Deep Neural Networks (DNNs) to create Artificial Intelligence (AI) at the edge, has successfully been used to teach smart systems to recognize or detect objects (Redmon et al., 2016 ; Krizhevsky et al., 2017 ; Shah and Kapdi, 2017 ; Yang and Song, 2018 ; Jiao et al., 2019 ), read texts (Jackel et al., 1991 ), and understand speeches (Xiong et al., 2018 ; Nassif et al., 2019 ). Contraints to such applications on edge devices derives from the inherent limitations of power consumption, memory, and little to no bandwidth. In addition, privacy and security concerns would recommend the data to be stored locally. In contrast, ANNs or DNNs systems that enable AI at the edge are getting larger to cope with the ever-increasing amount of data. Thus, resulting in more power consumption, memory, and bandwidth demand. Systems based on ANNs and Convolutional Neural Networks (CNNs) running on traditional von Neumann architectures require a large amount of memory, computational power, and bandwidth demand. While they perform well on expensive hardware such as GPUs (Pham et al., 2019 ), they are often unsuitable for smaller edge devices. Such a disconnect between the growing need in AI at the edge and limitations of processing hardware has compelled significant research efforts into beyond-von Neumann systems such as neuromorphic computing paradigms deployable at the edge (Bey, 2020 ; Kendall and Kumar, 2020 ). This paper focuses on an alternative, low-power, neuromorphic computing approach with Oscillatory Neural Networks (ONNs) (Raychowdhury et al., 2019 ; Csaba and Porod, 2020 ). The ONN is a system of coupled oscillators mimicking at circuit level the basic structure of the brain architecture. The key feature of ONNs is to encode the information on the phase relations between oscillators and to let them oscillate using their physical dynamics to compute. For example, the random start of five metronomes (similar to grandfather's clock) will make them oscillate in parallel (Met, 2013 ). After several cycles, they get synchronized in frequency while their phase relations can tell us if they are in- or out-of-phase. Contrarily to the classical computation based on voltage amplitude to determine a logic “1,” or “0,” in ONN we use the phase relations to determine the logic “1” (out-of-phase 180 o ) or “0” (in-phase 0 o ). Thus, working with parallel oscillators in the frequency and phase domains allows to reach fast and low-power computation (Roychowdhury, 2014 ; Shukla et al., 2016 ). This makes ONN an ideal solution to bring artificial intelligence on edge devices. ONN principle was first introduced in Hoppensteadt and Izhikevich ( 2000 ) where ONN showed good associative memory properties. Thus, there is a recent interest to exploit ONN for large-scale associative memory applications. While there is a lot of ongoing research on devices and analog architectures to implement ONN efficiently (Csaba and Porod, 2013 ; Jackson et al., 2015 ; Shi et al., 2016 ; Kumar and Mohanty, 2017 ; Corti et al., 2019 ; Velichko et al., 2019 ), we focus on addressing a more fundamental problem—can we perform relevant AI applications such as image recognition with ONN? We explore ONN at small-scale (up to 60 coupled oscillators) as a first attempt to show phase computation in the digital domain. Despite being a small-scale ONN, we investigate advantages and limitations on image recognition tasks suitable for AI applications on edge devices. To do so, we implement an FPGA-based digital ONN to serve as a proof-of-concept of the ONN computing paradigm for enabling AI at the edge. The rest of the paper is organized as follows. In section 2, we present materials and methods used for all experiments carried out for this work. In section 2.1, we introduce the ONN model and compare it with state-of-the-art associative memory models. Then, in section 2.2, we present the training methods we apply to the ONN. Afterward, in section 2.3, we describe the digital ONN design. section 2.4 presents methods used for ONN validation and characterization for design simulation and FPGA implementation. Next, section 2.5 exposes methods for a 10 × 6 ONN used for image recognition from a camera stream. section 3 reports on results related to ONN simulation, ONN FPGA implementation, and image recognition application using such 10 × 6 ONN. Finally, in section 4, we discuss the advantages and limits of ONN and future directions.",
"discussion": "4. Discussion Our work presents the design and performances of a novel fully digital ONN implementation. Further, we demonstrate ONN on FPGA implementation for image recognition applications. It is a proof-of-concept on the potential of ONN as an alternative computing paradigm. Here, we present the digital ONN design aspects and their advantages, limits, and future directions. 4.1. Advantages Presently, we developed ONN designs that exhibit good performances with 5 × 3 and 10 × 6 ONNs. An interesting feature of our approach in comparison with Jackson et al. ( 2019 ) is that by resorting to the 1-bit oscillation at the neuron's output, multipliers are avoided in the synapses block, while still retaining a multi-level neuron. This is possible because we encode the state in the neuron's oscillation phase. Another advantage of our approach is the easiness of training. Results reported in this paper have been obtained using simple Hebbian and Storkey rules. Given the patterns to be stored/recognized, weights are calculated offline by using matrix operations, while training other neural models can be a very time-consuming operation. Limited retrieval capacity has been obtained in some of the experiments described. However, the accuracy of the ONN can be increased by resorting to enhanced learning rules. Finally, a fundamental advantage of ONN is the fast computation. ONN parallel behavior allows the independence of the computation time and the network size. In this paper, we achieved more than 70000 FPS with 10 × 6 ONN and a serial initialization of the neurons. 4.2. Limitations and Future Directions The present digital ONN design has also limitations. First, our ONN is a preliminary design developed to validate the concept, and different optimization procedures at different levels are currently being carried out. Second, the limited FPGA resources introduce some constraints on the ONN size that can be implemented on FPGA (see Table 3 ). Such limited resources are the computation resources (LUTs). They are used to implement the combinatorial synapses block whose size increases quadratically with the number of neurons. The limited size does not allow for comparison with standard benchmark sets usually used to evaluate neural networks. For example, comparing ONN with SNN is not trivial because of the paradigm differences, such as the different network architectures and learning algorithms. To make a meaningful comparison, a common ground is necessary with a common application and benchmark. Comparison to other reported similar FPGA-based implementations of neural networks is meaningful only if the same ONN size are developed. The Associative Memory Neural Networks (AMNNs), such as HNNs, are the closest ANNs that can be compared with ONNs. In the existing literature, we found digital implementations of AMNNs such as Leiner et al. ( 2008 ), Mansour et al. ( 2011 ), and De Abreu de Sousa et al. ( 2014 ). The most relevant comparisons can be made with the digital design from De Abreu de Sousa et al. ( 2014 )'s work, which is also the most recent one. In this work, authors perform a frequency study for multiple HNN sizes, from 16 neurons to 32 neurons. Stored patterns resemble our stored patterns representing digits. For example, they use 8 × 4 representations of the letter U and the number 5. Tests are also similar as they use stored and corrupted patterns as inputs. Table 6 shows the comparison between (De Abreu de Sousa et al., 2014 )'s HNN and our ONN. Results show that ONN has a higher operating frequency than HNN. Such as in the 32 neurons case, the maximum HNN frequency is 33.55 MHz, but ONN can run faster, up to 60.61 MHz for 10 × 6 ONN. We do not have information about the computation time for the 32-neuron HNN, however, if we compare the 16-neuron HNN with the 5 × 3 ONN, we expect similar trends in frequencies and computation times. Though it is difficult to make any conclusive comparisons, it seems that our larger ONN can operate at higher frequencies than the cited HNN. Table 6 Frequency, computation time and resource comparison of ONN and HNN designs. \n Neural \n \n Size \n \n Frequency \n \n Computation \n \n Computation time \n \n Resources \n \n network \n \n (MHz) \n \n time (us) \n \n (clock cycles) \n \n (LUTs) \n \n Max - Avg \n \n Max - Avg \n HNN (*) 16 81.39 x - 1.3 x - 100 390 HNN (*) 32 33.55 x - x x - x 700 ONN - Hebbian 15 83.33 2.74 - 0.92 228 - 77 958 - Storkey 15 83.33 1.18 - 0.77 98 - 64 800 ONN - Hebbian 60 64.1 3.53 - 1.44 226 - 92 6,426 - Storkey 60 60.61 1.75 - 1.18 106 - 72 6,192 (*) Refers to De Abreu de Sousa et al. ( 2014 ). Note, HNN resource estimation comes from data extraced from De Abreu de Sousa et al. ( 2014 ) and FPGA documentation. Also note, HNNs are implemented on a Xilinx 3-series FPGA (4-input LUTs) while ONNs are implemented on a Xilinx 7-series FPGA (6-input LUTs) . Motivated by the potential of the proposed ONN, we are currently exploring optimizations in several directions like hardware resources, frequency, and accuracy. For example, using a faster internal clock than the oscillator frequency, a sequential implementation could be used to reduce the size of this resource-consuming block. Also, we can explore additional learning rules to increase accuracy. At this time, we have only studied Hebbian and Storkey, which are local and incremental with a limited storage capacity, but other learning rules will also be explored to have a better assessment of learning rules suitable for ONNs. For example, the pseudo-inverse rule (also called projection rule) can increase HNNs storage capacity and improve accuracy (Wu et al., 2012 ; Sahoo et al., 2016 ), but is not local nor incremental. Moreover, recent works have explored learning rules with self-feedback connections (non-0 diagonal), and have shown higher accuracy for a high number of stored patterns (Liou and Yuan, 1999 ; Folli et al., 2017 ; Rocchi et al., 2017 ; Gosti et al., 2019 ). In summary, despite the present limitations of the ONN, features in terms of FPS, computation time and training, are encouraging toward the exploration of a wider range of applications."
} | 3,670 |
37748078 | PMC10556619 | pmc | 1,955 | {
"abstract": "Significance Designing soft materials that permit the facile conversion of mechanical deformation into electricity is of significant technological importance. In this work, using fundamental science principles and the exotic phenomenon of flexoelectricity, we design a natural vegetable-based luffa sponge to “act” as a strong piezoelectric material suitable for energy harvesting, sensing and myriad other applications. We demonstrate applications of our developed materials in the context of sensing, actuation, and electrical energy harvesting for speech/voice recognition, pressure sensing, and powering light-emitting diodes. The natural luffa material is biodegradable, abundantly available and thus “green”. The developments we report constitute a paradigm for green and flexible sensors and energy harvesters with unique advantages of light weight, low cost, and full biodegradability.",
"conclusion": "Concluding Remarks We have investigated the flexoelectric effect in the natural LS, which is a low-cost, biodegradable, biocompatible, ultralight biomaterial. After the chemical treatment, the LS sample with the size of 20 × 20 × 6 mm 3 generates an electric current of 8 nA under 50% applied strain. The density-specific equivalent-specific piezoelectric coefficients of the LS are highest among known materials including the often-used PZT. We demonstrate potential applications of LSs in the field of wearable sensors for monitoring human movement (such as finger bending, vocal cord vibration, and speech recognition) and pressure detection. Moreover, the LS is a SGG, which can convert mechanical energy into electric energy to power devices."
} | 410 |
30479325 | PMC6258724 | pmc | 1,956 | {
"abstract": "Microbes in Guaymas Basin (Gulf of California) hydrothermal sediments thrive on hydrocarbons and sulfur and experience steep, fluctuating temperature and chemical gradients. The functional capacities of communities inhabiting this dynamic habitat are largely unknown. Here, we reconstructed 551 genomes from hydrothermally influenced, and nearby cold sediments belonging to 56 phyla (40 uncultured). These genomes comprise 22 unique lineages, including five new candidate phyla. In contrast to findings from cold hydrocarbon seeps, hydrothermal-associated communities are more diverse and archaea dominate over bacteria. Genome-based metabolic inferences provide first insights into the ecological niches of these uncultured microbes, including methane cycling in new Crenarchaeota and alkane utilization in ANME-1. These communities are shaped by a high biodiversity, partitioning among nitrogen and sulfur pathways and redundancy in core carbon-processing pathways. The dynamic sediments select for distinctive microbial communities that stand out by expansive biodiversity, and open up new physiological perspectives into hydrothermal ecosystem function.",
"introduction": "Introduction Microbial communities inhabit every environment and are comprised of a multitude of different phyla, the majority of which are uncultured 1 . Among these environments, marine sediments contain abundant and phylogenetically diverse microbial communities 2 – 4 . High diversity has been suggested to emerge as a strategy for survival of microbes under fluctuating environmental conditions in nature 5 , 6 . While single-gene surveys allow us to address the phylogenetic diversity of microbial communities, metagenomic analyses provide a connection between diversity and the functional potential encoded within sedimentary communities. Guaymas Basin (GB; Gulf of California, Mexico) is a young, active seafloor-spreading center characterized by high water column productivity and fast sedimentation rates, leading to the accumulation of massive layers of organic-rich sediments that cover the hydrothermal spreading center and ridge flanks 7 – 9 . The emplacement of hot basalt sills into organic-rich sediment transforms buried organic matter into CO 2 , H 2 , low-molecular-weight organic acids, ammonia, and hydrocarbons such as methane, ethane and benzene 8 , 10 , 11 . These compounds migrate to the sediment surface with rising vent fluids, where they fuel hydrocarbon-degrading microbial communities 11 , 12 . Among all hydrothermally generated hydrocarbons, methane has received considerable interest as greenhouse gas shaping global climate 13 . Porewater methane reaches millimolar concentrations while ethane ranges from 40-100 µM. Also present in these sediments are propane, n-butane and pentane, which accumulate at lower concentrations compared to methane. Altogether, hydrocarbons represent lucrative carbon sources for the resident microbial community 11 , 14 – 16 . Additionally, hydrothermal circulation and seawater in-mixing provide the upper sediments with electron acceptors, among which sulfate is widely available in millimolar porewater concentrations and rarely depleted within hydrothermal sediment cores 11 , 14 , 17 . In-situ microelectrode surveys detect small oxygen peaks within hydrothermal sediments near the mat-covered surface 18 , 19 . These results are consistent with short-term dynamics of hydrothermal flow within minutes and hours 17 . Additionally, short-term dynamics overlay with longer-term hydrothermal activity changes over months and years 18 . GB sediments have been shown to host diverse microbial communities with distinct roles in carbon cycling 11 , 17 , 20 . In particular, microbial consortia perform the anaerobic oxidation of methane (AOM) in a syntrophic interaction consisting of anaerobic methane-oxidizing archaea (ANME) and bacterial sulfate reducers, typically Deltaproteobacteria, but including other thermophilic bacterial lineages, such as Candidatus Desulfofervidus auxilii 21 – 23 . Anaerobic hydrocarbon degraders include Ca. Syntrophoarchaeum, which oxidizes butane in a syntrophic interaction with Ca. Desulfofervidus auxilii, or the butane- and propane oxidizing isolate BuS5, belonging to the Desulfosarcina - Desulfococcus cluster 15 , 16 . Other common archaeal lineages include Marine Benthic Group D and Bathyarchaeota, while bacterial phyla include Proteobacteria (Delta-, Epsilon- and Gammaproteobacteria), Bacteroidetes and Chloroflexi as well as several candidate phyla 11 , 17 , 24 . Within the GB hydrothermal area in the southern spreading center, a high degree of microbial community connectivity exists among hydrothermal vent sites and sediments within a few hundred meters 25 . A core microbiome is shared between microbial communities of GB hydrothermal sediments and cold seeps in the Sonora Margin, within a few km distance; this microbiome is thought to be involved in organic matter degradation as well as methane and carbon cycling, suggesting microbial exchanges across neighboring sites that share geochemical characteristics, such as abundant methane concentrations 26 . Previously, we employed metagenomic reconstructions of two GB sedimentary microbial communities, showing the interconnectivity of carbon, sulfur and nitrogen cycling among lineages 20 . However, despite these advances, we still have a limited understanding of the spatial biodiversity and full metabolic potential of microbes inhabiting the basin. Here we characterize the biodiversity and physiological capabilities of genomes from microbial communities inhabiting GB sediments. The highly localized hydrothermal gradients in surficial GB sediments are ideal to compare adjacent sites with distinct temperature and chemical regimes 18 , 27 . We selected samples from methane- and sulfate-rich hydrothermal sediments covering a wide thermal range, and contrasted them with cold, non-hydrothermal sediments, as well as with hot, oil-rich sediments. We hypothesize that microbial assemblages from hydrothermal sediments are phylogenetically distinct from those in the surrounding region and host a greater metabolic diversity. Therefore, we sequenced a total of ~4 billion genomic reads from eleven samples (two of which were from cool, background sediments) from GB. Altogether, these data add 22 branches to the tree of life and enabled to us determine the genetic repertoire and metabolic versatility of these extreme hydrothermal communities.",
"discussion": "Discussion In this study, we employed the largest genomic sampling of GB sediments to date to investigate the interplay of community composition and functional diversity. Compared to earlier work on Guaymas Basin sediments 20 , the higher sampling number and inclusion of background samples allowed to better describe the enhanced diversity present in these sediments and shed light on the drivers of community assembly. In contrast to previous studies showing that sulfidic- and methane-rich seep sediments host a lower microbial diversity compared to non-seep marine sediments 39 , 40 , we demonstrate that GB hydrothermal sediments contain a diverse community that is enriched in archaea compared to a less diverse, bacterial-dominated community found in nearby cold sediments. Therefore, the more extreme conditions in hydrothermal sediments, which include steep thermal and geochemical gradients 17 , 27 , appear not to inhibit microbial diversity. Due to difficulties in isolating sufficient amounts of DNA from deeper, hotter samples, we cannot exclude that diversity may decline in those sediments. Earlier work reported a decrease in cell numbers with increasing depth that did not necessarily correlate with a decrease in OTU numbers 25 , potentially explaining our difficulties in isolating sufficient amounts of DNA but supporting our assumption that steep temperature gradients do not necessarily inhibit microbial diversity. Especially samples from core 4569_9 experience a highly variable, fluctuating thermal regime over time, where even surficial layers can vary from 20 °C to 70 °C, as determined by multi-day continuous thermal logging (Supplementary Figure 1 ) 17 . In response to such conditions, microbes must either adapt, have a wide thermal optimum, as shown for some ANME-1 archaea 23 , or be able to recolonize the sediment after a temperature sweep from a surficial reservoir 41 . Here, we propose that the diverse communities inhabiting hydrothermal sediments could serve as a flexible seed bank for the deeper, hotter sediments as well as for highly fluctuating environmental gradients in shallow sediments 5 , 25 , 42 . The differences we observed in community composition across sites were not always translated into obvious changes in functional capacities of those communities. For example, we detected abundant genes for carbon cycling and fermentation across all sites, while other metabolic processes such as respiration, were limited to shallow sediments but present in both background and hydrothermal sediments. Respiratory processes were often partitioned among the community and only few genomes were encoding for full pathways. Metabolic handoffs have been observed in other microbial communities and could allow a flexible interchange of metabolites between changing populations 43 , 44 . Another metabolic feature that could allow for greater ecosystem stability could be metabolic plasticity, i.e. switching metabolic processes in response to changes in environmental conditions. We found indications for such plasticity in several bacterial genomes, especially within the Delta- and Gammaproteobacteria that might couple the reduction of sulfur with the oxidation of carbon, lipids or hydrocarbons. While we cannot determine which processes are active, enhanced genotypic diversity might provide an additional adaptation strategy to variable environmental conditions. The only functional categories that were consistently enriched across all hydrothermal sites and almost absent in background sediments were group 4g hydrogenases and pathways for methanogenesis and methane oxidation. Group 4g hydrogenases are not well characterized but are generally described to be membrane-bound hydrogenases that allow for energy-generation by establishing ion gradients over the membrane 45 . These complexes are often found in thermophiles, such as Pyrococcus furiosus 45 , and could potentially provide a selective advantage in hydrothermal sediments over other energy-generating systems. While trace concentrations of biogenic methane are present in background sediments (Supplementary Data 1 , Supplementary Methods), the inability to detect mcrA in these samples could be because of sequencing depth; in contrast detecting mcrA in hydrothermal sediments appears to be linked to microbial methane oxidation produced by pyrolysis of organic matter 17 . Within the phylogenetically and functionally diverse community inhabiting GB, the metabolic repertoire shows a high degree of functional redundancy across different phyla, i.e. different taxa encode the same metabolic function and thus might substitute for one another. Therefore, even if community composition varies, metabolic function is predicted to be relatively stable. Like phylogenetic diversity, functional redundancy could benefit the community when dealing with perturbations in environmental conditions and has been observed in other environments including the global marine or humane microbiome 46 , 47 . While any stressor, such as temperature, might result in the removal of a given taxon, functional redundancy across different lineages that are each tolerant to some degree of environmental fluctuations, and together cover a wide window of environmental conditions, ensures the stability of community function. This is consistent with the ‘it’s the song not the singer’ (ITSNTS) theory, which assumes that surviving taxa replace perturbed taxa (‘the singers’) and thereby allow nutrient cycles (‘the song’) to persist in the environment 48 . This theory is consistent with our findings, in which we not only observe phylogenetically diverse but also functionally redundant communities. Altogether, the phylogenetic diversity, metabolic partitioning as well as functional redundancy that we observe appear to be characteristics of microbial communities in these dynamic hydrothermal vent sediments. One question that arises when observing functional redundancy within a microbial community is whether this redundancy enhances species competition and de-stabilizes the community 49 , 50 . While it is not in the scope of this study to discern niche patterns, we would assume that the high redundancy in our dataset might still allow microbes to inhabit different niches. Two mechanisms that could allow co-existence of supposedly redundant microbes could be metabolic auxotrophies or heterogeneity in limiting resources and/or environmental conditions 50 – 52 . Amino acid auxotrophies can create community interdependencies, which could balance competition and thereby stabilize microbial communities 53 . We do see indications for such interdependencies in our dataset, where auxotrophies are common in small genomes belonging to CPR bacteria (Supplementary Data 10 ). Additionally, we assume that the diverse GB-inhabiting communities are stabilized by the high abundance of substrates present in hydrothermal sediments, which might reduce competition and allow taxa to coexist. Finally, while genes for core metabolic processes showed a high redundancy across our dataset, we hypothesize that enzymes involved in substrate degradation are undergoing substantial diversification with respect to their substrate spectra. The diversity of genes involved in carbohydrate ( mcrA , CAZYmes), lipid (acyl-CoA dehydrogenase) and peptide degradation and the expanding substrate range and diversity of hydrocarbon-degrading genes, such as mcrA , supports this notion 16 , 20 , 54 . A limitation of the current study that complicates a definite description of the diversity patterns and functional redundancy present in Guaymas sediments is the low sample number and limited number of bins recovered from a subset of samples (i.e. 4567_28 and 4488_9); given the limitations of deep-sea sampling, different habitat and sediment types are represented unevenly. Activity-based analyses of large sample numbers, i.e., metatranscriptomics, would more rigorously link genetic patterns to their environmental determinants. Guaymas Basin is a hotspot for microbial biodiversity and an ideal study site to investigate the functional diversity of hydrothermally influenced seafloor sediments. Here we establish that these hydrothermal sediments contain a large number of archaeal and bacterial lineages, including several uncultivated phylum-level lineages that have not been described from other habitats. Intriguingly, hydrothermal GB sediments hosted a greater diversity compared to surrounding non-hydrothermal sediments contrasting previous work on methane seep communities 39 , 40 . These differences are likely linked to the unique environment in GB sediments characterized by by convective mixing of fluids resulting in variable thermal regimes, and admixture of hydrothermal carbon and energy sources. Most functional properties were shared widely among different phylogenetic lineages across different sampling sites with a greater functional redundancy of metabolic processes found in hydrothermal sediments. One unique functional trait of hydrothermal compared to background sediments was the presence of methane cycling genes among novel lineages, including a new deep-branching Crenarchaeota group. We propose that the combination of dynamic seep and hydrothermal conditions in Guaymas Basin enhances microbial diversity, and sustains a distinctive microbial community, whose functional complexity and redundancy reflects the intricate and dynamic geochemical and thermal landscape of this habitat."
} | 3,999 |
31015996 | PMC6468064 | pmc | 1,957 | {
"abstract": "Abstract Climate change threatens coastal benthic communities on a global scale. However, the potential effects of ongoing warming on mesophotic temperate reefs at the community level remain poorly understood. Investigating how different members of these communities will respond to the future expected environmental conditions is, therefore, key to anticipating their future trajectories and developing specific management and conservation strategies. Here, we examined the responses of some of the main components of the highly diverse Mediterranean coralligenous assemblages to thermal stress. We performed thermotolerance experiments with different temperature treatments (from 26 to 29°C) with 10 species from different phyla (three anthozoans, six sponges and one ascidian) and different structural roles. Overall, we observed species‐specific contrasting responses to warming regardless of phyla or growth form. Moreover, the responses ranged from highly resistant species to sensitive species and were mostly in agreement with previous field observations from mass mortality events (MMEs) linked to Mediterranean marine heat waves. Our results unravel the diversity of responses to warming in coralligenous outcrops and suggest the presence of potential winners and losers in the face of climate change. Finally, this study highlights the importance of accounting for species‐specific vulnerabilities and response diversity when forecasting the future trajectories of temperate benthic communities in a warming ocean.",
"introduction": "1 INTRODUCTION From polar oceans to tropical seas, climate change dramatically affects marine ecosystems by influencing processes at all levels of biological organization (Doney et al., 2012 ; Poloczanska et al., 2016 ; Scheffers et al., 2016 ). Moreover, this anthropogenic pressure will continue to cause unprecedented impacts in the oceans during the next decades as global sea surface temperatures continue to rise and marine heat waves become more frequent and intense (Bellard, Bertelsmeier, Leadley, Thuiller, & Courchamp, 2012 ; Oliver et al., 2018 ). However, climate change effects have contrasting impacts on biotas (McKinney & Lockwood, 1999 ). Therefore, understanding how different species, populations and communities will respond to warming is key to developing specific conservation and management strategies aimed at enhancing the resilience of vulnerable marine ecosystems. Coastal benthic communities such as tropical and temperate reefs are among the most biologically diverse and socioeconomically valuable systems on the planet (Ballesteros, 2006 ; Bennett et al., 2016 ; Spalding, Ravilious, & Green, 2001 ). Nonetheless, when facing global warming, they are especially under threat. As migrating toward more thermally suitable conditions is not an option for most sessile species, most organisms from these communities will be compelled to rely on effective acclimatization (an adjustment of physiology via phenotypic plasticity) or adaptation (an increased abundance of tolerant genotypes over generations) processes to prevail. Although these two mechanisms that evolved for coping with environmental change will likely allow diverse species and/or populations to persist (Palumbi, Barshis, Traylor‐Knowles, & Bay, 2014 ), increasing evidence indicates that the unusually high rates of warming and the increasing frequency of extreme events may prevent many others from effectively doing so (Heron et al., 2017 ; Hoegh‐Guldberg, Poloczanska, Skirving, & Dove, 2017 ; Hughes et al., 2017 , 2018 ). In this situation, it is likely that as temperatures continue to rise species with lower thermal thresholds will more frequently be exposed to temperatures beyond their tolerance limits (especially during marine heat waves), potentially hindering adaption/acclimatization processes and favoring responses that range from sublethal effects to death and local extinction (Somero, 2010 ). The likely loss of such sensitive species would not only change the composition of benthic communities but also diminish the functions and services that they provide. However, if there is response diversity among functionally redundant organisms, the insurance hypothesis of biodiversity suggests that the overall ecosystem functionality may be stabilized through compensatory dynamics among species (Gonzalez & Loreau, 2009 ; Mori, Furukawa, & Sasaki, 2013 ; Yachi & Loreau, 1999 ). Exploring species‐specific thermal sensitivities among different components of benthic communities is, therefore, a key step toward forecasting the future composition and functionality of these communities in the face of climate change. However, while important efforts in this direction have been taken in shallow tropical reefs, thermotolerance analyses in temperate benthic communities largely lag behind (Kersting et al., 2015 ; Linares, Cebrian, Kipson, & Garrabou, 2013 ; Savva, Bennett, Roca, Jordà, & Marbà, 2018 ; Torrents, Tambuté, Caminiti, & Garrabou, 2008 ). In the Mediterranean, coralligenous assemblages are one of the most affected habitats by climate change. Coralligenous assemblages are biogenic formations built by the growth of crustose coralline algae and diverse calcareous macroinvertebrates at low irradiance levels and are characterized by their great structural complexity and species richness (harbouring ~10% of marine Mediterranean species) (Ballesteros, 2006 ). Most of the structural species of these habitats exhibit slow population dynamics and long life spans (+100 years; Garrabou & Harmelin, 2002 ; Linares, Doak, Coma, Diaz, & Zabala, 2007 ; Teixidó, Garrabou, & Harmelin, 2011 ); therefore, they are very sensitive to disturbances, including climate change (Balata, Piazzi, & Benedetti‐Cecchi, 2007 ; Ferrigno, Appolloni, Russo, & Sandulli, 2018 ; Garrabou et al., 2009 ; Montero‐Serra et al., 2015 ). In fact, more than 30 coralligenous species from different phyla and different structural roles have been affected in various mass mortality events (hereafter MMEs) associated with Mediterranean heat waves, suffering extensive tissue necrosis (partial and total mortality) and long‐term population declines (Cerrano et al., 2000 ; Crisci, Bensoussan, Romano, & Garrabou, 2011 ; Garrabou et al., 2009 ; Garrabou, Perez, Sartoretto, & Harmelin, 2001 ; Linares et al., 2005 ). Moreover, for some key habitat‐forming species, these population declines have been shown to potentially drive detrimental effects at the community level, such as the reduction of structural complexity and resilience (Linares et al., 2017 ; Ponti et al., 2014 ). However, while some species have been massively and recurrently affected during these warming events, other taxonomically and morpho‐functionally related organisms seem to remain unaffected, triggering the question of whether there could be different levels of thermal sensitivity within these communities in the context of climate change. This could have further implications for the future composition of these habitats and the loss (or maintenance) of the many associated functions and services they provide. In this study, we experimentally assessed the thermal response of 10 abundant, representative and widely distributed species from these communities that belong to different phyla and encompass contrasting growth forms. The main aim was to explore whether co‐occurring species of these highly diverse habitats differ in their thermal sensitivities, as field observations suggest, in view to discuss the implications of climate change on the composition and functioning of these key Mediterranean habitats. Our results contribute to filling the gap of thermotolerance data for coralligenous assemblages and suggest the presence of potential winners and losers in the face of ocean warming.",
"discussion": "4 DISCUSSION Forecasting temperature effects on ecological communities require a deep understanding of how temperature may influence the physiology of their different members. Thus, as oceans keep warming, community‐wide thermal sensitivity studies are becoming a powerful tool for reducing the uncertainty about the future composition, structure and functionality of marine communities facing climate change (Beveridge, Petchey, & Humphries, 2010 ; Fey & Cottingham, 2012 ; Iles, 2014 ; Savva et al., 2018 ; Stuart‐Smith, Edgar, Barrett, Kininmonth, & Bates, 2015 ). In this study, we explored the ranges of thermal sensitivity among structurally, functionally and taxonomically different components of Mediterranean coralligenous assemblages, showing contrasting responses to warming and suggesting the presence of potential “winners” and “losers” in the face of climate change. 4.1 Contrasting responses from experimental studies In our study, the experimental responses to thermal stress ranged from completely resistant species that did not suffer necrosis in any of the treatments (>21 days at 29°C) to sensitive species, which suffered necrosis after short‐term exposure (5–9 days) to the lowest‐temperature treatment of 26°C. In between these extremes, different levels of tolerance were found, including highly resistant species, for which 28°C represented the upper thermal limit that significantly reduced their probability of not suffering mortality, and intermediately tolerant species, which despite being affected at 26°C, did not suffer generalized necrosis until 1 week of exposure to 27°C. Bearing in mind that summer heat waves capable of sustaining temperatures between 26°C and 29°C for several days might become increasingly frequent during the next decades in diverse NW Mediterranean locations (Galli et al., 2017 ), our results suggest potential differences in climate change vulnerability among co‐occurring species dwelling in coralligenous assemblages. Indeed, differences between the members of the community were observed in our experiments to the extent that even the two morphotypes of Parazoanthus axinellae presented contrasting thermal responses. Regardless of the possible mechanisms behind these differences, which in the case of Parazoanthus axinellae could include the presence of highly bioactive secondary metabolites only in the “slender” morphotype as chemical defences induced for coping with environmental changes (Cachet et al., 2015 ; Reverter et al., 2016 ), our results represent a good example of the diversity of responses to warming found among structurally, functionally and taxonomically related organisms dwelling in coralligenous outcrops. Previous studies dealing with habitat‐forming emblematic species from these assemblages, such as the red gorgonian Paramuricea clavata (Risso, 1826), the white gorgonian Eunicella singularis (Esper, 1791), the red coral Corallium rubrum (Linnaeus, 1758) or the bryozoans Myriapora truncata (Pallas, 1766) and Pentapora fascialis (Pallas, 1766), have already pointed to such diversity (Crisci et al., 2017 ; Linares et al., 2013 ; Pagès‐Escolà et al., 2018 ; Torrents et al., 2008 ). However, the low number of studied species impeded the assessment of whether the response diversity was limited or widespread at the community level. Our results reinforce the latter possibility and suggest that regardless of their phyla and/or structural role, some species from these habitats could be living closer to their thermal limits than others and therefore might be more vulnerable under future warming scenarios. 4.2 Linking experimental and observational studies: evidence from MMEs The high variability of thermal responses observed in our study contributes to explaining why some coralligenous species were more affected than others in previous MMEs linked to warming (Cerrano et al., 2000 ; Garrabou et al., 2009 ; Linares et al., 2017 ; Perez et al., 2000 ). Moreover, the high (or low) resistance shown by the species in our thermal experiments was, in most cases, concomitant with the high (or low) vulnerability shown by these species in the past during these warming events (Table 1 ). For instance, species such as Axinella damicornis or Leptopsammia pruvoti that were highly resistant in our aquaria have never been reported as affected during previous warming‐induced MMEs that occurred in the NW Mediterranean Sea. In contrast, other species, such as Petrosia ficiformis, Crambe crambe, Alcyonium acaule or Parazoanthus axinellae, which have been greatly impacted during previous warming events (Cerrano et al., 2000 ; Cerrano, Magnino, Sarà, Bavestrello, & Gaino, 2001 ; Cerrano, Totti, Sponga, & Bavestrello, 2006 ; Garrabou et al., 2009 ; Linares et al., 2017 ; Parravicini et al., 2010 ; Perez et al., 2000 ), showed a higher vulnerability to thermal stress in our experiments. Thus, as we hypothesized, species‐specific thermal tolerances seem to play an important role in shaping divergent vulnerabilities in coralligenous species exposed to marine heat waves. Nonetheless, the unexpected differences in the responses among the aquaria thermotolerance experiments and observations in the field are also notable in some cases. In our experiment, Agelas oroides presented the highest resistance to thermal stress (> 21 days at 29°C) despite having sporadically been impacted during previous warming‐induced MMEs that were triggered at lower temperatures (Garrabou et al., 2009 ). Conversely, Dysidea avara was one of the most sensitive species in our experiment, while to our knowledge, no records of mass mortality linked to marine heat waves exist for this species in the NW Mediterranean. Such paradoxes have been noted in previous studies on mass mortality and bleaching events both in tropical and temperate species and have been attributed to the multifactorial nature of these events. A clear example of this is the Mediterranean coral Cladocora caespitosa (Ehrenberg, 1834), which, despite suffering recurrent warming‐induced mass mortalities in the field, showed resistance in single factor (temperature) experiments performed in aquaria while being impacted when exposed to additional factors such as the presence of invasive species (Kersting et al., 2015 ). Other factors that have been highlighted include food availability, pathogens, genetic differences or different physiologic processes (Arizmendi‐Mejía et al., 2015 ; Cebrian et al., 2011 ; Crisci et al., 2017 ; Linares et al., 2013 ; Pivotto et al., 2015 ). Therefore, bearing in mind the complex network of interacting factors that may ultimately determine vulnerability to warming in the field, determining the absolute thermal limits before which mortality of a given species should not be expected remains challenging. Our goal was, instead, to provide a ranking of thermal sensitivities among key components of coralligenous assemblages that could serve as a valuable baseline for better understanding the capacity of response and the trajectories of these species over broad temporal and spatial scales. Table 1 The thermal tolerances of the studied species confronted to observations from MMEs linked to warming in the NW Mediterranean Sea and reported in the scientific literature (1979–2017) Species Phylum Growth form Upper thermal limit in aquaria (°C) Resistance in aquaria Degree of damage in MMEs (NW Mediterranean) \n Agelas oroides \n Poriferan Massive >29°C (21 days) High * \n Cystodytes dellechiajei \n Tunicate Encrusting >27°C (21 days) High No reported damage \n Leptopsammia pruvoti \n Cnidarian Cup 28°C High No reported damage \n Axinella damicornis \n Poriferan Massive 28°C High No reported damage \n Axinella polypoides \n Poriferan Tree 27°C Medium No reported damage \n Alcyonium acaule \n Cnidarian Tree 27°C Medium * \n Crambe crambe \n Poriferan Encrusting 26°C Low *** \n Petrosia ficiformis \n Poriferan Massive 26°C Low *** \n Dysidea avara \n Poriferan Massive 26°C Low No reported damage \n Parazoanthus axinellae \n Cnidarian Encrusting 27°C “slender” Medium *** 26°C “stocky” Low The species have been ordered from the most to the least resistant according to their upper thermal limits obtained in aquaria (considered here as the first temperature significantly reducing their probability of remaining healthy without necrosis throughout the experimental period). No reported damage indicates that a given species has not been reported as being affected in any MME, whereas (*) refers to species that have been reported as being affected only in one MME and (***) refers to species that have been affected in multiple MMEs (>5 years and/ or locations). See Supporting information & Table S2 for further detail and references. John Wiley & Sons, Ltd 4.3 Response diversity in coralligenous assemblages facing climate change: consequences for ecosystem structure and functioning According to the insurance hypothesis of biodiversity (Gonzalez & Loreau, 2009 ; Yachi & Loreau, 1999 ), a high “response diversity” among functionally redundant organisms is essential for buffering the effects of environmental changes and ensuring the ecosystem functionality that prevents regime shifts (Mori et al., 2013 ). In the highly diverse coralligenous outcrops, the diversity of responses shown by the different studied species (including different phyla and different structural roles) suggests that rather than suffering dramatic shifts in response to climate change, many of these assemblages could be susceptible of changing their configuration in the future to less diverse but functionally similar systems, where thermally sensitive species might be replaced by more resistant species. However, the final outcomes will not depend only on species‐specific thermal tolerances. The “winners” and “losers” will also depend on the great variety of specific life histories and functional traits they present, and how these traits favor, or impair their success in a changing sea (Darling, Alvarez‐Filip, Oliver, McClanahan, & Côté, 2012 ; Hughes et al., 2018 ; Madin et al., 2016 ; Van Woesik, Sakai, Ganase, & Loya, 2011 ). For instance, populations of species that present faster growth and dispersal or higher reproduction may recover faster after disturbances and/or adapt more easily to rapid environmental changes over generations than populations of species with traits related to a close adaptation to their current environment, extremely slow dynamics or a low dispersal capacity (McKinney & Lockwood, 1999 ). Likewise, changes in ecological interactions may also determine the final result, as species might be ultimately favored or disfavored by the tolerant or sensitive response of others with which they are ecologically connected (Walther, 2010 ). For instance, the loss of key habitat‐forming species (such as some sensitive gorgonians) could potentially trigger cascading effects on the community that could result in an overall reduction in structural complexity and resilience (Ponti et al., 2014 ). Similarly, a decline in some massive slow‐growing sponges, such Petrosia ficiformis , could exert a major influence on the overall ecosystem functioning and stability (Bell, 2008 ). Therefore, whether the eventual “winners” will be able to replace the “losers” in such a way that the complexity and functioning of the coralligenous assemblages are maintained in the future warmed Mediterranean Sea remains to be seen. The contrasting responses to warming among the different components of these assemblages unravelled in this study indicate some promising capacity to buffer future warming effects. However, steering coralligenous outcrops in a way that their functionality is safeguarded in the face of climate change will be challenging and will necessarily depend upon an extensive understanding of the interplay between the functional and life history traits of coralligenous key species, their ecological interactions and their species‐specific vulnerabilities to climatic disturbances."
} | 4,973 |
24568605 | PMC3979089 | pmc | 1,959 | {
"abstract": "Microbial fuel cells (MFCs) are a\npromising technology for energy-efficient\ndomestic wastewater treatment, but the effluent quality has typically\nnot been sufficient for discharge without further treatment. A two-stage\nlaboratory-scale combined treatment process, consisting of microbial\nfuel cells and an anaerobic fluidized bed membrane bioreactor (MFC-AFMBR),\nwas examined here to produce high quality effluent with minimal energy\ndemands. The combined system was operated continuously for 50 days\nat room temperature (∼25 °C) with domestic wastewater\nhaving a total chemical oxygen demand (tCOD) of 210 ± 11 mg/L.\nAt a combined hydraulic retention time (HRT) for both processes of\n9 h, the effluent tCOD was reduced to 16 ± 3 mg/L (92.5% removal),\nand there was nearly complete removal of total suspended solids (TSS;\nfrom 45 ± 10 mg/L to <1 mg/L). The AFMBR was operated at a\nconstant high permeate flux of 16 L/m 2 /h over 50 days,\nwithout the need or use of any membrane cleaning or backwashing. Total\nelectrical energy required for the operation of the MFC-AFMBR system\nwas 0.0186 kWh/m 3 , which was slightly less than the electrical\nenergy produced by the MFCs (0.0197 kWh/m 3 ). The energy\nin the methane produced in the AFMBR was comparatively negligible\n(0.005 kWh/m 3 ). These results show that a combined MFC-AFMBR\nsystem could be used to effectively treat domestic primary effluent\nat ambient temperatures, producing high effluent quality with low\nenergy requirements.",
"introduction": "Introduction Growing concerns over\nthe large energy requirements needed for\neffective wastewater treatment has stimulated interest in the use\nof wastewater as a source of renewable energy. 1 Microbial fuel cells (MFCs) are being developed as a sustainable\nenergy technology, as they can directly produce electricity from wastewater\nallowing for energy recovery to offset the costs of wastewater treatment. 2 , 3 In an air-cathode MFC, organic matter in wastewater is oxidized\nby microorganisms, and electrons discharged to the anode travel through\nan external circuit to the cathode where they combine with oxygen,\nforming water. 4 , 5 Passive transfer of oxygen to\nthe air-cathode avoids the need for energy intensive aeration of the\nwastewater that is currently required for typical activated sludge\nor aerobic membrane bioreactor processes. In addition, MFCs have lower\nsludge production than conventional aerobic treatment processes, which\ncould reduce treatment costs and the challenges associated with sludge\ntreatment and disposal. 6 MFCs fed\nwith domestic wastewaters have shown promising performance\nin terms of achieving electricity generation with simultaneous organics\nremoval, 7 − 9 and there continue to be improvements in MFC designs\nthat have produced configurations more suitable for scaling up to\nlarger systems. 10 − 14 Capital costs of the materials used in MFCs are also being reduced,\nfor example, by using cathode catalysts such as inexpensive activated\ncarbon. 15 , 16 One operational aspect of using MFCs for\nwastewater treatment that has not been sufficiently addressed is the\nneed to meet stringent effluent quality requirements. Effluent chemical\noxygen demand (COD) concentrations with domestic wastewater in MFCs\nhave ranged from 23 to 164 mg/L in fed-batch tests, and 60 to 220\nmg/L in continuous flow tests, depending on influent COD concentrations,\nreactor configurations, and cycle time or hydraulic retention time\n(HRT). 8 , 11 , 14 One of the\nreasons for these high effluent CODs is likely inefficient removal\nof particulate organics, 17 as biofilm reactors\nsuch as MFCs and trickling filters are more effective for soluble\nthan particulate COD removal. Thus, post-treatment or integrated processes\nare needed to further improve the quality of the treated wastewater\nto meet discharge limits. One approach to improve the overall\nextent of wastewater treatment\nhas been to integrate the MFC with a membrane-based process in a single\nreactor. This approach has been referred to either as a membrane bioelectrochemical\nreactor (MBER) 18 or an electrochemical\nmembrane bioreactor (EMBR). 19 Although\nhigher treatment efficiencies have been obtained for both acetate\nsolutions and domestic wastewater in tests with this approach, energy\nconsumption has only been balanced with electrical energy production\nwhen acetate was used as the substrate. 18 , 19 The main challenges\nwith using both MFCs and membrane processes for domestic wastewater\ntreatment are obtaining high power production from the MFCs, while\nminimizing membrane fouling. 18 Using a\nshorter hydraulic retention time (HRT) in an MFC treating domestic\nwastewater will usually improve power production, 14 but a shorter HRT could mean a higher organic loading rate\non the membrane process, which could result in increased membrane\nfouling. 18 Membrane fouling control remains\nthe biggest challenge in the use of membranes in both aerobic 20 and anaerobic systems. 21 In previous membrane-based MFC studies, membranes inside the MFCs\nfouled in 15 days, and therefore these membranes would require frequent\ncleaning. 18 The high maintenance costs\ndue to cleaning processes could limit applications of integrated MFC\nand membrane bioreactor processes. 18 A new approach to obtain high quality effluent with low energy\nrequirements is proposed here based on using a second stage anaerobic\nfluidized bed membrane bioreactor (AFMBR) following wastewater treatment\nin the MFC. The AFMBR has recently been shown to be an effective approach\nfor achieving high quality effluent when used as a post-treatment\nmethod for an anaerobic fluidized bioreactor (AFBR). 1 , 22 In the AFMBR, membrane fouling is controlled by using granular activated\ncarbon (GAC) as the fluidized particles, as these particles can scour\nthe membrane and minimize fouling. 1 , 22 The properties\nof particles used in the fluidized bed are important, as spherical\nplastic particles have been shown to not be as effective as GAC. 23 The use of an MFC as the primary treatment process,\nas opposed to an AFBR, may be useful for several reasons. Electrical\nenergy is directly produced in the MFC, whereas in the AFBR electricity\nwould have to be produced in a separate process from biogas that might\nneed to be cleaned and purified to remove hydrogen sulfide and water\nto improve utilization efficiencies. 24 Any\nhydrogen sulfide generated in situ in an MFC would be expected to\nbe rapidly oxidized in the MFC as it is a good electron donor to the\nanode. 25 There should be very little methane\nin the MFC effluent compared to that produced by the AFBR, as organic\nmatter is mainly converted into current or lost to aerobic degradation\ndue to oxygen transfer across the cathode. It is important to remove\ndissolved methane, which can be supersaturated in these systems, to\nminimize its release into the atmosphere as it is a potent greenhouse\ngas. 26 , 27 In this study we examined domestic\nwastewater treatment using a\ntwo-stage MFC-AFMBR system, containing four MFCs and one AFMBR, at\nambient temperature. There were two separate flow lines into the AFMBR,\nwith two MFCs connected hydraulically in series (with separate electrical\ncircuits) in each flow line (Figure 1 ). The\nuse of two MFCs in series avoided large changes in COD concentrations\nin each MFC, as such large COD changes have previously been shown\nto adversely affect current generation. 14 , 28 Each pair\nof MFCs had a different electrode configuration in order to compare\ntwo design approaches: using a separator electrode assembly (SEA),\nwhere the electrodes are sandwiched together and a separator was placed\nbetween them to prevent short circuiting and reduce oxygen crossover\nfrom the cathode; and using a spaced electrode assembly (SPA), where\nthe electrodes are kept close to each other, but with sufficient space\nto avoid direct contact (no separator was used) (Figure S1, Supporting Information (SI) ). It has recently\nbeen shown that the SPA design can reduce treatment time compared\nto the SEA, although less energy may be recovered in the SPA configuration\ndue to the loss of organic matter to aerobic processes rather than\ncurrent generation. 29 Treatment efficiency\nwas evaluated in terms of COD and total suspended solid (TSS) removals,\nand energy efficiency was quantified for both processes in terms of\nproduction and demands, under continuous flow conditions. Figure 1 Schematic diagram\n(a) and photo (b) of the two-stage MFC-AFMBR\nsystem. ( U = the first upstream MFC, and D = the second downstream MFC prior to the AFMBR).",
"discussion": "Results and Discussion Performance of MFCs The start-up time needed for the\nSEA MFCs was shorter than that required for the SPAs. The SEA MFCs\nproduced a stable voltage of 0.59 ± 0.03 V (1000 Ω) after\n3 days, while the SPAs produced 0.51 ± 0.04 V after 3 days, and\nrequired 10 days to achieve a stable voltage of 0.58 ± 0.01 V.\nStable voltage production was indicated by a deviation between the\ndaily averaged voltage values that was <0.006 V (∼1% of\nthe daily averaged voltage) over three consecutive days. There was\nno appreciable difference in start-up time between the upstream or\ndownstream MFC within the individual flow paths (data not shown). The power produced by the SEAs and SPAs changed over time. Based\non the polarization data obtained after 1 month, the SEA-U MFC produced\na maximum power of 0.31 mW (89 mW/m 2 , normalized to the\ncathode projected surface area of 35 cm 2 ), which was comparable\nto that of SPA-U (0.33 mW) (Figure 2 ). Although\nthe same current was produced with these two configurations, the SEA-U\nhad better cathode performance but showed poorer anode performance\nthan the SPA-U (Figure 3 b and c). The downstream\nMFCs produced slightly lower maximum power than the upstream ones,\nwith 0.28 mW for SEA-D and 0.27 mW for SPA-D (Figure 2 ). The downstream MFCs generally had more positive anode potentials\nthan the first MFCs (Figure 3 b), likely due\nto the lower substrate concentrations in the downstream MFCs ( SI Table S1), as it was shown that the anode\npotentials became more positive at lower substrate concentrations\nin a previous study. 35 Figure 2 Power production of the\nSEA and SPA MFCs at different time after\nstart-up, after (a) 1 month and (b) 5 months. ( U =\nthe first upstream MFC, and D = the second downstream\nMFC prior to the AFMBR). Figure 3 Voltage, anode potential and cathode potential of the SEA and SPA\nMFCs at different time after start-up: (a) voltage, (b) anode potential\nand (c) cathode potential at 1 month, and (d) voltage, (e) anode potential\nand (f) cathode potential at 5 month. The letters “A”\nin (b) indicated the anodes, and “C” in (c) the cathodes.\nAll electrode potentials were reported versus the Ag/AgCl reference\nelectrode [+200 mV vs a standard hydrogen electrode (SHE); BASi]. After 5 months, the maximum power\ndensities of the SEA MFCs were\nrelatively unchanged (0.33 mW for SEA-U and 0.32 mW for SEA-D), and\nthe wastewater composition fed into the reactor was relatively unchanged\nbased on the influent tCOD concentrations (210 ± 11 mg/L at 5\nmonths, compared to 224 ± 17 mg/L at 1 month). However, the maximum\npower produced by SPA MFCs substantially decreased to 0.16 mW (SPA-U)\nand 0.18 mW (SPA-D). The reason for these decreases was a large reduction\nin cathode potentials (Figure 3 f), which was\nlikely due to biofouling. 15 , 36 While the cathodes\nused for the SEA configuration contained a separator that covered\nthe cathode, the SPA cathodes were directly exposed to the wastewater,\nand thus they were more prone to fouling (Figure 3 f). The maximum power density of 89 ± 6 mW/m 2 produced\nby the SEA MFC in these continuous flow tests was lower than the maximum\npower densities obtained in two other studies with domestic wastewater\nwhen the MFC was operated in fed-batch mode (120 mW/m 2 14 or 328 ± 11 mW/m 2 29 ). The lower power density here was likely due\nto a lower influent COD (217 ± 18 mg/L, compared to 275 ±\n71 mg/L 14 and 303 ± 69 mg/L 29 ), and operation under continuous flow conditions,\nwhere the average substrate concentration was lower than that in the\nfed-batch reactors at the beginning of the cycle. 28 The different electrode configurations (SEA or SPA)\ndid not appreciably\naffect the extent of COD removal. tCOD removals were 28 ± 7%\nfor SEA-U, and 34 ± 3% for SPA-U, which are comparable removals\nwithin the calculated standard deviations. The downstream MFCs had\nslightly lower tCOD removals than the upstream MFCs, with 17 ±\n5% for SEA-D and 19 ± 5% for SPA-D, likely due to the lower substrate\nconcentrations. Fed-batch tests with domestic wastewater have shown\nthat COD removal in MFCs is first order with respect to concentration\n(unpublished data). Thus, the reduction in COD concentration would\nhave reduced removal rates in the downstream reactors. sCOD removals\nshowed the same trends as tCOD, with greater removals in the upstream\nreactors (27 ± 10% for SEA-U, 32 ± 5% for SPA-U) than the\ndownstream ones (19 ± 5% for SEA-D, and 26 ± 7% for SPA-D).\nNote that these COD removals were based on the combination of all\ndata in tests at the different external resistances used in the polarization\ntests, as the effluent COD concentrations in these tests did not change\nsubstantially with the different external resistances (COD concentrations\nin the SI , Table S1). These COD removals\nwere lower than those obtained in previous studies operated in fed-batch\nmode using the same domestic wastewater source (62–93%), due\nto the short HRT (4 h) in this study compared to much longer fed-batch\ncycle times (12–36 h). 14 , 29 In additional tests,\nincreasing the HRT to 24 h increased COD removals to 67 ± 2%,\nwhich was about the same as that obtained in fed-batch mode with a\ncycle length of 24 h (65 ± 1%; 500 Ω resistance, data not\nshown). However, a long HRT is not desirable for efficient wastewater\ntreatment, and thus the shorter HRT was used here. The CEs increased\nin proportion to the current (lower resistance),\neven though the COD removals remained relatively constant with different\nresistances. At the maximum power density (0.31 ± 0.02 mW, 0.86\n± 0.02 mA), the overall CE of the SEA MFCs was 18% (13% for SEA-U\nand 28% for SEA-D). Over time, changes in the CEs paralleled those\non maximum power densities. The SEA and SPA MFCs had comparable overall\nCEs at 1 month, (range of 6–29%), but after 5 months the SEA\nMFCs remained relatively unchanged while those of the SPA MFCs decreased\n(range of 4–20%) with the decreased currents. CE values obtained\nhere under continuous flow conditions were comparable to those previously\nreported for fed-batch conditions (2–31%). 29 Overall, these results suggest that the SEA configuration\nwas superior to the SPA design on the basis of fixed HRTs as it maintained\nhigher power densities and CEs over time with the same level of wastewater\ntreatment. Wastewater Treatment with the Second Stage\nAFMBR The\ntwo-stage MFC-AFMBR system achieved excellent treatment levels in\nterms of COD and TSS removals. The AFMBR was first inoculated with\nanaerobic sludge and fed the effluent from MFCs for two weeks. After\nthat, the membrane module was installed, and preliminary tests were\nconducted to optimize the design and operation of the AFMBR over a\nperiod of approximately two months, with occasional system shutdown\nto address problems related to consistent flow and treatment. Following\nthis system optimization, a new membrane module was installed, and\nthe MFC-AFMBR system was operated continuously for 50 days in concert\nwith the MFCs operation over the last two of the five months. tCOD further decreased from the influent concentration to the AFMBR\nof 107 ± 10 mg/L to 16 ± 3 mg/L in the effluent, providing\nan overall tCOD removal for the two stages of 92.5% (49.1% for the\nMFCs, and 43.4% for the AFMBR) (Figure 4 and SI Table S2). The effluent sCOD and tCOD concentrations\nfrom the AFMBR were identical, and therefore there was a lower overall\nsCOD removal of 86.2% (influent sCOD to MFCs of 114 ± 8 mg/L,\ncompared to tCOD of 210 ± 11 mg/L) (Figure 4 and SI Table S2). A larger percent of\nsCOD was removed by the MFCs (50.3%) than by the AFMBR (35.9%), while\nparticulate COD removal was 47.9% for the MFCs compared to 100% for\nthe AFMBR. Changes in the forms of COD might occur through hydrolysis\nfrom particulate to soluble COD, and through biomass growth from soluble\nto particulate. The particulate COD removal in MFCs might be partially\ndue to settling in the reactor chambers, while that in the AFMBR was\nprimarily due to membrane filtration. The effluent contained <1\nmg/L of TSS due to filtration of the wastewater through the membrane,\nresulting in >99.6% TSS removal (Figure 4 and SI Table S2). These COD and TSS removals\nare\ncomparable to those obtained using a staged anaerobic fluidized membrane\nbioreactor (SAF-MBR) treating domestic wastewater. 22 There was little overall change in pH, as the influent\npH to the MFCs of 7.6 ± 0.1 decreased to 7.1 ± 0.1 in the\nMFCs effluent, but it increased to 7.5 ± 0.2 following treatment\nin the AFMBR ( SI Table S2). These pH changes\nmight result from losses of CO 2 and volatile fatty acids,\nfor example from methanogenesis processes occurring in the MFC-AFMBR\nsystem. Also there were no large changes in conductivity, with 1473\n± 33 mS/cm for the MFCs influent, 1457 ± 15 mS/cm for the\nMFCs effluent, and 1420 ± 19 mS/cm for the AFMBR effluent ( SI Table S2). Figure 4 Influent and effluent concentrations,\nand removals of tCOD, sCOD\nand TSS for the combined MFC-AFMBR system. The values inside the figures\nwere the percent of the influent concentration that was removed by\nthe MFCs, AFMBR, and the whole system. The AFMBR was operated continuously for 50 days at a high\nmembrane\nflux of 16 LMH, even without cleaning by backwashing or using chemicals.\nMost of the increase in the TMP, from 0.015 to 0.035 bar, occurred\nduring the first 8 days of operation (Figure 5 ). Thereafter, it slowly increased to 0.050 bar during the rest of\nthe test (Figure 5 ). Liquid (9 mL) was withdrawn\nfrom the AFMBR every 3.5 days (0.16% of the total influent flow) to\nremoval finer material and excess suspended solids from days 8 to\n50, as suggested in a previous AFMBR study. 1 The membrane flux of 16 LMH here is higher than that previously\nreported for the AFMBR following an AFBR (11 LMH), with a PVDF hollow-fiber\nmembrane with the same pore size as the one here (0.1 μm). 22 In that study, the TMP reached 0.25 bar in 3\ndays when the membrane flux was increased to 14 LMH, 22 which is much higher than the maximum TMP observed here.\nThe stable operation of the flux through the AFMBR without appreciable\nmembrane fouling was likely due to a combination of factors here that\nincluded the scouring effect of the GAC particles on the membrane\nsurface, intermittent filtration, and periodic removal of suspended\nsolids. The use of MFCs as the primary treatment process likely contributed\nto the high flux and stable performance of the AFMBR, due to the removals\nof COD and TSS in the MFCs. The improved flux with the first-stage\nMFC treatment, compared to that previously obtained with a first-stage\nAFBR treatment, suggests that MFCs might be a better first stage treatment\nthan AFBR, but this cannot be concluded without direct side-by-side\ntests of the two different systems. The operation of the AFMBR without\nwastewater pretreatment was not examined here as that would represent\na different treatment process, and one that would not allow for electrical\npower generation and recovery from COD removal. Figure 5 TMP for the AFMBR over\n50 days of operation. Energy Balance Energy usage for the two-stage MFC-AFMBR\nsystem was calculated as previously described. 1 , 22 All\nthe volumetric energy densities were reported on the basis of normalizing\nto 1 m 3 of wastewater treated. The energy requirements\nwere calculated as 0.0107 kWh/m 3 for fluidizing the GAC\nparticles, and 0.0014 kWh/m 3 for pumping permeate through\nthe membranes, resulting in a total electrical energy requirement\nfor the AFMBR of 0.0186 kWh/m 3 (Table 1 ). The electrical energy requirement for pumping liquid through\nthe MFCs was negligible compared to that needed for the AFMBR (Table 1 ). The higher energy requirement for the AFMBR was\nprimarily due to pumping needed for liquid recirculation to maintain\nthe GAC fluidization. The energy needed for this is proportional to\nthe total reactor flow rate and the hydraulic head loss of the system. 37 In an AFMBR reactor with a given configuration,\nthe minimum recirculation flow rate and the hydraulic head loss are\nfixed, and thus the energy requirement for recirculation is inversely\nproportional to the permeate flow rate or the HRT. 18 Therefore, the high permeate flux and low HRT achieved\nfor the AFMBR in this study were favorable for achieving a low energy\nrequirement of 0.0186 kWh/m 3 . This energy requirement is\nlower than previous reports using the AFMBR (0.027–0.040 kWh/m 3 ), 1 , 18 , 22 but there\nare many differences in these studies that preclude a direct comparison\nof these values. Table 1 Electrical Energy Requirements and\nProduction for the Two-Stage MFC-AFMBR System characteristic MFCs AFMBR system total Electrical\nEnergy required Energy for Hydraulic Loss reactor head loss, cm\nH 2 O 0.5 2.5 reactor\ninfluent plus recirculation flow rate, mL/min 1.1 171.1 hydraulic\nenergy requirement, kW a 0.001(10 –6 ) 0.699(10 –6 ) required pumping energy, kWh/m 3 b 0.00001 0.0107 0.0107 Energy for Permeate Extraction average TMP, cm H 2 O 50.8 permeate flow rate, mL/min 1.1 permeate energy requirement, kW 0.090(10 –6 ) required pumping energy,\nkWh/m 3 0.0014 total pumping energy required for system, kWh/m 3 0.00001 0.0121 0.0121 total electrical\nenergy required for pumps, kWh/m 3 c 0.000015 0.0186 0.0186 Electrical Energy Produced MFC maximum power, mW d 1.28 electrical energy production, kWh/m 3 0.0197 0.0197 electrical energy produced/required e 1.06 a Energy requirement =9.8 QE , where Q (m 3 /s) is flow rate and E (m\nH 2 O) is head loss. 1 b Energy per m 3 of wastewater\ntreated. c Assume energy efficiency\nof 65%\nin conversion of electrical energy to pump energy. 1 d Based on the\nmaximum power produced\nby the SEA MFCs in series. This maximum power output was quite similar\nto that obtained during steady operation, and therefore it represents\npower production that could be obtained during continuous treatment\ntests ( SI Figure S3). e The ratio of the electrical energy\nproduced to that required by the MFC-AFMBR system. Electrical energy could be produced\ndirectly from the MFCs. Based\non the maximum power that could be produced by the SEA configuration\nafter 5 months of operation (0.33 mW for SEA-U and 0.32 mW for SEA-D,\nFigure 2 ), the total power that could be produced\nby the four MFCs coupled to the AFMBR was 1.28 mW (four times 0.32\nmW) (Table 1 ). If all of this electrical energy\nwas recovered, the net electrical energy available for the system\noperation would be 0.0197 kWh/m 3 . This would be enough\nto supply the 0.0186 kWh/m 3 required to operate the system\nif these values could be maintained for larger-scale systems. However,\nthis energy balance would likely change as the size of the system\nincreases. Also, in practice there might be other energy losses that\nwould affect overall energy recovery, that have not been included\nhere. Direct electricity production by MFCs is advantageous, compared\nto methanogenic reactors that require combustion of the methane to\nproduce power, as the conversion efficiency of methane to electricity\nis typically only 33%. 1 However, there\nwill also be energy losses in converting the low voltage DC power\ninto higher voltage DC or AC power. 38 , 39 The\nadditional energy that could be recovered from the methane\nproduction in the AFMBR was not included in this energy balance as\nit would have been difficult to recover. Total methane production\nin the AFMBR was 1.67 mL/L liquid treated at ambient temperature and\npressure, with most of this present as dissolved methane (1.5 mL/L)\n(detailed calculations in the SI ). This\nconcentration was estimated to be 125% oversaturation relative to\nthe concentration of methane in the AFMBR headspace. The energy value\nof this amount of methane is 0.016 kWh/m 3 (methane combustion,\nassuming 800 kJ/mol), equivalent to electrical energy of 0.005 kWh/m 3 (33% energy recovery) which could theoretically add 27% more\nenergy production into the system. However, as most of this methane\nis dissolved, and at a low concentration, it would have been difficult\nto recover. The methane yield from the AFMBR was 0.75 mmol/g COD removed,\nindicating only 5% of the COD removal in the AFMBR could be attributed\nto methane generation. The amount of methane that theoretically could\nbe produced in the AFMBR was estimated as 17 mL/L, based on the COD\nremoval and assuming a conversion of 0.017 mol CH 4 /g COD\n(detailed information and calculations in the SI ). It is not clear what other processes occurred relative\nto COD removal as the methane production was only about 1 / 10 of that possible by methanogenesis alone. This subject\nwill require further study. More methane might be recovered in the\nfuture with improvements in the configuration and operation of the\nAFMBR. Outlook The two-stage MFC-AFMBR system was shown to\neffectively treat domestic wastewater (primary clarifier effluent)\nat ambient temperatures in terms of COD and TSS removals, producing\na high effluent quality with a near neutral (or net positive) energy\nrequirement, without the need for membrane cleaning even after 50\ndays of operation. tCOD was reduced from 210 ± 11 mg/L to 16\n± 3 mg/L, resulting in 92.5% overall COD removal with >99%\nremoval\nof TSS to a final effluent concentration of <1 mg/L. The energy\nrequirement of the AFMBR is much less than that needed for aerobic\nMBRs with internal membranes that require air sparging to control\nmembrane fouling. 1 The high permeate flux\n(16 LMH) and low HRT (1 h) here resulted in an overall low energy\nrequirement for the AFMBR of only 0.0186 kWh/m 3 . Thus,\nthe energy produced by only the MFCs (0.0197 kWh/m 3 ) was\ntheoretically sufficient here to meet the energy demands for the system\noperation, although the energy balance for a larger system would likely\nchange. An additional benefit of the MFC-AFMBR system should be a\nlow sludge production rate. Although sludge production was not measured\nhere (for an estimate, see the SI ), previous\nstudies have shown that the sludge production by anaerobic MFC and\nAFMBR processes are less than those of conventional aerobic processes\nsuch as activated sludge. 6 , 22 While feasibility\nof the combined process was shown here, additional work will be needed\nto optimize the performance of the individual and combined MFC and\nAFMBR reactors. More optimal treatment could likely be obtained by\nadjusting the HRTs of the two systems, and examining benefits of treatment\nrates compared to electrical power production. The current findings\nwere sufficient to show that the two-stage MFC-AFMBR system is useful\nfor treatment of low strength wastewater even at ambient temperatures.\nThe robustness of the system at other temperatures, particularly lower\nones, would need to be investigated. However, MFCs have been shown\nto function over a wide range of temperatures. 2 , 40 The\nassessment of nitrogen, phosphorus, and pathogen removals will be\nexamined in future studies to see to what extent the MFC-AFMBR system\ncan be used for these wastewater components. Other issues to address\nare capture and use(or destruction) of the methane to avoid its release\ninto the air, and efficient use of the electricity produced by MFCs.\nFollowing these optimization studies, it should be possible to better\nevaluate the economics of the system compared to traditional treatment\nsystems."
} | 6,998 |
36094206 | PMC9499035 | pmc | 1,961 | {
"abstract": "ABSTRACT Microorganisms in nature form multicellular groups called biofilms. In biofilms, bacteria embedded in the extracellular matrix (ECM) interact intensely due to their proximity. Most studies have investigated genetically homogeneous biofilms, leaving a gap in knowledge on genetically heterogeneous biofilms. Recent insights show that a Gram-positive model bacterium, Bacillus subtilis , discriminates between strains of high (kin) and low (nonkin) genetic similarity, reflected in merging (kin) and boundaries (nonkin) between swarms. However, it is unclear how kinship between interacting strains affects their fitness, the genotype assortment, and incorporation of the mutant lacking the main structural ECM polysaccharide (EpsA-O) into floating biofilms (pellicles). We cultivated Bacillus subtilis strains as mixtures of isogenic, kin, and nonkin strain combinations in the biofilm-promoting minimal medium under static conditions, allowing them to form pellicles. We show that in nonkin pellicles, the dominant strain strongly reduced the frequency of the other strain. Segregation of nonkin mixtures in pellicles increased and invasion of nonkin EpsA-O-deficient mutants into pellicles decreased compared to kin and isogenic floating biofilms. Kin and isogenic strains had comparable relative frequencies in pellicles and showed more homogenous cell mixing. Overall, our results emphasize kin discrimination as a social behavior that shapes strain distribution, spatial segregation, and ECM mutant ability to incorporate into genetically heterogenous biofilms of B. subtilis . IMPORTANCE Biofilm communities have beneficial and harmful effects on human societies in natural, medical, and industrial environments. Bacillus subtilis is a biotechnologically important bacterium that serves as a model for studying biofilms. Recent studies have shown that this species engages in kin discriminatory behavior during swarming, which may have implications for community assembly, thus being of fundamental importance. Effects of kin discrimination on fitness, genotype segregation, and success of extracellular matrix (ECM) polysaccharide (EpsA-O) mutant invasion into biofilms are not well understood. We provide evidence that kin discrimination depends on the antagonism of the dominant strain against nonkin by using environmental strains with determined kin types and integrated fluorescent reporters. Moreover, this antagonism has important implications for genotype segregation and for when the bacteria are mixed with ECM producers. The work advances the understanding of kin-discrimination-dependent bacterial sociality in biofilms and its role in the assembly of multicellular groups.",
"introduction": "INTRODUCTION Biofilms are multicellular assemblages of bacteria that engage in intense social interactions that bring about positive (cooperative) or negative (antagonistic) fitness effects ( 1 ). Biofilms in natural habitats are the most dominant forms of microbial life and are predominantly composed of genetically diverse microorganisms ( 2 – 4 ). Nevertheless, most studies have focused on genetically homogenous biofilms. The hallmark of biofilms is a matrix of self-generated extracellular polymeric substances (EPS) that glue cells together, mediate surface attachment, provide stability, and improve survival in harsh environments ( 5 ). Extracellular matrix (ECM) polymers serve as “public goods” because they are released and utilized by cells joined in biofilm collectives ( 6 ). However, we have limited understanding of ECM exploitation in genetically heterogeneous biofilms. According to the work of Hamilton ( 7 , 8 ), organisms apply kin discrimination to match their behaviors toward others according to their genetic similarity. Moreover, kin discrimination may limit the possibility of less similar genotypes exploiting costly public goods during group living, due to cells’ sorting, avoidance, or antagonism between nonkin ( 9 , 10 ). Studies on bacterial kin discrimination have mostly employed cooperative swarming as the test behavior to determine phenotypic responses of bacterial strains at the point of encounter, where a visible boundary between two strains was found to indicate nonkin interactions. In contrast, merging was found to be more common between swarms of very close kinship or between isogenic swarms ( 11 – 16 ). Recently, we showed that the nonpathogenic, spore-forming, Gram-positive soil bacterium Bacillus subtilis engages in kin discrimination during swarming, where the frequency of the boundary appearance or merging at the meeting point of two swarms correlates with genetic similarity. Genetic similarity between interacting strains was defined by comparing the identities of four housekeeping genes ( 13 ) and by average nucleotide identity (ANI) ( 17 , 18 ) of orthologous gene pairs shared between two microbial genomes. Specifically, B. subtilis kin strains with 99.93 to 99.99% ANI and isogenic strains (“self” pairings, 100% ANI) exhibited merging behavior, whereas less genetically similar strains (98.73 to 98.84% ANI), which we also refer to as nonkin strains, formed clear boundaries ( 18 ). Moreover, Lyons et al. showed that less genetically similar B. subtilis strains also differ in a combination of kin discrimination loci, which comprise genes for antimicrobials, contact-dependent toxin-immunity pairs, and the enzymes involved in the synthesis of major ECM polysaccharide EpsA-O ( 16 ). However, we lack information on how kin discrimination affects fitness and cell sorting of different genotypes in biofilms ( 19 ). B. subtilis is a model organism that is often used to investigate biofilm development on plant roots ( 13 , 20 , 21 ), on agar surfaces ( 22 ), or at the air-liquid (A-L) interface, where B. subtilis constructs intricate pellicles ( 23 – 26 ). Pellicle formation is dependent on biofilm ECM polysaccharide EpsA-O and the TasA amyloid protein fibers, anchored through TapA to peptidoglycan ( 24 , 27 – 29 ). B. subtilis ECM mutants form poor pellicles alone, but coculturing of the tasA and epsA - epsO operon (Δ epsA–O ) mutants that are isogenic in other loci compensates for the defect in pellicle formation ( 27 , 28 , 30 , 31 ). It is not known whether Δ epsA–O mutants can incorporate into pellicles when mixed with the phylogenetically more distant (nonkin) ECM-positive strains. We here test the hypothesis that low kinship will induce spatial segregation and will change the relative cell frequency of interacting B. subtilis strains in genetically mixed biofilms. Moreover, we predict that nonkin interactions between B. subtilis strains will limit the invasion of the Δ epsA–O mutant into the pellicle when cocultured with an ECM producer. To test these predictions, we use selected natural B. subtilis strains isolated from one 1-cm 3 soil sample ( 32 ) with previously determined genetic similarity and kin discrimination phenotype ( 13 , 16 , 18 ). We show that two nonkin B. subtilis strains behave differently from kin or isogenic mixtures in pellicles. In nonkin pellicles, the dominant strain decreases the fitness of the partner strain, whereas this dominance is not detected in kin and isogenic pellicles. Also, in nonkin pellicles, strain segregation is more prominent than in kin and isogenic pellicles, where two strains remain more intermixed. Moreover, nonkin interactions also restrict the incorporation of the Δ epsA–O mutant into pellicle when mixed with the ECM producer. The findings of this work make an important contribution to our understanding of how kin discrimination shapes bacterial biofilms, which is relevant for biofilm control and application.",
"discussion": "DISCUSSION Biofilms are composed of ECM-embedded bacteria that are capable of various social interactions that might affect fitness, cell assortment, and ECM exploitation and consequently structure and ecology of microbial communities ( 9 ). By using a subset of B. subtilis strains from micrometer soil aggregates ( 32 ) and with previously determined kin types ( 13 ), we here show strong interference competition between nonkin and the coexistence of kin and isogenic strains in pellicles. Nonkin strains also segregate into patches of different size and do not form well-mixed pellicles with EpsA-O-deficient strains as was shown for isogenic strains. We find a significant reduction in relative CFU frequency of one strain in mixed biofilms composed of two nonkin strains, with one strain showing a significant dominance over the other. For example, PS-216 won over PS-196 or PS-218 ( Fig. 1 ). In contrast, the relative frequency of both strains in kin and isogenic pellicles was preserved ( Fig. 1 ) (or was much less changed than in nonkin strain combinations when exponential-phase cells were used as inoculum) (see Fig. S3 in the supplemental material). The competitive reduction of one strain in nonkin pellicles is most probably a consequence of antagonism between nonkin. Antagonism between nonkin has been observed at the meeting point of two nonkin swarms of B. subtilis ( 13 , 16 , 18 ) but was absent if kin swarms interacted. Antagonism was suggested also for nonkin Proteus mirabilis ( 35 , 36 ) and Myxococcus xanthus ( 14 , 37 ) interactions. Our results are also similar to those of previous studies on the consequences of mixing genetically divergent Pseudomonas aeruginosa strains in static cultures ( 38 , 39 ). Although these authors did not specifically address kin discrimination, they reported antagonism between genetically different strains in biofilms. It is generally believed that clonemates cooperate whereas genetically different cell lineages of the same species compete ( 1 , 40 , 41 ). Competitive strategies between nonkin comprise contact-dependent killing ( 12 , 16 , 42 ) and local exchange of antagonistic molecules between swarms ( 16 , 43 ). Although swarms represent a different type of collectives than pellicles, it is possible that nonkin deploy similar types of interactions during swarming and pellicle development, although there are also pronounced differences in the opportunities for social interactions. During a swarming encounter assay, cells in the swarm are initially surrounded by their kin and only at the point of encounter of two nonkin swarming cells do they engage in direct cell-cell contact. At this point they are exposed to limitations stipulated by higher cell density and nutrient/space limitation. During pellicle formation, nonkin first engage in interactions in the liquid medium, where they exist as planktonic cells, interacting through secreted and diffusible factors and even EPS that engage cells in a loose network ( 5 ). Several hours later, probably when they sense a limitation of oxygen, they invade the A-L interface, start forming patches, and engage in direct cell-cell interactions ( 26 , 31 ). Our results demonstrate that in pellicles bacterial cells segregate into visible patches similar to those that have previously been shown for colonies ( 44 , 45 ). However, the assortment of nonkin cells results in more prominent patches of the dominant strain that are interspersed by smaller patches of the less competitive strain. This is different from cell organization in kin pellicles, where mixing is more pronounced and less heterogenous. This assortment is consistent with modeling experiments testing interstrain competition in spatially structured environments between a bacteriocin producer and nonproducer, which resulted in killing-driven assortment of two genotypes ( 46 , 47 ). In addition to social interactions, the biofilm heterogeneity is also shaped by steep vertical oxygen gradients that result from oxygen consumption by bacteria and slow oxygen diffusion into the medium ( 5 , 48 , 49 ). Pellicle construction is a strategy for effectively obtaining oxygen at the liquid surface ( 24 , 25 , 50 ). A possible explanation for the role of ECM in maintaining pellicles at the liquid surface is a decrease in the specific gravity or/and intrinsic hydrophobicity of cells glued by ECM, which counteracts selective forces faced by planktonic cells ( 50 ). We observed that the benefit of EpsA-O for occupying the A-L interphase was more important for the strains PS-196, PS-218, and NCIB 3610 than for PS-216 and PS-18, which even with inactivated epsA–O genes still formed weakly aggregated cells at the liquid surface ( Fig. 4 ), a finding that is in accordance with observations by Krajnc et al. (2022), who investigated biofilm formation of the PS-216 Δ epsA–O mutant and its parental strain over time in MSgg medium and a setting similar to what we used ( 26 ). The results also imply that PS-216 may encode additional polysaccharides that compensate for the lack of EpsA-O in the mutant. In fact, natural isolates of B. subtilis often carry the intact copy of the ypqP ( spsM ) gene, encoding a sugar epimerase likely to be involved in polysaccharide synthesis ( 51 ), which is inactivated due to the integration of the SPβ prophage in the standard biofilm NCIB 3610 strain ( 52 ), which also in our setting did not form a pellicle, while the spsM gene is intact in the PS-216 genome. Next, we studied incorporation of the EpsA-O-deficient strain into the pellicle at the air-liquid interface when mixed with the parental or nonkin strain. Confirming previous work performed with the NCIB 3610 strain ( 27 , 28 , 30 , 31 ), all tested Δ epsA–O mutants were able to incorporate into the pellicle and utilize EpsA-O produced by their parental strains. Although the success of incorporation into the pellicle slightly differed between selected Δ epsA–O strains in isogenic combinations, the difference was small. In contrast, the incorporation efficiency of mixing with the nonkin EpsA-O producer was very low. Most importantly, the results revealed that the dominant strains in the selected strain combinations outcompeted the other strain even if they were EpsA-O deficient ( Fig. 5 and Fig. S7), which resulted in more fragile or submerged biofilm of the mutant (Fig. S6). This result suggests that during pellicle formation the dominance of one strain over the other might be linked to antagonism against the less competitive strains regardless of the ECM production efficiency and that through this mechanism kin discrimination limits ECM exploitation between B. subtilis strains. However, antagonistic interactions might be only one of the mechanisms shaping nonkin interactions in bacteria, and our results do not exclude additional mechanisms that could contribute to kin discrimination outcomes. For example, bacteria may produce molecular determinants for kin recognition which ensure preferential grouping with kin strains that share these determinants. In conclusion, our results support the hypothesis that B. subtilis strains harbor antagonistic mechanisms directed toward nonkin strains, which operate through inhibition of pellicle formation of one strain and a modified assortment of nonkin groups in the floating biofilm. These mechanisms may limit the invasion of the WT groups by nonkin ECM nonproducers and thereby biofilm ECM exploitation."
} | 3,806 |
24006130 | null | s2 | 1,963 | {
"abstract": "Development of increasingly complex integrated cellular systems will be a major challenge for the next decade and beyond, as we apply the knowledge gained from the sub-disciplines of regenerative medicine, synthetic biology, micro-fabrication and nanotechnology, systems biology, and developmental biology. In this prospective, we describe the current state-of-the-art in the assembly of source cells, derived from pluripotent cells, into populations of a single cell type to produce the components or building blocks of higher order systems and finally, combining multiple cell types, possibly in combination with scaffolds possessing specific physical or chemical properties, to produce higher level functionality. We also introduce the issue, questions and ample research opportunities to be explored by others in the field. As these \"living machines\" increase in capabilities, exhibit emergent behavior and potentially reveal the ability for self-assembly, self-repair, and even self-replication, questions arise regarding the ethical implications of this work. Future prospects as well as ways of addressing these complex ethical questions will be discussed."
} | 290 |
28586408 | PMC5812522 | pmc | 1,964 | {
"abstract": "Abstract The main focus in development of yeast cell factories has generally been on establishing optimal activity of heterologous pathways and further metabolic engineering of the host strain to maximize product yield and titer. Adequate stress tolerance of the host strain has turned out to be another major challenge for obtaining economically viable performance in industrial production. Although general robustness is a universal requirement for industrial microorganisms, production of novel compounds using artificial metabolic pathways presents additional challenges. Many of the bio-based compounds desirable for production by cell factories are highly toxic to the host cells in the titers required for economic viability. Artificial metabolic pathways also turn out to be much more sensitive to stress factors than endogenous pathways, likely because regulation of the latter has been optimized in evolution in myriads of environmental conditions. We discuss different environmental and metabolic stress factors with high relevance for industrial utilization of yeast cell factories and the experimental approaches used to engineer higher stress tolerance. Improving stress tolerance in a predictable manner in yeast cell factories should facilitate their widespread utilization in the bio-based economy and extend the range of products successfully produced in large scale in a sustainable and economically profitable way.",
"conclusion": "CONCLUSIONS This review has discussed the main stress factors encountered by yeast cell factories used for production of bioethanol and bio-based chemicals and also how to engineer the cell factories for increased stress tolerance. Yeast cell factories are subject to a variety of stress factors during first- and second-generation fermentation processes. The industry is in dire need of microorganisms that can cope with the multitude of stress factors that can occur. A variety of yeast species possess specific stress tolerance characteristics that are of interest to the industry but lack other desirable traits. Saccharomyces cerevisiae is still considered to be one of the most robust and versatile microorganisms. Its long-standing track record in the fermentation industry makes it an obvious choice for further improvement of its stress tolerance properties to make it even more suitable for use in biofuel and bio-based chemical production with difficult feedstocks. However, S. cerevisiae is lacking the capacity for utilization or production of certain compounds, it can also make undesirable post-translational modifications in protein production and has lower tolerance to certain stress factors than other species. Therefore, other microorganisms could be favored for specific industrial production processes. Most information on the molecular-genetic basis of stress tolerance has been gained with laboratory yeast strains. This information is often not applicable to industrial yeast strains. Hence, more research is needed on the genetic basis of stress tolerance characteristics in industrial and natural yeast strains. These strains contain interesting alleles for targeted strain improvement while minimizing the risks of side effects on other industrially important traits. Most research has also concentrated on genetic elements that are required to maintain stress tolerance and much less on genetic modifications that can improve stress tolerance. Since industrial yeast strains are generally much more robust and stress tolerant than laboratory yeast strains, more research on genetic factors important for stress tolerance and especially enhancement of stress tolerance in industrial yeast strains is required. Genetic modifications that improve stress tolerance in yeast can compromise other properties that are important for industrial application, such as fermentation performance, propagation rate and tolerance to drying and storage conditions. Hence, it is important to evaluate newly constructed yeast strains under as many conditions as possible that are relevant in the industry. The recent advent of highly powerful and precise genome engineering technologies, especially the CRISPR-Cas9 technology, has enabled much faster engineering of industrial yeast strains and will hopefully stimulate the development of highly productive cell factories for various applications in the production of biofuels and bio-based chemicals. \n Conflict of interest. None declared.",
"introduction": "INTRODUCTION There has been a surge of interest in the commercial production of bio-based chemicals with microorganisms used as cell factories (Becker and Wittmann 2012 ; Chung et al. 2015 ; Tsuge et al. 2016 ). The yeast Saccharomyces cerevisiae has been a favorite organism in this respect because of its long-standing use in classical industrial applications, such as beer and wine production, its extensive toolbox for genetic modification and the vast knowledge on its physiology, molecular biology and genetics (Kampranis and Makris 2012 ). Although industrial yeast strains have great robustness, they often lack tolerance to specific stress factors when used as cell factories. A first near-universal stress factor is the toxicity of the end product, which has to be accumulated with high yield and maximal titer in order to ensure economic viability of the industrial process. Favored chemicals for cell factory production like organic acids exert strong inhibition on the metabolism of microorganisms, including yeast (Narendranath, Thomas and Ingledew 2001 ). Even the accumulation of very high levels of ethanol in biofuel production is toxic to the yeast (Stanley et al . 2010a ; Pais et al. 2013 ). A second source of stress factors is the composition of the substrate and its pretreatment process. Pure sugar streams can be used as feedstock for cell factory production of bio-based chemicals, but because of the low value of bulk chemicals and biofuels, the use of cheaper substrates is preferred. The latter are usually much more heterogeneous and often contain high levels of inhibitors, either present in the substrate itself or generated during the pretreatment process (Palmqvist and Hahn-Hägerdal, 2000 ). This necessitates the engineering of much higher tolerance against these inhibitors than is generally present in natural or industrial strains of species used as cell factories, including yeast. A third major stress factor is high temperature. Enzymatically catalyzed reactions and thus also microbial production processes for bio-based compounds proceed faster at higher temperature, which is favored because it enhances the productivity of the commercial plant and thus reduces capital expenses (capex). In addition, microbial fermentation processes are exergonic. They produce heat, and large-scale fermentors thus have to be properly cooled. In combination with changing environmental temperatures, this can cause temperature gradients and fluctuations in the fermentors that can compromise fermentation rate and productivity (Abdel-Banat et al. 2010 ). Additional stress factors include high osmolarity. Because of the need to achieve high product titers, very high gravity fermentations are required, increasing osmotic pressure during fermentation. Salt tolerance can be important because of high salt levels introduced during the pretreatment, the use of new feedstocks such as seaweed, the cleaning and water recirculation practices (Maiorella, Blanch and Wilke 1984 ; Silverstein et al. 2007 ; Wi et al. 2009 ; Chavez-Rodriguez et al. 2013 ; Wei, Quarterman and Jin 2013 ). These stress factors not only require a high level of general robustness of the cell factory microorganism but also much higher tolerance to specific stress factors than usually present even in the most robust industrial strains. The stress factors very often also reinforce each other making it even more difficult to reach the required level of fermentation performance under real industrial conditions. A rather unexpected outcome of the development of microbial cell factories is that the new artificially engineered metabolic pathways tend to be much more sensitive to stressful conditions than the intrinsic metabolic pathways of the organism. This is very clear for instance in the co-fermentation of glucose and xylose in second-generation bioethanol production, in which the artificially engineered xylose fermentation turns out to be much more sensitive to inhibitors such as acetic acid compared to glucose fermentation (Bellissimi et al. 2009 ). This is likely due to the extensive adaptation and selection that the microorganism has undergone during evolution when fermenting its natural substrates under myriads of different environmental conditions, whereas fermentation of the artificial substrate has never undergone a similar fine-tuned integration in the regulatory network governing microbial metabolism. This review provides an overview of the most important stress factors for yeast cell factories in industrial production processes (Fig. 1 ). Many of these stress conditions are general for all fermentations. However, due to the nature of the biomass and the pretreatment process used in second-generation bioethanol production, yeast cell factories are confronted with several additional stress factors in this process. Figure 1. Schematic overview of the different steps in the industrial production of bioethanol and bio-based chemicals with first- and second-generation substrates using yeast cell factories and the most common stress factors associated with the different subprocesses. To improve the yields and titers, superior alleles conferring tolerance to specific stress factors could be engineered in industrial strains using several methods, with one of the most powerful the recently developed, highly precise and scarless CRISPR-Cas9 genome editing approach. Also, several non-conventional yeast species display specific properties that are highly desirable in the fermentation industry such as thermotolerance, weak acid tolerance and osmotolerance ( Kluyveromyces marxianus , Zygosaccharomyces bailii and Debaryomyces hansenii , respectively) (Radecka et al. 2015 ). However, despite the superior growth of these yeast species under the specific stress condition, their implementation in industrial fermentations is often troublesome due to absent or very limited fermentation capacity and a low general robustness."
} | 2,606 |
37031262 | PMC10082802 | pmc | 1,965 | {
"abstract": "Photosynthetic algae have evolved mechanisms to cope with suboptimal light and CO 2 conditions. When light energy exceeds CO 2 fixation capacity, Chlamydomonas reinhardtii activates photoprotection, mediated by LHCSR1/3 and PSBS, and the CO 2 Concentrating Mechanism (CCM). How light and CO 2 signals converge to regulate these processes remains unclear. Here, we show that excess light activates photoprotection- and CCM-related genes by altering intracellular CO 2 concentrations and that depletion of CO 2 drives these responses, even in total darkness. High CO 2 levels, derived from respiration or impaired photosynthetic fixation, repress LHCSR3 /CCM genes while stabilizing the LHCSR1 protein. Finally, we show that the CCM regulator CIA5 also regulates photoprotection, controlling LHCSR3 and PSBS transcript accumulation while inhibiting LHCSR1 protein accumulation. This work has allowed us to dissect the effect of CO 2 and light on CCM and photoprotection, demonstrating that light often indirectly affects these processes by impacting intracellular CO 2 levels.",
"introduction": "Introduction A major challenge for photosynthetic organisms is to efficiently acclimate to highly dynamic light and nutrient conditions that occur in natural environments. While light provides the energy that fuels photosynthetic CO 2 fixation, excess light can cause oxidative damage and ultimately result in cell death. Therefore, light absorption must be precisely managed via photoprotective mechanisms that help integrate the use of light energy with CO 2 availability and the potential of the organism to grow and store fixed carbon. A dominant photoprotective mechanism, called qE (energy-dependent quenching), results in the harmless dissipation of excess absorbed light energy as heat 1 , 2 . Triggering qE requires the synthesis of specific proteins and pigments that are controlled both transcriptionally and post-transcriptionally. In the green microalga Chlamydomonas reinhardtii (hereafter Chlamydomonas ), qE depends on the nucleus-encoded, chloroplast-localized Light Harvesting Complex-Stress Related (LHCSR) proteins LHCSR1, LHCSR3 and Photosystem II Subunit S, PSBS, which are present in many algae and lower plants 3 and belong to the Light Harvesting Complex protein superfamily 4 . The LHCSR3.1 and LHCSR3.2 genes in Chlamydomonas encode identical LHCSR3 proteins 5 , while PSBS1 and PSBS2 encode proteins that differ by only one amino acid of the chloroplast transit peptide 6 . While LHCSR1 and LHCSR3 are present in algae but not in vascular plants, PSBS is present in both 4 . PSBS in Chlamydomonas is transiently expressed in cells exposed to high light (HL) 6 , 7 and accumulates in cells exposed to UV-B irradiation 8 . LHCSR3 is the main qE effector protein in HL 5 , although LHCSR1 can significantly contribute to qE under certain conditions 9 . In Chlamydomonas , expression of LHCSR3 has been reported to increase upon absorption of blue-light by the photoreceptor phototropin (PHOT1) 10 and involves calcium ion signaling 11 , active photosynthetic electron transport (PET) 10 , 11 and the transcriptional factor CONSTANS, which is also required for activation of the LHCSR1 and PSBS genes 12 , 13 . Similar to the dynamic light cue, the concentration of inorganic carbon (HCO 3 − , CO 2 and CO 3 2− , together designated Ci) in aquatic environments varies spatially and temporally; aquatic CO 2 levels can also fluctuate from extremely high (hyper-saturated) to extremely low 14 . Because low CO 2 levels limit photoautotrophic growth, microalgae have evolved a CO 2 Concentrating Mechanism (CCM) that elevates the level of CO 2 at the site of fixation by Ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO). Major components of the CCM are carbonic anhydrases (CAH), which facilitate interconversions among the different Ci species, and Ci transporters. The genes encoding many Ci transporters and CAHs are under the control of the zinc-finger type potential transcription regulator CIA5 (also CCM1) 15 , 16 , which is localized in the nucleus 17 and controls expression of low-CO 2 responsive genes. In addition to the use of CO 2 to support phototrophic growth in the light, Chlamydomonas can also use the two-carbon molecule acetate either in the dark to support heterotrophic growth, or in the light, to support photoheterotrophic or mixotrophic growth 18 . Acetate is incorporated into acetyl-CoA either in a one-step reaction catalyzed by acetyl-CoA synthetase (ACS), or in two steps that use acetate kinase (ACK) and phosphate acetyltransferase (PAT), which sequentially catalyze the formation of acetyl-phosphate and acetyl-CoA 19 . Acetyl-CoA can then enter the glyoxylate cycle, a shunt of the tricarboxylic acid (TCA) cycle 20 , recently characterized in Chlamydomonas 21 , where it can be converted to metabolites that are used for anabolic metabolism. Alternatively, acetyl-CoA enters the TCA cycle to feed the respiratory chain with reducing equivalents. Both, the glyoxylate cycle and respiration are essential for growth in the dark since Chlamydomonas mutants affected in either of these processes are unable to grow heterotrophically 21 , 22 . Despite the evident connection between light and CO 2 levels, the physiological responses to different light and CO 2 availabilities have been traditionally studied separately. However, several lines of evidence indicate that both acetate and Ci abundance impact not only qE but also the establishment of the CCM in Chlamydomonas 23 – 26 , while LHCSR3 transcripts accumulation have been reported to be CIA5-dependent 26 – 28 . Yet, the mechanism(s) associated with carbon-dependent regulation of qE and CCM induction and the intimate link between the two processes have still not been clearly defined. Here, using genetic and mathematical modelling approaches, we demonstrate that inhibition of LHCSR3 accumulation and CCM activity by acetate is at the level of transcription and a consequence of metabolically produced CO 2 . We also show that exposure of Chlamydomonas to HL triggers not only HL responses, but also low-CO 2 responses, and we report the discovery of a novel CO 2 - and CIA5-dependent mechanism that activates LHCSR3 gene expression even in complete darkness. Finally, we propose that PET is critical for the activation of LHCSR3 transcription because it sustains CO 2 fixation, consuming intracellular CO 2 and thereby relieving its inhibitory effect. This work emphasizes the importance of CO 2 in regulating photoprotection and the CCM, and demonstrates that light often indirectly affects these processes by altering intracellular CO 2 levels.",
"discussion": "Discussion In this work, we presented findings that advance our understanding of integration between CO 2 - and light-dependent signaling in Chlamydomonas . We propose that the intracellular level of CO 2 , defined by the equilibrium between light-driven CO 2 fixation in chloroplasts and the generation of CO 2 by mitochondrial metabolism (e.g. acetate assimilation), is a key regulator of two major processes in photosynthetic organisms: the CCM and photoprotection (Fig. 7 ). Fig. 7 CO 2 - and light-dependent signals converge to regulate photoprotection and CCM in Chlamydomonas. The intracellular levels of CO 2 , defined by the equilibrium between CO 2 fixation in chloroplasts and the generation of CO 2 by mitochondrial metabolism (e.g. acetate assimilation) is the key determinant of the regulation of gene expression controlling two major processes of photosynthetic organisms: CCM and photoprotection. Changes in light availability have a direct impact on intracellular CO 2 levels; exposure to HL increases CO 2 fixation rates leading to depletion of CO 2 and to activation of not only photoprotection- but also CCM-related genes. Conversely, depletion of CO 2 is sufficient to drive high expression levels of CCM genes and LHCSR3 even in complete darkness (indicated by the black arrows). High CO 2 levels, either exogenously supplied by sparging or metabolically produced via acetate metabolism or by inhibiting photosynthetic electron flow using DCMU, repress LHCSR3 and CCM genes while at the same time they stabilize LHCSR1 protein levels. The close interconnection of photoprotection and CCM is further corroborated by the fact that CIA5, the regulator of expression of genes associated with the CCM, also exerts control over LHCSR3 and to a lesser extent over PSBS mRNA levels and acts as repressor of LHCSR1 protein accumulation. Independent of CIA5, light strongly impacts expression of all of these photoprotective genes (yellow arrows). This impact can be the consequence of both photoperception (e.g. phototropin) and the production of reactive oxygen species. To better understand the role of CO 2 in regulating photoprotection and its integration with light, we designed experiments to separate the effects of the two signals (Figs. 5 – 7 ); we reduced the concentration of CO 2 in the microalgae medium by sparging it with VLCO 2 in complete darkness. This abrupt change in CO 2 levels experienced by the cultures in the dark may be considered a condition only encountered in the laboratory. However, in certain ecological niches, such as soil or catchments with elevated levels of organic matter 33 , Chlamydomonas would encounter changes in the levels of CO 2 that would be dependent on the microbes and the ratio between respiration and photosynthesis in the environment. Our experimental setup allowed us to observe a strong increase of LHCSR3 transcript levels when cells were shifted from air-CO 2 to VLCO 2 levels in the dark (Fig. 5 ), a very surprising result as the accumulation of LHCSR3 mRNA was considered so far to be strictly light-dependent 5 , 11 , 38 . Moreover, with this strategy we can disentangle light from CO 2 signalling effects; while dark induction of LHCSR3 under CO 2 -depletion was completely dependent on CIA5, light could still strongly impact expression of all photoprotective genes in the cia5 mutant, which was not the case for CCM gene expression that was completely abolished in the light or dark in the absence of CIA5 (Fig. 5 and Supplementary Fig. 6 ). This impact of light on qE gene expression may be the consequence of photoperception (e.g. PHOT1) 10 , but also the generation of light-dependent signals such as reactive oxygen species 28 . Furthermore, a CIA5-independent regulation (also observed in Fig. 3a ) explains LHCSR3 induction in high CO 2 -acclimated WT cells (cells in which CIA5 is not functional 15 – 17 ) as they transition from LL to HL (Fig. 1a ), which was not observed for CCM genes tested under identical conditions (Supplementary Fig. 3 ); it also explains why the CO 2 -mediated repression was more pronounced for most of the CCM genes relative to LHCSR3 (Fig. 1a, c , Supplementary Fig. 3 ). CO 2 and CIA5 appear to be of paramount importance in signal integration and transduction, regulating expression of both photoprotection and CCM genes. For instance, CO 2 represses the UV-B elicited, UVR8-mediated expression of LHCSR3 , and CIA5 is absolutely required for this expression 28 . Moreover, our results have shown that high CO 2 levels or the absence of CIA5 have a severe impact on LHCSR3 gene expression and, although HL can still induce LHCSR3 transcription, no protein is detected (Figs. 1 , 3 and 5 ). Besides transcriptionally controlling LHCSR3 , CIA5 post-transcriptionally controls LHCSR1. Our view on LHCSR1 regulation by light and CIA5 is as follows: under LL conditions, LHCSR1 protein accumulates in cia5 while it is non-detectable in WT and cia5-C (Fig. 3b ), suggesting that CIA5 suppresses LHCSR1 protein accumulation. Exposure to HL triggers a CIA5-independent LHCSR1 mRNA accumulation (Fig. 3a ), possibly driven by reactive species, previously shown to favor LHCSR1 mRNA accumulation 40 . As a result, LHCSR1 protein accumulates in WT and cia5-C in HL, despite the fact that suppression of LHCSR1 protein by CIA5 still occurs; indeed, LHCSR1 accumulates to higher levels in the cia5 mutant as compared to WT and cia5-C under HL conditions (Fig. 3b ). In line with the above observations in the cia5 mutant, high levels of LHCSR1 protein accumulate in WT under high CO 2 , conditions that inactivate CIA5 (Fig. 4b ). Put together, our findings unveil a multilevel role of CIA5 in regulating qE; inactivation of CIA5 in high CO 2 or by eliminating the CIA5 gene blocks LHCSR3 transcript accumulation, while it promotes LHCSR1 protein accumulation (Figs. 3 , 4 ). Further investigation will be required to explain how a single nuclear factor, CIA5, can control cellular processes happening in different cellular compartments; transcription in the nucleus and translation in the cytosol. Our results provide an interpretation of the findings that PET is required for LHCSR3 accumulation 11 , activation of the CCM and expression of CCM genes 41 . We propose that CO 2 , either provided directly or indirectly through metabolic generation, represents a critical link between PET and transcriptional regulation of LHCSR3 and the CCM genes (Fig. 6 ). Photosynthesis draws down cellular CO 2 levels, and therefore, blocking photosynthesis with DCMU leads to the accumulation of CO 2 (Fig. 6a ) which elicits LHCSR3 repression, while sparging DCMU-treated cells with VLCO 2 almost fully derepresses LHCSR3 (and partially CCM) expression (Fig. 6b ). DCMU also upregulates genes of the leucine degradation pathway 42 leading to the generation of acetoacetate and acetyl-CoA, which can lead to oxidative CO 2 production. Whether leucine itself has a regulatory role or CO 2 is the key regulator deserves further attention. It is tempting to propose that CO 2 is a retrograde signal that readily diffuses through the cell and impacts nuclear gene expression, which would integrate both mitochondrial and chloroplastic metabolic activities. The way in which Chlamydomonas senses CO 2 is not clear. Our data, i.e. accumulation of LHCSR3 and CCM genes in the dark, exclude the possibility that a metabolite produced by photorespiration plays a major signalling role, as previously proposed 43 . CO 2 itself might also serve as the metabolite being recognized by a putative sensor that could be controlled by carbamylation, a CO 2 -mediated post-translational modification that regulates, among others, the activation of Rubisco 44 . Furthermore, the large number of adenosine and guanylyl cyclases in Chlamydomonas 45 suggests that cyclic nucleotides play an important role in controlling various processes in this alga; these metabolites have been shown to be involved in mating 46 , regulation of flagellar beating and phototaxis 47 – 49 , in regulating inorganic nitrogen assimilation 50 and in restoring LHCSR3 accumulation in the absence of phototropin 10 . Cyclases have been shown to act as CO 2 sensors (as bicarbonate) in mammalian cells 51 , making it plausible that they can also serve as sensors in Chlamydomonas . As cyclic nucleotide signalling and calcium are tightly linked 51 , we anticipate an important role for calcium in CO 2 sensing; calcium signalling has already been shown to be involved in the regulation of both LHCSR3 and CCM genes 11 , 52 . Overall, our work shows that the intracellular CO 2 level is the main factor in regulating CCM genes and LHCSR3 in Chlamydomonas (Fig. 7 ). Exposure to HL increases the CO 2 fixation rate which causes a drop in intracellular CO 2 which, in turn, actives both photoprotection- and CCM-related genes. Depletion of CO 2 is sufficient to drive high expression levels of CCM genes and LHCSR3 even in complete darkness. On the other hand, high CO 2 levels, either generated through enhanced respiratory activity or impaired photosynthetic electron transport, repress LHCSR3 and CCM genes while at the same time stabilizing the LHCSR1 protein, which likely acts as a backup photoprotection protein under conditions where LHCSR3 is not expressed. Furthermore, our data reveals a closer interconnection of photoprotection and CCM as CIA5, the CCM master regulator, also exerts control over LHCSR3 and to a lesser extent over PSBS mRNA levels, while repressing LHCSR1 protein accumulation. Our findings highlight the need to develop an integrated approach that examines the role of CO 2 and light, not only as substrates of photosynthetic CO 2 fixation, but also as signals regulating photoprotection, CCM, and at a wider context genome-wide gene expression."
} | 4,168 |
27443913 | PMC4956755 | pmc | 1,966 | {
"abstract": "Brain-inspired computing architectures attempt to mimic the computations performed in the neurons and the synapses in the human brain in order to achieve its efficiency in learning and cognitive tasks. In this work, we demonstrate the mapping of the probabilistic spiking nature of pyramidal neurons in the cortex to the stochastic switching behavior of a Magnetic Tunnel Junction in presence of thermal noise. We present results to illustrate the efficiency of neuromorphic systems based on such probabilistic neurons for pattern recognition tasks in presence of lateral inhibition and homeostasis. Such stochastic MTJ neurons can also potentially provide a direct mapping to the probabilistic computing elements in Belief Networks for performing regenerative tasks.",
"conclusion": "Conclusions To conclude, researchers have explored MTJs as synapses 29 30 31 32 and for inter-neuron communication 30 previously. Further, previous research on utilizing spintronic devices as neurons 33 34 have been limited to emulating only thresholding operations of non-spiking neural computing models. On the other hand, spiking neurons offer a more biologically realistic perspective and are recently becoming popular computing models for implementing low-power, high accuracy recognition platforms in complex cognitive tasks 35 . Stochasticity exhibited by phase change memory 36 and spintronic devices 32 have been exploited previously in neuromorphic applications to implement learning functionality in synapses. However, the utilization of device stochasticity in nanoelectronic neural computing has been a relatively unexplored area. To the best of our knowledge, this is the first demonstration of mapping the stochastic leaky-integrate switching behavior of MTJs in presence of thermal noise to a probabilistic spiking neuron. An important point worth considering is whether other post-CMOS technologies 36 exhibiting stochastic switching characteristics could be potentially operated as neurons as well. A few words regarding the architecture of the pattern recognition system ( Fig. 5 ) are in order to outline the prospective opportunities offered by spintronic neurons. Neurons need to be interfaced with a crossbar array of resistive synapses for any pattern recognition system. Memristive devices are present at each cross-point to encode the synaptic weight. Input voltages are applied across each row and the current flowing through the memristors is weighted by its conductance and gets summed up along the column and passes as input to the neuron. However, this is true only when the input resistance of the neuron is sufficiently low since otherwise, the voltage drop across each memristor will be dependent on the voltage drop across the neuron which in turn, depends on the total amount of input synaptic current resulting in a coupled system. Low terminal voltage of MTJ neurons during “write” operation offers unique possibilities in this regard. Input synaptic current flows through the HM (with low resistance) and not through the oxide layer of the MTJ. Thus decoupled “read” and “write” current paths of the proposed neuron assist the neuron operation. In contrast, memristive devices are usually characterized by high threshold voltages (>1 V ) and high resistance values ( K Ω- M Ω 23 24 36 ). Hence, although intrinsic noise might be present in memristive devices, it will be potentially difficult to interface memristive synaptic crossbar arrays with memristive neurons. Although the impact of thermal noise on MTJ switching behavior has limited its scalability in memory applications, such noise effects can be potentially exploited to build probabilistic neural computing platforms that can perform Bayesian computation similar to the brain. Past research on hardware implementation of spiking neurons has mainly focused on the emulation of deterministic spiking neural characteristics and require area and power expensive CMOS implementations involving more than 20 transistors 2 3 . CMOS based stochastic neural models might be possible 37 but involve significant silicon area and power consumption since they do not offer a direct mapping to the underlying neuroscience mechanisms. However, the ultra-low current induced noisy switching characteristics of MTJs can efficiently mimic such stochastic spiking neural models and can potentially pave the way for neuromorphic systems that utilize noisy stochastic neurons as a computing element, such as Restricted Boltzmann Machines and Deep Belief Networks. We would like to conclude the paper by noting that the device stochasticity observed in such MTJ structures can be utilized to realize probabilistic learning functionality in single-bit synapses 32 which could be potentially interfaced with stochastic MTJ neurons resulting in an All-Spin neuromorphic architecture that leverages the underlying device stochasticity to perform neuromimetic computing."
} | 1,229 |
37520694 | PMC10384274 | pmc | 1,969 | {
"abstract": "Summary The development of biohydrogen as an alternative energy source has had great economic and environmental benefits. Hydrogen production from microalgae is considered a clean and sustainable energy production method that can both alleviate fuel shortages and recycle waste. Although algal hydrogen production has low energy consumption and requires only simple pretreatment, it has not been commercialized because of low product yields. To increase microalgal biohydrogen production several technologies have been developed, although they struggle with the oxygen sensitivity of the hydrogenases responsible for hydrogen production and the complexity of the metabolic network. In this review, several genetic and metabolic engineering studies on enhancing microalgal biohydrogen production are discussed, and the economic feasibility and future direction of microalgal biohydrogen commercialization are also proposed.",
"conclusion": "Conclusion To combat resource depletion and environmental concerns, microalgal biohydrogen must be commercialized. The application of genetic and metabolic engineering along with synthetic biology has been shown to make biohydrogen production more cost-effective. With the understanding of algal genome sequences and metabolic functions, as well as the development of advanced tools and software, the combination of genetic engineering and bioinformatics will no doubt help to develop microalgae strains that are more suitable for biohydrogen production than those available currently. Although measures have been taken to improve the competitiveness of biohydrogen production, the infant stage of this technology has not been applied to large-scale hydrogen production. A tremendous amount of effort is still required before this technology could occupy a large market share. Future directions and challenges As a sustainable energy source, microalgal biohydrogen has attracted increasing international attention. It can be produced in two ways, with the most common being hydrogen fermentation using microalgal biomass as raw material as a sustainable and robust feedstock for large-scale energy production. Genetic and metabolic engineering can be used to improve the efficiency of photosynthesis, improve carbon sequestration by driving carbon fluxes into energy-rich compounds that can be used as hydrogen energy sources, and develop robust microalgae strains to allow for low-cost large-scale cultivation to reduce operating costs. 115 , 182 Hydrogen can also be produced from microalgae through metabolism. There are two main processes involved in the production of biohydrogen; increasing biomass, and photolysis, photofermentation, or dark fermentation under anaerobic conditions. 183 Improving the efficiency of photosynthesis, rapidly accumulating carbohydrates, and maintaining cellular stability are important for the efficient use of water and organic matter. In addition, it is necessary to increase the activity of enzymes associated with hydrogen production through genetic manipulation, and reconstruct metabolic pathways through metabolic engineering to establish more efficient production strains. 121 , 165 Currently, there are bottlenecks in the large-scale production of biohydrogen, as production costs depend not only on the microorganisms used and metabolic pathways present 84 ; it is also necessary to improve the technology used for microalgae biomass cultivation and harvesting. Adjustments to culture conditions may have a large impact on the biomass and carbohydrate contents, and optimal production can be achieved by a proper combination of genetic manipulation and culture condition engineering. 115 Gene editing technology can overcome many obstacles, but the stability and safety of genetically-modified microalgae are also primary factors to consider, requiring further evaluation to ensure that they do not pose future risks to human and environmental health. 115 Another primary challenge in facing metabolic engineering is the limited number of available genes and the complex genotype-phenotypic relationships resulting from their combination with host organism metabolic pathways. 165 For downstream biohydrogen production, it is necessary to weaken the demand for light in downstream processes and consider more energy-efficient dark fermentation strategies. Biological hydrogen production also does not produce only one gas product, impacting hydrogen purity. Low-cost gas separation technologies need to be developed to remove by-products such as carbon dioxide. 184 For continuous production, periodicity is required, and the integration of steps in the production process is crucial. It is necessary to specifically understand the impact of each variable in each step on hydrogen production to achieve the highest possible hydrogen production. 108 Recommendations Application of genetic engineering in the field of biofuels is expanding rapidly. Genetic engineering tools solve many problems, and provide industrially capable microalgae strains. A systematic review of biofuel production in genetically modified (GM) algae between 2008 and 2019 found that one of the areas that received the least attention was environmental risk. 185 There are differences in biotechnology laws and regulations between countries; because of the lack of internationally harmonized regulations, the commercialization of genetically modified biofuels cannot be effectively applied. The use of genetic engineering can provide high yields, high growth rates, and insect resistance, but the impact of released modified plasmid or chromosomal DNA into surrounding water bodies has been overlooked. 186 Farming without a comprehensive risk assessment of GM strains can pose a serious threat to human and environmental health. More work should be done in regards to safety testing of GM organisms to assess the potential risks posed to the environment from the growth and processing of GM microalgae.",
"introduction": "Introduction Fossil fuels are the most important energy source globally and have brought great economic benefits. 1 Their excessive use, however, has negative environmental and human health impacts. At present, more than 80% of global energy use originates from fossil fuels. 1 With continuous population growth, fossil fuel demand will only increase, and the output of limited fossil fuel resources will inevitably decline in this century. 2 , 3 Excessive exploitation of fossil fuels has also caused deforestation, loss of farmland, and damage to ecosystems because of acid rain. 4 Global pollution caused by fossil fuels has multiple negative consequences 5 ; because the burning of fossil fuels contributes to about 75% of global carbon-dioxide emissions, it is estimated that carbon emissions from fossil fuels will increase to 39 billion tons by 2030. 6 Global warming caused by carbon emissions from human activities has resulted in a global temperature 1°C higher than that before industrialization. Global warming has also led to increased frequency and severity of ice melt, leading to rising sea levels and extreme weather. 7 The combustion of fossil fuels also negatively impacts human health, causing and exacerbating respiratory and nervous system diseases. 8 Reducing reliance on fossil fuels and finding alternative energy sources are therefore crucial for the long-term sustainability of human development. 9 Hydrogen has the highest energy content per unit of mass of all known fuels. It is the most abundant element in the universe and is a clean energy source because water is the only by-product of its combustion. 10 The high energy efficiency, high energy density, and safety of solid hydrogen demonstrate promise for scaling hydrogen energy. 11 As hydrogen is considered the most promising alternative energy source, hydrogen produced from renewable resources could be a sustainable and clean fuel. 12 , 13 Biomass-derived hydrogen can be generated through thermochemical and microbial processes. 14 , 15 The low concentration of hydrogen produced by thermochemical methods and the need for high-temperature complex operating conditions increase production costs and energy consumption, making them unsustainable. 16 , 17 Compared with chemical hydrogen production, biological hydrogen production by microbes can be carried out under lower temperatures and pressures. 18 , 19 Microorganisms can easily become hydrogen production sources. 20 Microalgae are a kind of biomass with great potential for hydrogen production. 21 Owing to their highly adaptable nature, they are widely distributed globally. 22 Microalgae are the basis of the global carbon cycle and absorb 50% of global carbon dioxide. 23 They also have rapid growth rates as well as high yields and carbohydrate contents. 24 They can be cultured in a variety of substrates including fresh water, seawater, domestic sewage, and wastewater. 25 , 26 As microalgae convert light into chemical energy and produce hydrogen by cracking water, they are considered a green production source. In addition, algal biomass can be cultured at a large scale, and could be used as a substrate fuel source for hydrogen production. 27 , 28 , 29 Novel strategies for microalgal hydrogen production have been explored to improve competitiveness. 30 Various pretreatments could also improve hydrogen recovery. 31 Despite these successes, there are some obstacles to developing large-scale hydrogen production using algae. 24 The production of hydrogen in these systems depends on interactions in the microalgal metabolic network, including photosynthesis, respiration, and fermentation; these natural processes are not enough to support large-scale industrial production. 32 , 33 With the application of genetic engineering, microalgae are becoming the most promising microorganisms for hydrogen production. 19 Directed transformation has been shown to improve hydrogen production capacity. 34 This review provides an overview of pretreatment techniques for microalgal biomass and discusses methods to improve biohydrogen production using genetic and metabolic engineering. Many strategies, including adjusting environmental tolerance, metabolic remodeling, and the expression of synthetic pathways from high-yield strains have been adopted to improve hydrogen production on an industrial scale. 35\n\nIntroduction of algal substrate The selection of the right substrate for industrial-scale fuel production is important because it accounts for 60–70% of the cost of production. 36 Traditional hydrogen production methods involving lignocellulose consumption lead to competition with farmland and utilization of land that could be used for other purposes. 18 The crystalline structure of lignocellulose limits its utilization by microorganisms and enzymes. The scarcity of land also hampers energy crops. 37 In contrast, microalgae have almost no lignin and low hemicellulose content, and the required pretreatment is not energy-intensive and does not require the use of ecologically toxic solvents that can cause secondary pollution. 38 Biofuel production has shifted from using sugar and starch to lignocellulosic materials, and finally to the third generation of algal biomass in recent years. 36 The transformation of microalgal biomass is an efficient and environmentally friendly strategy to produce hydrogen. 39 It can be converted into hydrogen by microbial fermentation. 40 ‘Microalgae’ is a general term for algae whose morphology can be distinguished under a microscope. 41 They are extremely diverse single-cell populations that can be as small as 0.2 μm. 42 Microalgae can be composed of one or a few cells, and there can also be structures formed by the aggregation of many cells. 35 , 24 The composition of algal cells also varies with species and the surrounding environment, with carbohydrate content reaching as high as 70%, making them ideal substrates for biohydrogen fermentation. 43 , 44"
} | 2,988 |
36643432 | PMC9835640 | pmc | 1,970 | {
"abstract": "Water-based superamphiphobic coatings that are environmentally\nfriendly have attracted tremendous attention recently, but their performances\nare severely limited by dispersibility and mechanical durability.\nHerein, a dispersion of poly(tetrafluoroethylene)/SiO 2 @cetyltrimethoxysilane&sodium\nsilicate-modified aluminum tripolyphosphate (PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP) superamphiphobic coatings was formed\nby mechanical dispersion of poly(tetrafluoroethylene) emulsion (PTFE),\nmodified silica emulsion (SiO 2 @CTMS), sodium silicate (Na 2 SiO 3 ), and modified aluminum tripolyphosphate (modified\nATP). The four kinds of emulsions were mixed together to effectively\nsolve the dispersity of waterborne superamphiphobic coatings. Robust\nwaterborne superamphiphobic coatings were successfully obtained by\none-step spraying and curing at 310 °C for 15 min, showing strong\nadhesive ability (grade 1 according to the GB/T9286), high hardness\n(6H), superior antifouling performance, excellent impact resistance,\nhigh-temperature resistance (<415 °C), anticorrosion (immersion\nof strong acid and alkali for 120 h), and heat insulation. Remarkably,\nthe prepared coating surface showed superior wear resistance, which\ncan undergo more than 140 abrasion cycles. Moreover, the composite\ncoating with 35.53 wt % SiO 2 @CTMS possesses superamphiphobic\nproperties, with contact angles of 160 and 156° toward water\nand glycerol, respectively. The preparation method of superamphiphobic\ncoatings may be expected to present a strategy for the preparation\nof multifunctional waterborne superamphiphobic coatings with excellent\nproperties and a simple method.",
"conclusion": "4 Conclusions In summary, the waterborne\ndispersion was successfully prepared\nby simple mechanical dispersion of emulsions (PTFE emulsion, dispersion\nof modified silica, sodium silicate, and modified ATP). The four kinds\nof emulsions that were mixed together can effectively solve the dispersity\nof waterborne superamphiphobic coatings. The dispersion was sprayed\nand cured at 310 °C for 15 min to obtain a surface with superior\nrepellency to water (CAs = 160 ± 0.7°, SAs = 4 ± 0.2°)\nand glycerol (CAs = 156 ± 0.9°, SAs = 10 ± 0.7°).\nThe PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP\ncomposite coating showed strong adhesive ability (grade 1 according\nto the GB/T9286), high hardness (6H), superior antifouling performance,\nexcellent impact resistance, high-temperature resistance (<415\n°C), and heat insulation. Remarkably, under a load of 200 g,\nthe composite coating still had remarkable repellency to water (CAs\n= 152 ± 0.8°, SAs = 11 ± 0.9°) and glycerol (CAs\n= 149 ± 0.7°, SAs = 22 ± 0.6°) after 140 cycles\nof wear. Furthermore, the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating was separately soaked in hydrochloric\nacid (pH 1) and potassium hydroxide (pH 14) for 120 h, and it still\nshowed excellent repellency to hydrochloric acid (pH 1) and potassium\nhydroxide (pH 14). With multifaceted robustness and scalability, the\nPTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite\ncoating possesses potential usage in impact resistance, heat insulation,\nhigh-temperature service, and chemical corrosion.",
"introduction": "1 Introduction Due to excellent repellency\nperformance for water and oil, 1 − 4 superamphiphobic coatings have wide application prospects\nin self-cleaning, 5 − 9 drag reduction, 10 , 11 anti-graffiti, 12 − 14 anticorrosion, 15 − 18 anti-icing, 19 − 21 petroleum pipeline, 22 etc.\nSuperamphiphobic surfaces with extremely excellent properties need\nto possess low surface energy and a hierarchical roughness structure. 23 , 24 Water-based superamphiphobic coatings are environmentally friendly\nand have attracted many researchers’ attention. 25 , 26 Nevertheless, the dispersion of hydrophobic substances in water\nis poor, which is not conducive to the performance of the coatings. 27 Therefore, improving the dispersibility is an\nimportant and critical factor for water-based superamphiphobic coatings\nto replace solvent-borne superamphiphobic coatings. 28 So far, many researchers have tried to solve the\nproblem of dispersion.\nAvijit et al. mixed water-dispersed clay sheets (6 wt %) with two\ndifferent functional silanes (1H,1H,2H,2H-perfluorooctyltriethoxysilane\nand 3-(2-aminoethylamino) propyltrimethoxysilane) and kept under vigorous\nstirring conditions for 6 h to obtain the coating dispersion. 29 Lozhechnikova et al. synthesized anionic wax\nparticles in combination with cationic ZnO nanoparticles by adjusting\npH to fabricate superhydrophobic coatings. 30 Furthermore, some works reported simple methods to develop aqueous\ndispersion. Silica nanoparticles were dispersed in an O/W emulsion\nand coated on different substrates to prepare aqueous superhydrophobic\nsurfaces. 31 − 33 However, there are few reports on the durability\nof superhydrophobic surfaces. Moreover, the dispersion of superamphiphobic\ncoatings is rarely discussed. The mechanical strength of a coating\nis extremely important, which\ndetermines the durability of the coating with a hierarchical roughness\nstructure. 34 Li et al. sprayed polyurethane\nand SiO 2 @HD-POS (hexadecylpolysiloxane-modified SiO 2 ) aqueous solution on a glass slide to prepare a superhydrophobic\nsurface. This coating owned self-repairing ability and wear resistance,\nbut the hardness of the coating was not mentioned. 35 Zhao et al. fabricated a superhydrophobic coating with\n6H by a silicone acrylate copolymer emulsion, and silica nanoparticles\nwere modified with 1H,1H,2H,2H-perfluorododecyltriethoxysilane in\nthe emulsion to air dry. But wear resistance has not been explored. 36 In addition, some works are attempted to solve\nthe durability of a coating with a hierarchical roughness structure,\nbut the methods are complex and damage the mechanical properties of\nthe substrate. 37 − 39 Therefore, it is particularly important to provide\na simple method to prepare superamphiphobic surfaces and consider\nthe overall properties of the coatings. In this paper, the dispersion\nof nanosilica was modified by cetyltrimethoxysilane\n(CTMS) to reduce coating surface energy and enhance adhesion. To solve\nthe dispersion and the durability of water-based coatings, the water-based\ndispersion was formed by the mechanical dispersion of the PTFE emulsion,\nmodified silica (SiO 2 @CTMS), sodium silicate, and modified\nATP. The four kinds of emulsions were mixed together to effectively\nsolve the dispersity of waterborne superamphiphobic coatings. Robust\nsuperamphiphobic surfaces with micro/nanohierarchical structures were\nsuccessfully fabricated by a simple one-step spraying method and curing\nat 310 °C for 15 min. Due to nanosilica lubrication and film\nhardness, the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating showed excellent wear resistance, which\ncan undergo 140 cycles of wear and still show high contact angles\n(CAs) and sliding angles (SAs) for water (CAs = 152 ± 0.8°,\nSAs = 11 ± 0.9°) and glycerol (CAs = 149 ± 0.7°,\nSAs = 22 ± 0.6°). Because of the refractories of modified\nsilica, sodium silicate, and modified ATP, 40 − 43 the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating has the characteristics\nof high-temperature resistance (within 415 °C) and insulation.\nFurthermore, it showed strong adhesive ability (grade 1), high hardness\n(6H), superior antifouling performance (100 cycles of slurry and glycerol),\nexcellent impact resistance, heat insulation, and anticorrosion (immersion\nof strong acid and alkali for 120 h). This work is expected to provide\na preparation strategy for multifunctional waterborne superamphiphobic\ncoatings with excellent properties and a simple method ( Figure 1 ). It not only enhances the\nhardness and wear resistance of the coatings but also takes into account\nimpact resistance, high-temperature resistance, heat insulation, and\nanticorrosion. Figure 1 Strategy of waterborne superamphiphobic coatings enhances\nthe hardness\nand wear resistance of the coatings and takes into account the impact\nresistance, high-temperature resistance, anticorrosion, and heat insulation.",
"discussion": "3 Results and Discussion 3.1 Characterization of Modified Silica Two kinds of modified silica particles (the average particle size\nwas 192 nm, and the average particle size was 360 nm) were synthesized.\nFrom the observation of the morphology of the two particle sizes,\nit was found that the modified silica with an average particle size\nof 192 nm was easier to agglomerate than the modified silica with\nan average particle size of 360 nm ( Figure 2 ). The composite coating was prepared by\n192 nm modified silica dispersion and 360 nm modified silica dispersion\nseparately. The CAs (glycerol, water) of the two kinds of particle\nsizes did not change significantly, but 360 nm modified silica showed\nrelatively high CAs ( Figure S3 ). Therefore,\nthe 360 nm particle size was used for the later experiment. Figure 2 Modified silica\nwith different particle sizes: (a, b) 192 nm modified\nsilica and (c–e) 360 nm modified silica. The surface of the modified silica particles was\nrelatively smooth\n( Figure 2 c). Cetyltrimethoxysilane\nwas uniformly coated on the surface of SiO 2 particles,\nwhich showed almost no agglomerated structure. The synthesized silica\nparticle had an obvious core–shell structure, and the shell\nthickness was about 20 nm ( Figure 2 d). The particle size was mainly distributed between\n345 and 375 nm (average particle size, 360 nm), showing that the size\ndistribution of modified silica was uniform ( Figure 2 e). Compared with the FT-IR spectra of silica\n( Figure 3 a), the FT-IR\nspectra of modified silica indicated that different absorption peaks\nwere located at 2855 cm –1 (symmetric tensile vibration\npeak) and 2924 cm –1 (antisymmetric tensile vibration\npeak), which belonged to methylene (CH 2 ) and methyl (CH 3 ), respectively. 44 , 45 The results testified\nthat cetyltrimethoxysilane had successfully modified silica. The mass\npercentage of silica-branched cetyltrimethoxysilane was measured by\nthermogravimetry. Compared with unmodified silica, the additional\nweight loss rate of modified silica was about 2.01% ( Figure 3 b). Figure 3 Characterization of modified\nsilica nanoparticles: (a) FT-IR spectra\nand (b) thermogravimetry of modified silica nanoparticles. 3.2 Wettability Analysis of the Coating The waterborne superamphiphobic coatings were prepared by spraying\nthe suspension of the filler/polymer with different proportions on\nthe surface of tinplate and curing at 310 °C for 15 min. Modified\nsilica increased the roughness of the coating and reduced the surface\nenergy. Therefore, modified silica was added to PTFE&Na 2 SiO 3 -ATP with different loadings varying from 15.52 to\n47.87 wt % to optimize the coating wettability. The water CAs were\ntransformed from 154 ± 0.8 to 160 ± 0.7° and the glycerol\nCAs from 153 ± 0.5 to 156 ± 0.9° as the modified silica\nfiller increased from 15.52 to 47.87 wt % ( Figure 4 a). The SAs of both water and glycerol droplets\nreduced gradually with increasing modified silica ( Figure 4 b). Because the composite coating\nhas a heterogeneous rough surface ( Figure 13 a) and low surface energy, making the composite\ncoating exhibit high CAs and low SAs. It conforms to the Cassie–Baxter\nmodel. 46 The droplets do not infiltrate\nthe rough structure and form a stable solid–liquid-vapor interface.\nThe composite coating shows excellent moisture resistance. A low solid–liquid\ncontact area fraction indicates high CAs and excellent hydrophobicity.\nThis is basically consistent with the results of the apparent contact\nangle ( Figure 4 a).\nWhen the content of modified silica in the coating system reaches\n52.43 wt %, the cross-linking curing of sodium silicate and modified\nATP is prevented to reduce the bonding strength of the coating, resulting\nin the cracking of the coating ( Figure S4 ). Generally considering, the coating has optimized superamphiphobicity\nwhen the modified silica filler is 35.53 wt %. It will be taken as\nthe further research object in the following and abbreviated as the\nPTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite\ncoating. Figure 4 Wettability of composite coating. (a, b) CAs and SAs of the composite\ncoating with different contents of modified silica nanoparticles to\nwater and glycerol. (c, d) CAs and SAs of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating to\ndifferent pH solutions. The repellency of corrosive droplets was tested,\nso as to comprehensively\nevaluate the scope of application of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. High CAs and low SAs\nfor the droplets in the range of pH 1–14 are shown in Figure 4 c,d. The composite\ncoating surface maintains superamphiphobicity for corrosive droplets.\nAccording to the results, it can be concluded that the prepared organic/inorganic\nhybrid superamphiphobic coatings have excellent liquid repellency. 3.3 Analysis of Morphology and Element The excellent performance of superamphiphobic coatings needs to meet\nthe requirements of the hierarchical roughness structure and extremely\nlow surface energy. Sodium silicate and the curing agent (modified\nATP) were added to the coating system, and the viscosity of coating\ndispersion was increased, which was easier to form aggregates to form\na rough structure in the process of spraying. After high-temperature\ncuring, a thickness of the composite coating of 40 μm was obtained\n( Figure S5a,c ), and the composite coating\nsurface had emulsion protrusions similar to the natural lotus leaf\n( Figure 5 a). The surface\nof the emulsion protrusion was embedded with modified silica, and\nthe rough morphology was similar to rice particles, which jointly\nformed nano/microhierarchical structures that intercepted air to form\na stable gas film ( Figure 5 b). 47 − 50 It was beneficial to fabricate the optimized superamphiphobic surfaces.\nAs displayed in Figure 5 c, elements were distributed on the surface of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. The\ncoating contained eight elements, among which fluorine was a low-surface-energy\nelement provided by PTFE. The composition analysis of the composite\ncoating is displayed in Figures S6–S8 and Table S1 . Figure 5 SEM of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. (a, b) SEM of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating\nand\nthe emulsion protrusion. (c) Element distribution of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. 3.4 Analysis of Antifouling Performance The most attractive feature of superamphiphobic coatings is the remarkable\npollution resistance. As shown in Figure 6 a,b, the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating was repeatedly soaked\nin a slurry (40 wt %) 100 times, and no sediment particles remained\non the coating surface (slurry dumping: Movie S1 ). Similarly, the coating was repeatedly immersed in glycerol\n100 times ( Figure 6 c,d), but the coating surface was intact, and no glycerol remained\n( Movie S2 , glycerol dumping: Movie S3 ). Furthermore, silica (50 nm) and sediment\nwere placed in the middle of the composite coating with an inclination\nof 30°, and a rubber dropper dropped deionized water from the\ntop of the sample plate. It was clearly observed that pollutants were\ntaken away quickly in the process of droplet falling ( Figure 6 e–h). The pollutant\nremoval method of the sample plate is the same as that of a lotus\nleaf, which proves that the surface of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating successfully imitates\nthe surface of a lotus leaf. Considering that the gas film on the\nsurface of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating will be renewed again during each coating\ndipping and pulling process, which will affect the judgment of antifouling\nperformance. Therefore, a more rigorous glycerol immersion test was\ncarried out to evaluate the antifouling performance of the coating.\nThe composite coating was soaked in glycerol for different times (2\ndays, 4 days, 6 days, 8 days, and 10 days) to measure the CAs and\nSAs (water, glycerol). It was worth mentioning that the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating\nstill had excellent repellency to glycerol (CAs = 151 ± 0.9°,\nSAs = 15 ± 1°) and water (CAs = 154 ± 2.7°, SAs\n= 9 ± 0.5°) after 10 days ( Figure 7 ). Figure 6 Self-cleaning and antipollution test. (a, b)\nSlurry (40 wt %) antipollution\ntest. (c, d) Glycerol antipollution test. (e, f) Self-cleaning test\nof 100 nm silica. (g, h) Self-cleaning test of silt. Figure 7 Glycerol immersion test. (a, b) CAs and SAs (glycerol,\nwater) of\nthe PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP\ncomposite coating soaked in glycerol for different times. 3.5 Analysis of High-Temperature Resistance The surface wettability after holding the composite coating at\ndifferent temperatures for 2 h is shown in Figure 8 . The PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating had excellent moisture\nresistance at 415 °C and maintained high CAs (water, glycerol).\nBut the SAs of glycerol were increased from 10 ± 0.3 to 15 ±\n0.4°. When the temperature reached 450 °C, the coating became\npowdered and lost superamphiphobic properties. The coating had moisture\nresistance in a wide temperature range on account of sodium silicate\nand modified ATP. 51 − 54 Figure 8 High-temperature\nresistance test. (a, b) CAs and SAs (water, glycerol)\nof the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating after treatment at different temperatures. 3.6 Analysis of Heat Transfer Performance In some application scenarios, superamphiphobic coatings require\nexcellent thermal insulation, such as the shell wall of the heat exchanger.\nWith the increase of the modified nanosilica filler, the thermal conductivity\nof the coating gradually decreased from 0.02 to 0.012 W/(m·k)\n( Figure 9 a). Figure 9 b shows the random\nparticle distribution of the composite coating with 47.87 wt % modified\nnanosilica. Due to the increase of the content of modified nanosilica\nwith low thermal conductivity, the path of thermal conductivity is\nformed in the coating to delay the heat conduction, which is beneficial\nfor the low thermal conductivity of the coating. 55 It proved the existence of the heat conduction path by\nfinite element analysis. The physical parameters of the model material\nare shown in Table S2 . By the heat flux\ndistribution, it is clearly observed that with the increase of the\ncontent of low thermal conductive particles, it is easier to form\na connected heat flux path inside the model ( Figure 10 d–f). In addition, the heat flux\nvalue of the modified nanosilica filler is lower than those of other\nparts. As shown in Figure 10 a–c, when the content of low thermal conductive particles\nincreases from 0 to 47.87%, it slows down the heat transfer rates\nand increases the temperature difference to 1.2 °C on two sides\nof the model. This slowing down of the heat transfer process is observed\nby infrared thermography ( Figure 9 c), and the filler with low thermal conductivity has\na certain effect on reducing the thermal conductivity of the coating.\nAccordingly, the results indicate that the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating is suitable for heat\ninsulation. Figure 9 Thermal conductivity and temperature distribution. (a) Thermal\nconductivity of the coatings with different contents of modified silica\nnanoparticles. (b) Heat conduction mechanism of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. (c)\nInfrared thermography. Figure 10 Finite element analysis. (a–c) The temperature\ndistribution.\n(d–f) The heat flux distribution. 3.7 Analysis of Mechanical Performance The mechanical performance of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating contains wear resistance,\nadhesion, and impact resistance. The wear resistance of the prepared\ncomposite coating was evaluated by the sandpaper abrasion test. The\nrough surface of sandpaper (800 mesh) contacted the coating surface,\nand the sandpaper was pulled back and forth, which was recorded as\none time ( Figure 11 ). After 10 cycles of the test, the PTFE/SiO 2 @CTMS composite\ncoating lost superamphiphobicity for glycerol (CAs = 149 ± 1°,\nSAs = 19 ± 0.5°) and water (CAs = 152 ± 1.5°,\nSAs = 11 ± 0.6°). Furthermore, after 20 cycles, the SAs\nof the PTFE/SiO 2 @CTMS composite coating increased rapidly\nfrom 6 ± 0.8 to 16 ± 1° for water and from 11 ±\n0.6 to 24 ± 1.8° for glycerol ( Figure 12 a,b). The hardness of the PTFE/SiO 2 @CTMS composite coating is regarded as 2B ( Figures S9 and S10a,b ). Even though the PTFE/SiO 2 @CTMS composite\ncoating possessed the wear-resistant lubrication effect of nanosilica,\nit was unable to match the suitable hardness of the coating to cannot\nresist the cyclic damage of the rough surface of sandpaper many times.\nHowever, compared with the PTFE/SiO 2 @CTMS composite coating,\nthe PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP\ncomposite coating still had high repellency to water (CAs = 152 ±\n0.8°, SAs = 11 ± 0.9°) and glycerol (CAs = 149 ±\n0.7°, SAs = 22 ± 0.6°) after 140 cycles ( Figure 12 c,d). The wear\nresistance of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating is about 14 times that of the PTFE/SiO 2 @CTMS composite coating. The excellent mechanical robustness\nof the composite coating was due to the introduction of sodium silicate\nand modified ATP, which made the hardness of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating reach\n6H ( Figure S10c,d ). Meanwhile, the modified\nnanosilica of wear-resistant lubrication was tightly embedded in the\nfilm surface ( Figure 5 b), and it was not easy to be taken away by the rough surface of\nsandpaper when sliding. 56 , 57 High hardness and lubrication\npromote the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating to show excellent wear resistance and durable\nmoisture resistance. As shown in Figure 13 c,d, the composite\ncoating surface was not damaged in a large area, and some emulsion\nprotrusions were slightly worn. According to the comparison of the\n3D surface morphology between the composite coating and the worn composite\ncoating (after the wear resistance test of 140 cycles; Figure 13 a,b), the arithmetic mean\ndeviation of the profile (Ra) had little difference. It was basically\nconsistent with the SEM of the composite coating before and after\nabrasion. The results indicated that the coating had excellent wear\nresistance. Figure 11 Process of the coating wear resistance test. Figure 12 Abrasion resistance test. (a, b) CAs and SAs (glycerol,\nwater)\nof the PTFE/SiO 2 @CTMS composite coating after different\ncycle wear times. (c, d) CAs and SAs (glycerol, water) of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating\nafter different cycle wear times. Inset: optical images of water,\npH 1, pH 14, and glycerol droplets on the coating after abrasion. Figure 13 Morphological analysis of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. (a) 3D surface morphology\nimages of the composite coating. (b) 3D surface morphology images\nof the composite coating after the wear resistance test. (c, d) SEM\nof the composite coating and the emulsion protrusion after the wear\nresistance test. The practical application of superamphiphobic coatings\nis inseparable\nfrom excellent adhesion and impact resistance. The coating adhesion\nis used to evaluate the adhesion index between the film and the substrate.\nFurthermore, the impact resistance is regarded as the ability to resist\nexternal deformation and cracking, which reflects the adhesion of\nthe coating for the side. According to the paints and varnishes—cross-cut\ntest ( Figure 14 f),\nto evaluate film adhesion, the surface of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating was\nmarked as a 2 mm × 2 mm small square by a wallpaper knife to\npress the surface of the composite coating with 3M test grade tape\nfor 3 min and then tore it quickly (repeated three times). It was\nclearly observed that the surface of the composite coating was not\ndamaged before and after the adhesion of 3M test grade tape, which\nwas classified as grade 1 adhesion ( Figure 14 a,b). The firm adhesion of the composite\ncoating is mainly attributed to the interlocking effect of the network\nstructure of PTFE, modified silica, sodium silicate, and modified\nATP. Moreover, the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating still had superamphiphobicity after\nthe cross-cut test ( Figure 14 c). Marks were left on the surface of PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating at\ndifferent heights ( Figure 14 d,e). With the increase of height (15–45 cm), deeper\nmarks were left on the coating surface, but the surface of the composite\ncoating was not cracked. It is indicated that the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating has\nexcellent cracking resistance and adhesion. Figure 14 Test of the mechanical\nproperties of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating. (a, b) PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating\nbefore and after the cross-cut tape test. (c) Optical images of water,\npH 1, pH 14, and glycerol droplets on the coating after the cross-cut\ntape test. (d, e) The composite coating was impacted by a heavy hammer\nwith different heights. (f) GB/T 9286. 3.8 Analysis of Corrosion Resistance For the sake of exploring the corrosion resistance of the superamphiphobic\ncoatings from a macroeconomic perspective, acid–base immersion\nexperiments were conducted. The PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating was soaked in solutions\nof hydrochloric acid (pH 1) and potassium hydroxide (pH 14) for different\ntimes (24, 48, 72, 96, and 120 h). The wettability of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating\nwas measured for hydrochloric acid (pH 1) and potassium hydroxide\n(pH 14) so as to evaluate the repellency of corrosive droplets. It\ndisplayed that the composite coating was soaked in hydrochloric acid\n(pH 1) after 120 h immersion, and the CAs of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating were\nstill greater than 150°, which were reduced from 158 ± 1.6\nto 152 ± 1° for hydrochloric acid and from 156 ± 0.9\nto 151 ± 1° for potassium hydroxide. The SAs were less than\n10°, which were increased for hydrochloric acid (5 ± 0.5\nto 8 ± 0.7°) and potassium hydroxide (6 ± 0.2 to 9\n± 0.5°, Figure 15 a,b). The changing trend of the test results soaked in potassium\nhydroxide (pH 14) was basically the same as that soaked in hydrochloric\nacid (pH 1) after 120 h ( Figure 15 c,d). The long-term macroscopic excellent liquid repellency\nwas closely related to the microscopic morphology. By observing the\nmorphology of the coating after immersing in hydrochloric acid and\npotassium hydroxide for 120 h, it was found that the overall morphology\nof the coating was not damaged, and it was consistent with the morphology\nbefore immersion ( Figure 16 ). It was indicated that hydrochloric acid (pH 1) and potassium\nhydroxide (pH 14) did not damage the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating with excellent corrosion\nresistance within 120 h. Therefore, the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating has excellent chemical\nstability in harsh corrosive environments. Figure 15 Acid–base immersion\ntest. (a, b) CAs and SAs (pH 1, pH 14)\nof the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating soaked in a pH 1 solution for different times.\n(c, d) CAs and SAs (pH 1, pH 14) of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating soaked in a pH 14 solution\nfor different times. Figure 16 SEM of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating after the acid–base\nimmersion\ntest. (a, b) SEM of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating after soaking in a pH 1 solution\nfor 120 h. (c, d) SEM of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating after soaking in a pH\n14 solution for 120 h. To further analyze the corrosion resistance of\nthe PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP\ncomposite coating,\nthe corrosion prevention efficiency of the composite coating in sodium\nchloride solution (3.5 wt %) was studied. The typical potentiodynamic\npolarization curve was obtained ( Figure 17 a). The electrochemical parameters were\nobtained from the polarization curve ( Table S3 ). Electrochemical parameters include corrosion potential ( E corr ), corrosion current density ( I corr ), corrosion rate (C.R.), and protection efficiency\n(P.E.). Compared with pure tinplate, the corrosion current densities\nof PTFE/SiO 2 @CTMSand PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coatings were reduced to 1.256\n× 10 –5 and 1.012 × 10 –5 A/cm 2 , respectively. Moreover, the corrosion potential\nof the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating has achieved a significant positive shift\nto −0.23 V. The Nyquist spectra of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating have\na large semicircle diameter, corresponding to a lower corrosion rate\n( Figure 17 b). Therefore,\ncorrosion resistance is the best. The electrochemical impedance spectra\nresults are consistent with the polarization curve results, both of\nwhich show that the corrosion resistance of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating is\nsuperior to that of the PTFE/SiO 2 @CTMS composite coating.\nThe prepared superamphiphobic coatings had corrosion resistance, which\nwas related to the internal structure of the coating and the formation\nof cavitation on the surface. The surface of the PTFE/SiO 2 @CTMS&Na 2 SiO 3 -ATP composite coating had\nmany protrusions with micro/nanostructures similar to the lotus leaf\nsurface. These protrusions with micro/nanostructures intercepted air\nand formed stable cavitation, which prevented the composite coating\nsurface from being wet and prevented corrosive substances from corroding\nthe substrate ( Figure 17 c). The modified silica as a physical barrier inside the coating\nprolonged the time for corrosive substances to pass through the coating\nand reach the substrate. The physical barrier and stable cavitation\nmake the film have excellent corrosion resistance. Hence, the ability\nof a composite coating to stand up to the severe corrosive environment\nis conducive to its application in a practical factory environment.. Figure 17 Corrosion\nresistance test. (a) Potentiodynamic polarization curves.\n(b) Nyquist plots (c) Anticorrosion principle."
} | 7,540 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.