pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
39338472 | PMC11434368 | pmc | 4,914 | {
"abstract": "Some sulfate-reducing bacteria (SRB), mainly belonging to the Desulfovibrionaceae family, have evolved the capability to conserve energy through microbial extracellular electron transfer (EET), suggesting that this process may be more widespread than previously believed. While previous evidence has shown that mobile genetic elements drive the plasticity and evolution of SRB and iron-reducing bacteria (FeRB), few have investigated the shared molecular mechanisms related to EET. To address this, we analyzed the prevalence and abundance of EET elements and how they contributed to their differentiation among 42 members of the Desulfovibrionaceae family and 23 and 59 members of Geobacteraceae and Shewanellaceae , respectively. Proteins involved in EET, such as the cytochromes PpcA and CymA, the outer membrane protein OmpJ, and the iron–sulfur cluster-binding CbcT, exhibited widespread distribution within Desulfovibrionaceae . Some of these showed modular diversification. Additional evidence revealed that horizontal gene transfer was involved in the acquiring and losing of critical genes, increasing the diversification and plasticity between the three families. The results suggest that specific EET genes were widely disseminated through horizontal transfer, where some changes reflected environmental adaptations. These findings enhance our comprehension of the evolution and distribution of proteins involved in EET processes, shedding light on their role in iron and sulfur biogeochemical cycling.",
"introduction": "1. Introduction Sulfate- and iron-reducing prokaryotes (SRPs and FeRPs) are microorganisms that play a vital role in the biogeochemical cycles of sulfur and iron, two essential elements for the functioning of life on earth [ 1 , 2 ]. A vast body of research highlights the impact of the interaction of these two groups across many ecosystems, including anaerobic soils and sediments, pristine or contaminated freshwater, groundwater and marine environments, and even intestinal systems [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 ]. The interactions between both functional groups have been widely reviewed in marine environments, in which their activity accounts for most of the anaerobic organic matter degradation in sediments at the global level [ 15 , 16 , 17 , 18 , 19 ]. Currently, marine sulfate (~29 mM) is the largest mobile sulfur reservoir on our planet, an oxidizing pool even greater than atmospheric oxygen [ 20 , 21 ]. The mechanical and chemical weathering of continental rocks releases sulfur into the seawater column, which, in turn, continuously supplies sulfate to marine snow and sediments, where it is reduced by SRPs [ 16 , 20 , 22 ]. Analogously, iron enters the ocean from different sources, mainly remaining as a redox-active element in sediments, where FeRPs use it as an electron acceptor [ 3 , 23 , 24 ]. H 2 , formate, acetate, and other volatile fatty acids produced by hydrolysis or fermentation are used as electron donors for both dissimilatory processes, making the biogeochemical cycling of carbon, iron, and sulfur tightly linked [ 10 ]. Major determining factors contribute to the balance of how iron and sulfate reduction are spatially and temporally organized [ 25 ]. First, there is a competition for electron donors that was initially addressed by studying the concept of competitive exclusion [ 11 , 26 , 27 , 28 , 29 , 30 ]. Second, these biogeochemical processes are constrained to those metabolisms that include dissimilatory pathways capable of interacting with insoluble (iron) and soluble (sulfate) electron acceptors and by the free energy released by each redox reaction. While dissimilatory iron-reducing metabolisms, which demand that electrons must be effectively transported from cytoplasmic donors to extracellular space and generate higher free energy, dominate surface environments where Fe (III) is available, reductions in sulfate, yielding less energy, are restricted to deeper layers of sediments [ 3 , 23 , 31 ]. Third, several chemical reactions regulate the bioavailability or toxicity of both substrates and by-products. For instance, Fe(II) produced by dissimilatory iron reduction tends to diffuse to an oxic/anoxic interface, where it is reoxidized back to Fe(III) [ 23 , 32 ]. In contrast, Fe(III) works as an oxidant for sulfide that is produced by deeper SRPs, producing iron sulfide (FeS) and, lately, pyrite [ 33 ], which competes with the incorporation of sulfide into organic matter [ 34 , 35 , 36 ]. Based on this evidence, it was initially assumed that iron and sulfate reductions occur in discrete zones [ 27 , 37 , 38 ]. However, more recent observations have challenged this notion, revealing that both processes can coexist simultaneously [ 3 , 13 , 14 , 39 ] and even interplay along different stages of mineral transformation [ 40 ]. Phylogenetically diverse species of SRPs can reduce Fe(III) and Mn(IV) as well as electrodes of bioelectrochemical systems, suggesting that these capabilities may be widespread in various clades, although few of them can conserve energy to support growth [ 23 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 ]. Several SRP strains have been involved in the corrosion of Fe-containing metals by a combination of different mechanisms, suggesting many possible pathways of interactions with extracellular electron donors/acceptors [ 49 , 50 ]. In addition, comparative genomic studies have shed light on the lasting role that genetic mobile elements play in the plasticity and evolution of genomes of members of the Desulfovibrionaceae , Geobacteraceae , and Shewanellaceae families [ 51 , 52 , 53 , 54 , 55 , 56 ]. Recent investigations have reported that genes encoding transmembrane electron conduits in Shewanella , MtrCAB and OmcA, have been disseminated through horizontal gene transfer within the same species [ 57 ], genus [ 58 ], and across distantly related genera [ 59 ], suggesting that the mobilome may play a role in acquiring sophisticated metabolic processes such as extracellular electron transfer (EET). This juxtaposition led to the hypothesis that as a result of co-localization, collaboration, and competition, as well as their metabolic plasticity, several SRPs may have also evolved the capability to reduce insoluble Fe(III) oxides coupled with the oxidation of low-concentration electron donors. The Geobacteraceae and Shewanellaceae families are FeRPs that are well known for their EET capabilities. The mechanisms of their model bacteria, G. sulfurreducens PCA and S. oneidensis MR-1, have been extensively studied by combining genomic, transcriptomic, and proteomic approaches coupled with functional genetic experiments, and they have been expanded to other strains [ 60 , 61 , 62 , 63 , 64 , 65 ]. Based on these well-studied mechanisms, we conducted a search to understand the prevalence of the orthologs of EET-related proteins in forty-two genomes of SRB belonging to the Desulfovibrionaceae family. The analysis included 130 proteins that encode c -type cytochromes, such as porin–cytochrome complexes (Pcc), b -type cytochrome complexes (Cbc), riboflavin biosynthesis genes, chemotaxis-related genes, and cell membrane components, and explored the mobilome of the three groups to assess the impact of horizontal gene transfer (HGT) on the acquisitions and losses of genes critical for EET by members of the Desulfovibrionaceae family. Through phylogenetic and orthology analyses, we identified key proteins involved in EET that are widely distributed across these three families. Most of these proteins display modular diversification, serving as components of multiple complexes engaged in respiratory mechanisms. We believe this pool may include the essential core features required for the proper physiological functioning of EET and could constitute a crucial aspect of evaluating its expansion to other SRPs. These findings enhance our understanding of the evolution and distribution of proteins related to extracellular electron transfer processes, shedding light on their role in microbial communities actively participating in iron and sulfur biogeochemical cycles.",
"discussion": "3. Results and Discussion 3.1. Phylogenomic Analysis of Sulfate-Reducing and Iron-Reducing Metabolism To identify the shared genomic features between SRPs and FeRPs, a phylogenomic analysis including 124 genomes belonging to the Desulfovibrionaceae , Geobacteraceae , and Shewanellaceae families was reconstructed using 109 single-copy orthogroups ( Figure 1 ). The analysis revealed that out of a total of 458,646 genes, 97.7% (448,320) were assigned to 15,156 orthologous groups (OGs). Of these, 289 OGs were present in all species, while 546 OGs were specific to certain species ( Figure S1 ). The remaining unassigned genes and species-specific orthogroups represented the unique genetic traits of each species. It was also found that the Desulfovibrionaceae , Geobacteraceae , and Shewanellaceae families share 2372 species-shared OGs and have 3300, 1737, and 4707 specific OGs, respectively. Interestingly, the Desulfovibrionaceae family shares a greater number of exclusive OGs with the Shewanellaceae family (1301) compared to the Geobacteraceae family (1088) ( Figure S1 ). The resulting phylogenomic reconstruction revealed three major clades, each corresponding to each family. The Desulfovibrionaceae clade includes 42 genomes, most of which have been isolated or recovered from pristine and contaminated ecosystems, as well as from a broad range of aquatic environments, from marine sediments to freshwater ( Tables S1 and S3 and Figure S2A ). The Desulfovibrionaceae clade contains strains with an average genome size of 3.9 Mbp and a G+C content of 60.8%, with the smallest genome corresponding to Desulfovibrio piger ATCC 29098 (2.87 Mbp) and the largest to Desulfovibrio inopinatus DSM 10711 (5.77 Mbp). The Desulfovibrionaceae clade is divided into two distinct subclades. Subclade I mainly comprises strains living in marine and estuarine ecosystems, whereas subclade II comprises strains from freshwater and engineered ecosystems ( Figure S2B ). The Geobacteraceae clade includes 23 species of the Geobacter genus, a group of genomes belonging to species that have been isolated or recovered from soil and freshwater sediments as well as polluted sites, where Geobacter species play an important role in the regulation of biogeochemical cycles [ 215 , 216 , 217 , 218 , 219 ]. The Geobacteraceae clade contains strains with an average genome size of 4.0 Mbp and a G+C content of 58.3%. The smallest genome of this clade is Geobacter benzoatilyticus Jerry-YX (3.58 Mbp), and the largest genome is Geobacter uraniireducens Rf4 (5.14 Mbp). The Geobacteraceae clade is divided into two subclades. Geobacter strains isolated/found in soils, sediments, groundwater, and engineered environments belong to subclade I, while strains isolated from polluted sites and freshwater ecosystems belong to subclade II ( Figure S2C ). The third clade comprises 59 species of the Shewanella genus, with genomes with remarkably low values of G+C content (45.3%) and an average genome size of 4.8 Mbp. The Shewanella clade includes genomes ranging from 3.9 Mbp to 6.4 Mbp of Shewanella aestuarii JCM 17801 and Shewanella psychrophila WP2, respectively. Unlike the Desulfovibrionaceae and Geobacteraceae clades, 56% (33) of the species belonging to this clade have been isolated or recovered from marine ecosystems, and secondly, from samples derived from soils and sediments (subclade I) as well as freshwater and human- and animal-associated environments (subclade II), where Shewanella has been recently found ( Figure S2D ) [ 52 , 220 ]. As anticipated, our findings revealed variations in genomic traits, including GC content and genome size, among genomes from different subclades ( Figure S3 ). In agreement with previous studies, the genomic GC content of strains of Shewanellaceae , a family within the class Gammaproteobacteria , was found to be lower than that of Desulfovibrionaceae and Geobacteraceae , which belong to Deltaproteobacteria [ 221 ]. Also, previous evidence has shown that genomes with higher GC content have more N in their proteomes [ 222 ]. Therefore, the lower GC content of Shewanellaceae strains may also be partially explained by the fact that several strains are primarily found in the ocean, an environment with a persistent limitation of N [ 223 ]. Another study reported that in the genome of Desulfovibrio vulgaris , mutations that convert GC to AT (GC->AT) were the most common, suggesting that a loss of GC content in this genome is slowly occurring [ 224 ]. In total, 34 strains out of the 124 surveyed had been implicated in some form of electron transfer to extracellular compounds, most of which belong to Geobacteraceae and Shewanellaceae clades. Both families have been the focus of a great extent of experimental evidence regarding their capabilities of extracellular electron transfer, which mainly relies on two types of mechanisms for electron transport across the outer surface. In Shewanella strains, substances that act as electron shuttles allow electrons to be transported from an intracellular enzymatic complex to the extracellular electron acceptor. This is the case with Shewanella oneidensis MR-1, which secretes small redox-active molecules for electron shuttling back and forth between cells and external electron acceptors [ 225 ]. In contrast, direct EET, which is prevalent in Geobacter strains, depends on the availability of redox-active enzymes and conductive appendages attached to the outer surfaces of the cells [ 226 ]. Four strains of the Desulfovibrionaceae clade, including Maridesulfovibrio frigidus DSM 17176 [ 46 ], Desulfocurvibacter africanus PCS [ 45 ], Desulfovibrio vulgaris str. Hildenborough [ 227 ], and Desulfovibrio desulfuricans DSM 642 [ 41 ], were shown to reduce Fe(III) and use it as an electron acceptor under experimental conditions. The extent to which this process takes place under environmentally relevant conditions as well as their molecular mechanisms remain to be explored. 3.1.1. Abundance of Multi-Heme Cytochromes Multi-heme c -type cytochromes ( c -Cyts) are proteins that harbor three or more hemes that have a central coordinated Fe atom that allows for the transfer of electrons. In species like Shewanella oneidensis MR-1 and Geobacter sulfurreducens PCA, c -Cyts play a fundamental role in EET to solid metal (hydro)oxides [ 228 , 229 , 230 ] as well as direct interspecies electron transfer [ 231 , 232 ]. c -Cyts are also very abundant in Desulfovibrionaceae [ 233 ]. To assess the diversity and prevalence of c -Cyts, we searched for the motifs CXXCH and CXXXCH across all strains, revealing a total of 9800 such proteins. On average, members of Geobacteraceae have 125.7 heme-containing proteins per genome, whereas Shewanellaceae and Desulfovibrionaceae have 77.1 and 56.1, respectively. The CXXCH motif was more common than the CXXXCH motif, representing between 84.4% and 94.3% of all proteins with heme motifs ( Figure 2 A). Interestingly, some c -Cyts exhibited both motifs, though this was more prevalent in Geobacter proteins (9.6% of proteins exhibited both motifs) and rare in Shewanella . Geobacter strains had an average of 10.6 extracellular predicted c -Cyts per genome, compared to 1.1 c -Cyts per genome of the Shewanellaceae family, and none of the Desulfovibrionaceae ( Figure 2 B). Regarding the number of motifs found per protein, most contain only one or two motifs. This ranges between 52.5% in the Geobacteraceae family and 77.5% in the Desulfovibrionaceae family. The abundance of multi-heme cytochromes in Geobacteraceae strains is particularly interesting, with an average of 59.7 per genome, which is significantly higher than other families, which contain 12.6 and 24.6 multi-heme proteins per strain, respectively ( Figure 1 and Figure 2 C). Some Geobacter strains, such as G. uraniireducens Rf4, G. sp. OR-1, G. daltonii FRC-32, and G. sp. DSM 9736, contain proteins with more than 40 heme motifs, whereas Shewanella and Desulfovibrio contain significantly less. In terms of cellular localization, it has been predicted that 9.5% of Geobacter ’s multi-heme cytochromes and 0.7% of Shewanella ’s are likely to be secreted from the cell. On the other hand, Desulfovibrionaceae strains do not seem to have extracellular multi-heme proteins or contain multi-heme proteins associated with the outer membrane ( Figure 2 ). These findings agree with previous reports, where Geobacter species, such as G. sulfurreducens , encode many c -type cytochromes in their genomes compared to Shewanella and Desulfovibrio [ 64 , 233 ]. 3.1.2. Similarity Network Analysis of Extracellular Multi-Heme Cytochromes A similarity network analysis was conducted to determine the phylogenetic relationships between the predicted multi-heme cytochromes located outside of cells. The analysis found 1807 sequences associated with 35 OGs (including some sequences that had not passed the extracellular localization filter). Geobacter strains had the majority of the sequences (918), followed by Shewanella strains (764) and Desulfovibrionaceae family strains (125). The similarity network had 130 sets of highly interconnected nodes (E-value threshold of 10 −40 ), known as communities, with 71 containing two or more nodes ( Figure 3 ). The network clusters exhibited a clade-specific pattern, indicating closer relationships based on their family of origin. Despite no clear correlation between the isolation source and clusters, it is evident that cluster proteins related to Shewanella strains are adapted to high-salinity conditions due to their origin from marine sources ( Figure S15 ). Some cytochromes involved in EET, including OmcA and MtrC from S. oneidensis MR-1, and OmcA, OmcS, OmcZ, and CbcA from G. sulfurreducens PCA, are exclusively grouped with cytochromes of the same family. However, the OmcI cytochrome of G. sulfurreducens PCA, and the DsmE and MtrA cytochromes of S. oneidensis MR-1, were grouped in the same cluster along with other cytochromes of the Shewanellaceae and Geobacteraceae families. On the other hand, the Desulfovibrionaceae family presents nodes related to six clusters, three of which contain at least two cytochromes belonging to Desulfovibrionaceae strains capable of Fe reduction. The first of these clusters (community N#37, Figure 3 B) is comprised exclusively of proteins of this family, whose products correspond to cytochrome c family proteins containing ten heme motifs. The second cluster ( Figure 3 C) contains 65 sequences (community N#33) encoding for a cytoplasmic membrane-bound cytochrome ubiquinol oxidase subunit I or c -type cytochromes. Both clusters related to OG 517 and OG 1638 also contain some Geobacter and Shewanella cytochromes predicted to be extracellular, and therefore, exploring their function in future studies may be interesting. 3.1.3. Comparative Genomic Analysis of Genes Related to Extracellular Electron Transfer Mechanisms Since several molecular mechanisms for which members of the Geobacteraceae and Shewanellaceae families interact with extracellular electron acceptors have been widely described, we analyzed the dataset of Desulfovibrio genomes to learn about the ubiquity and abundance of proteins involved in EET ( Figure 4 ). To conduct the analysis, we identified EET-related proteins in each genome by checking for their presence in the corresponding orthologous groups (OGs) from S. oneidensis MR-1 or G. sulfurreducens PCA. We evaluated 130 genes that encoded both outer-membrane and periplasmic c -type cytochromes, genes encoding porin–cytochrome complexes (Pcc) and b -type cytochrome complexes (Cbc), riboflavin biosynthesis genes, genes related to chemotaxis, and cell membrane components ( Table S2 ). Similarities with the EET Mechanism of G. sulfurreducens Investigations focused on Geobacter models, such as G.sulfurreducens and G. metallireducens , contributed to the comprehension of EET through multiple respiratory pathways [ 64 , 215 , 234 ]. These mechanisms mainly involve c - and b -type cytochromes in the inner membrane, ImcH and CbcL. The deletion of these genes impaired the ability to reduce electron acceptors with potentials above and below −0.1 V versus the standard hydrogen electrode (SHE) [ 235 , 236 ]. While both proteins have orthologs in all tested Geobacter species, only four Desulfovibrionaceae strains have CbcL homologs, including M. frigidus DSM 17176, a strain capable of Fe(III) reduction, but incapable of producing enough energy to support growth [ 46 ] ( Figure 4 ). The genome of G. sulfurreducens contains four gene clusters encoding inner-membrane cytochromes, including Cbc3 ( cbcVWX ), Cbc4 ( cbcSTU ), Cbc5 ( cbcEDCBA , where cbcC is also known as omcQ ), and Cbc6 ( cbcMNOPQR ), that play a role in EET [ 64 ]. The deletions of cbcV and cbcBA resulted in a considerable decrease in Fe(III) reduction, and a transcriptional study found that cbcT was upregulated on insoluble metal oxides versus Fe(III) citrate [ 64 , 237 , 238 ]. All these gene clusters are conserved and widely distributed in all Geobacter species [ 239 ], except for the Cbc6 cluster, which is incomplete in seven strains, mainly belonging to Geobacter subclade I ( Figure 5 ). The CbcOP subunits are CbcVW orthologous proteins from the Cbc3 cluster and are present in all Geobacter and Shewanella strains but only in a few strains of the Desulfovibrionaceae family. Among the periplasmic cytochromes, PpcA, MacA, and PccF have been characterized in more detail. It has been proposed that PpcA transfers electrons from the cytoplasmic membrane to the outer membrane, while MacA acts as a hydrogen peroxide reductase and transfers electrons to PpcA [ 240 , 241 ]. The expression of the gene encoding PccF was upregulated during growth on insoluble metal oxides, suggesting a possible role during EET [ 237 ]. Homologs of these three genes were heterogeneously distributed across the members of the three families. While several genes encoding PpcA were highly abundant in Geobacter (average of 4.9 genes per genome) and Desulfovibrio (average of 3 genes per genome), the gene encoding MacA was mostly shared between Shewanella and Geobacter strains ( Table S2 ). A porin–cytochrome complex (Pcc) capable of transferring electrons across a liposomal membrane is encoded by a periplasmic c -type cytochrome (OmaB/C), a porin-like protein (OmbB/C), and a reductase (OmcB/C). The Pcc protein complex reduces ferric citrate and ferrihydrite, similar to the MtrABC complex in S. oneidensis [ 242 ]. Three additional gene clusters encoding putative “electron conduits” involved in EET, including the porin–cytochrome (Pcc) complex extABCD, the porin–cytochrome (Pcc) complex extEFG, and the porin–cytochrome (Pcc) complex extHIJKL [ 243 ], were found to be highly prevalent in Geobacter species. However, no homologs were found in the Shewanellaceae and Desulfovibrionaceae family members ( Figure 5 ). A similar distribution was found in the plethora of multi-heme c -Cyts associated with the outer membrane, including OmcS, OmcZ, OmcE, OmcT, and PgcA, which play different roles along the EET for both Fe (III) and Mn(IV) oxide reduction and electrode respiration ( Figure 5 and Figure S4 ) [ 237 , 244 , 245 , 246 , 247 , 248 , 249 ]. In contrast, genes encoding the outer-membrane c -Cyts OmcI, a homolog of the CbcA subunit of G. sulfurreducens , and the outer-membrane protein J (ompJ), a channel known to influence the quantity and localization of cytochromes in the outer membrane [ 250 ], were found to be present in all the strains of Geobacteraceae and Desulfovibrionaceae , and partially in Shewanellaceae strains. Xap, an extracellular anchoring polysaccharide protein, has a crucial role in metal oxide attachment, cell–cell agglutination, and localization of essential c -Cyts. It possesses averages of 35, 29, and 23, high numbers of homologs, in the Desulfovibrionaceae, Geobacteraceae , and Shewanellaceae families, respectively [ 251 ]. Although recent evidence has highlighted the role of secreted riboflavins during DIET in Geobacter cocultures [ 252 ], this mechanism was not added to our analysis. Similarities with the EET Mechanism of Shewanella oneidensis EET is mediated by CymA, a six multi-heme c -Cyts, in S. oneidensis MR-1. CymA oxidizes quinol in the cytoplasmic membrane and transfers electrons to Fcc3 and STC, which transport electrons to MtrA [ 253 , 254 , 255 ]. MtrA, MtrB, and MtrC form a trans-outer-membrane complex to transfer electrons to the bacterial surface. MtrC and OmcA can physically interact with each other and transfer electrons directly to Fe(III) minerals [ 256 , 257 , 258 ], as well as associate with extracellular structures that were previously referred to as ‘nanowires’ [ 259 ]. S. oneidensis MR-1 employs endogenously produced flavin electron shuttles to enhance EET to minerals and electrodes during anaerobic respiration. Thus, released flavins are proposed to function as diffusive electron shuttles that transport electrons from MtrC and OmcA to mineral surfaces [ 225 , 260 , 261 ]. Homologs of CymA were found ubiquitously in all three families analyzed, with the exception of genomes of strains belonging to Geobacter subcluster I. In contrast, homologs of OmcA were found to be distributed in almost all the species of Shewanella and a few of Geobacter ( Figure 5 ). Whereas Fcc3 homologs were found mainly in Shewanella strains, homologs of cctA, the gene coding the tetraheme STC, were shared by Shewanella strains and all strains of Desulfovibrio desulfuricans ( Figure 5 and Figure S4 ). In addition to that, one guanosine triphosphate (GTP) and two ribulose-5-phosphate molecules are converted into one riboflavin molecule in a stepwise manner by the enzymes encoded by the ribA, ribB, ribD, ribH, and ribE genes [ 262 ]. This pathway seems to be ubiquitous in Shewanella species, although homologs of some of these genes, specifically ribB, ribD, and ribH, are also found in all species belonging to the Geobacteraceae and Desulfovibrionaceae families, which could imply some role of these molecules in their EET mechanisms. Further experimental research is required to investigate if flavins play a relevant role during EET by SRPs, as the addition of riboflavin and flavin adenine dinucleotide (FAD) showed the accelerated corrosion of carbon steel and stainless steel by D. vulgaris [ 263 , 264 ]. 3.2. Mobilome Analyses across Members of Desulfovibrionaceae, Geobacteraceae, and Shewanellaceae Families The genetic makeup of prokaryotic genomes is composed of DNA fragments from both vertical and horizontal gene transmission. The mobilome, a collection of mobile genetic elements, facilitates the transfer of genes and their corresponding functions within a community through horizontal gene transfer (HGT) [ 265 ]. Our analysis reveals that these SRP and FeRP families host various mobile genetic elements, such as plasmids, bacteriophages, integrons, insertion sequences (ISs), and integrative and conjugative elements (ICEs). According to a PHASTER search, 333 prophages were found in total, with 59 intact, 244 incomplete, and 30 questionable phages present in 122 strains ( Figure 6 A). The average number of prophages per strain was 3.1, 2.7, and 2.5 in Desulfovibrionaceae , Geobacteraceae , and Shewanellaceae , respectively. This variation is influenced by a combination of genomic, phenotypic, and environmental factors, including genome size, physiological status, and the specific habitat in which the strain resides [ 266 , 267 , 268 ]. Our findings revealed that strains from Desulfovibrio subclade II and Geobacter subclade I, predominantly present in soils, freshwater, subsurface, and engineered ecosystems, exhibited the highest prophage density (prophages per Mbp genome). In contrast, strains from Shewanella subclade I and Desulfovibrio subclade I, more prevalent in marine environments, displayed the lowest values ( Figure S5 ). Clustered regularly interspaced short palindromic repeats (CRISPR) is a system that allows the identification and cleavage of foreign DNA. The presence of CRISPR-Cas arrays constitutes a barrier to HGT, including natural transformation, transduction and conjugation [ 269 , 270 , 271 ]. Our analysis revealed a high prevalence of CRISPR arrays in the genomes of Desulfovibrionaceae and Geobacteraceae , while a significantly lower prevalence (less than 30%) was observed for Shewanellaceae strains ( Figure 6 B). This trend could be explained by the recent identification of phages with genes encoding proteins capable of inhibiting CRISPR-Cas function in several Shewanella strains [ 272 , 273 , 274 ]. Integrons are genetic units that capture genetic material in bacteria, adding novel features to the cell that contains it. They consist of an integron–integrase gene, an integration site, a promoter for gene cassettes, and up to 200 gene cassettes containing open reading frames flanked by attC recombination sites [ 275 , 276 ]. We found complete integrons in over 50% of strains of Shewanellaceae and subclade II of Geobacteraceae , while they were absent in the other clades. A similar pattern was found for the prevalence of insertion sequences (ISs), which are cryptic DNA segments containing passenger genes that contribute to the metabolic plasticity and evolution of microbial genomes [ 277 , 278 ]. Our results indicate that the prevalence and number of ISs are unevenly distributed across the different bacterial groups included in this study. The Shewanella and Geobacter genomes contained 584 (~10.8 IS per genome) and 234 (~13 IS per genome), respectively, while the Desulfovibrio genomes contained a total of 34 ISs (~2.3 ISs per genome), with 27 out of the 42 Desulfovibrio showing no detection of ISs ( Tables S4 and S5 ). With few exceptions, including Desulfovibrio vulgaris str. Hildenborough, the low prevalence of ISs in Desulfovibrio species is consistent with previous studies, suggesting limited genomic rearrangements by transposition in this genus [ 279 , 280 ]. This may be explained by the fact that some of the ISs found to belong to families (i.e., IS Dvu3 ) in which the control of transposase expression relies on stop codon read-through and, therefore, may be affected by other regulatory mechanisms [ 281 ]. Between 35 and 61% of genomes were found to contain integrative and conjugative elements (ICEs). In agreement with previous reports and in contrast to what was found with integrons and ISs, strains belonging to Shewanellaceae registered a lower prevalence of ICEs than the other two families [ 52 ]. The data from 124 genomes were used to conduct a Principal Component Analysis (PCA) of thirteen genomic and mobilome variables and metrics reflecting the prevalence of EET elements ( Figure 7 ). The analysis clustered the strains into three big groups on the PC1-PC2 plane (accounting for 50.88% of the total data variability), revealing a stronger correlation with the taxonomy of species than with their habitats ( Figure S6 ). Based on the number of copies of genes encoding CymA, PpcA, and riboflavins, the number of CRISPR arrays and integrons per genome, genome size, and GC content, the Shewanellacea group is farther from the Geobacteraccea and Desulfovibrionaceae groups. The latest two groups are separated along the PC2, which includes the number of total cytochromes and genes encoding OmpJ and Cbc-related genes, as well as PpcA, which is almost absent in Shewanellaceae strains ( Figure 5 , Figures S4 and S6 ). Interestingly, the number of copies of genes encoding proteins related to EET contributes to the differentiation within the three groups. While the number of copies of genes encoding for CymA and riboflavins contributed to the distinction between Shewanella , the number of copies of genes encoding for OmpJ contributed to the distinction of the Desulfovibrionaceae . In a similar manner, the number of copies of genes encoding for Cbc and total cytochromes contributed to the distinction of the Geobacteraceae and the other two groups. Thus, this result suggests that the prevalence and abundance of EET elements significantly contributed to the differentiation of these groups. 3.3. Evolutionary Relationship of the Most Prevalent EET-Related Genes in SRPs Many of the genomes belonging to the Desulfovibrionaceae family were found to possess genes homologous to crucial proteins involved in the EET mechanisms of S. oneidensis MR-1 and G. sulfurreducens. This includes the triheme periplasmic cytochrome PpcA, the outer-membrane protein OmpJ, the tetraheme c -type cytochrome CymA, the iron–sulfur cluster-binding protein CbcT (a subunit of the Cbc4 complex), and, to a lesser extent, the cytochrome c -type CbcC (a subunit of the putative Cbc5 complex). All of these exhibited widespread distribution within this family. To gain insights into the evolutionary relationship of these shared genes among the Desulfovibrionaceae , Geobacteraceae , and Shewanellaceae families, we conducted phylogenetic analysis and compared the genetic contexts of the select genes associated with the orthologous groups of these candidate proteins. 3.3.1. The Periplasmic Cytochrome PpcA, an Intermediary in Extracellular Electron Transfer PpcA is a periplasmic cytochrome that acts as an intermediary electron carrier for EET. Genetic studies have found that PpcA acts as a terminal reductase for anthraquinone-2,6-disulfonate (AQDS), Fe(III)-citrate, and Ferric nitrilotriacetate (Fe-NTA), although this gene was not differentially expressed when G. sulfurreducens was grown with Fe(III) citrate and Fe(III) oxide [ 240 , 282 , 283 ]. The genetic context of ppcA in G. sulfurreducens includes several adjacent genes encoding c -type cytochromes, as well as genes involved in their biogenesis, such as ResB and CcsB [ 284 ], and genes involved in the biosynthesis of menaquinones and ubiquinones, redox-active compounds involved in respiratory networks [ 285 , 286 ] ( Figure S7 ). The genomic context remains largely invariant across different species, and its phylogenetic relationship aligns with the species’ phylogenomic tree, suggesting vertical transmission rather than horizontal gene transfer. PpcA is part of a family of five periplasmic triheme cytochromes (including PpcB, PpcC, PpcD, and PpcE). Thermodynamic characterization of those cytochromes revealed differences in their heme reduction potentials, allowing for a wider range of redox partners and enhancing the adaptability of the respiratory mechanism [ 287 , 288 ]. Our findings revealed that ppcA homologs are prevalent in Geobacter species, with most strains containing between 5 and 6 homologous genes, averaging 4.9 genes per strain. Notably, ppcA homologs were also present in all examined members of the Desulfovibrionaceae family, with the majority of those having between 3 and 4 homologous genes (averaging about 3.4 genes per strain) ( Tables S2 and S6 ). The phylogenomic analysis identified four primary clades within the PpcA protein family ( Figure 8 ). Clade 1 exclusively consists of proteins from Geobacter strains. Notably, the five known homologous proteins from G. sulfurreducens (PpcA, PpcB, PpcC, PpcD, and PpcE) are distributed across different subclades within this group. Proteins similar to PpcA from the Desulfovibrionaceae family, found in the remaining three clades, exhibit significant diversification within their respective subclades. Similar to Geobacter , the presence of multiple variants and their diversification is likely the result of functional diversification associated with heme reduction. One example of evolutionary divergence is the gene encoding for the PpcA protein of Pseudodesulfovibrio mercurii (WP_014320801.1), which, based on genomic context, falls outside the four main clades ( Figures S7 and S8 ). The unique environmental characteristics of the mid-Chesapeake Bay estuarine sediments where P.mercurii was isolated, including complex geochemical processes, nutrient-rich reducing waters, and the presence of rare earth elements (REEs), such as Cerium (Ce) and Europium (Eu), might have been significant factors driving this divergence [ 289 , 290 ]. In contrast, homologs of ppcA in two Shewanella strains, S. atlantica HAW-EB5 and S. sediminis HAW-EB3, both isolated from marine sediments near Halifax Harbour, Canada [ 291 , 292 ] suggest a horizontal gene transfer (HGT) event from another bacterium in this geographical region. Furthermore, clade 2 mainly comprises homologous proteins from members of the Desulfovibrionaceae , except for a PpcA homologous from Geobacter sp. SVR (WP_239077329.1), suggesting potential horizontal gene transfer events based on an interruption in the phylogenetic tree topology. These assumptions are supported by the presence of transposase-encoding genes in the vicinity of these genes [ 293 , 294 ]. 3.3.2. OmpJ, an Integral and Widespread Outer Membrane Protein in the Desulfovibrionaceae Family Homologs of OmpJ were found to be widely distributed in the Desulfovibrionaceae family. Similar to Geobacter species, most of these strains have between one and two homologous ompJ genes. Interestingly, Desulfovibrio desulfuricans strains and those from the Solidesulfovibrio genus stand out, with most of their members having between eight and ten genes homologous to ompJ ( Tables S2 and S6 ). This substantial prevalence of OmpJ-like genes suggests that this protein may have a role in the physiology, adaptation, and capabilities of participating in sulfate reduction and metallo-reduction processes [ 41 ]. The distribution of the ompJ phylogenetic tree shows three clear groups: one from Geobacter and two from the Desulfovibrionaceae family ( Figure S9A ). The two branches of the Desulfovibrionaceae family consist mainly of species isolated from marine environments, pollution events, or industrial activity, while the other group primarily comprises species isolated from soil, freshwater, or animal sources ( Figure S10 ). One exception is Halodesulfovibrio aestuarii DSM 17919, isolated from shoal mud in Germany [ 295 ], which stands out as it harbors two similar proteins that are significantly different from the other genes. This discrepancy in the species tree may suggest a horizontal transmission event, since near these ompJ-homologous genes, several tRNA sequences are found. To date, there have been limited reports regarding the function of OmpJ. Therefore, it would be of interest to assess its potential role in signaling mechanisms and its impact on EET mechanisms. 3.3.3. CymA, a Common Branch Point in the Electron Transport Chain The c -type cytochrome CymA seems to be essential for facilitating the anaerobic respiratory adaptability of Shewanella . CymA plays a key role in transferring electrons from menaquinol to various systems responsible for reducing terminal electron acceptors, such as fumarate, nitrate, nitrite, dimethyl sulfoxide (DMSO), arsenate, and insoluble minerals like Fe(III) and Mn(IV) [ 253 , 296 , 297 , 298 ]. Homologous genes to cymA are widespread among the members of the analyzed families. The phylogenetic tree of CymA reveals three main clades: two from Shewanella strains and a third shared between strains of Geobacter and the Desulfovibrionaceae family. While Shewanella species contain between 1 and 5 homologs (with an average of 2.7), almost all members of the Desulfovibrionaceae family contain 1 or 2 homologs, with a few exceptions. Nearly half of the species analyzed in Geobacteraceae contain one to two genes homologous to cymA ( Tables S2 and S6 ). The variable presence of these genes in Geobacteraceae species suggests two potential evolutionary scenarios: the genes were either lost in most species or inserted into the genomes of the analyzed species. Notably, species containing these genes are mostly clustered within a specific clade in the phylogenetic tree ( Figures S11 and S12 ), hinting at the likely insertion of this gene through horizontal gene transfer into the common ancestor of these species, followed by its subsequent vertical transmission. Another notable sequence in the phylogenetic tree is one of the homologous copies from Geobacter hydrogenophilus DSM 13691 (WP_214187890.1), positioned between two subclades of Desulfovibrionaceae family proteins ( Figure S9B ). The context of this sequence suggests it was likely acquired via horizontal gene transfer, especially considering the proximity of genes encoding site-specific integrases and recombinase family proteins. Interestingly, when investigating the gene contexts of cymA homologs, several members of the Desulfovibrionaceae and Geobacteraceae families feature one of these homologous genes located near the gene encoding ammonia-forming cytochrome c nitrite reductase subunit c552 ( Figures S11 and S12 ). The similarities between the genetic contexts of both families, along with their closer phylogenetic proximity in comparison to Shewanellaceae species ( Figure S9 ), suggest that the potential insertion into Geobacteraceae genomes might have originated from horizontal transmission from a member of the Desulfovibrionaceae family or closely related species. 3.3.4. Inner-Membrane Quinone Oxidoreductase Protein Complexes: CbcC and CbcT Subunits Provide Plasticity and Modularity to Different Complexes Involved in EET Cytochrome bc1 complexes are membrane protein complexes found in the electron transfer chains of bacteria using oxygen, nitrogen, and sulfur compounds as electron acceptors. These enzymes transfer electrons from ubiquinol to cytochrome c and move protons across the membrane. Despite transcriptomic and proteomic studies revealing differential expression patterns of Cbc-like gene clusters in G. sulfurreducens in response to electron acceptor availability, there is still limited information available regarding these complexes [ 237 , 244 , 299 , 300 ]. CbcT homologs are present in high abundance across all three families. In the Desulfovibrionaceae family, which has the highest average number of cbcT homologs (9.8), Pseudodesulfovibrio mercurii is noteworthy with 17 copies ( Tables S2 and S6 ). G. uraniireducens Rf4 and S. sediminis HAW-EB3, from the Geobacteraceae and Shewanellaceae families, respectively, stand out with the largest numbers of CbcT orthologs, at 11 and 19 genes, respectively, except for S. denitrificans , which does not contain these genes. The intricate topology of the phylogenetic tree mirrors the abundance of this gene. Specifically, distinguishing clades by family is challenging due to their phylogenetically interwoven sequences. Moreover, there does not appear to be a relationship between abundance and sources of isolation, as the amount varies across all environmental classifications ( Figure S9 ). The cbcSTU operon is highly conserved among Geobacteraceae species ( Figure S12 ). Among the strains analyzed, only Geobacter sp. FeAm09 presents homologs of cbcT, but neither cbcS not cbcU. Several Shewanellaceae strains exhibit a closely related cluster, featuring an orthologous of cbcT, which is a homolog to the sirC gene in S. oneidensis MR-1. SirC is a 4Fe-4S ferredoxin that, together with its partner SirD, encoding an NrfD/PsrC-type quinol dehydrogenase, has the ability to transfer electrons from quinols to the same respiratory pathways as CymA, except for nitrate. As a consequence, this quinol dehydrogenase complex (SirCD) can functionally replace CymA in the respiratory pathways for fumarate, DMSO, and ferric citrate as the electron acceptor [ 301 ]. It is worth noting that in most of the Shewanellaceae strains, there are two genes adjacent to one of the homologous cbcT genes. These two genes do not have homology to cbcS and cbcU but share the same annotation as the latter two genes, and are arranged in a similar spatial disposition. Specifically, the gene encoding the cytochrome c3 family protein (similar to CbcS) is homologous to the outer-membrane lipoprotein c -type cytochrome OmcI in G. sulfurreducens PCA. In this case, similar to the cbcU gene in G. sulfurreducens , there is a gene encoding a cytochrome c nitrite reductase subunit NrfD, although it does not belong to the CbcU orthogroup. These observations suggest a convergent evolution event in forming these modular membrane complexes, which appear to function similarly in electron transfer from the quinol pool in different respiratory pathways [ 302 , 303 ]. Interestingly, S. sediminis HAW-EB3, which presents the greatest number of cbcT homologs (and which acquired a ppcA homolog), also presents a high number of copies of other genes involved in EET mechanisms, including cbcA/omcI (12) and mrtA (12) homologs. This strain, isolated from an unexploded ordnance dumping site in the Atlantic Ocean near Halifax Harbour, Nova Scotia, Canada, can degrade RDX, nitrate, and nitrite. However, it does not demonstrate a reduction in Fe(III) or elemental sulfur [ 291 ]. These distinctive characteristics are likely a result of the environmental pressures in its natural habitat, which gives it unique features compared to other species. Furthermore, these observations suggest that these proteins are not limited to iron- or sulfur-reduction pathways from natural sources, but to other compounds, including chemicals from industrial activity and pollution events, to which bacteria have had to adapt. Considering its genetic features, it would be interesting to conduct a more in-depth characterization to investigate the functional potential of these mechanisms. In contrast, S. violacea DSS12 and S. denitrificans OS217 either lack or have only one homolog of cbcT. These strains also lack other crucial genes involved in EET mechanisms, such as OmcA, CctA, FccA, and the MtrCAB complexes. Recently, Baker et al. (2021) also reported the absence of this last complex in both species and attributed it to an environmental pressure effect, which was linked to the transition to an aerobic environment, facilitated by their habitat at the oxygenated sediment-water interface [ 59 , 304 , 305 ] Regarding the Desulfovibrionaceae family, 24 of the 42 strains analyzed contain at least one homolog of cbcC. This distinct group forms a well-differentiated clade that exhibits significant evolutionary divergence from Geobacteraceae and Shewanellaceae , suggesting adaptation to the metabolic needs of these bacteria since their last common ancestor. In fact, this group of cbcC homologs coincides with one of the clusters of the multi-heme cytochrome similarity network ( Figure 3 B), specifically community N#37. This cluster comprises cytochrome c family proteins with 10 heme motifs, forming a separate cluster from other nodes in the network, thereby supporting the observed evolutionary divergence in the phylogenetic tree. Notably, in these species, at least one of the cbcC homologs was found alongside the rnfABCDGE operon, which was initially identified in Rhodobacter capsulatus [ 306 ]. The RNF complex is composed of six subunits, including four membrane proteins (RnfA, RnfD, RnfE, and RnfG) and two iron–sulfur proteins (RnfB and RnfC), and encodes a membrane-bound NADH:ferredoxin dehydrogenase [ 306 ]. Our results revealed that many members of the Desulfovibrionaceae family exhibit the loss of the rnfB gene and instead have a gene encoding an FAD-dependent oxidoreductase at the end of the operon ( Figure 9 ). In the Bacteroidota/Chlorobiota group, the loss of the rnfB gene has been documented, along with the recruitment of a reductase subunit from aromatic monooxygenases (AMOr protein), resulting in the emergence of the sodium-dependent NADH:ubiquinone oxidoreductase (Na + -NQR). This complex is commonly associated with the aerobic respiratory metabolism of pathogenic bacteria. A key distinction between Rnf and Na + -NQR is the mechanism of electron incorporation into the complex, suggesting an alternative mechanism for electron transfer in the presence of this oxidoreductase in the Rnf complex in Desulfovibrionaceae family strains [ 307 ]. The presence of this oxidoreductase in the Rnf complex in strains of the Desulfovibrionaceae family may indicate an alternative mechanism for electron transfer. Interestingly, the strains exhibiting this substitution are typically found in marine or impacted environments, which are known for their harsh conditions compared to the habitats of strains with the RnfB subunit (mainly found in soil, animals, and freshwater). These findings imply that environmental factors such as oxygen levels and the presence of metallic and organic electron acceptors or donors may have driven the modification of the rnf operon. It is important to note that in almost all strains of the Desulfovibrionaceae family that do not have cbcC homologs, the other subunits of the Rnf complex are also missing. This supports a hypothesis regarding the integration of the CbcC homolog as an accessory protein within the complex. The only exceptions are Desulfovibrio cuneatus DSM 11391, which possesses the rnf operon minus the RnfB subunit and the CbcC homolog, and Halodesulfovibrio aestuarii DSM 17919, which harbors the complete rnf operon and a cytochrome c, not homologous to CbcC ( Figure S14 ). All other strains with the rnf operon contain the cbcC homolog, likely indicating an ancient evolutionary incorporation event before these species diverged. Furthermore, the sporadic occurrence of this complex in some family members strongly suggests horizontal gene transfer as a mode of acquisition. Previous research indeed indicates the spread of this complex among various lineages by HGT, including several species within the phyla Pseudomonadota , Chlamydiota , and Planctomycetota , which subsequently led to the rise in other complexes, such as Na + -NQR mentioned above [ 308 ]. In this context, the CbcC-like cytochrome represents a unique module that has become integrated into various membrane complexes involved in electron transfer through evolution and species diversification. In Geobacter , it is part of the CbcEDCBA cluster where it is predicted to form menaquinol:ferricytochrome c oxidoreductase [ 237 ], and in Desulfovibrionaceae family members, it forms part of the Rnf complex, with a predicted cytoplasmic location, where it likely engages in coupling to facilitate efficient electron transfer ( Figure 9 ). These findings indicate that this subunit, adapted by Desulfovibrionaceae family members with significant evolutionary divergence, might offer a potential new catalytic innovation. By incorporating it into the Rnf complex, these bacteria could potentially broaden the detection of redox potentials and access an alternative electron transfer pathway, thereby enhancing growth efficiency in variable environments."
} | 12,865 |
28691092 | PMC5479648 | pmc | 4,916 | {
"abstract": "A strategy for bonding water-rich hydrogels to diverse materials for electronic skins, energy storage, and soft optics is reported.",
"conclusion": "CONCLUSION In summary, we introduce here a facile, universally applicable method for instant tough bonding of hydrogels to a wide variety of materials—from soft to hard—with unprecedented interfacial toughness exceeding the intrinsic fracture strength of the gels. We apply our approach to create a new set of soft machines and electronics and to demonstrate instant healing, adaptive optics, soft actuators and generators, tough batteries, and hydrogel electronic skins. Applications range from robotics, energy harvesting from renewable sources, consumer electronics, and wearables to a new class of medical tools and health monitors.",
"introduction": "INTRODUCTION Hydrogels are versatile building blocks of life—living beings are in essence gel-embodied soft machines. The intricacy and vast diversity found in nature, from entirely soft Cnidaria or mollusks to hybrid vertebrates, rely on insightful merging of a wide variety of biological materials ( 1 ) forming gels, tissues, fibers, muscles, tendons, and skeletal structures ( 2 ). Although technologically inimitable in full complexity, the unique properties of engineered, tough hydrogels ( 3 , 4 ) advance soft machines ( 5 ) and electronics ( 6 ), sparking a new generation of biomimetic systems. A high level of functionality in the artificial counterparts is achievable through hybrid combinations of hydrogels and soft-to-solid materials, from elastomers to polymers, metals, or mineralized tissue. Without addressing issues of adhesion, impressive demonstrations of tough ionic hydrogels and elastomer membranes for transparent loudspeakers ( 7 ) and cables ( 8 ), touch panels ( 7 , 9 , 10 ), and stretchable electroluminescent displays ( 11 , 12 ) were realized. Overcoming the low interfacial toughness when bonding diverse substances to water-rich hydrogels is challenging ( 13 , 14 ). Recent approaches realizing strong bonds between tough gels and nonporous solids ( 15 ) involve surface functionalization via self-assembled monolayers; bonding to elastomers was achieved through ultraviolet (UV)–initiated polymerization ( 16 ). However, all of these methods require extended curing processes difficult to implement in the vicinity of sensitive target surfaces. They are not generally applicable over a wide range of antagonistic material classes and lack an instant character, essential in fracture healing and other time-critical applications such as rapid prototyping and high-throughput manufacturing ( 17 ).",
"discussion": "RESULTS AND DISCUSSION We introduce here a set of concepts, material approaches, and design rules for wide-ranging classes of soft, hydrogel-based electronic, ionic, and photonic devices based on soft and soft-to-hard hybrid architectures. Robust interfaces between the individual layers prevent delamination and are essential for mechanically tough yet soft, highly functional and compliant systems. By mixing synthetic adhesives that rapidly polymerize in the presence of hydroxyl ions with suitable nonsolvents, we achieve easy-to-use dispersions with tailorable properties. The general principle is outlined in Fig. 1A , where our adhesive dispersion is applied to form tough, instant bonds between (i) separate hydrogels, (ii) hydrogels and elastomers, and (iii) hydrogels and solid surfaces (illustrated here with polymer foils bonded to prestretched gels). The use of a bonding agent where the adhesive is dispersed in a nonsolvent, instead of the pure or solvent-diluted adhesive, overcomes several difficulties. Alignment of the surfaces to bond requires sufficiently delaying the water-triggered polymerization reaction, achievable with the volatile nonsolvent acting as temporal protecting shell. Our dispersion does not reduce the bonding strength, an effect we observe when using simple dilutions, and allows for precise control of bonding agent quantities such that no hard and brittle interfaces form in soft hybrids. With cyanoacrylate-based adhesives dispersed in alkanes (see the Supplementary Materials for details), the bonding process, including application of the dispersion, alignment, and curing, is typically completed in less than a minute, easing design iterations and upscaling. We demonstrate the versatility of our approach on a wide class of fully swollen hydrogels, including conventional [polyacrylamide (PAAm), poly(2-hydroxyethyl methacrylate) (PHEMA), and poly(vinyl alcohol) (PVA)] and double-network PAAm/alginate with internal dissipation ( 18 , 19 ). The bonding agent polymerizes such that instant tough bonding is achieved, for example, between a tough hydrogel containing covalently and dissipatively cross-linked subnetworks and a single-network elastomer, hydrogel, or polymer ( Fig. 1B ). Glues of the cyanoacrylate family were previously applied to fixate hydrogels ( 16 , 20 ) but without investigating the mechanical properties and interfacial bonding strength, likely due to the rapid formation of bulky, brittle acrylic resin when the monomers polymerize in direct contact to the water-containing hydrogels. Our dispersions of cyanoacrylate monomers and 2,2,4-trimethylpentane overcome these issues, because they allow diffusion of the adhesive into the hydrogel and into elastomers, leading to tough bonds but without forming rigid resin interlayers. We evidence these effects with spatially resolved Raman spectroscopy (diffusion of adhesive; figs. S1 and S2) and with practical demonstrations of highly stretchable hydrogel-hydrogel and hydrogel-elastomer heterostructures. We suspect that the synergistic effects of physical entanglement due to adhesive interdiffusion and the ready formation of van der Waals and hydrogen bonds of cyanoacrylates ( 21 ) with a wide variety of materials result in the high interfacial toughness of our adhesive interfaces. Here, the hydrophobic nature of the alkanes may additionally support diffusion of the adhesive into the hydrogel due to the formation of a protective shell, sufficiently slowing down the water-induced polymerization of cyanoacrylates. Instant healing of fractures in hydrogels, including ionic conductors, is a clear benefit of our approach. We are able to restore the mechanical and electrical properties almost completely within seconds, even for severed structures ( Fig. 1C and video S1). The healed conductors endure large stretch ratios λ with electrical resistance almost identical to the pristine one ( Fig. 1D ). Although requiring the application of cyanoacrylates as polymerization agent, the instant character of our method is of advantage for time-critical applications, whereas autonomously self-healing, highly stretchable ionic conductors ( 22 ) typically require longer times for fully restoring their mechanical properties. Instant tough bonding of elastomers and hydrogels without impairing the overall elasticity is challenging, so far only possible after extensive pretreatment of both elastomer and hydrogel surfaces followed by postcuring processes ( 16 ). In contrast, our dispersions facilitate rapid adhesion of hydrogels on elastomers following a simple casting or spin-coating process. The formed interface remains highly elastic and enables the sandwich to be stretched reversibly without delamination by more than 2000% in area, whereas physically attached gels slip ( Fig. 1 , E to G, and figs. S3 and S4). In our approach, stretchability is essentially only limited by the intrinsic rupture strength of the hydrogel. Enhancing functionality and complexity is, for example, feasible though the integration of flexible ( 23 , 24 ) and imperceptible ( 25 ) electronic foils. We work with prestructured, fully polymerized hydrogels that are mechanically deformable before linking, allowing us to exploit wrinkling instabilities when targeting stretchable hydrogel electronics. This approach is illustrated with a structured conductive copper electrode on a thin polymer foil, intimately bonded to a prestretched PVA hydrogel ( Fig. 1 , H to J). Fig. 1 Instant tough bonding of hydrogels to a wide range of materials. ( A ) Schematic illustration of the bonding method. An adhesive dispersion is used to instantly tough-bond (i) heterogeneous hydrogels, (ii) hydrogels and elastomers, and (iii) rigid to flexible foils to (prestretched) hydrogels in less than a minute. ( B ) The bonding agent links tough hydrogels with covalent and dissipative cross-links to elastomers, and the formed interface is instant and tough yet remains stretchable. ( C ) Instant healing of a conductive hydrogel rod used to light up a light-emitting diode (LED) circuit. Both mechanical and electrical properties are restored after complete incision. ( D ) Normalized resistance versus uniaxial strain before (blue trace) and after (red trace) healing a hydrogel conductor. ( E to G ) Hydrogel square (colored blue for visibility) bonded to a transparent elastomer. The soft hybrid is stretched more than 1000% in area without delamination. ( H to J ) A conducting Cu-coated poly(ethylene terephthalate) (PET) foil is tough-bonded to a prestretched hydrogel. Upon release of the prestrain, out-of-plane wrinkles form in the foil, creating a reversibly stretchable soft-hard hybrid. Tough hydrogel bonding characterization We substantiate our initial findings with systematic 90° peeling tests ( Fig. 2A ; see details in Materials and Methods) ( 15 , 26 ) on various hydrogels (PHEMA, PAAm/alginate, and PVA), where tough versions dissipate a substantial amount of mechanical energy through the breakup of noncovalent bonds in their bulk. Following initial elastic deformation of hydrogel and, where applicable, elastomer, a stable plateau region in the displacement versus normalized peeling force (identical to interfacial energy; Fig. 2B ) evidences bulk rupture of the hydrogel without bonding failure at the gel/target surface interface (see also video S2). We observe this behavior for all investigated interface classes [polymers, elastomers, leather, bone, and chromium-coated metals and glass ( 27 )] and thus find an average fracture toughness of 736 ± 112 J/m 2 for PHEMA, 1427 ± 89 J/m 2 for PAAm/alginate, and 2208 ± 186 J/m 2 for PVA ( Fig. 2C ). PAAm hydrogels, although having bonding mechanisms similar to their tough counterparts, typically rupture at around 30 J/m 2 because of their much lower bulk toughness (fig. S5A). The observed independence of fracture toughness on substrate material for all hydrogels further corroborates that the interfacial strength achieved with our method exceeds, in any case, the intrinsic fracture toughness of the constituent gels. Consequently, peeling tests on hydrogel-hydrogel (PHEMA-PHEMA and PHEMA-PVA) heterostructures result in bulk rupture of the weaker gel (fig. S5B). These results are comparable with the bulk fracture toughness of 489 ± 47 J/m 2 for our PHEMA and 1472 ± 440 J/m 2 for our PVA hydrogels (see fig. S6) obtained from pure shear notch tests ( 3 , 28 ). The deviations of bulk and interfacial toughness stem from differences in sample and test setup geometries. We find that the interfacial bonding strength is over a wide range nearly independent on the mixing ratio of the cyanoacrylate with the nonsolvent (1:1 to 1:15 in volume; fig. S5C). However, the optical and elastic properties of the composite depend on the total amount of applied adhesive per unit area, where stretchable, transparent hydrogel-elastomer heterojunctions are achieved with mixtures of 1:10 to 1:15. In a similar manner, the choice of alkane has little influence on the interfacial toughness (fig. S5D); here, the dynamics of the bonding process is influenced by the volatility of the alkane. High–vapor pressure 2,2,4-trimethylpentane will more rapidly evaporate, and longer-chain 1-octadecene or paraffin oil is nonvolatile and delays (on the order of seconds) the polymerization of cyanoacrylates, providing more time for specimen alignment. Despite its simplicity, our approach leads to bonding strengths surpassing previously reported methods. The limits are not yet reached; designing hydrogels with higher internal fracture toughness is expected to further increase resistance to fracture and tear of these soft systems. We additionally visualize the instant character of our bonding method by assembling a model vertebral column with three-dimensional (3D)–printed vertebrae and hydrogel intervertebral discs (fig. S7 and video S3), and the toughness of the interface by lifting a 1-kg metal weight instantly attached to a tough hydrogel rod (fig. S8 and video S4). Adaptive optics, soft photonics, and camouflaging benefit from near-perfect transparency in the visible range ( 29 ), an intriguing feature of some hydrogels and certain elastomers ( 7 ). Not impairing this advantage by the bonding method is paramount; we demonstrate this for PAAm/VHB double layers that remain perfectly transparent throughout the visible spectrum from 400 to 800 nm. Near-zero absorption within the materials and approximately 93.5% transmission through the stack ( Fig. 2D ), where scattering at rough surfaces and reflection due to refractive index mismatch on the air/hydrogel and elastomer/air interfaces cause the main losses (for reflectance measurements, see fig. S9), is maintained even when the soft structures are undergoing areal extension by more than 2000% ( Fig. 2E ). Here, using our adhesive dispersions is crucial, because pure cyanoacrylates will result in the formation of a diffuse scattering layer at the bonding interface (fig. S10). Fig. 2 Interfacial toughness and optically transparent, stretchable bonds. ( A ) Illustration of the 90° peeling test, with the hydrogel instantly tough-bonded to the bulk substrate. A liner serves as stiff backing. Crack propagation in the hydrogel occurs perpendicular to the peeling direction (detail and photograph with colored hydrogel). ( B ) Measured interfacial toughness for PVA-isoprene (turquoise trace), PAAm/alginate–PET (light blue trace), and PHEMA-Ecoflex (purple trace) instant tough bonds. ( C ) Interfacial toughness of PHEMA (purple), PAAm/alginate (light blue), and PVA (turquoise) hydrogels instantly tough-bonded to plastics, elastomers, leather, bone, and chromium-coated metals (aluminum and copper) and glass. Mean values and variance for at least three individual peel tests are shown. PDMS, poly(dimethylsiloxane); PI, polyimide. ( D ) Transmission and absorption spectra of the acrylic elastomer VHB (yellow trace), PAAm hydrogel (blue trace), and PAAm hydrogel instantly bonded to VHB (red trace). Bonding does not deteriorate the optical properties of the soft hybrid. ( E ) Nearly strain-independent transmission (red trace) at 600 nm for a biaxially stretched VHB/PAAm stack up to 2000% areal expansion. Inset: Photograph of the sandwich at 1000% strain with a color wheel in the background. The dashed red line outlines the PAAm hydrogel. Hydrogels for soft actuators, generators, and power sources Encouraged by these results, we use our methods in demonstrating tough, stretchable adaptive optics, soft machines, and generators. Soft lenses with electrically tunable focal length ( 29 , 30 ) may have applications ranging from consumer electronics to biomimetic robots ( 31 – 33 ). Our approach ( Fig. 3A ) uses transparent PAAm hydrogel sheets as ionic conductors, bonded to a convex lens formed from acrylic elastomer filled with NaCl solution. Applying a dc voltage causes the whole lens to thin, thereby changing the focal length by up to 110% at a voltage of 6.5 kV ( Fig. 3 , B and C). Here, the voltage drop across the hydrogel-dielectric-water capacitors is large, preventing electrochemical side reactions at the high-capacitance interface between hydrogel and metal wiring to the power supply ( 7 ). The response time of this centimeter-scale soft, tunable lens is in the 300-ms range, demonstrated by diverting a laser beam guided through a dye-filled water basin ( Fig. 3 , D and E, and video S5). Here, the mass of the lens dominates the time constant and is readily reducible by downscaling the dimensions to form microlenses. Crucial not only for powering soft electronics and machines, energy generation from mechanical sources on small (that is, walking) and large (that is, ocean waves) scales is currently pursued with soft, dielectric elastomer generators ( 34 ) due to their promise of high specific energy of conversion ( 35 ). New concepts for operating such devices in contact with seawater are required ( 29 ), rendering ionic hydrogels a promising electrode material. Dielectric elastomer actuators and generators are highly deformable capacitors ( Fig. 3F ), allowing the conversion of electrical to mechanical energy and vice versa ( 36 ). We have developed test setups and methodologies for assessing key material parameters ( 37 ). Electromechanical coupling is most efficient if the soft capacitor is deformed equibiaxially, that is, if the capacitance C scales with the fourth power of stretch ratio λ: C ∝ λ 4 . The capacitance of our strongly bonded hydrogel-VHB-hydrogel membranes scales accordingly ( Fig. 3H ). A simple method of achieving equibiaxial stretch, at least in good approximation at the apex, is realized by inflating a balloon. We exploit this for our soft hydrogel generator ( Fig. 3G ) and transform approximately 500 mJ of mechanical into 54 mJ electrically usable energy per cycle ( Fig. 3 , I and J), yielding an overall conversion efficiency of 11%. Improvements are expected with low-viscosity elastomers and the use of stacked multilayer capacitors with shared hydrogel electrodes. Energy storage is an equally central issue in the quest for fully autonomous soft machines to date, which is, for example, beautifully addressed through the catalytic decomposition of onboard fuel ( 5 ). Powering embedded electronics including logic and sensors will, however, require current sources. Soft batteries are a natural choice; we have developed the first concepts for such intrinsically stretchable electrochemical cells ( 38 ). Briefly, realizing stretchable batteries requires all electrochemical components (anode and cathode materials, electrolyte, and current collectors) in the form of pastes or gels embedded in an elastomeric casing. These early concepts required lateral separation of anode and cathode chambers to enable mechanical compliance, in turn reducing power density and short-circuit currents. Now state-of-the-art ( 39 – 42 ), here we radically redesign this approach using tough, electrolyte-containing hydrogels as stretchable separator ( Fig. 4A ). Tough bonding to the elastomer matrix enables a “top configuration” of anode, separator, and cathode, greatly reducing internal resistance ( Fig. 4B and fig. S11) and increasing specific capacity to 1.6 mA·hour/cm 2 (fig. S12) when discharging to 0.9 V for a zinc-manganese dioxide primary battery. In conjunction with imperceptible electrodes as current collectors, not only are our tough batteries highly stretchable, but also their performance improves when stretched as the electrolyte resistance decreases with thinning of the separator gel ( Fig. 4C ). We use our soft batteries for self-powered stretchable circuits directly integrated atop the compliant power source ( Fig. 4D ). Imperceptible circuit boards ( 43 ) equipped with off-the-shelf surface mounted device (SMD) electronics (fig. S11) including dc-dc step-up converters and LEDs are laminated to the prestretched battery, forming an out-of-plane wrinkle structure when the strain is released ( Fig. 4E ). These devices are tough soft-hard hybrids that endure repeated twisting and stretching without impairing performance ( Fig. 4F and fig. S11). Future work should focus on tough, secondary batteries, a fully biocompatible chemistry ( 44 , 45 ), and integrated wireless or solar charging ( 41 ). Fig. 3 Soft adaptive lens and energy harvester with instantly tough-bonded hydrogel electrodes. ( A ) Working principle of the tunable lens. Transparent hydrogel electrodes were used to electrically deform the lens, defined by a liquid reservoir embedded within a VHB elastomer. ( B ) Photograph of the soft lens without and with applied voltage, demonstrating electrical tuning of focal length. ( C ) Focal length and change of focal length versus voltage, illustrating a large voltage-induced focal length change of 110% at 6.5 kV. ( D and E ) Visualization of the voltage-controlled focal length change by showing laser light traces in a basin filled with a diluted rhodamine/water solution. ( F ) Illustration of mechanical into electrical energy conversion with a deformable elastomer/hydrogel capacitor. ( G ) Photographs of the deformable balloon-shaped elastomer/hydrogel capacitor. ( H ) Normalized capacitance of the deformable capacitor versus stretch ratio λ, demonstrating a λ 4 behavior. ( I and J ) Exemplary energy harvesting cycle in pressure-volume and voltage-charge work conjugate plots. The enclosed areas in the two diagrams illustrate the mechanical energy input (497 mJ) and electrical energy output (54 mJ) of the cycle, resulting in a conversion efficiency of more than 10%. Fig. 4 Tough, stretchable battery and self-powered stretchable circuit. ( A ) Scheme of a stretchable battery in top configuration, enabled by the tough hydrogel separator instantly tough-bonded to the elastomer matrix. Imperceptible 1.4-μm-thick electrodes serve as current collector. Cu-coated PET foil with Zn paste is the anode, and MnO 2 paste contacted with Au-coated PET is the cathode. ( B ) Nyquist plot for top (red dots) and lateral (blue dots) configuration of a stretchable battery, showing reduced internal resistance due to shorter ionic paths through the separator in top configuration. ( C ) Voltage versus discharge current at 0% (red squares) and 50% (blue triangles) strain. Internal resistance decreases from 8.9 to 6.7 ohms with stretching due to thinning of the hydrogel separator. ( D ) Illustration of the self-powered stretchable LED circuit with a 6-μm PET circuit board and SMD elements. A dc-dc converter boosts the voltage of the battery to power three LEDs. ( E and F ) Stretchable circuit atop a hydrogel-based battery in relaxed (E) and strained and twisted state (F), without impairing function. Hydrogel electronics Tough hydrogels currently find applications in soft transducers, exploiting their good ionic conductivity and high stretchability ( 7 – 12 ). However, advanced soft systems, from prosthetic skins ( 23 ) to smart medical implants ( 46 – 49 ), benefit from the pluripotency of electronic sensors, actuators, and logic circuits. Pairing such systems with hydrogels is in an early state, with initial attempts toward basic circuit elements including stretchable serpentine-based conductors and SMD LEDs, as well as designs for microparticles and microfluidic channels allowing drug delivery ( 6 , 16 , 50 ). Relying on our instant tough bonding approach, we demonstrate here design strategies and methods for untethered hydrogel-based electronic patches including power supply, control/readout electronics, wireless communication, and stretchable actuators and sensors ( Fig. 5A ). In a modular architecture, we assemble a reusable power, logic, and radio-containing unit on a flexible circuit board that connects to the stretchable, imperceptible transducer array consisting of four temperature sensors and heater elements. The electronic unit is then bonded to a tough, biocompatible PVA hydrogel matrix, serving as interface to biological tissue ( Fig. 5B and fig. S13, with details on the circuit layout). In contrast to thin polymer foils ( 24 , 25 ) or elastomers ( 51 ) commonly used for electronic skins ( 23 , 52 , 53 ), hydrogels more closely mimic mammalian skin, with the possibility of providing nourishment, waste removal, or drug delivery through convection and diffusion throughout the gel matrix or embedded diffusive microchannels ( 6 , 16 ). In particular, thermally triggered release of drugs stored in lipid-coated reservoirs within the hydrogel body may turn out advantageous for chronic wound treatment, easily controllable via smartphone apps or computers. We therefore initially design an array of four heaters and resistive-type temperature sensors on 1.4-μm PET foil, strongly bonded to a sheet of a 2-mm-thick PVA hydrogel. Finite element simulations of the heat distribution throughout such a patch ( Fig. 5C ) agree well with IR measurements ( Fig. 5D ). We then integrate the stretchable transducer array with the flexible power, communication, and control unit. This wireless hydrogel electronic skin is stretchable to about 20% (fig. S14) and intimately conforms to human skin, with the biocompatible PVA gel at the biotic/abiotic interface. We demonstrate untethered, fully autonomous operation including data logging with a smartphone app ( Fig. 5E and fig. S15) while, in addition, monitoring the heater function with an IR camera ( Fig. 5F and video S6). The temperature response of the four heater elements is precisely monitored by the integrated temperature sensors and agrees excellently with IR camera control measurements ( Fig. 5 , G and H). Physical attachment is sufficient for wearing our devices cutaneously. However, it is possible to directly bond hydrogel electronics to mammalian tissue with U.S. Food and Drug Administration–approved ( 54 ) octyl cyanoacrylate (fig. S16A); we initially assess the biocompatibility of this approach with cell viability studies (fig. S16B). Thermally triggered release of drugs becomes possible through lipid-coated reservoirs atop a heater element ( Fig. 5I ). Once breaching of the lipid shell is induced by a short temperature increase, the water-soluble drug is enabled to diffuse throughout the hydrogel ( Fig. 5J ) with an initial diffusion constant of 2.1 × 10 −9 m 2 /s (see fig. S17), visualized here with a green food colorant. Water loss in hydrogel electronics may be an issue for prolonged operation. Several options exist to address this, from adding hygroscopic salts to designing (stretchable) encapsulations ( 16 , 55 ). The latter is challenging; however, our hydrogel electronic skins are in direct contact with skin on one side and have a PET encapsulation bearing the sensor electronics on the other side. In addition, a modular approach that allows replacement of the hydrogel part after use may alleviate these issues and additionally addresses concerns of hygiene. Fig. 5 Hydrogel electronic skin. ( A ) Concept of a hydrogel smart skin, with a flexible unit bearing power supply, control, readout and communication units, and a stretchable transducer batch. PCB, printed circuit board. ( B ) Photograph of an untethered electronic hydrogel with four stretchable heating elements and adjoined temperature sensors strongly bonded to a PVA hydrogel. Battery, control, readout, and Bluetooth low energy communication electronics are hosted on a flexible circuit board. ( C ) Finite element (FE) simulation of the transducer batch with four heating elements enabled. ( D ) Corresponding infrared (IR) thermography image of the freestanding device. Dashed lines in (C) and (D) outline the heater and sensor metal traces. ( E ) Autonomous hydrogel electronic skin controlled and read out continuously via mobile phone, worn on a human wrist with all heaters activated. ( F ) Corresponding IR image after ~450 s with all heater elements activated. ( G ) Measured temperature evolution of the heater elements measured via IR thermography and ( H ) the sensor elements. Switching of the heating elements is indicated by vertical lines. Recorded temperature traces of the sensor elements (solid lines) are in excellent agreement with data taken from IR images (open squares). ( I ) Concept of thermally triggered drug delivery where the substance is enclosed in a lipid shell. Temperature increase melts the shell, releasing the drug into the hydrogel. ( J ) Thermally triggered diffusion of a green food colorant throughout the hydrogel matrix."
} | 7,046 |
33154962 | PMC7591714 | pmc | 4,918 | {
"abstract": "Synthetic biology has played a major role in engineering microbial cell factories to convert plant biomass (lignocellulose) to fuels and bioproducts by fermentation. However, the final product yield is limited by inhibition of microbial growth and fermentation by toxic phenolic compounds generated during lignocellulosic pre-treatment and hydrolysis. Advances in the development of systems biology technologies (genomics, transcriptomics, proteomics, metabolomics) have rapidly resulted in large datasets which are necessary to obtain a holistic understanding of complex biological processes underlying phenolic compound toxicity. Here, we review and compare different systems biology tools that have been utilized to identify molecular mechanisms that modulate phenolic compound toxicity in Saccharomyces cerevisiae. By focusing on and comparing functional genomics and transcriptomics approaches we identify common mechanisms potentially underlying phenolic toxicity. Additionally, we discuss possible ways by which integration of data obtained across multiple unbiased approaches can result in new avenues to develop yeast strains with a significant improvement in tolerance to phenolic fermentation inhibitors.",
"conclusion": "Future Outlook and Concluding Remarks Overall, different systems biology approaches have been used to track global phenolic stress responses in S. cerevisiae . While common themes or mechanisms coincide among multiple studies, the different approaches provide alternative pathways and biological processes that can be exploited for strain improvement. Moving forward, since most of the long list of genetic hits reported in the various studies has not been validated, significant effort is required to confirm their actual role in tolerance or sensitivity to phenolics. This is particularly crucial for the transcriptomics data because the fact that a gene is upregulated or enriched during stress does not necessarily mean an overexpression of that gene will result in tolerance ( Evans, 2015 ). Gene enrichment could merely be a stress response and not a tolerance mechanism. If possible, the role of enriched genes associated with phenolic tolerance should be confirmed by deleting and/or overexpressing target genes in cells grown in the presence phenolic inhibitors. Such confirmed genes should be cataloged in a “phenolic stressome” database similar to the yStreX ( Wanichthanarak et al., 2014 ) as a repository where synthetic biologists can search for genetic targets to engineer tolerance to different phenolic compounds. By applying synthetic biology tools, such as the CRISPR/Cas technology, the expression of single or multiple genes identified in the “phenolic stressome” can be regulated in order to improve tolerance to phenolic compounds. Finally, establishing the metabolomic profile of S. cerevisiae that are tolerant to a wide spectrum of individual phenolics may guide the development of biosensors to detect “signature metabolites” characteristic of tolerant and high-performing strains. Again, using synthetic biology, biosensors can be constructed with promoters (that are responsive to metabolites characteristic to tolerant and high-performing strains) and a reporting system (e.g., GFP), and inserted into a library of yeast mutants. Next, by applying microfluidics, a pool of heterogenous yeast mutants can be sorted to isolate phenolic tolerant strains that can be used in fermentation-based biomanufacturing to increase product yield and titers.",
"introduction": "Introduction Biomanufacturing is transforming how new and existing platform chemicals are made in a way that is environmentally friendly, renewable, and sustainable. To make bio-derived chemicals competitive to fossil-derived chemicals, high productivity and cost reduction are a major consideration. Therefore, there has been a growing interest in using cheap and readily available feedstocks, such as plant material (lignocellulose) obtained from agricultural and forestry wastes. Lignocellulose is an abundant and ubiquitous biomass feedstock that can be hydrolyzed to yield simple sugars which are fermented by yeast to produce bioethanol, fine chemicals, and other bioproducts ( Abo Bodjui et al., 2019 ). However, converting lignocellulose to these products during biomanufacturing has its challenges. The sugars in lignocellulose exist as long polysaccharide chains in the form of cellulose and hemicellulose which are held together by lignin ( Becker and Wittmann, 2019 ). In order to make the cellulose and hemicellulose polymers accessible for hydrolysis to release sugars for fermentation, a pre-treatment step is required to dissolve the lignin fibers holding the sugar polymers. While physical (pyrolysis), physicochemical (ammonia fiber explosion) and biological methods exist for pre-treating lignocellulose, chemical pre-treatment methods are commonly used since they are simple and efficient ( Becker and Wittmann, 2019 ). Chemical pre-treatment involves the use of dilute acid or alkali to break down the lignin. As a result, phenolic compounds which are monomeric subunits of lignin are produced during the pre-treatment step ( Palmqvist and Hahn-Hägerdal, 2000 ). Phenolic compounds inhibit enzymes used to hydrolyze cellulose ( Qin et al., 2016 ) and in effect, limit the amount of sugars available for fermentation. Phenolics are also extremely toxic to yeast even in minute quantities and significantly inhibit yeast growth and fermentation ( Ando et al., 1986 ; Adeboye et al., 2014 ) thus, reducing the product yield and increasing the cost of fermentation. Phenolic compounds exist in different forms in lignocellulosic hydrolysates as phenolic acids (e.g., ferulic acid), phenolic aldehydes (e.g., vanillin), phenolic ketones (e.g., 4-hydroxyacetophenone), and phenolic alcohols. The concentrations of each of these compounds in hydrolysates vary depending on the plant material and the pre-treatment method used. They appear to have different toxic effects on the cell with phenolic aldehydes being the most toxic and completely inhibit yeast growth at concentrations as low as 1 and 5 mM for coniferyl aldehyde and vanillin, respectively ( Adeboye et al., 2014 ). The different levels of toxicity of multiple phenolic compounds were confirmed in a study which showed that the chemical nature of phenolic compounds determine their toxicity and the physiological impact they have on the cell ( Adeboye et al., 2014 ). This study was backed by another report that demonstrated that ferulic acid and coniferyl aldehyde, though structurally similar with the only difference being the functional group, presented very distinct chemogenomic profiles and inhibited yeast growth using specific mechanisms ( Fletcher et al., 2019 ). Apart from converting lignocellulosic materials to bioethanol and other chemicals by fermentation, there has been a recent interest in valorizing lignin in lignocellulose to produce precursors and final products for the fine chemicals industry ( Becker and Wittmann, 2019 ; Li et al., 2019 ; Ponnusamy et al., 2019 ). Vanillin is an example of a valuable phenolic compound in the fine chemicals industry mainly used as flavor or scent in food, pharmaceuticals and cosmetics ( Luziatelli et al., 2019 ). While a process has been developed for fermenting glucose to vanillin ( Brochado et al., 2010 ), ferulic acid is an important precursor which can be converted to vanillin by engineering microbial cell factories to express feruloyl-CoA synthase and feruloyl-CoA hydratase ( Luziatelli et al., 2019 ). Other phenolic compounds, such as eugenol present in grains and cereals can be converted to ferulic acid and subsequently to vanillin ( Overhage et al., 2002 ; Di Gioia et al., 2009 ). Again, a major limitation of using engineered yeasts for ferulic acid conversion to vanillin is the issue of toxicity of both vanillin and its ferulic acid precursor. It is possible to remove phenolic compounds from lignocellulosic hydrolysates as they form to prevent toxicity to the yeast cells ( Carter et al., 2011 ; Xue et al., 2018 ) but this comes at an extra manufacturing cost. Therefore, to cost-effectively achieve high yields of bioethanol and other bioproducts from lignocellulose by fermentation, there is the need to improve tolerance to phenolic fermentation inhibitors in yeast cell factories that are used for the bioconversion. A thorough understanding of the mechanisms that modulate phenolic compound toxicity is required to engineer yeast strains that are tolerant to individual phenolic compounds and/or a complex mix of phenolics found in hydrolysates. As inhibitor tolerance is a multigenic complex trait ( de Witt et al., 2019 ) global cellular approaches are required to identify key determinants associated with phenolic compound tolerance. Advances in systems biology approaches have revolutionized our ability to assess how cells respond to phenolic toxicity. The use of genome-wide approaches have given insight into how the cell responds to individual phenolics and identified genetic and metabolic targets that can be engineered to improve tolerance to toxic phenolic fermentation inhibitors. However, a comprehensive understanding of the phenolic tolerance pathway remains lacking since data from the individual studies have not been fully integrated. Here, we review several unbiased functional genomics and transcriptomic approaches to identify general and specific genetic targets that modulate phenolic compound toxicity in S. cerevisiae . We also highlight the potential of exploiting proteomics and metabolomics approaches, which remain underutilized in the field. Finally, synthetic biology approaches and future developments that can rapidly be used to generate yeast tolerant to phenolic fermentation inhibitors are discussed."
} | 2,451 |
35346686 | PMC9062432 | pmc | 4,919 | {
"abstract": "The actinobacterium Rhodococcus jostii RHA1 grows on a remarkable variety of aromatic compounds and has been studied for applications ranging from the degradation of polychlorinated biphenyls to the valorization of lignin, an underutilized component of biomass. In RHA1, the catabolism of two classes of lignin-derived compounds, alkylphenols and alkylguaiacols, involves a phylogenetically distinct extradiol dioxygenase, AphC, previously misannotated as BphC, an enzyme involved in biphenyl catabolism. To better understand the role of AphC in RHA1 catabolism, we first showed that purified AphC had highest apparent specificity for 4-propylcatechol ( k cat / K M ∼10 6 M −1 s −1 ), and its apparent specificity for 4-alkylated substrates followed the trend for alkylguaiacols: propyl > ethyl > methyl > phenyl > unsubstituted. We also show AphC only poorly cleaved 3-phenylcatechol, the preferred substrate of BphC. Moreover, AphC and BphC cleaved 3-phenylcatechol and 4-phenylcatechol with different regiospecificities, likely due to the substrates’ binding mode. A crystallographic structure of the AphC·4-ethylcatechol binary complex to 1.59 Å resolution revealed that the catechol is bound to the active site iron in a bidentate manner and that the substrate’s alkyl side chain is accommodated by a hydrophobic pocket. Finally, we show RHA1 grows on a mixture of 4-ethylguaiacol and guaiacol, simultaneously catabolizing these substrates through meta- cleavage and ortho- cleavage pathways, respectively, suggesting that the specificity of AphC helps to prevent the routing of catechol through the Aph pathway. Overall, this study contributes to our understanding of the bacterial catabolism of aromatic compounds derived from lignin, and the determinants of specificity in extradiol dioxygenases.",
"discussion": "Discussion This study establishes that AphC has a strong substrate specificity for 4-alkylcatechols, cleaving 4-propylcatechol with an apparent k cat / K M approximately three orders of magnitude higher than catechol. This specificity is consistent with the enzyme’s role in the catabolism of 4-alkylphenols and 4-alkylguaiacols by R. jostii RHA1 ( 13 , 14 ). A crystal structure of AphC in complex with the substrate 4EC revealed that the enzyme’s catalytic machinery is essentially identical to that of other type I extradiol dioxygenases and that the alkyl substituent is accommodated by a large hydrophobic pocket that is contiguous with the catechol binding site. The catechol binding site is similar between AphC and BphC LB400 , whose best substrate is 3-phenylcatechol, but the proximal pocket is extended either out from the catechol -C4 or -C3 carbon, respectively. The orientation of this proximal hydrophobic pocket appears to be a major determinant of substrate specificity. The substrate specificity of AphC is unique among characterized extradiol dioxygenases ( Table S2 ). For example, XylE ( 25 ) and AtdB ( 26 ), enzymes whose physiological role is also to cleave alkylcatechols, have highest specificity for unsubstituted catechol. Interestingly, AphC is as phylogenetically distant from these two enzymes as it is to BphC ( Fig. S1 ) ( 27 ). Notably, the apparent specificities ( k cat / K M ) of these enzymes for their preferred substrates are one to two orders of magnitude higher than that of AphC for 4-propylcatechol. LapB, involved in the catabolism of 4- n -alkylphenols (C 1 –C 5 ), has similar kinetic parameters for 4-methylcatechol, a physiological substrate, as those of AphC for 4-alkylcatechols ( 28 ). However, this enzyme is also phylogenetically distant from AphC, suggesting that the similar specificities of these enzyme may have arisen independently. Interestingly, AphC is closely related to a family of predicted catechol 2,3-dioxygenases that occur in Chloroflexi, exemplified in Fig. S1 by the sequence identified as PheC Ca from Chloroflexus auranticus . The basis for the different regiospecificities of AphC and BphC for 3-phenylcatechol and 4-phenylcatechol is unclear. In principle, the different regiospecificity of these enzymes could arise in one of two ways. In the case of 3-phenylcatechol, AphC would either have to accommodate the substrate in a flipped orientation versus BphC or direct O 2 attack to C1 of the catecholic substrate instead of C2, as proposed for BphC ( 23 ). The architecture of the catalytic machinery is highly conserved in these homologs. The AphC iron ion, two His-carboxylate triad, His200, His247, and Tyr258 overlay with r.m.s.d values of 0.54 Å and 0.78 Å over all heavy atoms to BphC and DoxG, respectively. Given the conserved orientation of catalytic residues between AphC, BphC, DoxG and indeed other well-characterized enzymes such as 2,3-HPCD ( 18 ), particularly His249 and Tyr258 involved in catechol deprotonation and His200 involved in protonating the activated oxygen intermediate, the mode of O 2 attack would almost certainly be conserved. This suggests that the differences in regiospecificity are due to 3-phenylcatechol binding in a flipped orientation. Our substrate modeling studies indicate that the binding of 3-phenylcatechol to AphC result in steric clashes regardless of its orientation, suggesting that some structural change occurs for activity. The low specific activity of AphC for 3-phenylcatechol is consistent with the poor fit of the substrate to the enzyme’s binding site. DoxG has an ∼260-fold higher specificity constant for 4-phenylcatechol over 3-phenylcatechol, and a much lower K m for 4-phenylcatechol ( 24 ). The DoxG substrate binding pocket is larger than that of AphC and BphC LB400 , extending open from both the catechol C3 and C4 positions, so the preference for 4-phenylcatechol may be driven by favorable stacking of the phenyl against Tyr178 or other interactions in the hydrophobic pocket ( Fig. 2 E ). The presented data establish that RHA1 simultaneously catabolizes similar aromatic compounds by ortho -cleavage and meta -cleavage pathways, respectively. Thus, RHA1’s growth on a mixture of 4-ethylguaiacol and guaiacol was monophasic, and the two substrates were depleted with similar kinetics during this growth. Moreover, both pathways were required for growth on the mixture, and both intradiol and extradiol dioxygenases were present under these conditions. This simultaneous growth is somewhat unexpected given the potential of forming dead end metabolites such as 4-methylmuconolactone if 4-alkylcatechols are subject to ortho -cleavage ( 17 ). Our data suggest that the respective specificities of AphC and CatA play a major role in ensuring that the catechols are correctly routed. In this respect, it is noted that the relative specificity of AphC for 4EC over catechol (∼140) is less than that of BphC LB400 for 3-phenylcatechol over catechol (∼350 ( 27 )) which are routed through meta -cleavage and ortho -cleavage pathways, respectively, in P. xenovorans LB400 and other biphenyl-degrading bacteria. However, it is possible that compartmentalization plays a role in correctly routing the catechols. In conclusion, this study establishes the substrate specificity of a phylogenetically distinct alkylcatechol-cleaving extradiol dioxygenase and identifies a major structural determinant of that specificity. In addition, the enzyme’s specificity appears to play a major role in the correct routing of aromatic compounds in Rhodococcus ."
} | 1,854 |
35493553 | PMC9042265 | pmc | 4,920 | {
"abstract": "Introducing double physical crosslinking reagents ( i.e. , a hydrophobic monomer micelle and the LAPONITE® XLG nano-clay) into the copolymerization reaction of hydrophilic monomers of N , N -dimethylacrylamide (DMAA) and acrylamide (AM) is reported here by a thermally induced free-radical polymerization method, resulting in a highly tough and rapid self-healing dual-physical crosslinking poly(DMAA- co -AM) hydrogel. The mechanical and self-healing properties can be finely tuned by varying the weight ratio of nanoclay to DMAA. The tensile strength and elongation at break of the resulting nanocomposite hydrogel can be modulated in the range of 7.5–60 kPa and 1630–3000%, respectively. Notably, such a tough hydrogel also exhibits fast self-healing properties, e.g. , its self-healing rate reaches 48% and 80% within 2 and 24 h, respectively.",
"conclusion": "4. Conclusions We have innovatively combined the appropriate raw materials selection and synthesis protocol to generate a highly tough and rapidly self-healing dual-physical crosslinking hydrogel. The dual-physical crosslinking network is formed by using hydrophobic monomer micelles and nanoclay. The central role of the nanoclay inclusion in significantly improving the mechanical properties and self-healing performance is elucidated. Incorporating an appropriate content of nanoclay gives rise to the homogeneous distribution of the physical crosslinking points, thus facilitating the efficient load stress transfer out of the hydrogel substrate and significantly reducing the destructive impact. The clay nanoadditive can be effectively incorporated into the polymer matrix, as proven by the coarsened surface of the hydrogel pore wall under SEM, thereby strengthening the wall and promoting the mechanical performance of the hydrogel material. Nevertheless, the over-load of the nanoclay degrades the mechanical properties of the NC hydrogel in that the highly concentrated nanoclay would much constrain the polymer chain motion, lowering the toughness. Notably, the optimized NC hydrogel exhibits high toughness and rapid self-healing properties, as evidenced by various characterization techniques such as metallographic microscope observation, tensile tests, dynamic rheological curve measurements. This work opens up a new avenue to construct highly tough and rapidly self-healing hydrogel materials for many practical applications of significance, such as chemical and biochemical engineering and biomedical scaffolds.",
"introduction": "1. Introduction With chemically or physically crosslinked three-dimensional (3D) network structures, polymer hydrogels possess the ability to absorb large amounts of water and resist dissolution. 1–3 The excellent biocompatibility, porous structure, tunable stiffness, and biological tissue-like elasticity of polymer hydrogels impart widespread use as biomaterials in the fields of tissue engineering, 4–6 gene and drug delivery systems, 7–9 cell cultures, 10–12 superabsorbents, 13–16 biosensors, 17–20 artificial e-skins, 21–23 wound healing, 24–26 etc. However, conventional chemically crosslinked hydrogels are always brittle and exhibit poor mechanical performance due to the absence of an energy dissipation mechanism, limiting their widespread applications. 27–30 Functional biomimetic self-healing hydrogels possess remarkable performance, such as fast self-healing properties and high mechanical strength and toughness, which greatly extend their service life, especially in artificial e-skins, biosensing, and drug delivery. 31–35 Different approaches for repairing the damaged network structures of hydrogels have been explored to obtain highly self-healing hydrogels. The strategies to design self-healing hydrogels are mainly based on healing agents, dynamically reversible covalent bonds, or dynamically non-covalent bonds. 36–39 Up to now, the frequent use of non-covalent bonds to prepare self-healing hydrogels has primarily been based on metal–ligand interactions, 40–43 host–guest interactions, 44–46 hydrophobic interactions, 47–50 hydrogen bonding, 51–57 ionic interactions, 58–62 etc. Since non-covalent bonds are generally more susceptible to the external environment than dynamic covalent linkage, the self-healing hydrogel systems based on non-covalent interactions, or the self-healing supramolecular hydrogel systems, will be more intelligent and easier for structural control than those based on dynamically reversible covalent bonds. 63,64 Among various non-covalent bonding, hydrophobic bonding is the most common for bulk self-healing materials, thus playing a dominant role in forming large biological systems. 65–67 Therefore, applying hydrophobic interactions to construct robust hydrogels with superior self-healing ability and remarkable mechanical properties is highly desirable. In particular, Okay and co-workers 47 presented a simple strategy to create strong hydrophobic interactions by incorporating hydrophobic chains into hydrophilic polymer network chains, resulting in self-healing hydrogels. In addition, the finite lifetime of hydrophobic associations between stearyl methacrylate (C18) blocks made the resultant hydrogel highly tough. The breakage of the hydrogel samples bearing C18 blocks occurred at elongation ratios of 3600%, and a self-healing efficiency to elongation at break of about 100% was achieved. 47 However, these hydrogels suffer from low tensile strength and resilience, thus impeding their practical applications. Hydrogels based on nanocomposite (NC) 68 and dual networks (DN) 69 are regarded as novel and functional self-healing materials, and NC hydrogels have drawn the most widespread attention because of their high strength and versatility. NC hydrogels can be prepared by physically crosslinking polymer chains with nanoparticles such as inorganic clay LAPONITE® XLG. The unique polymer/clay composite network structure provides NC hydrogels with self-healing capability. 70–76 The interactions between the polymer chains and the adjacent clay particles are considered non-covalent or hydrogen bonding. Adjoining two such surfaces facilitates grafted chains at the two surfaces to diffuse into each other and interact with neighboring clay platelets by hydrogen bonding. Consequently, many new crosslinks are generated across the interface, which leads the cut surfaces to rejoin. Thus, the macroscopic self-healing of NC hydrogels arises from the reconstruction of the network at the interface due to the mutual diffusion of long grafted chains and their subsequent interactions with clay. 70,75 This study judiciously combined the appropriate materials selection and synthetic processing to construct a highly tough and rapidly self-healing dual-physical crosslinking hydrogel material. We introduced a nanoclay (LAPONITE®) and a hydrophobic monomer micelle into the copolymerization system of acrylamide (AM) and N , N -dimethylacrylamide (DMAA) to generate a dual-physical crosslinking network through a thermally induced free-radical polymerization process. The synthetic process of the dual-physical crosslinking hydrogel is presented in Fig. 1 . The generation of a dual-physical crosslinking network is realized by introducing inorganic nanoclay (leading to one physical crosslinking network) and organic hydrophobic monomer micelles (resulting in the other physical crosslinking network) into the copolymerization reaction system of hydrophilic DMAA and AM monomers. Hydrophobic and ionic interactions within the NC hydrogel matrix render it highly tough and rapidly self-healing. Fig. 1 Schematic illustration of the poly(DMAA- co -AM) hydrogel synthesis with dual-physical crosslinking network as generated by introducing the nanoclay and hydrophobic monomer micelles.",
"discussion": "3. Results and discussion 3.1. Structural characterization We first carried out a preliminary screening test on the prepared series of hydrogel samples and selected representative ones for the subsequent systematic characterizations, and the selected three samples were designated as 0.15DPNC, 0.2DPNC, and 0.3DPNC, where the number represents the content of the nanoclay ( vs. 2.0 g DMAA, at the DMAA-to-AM mass ratio of 8 : 1). Fig. 2 shows the FITR spectra of these representative samples. While the peak at 1603 cm −1 corresponds to N–H (amide II) bending vibrations, the peaks at 3393, 2922, 1252, and 1660 cm −1 can be assigned to the stretching vibrations of N–H, CH 2 –, C–O, and 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 (amide I), respectively. These absorption peaks present the characteristics of the poly(AM- co -DMAA) skeletons, thus verifying the successful synthesis of the poly(AM- co -DMAA) hydrogel. With a change in the nanoclay content, no significant variation can be noted in the FTIR spectra, suggesting that physical rather than chemical interactions exist between the nanoclay and poly(AM- co -DMAA) chains. The chemical structure of the synthesized the poly(AM- co -DMAA) hydrogel was also characterized by 1 H NMR analyses. Fig. S1–S3 † display the 1 H NMR spectra of 0.15DPNC, 0.2DPNC, and 0.3DPNC, respectively. For all these hydrogels, the spectra can be divided into three main regions: 2.8–3.6 ppm (6H) attributed to the methyl protons of the amide function, 2.2–2.6 ppm (1H) indexed into the poly(AM- co -DMAA) backbone methine protons, and 0.8–1.9 ppm (2H) corresponding to the poly(AM- co -DMAA) backbone methylene protons. 77–79 Fig. 2 FTIR spectra of the poly(AM- co -DMAA) hydrogels: (a) 0.15DPNC, (b) 0.2DPNC, and (c) 0.3DPNC. 3.2. Swelling properties The swelling behavior of the hydrogels is one of the critical performance indexes for biomaterials. The swelling ratio is usually adopted to evaluate the absorbability and structural stability of hydrogel materials. The swelling behavior of the poly(AM- co -DMAA) hydrogels is evaluated through a long-term immersion process in water, and the results are provided in Fig. 3 – 5 . A 4-fold increase in the volume of the swelled hydrogel relative to the as-obtained hydrogel can be noted. The drying treatment of the as-obtained hydrogel leads to the formation of a plastic sheet ( Fig. 3 ). These findings indicate that the prepared hydrogel possesses an appreciable water absorption capacity. With an increase in the clay content, the hydrogel swelling ratio increases initially and then decreases. The hydrogels swelling ratio in this process is shown in Fig. 4 . After swelling in water for 1 h, the swelling ratio of the 0.2DPNC hydrogel reaches 100% of the original mass, much higher than that of other poly(AM- co -DMAA) hydrogels, thus implying that the 0.2DPNC hydrogel absorbs water at a higher speed. The EWCs calculated for the typical hydrogels are shown in Fig. 5 . The 0.2DPNC sample exhibits the highest EWC (99%) among the investigated samples, confirming the excellent water absorption capacity that likely results from the compact double crosslinking network (formed by the inorganic nanoclay and hydrophobic organic monomer micelles) and the presence of a microporous structure with the ability to hold a large amount of water. A further increase in the nanoclay content yields the 0.3DPNC sample, but the water absorption capacity becomes lowered because the nanoclay overload reduces the number of macropores. Fig. 3 (a–f) Photo images of the as-prepared 0.2DPNC hydrogel sample (a and d), the swelled 0.2DPNC hydrogel sample after water absorption equilibrium (b and e), and the corresponding oven-dried sample (c and f). Fig. 4 Plots of the swelling rate of the hydrogels as a function of the swelling duration. Fig. 5 EWCs estimated for the three representative 0.15DPNC, 0.2DPNC, and 0.3DPNC hydrogel samples. 3.3. Morphology observation The SEM morphologies of the cross-section of the typical poly(AM- co -DMAA) hydrogels are presented in Fig. 6 . It can be noted that these hydrogel samples exhibit a honeycomb-like porous structure. Having an average pore size of approximately 50 μm, the porous structure exhibits uniformity, homogenous distribution, and high density. The higher concentration of the crosslinker would give rise to a more compact crosslinking network. The incorporation of the nanoclay can improve the composite hydrogel toughness and, consequently, contribute to maintaining structural integrity during the swelling process. The compact crosslinking network contributes to the impressive toughness of the NC hydrogel. Notably, incorporating the nanoclay into the poly(AM- co -DMAA) makes the pore wall surface rougher, revealing the dispersion of the inorganic nanoclay within the polymer matrix. As a result, the strengthening effect of the inorganic nanoclay can be effectively imparted to the composite hydrogel. Fig. 6 (a–c) SEM images of the representative 0.15DPNC (a), 0.2DPNC (b), and 0.3DPNC (c) hydrogels. 3.4. Mechanical property evaluation The mechanical properties of the hydrogel materials usually dictate their practical application potential. In this regard, we first tested the knotted long cylinder hydrogel sample by stretching with hands to see whether it would rupture at the knotted site and evaluate the toughness ( Fig. 7 ). The good toughness can be revealed by the finding that the knotted long cylinder hydrogel can be stretched to a significant extent that it becomes many times longer than that before stretching. Fig. 7 Photo images illustrating the robustness of the as-prepared 0.2DPNC hydrogel: the 0.2DPNC hydrogel strip-tied knot before (left) and after (right) stretching with hand. A universal tensile machine was employed further to assess the mechanical properties of the typical samples. The specimens with a rectangular shape and a size of 20 mm × 5 mm × 1 mm were prepared for the uniaxial extension test ( Fig. 8 ). The photo image (on the right side) presents that the typical hydrogel can be highly stretched. The strain–stress plots in Fig. 8 show that the 0.15DPNC sample with the lowest nanoclay content performs the worst in the tensile test by considering its lowest tensile strength as a result of the smallest number of the crosslinking sites afforded by the nanoclay. With an increase in the nanoclay content, the tensile strength of the resulting 0.2DPNC sample is markedly increased, along with a noticeable improvement in elongation at break (exceeding 3000%). A further increase in the nanoclay content degrades the tensile properties of the resultant 0.3DPNC sample, as caused by the overhigh content of the nanoclay that would substantially increase the contact area between the polymer chains and nanoclay, consequently restraining the molecular chains to a great extent. Fig. 8 Strain–stress plots for the typical composite hydrogel samples with different nanoclay contents. The photo image on the right side presents the typical 0.2DPNC hydrogel before and after the uniaxial extension test. 3.5. Self-healing and mechanical stretching behaviors Smart hydrogels have become the research focus of many important areas, such as biomedicines, biochemical engineering, and environmental remediations. Self-healing hydrogels are under the spotlight due to their viability for practical biochemical engineering applications. Notably, the present hydrogel exhibits excellent self-healing performance ( Fig. 9 ). The parts cut from the hydrogel can be healed along the contact area within 1 h, yielding an integrated hydrogel that can be highly stretched. We dyed one portion in red and connected to the other along the cut surface to present the self-healing performance. The efficient self-healing performance most likely results from the self-assembling capability of the hydrophobic octadecyl methacrylate sections in the hydrophobic micelles. The non-covalent interactions occur between the self-assembled sections/DMAA chains and the nanoclay surface, resulting in physical crosslinking points. The dual-physical crosslinking networks provide the composite hydrogel with rapid self-healing capacity. Fig. 9 Photo image demonstrating the self-healing performance of the typical 0.2DPNC hydrogel. A metallographic microscope was further employed to observe the self-healing effect of the typical sample at the microscopic level ( Fig. 10 ). A scratch was created on the thin hydrogel slice using a sharp blade, which was then monitored under the metallographic microscope. Of note, the scratch almost disappeared within 3 min. We also employed the universal tensile machine to measure the tensile properties of the self-healed samples prepared by contacting the cut parts for 2 h and 24 h. The tensile plots of the samples self-healed for 2 h and 24 h are presented in Fig. 11 . After self-repaired for 2 h, the integrated hydrogel sample exhibits an elongation at break of 1503%, corresponding to the self-healing rate of 48.7%. In contrast, self-healing for 24 h yields the integrated hydrogel possessing an elongation at break of 2334%, corresponding to the self-healing rate of as high as 80%, basically the same as the original sample. Thus, these results prove the rapid self-healing properties of the present poly(AM- co -DMAA)-based nanocomposite hydrogel, different from a similar poly(AM- co -DMAA)-based nanocomposite hydrogel without self-healing properties albeit with good mechanical strength. 80,81 Notably, our dual-physical crosslinking poly(DMAA- co -AM) hydrogels incorporating a small amount of clay exhibit both excellent elongation at break and self-healing properties. The properties of similar poly(AM- co -DMAA)-based nanocomposite hydrogel are listed in Table 1 . Fig. 10 Metallographic microscope images showing the self-healing process of the typical 0.2DPNC composite hydrogel for different durations: (a) 0 min, (b) 1 min, (c) 2 min, and (d) 3 min test. Fig. 11 Tensile test for the 0.2DPNC hydrogel integrated by self-healing for 2 h (black line) and 24 h (red line). The properties of similar poly(AM- co -DMAA)-based nanocomposite hydrogels Similar hydrogels Tensile strength Elongation at break Self-healing TAD gels 80 50–367 kPa 500–1760% No NCP gels 81 18–313 kPa About 1600–3205% No DPNC gels 7.5–60 kPa 1630–3000% Yes 3.6. Rheological property investigation The rheological properties of the samples were examined by a dynamic rheology test. A strain amplitude sweep measurement was performed to analyze the storage modulus G ′ and the loss modulus G ′′ of this hydrogel vs. the oscillatory strain amplitude, with the results shown in Fig. 12 . When the parameter γ is smaller than 100%, the hydrogel exhibits an elastic characteristic as reflected by the constant value of both G ′ and G ′′ ( G ′ > G ′′). The G ′ value decreases rapidly with the parameter γ larger than 100%, making G ′′ approach the value of G ′ and thus implying the collapse of the gel network. Fig. 12 \n G ′ and G ′′ values of the typical hydrogels (0.15DPNC, 0.2DPNC, and 0.3DPNC), as obtained based on the strain amplitude sweep. Rheological characterization was also performed further to demonstrate the robust self-healing properties of these hydrogels. Fig. 13 shows the repeated recovery of the mechanical properties of these hydrogels after being subjected to disruptive mechanical shearing forces. Initially, a small amplitude oscillatory shear (a strain of 1% at a frequency of 1 rad s −1 ) is applied to these hydrogels for 120 s. Under this condition, the G ′ value is greater than that of G ′′, and both do not change with time. This implies that the hydrogel network remains intact under small oscillatory strains. Afterward, the hydrogel is subjected to a large amplitude oscillatory shear (a strain of 500% at a frequency of 1 rad s −1 ) for 40 s. In this case, the G ′ value decreases drastically to approach the G ′′ value. This is indicative of the rupture of the gel network. Subsequently, a low amplitude strain of 1% was applied to the hydrogel within the constant regime shown in Fig. 13 . The gel-like character (featuring G ′ > G ′′) was recovered within 10 s, along with the recovery of G ′ and G ′′ to their initial values within 160 s. The complete recovery of the destructed hydrogel network confirms the self-healing capacity of the hydrogel. Fig. 13 \n G ′ and G ′′ values estimated based on a continuous strain sweep with the hydrogels alternatively subjected to small (1% strain) and large (500% strain) oscillation forces."
} | 5,205 |
27387283 | PMC5001620 | pmc | 4,921 | {
"abstract": "DNA nanoassemblies have demonstrated wide applications in various fields including nanomaterials, drug delivery and biosensing. In DNA origami, single-stranded DNA template is shaped into desired nanostructure by DNA staples that form Holliday junctions with the template. Limited by current methodologies, however, mechanical properties of DNA origami structures have not been adequately characterized, which hinders further applications of these materials. Using laser tweezers, here, we have described two mechanical properties of DNA nanoassemblies represented by DNA nanotubes, DNA nanopyramids and DNA nanotiles. First, mechanical stability of DNA origami structures is determined by the effective density of Holliday junctions along a particular stress direction. Second, mechanical isomerization observed between two conformations of DNA nanotubes at 10–35 pN has been ascribed to the collective actions of individual Holliday junctions, which are only possible in DNA origami with rotational symmetric arrangements of Holliday junctions, such as those in DNA nanotubes. Our results indicate that Holliday junctions control mechanical behaviors of DNA nanoassemblies. Therefore, they can be considered as ‘mechanophores’ that sustain mechanical properties of origami nanoassemblies. The mechanical properties observed here provide insights for designing better DNA nanostructures. In addition, the unprecedented mechanical isomerization process brings new strategies for the development of nano-sensors and actuators.",
"conclusion": "CONCLUSIONS Serving as essential components inside DNA origami nanoassemblies, DNA staples offer stability to DNA nanoassemblies by forming Holliday junctions with the single-stranded DNA template. Using optical tweezers, we have quantified and explained two emergent properties of DNA nanoassemblies not seen in their individual components from mechanical perspective. First, we have found the mechanical stability of DNA origami structures can be geometry dependent, which is determined by the effective density of Holliday junctions along a particular stretching direction. Second, we have quantitatively ascribed the cooperative transition between short and long DNA nanotubes to the collective mechanical isomerizations of many individual Holliday junctions. These observations indicate that the Holliday junctions serve as mechanophores in DNA origami nanoassemblies investigated here. Further testing is required to validate this point for more complex DNA nanoassemblies such as DNA helix bundles ( 2 , 21 , 22 , 45 ). We anticipate our new findings can provide unprecedented guidelines to design DNA nanoassemblies with better mechanical properties in both thermodynamic and kinetic aspects.",
"introduction": "INTRODUCTION Because of the highly stable and specific recognition between two complementary DNA strands, DNA has been used as an attractive component in nanoassembly. In the DNA origami nanoassembly, a long ssDNA serves as a template to fold into nanostructures through hundreds of Holliday junctions formed between short DNA staples complementary to the template sequences at particular locations ( 1 ). The simple, robust and highly efficient synthesis strategy of DNA origami has established this structure as a highly potent nanomaterial ( 2 – 10 ) yet to be fully characterized for its properties. Among these unknown territories, mechanical property is certainly a notable missing link with high significance. Mechanical stability of the connecting regions in DNA nanoassemblies is essential to sustain robust interactions between biomolecules ( 11 ) and nanoassemblies. Likewise, mechanical rigidity of DNA nanocavities is critical to define the morphology of inorganic nanoparticles contained within ( 12 , 13 ). In the sensing applications using DNA nanoassemblies, ( 14 – 18 ), the mechanical rigidity and stability of the DNA nanoassemblies can directly affect the accuracy in the signal output and the sensitivity in the analyte recognition ( 5 ). So far, only a handful of investigations have been reported with a main focus on the mechanical functionalities of the DNA nanoassemblies ( 14 , 19 – 22 ). The insufficient information on the mechanical rigidity and stability of DNA nanoassemblies in general and DNA origami in particular hinders rational design of DNA nanomaterials for applications that exploit their mechanical properties. One reason for this lack of knowledge lies in the difficulty in the characterization of individual nano-objects for their mechanical properties. Special tools with high resolution for mechanical force measurement, such as AFM and optical tweezers, must be employed to carry out the characterization ( 22 , 23 ) after immobilization of nanoparticles with different shapes, which is another challenging practice. We reasoned that due to the flexible nature in the design of a DNA origami with the single-nucleotide precision, it is possible to introduce two duplex DNA handles to tether a DNA origami. These handles are then linked to the two optically trapped polystyrene beads in laser tweezers, which allow the quantification of mechanical properties of DNA nanoassembly. Previously, with the aim to develop origami based nanomechanical devices, Sugiyama and coworkers have found that tubular designs of origami structures can have two stable conformations: a short and a long tubular forms that may stem from different isomers of Holliday junctions contained inside the origami devices ( 10 ). Given the potential applications of controlling conformations of DNA origami structures by mechanical factors as discussed above, here we wish to understand whether it is a unique feature for the tubular-shaped origami to demonstrate different structural isomers. To this purpose, we applied two duplex DNA handles to the tubular DNA origami structures for the comparison of their mechanical properties with those from other nanoassemblies, DNA nanopyramids and DNA nanotiles. We found mechanical stabilities of DNA origami structures are correlated with the effective density of Holliday junctions in a particular nanoassembly. The mechanical stability is anisotropic in nature with the short axis of a DNA nanotube resisting higher external stress than the long axis. Interestingly, mechanical isomerization between two conformations of a DNA nanoassemby at 10–35 pN external force was observed only in DNA nanotubes, which have unique symmetric arrangements of Holliday junctions. Similar to the mechanical stability, the mechanical isomerization also showed anisotropic behavior. Given that individual Holliday junctions have isomerization force in the sub-picoNewton range ( 24 ) and they are anisotropically arranged in DNA nanotubes, we have attributed the mechanical isomerization of the DNA nanotubes to the collective actions of many Holliday junctions that experience similar microenvironment. All these results indicate that Holliday junctions in DNA origami structures serve as mechanophores ( 25 , 26 ), which determine the mechanical property of DNA nanoassemblies, similar to chromophores and fluorophores that carry spectroscopic information in a molecule. We anticipate these new findings are instrumental to optimize the mechanical strength of DNA nanoassemblies.",
"discussion": "RESULTS AND DISCUSSION Mechanical isomerization and mechanical disassembly of DNA tubes Optical tweezers have been used to apply and measure forces in picoNewtons ( 27 – 29 ). Compared to AFM, it has better force resolution and therefore is particularly suitable to characterize mechanical stability of macromolecules such as proteins and nucleic acids. Here, we used a home-built laser tweezers instrument ( 30 , 31 ) to characterize the mechanical properties of individual DNA origami nanoassemblies. To this end, we introduced two double-stranded (ds) DNA handles to DNA origami structures (see Figure 1A , Supplementary Figure S1A and B and Materials and Methods) ( 14 ). DNA handles were inserted into the origami structure via overhang single-stranded staple sequence, which contains 40 nucleotides complementary to the single-stranded M13mp18 template (it has been reported that the shearing force of 30 bp duplex DNA is >60 pN) ( 32 ). The other end of the handle was labeled with a digoxigenin or biotin molecule. These DNA handles were mixed and incubated with rest of the DNA components for the origami construction (see Materials and Methods and Supplementary Figures S1–S7). AFM images have revealed successful incorporation of the handles (see Figure 1B and C , Supplementary Figures S8 and S9 for the handles attached to the DNA nanotubes; see Supplementary Figures S6 and S7 for handle incorporation of other origami structures). Figure 1. Mechanoanalytical characterization of DNA origami nanotubes. ( A ) Schematic of the experimental set up for the mechanical isomerization of DNA origami tubes. DNA origami tube is sandwiched between two long dsDNA handles and the whole construct is tethered between two optically trapped beads with affinity linkages. The mechanochemical property of DNA tubes is revealed by moving one of the beads away from another at a loading rate of ∼5.5 pN/s. AFM image of the eight-tube DNA for longitudinal ( B ) and horizontal stretching ( C ). Schematic pictures are shown to the right of corresponding AFM images. Using the digoxigenin and the biotin molecules labeled at the free end of the DNA handles, the modified DNA origami construct was tethered to the two optically trapped particles coated with digoxigenin antibody and streptavidin through respective affinity interactions ( 14 ). The mechanical stability of the origami was probed by force ramping experiments in which one of the optically trapped particles was moved away from the other at a loading rate of ∼5.5 pN/s in laser tweezers. Due to the unique behavior of isomerization displayed in DNA origami tubes ( 10 ), we chose these origami structures as our first samples to evaluate mechanical properties of DNA origami structures. Using the reported procedures ( 10 ), we prepared eight-tube and six-tube DNA origami nanoassemblies that respectively consist of eight and six Holliday junctions in each of the circular layers that are spirally arranged into desired tubes. Previously, AFM images ( 10 ) have revealed that each type of tube has two equilibrated isomers, long and short tubes (similar to those shown in Figure 1B and C , Supplementary Figures S8 and S9). Due to the limited time resolution in AFM, the isomerization between the two conformers has not been observed in real time. The dynamic nature of mechanical unfolding at the single nanostructure level may provide an unprecedented opportunity to investigate this isomerization process. To this end, by using the procedures reported earlier ( 14 ), either eight-tube (short and long) or six-tube (short and long) DNA origami structures were obtained and tethered to two optically trapped polystyrene beads as described above. In the first set of experiments, we mechanically stretched DNA origami samples from longitudinal direction of the tube (see the eight-tube DNA in Figure 1 as an example). We observed two types of rupture transitions (Figure 2A and B ), which were assigned to mechanical isomerization and disassembly of origami tubes on the basis of the following aspects. First, the two types of transitions can be separated according to the low (10–35 pN) and high (35–60 pN) force of transitions (Figure 2C , two left panels). The hierarchical cluster analysis (HCA) ( 33 ), which is a non-supervised method ( 34 ) to classify groups using the distance between two data points, confirmed the presence of these two populations (Figure 2C , two right panels). Here, we used rupture force as a distance variable in the HCA calculation (see Materials and Methods for details). When we scrutinized the properties of the transition in these two groups, we found that the single or multiple transitions in the lower force region (10–35 pN) were reproducibly observed after incubating the construct for 90 s before each force–extension cycle (Supplementary Figure S10). On the other hand, we observed successive transition events in the higher force region (35–60 pN) that were not reversible after the same incubation time. These high-force transitions, all centered in a similar force region (35–60 pN) even for different DNA tubes (Supplementary Figures S11 and S12), suggesting the disassembly of the structures with similar mechanical stabilities. Since the force range we observed here fell into that for the slide-opening of DNA duplex ( 35 ), we surmised these transitions may be due to the force-induced disassembly of Holliday junctions (shearing breakage), whose re-assembly is slow. Consistent with this scenario, when the tension for the origami in the higher-force region was relaxed to 0 pN, we did not observe the same trajectory in subsequent force-ramping cycles. Instead, the contour length of the origami became longer even after incubation up to 90 s at 0 pN (see curves in the disassembly region in Figure 2B ). Similar events were observed during the mechanical stretching of DNA nanotiles, which have been ascribed to the disassembly of DNA origami structures ( 14 ). Therefore, these high-force transitions likely represent the disintegration of a DNA origami structure through the irreversible breakage in the Holliday junctions during the time scale of the experiment. Figure 2. Mechanochemical properties of DNA origami tubes. ( A ) A typical force–extension ( F – X ) trace for the longitudinal stretching of a eight-tube DNA in optical tweezers experiment. Dotted circles depict the isomerization and disassembly processes. ( B ) A set of F – X traces for another eight-tube DNA origami. The first transition event in the red curve depicts mechanical isomerization. The relaxing curve after mechanical isomerization is shown in black. After incubation, the next stretching curve is identical with the first red curve. After mechanical disassembly, the traces do not overlap with the first red curve after incubation at 0 pN. Instead, they depict longer contour lengths (blue). ( C ) Distribution plots for the transition forces (left two panels) and the hierarchical cluster analyses (right two panels) clearly indicate the presence of two populations. The red population with lower transition force is due to isomerization of the nanotube while the blue population with higher transition force is due to nanotube disassembly. Solid curves in the left panel represent Gaussian fittings. The green data points are identified as stand-alone groups in HCA. We assigned these points to each of the populations based on force. ( D ) Schematic of mechanical isomerization of the short to long eight-tube DNA origami structures followed by disassembly. Magnified sections show the rearrangement of Holliday junctions before and after mechanical isomerization. ( E ) Schematic of the molecular rearrangements of Holliday junctions in a DNA tube during mechanical isomerization and disassembly processes. Numbers in black and green respectively indicate the duplex layers and the Holliday junctions in DNA origami tubes. To explain the nature of the low-force transitions, we analyzed the population of the origami that yielded reversible transitions in the lower force region. We found that 42% (Supplementary Figure S13B) and 75% (Supplementary Figure S10) of the origami showed these transitions in the six-tube and eight-tube DNA stretched from the longitudinal direction, respectively. During longitudinal stretching, only short tubes can isomerize into the long tubes (Figure 2D ), therefore, 42% and 75% populations represent short tube isomers. These values are close to that obtained from AFM images, which showed 45% and 85% (Table 1 ) for the short isomers in the six-tube and eight-tube DNA, respectively. The striking agreement between the mechanical unfolding and AFM analyses strongly suggests that these low-force transitions are likely force-induced isomerizations, which are expected to be reversible after incubation as the origami structures remain intact during the process (Figure 2D and E ). To confirm this scenario, we measured the changes-in-extension due to the rupture transitions in the lower force region. We found that the measured values were closely matched with the calculated difference in length between the short and long isomers in each type of the DNA tube (see next section for details). It is noteworthy that in the force range (10–35 pN) we observed for the isomerization, the changes-in-extension for different DNA nanotubes do not vary with force significantly (calculations see SI). Table 1. Mechanochemical properties of tubular DNA origami structures \n \n Based on these facts, we assigned the lower-force transition as mechanical isomerization while the higher-force transition as disintegration of origami structures (see schematic models in Figure 2D (nanostructure level) and Figure 2E (molecular level)). Given that each Holliday junction consists of Watson–Crick base pairing that can be taken apart by different mechanical force from varying directions, ( 24 , 36 ) it is interesting to see whether or not the same force is required to disintegrate the DNA origami tube from different directions. To test this scenario, we stretched the same type of tube from horizontal direction by using two dsDNA handles facing each other in the middle of the eight-tube DNA (Figure 1C and Supplementary Figure S3). Similar to what was observed for the longitudinal stretching, two regions were found for the horizontal stretching. However, the disassembly process was abrupt. While the lower force region (10-35 pN) showed reproducible features after incubation at 0 pN, the higher force region (35–60 pN) again revealed irreversible disassembly, likely due to disintegration of staples. Similar characteristics were observed during the longitudinal and horizontal stretchings of the six-tube DNA (see below for detailed discussion). Mechanical disassembly (or stability) of a DNA origami structure is correlated with the effective density of Holliday junctions When we scrutinized the higher-force mechanical disassembly region that reflects the mechanical stability of a DNA nanoassembly, we found that the mechanical stability for horizontal stretching of the eight-tube DNA was higher (49 pN, see Supplementary Figure S13A for F – X curves and Figure 3B for histogram) than longitudinal stretching (42 pN, Figure 2A and B for F – X curves and Figure 3A for histogram). Similar anisotropy in mechanical stability was observed for the six-tube DNA origami (see Supplementary Figures S4 and S5 for its structure and preparation and Supplementary Figure S9 for AFM images). While the horizontal mechanical stability was 49 pN, that of longitudinal was lower (44 pN, summarized in Table 1 and Figure 3C , see representative F – X curves in Supplementary Figure S13 B&C). During disassembly, gradual deviation in the extension that corresponds to the sequential rupture of Holliday junctions was not observed in the F – X curves for either horizontal or longitudinal pulling direction. Instead, sudden and big (80–100 nm, see Figure 2A and B ) rupture events that were consistent with the layer-by-layer disintegration during longitudinal pulling were observed (Figure 3E ). Peeling off each layer during layer-by-layer disassociation in the eight-tube DNA requires the breakage of eight Holliday junctions (see black arrows in Supplementary Figure S2), while it requires breaking six Holliday junctions for the same effect in the six-tube DNA (see black arrows in Supplementary Figure S4). On the other hand, with horizontal stretching, the layer-by-layer peeling is rather difficult (Supplementary Figures S3 and S5) and the disassembly of the structure is likely due to force induced melting ( 35 , 37 ) of Holliday junctions (Figure 3F ). As a result, complete disassembly of many more Holliday junctions is necessary to break the tube. During stretching experiments, it is more likely that the loss of the Holliday junction connected to the DNA handles would cause the sudden breakage of the tether (Supplementary Figure S13A and C). These breakage forces are used to represent the lower end of the disassembly forces of the origami tubes at the horizontal direction (Figure 3C ). This geometry difference also predicts that during longitudinal stretching, more intermediates can be accommodated compared to those from horizontal direction. In fact, this has been confirmed during experiments, in which on average 6 intermediates were observed in longitudinal disassembly while it contained only one intermediate in horizontal direction (Figure 3D ). Figure 3. Disassembly processes of DNA origami structures. ( A and B ) Histograms of the mechanical disassembly forces measured along longitudinal and horizontal stretching directions of DNA eight-tubes, respectively. The disassembly force was determined at the first disassembly event (the red arrow in Figure 2B ). The insets depict the stretching directions. N and n respectively indicate the number of F – X traces and the number of molecules. ( C ) Comparison of the disassembly force of DNA nanotiles (Tiles), DNA nanopyramid (Py), eight-tube DNA (8T) and six-tube DNA (6T). ( D ) Average number of the disassembly events observed during the mechanical disintegration of the 8-tube DNA (8T) in longitudinal (L) or horizontal (H) stretching. ( E and F ) Schematics of the disassembly with respect to Holliday junctions during longitudinal and horizontal stretching directions, respectively. Dotted yellow lines depict Holliday junction layers in different stretching orientations. For clarity, only one complete Holliday junction is shown in the white circle in each scheme. To further characterize the mechanical stability of DNA origami structures, we performed stretching experiments of origami nanotiles ( 14 ) (see Supplementary Figure S6 for design and AFM image, S13D for F – X curves) and pyramid-shaped origami (see Materials and Methods and Supplementary Figure S7 for the design and AFM image, Supplementary Table S1 for the staple sequences and Supplementary Figure S13 E for F – X curves). As shown in Figure 3C , DNA nanotiles have shown the weakest mechanical stability (disassembly force 28 pN), which was followed by a 3D structure, DNA nanopyramid (35 pN). The most compact 3D structures, horizontal eight-tube and six-tube DNA, demonstrated the strongest stabilities (49.4 and 48.5 pN respectively). Due to the open geometry in the 2D origami nanotiles ( 14 ), disassembly of the structure via stretching only involves the breakage of Holliday junctions located close to the axis of the stress. In contrast, in the 3D DNA assemblies, the stability of a structure can be contributed from distally located Holliday junctions that come close to the stress axis through the long-range, 3D arrangement. In fact, the density of Holliday junctions along the longitudinal pulling for the 3D tubular structure increases significantly as the Holliday junctions in the same layer becomes more compact after isomerization (see Supplementary Figure S15). Therefore, the effective density of the Holliday junctions along the stretching direction becomes a decisive factor for the mechanical stability of an origami structure along that direction. Here, we defined the density of Holliday junctions as the number of contributing Holliday junctions per nanometer of the force axis in the nanostructure. For nanotiles and nanopyramids, the effective Holliday junction densities are 0.17 and 0.18 HJ/nm, respectively. For the eight-tube and the six-tube DNA along the longitudinal pulling direction, due to the symmetric geometries, all the Holliday junctions contribute equally to the mechanical property of the origami. Therefore, they are all counted in the density calculation, which leads to much higher densities (1.43 HJ/nm for the eight-tube and 1.09 HJ/nm for the six-tube). Corresponding to this trend, the disassembly forces of the isomerizable structures along the longitudinal direction for both six- and eight-tube DNA origami are higher (35–60 pN) than those of the non-isomerizable structures such as nanotiles and nanopyramids (28–40 pN). This clearly shows that the mechanical stability of DNA origami structures may depend on the effective density of the Holliday junctions. Therefore, the observed data of mechanical stability suggest that Holliday junctions serve as mechanophores ( 26 , 38 , 39 ) to impart the mechanical stability to the entire DNA origami nanoassemblies. Holliday junctions in DNA origami tubes are responsible for mechanical isomerization between nanotube isomers Next, we investigated the mechanical property of the DNA origami at the lower force region (Figures 2A – C and 4 ), in which transition events were reproducibly observed during consecutive stretching of the DNA nanotubes (see Supplementary Figure S10). F – X curves obtained during the stretching of the 8-tube DNA showed characteristic changes-in-extension along the longitudinal (Figure 4C ) or horizontal direction (Figure 4D ). This confirms that the low-force transition events represent isomerization of the DNA tubes. To determine the nature of these rupture transitions, we measured the change-in-extension during each event of isomerization transition. The change-in-extension is described as the isomerization length ( L iso ) in Figures 4 and 5 . For the eight-tube DNA, we found two major transitions of ∼62 and ∼26 nm during the longitudinal stretching (Figure 4E ) and one transition of ∼11 nm during the horizontal stretching (Figure 4F ). These change-in-extension histograms show broad distributions, which is expected as the transition events are likely mediated by one or more intermediates presented in DNA origami structures with large sizes ( 40 , 41 ). It is also possible that some reversible disassembly or melting processes may contribute to this broad feature. For example, it has been observed that under mechanical stress, force induced melting can weaken Watson-Crick base pairing ( 32 ), which may lead to transitions with longer-than-expected sizes. To clearly determine the most likely change-in-extension values, we used Population Deconvolution at Nanometer resolution (PoDNano), a statistical tool based on bootstrapping analysis as described in literature ( 42 , 43 ). With a total of 5000 random resampling processes, we confirmed the presence of the ∼62 and ∼26 nm transitions in the change-in-extension histogram with the broadest distribution (Figure 4E inset). In AFM imaging, we have observed two different DNA tube isomers, long and short tubular structures (Supplementary Figure S8) ( 10 ). Compared to the expected change-in-extension between these two isomers (63 and 14 nm for longitudinal and horizontal stretching directions, respectively, see Supplementary Figure S16 for detailed calculation), the most probable values obtained from the histograms of the change-in-extension (Table 1 ) strongly suggest that the transitions are due to the force induced isomerization between the long and short DNA eight-tubes (Figure 4A and B ). Stretching of the six-tube DNA origami confirmed that the transitions in the low force region are indeed due to the mechanical isomerization between the long and short DNA tubes (80 and 9.2 nm in sum of all transition events vs expected values of 83 and 12 nm for the longitudinal and horizontal stretching directions, respectively, see Figure 5 , Supplementary Figures S16 and S17). It is noteworthy that multiple transitions were observed in every longitudinal stretching experiment. This suggests the presence of intermediates during isomerization. For example, a stable intermediate at ∼46 nm (Figure 5E ) can be attributed to a conformation trapped during incomplete isomerization (Supplementary Figure S18). Figure 4. Mechanical isomerization of the eight-tube DNA origami. Mechanical isomerizations between the long and short DNA tubes are explained from the perspective of Holliday junctions in the origami structure during longitudinal ( A ) or horizontal ( B ) stretching. For clarity, only one complete Holliday junction is shown in the middle of each structure. Dotted yellow lines depict Holliday junction layers in different stretching orientations. Representative F-X curves were obtained during longitudinal ( C ) or horizontal ( D ) stretching of the eight-tube DNA. Arrow heads in both F-X curves indicate the isomerization events (see insets for respective schematic diagrams). The blue/red and green traces depict stretching and relaxing curves, respectively. Due to the nature of the mechanical isomerization, only the transitions from the short tube to the long tube in the longitudinal pulling, and the long tube to the short tube in the horizontal pulling can be observed. Histograms of the isomerization length ( L iso ) of the eight-tube DNA are plotted for longitudinal ( E ) or horizontal ( F ) stretching (inset in E depicts the most probable populations using PoDNano approach, see text). Histograms of the isomerization force ( F iso ) of the same DNA tube measured during longitudinal ( G ) or horizontal ( H ) stretching. Solid curves represent Gaussian fittings. N and n respectively depict the number of F – X traces and number of molecules in experiments. Figure 5. Mechanical isomerization of the six-tube DNA origami. AFM images of the DNA nanoassembly with longitudinal ( A ) or horizontal ( B ) dsDNA handles. Representative F-X traces were obtained by longitudinal ( C ) or horizontal ( D ) stretching of the six-tube DNA. The arrowheads in both F-X curves indicate isomerization events (see insets for respective schematic diagrams). The blue/red and green traces depict stretching and relaxing curves, respectively. ( E ) Histogram of the isomerization length ( L iso ) measured from the longitudinal stretching shows individual transitions that involve different intermediates. Inset depicts the histogram of the sum of all transitions observed in individual F-X curves. While the 46.5 nm population suggests transitions involving intermediates, the 80 nm population is consistent with the full transition from the short to long DNA tubes (expected value: 83 nm). ( F ) Histogram of the isomerization length ( L iso ) of the six-tube DNA during horizontal stretching. Isomerization force ( F iso ) histograms of the six-tube DNA during longitudinal ( G ) or horizontal ( H ) stretching. Solid curves represent Gaussian fittings. N and n respectively depict the number of F – X traces and number of molecules in experiments. Interestingly, during horizontal stretching, the mechanical isomerization involves fewer intermediate states compared to the longitudinal stretching (compare between Figure 4E and F ; and between Figure 5E and F ). This reflects different origami arrangements along various stretching orientations. During longitudinal stretching, it travels longer (62 nm for the eight-tube and 80 nm for the six-tube) from the shorter DNA tube to the longer tube with respect to the horizontal stretching, in which the distance is much shorter between the two isomers (11 nm for the eight-tube and 9.2 nm for the six-tube). The extra transition distance in the former case (longitudinal) renders a more elastic energetic profile that allows more intermediates to populate. In the latter case (horizontal), however, the entire process becomes more cooperative due to the shorter and therefore more rigid isomerization pathways that cannot host as many intermediate populations. Such an observation is consistent with the disassembly of the origami structures at the high force region discussed above. While the DNA tube experiences more intermediates during the longitudinal, layer-by-layer disintegration, it has fewer intermediates during the horizontal disassembly with increased cooperativity (Figure 3D ). Similar observations have been previously obtained in the mechanical unfolding of short and long DNA duplexes. While the short DNA duplexes such as hairpins almost do not present intermediates, ( 44 ) the saw-teeth features observed in force–extension traces demonstrate the existence of many intermediates in long DNA duplexes ( 41 ). During the longitudinal isomerization of the eight-tube DNA origami, the isomerization force ( F iso = 26.4 pN, Figure 4G and Table 1 ) is comparable to that of the horizontal F iso (27.8 pN, Figure 4H and Table 1 ). Similar forces between the two stretching orientations have been observed in the isomerization of the six-tube DNA origami ( F iso = 27.3 pN for longitudinal and 29.7 pN for horizontal, see Figure 5G and H , respectively). We reasoned that the isomerization of the DNA nanotubes can be a result of collective isomerization of individual Holliday junctions at the microscopic level. Based on this assumption, we calculated the isomerization force for individual Holliday junctions as 0.12–0.14 pN (see SI). This value is in agreement with the isomerization force of a single Holliday junction (0.1–0.3 pN) predicted by Hohng et al . ( 24 ). It is noteworthy that mechanical isomerization was not observed in DNA nanotiles or nanopyramid structures. Close inspection on all three types of origami nanoassemblies shows that the latter two structures are rather different from the tubular structures. While the DNA nanotiles have a planar geometry, nanopyramids consisting of four triangular tiles fold into an object in which the perimeter of each layer gradually reduces from the base to the apex of the pyramid. In DNA nanotubes, the circular geometry with the same tube diameter gives rise to less hierarchical contributions between the distal and proximal regions with respect to the force axis, which bring a similar microenvironment to all Holliday junctions. However, due to the loss of symmetry in either DNA nanotiles or nanopyramids, individual Holliday junctions experience different environment and therefore, behave differently. As a result, the mechanical isomerization between two origami isomers can only be carried out by many identical Holliday junctions as a collective action in DNA nanotubes, but not in other structures."
} | 8,607 |
33683803 | PMC8085966 | pmc | 4,922 | {
"abstract": "Summary Marine photosynthetic microalgae are ubiquitously associated with bacteria in nature. However, the influence of these bacteria on algal cultures in bioreactors is still largely unknown. In this study, eighteen different bacterial strains were isolated from cultures of Nannochloropsis sp. CCAP211/78 in two outdoor pilot‐scale tubular photobioreactors. The majority of isolates was affiliated with the classes Alphaproteobacteria and Flavobacteriia . To assess the impact of the eighteen strains on the growth of Nannochloropsis sp. CCAP211/78, 24‐well plates coupled with custom‐made LED boxes were used to simultaneously compare replicate axenic microalgal cultures with addition of individual bacterial isolates. Co‐culturing of Nannochloropsis sp. CCAP211/78 with these strains demonstrated distinct responses, which shows that the technique we developed is an efficient method for screening the influence of harmful/beneficial bacteria. Two of the tested strains, namely a strain of Maritalea porphyrae (DMSP31) and a Labrenzia aggregata strain (YP26), significantly enhanced microalgal growth with a 14% and 12% increase of the chlorophyll concentration, respectively, whereas flavobacterial strain YP206 greatly inhibited the growth of the microalga with 28% reduction of the chlorophyll concentration. Our study suggests that algal production systems represent a ‘natural’ source to isolate and study microorganisms that can either benefit or harm algal cultures.",
"conclusion": "Conclusion In this study, we isolated 18 bacterial strains from two outdoor photobioreactors for cultivation of microalgae. A Maritalea porphyrae strain and a Labrenzia aggregata strain significantly promoted growth of Nannochloropsis sp. CCAP211/78 in liquid cultures in well plates (14% and 12% increase of the maximum chlorophyll concentration compared to the controls, respectively), and the Labrenzia aggregata strain also notably increased growth of the alga on agar plates. In addition, one strain most closely related to Aquaticitalea lipolytica significantly reduced the chlorophyll content with 28% compared to the axenic and non‐axenic controls. Our results suggest that some bacteria from algal production systems may have pronounced impacts on algal growth under controlled laboratory conditions, an effect that should be verified for larger‐scale algae cultures. Our results indicate that in the practice of improving the production of microalgae, the bacterial community in algal inocula should be considered. If harmful bacteria are present, the inoculum should be replaced by an inoculum where these bacteria are absent to increase the cultivation success. Perhaps even more interesting, beneficial bacterial strains may be supplemented as a new means to improve algal productivity and culture stability.",
"introduction": "Introduction Microalgae show great potential in producing numerous sustainable bioproducts as alternatives to fossil feedstocks (Ruiz et al ., 2016 ; Wijffels and Barbosa, 2010 ; Berthold et al ., 2019 ). A long‐neglected aspect in algal biomass production is the role of bacteria that are co‐occurring in algae cultivation systems (Cho et al ., 2014 ; Biondi et al ., 2018 ). Algal cultures are axenic in only a few applications, whereas all microalgae mass production systems inevitably contain a number of non‐target organisms (contaminants), including bacteria (Zittelli et al ., 2013 ; Newby et al ., 2016 ). Bacteria are introduced in algae cultivation systems as algae stocks used as starter cultures are often not axenic (Biondi et al ., 2017 ; Biondi et al ., 2018 ; Gouveia et al ., 2019 ). On the other hand, bacterial contaminants may enter cultivation systems through multiple operation processes, such as the supplementation of unsterilized medium or simply as airborne invaders in open algal cultures. Microalgae–bacteria interactions are prevalent in natural aquatic environments, where microalgae release exudates into the phycosphere, the region immediately surrounding individual cells. Chemotaxis drives multiple bacteria to the phycosphere (Smriga et al ., 2016 ), and metabolites are readily exchanged between algae and bacteria (Seymour et al ., 2017 ). Although the phycosphere represents only a tiny area that can be as small as 1 µm surrounding the algal cell, it represents the hotspot for most of the algal–bacterial interactions that can profoundly affect the productivity and stability of aquatic ecosystems (Amin et al ., 2012 ; Seymour et al ., 2017 ). Recent research on algal–bacterial interactions has usually been centred around the competitive or antagonistic aspects, which often involve competition for nutrients (Liu et al ., 2012 ; Wang et al ., 2016 ; Le Chevanton et al ., 2016 ) or algicidal activities (Paul & Pohnert, 2011 ; Seymour et al ., 2017 ). For instance, in a microcosm experiment it was found that bacteria were more efficient than algae in the uptake of phosphorus (Liu et al ., 2012 ). The advantage for bacteria is especially evident under phosphorus‐limiting conditions (Zubkov et al ., 2007 ). Apart from competing for nutrients with algae, some bacteria are known to inhibit algal cell division (van Tol et al ., 2017 ) or cause algal cell lysis via secretion of algicidal compounds (Seyedsayamdost et al ., 2011 ; Wang et al ., 2012 ; Zheng et al ., 2013 ). In contrast to early views that bacteria mostly affect microalgae negatively, it has been demonstrated that mutualistic relationships between microalgae and bacteria are also prevalent, or even more common than antagonistic interactions (Seymour et al ., 2017 ; Lian et al ., 2018 ). Proof has been found from frequent observations that the absence of bacteria in algal cultures negatively affects algal physiology and growth (Bolch et al ., 2011 ; Windler et al ., 2014 ). In exchange for dissolved organic matter from microalgae, bacteria fix nitrogen (Foster et al ., 2011 ; Thompson et al ., 2012 ) and synthesize a wide range of molecules, including vitamins (Xie et al ., 2013 ; Grant et al ., 2014 ), the growth‐promoting hormone indole‐3‐acetic acid (Amin et al ., 2015 ; Dao et al ., 2018 ) and the siderophore vibrioferrin (Amin et al ., 2007 ; Lupette et al ., 2016 ). The division of labour and close cooperation enables the holobiont to better adapt to and grow in changing aquatic environments, which has also triggered a growing interest for applications in industrial settings (Hom et al ., 2015 ; Lutzu and Turgut Dunford, 2018 ; Yao et al ., 2019 ). Contrary to extensive tests of effects of environmental and chemical factors (irradiation, temperature, pH, nutrients, etc.) on algal growth in industrial photobioreactors, only a few studies have considered the effects of biotic factors such as associated bacteria. In order to assess the effects of co‐occurring bacteria on microalgae in algal cultivation systems, we isolated and characterized bacteria from two pilot‐scale outdoor tubular photobioreactors. Subsequently, a 24‐well plate‐based co‐cultivation device was used to evaluate algal growth with addition of the isolated bacterial strains to axenic microalgae. Effects of bacteria on microalgae were further tested on a double‐layer agar plate to verify algal–bacterial interactions.",
"discussion": "Results and discussion Bacterial isolation and identification In order to recover as many different bacteria as possible from outdoor bioreactors, eight carbon sources were used for bacterial isolation. In total, we picked and sequenced 138 bacterial isolates from four samples from two outdoor photobioreactors with Nannochloropsis sp. CCAP211/78. All isolated bacteria were classified as Proteobacteria or Bacteroidetes and encompassed sixteen genera (Table S2 ). Two bacteria, closely related to Celeribacter sp. and Maritalea porphyrae , were the most frequently isolated and were recovered from all media (Table S1 ). Six bacteria were recovered from multiple media, while ten bacterial strains were recovered from only one medium. From medium YP (yeast and peptone extract), more bacterial species (11 out of 18) were recovered than from any of the other carbon sources also because many more colonies were obtained and picked (43 out of 138) from agar plates with YP. We then chose 18 representative bacterial isolates for co‐cultivation experiments. Of the 18 isolates, 11 belong to the class Alphaproteobacteria and five to Flavobacteriia . In addition, single isolates were obtained from the classes Cytophagia and Saprospiria (Table 1 ; Fig. S4 ). At the family level, isolates were mainly classified into three families: Hyphomicrobiaceae , Rhodobacteraceae and Flavobacteriaceae . It has also been corroborated by global surveys that phytoplankton‐associated bacterial communities are often restricted to only a few bacterial classes including Alphaproteobacteria ( Rhodobacteraceae ), Gammaproteobacteria ( Alteromonadaceae ) and Flavobacteriia ( Flavobacteraceae ) (Amin et al ., 2012 ; Teeling et al ., 2012 ; Goecke et al ., 2013 ; van Tol et al ., 2017 ). Within Alphaproteobacteria , bacteria from the family Rhodobacteraceae are frequently associated with algae, of which the most studied ones are Phaeobacter gallaeciensis (Seyedsayamdost et al ., 2011 ), Dinoroseobacter shibae (Wang et al ., 2015 ), Sulfitobacter sp. (Amin et al ., 2015 ) and Ruegeria pomeroyi (Durham et al ., 2015 ). These apparently widespread patterns imply that the lifestyle of some bacteria within these groups is substantially related to that of algae. Table 1 Bacterial strains isolated from Nannochloropsis cultures. Strain Accession Number (bacterial isolate) Class Family Blast result \n a \n \n Identity [%] Accession number (Genbank best hit) OTUs in bioreactors \n b \n \n Identity [%] GLU107 MH843917 \n Alphaproteobacteria \n \n Erythrobacteraceae \n \n Porphyrobacter sanguineus \n 100 LC349792 OTU247 100 PRO103 MH843918 \n Alphaproteobacteria \n \n Hyphomicrobiaceae \n \n Algimonas arctica \n 98 NR_137369 OTU321 100 DMSP31 MH843919 \n Alphaproteobacteria \n \n Hyphomicrobiaceae \n \n Maritalea porphyrae \n 99 AB583776 OTU327 100 DMSP20 MH843920 \n Alphaproteobacteria \n \n Hyphomicrobiaceae \n \n Maritalea sp. 99 AB758563 OTU331 100 PRO34 MH843921 \n Alphaproteobacteria \n \n Hyphomicrobiaceae \n \n Maritalea sp. 96 KP301112 OTU343 100 YP210 MH843922 \n Alphaproteobacteria \n \n Phyllobacteriaceae \n \n Pseudohoeflea suaedae \n 100 LT600545 OTU490 100 YP18 MH843923 \n Alphaproteobacteria \n \n Rhodobacteraceae \n \n Celeribacter sp. 100 MF045112 OTU582 100 YP26 MH843924 \n Alphaproteobacteria \n \n Rhodobacteraceae \n \n Labrenzia aggregata \n 100 MG273739 OTU247 100 YP29 MH843925 \n Alphaproteobacteria \n \n Rhodobacteraceae \n \n Roseovarius mucosus \n 99 CP020474 OTU585/709 100 YP202 MH843926 \n Alphaproteobacteria \n \n Rhodobacteraceae \n \n Sulfitobacter sp. 99 KY272045 OTU143/289 100 PAL103 MH843927 \n Alphaproteobacteria \n \n Sphingomonadaceae \n \n Sphingorhabdus sp. 99 KT325114 OTU259 98 DMSP2‐Y MH843928 \n Cytophagia \n \n Cytophagaceae \n \n Emticicia sp. 99 KP265953 OTU574 100 YP206 MH843929 \n Flavobacteriia \n \n Flavobacteriaceae \n \n Aquaticitalea lipolytica \n 99 NR_149769 OTU532/533 94 ALG110 MH843930 \n Flavobacteriia \n \n Flavobacteriaceae \n \n Arenibacter sp. 98 JX529985 OTU582 100 PAL10 MH843931 \n Flavobacteriia \n \n Flavobacteriaceae \n \n Cellulophaga lytica \n 100 MG456766 OTU519 96 PAL110 MH843932 \n Flavobacteriia \n \n Flavobacteriaceae \n \n Maribacter sp. 99 KT731371 OTU525 96 SUC105 MH843933 \n Flavobacteriia \n \n Flavobacteriaceae \n \n Muricauda sp. 99 KJ188010 OTU512 100 PRO13 MH843934 \n Saprospiria \n \n Saprospiraceae \n \n Phaeodactylibacter xiamenensis \n 99 NR_134132 OTU579 100 a The best hit (highest per cent identity) in NCBI Genbank. b The best hit of photobioreactor OTUs. John Wiley & Sons, Ltd When Sanger‐sequenced 16S ribosomal RNA (rRNA) genes of the bacterial strains were compared to the 138 operational taxonomic units (OTUs) present in the four original bioreactor cultures, fourteen out of 18 bacterial strains had an identical match with OTUs encountered in the reactors, while four isolates had not (Table 1 ). The cultivable bacteria isolated in this study accounted for approximately 10% of the total OTUs (14 of 138) present in the original photobioreactor samples and represented nearly 7% of the total reads (11 820 of 152 260) in the bioreactor samples. Thus, a substantial fraction of bacteria in algal cultures remained uncultured under the conditions applied in our experiment. We observed sixteen OTUs with high relative abundance (≥ 5%) in our algal cultures (Table S2 ), of which four (OTU533, 579, 327, 331) were successfully cultured. It is noticeable that although Gammaproteobacteria was one of the most abundant classes in two of four bioreactor cultures based on cultivation‐independent assessment of bacterial diversity, no strains belonging to this class were recovered (Table S2 and Fig. S2 ). Effect of bacteria on the growth of algae To examine potential interactions between Nannochloropsis and the bacterial isolates, the bacterial isolates were re‐introduced to axenic microalgae. All the cultures except the ones supplemented with strain YP206 had a similar growth pattern, that is, after rapid growth for nearly 5 days, the stationary phase was reached, which continued until the end of the experiment at day 11 (Fig. S3 ). No significant difference was found in relative fluorescence between axenic and non‐axenic control cultures of Nannochloropsis . Addition of bacteria to the axenic Nannochloropsis sp. cultures had no significant impact on algal growth rates in the first six days (Fig. S3 ), except for YP206 where the growth rate was significantly lower (Fig. 1B ), but mostly resulted in a slight decrease of the maximal fluorescent intensity reached at the stationary phase (Fig. 1A and Fig. S3 ). Fig. 1 Relative Fluorescence (~ algal biomass) of Nannochloropsis sp. CCAP211/78 co‐cultured with individual bacterial strains. A. Relative Fluorescence Unit (RFU) for Nannochloropsis sp. CCAP211/78 was calculated as maximal fluorescent intensity and compared to RFU of the axenic culture. Error bars represent standard deviation. Results of the statistical analysis are indicated by NS ( P .adjust > 0.05), * ( P .adjust ≤ 0.05), and ** ( P .adjust ≤ 0.01), respectively. The statistical results of pair‐wise comparison against non‐axenic culture (not shown) are the same as for the comparison to the axenic culture. B. Growth curves of Nannochloropsis sp. with bacteria that significantly affected the growth (DMSP31, YP26 and YP206) and the axenic and non‐axenic controls. For strain YP206 ( Flavobacteriia ), Nannochloropsis growth was strongly inhibited, leading to a reduction by more than 28% in fluorescent intensity. Flavobacteriia have repeatedly been reported to have antagonistic relationships with algae. For instance, Kordia algicida was shown to excrete an extracellular protease to lyse algal cells to acquire their dissolved organic carbon (Paul and Pohnert, 2011 ), and Croceibacter atlanticus was observed to release an unidentified molecule to arrest diatom cell division and increase secretion of organic carbon (van Tol et al ., 2017 ). The closest relative of strain YP206 is Aquaticitalea lipolytica (99% identity of the 16S rRNA gene) that was isolated from Antarctic seawater and known to hydrolyse lipids (Xamxidin et al ., 2016 ). However, when YP206 was co‐cultured on agar plates with Nannochloropsis , the growth inhibition observed in liquid culture was not observed (Figs 1 and 2 ). Although mechanistic insight requires future research, one can speculate that the incubation time (7 days) used in the agar plate experiments described here was too short or that the algal density was still too low on the agar plate for the bacterial inhibition to take place, as some algicidal bacteria have been shown to only kill senesced algal cells in the stationary phase or decline phase (Seyedsayamdost et al ., 2011 ; Wang et al ., 2015 ). This has previously been explained by competition for limiting nutrients such as nitrogen (Meseck et al ., 2006 ) and phosphorus (Danger et al ., 2007 ; Liu et al ., 2012 ). However, that is not likely to be the case for our results as nitrogen and phosphorus concentrations added would support much higher algae concentrations than those present in the stationary phase, and for nitrogen, it was confirmed in the stationary phase that it was not depleted (data not shown). Alternatively, release of toxic compounds by bacteria could contribute to the inhibitory effects observed at stationary phase (Fukami et al ., 1997 ; Mitsutani et al ., 2001 ). Many bacteria belonging to the family Flavobacteriaceae are able to glide on solid surfaces and decompose agar (Nedashkovskaya et al ., 2004 ). PAL10 and PAL110 displayed these features and formed larger and concave colonies on the agar surface (Fig. 2 ). Although both strains showed no significant effects on algal growth in liquid co‐culture, they apparently enhanced the growth of Nannochloropsis sp. in the agar plate assay (Fig. 2 ). A possible explanation for the growth promotion on solid media could be that Nannochloropsis cells consumed the by‐products from the agar degradation by the bacteria. For instance, Cellulophaga lytica (PAL10) has previously been shown to synthesize different kinds of agarases (Lee and Choi, 2017 ), and the enzymatic hydrolysis of agar yields monomeric sugars, such as d ‐galactose, 3,6‐anhydro‐ l ‐galactose and l ‐galactose‐6‐sulphate (Chi et al ., 2012 ). Research has shown that supplementation with galactose increases the growth rate of Nannochloropsis salina by nearly 10% (Velu et al ., 2015 ). Fig. 2 Co‐cultivation of Nannochloropsis sp. CCAP211/78 and bacterial strains on double‐layer agar plates after seven days. ESW (Enriched natural seawater medium) and ESW‐YP (ESW medium with peptone and yeast extract) were used as controls. The labelled names referred to the added bacteria. YP26 was added on three different plates as replicates indicated by arrows. In contrast to the inhibition of algal growth by YP206, two bacterial strains (YP26, DMSP31) resulted in significantly better growth of the algae, and the intensity of algal fluorescence increased by 12–14% compared to the axenic control (Fig. 1 ). Members of the genus Labrenzia (YP26) have been isolated from a wide range of habitats and found to be frequently associated with other marine organisms (Weber and King, 2007 ; Coates and Wyman, 2017 ). These organisms include invertebrates such as molluscs, corals and sponges, and a wide variety of photosynthetic partners including seaweeds, diatoms, dinoflagellates, green and red algae (Boettcher et al ., 2000 ; Groben et al ., 2000 ; Sfanos et al ., 2005 ; Weber and King, 2007 ). Labrenzia aggregata has also been isolated previously from Nannochloropsis oculata and Nannochloropsis gaditana (Han et al ., 2016 ). A recent study revealed that Labrenzia sp. increased the biomass accumulation of the marine microalga Isochrysis galbana by 72% and the growth rate by 18% (Sandhya and Vijayan, 2019 ). On the other hand, it has been reported that a bacterial isolate (KD531) with 100% similarity to the partial 16S rRNA gene of our Labrenzia aggregata isolate had an algicidal effect on Chlorella vulgaris (Chen et al ., 2014 ). The addition of bacterial lysate of KD531 to Chlorella vulgaris cultures caused nearly 20% reduction in biomass dry weight and nearly 60% reduction in lipid content. The contradiction between these and our observations may be due to strain‐specific differences between isolates of L. aggregata , and/or different interactions of the bacterium with different algal hosts. Prior research has shown that some bacteria that are mutualistic to their native algal partner can be parasitic to foreign algae, which hints at co‐adaption and evolution of algae and their associated microbiome (Sison‐Mangus et al ., 2014 ). In addition, we added live bacteria rather than a bacterial lysate, which may lead to a different effect. Nannochloropsis sp. also appeared to grow faster and denser on a solid growth medium in the presence of Labrenzia aggregata (Fig. 2 ). Agar plates have been the most commonly used method to study algal–bacterial interactions (Kazamia et al ., 2012 ; Hertweck et al ., 2017 ). For example, the vitamin B 12 ‐dependent microalga Lobomonas rostrata could grow on agar plates only when vitamin B 12 or a vitamin B 12 ‐synthesizing bacterium ( Mesorhizobium loti ) was added (Kazamia et al ., 2012 ). Therefore, it is tempting to speculate that the growth promotion observed here for Labrenzia aggregata could be related to inorganic nutrient exchange or algal acquisition of growth factors released by bacteria. Although the growth increase of Nannochloropsis in the presence of Maritalea porphyrae (DMSP31) was significant in liquid cultures, this beneficial effect disappeared on the agar plate. It is interesting to note that Maritalea porphyrae (DMSP31) has been previously isolated from the thalli of the red alga Pyropia yezoensis (Fukui et al ., 2012 ). However, experimental evidence showed that these bacteria exhibited no apparent morphogenetic effects on the red alga (Fukui et al ., 2014 ), and therefore, the nature of a symbiotic relationship – if any – with the phototroph remains unknown. Some other bacterial isolates for which we did not find a significant effect have been previously associated to microalgae. For example, the family Saprospiraceae (strain PRO13) was the most prevalent taxon and also the most abundant one in industrial cultures of Nannochloropsis salina (Fulbright et al ., 2018 ). In addition, the 16S rRNA gene of strain PRO13 was identical to OTU579 found in the outdoor photobioreactors, particularly in sample HD0105 where this bacterium made up nearly 25% of the whole bacterial community (Table S2 ). In spite of this strikingly high relative abundance, co‐culturing with strain PRO13 had no significant effect on the growth of Nannochloropsis sp. either in liquid co‐cultures or on agar plates (Figs 1 and 2 ). Similarly, the study by Fulbright et al . ( 2018 ) reported that there was no correlation between the abundance of Saprospiraceae and growth of N. salina . However, the prevalence of this bacterium suggests it may have other functions in algal cultures, and the lytic capability of members of this bacterial family may relate to degrading cell debris for nutrient recycling (Fulbright et al ., 2018 ). DMSP2‐Y is closely related to Emticicia sp., and species from the genus Emticicia have been recorded to live with Chlorella vulgaris (Otsuka et al ., 2008 ) and the macroalga Cladophora glomerata (Zulkifly et al ., 2012 ). Emticicia sp. was found to slightly reduce the growth rate of axenic Chlorella vulgaris in co‐cultivation, but the co‐culture revealed prolonged stationary phase (Vu et al ., 2010 ). For a number of strains (e.g. YP206, PAL10, PAL110, DMSP31), the observed effect of addition of the strain to liquid cultures of Nannochloropsis sp. CCAP211/78 (Fig. 1 ) was not in line with the trend observed for the same combination on solid agar (Fig. 2 ). This discrepancy between two screening methods corroborates that algae–bacteria interactions are complex and may vary under different culture conditions. Therefore, preliminary screening results should be confirmed by other methods such as flask cultures or bioreactors before claims regarding beneficial effects of bacteria on large‐scale algal growth can be made."
} | 5,959 |
32393858 | null | s2 | 4,923 | {
"abstract": "Plant pathogenic bacteria cause high crop and economic losses to human societies"
} | 20 |
31281294 | PMC6595629 | pmc | 4,927 | {
"abstract": "The bio-economy relies on microbial strains optimized for efficient large scale production of chemicals and fuels from inexpensive and renewable feedstocks under industrial conditions. The reduced one carbon compound methanol, whose production does not involve carbohydrates needed for the feed and food sector, can be used as sole carbon and energy source by methylotrophic bacteria like Methylobacterium extorquens AM1. This strain has already been engineered to produce various commodity and high value chemicals from methanol. The toxic effect of methanol limits its concentration as feedstock to 1% v/v. We obtained M. extorquens chassis strains tolerant to high methanol via adaptive directed evolution using the GM3 technology of automated continuous culture. Turbidostat and conditional medium swap regimes were employed for the parallel evolution of the recently characterized strain TK 0001 and the reference strain AM1 and enabled the isolation of derivatives of both strains capable of stable growth with 10% methanol. The isolates produced more biomass at 1% methanol than the ancestor strains. Genome sequencing identified the gene metY coding for an O -acetyl- L -homoserine sulfhydrylase as common target of mutation. We showed that the wildtype enzyme uses methanol as substrate at elevated concentrations. This side reaction produces methoxine, a toxic homolog of methionine incorporated in polypeptides during translation. All mutated metY alleles isolated from the evolved populations coded for inactive enzymes, designating O -acetyl- L -homoserine sulfhydrylase as a major vector of methanol toxicity. A whole cell transcriptomic analysis revealed that genes coding for chaperones and proteases were upregulated in the evolved cells as compared with the wildtype, suggesting that the cells had to cope with aberrant proteins formed during the adaptation to increasing methanol exposure. In addition, the expression of ribosomal proteins and enzymes related to energy production from methanol like formate dehydrogenases and ATP synthases was boosted in the evolved cells upon a short-term methanol stress. D-lactate production from methanol by adapted cells overexpressing the native D-lactate dehydrogenase was quantified. A significant higher lactate yield was obtained compared with control cells, indicating an enhanced capacity of the cells resistant to high methanol to assimilate this one carbon feedstock more efficiently.",
"introduction": "Introduction Methanol is a highly available industrial compound that can be produced from the greenhouse gas carbon dioxide by chemical or electrolytic reduction processes ( Ganesh, 2014 ). A growing number of methanol based biotechnological production pathways are developed as alternatives to industrial fermentations relying on sugar as feedstock ( Schrader et al., 2009 ; Bennett et al., 2018 ). Among methylotrophic organisms capable of growing on methanol as the sole carbon and energy source, the facultative methylotrophic alpha-proteobacterium Methylobacterium extorquens AM1 is widely used as catalyst and has been engineered for the production of a variety of value-added chemicals or biofuels ( Hu and Lidstrom, 2014 ; Hu et al., 2016 ; Zhu et al., 2016 ). The central carbon and energy metabolism of M. extorquens AM1, which belongs to the serine-cycle methylotrophs, has been studied in detail ( Chistoserdova et al., 2003 ; Ochsner et al., 2015 ). Genomic sequences ( Vuilleumier et al., 2009 ; Marx et al., 2012 ) and metabolic models ( Marx et al., 2005 ; Peyraud et al., 2011 ) are also available for this and other members of the Methylobacterium genus. With the objective to construct platform strains for the optimal use in biotechnological applications and to answer fundamental questions of evolution, attempts have been made to optimize methanol assimilation through rewiring central steps of formaldehyde conversion ( Carroll et al., 2015 ) or to broaden the set of one carbon (C1) compounds used as substrates ( Michener et al., 2014 ). A limitation of industrial scale methanol fermentation is the low tolerance of most of the methylotrophic organisms toward methanol. The toxicity of methanol is, at least partially, attributable to its properties as organic solvent ( Busby et al., 1999 ; Hwang et al., 2011 ). M. extorquens AM1 has a growth maximum with 1% (v/v) methanol. Low solvent tolerance is a common hurdle of biotechnology, notably for processes of biofuel production. Since solvent stressors usually have pleiotropic effects on cell integrity and proliferation ( Stephanopoulos et al., 2004 ), directed evolution rather than rational design approaches are followed to select for enhanced tolerance. Numerous reports relay the adaptation of the natural producer strain Saccharomyces cerevisiae to higher ethanol tolerance ( Ma and Liu, 2010 ). Likewise, mutants of the reference strain Escherichia coli were obtained exhibiting resistance to ethanol ( Yomano et al., 1998 ), n-butanol ( Rutherford et al., 2010 ; Reyes et al., 2011 , 2013 ), and isobutanol ( Atsumi et al., 2010 ). Phenotypic analysis of resistant cells, including genomic, transcriptomic, and metabolomics studies, often revealed complex patterns of adaptations, which depended on the nature of the solvent. Changes included the lipid content of cell membranes, the enhanced synthesis of peptidoglycan precursors, and overexpression of chaperones or oxidative stress responses. In some cases, stress resistance could be attributed to point mutations in particular genes. This was the case for M. extorquens AM1 mutants growing with increased n-butanol concentrations, where a potassium-proton antiporter was found to partially account for the adaptive phenotype ( Hu et al., 2016 ). In addition to direct effects of methanol as alcoholic solvent, the molecule can exhibit indirect toxicity due to its metabolic conversion to formaldehyde and formate ( Liesivuori and Savolainen, 1991 ; Oyama et al., 2002 ; Chen et al., 2016 ). Resistance of Bacillus methanolicus toward high methanol was found to depend on the overexpression of the genes of the ribulose monophosphate cycle converting formaldehyde into biomass ( Jakobsen et al., 2006 ). Besides toxicity via oxidation, for some strains experimental evidence pointed to the conversion of methanol at high concentrations to the toxic methionine analog methoxine ( Leßmeier and Wendisch, 2015 ; Schotte et al., 2016 ). In this work, parallel experiments of directed evolution were performed in GM3 continuous culture automatons ( Marlière et al., 2011 ) to adapt the recently characterized M. extorquens strain TK 0001 ( Belkhelfa et al., 2018 ) and strain AM1 to growth on methanol concentrations of up to 10% (v/v). Genome sequencing of isolates of both lineages were conducted to identify common mutations potentially relevant for the methanol tolerant phenotype. The metY gene coding for O -acetyl- L -homoserine sulfhydrylase was found to harbor missense mutations in all isolates sequenced. A causal genotype-phenotype relation was experimentally demonstrated for this mutated locus. Furthermore, a transcriptomic analysis revealed differences in expression patterns of wildtype and methanol tolerant cells provoked by methanol stress. Finally, quantifying D-lactate production by an engineered methanol resistant TK 0001 mutant further demonstrated the usefulness of directed evolution for the selection of industrial production chassis.",
"discussion": "Discussion In the present study, an evolutionary approach was followed to obtain high methanol tolerant derivatives of two closely related methylotrophic strains, M. extorquens TK 0001 and AM1. For these strains, methanol serves as the sole carbon and energy source, but inhibits growth at concentrations above 1% (v/v). This limitation challenges the suitability of these organisms as production strains in industrial methanol fermentations. The GM3 technology of continuous culture was used to perform the stepwise adaptation of cultures of the two strains to grow with up to 10% methanol. Medium swap culture regimes were employed to select for higher methanol tolerance. This regime enables an incremental increase of the concentration of a stressing compound in the culture. The increment depends on the enhanced growth capacities of the cells, thus maintaining a constant selection pressure. An analogous functioning is the basis of the morbidostat, a continuous culture device used to study bacterial drug resistance ( Toprak et al., 2013 ). The medium swap regime which resembles a chemostat modified to enable the dilution of the culture by two growth media, has been found to permit lineages with different genotypes and consequently differing in fitness to prevail in the population ( Marlière et al., 2011 ). For this reason, following a swap adaptation, the GM3 were run under turbidostat mode. During this culture phase, through selection of fastest growing cells, the populations are genetically homogenized and generally establish a stable generation time. Above 7% methanol (1.7 M), the selection of very high tolerance was performed through incrementing the methanol concentration by 1% until 10% in turbidostat. This adaptation protocol, applied to both strains, was chosen to avoid prolonged swap periods and to expose the cells to high methanol at a stable concentration. A different inoculation strategy was pursued for the two cultures. In an attempt to accelerate the adaptation, a best growing isolate was selected at 5, 7, and 8% methanol resistance to relaunch the TK 0001 culture. By contrast, the AM1 evolution was performed from a single inoculum. However, both cultures progressed in a similar manner toward high methanol tolerance. Differences were seen for the medium swap adaptation periods (1–5%; 5–7%), which were longer for AM1, possibly reflecting the larger and more complex genome of this strain compared to TK 0001 ( Belkhelfa et al., 2018 ). Only three mutations were identified for the six TK 0001 isolates obtained after the 1–5% medium swap and found to be shared. No additional mutations common to all isolates analyzed were found for the adaptations to 7% in medium swap regime and the subsequent turbidostat adaptation to 8% methanol. Toward the end of the adaptation, however, the number of mutations increased cumulating to eight common mutations for the final six isolates. Higher mutation variability was found for the AM1 isolates obtained at the four resistance levels, reflecting the fact that the adaptation was performed from a single inoculum. However, the number of common mutations was limited to five for the final six isolates. Given the strong gain in methanol resistance, the adapted isolates carried surprisingly few mutations. Possibly, the tolerance to high methanol does not require a large number of adaptive mutations, due to its relatively low potential to perturb protein structure, as compared with larger alcohols like ethanol or butanol ( Perham et al., 2006 ). In addition, non-beneficial mutations did not accumulate due to the turbidostat selection periods where only the fittest cells stay in the culture constantly kept at exponential growth. This setup prevents mutation events to occur which are known to be specific for the stationary phase ( Zambrano et al., 1993 ). The metY gene specifying O -acetyl- L -homoserine sulfhydrylase carried a non-synonymous mutation in all isolates sequenced. The accumulation of loss of function mutations not affecting cell viability at this locus raises the question of the utility of this gene for the M. extorquens strains. As deduced from annotated genes, methionine biosynthetic routes involving trans-sulfurylation reactions are functional in both M. extorquens TK 0001 and AM1, what may explain the non-essential character of the MetY activity ( Supplementary Figure S5 ). Possibly, a functional O -acetyl- L -homoserine sulfhydrylase widens the methionine precursor reservoir to compounds like methanethiol and dimethyldisulfide which can be used by the enzyme to directly add the terminal S-CH 3 group to O -acetyl- L -homoserine to form methionine, as has been shown for MetY from C. glutamicum ( Bolten et al., 2010 ). Methanol at concentrations toxic for the cells through the MetY side reaction is not likely to be found in the cell’s natural habitat, allowing the retention of the wildtype form of the enzyme. In isolates of both evolved lineages, missense mutations were found in gene rpsL coding for ribosomal protein S12 (R94C in TK 0001 and Y95C in AM1). This component of the 30S subunit is known to play a crucial role in translation accuracy. A large number of E. coli S12 variants containing altered residues have been constructed and the decoding phenotypes tested ( Agarwal et al., 2011 ). The mutation of residues R94 and Y95 caused a restrictive phenotype increasing the accuracy of translation. Most probably, growth in high methanol media caused increasing translational misreading leading to the selection of these mutations to lower the error rate. Solvent stressors have been described to diminish the fidelity of the ribosomal polypeptide synthesis in vitro ( Phoenix et al., 1983 ). Furthermore, genes of ribosomal proteins carried mutations in E. coli cells adapted to high ethanol ( Haft et al., 2014 ). It is also noteworthy that no mutations occurred in the genes coding for proteins of formaldehyde detoxification. This is in accordance with growth phenotypes of the methanol tolerant strain G4105 on formaldehyde, which was not significantly altered with respect to the wildtype strain ( Supplementary Figure S4 ). Besides chromosomal mutations, evolution of microorganisms to resist solvent stressors commonly provokes alterations of gene expression suggesting adaptive mechanisms at the level of transcription ( Reyes et al., 2013 ). The continuous culture adaptation caused broad effects on the transcription of genes involved in multiple cellular functionalities, revealed by pairwise comparisons between the wildtype and the evolved strain. Chaperones and proteases implicated in the maintenance of protein homeostasis were upregulated in the mutant, as was the case for mRNAs of ribosomal proteins. The high abundance of these genes in the adapted cells was apparent when the transcriptomes of cells grown at permissive conditions were compared, probably reflecting the better growth and higher biomass production of the mutant strain at 1% methanol. After a 5 min growth period with 5% methanol, inter-strain comparison revealed an additional increase in the number of higher expressed genes coding for ribosomal proteins in the mutant. Obviously, the adaptation to high methanol enabled the cells to form more ribosomes upon an increased supply of methanol, their sole carbon and energy source. Concomitantly, formate dehydrogenases were also overexpressed, likely to provide the energy necessary for the enhanced biomass production. No significant overexpression [log 2 (FC) greater than 2 or less than −2] was noted for enzymes catalyzing the assimilation of methanol. It was observed that the corresponding genes, like mxaF coding for methanol dehydrogenase or fae coding for formaldehyde activating enzyme, were among the most highly expressed in both strains for all growth conditions tested, suggesting that significant upregulation of these transcripts was not feasible for the cells. However, nearly all enzymes of the central carbon route were found to be moderately overexpressed in the adapted cells [log 2 (FC) between 0.9 and 2, Supplementary Table S4 ]. Therefore, enhanced carbon flux and energy production through the central metabolism is a likely adaptation consequence. Evolved cells overexpressing D-lactate dehydrogenase from a plasmid grown in 1% methanol did produce more D-lactate than wildtype control cells, showing that the methanol tolerant cells can provide a genetic context for industrial production of chemicals through implemented synthetic pathways."
} | 4,016 |
38944415 | PMC11258901 | pmc | 4,928 | {
"abstract": "Abstract \n Corynebacterium glutamicum ATCC 13032 is a promising microbial chassis for industrial production of valuable compounds, including aromatic amino acids derived from the shikimate pathway. In this work, we developed two whole-cell, transcription factor based fluorescent biosensors to track cis,cis -muconic acid (ccMA) and chorismate in C. glutamicum . Chorismate is a key intermediate in the shikimate pathway from which value-added chemicals can be produced, and a shunt from the shikimate pathway can divert carbon to ccMA, a high value chemical. We transferred a ccMA-inducible transcription factor, CatM, from Acinetobacter baylyi ADP1 into C. glutamicum and screened a promoter library to isolate variants with high sensitivity and dynamic range to ccMA by providing benzoate, which is converted to ccMA intracellularly. The biosensor also detected exogenously supplied ccMA, suggesting the presence of a putative ccMA transporter in C. glutamicum , though the external ccMA concentration threshold to elicit a response was 100-fold higher than the concentration of benzoate required to do so through intracellular ccMA production. We then developed a chorismate biosensor, in which a chorismate inducible promoter regulated by natively expressed QsuR was optimized to exhibit a dose-dependent response to exogenously supplemented quinate (a chorismate precursor). A chorismate–pyruvate lyase encoding gene, ubiC , was introduced into C. glutamicum to lower the intracellular chorismate pool, which resulted in loss of dose dependence to quinate. Further, a knockout strain that blocked the conversion of quinate to chorismate also resulted in absence of dose dependence to quinate, validating that the chorismate biosensor is specific to intracellular chorismate pool. The ccMA and chorismate biosensors were dually inserted into C. glutamicum to simultaneously detect intracellularly produced chorismate and ccMA. Biosensors, such as those developed in this study, can be applied in C. glutamicum for multiplex sensing to expedite pathway design and optimization through metabolic engineering in this promising chassis organism. One-Sentence Summary High-throughput screening of promoter libraries in Corynebacterium glutamicum to establish transcription factor based biosensors for key metabolic intermediates in shikimate and β-ketoadipate pathways.",
"introduction": "Introduction Nature has a vast variety of microorganisms with unique phenotypes fully adapted to survive in extreme conditions and exhibiting unique and versatile metabolisms that are attractive as potential industrial platforms for bioconversion (Calero & Nikel, 2019 ). Despite their versatility, these microorganisms often require development and implementation of genetic engineering tools for their “domestication” to enable their productive use as industrial strains (Riley & Guss, 2021 ). One such organism is Corynebacterium glutamicum ATCC 13032 (hereafter referred to as C. glutamicum ), a non-pathogenic, Gram-positive soil bacterium that is a promising candidate for production of industrially relevant chemicals, including biofuels, cosmetics, pharmaceuticals, and a wide range of acids that include amino acids and various other organic acids (Becker, Rohles et al., 2018 ; Inui & Toyoda, 2020 ; Lin et al., 2022 ). To realize the full potential of this organism as an industrial chassis for the production of biochemicals, advancement in metabolic engineering tools such as genome editing (Wang et al., 2021 ) and high-throughput screening (described below) are needed. Biosensors have gained substantial interest in metabolic engineering (Kaczmarek & Prather, 2021 ; Mahr & Frunzke, 2016 ) for screening libraries of microbial strains, dynamic regulation of metabolic pathways, single-cell analysis, and screening of microbial populations generated from adaptive laboratory evolution (Bentley et al., 2020 ; Seok et al., 2021 ; Zhang et al., 2012 ). Whole-cell biosensors usually contain a sensor–reporter gene circuit, which responds to the presence of the molecule of interest by generating a survival, bioluminescent, colorimetric, or fluorescent response. Existing biosensors for C. glutamicum are mainly comprised of native transcription factors (TFs) that detect and respond to small molecule targets and have been applied for high-throughput screening of various libraries generated rationally or through adaptive laboratory evolution (Han et al., 2020 ; Mahr et al., 2015 ; Schendzielorz et al., 2014 ; Schulte et al., 2017 ; Siedler et al., 2017 ; Sonntag et al., 2020 ; Steffen et al., 2016 ). Recently, a functional shikimate biosensor was constructed from a LysR-type transcriptional regulator ShiR, a native TF from C. glutamicum that exhibited a detection range of 1–100 mM and was used to monitor shikimate production and to optimize ribosome binding sites to increase the carbon flux in the shikimate pathway (Liu et al., 2018 ). In this work, we focused on the development and implementation of two TF-based fluorescent biosensors in C. glutamicum targeting metabolic intermediates in the shikimate and β-ketoadipate pathways (Fig. 1 ), with the aim of adding to the synthetic biology toolset for this highly promising microbial chassis. Firstly, we established a cis,cis -muconic acid (ccMA) biosensor in C. glutamicum using CatM, a LysR-type transcription regulator (LTTR) (Maddocks & Oyston, 2008 ) from the soil bacterium Acinetobacter baylyi ADP1 (Ezezika et al., 2006 ). CatM responds to the effector molecule ccMA, and the experimentally determined structure of CatM confirms ccMA bound to CatM in the interdomain pocket (Ezezika et al., 2007 ). The CatM promoter was previously optimized for a ccMA biosensor in Pseudomonas putida by screening a CatM regulated promoter library and was shown to respond specifically to ccMA (Bentley et al., 2020 ). That same promoter library was used in this study and screened for biosensing performance in C. glutamicum . Secondly, we targeted the detection of chorismate, which is the final product of the shikimate pathway and a key intermediate for the synthesis of a wide range of aromatic compounds, including aromatic amino acids as well as various coenzymes (Averesch & Krömer, 2018 ; Masuo et al., 2016 ). By utilizing a native chorismate-dependent transcription regulator QsuR (Kubota et al., 2014 ) and screening a diversified P qsu promoter regulating a fluorescent reporter, we established a chorismate biosensor with improved sensitivity and dynamic range in C. glutamicum when compared to the simpler strategy of combining the native P qsu promoter with the same fluorescent reporter. While the biosensors are not precise enough to replace more quantitative tools like high-performance liquid chromatography (HPLC) (for extracellular titer of a metabolite) and mass spectrometry (MS)-based metabolomics (for intracellular metabolite concentration), the utility of theses biosensors spans from estimating relative carbon flux through their respective metabolic nodes to high-throughput screening of strains for productivity. Considering that the conversion of glucose to ccMA results from diversion of carbon from the shikimate pathway, dual sensing of chorismate and ccMA will be an attractive approach for estimating and optimizing the distribution of carbon flux between the shikimate pathway and the shunt pathway, respectively. Fig. 1. A schematic of the shikimate and β-ketoadipate pathways. In the C. glutamicum strain used in this study, the shikimate and β-ketoadipate pathways are connected by the native qsuB encoding 3-dehydroshikimate dehydratase (DHSase), as well as by heterologous enzymes, including chorismate pyruvate-lyase (UbiC, which converts chorismate to 4-hydoxybenzoate) and PCA decarboxylase (AroY, which converts PCA to catechol). Two of the industrially relevant molecules in these pathways are chorismate from the shikimate pathway and cis, cis -muconate (ccMA) from the β-ketoadipate pathway. Abbreviations: PEP, Phosphoenolpyruvate; E4P, Erythrose 4-phosphate; DAHP, 3-deoxy-D-arabino-heptulosonate 7-phosphate; 3DHQ, 3-Dehydroquinate; 3DHS, 3-Dehydroshikimate; 5-EPSP, 5-Enolpyruvylshikimate 3-phosphate; 4HB, 4-hydroxybenzoate; PCA, Protocatechuate; p-CA, p-Coumarate; βCMA, β-Carboxymuconate; βKA, β-ketoadipate; TCA, Tricarboxylic acid; CAT, Catechol. Created with BioRender.com",
"discussion": "Discussion Transcription factors are promising scaffolds for establishing biosensors in an organism. Although their promoters are active for native function, they usually need to be optimized for sensitivity, linear detection range, and contrast ratio (dynamic range) to increase their utility for biotechnological applications. There are several key regions in a TF and promoter that can be mutated for gain-of-function in a biosensor. Mutations in the −35/−10 and/or operator sites have led to increases in sensitivity and dynamic range and made TFs usable in new host organisms (Jha et al., 2014 , 2018 ; Bentley et al., 2020 ; Pardo et al., 2020 ). More specifically, because the mechanism of transcription activation and controlling features of various LTTRs are often reasonably conserved, they are good targets for semi-rational library design for biosensor development. An LTTR is a homo-tetramer that binds to at least two locations near the transcriptional initiation site (TI). The regulatory binding site typically positioned upstream from the TI binds the LTTR, which remains strongly anchored in either the apo form or when bound to the co-inducer. The other arm of the tetramer binds to an activation binding site with lower affinity. Conformational changes in the protein upon binding to the co-inducer shifts the binding position of the second arm marginally in some cases, or to a third LTTR-binding motif entirely in other cases. These shifts can alter DNA bending, expose important promoter regions or allow direct interaction between the LTTR and the RNA polymerase complex (Maddocks & Oyston, 2008 ). In the current study, the TFs belonged to the LTTR family and the knowledge of this class of TFs was applied for establishing biosensors in C. glutamicum ATCC 13032. In the current work, we pursued semi-rational design of promoter libraries for the development of biosensors for ccMA and chorismate (Masuo et al., 2016 ). We used a heterologous LTTR, namely CatM from A. baylyi ADP1 as a ccMA biosensor, while the native QsuR, also belonging to the LTTR family, was used for engineering a chorismate biosensor. Libraries of size on the order of ∼10 5 were built and screened using flow cytometry in both biosensor development workflows. While direct transfer of a biosensor cassette from one host to another can compromise its function, we were able to exploit a CatM regulated promoter library developed for P. putida to screen and isolate a promoter variant with much improved response in C. glutamicum in an expedited manner. The native P cat promoter from A. baylyi ADP1 failed to show any detectable dose-response with ccMA precursors, but seven mutations in the promoter and operator regions along with a strong RBS generated high sensitivity and dynamic range in response to ccMA (Fig. 2e ). In the case of the chorismate biosensor, a new promoter variant with six mutations and a strong RBS helped increase the sensitivity and dynamic range in response to feeding of quinate (chorismate precursor). Interestingly, and contrary to the functional ccMA biosensor in P. putida , we observed a response of this biosensor in C. glutamicum to the addition of exogenous ccMA to the growth medium. Although there are substantial differences in the membrane composition of P. putida (Gram-negative) and C. glutamicum (Gram-positive), it is still unlikely that any sufficient diffusion of ccMA across the membrane took place as it will carry a net −2 charge at the physiological pH. Thus, we hypothesized that the biosensor response was due to ccMA import facilitated by an active transporter. The fact that the ccMA biosensor in C. glutamicum showed response to extracellular ccMA at 10-fold to 100-fold higher concentrations than observed using the ccMA precursors benzoate and catechol (Fig. 2d ), a wide range of transporters might be promiscuously operating to transport ccMA, albeit poorly. Our investigation was biased towards MFS transporter-mediated ccMA transport in C. glutamicum since an annotated ccMA transporter, MucK belongs to the same family. A survey into the genome of C. glutamicum ATCC 13032 using MucK from A. baylyi ADP1 (Sequence ID P94131.1) revealed the presence of a relevant MucK homolog with only ∼27% sequence identity (Sequence ID BAC00372.1, Supplementary Information SI 2.2) in this strain, a protein regarded as an MFS transporter. Close homologs of this MFS transporter identified in other Corynebacterium strains were also annotated as a hypothetical protein (Sequence ID BAF55884.1, Supplementary Information SI 2.4.1) or as a sugar phosphate permease (Sequence ID WP_047263143.1, Supplementary Information SI 2.4.2) We surmise that the transporters of other families could also achieve the same task, as suggested by the discovery of several C 4 dicarboxylate transporters, such as DccT and DctA in C. glutamicum (Youn et al., 2008 , 2009 ). Future work that further investigates the role of such genes in transport of dicarboxylic acids will provide insights and gene knockouts could be used to measure the effect on ccMA transport using the ccMA biosensor. Shikimate is an intermediate metabolite in the shikimate pathway upstream of chorismate and can be used as a precursor to alter the intracellular chorismate pool. A shikimate biosensor was previously developed in C. glutamicum RES167 (derivative strain of C. glutamicum ATCC 13032) that showed a dose-dependent response to shikimate only when a heterologous gene shiA encoding for the shikimate transporter and natively present in C. glutamicum strain R (JCM18229) (Kubota et al., 2015 ), but not in C. glutamicum RES167, was included in the biosensor plasmid (Liu et al., 2018 ). In that study, cells grew poorly with shikimate as the sole carbon source, and the biosensor responded only weakly to exogenous shikimate until shiA was introduced on the biosensor plasmid, which improved cell growth and increased the fluorescence response. This may be why we did not observe a dose-dependent response to exogenously added shikimate. Like Liu et al. (Liu et al., 2018 ), we were unable to find a shiA homologue, nor any other suitable gene target that could transport shikimate in C. glutamicum ATCC 13032. We expected some cross reactivity in functions of transporter proteins for quinate and shikimate (due to similarity of the two molecules), specifically that the quinate transporter QsuA could act as a transporter for shikimate, but we failed to detect any increase in the chorismate biosensor response upon exogenous shikimate addition (Fig. 3c ). By contrast, Kubota et al. observed activation of the QsuR-controlled operon by shikimate in a C. glutamicum strain R and but this be attributed to the presence of the shiA gene coding for a shikimate transporter in that particular strain. Both biosensors have their limitations. As with most TF based fluorescence biosensors, they become unreliable below a certain sensitivity threshold or above a certain saturation threshold. In this study, the biosensors were induced with the external addition of precursor metabolites rather than directly with the ligands for the TFs for various transport and stability reasons. As such, the exact dynamic range with respect to intracellular concentrations of the actual ligands is still unknown. Future studies might compare the biosensor responses to intracellular concentrations by liquid chromatography mass spectrometry (LC-MS) evaluation of cell lysates or performing assays in cell free conditions. These biosensors are not expected to be quantitative bioassay replacements for analytical techniques such as HPLC or LC-MS, but they are well suited to high throughput screening of adaptive laboratory evolution populations or mutation libraries, where the fluorescence response correlates to the production rate or intracellular pool of the metabolites. In such applications, the biosensors are used for initial screening, while the actual performance of the selected subset of strains is validated by HPLC. Further optimization of the biosensors is possible by targeted modifications to the promoter sequence or to the TF proteins, but this study highlights the utility of these biosensors and that their dynamic ranges are such as to respond to some biologically relevant perturbations such as metabolic knockouts or enzyme activity changes upstream or downstream of ccMA and chorismate nodes. Dual sensing for metabolites using genetically encoded biosensors in C. glutamicum was demonstrated in this study. We showed that by feeding different precursors for ccMA and chorismate (i.e., benzoate and quinate, respectively) to biosensor strains of C. glutamicum , green and red fluorescence signals could be used to detect intracellular ccMA and chorismate, respectively. We also observed an indirect effect of quinate (chorismate precursor) on the ccMA biosensor. Given the lack of PCA decarboxylase (aroY) in C. glutamicum , it is unlikely that quinate was being shunted through PCA and into ccMA production, especially considering that the ccMA biosensor does not respond to exogenous PCA (Fig. 2c ). A strain with gene knockout (Δ qsuB ) that disconnects the shikimate pathway from the β-ketoadipate pathway did not show the indirect effect of quinate on the ccMA biosensor (Fig. 4f ). At this stage, we could not conclude if the effect was due to any metabolic advantage conferred by quinate or any regulatory crosstalk between the shikimate/quinate pathway intermediate and β-ketoadipate pathway. In summary, we built promoter libraries in C. glutamicum ATCC 13032 and screened them to develop and optimize biosensors for a central intermediate (chorismate) and an industrial chemical precursor (ccMA). In the process, we observed what we believe is ccMA transported into C. glutamicum ATCC 13032. This work will expand the toolkit for C. glutamicum , which is recognized as a promising microbial chassis for bio-based production of a wide range of chemicals, fuels, polymer precursors and healthcare products (Becker, Rohles, et al., 2018 )."
} | 4,647 |
38961629 | PMC11222294 | pmc | 4,931 | {
"abstract": "Abstract Ciliate protozoa are an integral part of the rumen microbial community involved in a variety of metabolic processes. These processes are thought to be in part the outcome of interactions with their associated prokaryotic community. For example, methane production is enhanced through interspecies hydrogen transfer between protozoa and archaea. We hypothesize that ciliate protozoa are host to a stable prokaryotic community dictated by specific functions they carry. Here, we modify the microbial community by varying the forage‐to‐concentrate ratios and show that, despite major changes in the prokaryotic community, several taxa remain stably associated with ciliate protozoa. By quantifying genes belonging to various known reduction pathways in the rumen, we find that the bacterial community associated with protozoa is enriched in genes belonging to hydrogen utilization pathways and that these genes correspond to the same taxonomic affiliations seen enriched in protozoa. Our results show that ciliate protozoa in the rumen may serve as a hub for various hydrogenotrophic functions and a better understanding of the processes driven by different protozoa may unveil the potential role of ciliates in shaping rumen metabolism.",
"conclusion": "CONCLUSION Our study sheds light on the underexplored ecological dynamics of ciliate protozoa in the rumen, emphasizing the significant disparity in their response to environmental changes, in this case, dietary shifts, when compared to prokaryotic communities. While the prokaryotic community exhibits a strong collective response to alterations in diet, the protozoa population size is influenced by dietary changes but shows no significant shifts in community composition. This observation underscores the stability of protozoa in the rumen and hints at diverse, unexplored ecological forces governing their community dynamics. Our findings reveal that although the ciliate protozoa‐associated prokaryotic community is largely defined by their surrounding prokaryotic community, the retention of significant differences in the protozoa‐associated community, irrespective of diet, suggests a level of selection occurring within this community, which may be the result of mutualistic interaction or selective predation. Furthermore, our investigation into key hydrogen‐utilizing pathways and associated genes indicates that ciliate protozoa in the rumen may serve as a hub for various hydrogenotrophic functions, extending beyond methanogenesis. The potential metabolic interactions between protozoa and nitrate/nitrite‐reducing bacteria suggest a complex relationship that could contribute to nitrate reduction and may have implications for methane mitigation strategies. Overall, our results reveal the potentially impactful role of ciliate rumen protozoa as stable interaction partners and as a source of energy and habitat for diverse prokaryotic species. As we unveil the intricate dynamics of protozoa in the rumen, further research into their ecological roles and interactions holds promise for advancing our understanding of microbial communities and their potential contributions to addressing environmental challenges, such as methane reduction.",
"introduction": "INTRODUCTION The rumen microbial community is considered to be one of the most diverse and complex among host‐associated communities (de Jonge et al., 2022 ). Unlike monogastric animals in which microbial fermentation represents but a fraction of the total energy balance of the host, the rumen microbiome is responsible for the vast majority of the carbon and nitrogen requirements of the ruminant animal (Bergman, 1990 ). This is performed by a vast microbial community encompassing all domains of life; bacteria, archaea and microbial eukaryotes. The feed ingested by the host is broken down by the microbes involving complex cascades of cross‐feeding and interactions that lead to the production of end products for the animal in the form of volatile fatty acids as well as the production of the unusable methane. Although the understanding of the rumen microbial ecology as well as the role and interactions between the rumen bacterial community has increasingly been studied, the eukaryotic portion of the rumen microbiome, their dynamics in the rumen and interactions remain largely neglected (Firkins et al., 2020 ; Mizrahi & Jami, 2021 ; Solomon & Jami, 2021 ). The largest fraction of the rumen eukaryotic community is the ciliate protozoa, which account for 2% of all microbial species and can account for up to 50% of the biomass (Cottle et al., 1978 ; Newbold et al., 2015 ). Several studies, such as Henderson et al, surveyed ruminant species across different geographies and diets, emphasized that ciliate protozoa may be more diverse than originally expected and that they may not be constrained by the same ecological forces as the prokaryotic community (Cui et al., 2022 ; Henderson et al., 2015 ). In addition to their still elusive ecological dynamics, ciliate protozoa have also been suggested to have many metabolic and physiological functions in the rumen, most of which have not yet been elucidated (Li et al., 2022 ). Some of these functions are further suggested to be a direct consequence of their interactions with the prokaryotic community (Ushida et al., 1997 ; Villar et al., 2020 ). The most studied functional interaction is their role in enhancing methane production via interspecies hydrogen transfer to methanogenic archaea (Ushida et al., 1997 ). Such interaction was observed across a wide range of in‐vivo and in‐vitro studies, with the estimated enhancement of methane due to the presence of protozoa ranging from 11% to 37% (Newbold et al., 2015 ; Ranilla et al., 2007 ; Solomon et al., 2021 ; Ushida et al., 1997 ), and not limited to the rumen environment (Treitli et al., 2023 ; Yamada et al., 1997 ). Additional evidence for the interaction stems from the physical association observed between ciliate protozoa and methanogens (Levy & Jami, 2018 ; Lloyd et al., 1996 ; Park & Yu, 2018 ). These studies note that, in addition to methanogens, protozoa likely harbour a diverse bacterial community for which as of yet no type of interaction has been proposed. In this study, we investigate the effect of host diet on the protozoa composition and the composition of their associated prokaryotic community. We find that the protozoa community is less affected by dietary changes compared to the prokaryotic community. We also discover that while the protozoa‐associated prokaryotic community is dependent on the free‐living (FL) community and thus changes with diets, the protozoa microenvironment retains a unique community of enriched taxa independently of their abundance in the FL community. The enrichment of these taxa may be the result of functional interactions similar to methanogens, driven by their production of hydrogen.",
"discussion": "DISCUSSION The ecological dynamics of ciliate protozoa in the rumen remains seldom analysed when compared to bacteria and archaea. This phenomenon is not only true for the rumen environment but has been true for most other environments leaving a gap in our understanding of complex microbial communities (Caron et al., 2009 ; Gao et al., 2019 ; Solomon & Jami, 2021 ). Our results show that while the rumen prokaryotic community responds strongly to changes in fibre content in the diet as a community, the protozoa do not respond to changes in their environment caused by the differing diets. More specifically, while the protozoa population size increases with decreased fibre content in the feed as previously observed (Hook et al., 2011 ; Towne et al., 1988 ), there were no significant changes in community composition, as opposed to the bacterial community. Henderson et al, which assessed the protozoa community across different cohorts of different species of ruminants, similarly reported a strong host individuality in protozoa communities even within specific cohorts (Henderson et al., 2015 ). Furthermore, a recent study in the yak rumen found several differences in the relative abundance of some protozoa species in response to feed type, but overall community structure as measured by Beta diversity was not significantly different as opposed to bacterial or the fungal community (Cui et al., 2022 ). Our results further reveal a higher stability of protozoa species compared to prokaryotic species in the rumen, when examining the community changes under varying diets (Figure 1C ). This observation was in line with previous results showing that changes in fibre to concentrate ratio significantly affected bacterial, archaeal and fungi composition, but not the protozoa community structure (Tapio et al., 2017 ). In contrast, supplementation of the animal feed with lipids was shown to have an effect on the protozoa community and protein expression (Andersen et al., 2023 ; Tapio et al., 2017 ). Interestingly, the study by Andersen et al. ( 2023 ) also noted that when a high starch diet is given, protozoa species with well‐established starch degradation functions, are not necessarily more active, which they propose is the result of competition with starch‐degrading bacteria or sub‐optimal pH conditions in the rumen. The large difference in the response of protozoa and prokaryotes to stark changes in the rumen environment such as animal diet hints to diverse unexplored ecological forces driving the community of protozoa in the rumen. The individuality in host animals may be driven by genetic factors (Wallace et al., 2019 ), or alternatively, differential early acquisition may dictate protozoa composition later in life. The latter alternative may carry important implications as to the possibility of early intervention in order to stably modulate the protozoa community (Mizrahi & Jami, 2021 ). Studies on association between protists and prokaryotes in general and specifically ciliate protozoa have shown that the nature of interaction can be diverse, ranging from antagonistic interaction such as predation, to commensal, all the way to mutualistic interactions (Gao et al., 2019 ; Gast et al., 2009 ; Paisie et al., 2014 ; Pernthaler, 2005 ). Our results indicate that the ciliate protozoa‐associated prokaryotic community is strongly driven by the community surrounding them, and changes induced in this community result in different associated communities. This was seen on two levels, with different diets modulating the protozoa‐associated bacterial community and on the individual level, as the composition reflected the FL community in each individual host. However, our results also show that the protozoa‐associated community retains significant differences when compared to the FL community regardless of diet, suggesting that a certain degree of selection may occur in this community. This is particularly evident for taxa belonging to the Proteobacteria phylum, methanogens and Oscillospirales. The differences observed could stem from preferential predation for which inconclusive evidence exists for rumen protozoa (Coleman, 1964 ; de la Fuente et al., 2011 ; Gutierrez, 1958 ), or from the provision of habitats and nutrients for mutualistic exchange with their associated bacteria. It is notable that none of the hydrogen‐utilizing gene sequences obtained belonged to the γ‐Proteobacteria phylum, despite it being highly enriched in the protozoa‐associated community. Furthermore, in a previous study, γ‐Proteobacteria were also highly enriched in the presence of protozoa in in‐vitro rumen microcosms compared to microcosms incubated without protozoa (Solomon et al., 2022 ). As these taxa were shown to carry genes conferring them resistance to predation (Gong et al., 2016 ), their presence may reflect their persistence and accumulation in the prokaryotic community associated with protozoa. As well‐established hydrogen‐producers, protozoa are suggested to create an attractive micro‐environment for taxa such as methanogens that utilizes hydrogen to reduce CO 2 to methane (Newbold et al., 2015 ; Ushida et al., 1997 ). By quantifying key genes belonging to hydrogen utilizing pathways, other than methanogenesis, known to exist in the rumen, we found that the community associated with ciliate protozoa is significantly enriched in those genes. In the case of potential nitrite and sulfate‐reducing taxa, the gene sequences obtained could be attributed to taxonomic affiliations seen as enriched in association with protozoa via amplicon sequencing. Our results thus suggest that ciliate protozoa in the rumen may be a hub for various hydrogenotrophic functions, beyond methanogenesis. This observation is not without precedents and different types of hydrogen transfer‐based putative mutualism have been suggested between prokaryotes and ciliate (Bernhard et al., 2000 ; Gast et al., 2009 ; Ott et al., 2004 ). In termites, putative bacterial symbionts from the Myxococcota (formerly Deltaproteobacteria ) and Spirochaetota phylum were shown to have a wide range of hydrogenotrophic function which appear to be the nature of the interaction between them and their flagellate protozoa host (Ikeda‐Ohtsubo et al., 2016 ; Kuwahara et al., 2017 ; Ohkuma et al., 2015 ). Likewise, in anoxic fresh water lake, a ciliate endosymbiont was found to supply its ciliate host with energy via denitrification performed by hydrogen transfer (Graf et al., 2021 ). In the rumen, metabolic interactions between ciliate protozoa and nitrate/nitrite‐reducing bacteria have been proposed (Lin et al., 2011 ; Roman‐Garcia et al., 2019 ; Villar et al., 2020 ; Welty et al., 2019 ). When co‐cultured with bacteria, rumen protozoa were reported to accelerate nitrate reduction (Villar et al., 2020 ), and the protozoal fraction, which likely included both protozoa‐associated prokaryotes showed a greater ability to reduce nitrate without accumulation of nitrite (Lin et al., 2011 ), than the prokaryotic community alone. The possibility that, combined nitrate reduction to nitrite by protozoa, for which genomic evidence exists, with production of hydrogen may lead to the enrichment of nitrite‐reducing bacteria (Roman‐Garcia et al., 2019 ; Welty et al., 2019 ). Considering the observation that exogenous addition of nitrite (but not nitrate) severely inhibits motility and chemotaxis of protozoa (Roman‐Garcia et al., 2019 ), we can speculate that its clearance by associated bacteria might encompass the mechanism of mutualistic interaction between them. Our results suggest that the protozoa‐associated microbiome may possibly complete the reduction of nitrate. Successful decoupling of protozoa and associated bacteria would be required to assess such. As nitrate and other nitrogenous compounds are being evaluated for their potency to reduce methane, the role of protozoa and their associated microbes in metabolizing these compounds and their contribution to the mitigation of methane is an interesting avenue for further research (Morgavi et al., 2023 ; Yang et al., 2016 )."
} | 3,764 |
31663059 | PMC6787351 | pmc | 4,932 | {
"abstract": "Biofilm research is usually focused on the prevention or control of biofilm formation. Recently, the significance of the biofilm mode of growth in biotechnological applications received increased attention. Since biofilm reactors show many advantages over suspended cell reactors, especially in their higher biomass density and operational stability, bacterial biofilms have emerged as an interesting approach for the expression of specific proteins. Despite the potential of biofilm systems, recombinant protein production using biofilms has been scarcely investigated for the past 25 years. Our group has demonstrated that E. coli biofilms were able to produce a model recombinant protein, the enhanced green fluorescent protein (eGFP), at much higher levels than their planktonic counterparts. Even without optimization of cultivation conditions, an attractive productivity was obtained, indicating that biofilm cultures can be used as an alternative form of high cell density cultivation (HCDC). E. coli remains one of the favorite hosts for recombinant protein production and it has been successfully used in metabolic engineering for the synthesis of high value products. This review presents the advantages and concerns of using biofilms for the production of recombinant proteins and summarizes the different biofilm systems which have been described for this purpose. The relative advantages and disadvantages of the four microbial hosts tested for recombinant protein production in biofilms (two bacteria and two filamentous fungi) are also discussed.",
"conclusion": "5. Conclusions Biofilm reactors present many advantages over suspended cell reactors, including higher cell densities and long-term stability as required for continuous processing. Furthermore, according to the work published on recombinant protein production, microbial biofilm systems are sometimes able to produce recombinant proteins at attractive levels. Nevertheless, several elements that affect the overall efficiency of biofilm reactors, especially limitations in the diffusion of nutrients and oxygen, call for careful implementation of strategies specifically targeted for increasing recombinant protein production in these systems. Additionally, there are several hosts that can be chosen for recombinant protein production in biofilms. However, the selection of the most appropriate expression host is limited to a few options given the characteristics of the protein and the process to be developed. In this review, we focused on two bacteria ( E. coli and B. subtilis ) and two filamentous fungi ( A. niger and A. oryzae ) since they have already been tested for recombinant production in biofilms. E. coli remains one the favorite hosts for recombinant protein production at lab and industrial scale since it offers many advantages over the other host systems, namely fast growth at high cell densities, well-characterized genetics, and availability of a large number of cloning vectors.",
"introduction": "1. Introduction Recombinant proteins are synthesized in a host cell which is usually of a different species from the source of the DNA encoding them [1] and in that case they can be called heterologous proteins. Recombinant proteins have wide applications in medicine, research and biotechnology. With the development of recombinant protein methodology, it is possible to clone genes encoding proteins from different organisms and express them in other organisms at much higher levels than those naturally achieved [1] , [2] , leaving behind the necessity of huge amounts of animal and plant tissues or volumes of biological fluids [3] . To obtain recombinant proteins, the gene encoding the protein is isolated and introduced into an expression vector, afterwards it is transformed into the chosen host system [1] , [3] . An important step in recombinant protein production is the choice of the ideal expression system. A large number of protein expression hosts are available, such as bacteria, yeast, filamentous fungi, and mammalian, plant and insect cells [4] , [5] . It is also important to select the most suitable expression vector, which is composed by a set of genetic elements that affect both transcriptional and translational steps of protein production, namely the origin of replication, promoter, ribosome binding site, start codon, transcriptional terminator, and selective marker [3] . Recombinant proteins have been mostly produced in suspended cultures. The insertion of the gene of interest into a multicopy plasmid may result in high recombinant protein expression [1] . Nevertheless, this may impose a metabolic burden on the host cell due to the energy and metabolites needed for the replication of plasmid DNA and synthesis of recombinant proteins [1] , [6] . In planktonic cells, this added metabolic burden may decrease the cellular growth rates and biomass yield, besides affecting the yield and activity of the desired protein [6] . In order to overcome these problems, a strategy based on the use of biofilm systems has been studied. Microbial biofilms are communities of microorganisms, single or multispecies, attached to surfaces which are embedded in a self-produced matrix of extracellular polymers [7] – [9] . Biofilms are usually known for their negative effects on health and industrial sectors as they can cause diseases, equipment corrosion, local clogging, heat transfer resistance and product contamination in food processing environments [7] , [10] . However, biofilms have beneficial use in wastewater treatment [11] and are being tested for the production of solvents, organic acids and enzymes [12] – [14] . The biological organization of biofilms provides them with many advantages over the suspended cells, including high cell density and protection against hostile conditions [15] . Furthermore, the presence of expression vectors in the sessile cells has shown to increase biofilm formation and lead to higher production of recombinant proteins compared to planktonic cells [16] – [19] . This review will focus on the production of recombinant proteins by biofilms. The main studies using biofilm systems for the production of recombinant proteins will be explored, as well as the differences between producing recombinant proteins in suspended and biofilm cultures. The advantages and disadvantages of using different microbial hosts in the production of recombinant proteins, namely Escherichia coli , Bacillus subtilis , Aspergillus niger and Aspergillus oryzae , will also be addressed."
} | 1,633 |
24053676 | PMC3851129 | pmc | 4,933 | {
"abstract": "Background Understanding the process of amino acid fermentation as a comprehensive system is a challenging task. Previously, we developed a literature-based dynamic simulation model, which included transcriptional regulation, transcription, translation, and enzymatic reactions related to glycolysis, the pentose phosphate pathway, the tricarboxylic acid (TCA) cycle, and the anaplerotic pathway of Escherichia coli . During simulation, cell growth was defined such as to reproduce the experimental cell growth profile of fed-batch cultivation in jar fermenters. However, to confirm the biological appropriateness of our model, sensitivity analysis and experimental validation were required. Results We constructed an l -glutamic acid fermentation simulation model by removing sucAB, a gene encoding α-ketoglutarate dehydrogenase. We then performed systematic sensitivity analysis for l -glutamic acid production; the results of this process corresponded with previous experimental data regarding l -glutamic acid fermentation. Furthermore, it allowed us to predicted the possibility that accumulation of 3-phosphoglycerate in the cell would regulate the carbon flux into the TCA cycle and lead to an increase in the yield of l -glutamic acid via fermentation. We validated this hypothesis through a fermentation experiment involving a model l -glutamic acid-production strain, E. coli MG1655 Δ sucA in which the phosphoglycerate kinase gene had been amplified to cause accumulation of 3-phosphoglycerate. The observed increase in l -glutamic acid production verified the biologically meaningful predictive power of our dynamic metabolic simulation model. Conclusions In this study, dynamic simulation using a literature-based model was shown to be useful for elucidating the precise mechanisms involved in fermentation processes inside the cell. Further exhaustive sensitivity analysis will facilitate identification of novel factors involved in the metabolic regulation of amino acid fermentation.",
"conclusion": "Conclusions In this study, we evaluated a literature-based dynamic metabolic pathway model of E. coli by computational analysis and verified it experimentally. Our kinetic metabolism model was particularly useful for analysis of feedback regulation systems, and it was stable and robust against perturbation. In future, we would need to improve cell growth modelling to improve the flexibility and quantitative prediction capability of the current model. This resource will contribute to metabolic engineering that predicts the key factors for substance production.",
"discussion": "Discussion In our previous study, we constructed a dynamic simulation model of E. coli based on biological knowledge and reproduced the experimental cultivation results by parameter fitting [ 10 ]. In this study, we attempted to elucidate novel factors that affect l -glutamic acid fermentation by using dynamic simulation based on a computer-aided rational design of biochemical networks. First, we refined the model with respect to biomass production through model validation. Then, a precise sensitivity analysis was performed and revealed many factors that would be important for l -glutamic acid fermentation. For example, an increase in the expression of gltA , which encodes citrate synthase, icdA , which encodes isocitrate dehydrogenase, and a combined increase in the expression of both these genes were predicted to have a high impact on l -glutamic acid production. In fact, an increase in the expression of gltA or icdA enhanced l -glutamic acid production in E. coli (Table 2 ). These genes have already been utilized to optimize l -glutamic acid fermentation in an industrial strain, C. glutamicum [ 20 , 21 ], thus supporting the accuracy of our dynamic simulation model for understanding E. coli metabolism. In a previous study, we proposed that the putative transcriptional regulator YdcI controls carbon flux into the TCA cycle in E. coli [ 31 ]. We observed that an increase of citrate synthase activity by deletion of ydcI led to an increase in l -glutamic acid production, and a decrease of citrate synthase activity due to ydcI amplification led to a decrease in l -glutamic acid production. Our sensitivity analysis results support a clear relationship between YdcI and l -glutamic acid production, because both sensitivity analysis and experimental results clearly showed that a change in citrate synthase expression levels exert a significant effect on l -glutamic acid production in E. coli . Integration of the transcriptional regulator YdcI into our dynamic simulation model is a subject for future studies. In theoretical flux analysis, the metabolic flux distribution, which facilitates the maximal theoretical yield of l -glutamic acid in E. coli , indicated that flux through α-ketoglutarate dehydrogenase (encoded by sucAB ) and the glyoxylate shunt pathway (encoded by aceBA ) should be 0. Thus, to achieve maximal l -glutamic acid production from the MG1655 Δ sucA strain, deletion of the aceBAK operon, which encodes enzymes in the glyoxylate shunt pathway and isocitrate dehydrogenase kinase/phosphatase, or amplification of iclR , which encodes a negative regulator of the glyoxylate shunt, would be a plausible approach. In contrast, using the dynamic simulation model in addition to sensitivity analysis, we identified unexpected factors. This approach indicated that amplification of pgk and attenuation of gpmA and/or eno would induce accumulation of 3-phosphoglycerate, which inhibits the phosphorylation of isocitrate dehydrogenase, and consequently results in increased glutamate production. Furthermore, the effects of amplification of pykF or pdhR on l -glutamate production could be explained according to a different mechanism. This type of working hypothesis cannot be generated using conventional theoretical flux analysis because it requires that the modification systems that control enzyme activity be taken into account. Thus, a critical advantage in using the dynamic simulation model is to be able to take various modes of regulation into consideration simultaneously and comprehensively. In biotechnology, both production yield and productivity are important. In general, it is not simple to maintain productivity while improving production strains because the achievable yield and productivity can vary, depending on the strains and production conditions. From an engineering point of view, improving yield or productivity from a current production strain is a practical issue. One of the key features that affect productivity is biomass formation. In this study, we improved our simulation model to describe biomass production as precisely as possible; however, it was not sufficiently accurate to predict productivity. In E. coli , the carbon flux through the TCA cycle is known to affect biomass production directly [ 32 ]. Our dynamic simulation predicted the changes in concentrations of each molecule with response to gene amplification or deletion, and based on simulation results, we could speculate whether these perturbations would affect cell growth through carbon flux into the TCA cycle. In this study, we predicted that the concentration of 3-phosphoglycerate could play an important role in controlling the carbon flux through the TCA cycle; we subsequently validated this hypothesis experimentally. We also estimated that metabolic regulation through 3-phosphoglycerate would contribute to the changes in biomass production and fermentation productivity. In future, we plan to describe these changes more precisely in our model. Our model requires further refinement. When we perturbed the copy number of ppc , which encodes PEP carboxylase, the sensitivity was 0.02138 and the ranking of this gene was 17 th . In a previous study on C. glutamicum, it had been experimentally shown that an increase in PEP carboxylase activity led to an increase in l -glutamic acid yield, with a reduction in organic acid byproducts [ 33 ]. We speculated that one of the reasons for the low sensitivity of ppc was related to the process of parameter tuning because the catalytic constant of PEP carboxylase was modified to 100 times greater than the reported value. Thus, an increase in PEP carboxylase was not sensitively related to l -glutamate concentration in our model [ 10 ]. However, there is a difference in anaplerotic pathway enzymes and their regulation in E. coli and C. glutamicum [ 34 ]. Together with these facts, we speculate that unknown mechanisms, related to PEP carboxylase or an anaplerotic pathway in E. coli , could exist and result in inadequate modeling. The citrate synthase reaction requires both oxaloacetate and acetyl-CoA as substrates. An increase in PEP carboxylase expression will intensify the carbon flux toward oxaloacetate. However, to enhance citrate synthase activity, acetyl-CoA would be required. In our simulation model, the current acetate production model may be too simplistic to describe the dynamic changes in acetyl-CoA concentration inside the cell because the regulation of acetate metabolism, including formation, excretion, uptake, and utilization of this substance, is quite complex in reality [ 10 ]. In the future, improvement in simulating the metabolism of acetate, including its excretion, is a priority for refining our model. We recognize that our model is currently limited in terms of quantitative predictive power. According to our sensitivity analysis, amplification of pgk would increase glutamate production yield by 106%, whereas, experimentally, we found a 120% increase. One of the factors that affect the prediction seems to be the difference between the experimental conditions considered for simulation and those used for validation. In the simulation model, data for the parameters were obtained using jar fermenters; however, experimental validation involved the use of shake flasks [ 10 ]. We assume that the most important issue related to the predictive power is the modeling of cell growth. In l -glutamic acid production using E. coli MG1655 Δ sucA , the succinyl-CoA used for biomass formation should be supplied through the glyoxylate shunt. Amplification of pgk in E. coli MG1655 Δ sucA decreases the carbon flux toward the glyoxylate shunt pathway. Consequently, we observed 2 phenomena, viz., increased l -glutamic acid production and decreased biomass production. At present, this trade-off has not been accommodated in our simulation because the cell growth profile is fixed in accordance with experimental results obtained using jar fermenters [ 10 ]. To further refine our model, the cell growth profile should be allowed to be variable, based on the concentration of biomass precursor molecules. Experimentally verified biological evidence will continue to be appropriately incorporated into our model during further refinements."
} | 2,730 |
29062948 | PMC5625796 | pmc | 4,935 | {
"abstract": "The rapid development of synthetic biology enables the design, construction and optimization of synthetic microbial consortia to achieve specific functions. In China, the “973” project-“Design and Construction of Microbial Consortia” was funded by the National Basic Research Program of China in January 2014. It was proposed to address the fundamental challenges in engineering natural microbial consortia and reconstructing microbial consortia to meet industrial demands. In this review, we will introduce this “973” project, including the significance of microbial consortia, the fundamental scientific issues, the recent research progresses, and some case studies about synthetic microbial consortia in the past two and a half years.",
"introduction": "1 Introduction In natural environments, 99% microorganisms exist in the form of microbial consortia. However, some defects of naturally occurring microbial consortia, such as difficulty in culturing, long operation cycle, low conversion efficiency, and poor stability and controllability, limited their practical applications in biotechnology industries. Synthetic microbial consortia constructed via synthetic biology approaches would be an alternative for programming novel complex behaviors and optimal features for practical biotechnology applications. Arnold [1] and Weiss [2] , [3] pointed out that synthetic microbial consortia could perform even more complicated tasks and endure more changeable environments than that of monocultures, thus providing an important new frontier for synthetic biology. A better knowledge of the multicellular systems that drive cell-cell interactions in the consortia was highly needed [4] , [5] . Engineering novel cell-cell interaction capabilities became crucial in the nascent field of synthetic biology [2] . Recently, scientists have made great progresses about analysis, design and construction of microbial consortia ( Fig. 1 ). Stephanopoulos et al. constructed an Escherichia coli- Saccharomyces cerevisiae consortium to successfully produce oxygenated taxanes [6] , and an E. coli- E. coli consortium to produce muconic acid [7] , [8] and 3-amino-benzoic acid [9] . Jones and his colleges [10] constructed an E. coli- E. coli co-culture for the efficient production of flavonoids. Lin et al. [11] designed and constructed a fungal-bacterial consortium to efficiently produce isobutanol from cellulose. Shou et al. [12] , [13] , [14] have been focused on engineering and analyzing the underlying mechanisms of cell-cell communication for many years. A series of synthetic syntrophic communities were constructed to probe the metabolic cross feeding principles underlying the complex microbial consortia [15] , [16] , [17] , [18] , [19] . Fig. 1 Applications of synthetic microbial consortia. Part 1 is adapted by permission from Nature Biotechnology [6] © . Part 2 is adapted by permission from Metabolic Engineering [10] © . Part 3 is adapted by permission from Proc Natl Acad Sci USA [7] © . Part 4 is adapted by permission from Proc Natl Acad Sci USA [11] © . Part 5 is adapted by permission from Proc Natl Acad Sci USA [12] © . Part 6 is adapted by permission from Proc Natl Acad Sci USA [15] © . Fig. 1 In the USA, Defense Advanced Research Projects Agency (DARPA) announced a funding entitled “Biological Robustness in Complex Settings (BRICS)” in August 2014. The BRICS program aimed to design synthetic communities consisting of multiple organisms and to elucidate the design principles of engineering robust microbial consortia. The end-program objective is to engineer robust, stable, and safe bio-systems. In China, the “973” project-“Design and Construction of Microbial Consortia”, funded by the National Basic Research Program of China in January 2014 was proposed to address fundamental challenges in engineering natural microbial consortia and reconstructing artificial microbial consortia to meet industrial demands. Great progresses were made in China about the analysis, design, construction of microbial consortia related to microbial fuel cells (MFCs) [20] , vitamin C fermentation [21] , [22] , polyhydroxyalkanoate (PHA) production [23] , methane production [24] , wastewater treatment [25] , biodegradation [26] , etc. For further design and construction of synthetic microbial consortia, a synthetic biology module library SynbioML@TJU ( http://www.synbioml.org/ ) that contains more than 5000 artificial synthetic genes and functional modules for diverse products was established by Tianjin University, China. SynbioML covers synthetic genes and functional modules for biosynthesis of natural products (terpenes, flavonoid, polyketones, alkaloids, etc), chemical products (steroids, aminoglycosides, polypeptides, etc), nutrition and health care products (fatty acids, vitamins, etc), biofuels (bioethanol, aliphatic alcohols, butanediol, etc), environmental biological sensors, microbial fuel cells, etc. All of the physical modules, preserved at Tianjin University, can be freely obtained through SynbioML website. Users can search for modules in the website by gene name, protein name, E.C. number, metabolic pathway, enzyme reaction etc."
} | 1,299 |
29062958 | PMC5625795 | pmc | 4,938 | {
"abstract": "Cell-free synthetic biology emerges as a powerful and flexible enabling technology that can engineer biological parts and systems for life science applications without using living cells. It provides simpler and faster engineering solutions with an unprecedented freedom of design in an open environment than cell system. This review focuses on recent developments of cell-free synthetic biology on biological engineering fields at molecular and cellular levels, including protein engineering, metabolic engineering, and artificial cell engineering. In cell-free protein engineering, the direct control of reaction conditions in cell-free system allows for easy synthesis of complex proteins, toxic proteins, membrane proteins, and novel proteins with unnatural amino acids. Cell-free systems offer the ability to design metabolic pathways towards the production of desired products. Buildup of artificial cells based on cell-free systems will improve our understanding of life and use them for environmental and biomedical applications.",
"conclusion": "5 Conclusions Cell-free synthetic biology proves a promising tool to overcome inherent limitations of living cells. Its open nature enables flexible biological engineering at both molecular and cellular levels. Because cost remains a top concern in industry, cell-free biosynthesis methodology is well suited for the development of high-value biopharmaceuticals. It is believed that cell-free systems would become more commonly used for basic and applied research in the future. Despite the promising features of cell-free synthetic biology, challenges still remain. The very first is protein post-translational modification, which is critical in the biology studies and the disease treatment. These modifications include glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation, lipidation and proteolysis. The second is how to expand the genetic code to incorporate multiple different uAAs into a single protein. The next is reuse of the cell-free systems. To address these challenges, cell-free synthetic biosystems must be further optimized to flexibly regulate the transcription and translation by gene editing and addition of exogenous substances. A proposed solution for reusing the cell-free system is designing a membrane bioreactor to extend the lifetime of cell-free system by continuously removing the inhibitory molecules. To broaden the applications, cell-free synthetic biology needs to be integrated with other cutting-edge technologies, such as stem cell, 3D printing, microbiome, neuroscience, and artificial intelligence.",
"introduction": "1 Introduction Advances in DNA sequencing and gene editing technologies have endowed the synthetic biologist with unprecedented power to program cells at will. The ability of synthetic biology to engineer biological functions holds great promises for applications ranging from biomedical to biofuel research. For the most part, synthetic biology is still tied to the living cell. One major advantage of using the living cell is its self-reproduction. However, the daunting complexity of living cells and the barriers of cell membrane make engineering difficult, and therefore make synthetic biology face four insurmountable challenges [1] : hard to standardize, unwieldy complexity, incompatibility and variability. From the standpoint of synthetic biology, it is highly desirable for these problems to be overcome using a standardized set of better engineering solutions. To address these challenges, an emerging interdisciplinary approach has been adopted: cell-free synthetic biology. Cell-free synthetic biology system activates biological machinery without the use of living cells. It allows direct control of transcription, translation and metabolism in an open environment. Three types of cell-free systems have been well developed. One is extract-based system. The system is composed of crude extract with basic transcription and translation functions, DNA templates, energy regeneration substrates, amino acids, nucleotides, cofactors, and salts. Most commonly used organisms providing the extracts are Escherichia coli \n [2] , Saccharomyces cerevisiae \n [3] , rabbit reticulocyte [4] , wheat germ [5] , and insect cell [6] . The other one is purified system, such as the PURE system which consists of a toolbox of purified E . coli translational components [7] . The third one is synthetic enzymatic pathway system, which consists of numerous enzymes for implementing complicated bioreactions [8] . The ability of cell-free systems to harness a cell's capabilities unimpeded by cells opens new opportunities for the academic research and industry applications. Briefly, reduced dependence on cells drives the increase in engineering flexibility. As a result, in vitro cell-free systems have many advantages over traditional in vivo cell systems ( Table 1 ) [8] , [9] , [10] , which include controllable transcription, translation and post-translational modification, convenient high-throughput screening format, accelerated design-build-test-learn cycle, high synthesis rate and product yield, easy production of soluble membrane proteins and complex proteins, easy incorporation of unnatural amino acids (uAAs), high tolerance for toxic substrates or products, and good ability to focus on particular metabolism. Table 1 Comparison of in vitro cell-free systems and traditional in vivo cell systems. Table 1 Feature In vitro cell-free system In vivo cell system Manipulation of transcription and translation Easy to control in an open environment Hard because of cell membrane as the barrier Post-translational modification Hard Easy Self-replication Hard Easy DNA template Plasmids or PCR products Plasmids or genomes Synthesis of membrane proteins and complex proteins Easy synthesis by adding surfactants or adjusting the system environment Hard synthesis due to limited intracellular environment Incorporation of unnatural amino acids into proteins Easy Hard Ability to only produce the desired products Easy achievement by focusing on the target metabolic pathways Hard achievement due to complicated cellular metabolism Toxic tolerance High Low Integration with materials Easy Hard Design-build-test-learn cycle Two days Two weeks Biomanufacturing High production rate Modest production rate High product yield Modest product yield Easy purification process without cell lysis Cell lysis prior to product purification Cost Modest to high Low to modest These features make cell-free synthetic biology serve as a versatile platform for engineering biological parts at three different levels of protein, metabolism and cell ( Fig. 1 ). Cell-free synthetic biology can be an enabling technology for innovating medical diagnostics and therapeutics, developing complex metabolic system, making functional biomolecules, and producing sustainable bioenergy and biochemicals. Fig. 1 Engineering protein, metabolism and artificial cell in the open cell-free system. Fig. 1"
} | 1,751 |
31680818 | PMC6803503 | pmc | 4,939 | {
"abstract": "Purkinje cell is an important neuron for the cerebellar information processing. In this work, we present an efficient implementation of a cerebellar Purkinje model using the Coordinate Rotation Digital Computer (CORDIC) algorithm and implement it on a Large-Scale Conductance-Based Spiking Neural Networks (LaCSNN) system with cost-efficient multiplier-less methods, which are more suitable for large-scale neural networks. The CORDIC-based Purkinje model has been compared with the original model in terms of the voltage activities, dynamic mechanisms, precision, and hardware resource utilization. The results show that the CORDIC-based Purkinje model can reproduce the same biological activities and dynamical mechanisms as the original model with slight deviation. In the aspect of the hardware implementation, it can use only logic resources, so it provides an efficient way for maximizing the FPGA resource utilization, thereby expanding the scale of neural networks that can be implemented on FPGAs.",
"conclusion": "Conclusion In this work, we present an efficient implementation of a modified cerebellar PC using the CORDIC algorithm with recently found new dynamic performance. Through the analysis of various errors of the two single-neuron models and the comparison of waveforms and network behaviors from different aspects, it can be concluded that the original model and the CORDIC-based model are consistent in biological activities and dynamic mechanisms. After that, we use the non-multiplier and non-LUT methods and implement the CORDIC model on the LaCSNN system. The implementation results are observed on the oscilloscope through the DA conversion module, which are also consistent with the results of the software simulation. By comparing the resource utilization of the original model and the CORDIC model in FPGA implementation, we can conclude that the method used in this paper can transform the use of multiplier resources and memory resources into logical resources, so as to maximize the utilization of FPGA on-chip resources and expand the network scale that can be achieved. This work provides an effective method for realizing large-scale spiking neural networks of cerebellum or many other spiking neural networks on FPGAs.",
"introduction": "Introduction The cerebellum is a very important part of the human brain and associated with many important functions with a large number of incoming and outgoing connections between the brain, brainstem, and spinal cord. These functions are not only relevant to motor control including error correction (Doya, 2000 ; Llinas, 2009 ), tracking movements (Paulin, 1993 ; Miall et al., 2000 ), and coordinated movements (Thach et al., 1992 ; Heck et al., 2007 ) but also relevant to many non-motor functions such as linguistic prediction, word generation, emotional control, and so on (Leiner et al., 1993 ; Schmahmann and Caplan, 2006 ; Pleger and Timmann, 2018 ). Purkinje cells (PCs) make up the middle layer of the cerebellum, Purkinje layer, which is responsible for receiving information from the cerebellar granule cell (GC) synapses through parallel fibers (PF) and climbing fibers (CF) in brainstem. In addition to being all the constituent cells of the cerebellar Purkinje layer, PCs also directly connect to deep cerebellar nuclei cells, which are the main output cells of cerebellum. So, it is obvious that PCs play the most important role in the information processing of the cerebellum. Besides, PCs are responsible for cerebellar motor learning (Gilbert and Thach, 1977 ) with the information stored in the synapses with granule cells. The information is presented as the variation of synaptic strength according to the error signals carried by CFs through spike timing-dependent plasticity (STDP), which consists of long-term potential (LTP) and long-term depression (LTD) (Ito and Kano, 1982 ; Han et al., 2000 ; Medina et al., 2000 ). This learning mechanism can be obviously observed in classical eyeblink conditioning experiments (Bao et al., 2002 ) and cerebellar vestibulo-ocular reflex (VOR) (Blazquez et al., 2003 ; Masuda and Amari, 2008 ), which are mainly caused by the function of PCs. There are two calculation modes for simulation spiking neurons or spiking neural networks, serial computing mode, and the parallel computing mode (Yang S. M. et al., 2019 ). The serial computing mode is mainly based on some computer simulation software that is incompatible with the parallel computing features of real neural systems. In order to achieve these in a more biological way, more and more neuroscientists prefer to implement neurons and neural networks with parallel computing mode. Analog very Large-Scale Integration (VLSI), Graphics Processing Unit (GPU), and Field Programmable Gate Array (FPGA) are the three most used platforms with parallel computing capacity. Analog VLSI is an efficient analog-based method for hardware implementation of spiking neurons and neural networks because it can realize the non-linear function directly (Han, 2005 ; Hsieh and Tang, 2012 ). However, it cannot be flexibly changed once formed, so it is more suitable for well-defined circuits. In addition, its high cost and long development cycle also limit the application range. GPU provides a digital implementation method for spiking neurons and neural networks with its powerful parallel calculation ability and many researches have been carried on GPUs (Igarashi et al., 2011 ; Yamazaki and Igarashi, 2013 ). However, the kernel-launch method used on GPU and the limited bandwidths are obstacles for dealing with a lot of data. Compared to the two methods above, FPGA has many advantages for realizing the neural circuits. On one hand, the flexible reconfigurability and parallel computing architecture can perfectly meet the requirements for exploring characteristics of not only spiking neurons but also the large-scale spiking neural networks; on the other hand, its low area and power consumption also make it popular in neurosciences (Yang et al., 2017 , 2018a ). In this work, the neuron is implemented on the Large-Scale Conductance-Based Spiking Neural Networks (LaCSNN) system first proposed by Yang S. et al. ( 2019 ). The system consists of six Altera EP3SL340 FPGAs and is designed to simulate large-scale spiking neural networks with digital neuromorphic architecture. Its powerful storage capacity, high calculation speed, and sufficient resources make it an effective tool for neuroscience researches. Although the advantages of FPGA are very prominent, the disadvantages are also distinct. Most of the resources on FPGA are logic resources; the lack of memory and multiplier resources often limits the scale when implementing neural networks. As a kind of digital systems, it is difficult to implement the non-linear functions directly. To solve these problems, many methods have been proposed. One of the most frequently used methods is to store the function values in a storage area with continuous address space in advance, which is called look up table (LUT) realization. When used, the function value can be obtained by addressing. This method is very easy but costs much memory resources. Besides, the use of LUTs increases the duration of reconstruction when changing model parameters. Another method, Taylor series approximation, is to replace the non-linear function in the neighborhood with an n -order polynomial approximation for a certain error. This method can make a trade-off between LUT resources and multiplier resources with different approximation order, but it still needs these resources (Lee and Burgess, 2003 ). The piece-wise linear (PWL) approximation (Julian et al., 1999 ) is a more efficient method to solve these problems but there are two main cons: one is there will be unavoidable error due to the use of several linear segments; the other is that it needs to recalculate when the non-linear function changes. So, in this work, we propose a non-multiplier and non-LUT method with the CORDIC algorithm for implementing the cerebellar Purkinje model on FPGA. One of the main reasons for implementing single neurons with optimization algorithms on FPGA is to lay a foundation for realizing large-scale spiking neural networks. Many researches have been carried out in recent years. Yang et al. ( 2018b ) propose a series of techniques for implementing a conductance-based neuron model that is beneficial for building large-scale neural networks. Soleimani et al. ( 2012 ) implement a classic Izhikevich model using PWL method to prove that the method can simplify the hardware implementation with showing similar dynamic behaviors. Ambroise et al. ( 2013 ) also implement an Izhikevich model on FPGA, but it is mainly to propose an architecture to reproduce a neural network with only one computation core (one neuron) based on one multiplier. Bonabi et al. ( 2012 ) implement a Hodgkin–Huxley (H–H) single neuron with the CORDIC algorithm and some LUTs that show high precision with more compact used logic. There are also many researches about implementing the CORDIC algorithm on FPGA. Valls et al. ( 2002 ) evaluate some methods for the CORDIC algorithm and realize a variable precision method using conventional arithmetic on FPGA. Liu et al. ( 2014 ) implement a modified CORDIC algorithm that reduces the utilization of ROM resources and power consumption. Garcia et al. ( 2006 ) realize a pipelined CORDIC architecture with solution for overflow and quadrant correction and successfully generating sine and cosine waves. Muñoz et al. ( 2010 ) propose a floating-point CORDIC FPGA implementation for calculating transcendental functions. The FPGA implementation of the CORDIC algorithm can give full play to the advantages of FPGA and utilize hardware resources to realize an optimization scheme combining hardware and algorithm. The pipelined computational structure of FPGA can also enhance the real-time performance of the CORDIC algorithm, minimizing the computational delay due to the iterative operations. Therefore, the CORDIC algorithm can be widely applied to real-time high-quality signal processing with high-performance requirements. The remaining parts of this work are arranged as follows. In section Neuron Model, the original model and modified CORDIC model of cerebellar PC are presented. The CORDIC algorithm used is also introduced in this section. Section Hardware Implementation Based on LaCSNN describes the details of hardware implementation. The results of software simulation and hardware simulation are shown in section Results. We also compare and analyze the result between the original model and the CORDIC model with various evaluation indicators for both the two simulations. The behaviors of a network with this neuron are also presented. section Discussion illustrates the discussion and conclusion for this work.",
"discussion": "Discussion There is a bottleneck for realizing a large-scale neural network with high biological precision neurons such as the model in this paper based on the H–H neuron model. These models have many conductance-based ionic currents that usually contain many non-linear functions and greatly increase the computational complexity. To solve this problem, many previous studies are working on FPGA resource optimization for spiking neurons with different methods (Ahmadi and Zwolinski, 2010 ; Bonabi et al., 2014 ; Hayati et al., 2016 ; Akbarzadeh-Sherbaf et al., 2018 ). Ahmadi and Zwolinski ( 2010 ) propose a method with PWL approximation for implementing the Izhikevich model. The non-linear operations in the model are only multiplications for there are no detailed ionic currents. The model complexity is relatively simple so the reference meaning for building high biological precision neurons is limited. Bonabi et al. ( 2014 ) implement an H–H-based model and a two-mini-column network with the CORDIC algorithm but it is only used for calculating exponent operations, but there are still some things to do to implement a large-scale neural network, because the multiplication and division operations account for a large proportion of the model and they still need multipliers and memory resources. Besides, there is no simplification for the iterative structure as we have done. Akbarzadeh-Sherbaf et al. ( 2018 ) use a general PWL approach to implement a randomly connected network with H–H models. If we just focus on one H–H model, the PWL approach can successfully realize the non-linear functions and improve the working frequency, but the precision is lower than the CORDIC algorithm for a sharp curve will certainly appear at the junction of the two linear sections. Besides, the approximate range of each linear part is only applicable to that set by the designer, so the linearization must be redesigned each time the model changes, and any unexpected values may get unexpected behaviors. As for the GPU platform, there may not be many researches on implementing a single neuron on it, but many researches have been carried on for the comparison between GPUs and FPGAs about implementing spiking neural networks (Cheung et al., 2012 , 2016 ; Luo et al., 2016 ). The results show that GPUs can speed up the simulations with multi-core processors and parallel computing capacity, but compared to FPGA, two obvious cons still exist. One is the small on-chip memory and bandwidth, which limits the scale, the other is the high-power consumption of the desktop system. Besides, the calculation speed of GPUs is also lower than FPGAs in these works. In order to save multiplier resources on FPGA, many multiplier-less methods have been proposed with different application ranges. Both Jokar and Soleimani ( 2017 ) and Hayati et al. ( 2016 ) propose a multiplier-less structure with the PWL approach that needs to linearize each function that contains multiplication of variables. The multiplier-less implementation in Agostini et al. ( 2005 ) and Gomar and Ahmadi ( 2014 ) are simple for there are all constant number multiplications in their models, which can be easily replaced by adders and shifters. Thomas and Luk ( 2013 ) replace the multipliers with LUTs and block RAMs, which use more LUT resources to save multiplier resources. Our work presents an FSM, which is common to all multiplication operations and easy to use. With this method, users do not need to redesign the whole approximation using the PWL approach, and all of the multiplications can be realized just by adjusting the supported bit width, even simpler than the method implementing the constant number multiplications. The working frequency of the FSM is 195.92 MHz as shown in Table 5 , so the lower working frequency of the cell model compared to the model mentioned above is only due to the unavoidable iterative structure of the CORDIC algorithm and the complexity of this model. This paper presents a multiplier-less and LUT-less CORDIC method to realize the conductance-based cerebellar Purkinje model on FPGA. This can be used for the trade-off among logic resources, memory resources, and multiplier resources, which can be adopted to make full use of the FPGA resources to build a large-scale neural network. All of the calculation modules in our work, the FSM, CDI, and ECEXP, can be directly used for any other models without any extra operation. Besides, the modified pipelined parallel CORDIC algorithm can significantly reduce the resource consumption and the complexity of the hardware implementation architecture."
} | 3,874 |
24288334 | null | s2 | 4,940 | {
"abstract": "In photosynthetic organisms, photons are captured by light-harvesting antenna complexes, and energy is transferred to reaction centers where photochemical reactions take place. We describe here the isolation and characterization of a fully functional megacomplex composed of a phycobilisome antenna complex and photosystems I and II from the cyanobacterium Synechocystis PCC 6803. A combination of in vivo protein cross-linking, mass spectrometry, and time-resolved spectroscopy indicates that the megacomplex is organized to facilitate energy transfer but not intercomplex electron transfer, which requires diffusible intermediates and the cytochrome b6f complex. The organization provides a basis for understanding how phycobilisomes transfer excitation energy to reaction centers and how the energy balance of two photosystems is achieved, allowing the organism to adapt to varying ecophysiological conditions."
} | 228 |
38313259 | PMC10836074 | pmc | 4,943 | {
"abstract": "Microbiomes are generally characterized by high diversity of coexisting microbial species and strains that remains stable within a broad range of conditions. However, under fixed conditions, microbial ecology conforms with the exclusion principle under which two populations competing for the same resource within the same niche cannot coexist because the less fit population inevitably goes extinct. To explore the conditions for stabilization of microbial diversity, we developed a simple mathematical model consisting of two competing populations that could exchange a single gene allele via horizontal gene transfer (HGT). We found that, although in a fixed environment, with unbiased HGT, the system obeyed the exclusion principle, in an oscillating environment, within large regions of the phase space bounded by the rates of reproduction and HGT, the two populations coexist. Moreover, depending on the parameter combination, all three major types of symbiosis obtained, namely, pure competition, host-parasite relationship and mutualism. In each of these regimes, certain parameter combinations provided for synergy, that is, a greater total abundance of both populations compared to the abundance of the winning population in the fixed environments. These findings show that basic phenomena that are universal in microbial communities, environmental variation and HGT, provide for stabilization of microbial diversity and ecological complexity.",
"discussion": "DISCUSSION In this work, we developed a mathematical model to explore the role of HGT in the interaction between two cohabiting populations in an oscillating vs a fixed environment. We explored what seems to be the simplest possible model in which two populations differed by a single gene allele, such that each entity in the pool carried only one allele of the given gene that is subject to HGT through copying the gene in the donor cell, and then, substituting the existing gene of the recipient cell, which corresponds to homologous recombination [ 14 , 18 ]. Notwithstanding the ultimate simplicity of this scheme, it is biologically realistic, for example, in the context of acquisition of multi-drug-resistance [ 64 ]. We start with the stochastic description of all possible interactions between two populations, namely, reproduction and death of each population, competition within and between populations, and gene transfer between the populations. Then, we take the continuous limit and obtain the deterministic description of the time variations of the size of each population (see Additional File1 , [ 53 , 54 ]). The dynamical system obtained by this procedure is a variation of the well-known replicator dynamics [ 55 – 59 ], with varying total abundance of both populations. The composition and the total abundance of both populations in the equilibrium states are found from the rest points of this system. By examining this mathematical model, we show that stable coexistence of the two populations in a fixed environment is unattainable within the assumption that the gene allele replacement via HGT only affects the reproduction rate, the competing abilities of both populations are the same within and between populations, and HGT between the populations is balanced, that is, there is no preferred direction in the gene flow between the populations. These results reflect the classic competitive exclusion (Gauze) principle according to which, in a spatially homogeneous environment, a population with an inferior reproduction rate will inevitably go extinct when competing with a fitter population for the same resource, within the same niche [ 60 – 62 ]. Notable, however, if the HGT is unbalanced, that is, the rates of gene transfer in the two directions are unequal, the exclusion principle does not apply anymore, and the two populations can coexist. Evolutionary outcomes in a time-varying environment can drastically differ from those in the fixed environment, and despite the simplicity of our model, the emerging dynamics is complex, with the outcomes critically depending on the parameter combination. Environmental oscillations are incorporated into the model by assuming that the reproduction and gene transfer rates are time-periodic functions, with the implicit assumption that these rates are affected by the changes in the environment. The oscillations were set up such that the averages across environmental variations coincided with the constant rates in the fixed environment, providing for a fair comparison of the dynamics in the two types of environment. The environmental oscillations occur on a fast time scale, whereby the populations are exposed to numerous changes in the environment before approaching any potential equilibrium. We adopted a coarse-graining approach to account for the environmental variations whereby the solution for the composition and total abundance of both populations was obtained as a combination of two components, one delineating the slow-time (coarse grained) behavior and the other capturing the oscillating component [ 44 , 45 ]. By averaging over the period of environmental variations and retaining the first two leading-order terms of the varying quantities, we obtained the replicator dynamics model with varying total abundance that features fitness contributions derived from two distinct games. The first game represents the fixed/averaged environment whereas the second one arises due to environmental fluctuations, with the payoffs in this game determined by the intersection of the oscillating components of the reproduction and gene transfer equilibrium rates. These new payoffs are such that the fitness of a particular population is defined by the average of the product of the competitor’s reproduction rate and the gene transfer equilibrium rate between the populations throughout environmental oscillations. For the emergence of the new game, reproduction and gene transfer balance rates must be differentially affected by the environment. Notably, variations either in the reproduction rates or in the gene transfer rate alone did not result in new evolutionary scenarios compared to the fixed environment case. The emerged game maintains a straightforward payoff structure: once the environmental variations of reproduction and gene transfer rates are given, then either one of the strategies consistently dominates, or there exists a unique non-trivial equilibrium for any given total abundance of both populations. With these adjusted fitness terms, stable coexistence between the two populations becomes possible within a broad range of model parameters. Moreover, depending on the combination of the time-dependent reproduction and gene transfer balance rates, the co-existence of the two populations manifests as all major types of symbiotic relationships. Two populations can coexist in a purely competitive symbiosis, where the fitness of each is negatively impacted by the presence of the other population; a host-parasite relationship, whereby the impact of the second population is positive for one and negative for the other population; and in mutualistic symbiosis, where the presence of the other population is reciprocally beneficial. We further analyzed the behavior of the total abundance of both populations and defined the regions of synergistic interactions between the competing populations. In this case, synergy is observed, that is, the total abundance of the two populations at the stable coexistence in the oscillating environment is greater than the total abundance in the fixed environment, that is, the population size of the winner at equilibrium, under the same model parameter values. As could be expected, mutualism necessarily entails synergy and yields the greatest total abundance of both populations among all the regimes by increasing the resource utilization in the environment. Notably, however, under certain combinations of the reproduction and HGT rates, the synergistic effect was observed also in the cases of parasitism and even pure competitive symbiosis. Outside of the coexistence regimes, the outcomes of the competition between two populations in a stable environment can persist in the oscillating environment such that the winner in the fixed environment wins in the oscillating environment as well. However, in another region of the phase space, the outcome can be reversed due to environmental oscillations. Moreover, there is also a bistability regime where one or the other population goes extinct depending on the initial conditions, a scenario precluded in the fixed environment assuming equal competing abilities and balanced gene transfer. Despite our prior discussion on coexistence and potential outcomes in an oscillating environment with balanced gene transfer in the fixed environment, the introduction of oscillations can notably alter the composition at the coexistence equilibrium, even if both populations coexist in the fixed environment due to non-zero gene transfer balance. Environmental variations, in this case, can even destroy the stable coexistence of the two populations, attained in the fixed environment with unbalanced HGT. From the methods point of view, it is worth pointing out that coarse graining was an essential ingredient of the present analysis because without it, the symbiotic relationships between two populations would be impossible to elucidate because the fitness gradients can change their signs throughout environmental oscillation due to the oscillating rates. From the biological standpoint, environmental oscillations provide for the synergy between the two populations by lowering the intensity of the inter-population competition below the level of the intra-population competition. This work shows that stabilization of strain diversity and increase of ecological complexity via HGT in an oscillating environment is an intrinsic feature of even the simplest microbiomes that emerges under minimal assumptions on the basic processes occurring within a microbial community. Given that both environmental variation and HGT are ubiquitous phenomena that affect any microbiome [ 65 , 66 ], these conclusions appear to be broadly applicable. Notably, all emerging coexistence regimes provide for synergy between the populations indicating that HGT in an oscillating environment is favorable for a microbial community perceived as an integral whole. Evidently, the model used in this work is grossly (and deliberately) over-simplified. Real microbiomes encompass interactions among thousands of microbial strains and species, HGT of multiple genes via different routes and many other processes [ 14 , 50 , 66 ]. Nevertheless, the general principle established here should apply, amplified by the microbiome complexity. Characterization of the conditions for stabilization of microbiome diversity and the factors that can perturb it is crucial for understanding the role of the microbiome in health and diseases as well as the ecology of microbial communities."
} | 2,753 |
26285202 | PMC4540456 | pmc | 4,944 | {
"abstract": "Taxonomic marker gene studies, such as the 16S rRNA gene, have been used to successfully explore microbial diversity in a variety of marine, terrestrial, and host environments. For some of these environments long term sampling programs are beginning to build a historical record of microbial community structure. Although these 16S rRNA gene datasets do not intrinsically provide information on microbial metabolism or ecosystem function, this information can be developed by identifying metabolisms associated with related, phenotyped strains. Here we introduce the concept of metabolic inference; the systematic prediction of metabolism from phylogeny, and describe a complete pipeline for predicting the metabolic pathways likely to be found in a collection of 16S rRNA gene phylotypes. This framework includes a mechanism for assigning confidence to each metabolic inference that is based on a novel method for evaluating genomic plasticity. We applied this framework to 16S rRNA gene libraries from the West Antarctic Peninsula marine environment, including surface and deep summer samples and surface winter samples. Using statistical methods commonly applied to community ecology data we found that metabolic structure differed between summer surface and winter and deep samples, comparable to an analysis of community structure by 16S rRNA gene phylotypes. While taxonomic variance between samples was primarily driven by low abundance taxa, metabolic variance was attributable to both high and low abundance pathways. This suggests that clades with a high degree of functional redundancy can occupy distinct adjacent niches. Overall our findings demonstrate that inferred metabolism can be used in place of taxonomy to describe the structure of microbial communities. Coupling metabolic inference with targeted metagenomics and an improved collection of completed genomes could be a powerful way to analyze microbial communities in a high-throughput manner that provides direct access to metabolic and ecosystem function.",
"conclusion": "Conclusions We’ve described a framework for inferring microbial metabolic structure, as defined by the abundance of metabolic pathways, from 16S rRNA gene data using a phylogenetic placement approach [ 12 ] and the MetaCyc ontology [ 13 ]. This metabolic inference framework is complementary to metagenomic analysis, and should be paired with genome sequencing and metagenomics to reasonably constrain the metabolisms present in any environment. Although metagenomics can provide a less biased profile of potential metabolism, insufficient metagenomic sequencing depth or highly variable coverage will make for an incomplete metabolic profile [ 4 ]. In addition we argue that it is neither necessary nor desirable to produce detailed, high quality metagenomes in a high-throughput fashion (i.e. on hundreds or thousands of samples required for a single ecological study). Even with the limited number of completed genomes currently available, however, it is possible to infer metabolism to a level that closely reproduces patterns observed by 16S rRNA gene comparison. Furthermore, by converting 16S rRNA gene community data directly to metabolic structure data, differences between samples can be evaluated in the context of changes to the ecosystem function of the microbial community. Although our use of complete pathways to describe metabolism is necessarily conservative, as a metabolism cannot be predicted unless a complete pathway has been described, presentation of the data in this fashion makes it most compatible with other ‘omics analyses. Gene expression and metabolite concentrations, for example, can be readily mapped to the PGDBs by available tools [ 22 ]. As with other metabolic inference techniques, the framework introduced here is only as good as the collection of completed genomes available in the public repositories and our knowledge of gene function. Although draft genome assemblies from metagenomes and single cell sequencing efforts have allowed investigators to access much of the metabolic potential of (as yet) unculturable marine microbes, few investigators take the time to complete the difficult task of closing the draft genomes produced by these analyses. Although our framework could be easily modified to include draft genomes, which would result in improved taxonomic resolution, the resulting inference would be less informative than if the genomes were complete. This is due to the low quality scores such an analysis would produce, since genomic plasticity would be artificially elevated and core genome size would be artificially small. Second, geographical variations in the proportion of microbial dark matter (the functionally uncharacterized portion of the microbial assemblage) [ 43 ] and in the degree of genomic plasticity require validation with thorough, high quality metagenomes. Differences between a metabolic inference and metagenomics analysis would highlight clades that are not well represented by sequenced genomes, or that are exceptionally plastic (and thus require a high degree of taxonomic resolution). These clades could then be targeted for isolation, genome sequencing, and phenotyping, or for genome assembly from a deep, targeted metagenome.",
"introduction": "Introduction Biological communities are structured by a variety of physical, chemical, and ecological environmental factors. For the marine microbial community, these include the availability of dissolved organic carbon (DOC), the distribution of bioavailable nitrogen and phosphorous, light, and temperature, among numerous other biological, chemical, and physical factors. Although microbial community structure is often described in terms of taxonomy, with clear correlations between the taxonomic composition of various microbial communities and different environmental settings [ 1 , 2 ], these environmental conditions are more directly linked to metabolic structure. The cyanobacterial genus Trichodesmium , for example, is associated with the low concentration of bioavailable nitrogen in the tropical and subtropical oceans. It is the metabolic properties (e.g. diazotrophy) of the genus and not its taxonomy, however, that afford a direct link with environmental conditions. To understand the role of microbial communities in biogeochemical processes it is preferable to consider the metabolic structure of a community over its taxonomic structure. The correlation between taxonomy and metabolic function [ 3 , 4 ] is the basis for the considerable body of work focused on the identification of community structure and composition through taxonomic marker gene analysis, namely the 16S rRNA gene. Although sometimes criticized as “stamp collecting” [ 5 ], these marker gene studies have enabled microbial ecologists to identify complex patterns of microbial diversity in a large number of geographic locations, and under widely varying environmental conditions. In contrast to the ease with which large 16S rRNA gene libraries can be generated however, it is not practical for a team of investigators to manually and exhaustively explore the metabolisms known to associate with all the observed operational taxonomic units (OTUs). More recently metagenomics studies, which profile not just a marker gene but all genes within a microbial community, have made it possible to explore total metabolic potential—though this type of analysis introduces another set of challenges, including cost, dataset size and throughput, and, in some cases the poor knowledge of gene function. In a process analogous to the cellular organization of metazoans and plants, microbial communities partition metabolites and metabolic transformations within individuals, though this compartmentalization is imperfect. Although in some cases it is possible to reconstruct partial [ 6 ] or even full genomes [ 7 ] from metagenomics datasets, thus reproducing cellular partitions, this datatype does not lend itself to high confidence metabolic reconstruction; for example it is difficult to exclude possible chimeric metabolic pathways unless an investigator is very conservative and the assembly particularly robust. Furthermore, despite a precipitous decrease in the cost of sequencing since the first application of so-called next generation sequencing (NGS) to environmental samples in 2006 [ 8 ], it remains prohibitively expensive for most investigators to produce sufficiently deep, high-quality metagenomes on tens, let alone hundreds, of samples. As a result of these limitations new tools and a new conceptual framework are needed to bridge the gap between marker gene studies, which are economical but indirectly linked to community function, and potential metabolism, which is costly and labor intensive to analyze but is directly linked to function. Several recent studies have begun to make progress in this direction. Okuda et al. [ 9 ] undertook the construction of artificial metagenomes from collections of 16S rRNA sequences, finding a high degree of similarity between a reconstructed and actual metagenome. More recently Langille et al. [ 4 ] introduced an open-source tool, PICRUSt, to infer functional gene and pathway profiles from 16S rRNA gene data. In an extensive evaluation PICRUSt based predictions agreed well with the results of metagenomic analyses for diverse environments, including the human microbiome [ 4 ]. Together these studies suggest that metabolic inference from 16S rRNA gene libraries has a sound theoretical basis. Despite the demonstrated success of this method there are limitations to how well microbial metabolism can be inferred from taxonomy. First, the connection between inferred metabolism and taxonomy is only as good as the collection of published sequenced genomes. At the time of analysis over 2,700 finished genomes were available on Genbank, along with more than 20,000 draft genomes. For strains that are very closely related to these genomes a reasonable inference of metabolism can be made, though incomplete, ambiguous, or incorrect annotations will lead to inaccuracies. Second, genomic plasticity is known to be high for many clades of the Bacteria and Archaea. Genetic composition, and thus metabolic potential, can vary widely between strains with a nearly identical 16S rRNA gene sequence [ 10 ]. Third, phenotypic plasticity means that even clonal strains encountering similar environmental conditions will drift in their specific metabolic response [ 11 ]. Here we present a novel framework for inferring complete metabolic pathways from 16S rRNA gene sequence collections, based on a phylogenetic placement approach [ 12 ] and the MetaCyc pathway ontology [ 13 ]. Our framework includes a method for quantifying genomic plasticity, which we apply to all finished genomes in the Genbank collection. We evaluated our metabolic inference methods against a marine metagenome, and applied both PICRUSt and our framework to 16S rRNA gene sequence libraries from the coastal West Antarctic Peninsula (WAP), inferring metabolic pathways in 18 samples separated by location, depth, and season. Both metabolic and taxonomic structure showed clear separation by sample type, suggesting that the abundance of metabolic pathways can be used in place of the abundance of OTUs to describe inter-sample relationships. This approach has the additional advantage of identifying the metabolisms most relevant to microbial community function in each environment. Although it is beyond the scope of the current work, our framework allows for a further analysis of community function by metabolic flux analysis [ 14 ]. This modeling of chemical conversions within and between cellular spaces has the potential to resolve connections between the composition of a microbial community and its ecosystem functions.",
"discussion": "Results and Discussion Genetic plasticity The comparison of 16S rRNA gene and compositional vector distance revealed distinct patterns of genomic plasticity across the analyzed genomes ( Fig 1 ). The most unstable genomes in our analysis belonged to a clade of insect symbionts that include Candidatus Tremblaya princeps PCVAL, Oxalobacteraceae symbiotic bacterium CARI, Candidatus Nasuia deltocephalinicola NAS ALF, and Candidatus Carsonella ruddii . This high degree of instability is consistent with expectations for obligate symbionts, which often have highly streamlined genomes [ 30 ] and high rates of horizontal gene transfer [ 31 ]. Other regions of high genomic plasticity were also dominated by symbionts, including Candidatus Hodgkinia cicadicola DSEM, Candidatus Sulcia muelleri , Candidatus Portiera aleyrodidarum , Buchnera aphidicola , and Nanoarchaeum equitans . The Mycobacteria and Mycoplasma also had exceptional levels of genomic plasticity. 10.1371/journal.pone.0135868.g001 Fig 1 Relative genomic plasticity of the genomes used for metabolic inference. Colors correspond to the level of plasticity, which is also given by the y-axis. X-axis is labeled according to terminal node order for the representative 16S rRNA gene from each genome on a phylogenetic tree rooted at the node ancestral to the Archaea and Planctomyces ( S1 File ). Note that terminal node order on this rooted tree does not correspond to edge numbers on the unrooted tree used for phylogenetic placement ( S2 File ). Arrows indicate strains or clades with exceptional plasticity, the majority of which are known symbionts. I) Nanoarcheum equitans II) the Mycobacteria III) a butyrate producing bacterium within the Clostridium IV) Candidatus Hodgkinia circadicola V) the Mycoplasma VI) Sulcia muelleri VII) Portiera aleyrodidanum VIII) Buchnera aphidicola , IX) the Oxalobacteraceae . Comparison with metagenome Our metabolic inference method predicted 891 metabolic pathways for the South Orkney Islands metagenome. A lesser number of pathways were identified directly from functional genes represented in the metagenome; 612 and 690 for the partitioned and non-partitioned methods respectively. Most of the 251 pathways predicted by metabolic inference, but not identified in the metagenome, were associated with low abundance taxa ( Fig 2 ) suggesting that their absence may be the result of incomplete coverage. Several of these pathways however, had an abundance greater than 4,500 (14.4% of 16S rRNA read abundance), including the pentose phosphate pathway (oxidative branch) II, indole-3-acetate biosynthesis V (bacteria and fungi), lysine degradation I, and cysteine biosynthesis/homocysteine degradation. The greater abundance of these pathways in the metabolic inference does not necessarily preclude incomplete coverage of the metagenome as a reason for their absence in that dataset, alternatively, they may represent pathways inferred for but not present within the sampled community. 10.1371/journal.pone.0135868.g002 Fig 2 Comparison of metabolic inference and metagenomic analysis. A) Metabolic pathways identified (metagenomes) or inferred (PICRUSt, this method) with each method. PICRUSt pathways (indicated by white bar) are based on the KEGG ontology, thus the number of pathways inferred is not directly comparable to the other methods which are based on the MetaCyc ontology. B) The abundance of pathways inferred for but not identified in the metagenome. Metabolic inference with PICRUSt yielded 328 metabolic pathways for the metagenome, however, it is important to recognize that our PICRUSt analysis relied on the Kyoto Encyclopedia of Genes and Genomes (KEGG) ontology [ 32 ] which is not directly comparable to the MetaCyc ontology [ 13 ], nor are we aware of any method for converting between ontologies. Thus the lower number of predicted pathways for PICRUSt is not a reflection of lower sensitivity, but it may reflect a lesser degree of specificity in the final result. Application to marine samples from the WAP To test our metabolic inference method on 16S rRNA gene amplicon datasets we selected a sample set obtained from the Palmer Long Term Ecological Research site (PAL-LTER) off the WAP [ 19 ] ( Fig 3 ). Using an OTU based approach on these samples, Luria et al. [ 19 ] had previously found that samples from like environments grouped together by hierarchical clustering and nonmetric multidimensional scaling (NMDS), although there was significant variation between samples along all gradients (north vs. south, inshore vs. offshore, surface vs. deep). Inshore surface samples, which also host large populations of eukaryotic Cryptomonads [ 19 ], showed the largest difference from other sample types. Winter surface samples had some similarity to deep summer samples, supporting the hypothesis that remnant “winter water” drives community structure at depth [ 19 ]. 10.1371/journal.pone.0135868.g003 Fig 3 Sample locations in the West Antarctic Peninsula (WAP). Summertime surface (10 m) and deep (100 m) samples are analyzed from two inshore and two offshore samples, organized along a North to South gradient. Winter surface water samples were analyzed from the northern, inshore station (NE). Our phylogenetic placement approach to evaluating community structure broadly supports the findings of Luria et al. [ 19 ], however, we observed a clearer separation between winter and deep samples and summer surface samples ( Fig 4 ), suggesting that a phylogenetic placement approach can improve the sensitivity of β-diversity analysis. To identify what edges were most responsible for variation between samples we applied PCA to our edge abundance data ( Fig 5 ). Variance was primarily driven by low-abundance taxa; for the top 30 edges ranked by magnitude in PC1 and PC2 the most abundant edge in any one sample was edge 2216 ( Caldisericum exile AZM16c01), at 76 occurrences in summer_sw_deep_b.1 (5.7% of the total population). Much of the variance between samples was driven by rare edges occurring more often in winter and deep samples than summer surface samples, including 2216 ( Caldisericum exile AZM16c01), 1917 ( Bifidobacterium ), 313 ( Nitrosospira multiformis ATCC25196), and 3069 ( Maricaulis maris MCS10). Consistent with this we found the winter and deep samples to have greater predicted edge richness by Chao1 [ 33 ] as determined with a Mann-Whitney test (Chao1: mean summer surface = 118.9, mean winter and deep = 197.3, w = 13, p = 0.016). This increase in richness is consistent with other studies that observed greater richness in high latitude winter samples [ 19 , 34 , 35 ]. Overall the most abundant edges across all samples were 3324 ( Candidatus Pelagibacter ubique HTCC1062), which comprised up to 22.3% of the total community (mean = 15.8%, sd = 4.0%), and 268 (an ancestral node to the Gammaproteobacterial clade Franscella ), which comprised up to 15.6% of the total community (mean = 11.6%, sd = 3.0%). 10.1371/journal.pone.0135868.g004 Fig 4 Hierarchical clustering of samples by pathway and edge abundance. Clustering used the Ward algorithm on Bray-Curtis distance. A) Hierarchical clustering by pathway after normalization to the maximum abundance of each pathway. B) Hierarchical clustering by edge abundance after normalization to the total abundance of each sample. C) Linear model (red line) for distance by edge abundance as a function of distance by pathway abundance, R 2 = 0.65, df = 75, p ≈ 0. D) Linear model (red line) for distance by OTU abundance as a function of distance by pathway abundance, as predicted using PICRUSt (see text) [ 4 ]. 10.1371/journal.pone.0135868.g005 Fig 5 Metabolic pathways and edges accounting for the most variance between samples. A) PCA of normalized metabolic pathway abundance (see text). Arrows are vectors for the top twenty pathways ordered by their magnitude in PC1 and PC2. B) Heatmap of abundance for the top twenty pathways ordered by their magnitude in PC1 and PC2. Hierarchical clustering on the y-axis uses the defaults of the heatmap command in R; Euclidean distance and the complete linkage clustering method. C) PCA of normalized edge abundance (see text). Arrows are vectors of the top twenty edges ordered as for A. D) Heatmap of abundance for the top twenty edges ordered as for B. Hierarchical clustering on the y-axis is as for B. Values on the color bar are raw (not normalized) values. Clustering by pathway abundance agreed strongly with clustering by edge abundance, with the pairwise distance between all sample strongly correlated (R 2 = 0.70, p ≈ 0) and only minor differences in cluster membership ( Fig 4A–4C ). The correlation between samples by PICRUSt metabolic inference was weaker than with our method, but was still highly significant (R 2 = 0.40, p ≈ 0) ( Fig 4D ). Unlike in the description of community structure by edge abundance however, the pathways accounting for the most variance between samples were not necessarily low in abundance, with high variance pathways falling into high, middle, and low abundance groups ( Fig 5 ). To further explore the implications of metabolic pathway distribution on microbial ecosystem function we considered pathways involved in the degradation of carbon substrates and in energy acquisition, evaluating the normalized abundance of these pathways as a standard anomaly (the difference between summer surface and winter and deep samples divided by the normalized abundance in both groups) ( Fig 6 ). Some pathways involved in the degradation of plausible DOC components were differentially present between groups, including pathways for the degradation of phenylacetate, ethanol, oxalate, amino acids, and nucleosides and their derivatives. A differential distribution of degradative pathways is consistent with the structuring of microbial communities around the composition of the DOC pool, a concept that has been explored in numerous studies [ 36 – 38 ]. Interestingly, two nitrate reduction pathways were present at a greater normalized abundance in the summer surface samples. Anaerobic environments are known to form in particles and aggregates in the otherwise oxic photic zone [ 39 ]. In the WAP these processes could be linked to anaerobic microenvironments that form during periods of high respiration and suggest an additional sink for nitrate. 10.1371/journal.pone.0135868.g006 Fig 6 Metabolic pathways differentially present between summer surface samples and winter and deep samples. Color gives the p-value for a Mann-Whitney test between sample groups. X-axis gives the anomaly, calculated as the difference in sample group means divided by the sum of the sample group means. Metabolic pathway abundance and Choa1 richness were not significantly different between summer surface and winter and deep samples by the Mann-Whitney test at the 95% confidence level (abundance: mean summer surface = 184,852, mean winter and deep = 208,951, w = 40, p = 1; Chao1: mean summer surface = 640.29, mean winter and deep = 674.57, w = 20, p = 0.08). Sample quality scores were significantly higher for the summer surface than winter and deep samples (mean summer surface = 0.61, mean winter and deep = 0.56, w = 67, p = 0.02), suggesting that the observed disparity in taxonomic and metabolic diversity between seasons and depths could be influenced by lower quality metabolic inferences for the winter and deep samples. This finding makes sense in the context of historical sampling; there is a sampling bias toward the surface ocean and, for high latitude waters, the summer season. Genomes from Bacteria and Archaea that are specialists below the photic zone and in dark winter surface waters are underrepresented in culture collections and in the database of completed genomes. Functional redundancy The region of the WAP is undergoing rapid climatic and ecological change [ 40 ]. Although it is not clear what impact this change will have on Bacterial and Archaeal communities, or even how persistent these communities are from year to year in the absence of rapid environmental change, changes to phytoplankton community structure suggest that microbial communities are prone to shift with the changing environment [ 41 ]. Of particular interest in our analysis are pathways that are present in very few strains, suggesting low functional redundancy. These pathways—and their corresponding ecosystem functions—could be lost, at least temporarily, if environmental conditions change rapidly. Examples of such rapid environmental change include deep mixing events, rapid glacial outflow, and the rapid retreat or advance of sea ice. A recent example of the rapid loss of a microbial ecosystem function was described by Steinle et al. [ 42 ] over the Arctic continental shelf. There, a rapid change in the structure of the water column temporarily eliminated a methanotrophic community associated with seafloor methane seeps. After 11 days, and despite the fact that the methanotrophic niche remained open, that particular function had not been fully restored to the ecosystem. To identify pathways in these samples that might have high ecological importance but low functional redundancy we considered abundant pathways that were present in very few genomes (edges) in any one sample ( Fig 7 ). Although pathway abundance is an imperfect proxy of environmental significance as reaction rates, expression rates, and the concentration of substrates and products required to achieve significance are all independent of abundance, abundance is an indicator of potential significance. Low redundancy, high abundance pathways suggest metabolisms that could connect to major ecosystem shifts. 10.1371/journal.pone.0135868.g007 Fig 7 Pathway abundance as a function of the number of edges pathway appears in for a given sample. Inset is a histogram for abundance at the edge richness of 1 (i.e. the abundance of pathways with the lowest redundancy). Abundance and edge richness are linearly correlated (red line; R 2 = 0.78, df = 11,623, p ≈ 0). Across all samples 140 pathways were represented by only one edge. The fewer edges a pathway was predicted in the higher the likelihood that the pathway was predicted incorrectly, thus some of these nonredundant pathways are erroneous and not expected in the domain Bacteria. One example is the glycerol-3-phosphate shuttle, a mitochondrial energy carrier that is not expected outside the Eurkaryota, but that appears as a nonredundant pathway in 10 of 18 samples due to a prediction for edge 3324 ( Candidatus Pelagibacter ubique HTCC1062). Although the logic for the prediction is sound; the annotation for the Candidatus Pelagibacter ubique HTCC1062 genome includes proteins with the BRENDA numbers EC 1.1.5.3 and EC 1.1.1.8, the requisite enzymes for this pathway, within the Bacteria these enzymes function in different but biochemically related processes. This overprediction highlights one of the current limitations of the method. Methodological improvements, including improved pathway specificity, will be priorities in future work. Aside from these probable erroneous classifications, we identified several nonredundant pathways of potential ecological significance ( Table 1 ). These included glycine betaine degradation (edge 3324, Candidatus Pelagibacter ubique HTCC1062), isopenicillin N biosynthesis (edge 268, Francisella spp.), acrylonitrile degradation (edge 201, Teredinibacter turnerae T7901), and nitrate reduction IV (edge 1242, Syntrophomonas wolfei Goettingen). Conclusions We’ve described a framework for inferring microbial metabolic structure, as defined by the abundance of metabolic pathways, from 16S rRNA gene data using a phylogenetic placement approach [ 12 ] and the MetaCyc ontology [ 13 ]. This metabolic inference framework is complementary to metagenomic analysis, and should be paired with genome sequencing and metagenomics to reasonably constrain the metabolisms present in any environment. Although metagenomics can provide a less biased profile of potential metabolism, insufficient metagenomic sequencing depth or highly variable coverage will make for an incomplete metabolic profile [ 4 ]. In addition we argue that it is neither necessary nor desirable to produce detailed, high quality metagenomes in a high-throughput fashion (i.e. on hundreds or thousands of samples required for a single ecological study). Even with the limited number of completed genomes currently available, however, it is possible to infer metabolism to a level that closely reproduces patterns observed by 16S rRNA gene comparison. Furthermore, by converting 16S rRNA gene community data directly to metabolic structure data, differences between samples can be evaluated in the context of changes to the ecosystem function of the microbial community. Although our use of complete pathways to describe metabolism is necessarily conservative, as a metabolism cannot be predicted unless a complete pathway has been described, presentation of the data in this fashion makes it most compatible with other ‘omics analyses. Gene expression and metabolite concentrations, for example, can be readily mapped to the PGDBs by available tools [ 22 ]. As with other metabolic inference techniques, the framework introduced here is only as good as the collection of completed genomes available in the public repositories and our knowledge of gene function. Although draft genome assemblies from metagenomes and single cell sequencing efforts have allowed investigators to access much of the metabolic potential of (as yet) unculturable marine microbes, few investigators take the time to complete the difficult task of closing the draft genomes produced by these analyses. Although our framework could be easily modified to include draft genomes, which would result in improved taxonomic resolution, the resulting inference would be less informative than if the genomes were complete. This is due to the low quality scores such an analysis would produce, since genomic plasticity would be artificially elevated and core genome size would be artificially small. Second, geographical variations in the proportion of microbial dark matter (the functionally uncharacterized portion of the microbial assemblage) [ 43 ] and in the degree of genomic plasticity require validation with thorough, high quality metagenomes. Differences between a metabolic inference and metagenomics analysis would highlight clades that are not well represented by sequenced genomes, or that are exceptionally plastic (and thus require a high degree of taxonomic resolution). These clades could then be targeted for isolation, genome sequencing, and phenotyping, or for genome assembly from a deep, targeted metagenome."
} | 7,647 |
38136504 | PMC10742641 | pmc | 4,945 | {
"abstract": "The increase in ecosystem biodiversity can be perceived as one of the universal processes converting energy into information across a wide range of living systems. This study delves into the dynamics of living systems, highlighting the distinction between ex post adaptation, typically associated with natural selection, and its proactive counterpart, ex ante adaptability. Through coalescence experiments using synthetic ecosystems, we (i) quantified ecosystem stability, (ii) identified correlations between some biodiversity indexes and the stability, (iii) proposed a mechanism for increasing biodiversity through moderate inter-ecosystem interactions, and (iv) inferred that the information carrier of ecosystems is species composition, or merged genomic information. Additionally, it was suggested that (v) changes in ecosystems are constrained to a low-dimensional state space, with three distinct alteration trajectories—fluctuations, rapid environmental responses, and long-term changes—converging into this state space in common. These findings suggest that daily fluctuations may predict broader ecosystem changes. Our experimental insights, coupled with an exploration of living systems’ information dynamics from an ecosystem perspective, enhance our predictive capabilities for natural ecosystem behavior, providing a universal framework for understanding a broad spectrum of living systems.",
"conclusion": "4. Conclusions In this study, we engaged with the significant question of how ecosystems change, discussing the information dynamics in living systems ranging from molecules to ecosystems from an ecosystem standpoint. Specifically, we utilized coalescence experiments in synthetic ecosystems to elucidate the quantitative relationship between biodiversity and competitive stability. We revealed that ecosystems with larger species richness were more stable in the disturbances by coalescence. Moreover, we found that species richness holds more robustness compared to population sizes in response to the process of coalescence, which was similar to a natural wetland ecosystem and the human gut microbiota. These results inferred that the information carrier of ecosystems was species composition or merged genomic information, and the functional unit of ecosystems was species abundance. The distinction between species composition and species abundance may be interpreted as a difference between species-level and individual-level parameters. Our experiments also suggested the potential applicability of the DMH to ecosystems, proposing that fluctuations in steady state, instantaneous responses to environmental changes, and long-term shifts are all constrained within the same lower dimensionality. These outcomes, combined with explanations of general aspects of adaptability, would contribute to the understanding and forecasting dynamics of not only ecosystems but also a wide range of living systems.",
"introduction": "1. Introduction Living systems can be perceived as systems that convert energy into information. The increase in the biodiversity in ecosystems is seen as one of the conversion processes. Therefore, understanding ecosystem dynamics from the perspective of information, the focus of this special topic is important for both preventing ecological crises and grasping the fundamental nature of living systems. While it is argued that organisms increase their systemic information through “adaptation by natural selection,” ecosystems lacking overt natural selection mechanisms require a distinct framework to understand phenomena that appear to enhance their information. “Adaptability” is posited as a concerted counterpart to natural selection, embodying the essence of information processing in living systems [ 1 ]. From the viewpoint of ecosystem framework, modeling living systems has propelled our comprehension of how they augment their information by increasing the entropy of the universe [ 1 , 2 ]. In systems ecology, numerous measurable macroscopic parameters encompassing information have been introduced [ 3 ]. Theoretical ecology has identified pivotal challenges [ 4 ] and offered various mechanisms addressing them in the context of adaptation [ 5 , 6 ]. Despite these advances, a comprehensive quantitative understanding of the dynamics of ecosystems and broader living systems remains elusive. The interdisciplinary nature of this field presents challenges in fostering idea exchange among researchers, potentially impeding progress [ 7 ]. A contributing factor may be the need for explanations that are more extensive than those typically found in ordinary papers. This study, which centers on the changes in ecosystems, approaches a wide array of living systems from the perspective of an ecosystem framework. Aligned with the goals of this special issue—to encourage interdisciplinary dialogue—this study offers both a thorough and accessible introduction as well as preliminary experimental findings for greater unseen ideas. The extended introduction is designed to elucidate the relevance of the experiments conducted in this study, demonstrating their significance not only for understanding ecosystems but also for providing insights into a broad spectrum of living systems. 1.1. Ecosystem Framework and Macroscopic Parameters Grasping the overall changes in ecosystems through the lens of macroscopic parameters, such as entropy and information, is beneficial for comprehensive understanding. If the alterations within various ecosystems can be encapsulated by a limited set of macroscopic parameters, it not only facilitates predictive modeling but also indicates the presence of robust constraints, effectively reducing the substantive dimensionality of the systems. While this approach diverges from the conventional definition of “ecosystem,” expanding the concept of ecosystems to encompass lifeless environments allows for a seamless integration from molecular to ecosystems. Often, even in typical ecosystems, boundaries are indistinct and defined abiotically. To circumvent confusion with ecosystems, we introduce the term “panecosystems” to describe systems that include ecosystems but also extend to contexts devoid of living entities, thereby enabling analysis through an ecosystem framework. To fully comprehend the significance of ecosystem stability and its underlying mechanisms, it is imperative to consider the entropy or information within these panecosystems. Let us consider, as a hypothetical exercise far removed from practical reality, the four panecosystems depicted in Figure 1 A. These are represented by focusing solely on the spatial distribution of an equal total number of identical atoms in a scenario where other factors, such as chemical energy, are disregarded. In the monomer system (at the leftmost end of Figure 1 A), each monomer possesses a degree of freedom in its state (position, velocity, etc.). This macrostate encompasses numerous microstates. Statistical entropy ( S B ) has been defined as a quantifier of the number of possible microstates ( W ) expressed as S B = k B l n W , where k B is the Boltzmann constant. Note that understanding formulas is not essential to grasp the concepts presented in this study. In the polymer system (positioned to the right of the monomer system in Figure 1 A), defining the state of one monomer inherently limits the potential states of the remaining nine monomers within a certain proximity. As a result, the polymer system demonstrates greater order and constraints relative to the monomer system, manifesting in a reduced number of possible microstates or a lower statistical entropy. A similar trend of diminishing statistical entropy can be observed when moving rightward in Figure 1 A. This entropy, S B , is commonly understood to be analogous to the equilibrium case of Shannon entropy (or expected information) for microstates x , expressed as S H = − ∑ x p x l n p ( x ) , where p ( x ) is the probability of state x . This relationship can also be extended to non-equilibrium states [ 8 , 9 , 10 , 11 ]. Entropy is recognized not only as an indicator of the direction in which a state will evolve but also as a fundamental link to information, energy, and work. In this study, we do not dwell on the distinctions but rather adhere to the prevailing conventions of information thermodynamics. We employ S H , which is applicable to non-equilibrium states, as a surrogate for entropy ( S ) in the broader contexts of thermodynamics and statistical mechanics. The amount of system information could be simply defined as the decrease in the number of possible microstates, i.e., the difference in statistical entropy [ 1 ]: I S = S initial − S . This metric interprets the extent to which states are constrained by order. In this context, as the statistical entropy decreases as one proceeds rightward in Figure 1 A, the information correspondingly increases. This definition of information is simple, intuitive, and theoretically convenient as it directly relates S , despite the unclear initial state S initial . Therefore, in this study, “information” means I S unless otherwise specified. However, their practical application to complex systems presents significant challenges, and actual measurement is fraught with difficulties [ 1 ]. Furthermore, note that this definition is sometimes not appropriate when considering information about living systems. For example, if all internal components in the system disappear, S will be zero, and I S will be maximum. Therefore, more suitable definitions of information exist, contingent upon the specific situation. For instance, the relative entropy between the current state and the steady state is thought to exemplify the concept that ecological communities acquire information from the environment as they near equilibrium [ 12 ]. It has also been demonstrated that relative entropy can depict state stability in models of the adaptive dynamics of ecological communities [ 13 ]. In the field of systems ecology, information is addressed as the “components and connections of system organization” [ 3 ]. This definition intuitively aligns with the aforementioned concept of I S . Although translating this definition into mathematical terms poses a challenge, systems ecology proposes some measurable indexes that include the concept of information, such as “E m ergy” [ 3 , 14 , 15 ]. This term represents the summation of energy required to generate a system and has been utilized as an indicator of ecosystem sustainability. Intuitively, incinerating 50 kg of humans or bacteria may yield a similar amount of energy. However, the creation of a 50 kg human would consume more energy. This discrepancy suggests the involvement of information, positioning Emergy as a substantive macroscopic parameter encapsulating information. Moreover, the concept of “Transformity” is defined as Emergy divided by available energy, potentially drawing it closer to the concept of information. Distinct from the information I S , the Shannon–Wiener index, frequently employed as a diversity index within ecosystems [ 16 ], denotes the Shannon entropy of species [ 17 ] as H ′ = − ∑ i p i l n p i , where p i is the proportion of individuals belonging to the i -th species. This index is zero for all panecosystems in Figure 1 A involving only a single species or type. It is important to note that while various terms with “information”, “entropy”, and “diversity” are prevalent, there are both nuanced similarities and even conceptual inversions between them. However, adopting a slightly more realistic perspective with diversity, as depicted in Figure 1 B, the diversity index H′ increases as one moves to the right, aligning with the direction of I S . This co-direction suggests that the constraints by the “realistic perspective” of the living system somehow connect H′ and I S . In other words, understanding these relationships will directly lead to an understanding of living systems. Though diversity takes various forms, it tends to align with the direction of I S in reality. Diversity’s relationship with thermodynamic indicators like I S or Emergy is typically more tenuous, yet it is often more convenient and allows for more straightforward measurement. No single form of diversity would be inherently superior, and even combinations of various diversity indexes lead to a dramatic reduction in system dimensions. Ecological studies have characterized ecosystems using various diversity indices [ 18 , 19 , 20 ]. For instance, Species Richness, which simply counts the number of different species in an ecosystem, is the most intuitive and commonly used measure of biodiversity. However, this indicator does not account for species abundance, rendering it a limited expression of biodiversity and a challenging metric to estimate accurately from natural observations. Consider an ecosystem with three species, each comprising 1000 individuals, totaling 3000. Contrastingly, another ecosystem might have 2997 individuals of one species and one individual from each of three species, totaling 3000 individuals across four species. While the latter demonstrates greater Species Richness, the presence of a single individual species may have minimal impact on the ecosystem’s characteristics. Moreover, observing a solitary individual is highly probabilistic, and Species Richness is greatly influenced by the scope of observation. The above-mentioned H′ index is a diversity measure that incorporates the likelihood of encountering individuals based on probability theory [ 18 ]. The Hill number is a more generalized indicator that enables multi-faceted evaluation, used in the Results and Discussion section below. Utilizing profiles with multiple indicators is considered preferable to selecting a single or limited measure [ 20 ]. It is also crucial to consider not only species differences but their phylogenetic disparities and functional diversity [ 19 ]. Functional diversity, in particular, is key to comprehending ecosystem processes and their responses to environmental stresses and disturbances, marking a rapidly evolving research area. As stated, there are various diversity indicators, each with its merits and limitations, and the field continues to grow with new insights and measures. These can be regarded as realistic macroscopic parameters of ecosystems. Those macroscopic parameters, such as S and I S , Emergy, Transformity, and various diversity indexes, are applicable beyond ecosystems. Applying them to a wide range of living systems as panecosystems seamlessly will highlight the characteristics of each system and provide an integrated understanding of living systems. 1.2. Information Dynamics in Living Systems: Macroscopic and Microscopic Perspectives How do living systems accumulate information? As depicted in Figure 1 B, according to the Second Law of Thermodynamics, in an isolated system—one devoid of external material or energy inputs—the entropy S would increase (signifying a movement to the left in the figure), indicating a decrease in information, and ultimately reaching a state of equilibrium. However, when considering living systems, it is important to note that even Earth is not an isolated system but rather a closed system. In the context of these living systems, the universe represents the sole example of an isolated system. Thus, living systems, being subject to external energy inputs, have the capacity to increase information. This process does not contradict the Second Law of Thermodynamics as long as the increase in information within living systems is offset by an overall increase in entropy within the universe. Nonetheless, it is not a given that energy input always increases information. For instance, simply raising or lowering the system’s temperature or altering its volume by expansion or contraction would not typically result in a continuous growth of information. The continuous information growth necessitates “agents” capable of information processing, akin to Maxwell’s demon [ 21 ] that can manipulate internal components. In ecosystems, the organisms residing internally serve as agents capable of processing information. This ability is not limited to higher organisms, such as humans; even bacteria possess systems enabling them to respond optimally to their environment [ 22 ]. In the case of organisms with explicit self-replication capabilities, it can be posited that those with higher informational content may have enhanced survival prospects through mutation and natural selection. However, this does not necessarily imply an average increase in the system’s informational content; rather, it may lead to a decrease in information among systems that are not selected. Moreover, it is also difficult to grasp the reproduction or disappearance of the system itself in a reversible manner. By considering systems at the unit level of ecosystems like panecosytems and addressing processes like polymer synthesis, cellular replication, and their respective reverse reactions through an information thermodynamics analysis, conditions for self-replication have been mathematically formulated from a statistical entropy standpoint [ 23 ]. This framework suggests that if the system changes in a way that increases the entropy of the universe more effectively, then the energy input to the system will result in a generation or enhancement of the system information. It can be used to explain how pre-living molecular systems gained the function of self-replication or how the Earth has given rise to a variety of species. In other words, considering the time scale of billions of years or infinity, it might be possible to think that it is no coincidence that Figure 1 B moves to the right. However, this framework has not yet provided dynamics in a specific timescale and upper limits, such as the maximum amount of information, or eternal stability, such as immortality. Regardless of whether it is a simple chemical reaction system or a complex living system, the behavior of the equilibrium state can be described using free energy, incorporating both energy and statistical entropy. However, in a non-equilibrium state, various dynamics can occur in high-dimensional complex systems, making them extremely difficult to understand and predict. Moreover, the emergence of sequence information in polymers such as DNA and proteins further complicates the analysis. In order to understand the characteristics of information carriers or other specific hardware, the interaction mechanism of elements within a system, and the corresponding dynamics on a specific time scale, it is necessary to consider not only the macroscopic perspective but also the mechanism of microscopic dynamics. As for information carriers, speaking broadly, the simple answer for organisms would be genomic DNA, although it is known that genomic DNA alone is insufficient to represent the heritable information [ 24 , 25 , 26 ]. However, for more general living systems, it is useful to consider the process by which information carriers are born in dynamics. Theoretical research has revealed that competition in two hierarchical layers, intracellular and intercellular, make two originally symmetrical elements asymmetrical into “information carriers,” which is not directly functional and become an information source for the functional units, and “functional units,” which is directly functional and does not become information source [ 27 ]. This research shows that the characteristics of each element become differentiated regardless of their original characteristics. Thus, the framework can be applied not only to the differentiation of DNA and proteins, i.e., the origin of the central dogma but also to cell differentiation or the division of roles in social animals. Extended to an extreme, it may be possible to consider that two system parameters acquire characteristics appropriate for the roles of information carriers, e.g., robust and not directly functional, and functional units, e.g., flexible and directly functional, regardless of their original characteristics. This assumption may provide clues to elucidate the information carrier of ecosystems in experiments in this study. The fidelity of replication of sequence information in organisms has been extensively studied in the context of the “error catastrophe” concept, wherein excessive copying errors can lead to inviability [ 28 ]. While this concept primarily addresses a high level of accuracy in genetic self-replication, it might be valuable when considering other systems, such as ecosystems. The perspective of the dynamics of sequence information also has features in common with other systems, such as ecosystems. Theoretically, the synthesis and destruction process of polymer sequence information has been analyzed from the perspective of the hardness of the processes [ 29 ], which would be similar to the concept of Emergy in the ecosystem mentioned above. Also, it is known that slow kinetic synthesis produces complex polymer sequences [ 30 ], which would be related to the trends of ecosystems in that developed ecosystems are slow [ 31 ] and that mutualism and diversity are enhanced in slower environments [ 32 ]. From this microscopic dynamic perspective, there arises a potential for macroscopic parameters that can describe the stability of conditions and the direction of change from molecules to ecosystems. 1.3. Diversity and Information Dynamics in Ecosystems: Necessity of Adaptability From the information thermodynamics perspective above, it may be natural that ecosystems with higher diversity are more stable at Earth-level sizes and very long timescales. However, understanding the human-level time scale of each ecosystem would require more specific mechanisms. Note that the consideration below ignores details and roughly assumes that more complex or diverse systems have a larger amount of information. The relationship between ecosystem stability and diversity is a paramount topic in ecology. Ecosystems are posited to develop towards a stable state and accrue information following significant disturbances [ 31 ]. Empirical observations have led to the hypothesis that complex ecosystems tend to be more stable [ 33 ]. Conversely, mathematical models indicate that as the number of species increases in simple random networks, the stability correspondingly diminishes [ 4 ]. This principle holds across networks of various structures, not limited to random configurations. However, network models that include adaptation [ 5 ] or network assembly models by adding new species, i.e., not organism-level adaptation but ecosystem-level adaptation [ 6 ], can exhibit enhanced stability with growing diversity. These insights suggest that “adaptation” is a key when considering ecosystem diversity and information dynamics. It is necessary to consider this “adaptation” in ecosystems more deeply. First, in the genetic adaptation of natural selection, organisms are systems that are selected. The system information does not consistently increase, and many systems disappear. However, if we consider the population as a panecosystem, it is possible for the system to consistently increase information. In other words, as living organisms act as information-processing agents through natural selection, even if many individual organisms disappear, the population can consistently increase in information. However, this is true within one population and does not necessarily increase the information of ecosystems with diverse populations. Natural selection alone is not sufficient to explain the increase in ecosystem information. Ecosystems, unlike organisms, do not have clear boundaries or solid information carriers and, therefore, are not subject to sophisticated selection. In the ecosystem itself, it is impossible to randomly make various copies and end up with the best ones remaining. Note that in the framework of natural selection, adaptation and fitness are determined post hoc. While fitness is often predefined for each organism in theoretical studies, it actually varies depending on the environment or situation. Therefore, fitness cannot be definitively established before the specific environment or situation is encountered. Similarly, adaptation is fundamentally the result of these environmental interactions. Thus, what is required for the ecosystem per se is not only a serendipitous adaptation by processing results but also a successful adjustment that preserves and increases information even in the face of all unexpected disturbances, i.e., “adaptability.” In this study, we adopt the definition of adaptability in the previous study [ 1 ] as “the ability of a system to cope with unexpected disturbances in the environment.” Adaptability can be thought of as the ability of a system to use energy to retain or increase information. The concept of the distinction between adaptation and adaptability does not imply that organisms only adapt due to natural selection, but organisms also have adaptability [ 1 ]. Natural selection was not considered the only mechanism for species modification even when it was proposed [ 34 ]. The adaptability of organisms encompasses adaptive phenotypic plasticity in response to unforeseen situations that is indeed observed in microbial experiments [ 35 , 36 ] and is considered to be an efficient exploratory dynamical process inherent even in cells and organisms [ 36 ]. Nevertheless, the mechanism that achieves adaptability and information increase in ecosystems remains unclear. 1.4. Mechanism for Information Increase and Identifying Information Carriers in Ecosystems When considering a mechanism for information increase in ecosystems, other than natural selection, it would be easiest to first consider some kind of “competition between ecosystems” as an information selection. Figure 2 A shows the relationship between ecosystems L and H, depicted in Figure 1 B. If these ecosystems were a closed system, a shift to the left could occur rapidly by species extinction, while a shift to the right would necessitate a long time for evolution. If the spatiotemporal scale is limited, unlike Earth, and evolution is negligible, ecosystem L would be more stable, and information would decrease. However, because ecosystems are open systems, the reintroduction of extinct species from external sources is feasible, making a shift to the right possible. This leads to important discussions about the competition between which information will remain when two ecosystems interact. Consider competition between two ecosystems when the ecosystem is open. At the extreme, this is a question of what kind of information will result when the two are mixed ( Figure 2 B). Although complete coalescence is unrealistic in natural ecosystems, similar phenomena are likely commonplace at the boundaries between distinct ecosystems. This competition by coalescence is not like natural selection as a competition within a species population. Rather, in an organism-level analogy, it would be like competition between species, i.e., what happens when two organisms exist in the same place. In a predator–prey relationship, information from the prey, primarily utilized to sustain the predator’s information, would diminish. In the case of symbiosis, both information would remain, resulting in merged information. Next, consider the competition between ecosystems H and L in a meta-ecosystem consisting of multiple ecosystems H and L ( Figure 2 C). Let us assume that ecosystem H is stronger in competition than ecosystem L and that when those interact, ecosystem L becomes H. If every ecosystem is completely closed and there is no interaction, all ecosystems H become L, as discussed above ( Figure 2 C(i)). Conversely, if every ecosystem is completely open, i.e., the meta-ecosystem becomes a single merged ecosystem, the merged ecosystem once becomes ecosystem H, but it eventually becomes ecosystem L, assuming the scale difference between the meta-ecosystem and each ecosystem is negligible ( Figure 2 C(ii)). Only if the ecosystem is moderately open can information on ecosystem H be preserved ( Figure 2 C(iii)). In natural ecosystems, for instance, this moderately open scenario might correspond to a meta-ecosystem separated by rivers that occasionally intermix due to relatively rare events such as typhoons. Alternatively, simply, it might be that the rate of transition between ecosystems is low, but this was not experimentally confirmed in this study. In this way, ecosystem information can be preserved through ecosystem competition at moderate openness. It is possible to infer the information carrier and functional unit of ecosystems if it is indeed possible to maintain or increase information through such competition between ecosystems. As discussed above, let us assume that the two system parameters differentiate into those that are robust and not directly functional and those that are flexible and directly functional, as the information carrier and functional unit, respectively, regardless of their original characteristics. Then, simply examining what parameter is more robust or flexible during the competitions may provide clues to identify the two roles, even without knowing their properties. This method of identifying information carriers is not surprising when considering a typical example. One of the most notable examples of adaptability in living systems is the brain, which utilizes energy for information processing. Let us consider two parameters in the brain: synaptic weight and neural activity. When mathematically modeled as artificial neural networks, these are often expressed as w and x vectors, respectively, and function as w∙x [ 37 ]. Thus, the characteristics of those two parameters are often symmetrical. However, there is a clear difference between the two parameters in both the brain and artificial neural networks: synaptic weight and neural activity are more robust and flexible as information carriers and functional units, respectively. In artificial neural network models, synaptic weights are developed gradually through a learning process. Often, these weights remain unchanged when the model is utilized for inference. Therefore, synaptic weight can be regarded as an information carrier that defines the model’s characteristics. In contrast, neural activity is a variable that changes at each instance of inference. It is a functional parameter that varies in response to input and generates output. Consequently, in artificial neural networks, synaptic weight and neural activity can be interpreted as information carriers and functional units, respectively. While it cannot be claimed that these are identical to the brain, artificial neural networks were initially devised as brain models and are believed to replicate similar characteristics. Additionally, there are some similarities between ecosystems and the brain, as both are living systems that use energy to process information without natural selection [ 38 ]. A typical example of the brain’s adaptability to respond to unexpected situations is inspiration with the Eureka effect, which is an ability to come up with the correct answer to unlearned tasks without any learning by taking a relatively long time to think [ 39 , 40 ]. Also, the brain can learn a new concept from one example of Hebbian learning, a simple process of strengthening the synaptic connections that are used [ 41 ]. Conversely, it is theoretically shown that ecosystems can have adaptability akin to Hebbian learning in neural networks [ 42 ]. This notion of adaptability as an inherent characteristic of living systems also aligns with the “free energy principle,” which assumes that brains and organisms inherently predict their state, ensuring their survival by minimizing the “prediction error” using sensory feedback from the environment [ 43 ]. This convergence suggests that both organisms and ecosystems are inherently grounded in principles of adaptability, ensuring resilience in the face of uncertain disturbances. 1.5. Freedom and Constraints, Homeostasis and Homeorhesis Adaptability, which is an essential ability of living systems, is thought to require a balance between freedom and constraints. It would be reasonable for information-processing systems to have strong orders and keep large amounts of information, that is, constraints of the system states in the context of statistical entropy. On the other hand, responding to unexpected disturbances requires not only constraints but also freedom. Thus, a balance between the opposing facets of freedom and constraint is considered to be a fundamental demand on living systems [ 1 , 44 ]. The property that expresses the balance in an easy-to-understand manner is “homeorhesis” [ 45 ]. Homeorhesis is a property of a dynamical system that maintains a particular trajectory despite perturbations from the environment. This is similar to homeostasis, which describes the property of maintaining a stable state [ 46 ], but it is different in that a system with homeorhesis is constantly changing. It should be noted that living systems exhibit characteristics that can be identified as both homeostasis and homeorhesis, and it is not feasible to distinctly differentiate between these two in actual phenomena. These terms merely serve as useful concepts for emphasizing a particular characteristic of a dynamic system. To discuss the necessity of homeorhesis, let us consider, conversely, that adaptation through natural selection is possible even without homeorhesis. In adaptation through natural selection, freedom and constraint can be provided to the organisms independently ( Figure 3 A). First, organisms need a stable phenotype that corresponds to a static information carrier for being selected by the environment. This is homeostasis, which is a property of system dynamics as constraints. On the one hand, freedom in variation is required for the adaptation, which is provided as random mutations, independent of the properties of the dynamics. Therefore, as a property of dynamics, as long as there is homeostasis to maintain a stable state, adaptation through natural selection is possible. On the other hand, in the case of an ecosystem without a static information carrier and its mutation, freedom in variation for adaptation is also required as a characteristic of the system dynamics. Therefore, the system must not only have stability but also remain degrees of freedom to change. As mentioned above, if this change were completely free, the system would not be able to continuously retain or increase its information. Therefore, adaptability requires both strong constraints and a certain degree of freedom for changes as a property of the system dynamics, which is exactly homeorhesis ( Figure 3 B). Note that homeostasis and homeorhesis are not mutually exclusive properties. At least, homeorhesis in organisms and ecosystems would encompass random changes within the sublevels of their hierarchical structure. Consequently, homeorhesis may be considered a property of a higher hierarchy than homeostasis. In systems ecology, it is postulated that systems at sub-organism levels in the hierarchies of living systems, such as organs or molecules, mainly demonstrate homeostasis, whereas super-organism levels, such as populations or ecosystems, mainly display homeorhetic properties [ 3 ]. For instance, molecular systems have an equilibrium or steady state they can maintain, whereas ecosystems are far from any equilibrium or steady state and continuously changing. Organisms can play genetic evolution through natural selection without homeorhesis; however, a theoretical study has suggested that continuous evolution inevitably leads to the acquisition of homeorhesis [ 47 ]. More specifically, as a result of continuous evolutionary processes, the behavior of high-dimensional organisms was constrained to a small number of dominant mode dimensions, which correspond to the dimensions of proliferation rate. The changes due to steady fluctuations, responses to environmental changes (i.e., phenotypic plasticity for organisms), and long-term changes (evolution) are constrained into the dominant mode ( Figure 3 B). Therefore, for example, the direction of an adaptive change in the state space is predictable from the fluctuation, like a fluctuation-response relationship [ 48 , 49 ]. In this study, this hypothesis proposing the existence of such a dominant mode as a balance of strong constraints and a small degree of freedom is referred to as the Dominant Mode Hypothesis (DMH). The DMH demonstrates the emergence of homeorhesis and proposes that even if a genetic mutation is random, changes in the system are never random and fully controlled. This property depicts the adaptability of dynamics that facilitates ex ante adaptation rather than ex post adaptation as ordinarily considered in natural selection. Indeed, the DMH was experimentally tested using bacterial adaptation [ 47 , 50 ]. For instance, various types of environmental changes, such as osmotic pressure, high temperature, and starvation, seemed to induce similar changes in the transcriptome [ 51 ]. Similar observations have been made in proteomes under varying environmental conditions, including different nutrients and cultivation methods [ 47 ]. Additionally, it has been observed that responses to environmental changes and alterations resulting from adaptive evolution (laboratory evolution) were also similar [ 52 ]. Constraints to low-dimensional phenotypic states were further confirmed in a high-throughput laboratory evolution for various drug resistances [ 50 ]. Considering that ecosystems do not undergo clear natural selection, unlike living organisms, and are thought to exhibit stronger characteristics of homeorhesis rather than homeostasis [ 3 ], it can be proposed that the DMH may also be applicable to ecosystems. The DMH will be a powerful tool for predicting system changes, e.g., the response to the global warming of an ecosystem can be predicted from its daily fluctuations. Note that there is another useful idea of using the resilience of dynamical systems to describe changes in a steady state in a low-dimensional “efficient dimension” [ 53 , 54 ]. This idea focuses only on homeostasis and is fundamentally different from the DMH. More generally, the DMH suggests an inevitable existence of strong dimensionality reduction for systems with adaptability, including all living systems. Living systems can fundamentally be viewed as dynamic networks comprising numerous dimensions. For instance, even in the bacterium Escherichia coli , there are over 1000 types of gene expression levels, in addition to the amount and spatial arrangements of many other molecules. From a control theory standpoint, managing scale-free networks, which typify biological networks, is exceedingly challenging [ 55 ]. However, it is posited that this difficulty is managed due to the interrelations among many dimensions and their constraint to fewer dimensions [ 56 ]. This dimensionality reduction has also been observed in the brain, which is a typical example of a biological information-processing system [ 57 , 58 ]. The dimensionality reduction per se would be an inherent feature of information-processing systems and may also appear as a topological constraint, such as biological “Bowtie” structures [ 59 ] or autoencoders in artificial neural networks [ 60 ]. Therefore, the DMH would be useful not only for understanding living systems with adaptability but also for constructing dynamic artificial information-processing systems as an application. 1.6. Using Experimental Ecosystems as a Phenomenological Approach Despite numerous approaches to understanding the changes and stability of complex living systems, a comprehensive understanding at both macroscopic and microscopic levels remains elusive [ 61 , 62 ]. This gap indicates a potential shortfall in phenomenological approaches similar to those employed in thermodynamics. Since living systems have many commonalities and strong universality, model experimental systems are useful. For example, the understanding gained with one of the simplest model organisms, Escherichia coli , has helped us understand many other organisms [ 63 ]. Such universality suggests common constraints and reduced dimensionality and, therefore, makes us expect that the systems can be described with a small number of macroscopic parameters. In systems ecology, species-defined experimental ecosystems by synthetic assemblages of microorganisms, designated here as “synthetic ecosystems”, were proposed for experimental model ecosystems [ 64 , 65 ]. Experimental studies concerning ecosystem diversity, stability, and ecosystem services include outdoor systems such as Cedar Creek [ 66 , 67 ] and laboratory-level systems like Ecotron [ 68 , 69 ]. Even microcosms composed of microorganisms, despite their limited scale, are considered valuable for addressing global ecological issues [ 70 ]. However, assessing ecosystem stability poses various challenges, including issues related to replicability. Furthermore, exploring system characteristics sometimes requires studying conditions not present in natural ecosystems. For instance, in assessing a system’s resilience to external disturbances, it is crucial to impose various types of disturbances that are too extreme for the system’s continued existence. Additionally, to ascertain if the core characteristic of a sustainable ecosystem is “something”, it is vital to compare systems that vary only in the presence or absence of that “something,” ensuring that the system lacking this “something” cannot exist as a natural ecosystem. To address these challenges in microcosms, we previously developed a high-throughput experimental system of a synthetic ecosystem consisting of only model microorganisms, comprising three important functional groups of the ecosystem: producers, decomposers, and consumers [ 71 ]. The synthetic ecosystem includes fundamental ecological processes like photosynthesis, predator–prey interactions, competition, and cooperation. Each species within this system is amenable to cryopreservation, ensuring experimental replicability. This experimental model ecosystem allows for systematic ecosystem experimentation under various conditions, including those unattainable in natural environments, akin to the role of E. coli as a model for various organisms. Specifically, our model ecosystem facilitates the investigation of inter-ecosystem competition and the identification of ecosystem information carriers, as depicted in Figure 2 , as well as the examination of ecosystem constraints in the form of homeorhesis, illustrated in Figure 3 . The mixture of two ecosystems shown in Figure 2 B can be easily tested systematically in our model ecosystem. While such complete coalescence is unrealistic in macroscopic ecosystems, thereby lacking research, it is considered to be frequent in microbial ecosystems, known as “community coalescence” [ 72 ]. Experiments involving merged microbial communities demonstrated the strong influence of dominant species and support of dominant species by other species [ 73 ]. Therefore, these results suggested that the characteristics of dominant species, rather than diversity, are important for representing the system dynamics. The study represents an important step in investigating commonalities and discrepancies in various ecosystems; however, the tested experimental ecosystems lack predatory factors, a feature crucial in general ecosystems. Moreover, it is also possible to address the DMH, i.e., homeorhesis and constraints of ecosystems, in our model ecosystem. The homeorhesis has been observed in the process of leading ecosystems towards stable states, called ecological succession, in an exceedingly simple synthetic ecosystem [ 74 ]. In similar synthetic ecosystems, it has also been demonstrated that stochastic fluctuations within the system adhere to a power law as a constraint [ 75 ]. Furthermore, in experimental ecosystems using more complex, field-collected microbial communities, even in microbial ecosystems with considerable population changes, the functional structure has been found to remain stable [ 76 ]. Similar experimental ecosystems have shown that even with changes in species, the overall phylogenetic structure is robust [ 77 ]. Those studies might suggest the existence of the homeorhesis and constraints in experimental ecosystems; however, the applicability of the DMH for ecosystems has not been empirically demonstrated. Experimentally demonstrating the DMH necessitates precise and numerous replicated experiments, a challenge in natural ecosystems or experimental ecosystems where high-throughput experimentation is arduous. Moreover, testing dimensionality reduction is unfeasible with an overly simplistic synthetic ecosystem. Nonetheless, our synthetic ecosystem possesses the potential to demonstrate the DMH as a high-throughput experimental system despite possibly having too few species. If the DMH is experimentally shown to be applicable to ecosystems, it will be of great help in proactive biodiversity conservation and ecosystem management. Understanding ecosystem dynamics requires not only a macroscopic view but also microscopic insight, with a particular emphasis on comprehending evolution and population dynamics [ 78 ]. Microbial experimental systems are powerful tools for elucidating evolutionary processes, as demonstrated by many studies [ 79 , 80 , 81 ]. Our model ecosystem is also conducive to evolutionary research, and some pertinent results related to evolution are shown below, but this study will not delve deeply into discussions of evolution due to limited experimental data. In studies concerning evolution within similar synthetic ecosystems, numerous significant findings have been presented. For instance, evidence has been shown of species diversifying their survival strategies [ 82 ] and instances where free-living algae have shifted towards a more endosymbiotic existence [ 83 , 84 ]. These examples are being discussed from a broader perspective [ 85 ], suggesting experimentally that ecosystems tend toward an increase in information. 1.7. Experiments in This Study We have demonstrated, using the model synthetic ecosystems, scenarios such as the coalescence of two ecosystems and the constraints of ecosystems shown in Figure 2 and Figure 3 , respectively. Specifically, we quantified which ecosystem was more competitive by merging two ecosystems. The results showed that ecosystems with higher diversity were more competitively stable, and the species composition was more robust and had a higher ability to explain the dynamics than the population of dominant species that was flexible. Therefore, the scenario depicted in Figure 2 C was valid in these experiments, and it was speculated that the information carrier and functional units of ecosystems were species composition, i.e., merged genomic information and species abundance, respectively. Moreover, we investigated the response to temperature changes and long-term changes in ecosystems. The results have suggested that the DMH is also applicable to ecosystems. While these results are not comprehensive enough to substantiate theories, those outcomes are appropriate for fostering discussion in this special issue and for demonstrating the potential for future research using our model synthetic ecosystem for connecting a wide range of living systems.",
"discussion": "3. Results and Discussion 3.1. Ecosystems Used as Initial State We prepared eight ecosystems, each with a certain degree of stability, and conducted coalescence experiments by mixing them pairwise, as detailed below. In the previous research, experimental ecosystems composed of a mix of 11 species were divided into 72 replicates and cultivated under identical conditions for six months. Stochastically, these ecosystems were separated into roughly seven patterns [ 71 ]. For this study, we utilized 8 of these 72 ecosystems. Out of the 11 species, 5 species could not survive in any of the ecosystems, leaving 6 species that persisted in at least across one of the seven patterns. These six species are listed in Table 1 , and we refer to each of them by their abbreviated names shown in Table 1 , i.e., Ecoli, Tetra, CyanoA, CyanoS, AlgaR, and AlgaC, in this study. These species include three important functional groups of ecosystems: producers, decomposers, and consumers. Some species exhibit mutualistic relationships, enabling their coexistence with multiple species, as they could not survive alone [ 71 , 89 ]. Moreover, Tetra had predator–prey interactions with Ecoli and CyanoS [ 90 ]. The producers include four species, with both prokaryotes and eukaryotes represented by two species each, which have potentially competitive relationships. Note that not all six species coexisted within a single ecosystem, but each ecosystem contained between two and five species. Figure 4 illustrates the species composition of the eight ecosystems used as the initial states (designated as Ecosystems E0 through E7). Figure 4 A represents the original ecosystems, with 4B depicting the initial states that were achieved by diluting and aliquoting the original ecosystems into four replicates. Subsequently, the outcomes after approximately seven serial transfers, conducted approximately every two weeks, are shown in 4C (in other words, those that survived through 10 7 to 10 8 dilutions, roughly four months after the coalescence). While the ecosystems are stable overall, not all of them are entirely so. Specifically, the population of AlgaR tended to decrease gradually, and in some ecosystems, it fell below the detectable limit after four months. For the following analyses, the values from Figure 4 B were utilized as the initial conditions, representing the states before coalescence. 3.2. Ecosystem Coalescence Experiments for Investigating Competitive Stability and Information Carrier We mixed the above eight distinct ecosystems in a comprehensive pairwise manner, incorporating two of each, leading to an all-versus-all combination. Figure 5 A illustrates the outcomes four months post-coalescence for these pairwise combinations. It encompasses the results of the 36 distinct ecosystems, considering both the 8 C 2 combinations and the 8 original ecosystems. For the latter, identical ecosystems were mixed to align experimental conditions. Observationally, when ecosystems with lower and higher species richness (the number of species) were merged, the resulting species richness seemed to tend to gravitate towards the values of the higher species richness (see below for quantitative analyses). Figure 5 B represents the outcome of the 36 ecosystems after 4 months using a Principal Component Analysis (PCA) performed on the logarithm of the population sizes of the six species. The results for the unmixed eight ecosystems are indicated by text. An immediate observation is that the ecosystem consisting solely of prokaryotes (E0) was dramatically altered from its own state in every combination. Additionally, there appears to be a clustering toward ecosystems E4, E5, E6, and E7. We investigated which ecosystems maintained their state stably. In this context, we introduce the concept of a competitive stability index ( Θ ) as a metric to assess the extent to which an ecosystem sustains its population composition post-coalescence. The competitive stability index of each i -th ecosystem is defined as Θ i = 1 / ∑ j ∑ k x a f t e r , k , j − x i n i t , k , i 2 , where x after ,k,i and x init ,k,i denote the logarithm of the population of species k in the i -th ecosystem at initial and 4 months, respectively. As x is a logarithm value, we used x = 0 for the population not detected. Figure 5 C illustrates the relationship between Θ values and diversity indexes (α-diversity) of the eight ecosystems. We considered three simple measures of α-diversity: species richness ( 0 D , the number of species), the Shannon–Wiener index ( H′ ), and the biomass-corrected Shannon–Wiener index ( BH′ ). In BH′ , the probability of each species’ population ( p i ) in H′ is substituted with the relative biomass of each species, which is the proportion of biomass represented by each species [ 91 , 92 , 93 , 94 ]. The values of approximate volume, shown in Table 1 , were used as biomass values for each species. The results indicated the highest correlation between BH′ and Θ , with R = 0.86, p = 0.007, hereafter α = 0.05. Note that this relationship was somewhat influenced by the formulation of Θ . For instance, while species richness did not significantly correlate with Θ ( R = 0.59, p = 0.12), its inverse (1/ Θ ) showed a significant negative correlation ( R = −0.72, p = 0.04), similar to that of BH′ ( R = −0.73, p = 0.04). BH′ normalizes the disparities between populations of larger and smaller organisms, making it closer to a measure of species richness. Therefore, for a simple understanding, the larger the species richness, the more stable the ecosystem was, i.e., having better adaptability. These results support the mechanisms illustrated in Figure 2 C that explain the sustainability or increase in ecosystem information. Note that the objective of this comparison of indicators is not to determine which is superior in representing natural ecosystems but to contrast the characteristics of stable ecosystems using straightforward indicators. For example, species richness, although deemed overly simplistic and problematic in depicting a natural ecosystem requiring estimation [ 20 ], has the benefit of involving just a single parameter, considerably fewer than other indicators. Therefore, if a phenomenon can be effectively explained by species richness, it is beneficial from an information criterion standpoint. Conversely, H′ failed to account for the competitive stability ( R = 0.03, p = 0.95). This shortfall likely arises because H′ inherently underrepresents species with larger biomass but smaller populations, thereby reducing their contribution. In systems ecology, larger individuals are often considered to carry more information [ 31 ], which is expressed in the opposite way in H′ . The obtained fact that ecosystems with a larger richness are more stable suggests that the larger richness of the two pre-coalescence ecosystems could more accurately predict the post-coalescence richness than the smaller one. However, it is not clear whether information from the ecosystem with smaller richness remains in the post-coalescence ecosystems. Using an analogy with organisms ( Figure 2 B, lower), it is necessary to clarify whether only information about ecosystem H remains, like predator–prey relationships or information that merges both ecosystems remains, like symbiosis. We investigated which data set from the two ecosystems or their combination could better forecast the outcome. Specifically, we employed richness as the most fundamental indicator with the fewest number of parameters and determined how much the pre-coalescence richness could dictate the post-coalescence richness ( Figure 5 D). It was found that the larger richness had a greater coefficient of determination than the smaller or mean richness of the two ecosystems. This implies that the results are closer to the larger richness, consistent with the aforementioned competitive stability of ecosystems with higher richness. Additionally, the richness calculated from the merged two ecosystems (representing the gamma diversity of the two ecosystems, which is the same as the initial state of the merged ecosystem) had the greatest coefficient of determination. This suggests that the initial richness upon coalescence remains relatively unchanged, indicating that the information from the ecosystem with smaller richness was not lost but was influential in the resultant richness; thus, when comparing with the cases of organisms in the aforementioned Figure 2 B, the observation that mixing two ecosystems results in the retention of information from both may suggest that the relationship between these ecosystems can be interpreted as not predatory but rather symbiotic. In the results above, the richness was able to adequately explain the outcomes, whereas H ’, i.e., population information, was less explanatory. This may differ from previous microbial coalescence studies where the dominant species could explain the outcomes [ 73 ]. Conversely, in a natural wetland ecosystem, it is known that an ecosystem state index NDVI (Normalized Difference Vegetation Index) can predict species richness more accurately than dominant populations [ 95 ], suggesting a potentially similar situation. To quantitatively verify this in our case, we employed an approach akin to the previous study of the natural wetland ecosystem [ 95 ], examining the predictability of outcomes by varying the order parameter q in the widely applicable diversity index known as Hill numbers D q = ∑ i p i q 1 / ( 1 − q ) , where p i is the proportion of individuals belonging to the i -th species. When q = 1, the formulation is undefined, but the mathematical limit as q approaches 1 is defined as D 1 = e x p − ∑ i p i l n p i , i.e., the exponential of H′ . Specifically, using a certain value of q , we calculated the q D from the population of two pre-coalescence ecosystems and used this as the explanatory variable, with the q D of the post-coalescence ecosystem after four months as the dependent variable to determine the coefficient of determination. This process was repeated with varying q values, and we obtained the q profile of the coefficient of determination ( Figure 5 E). Note that the interpretation of Hill numbers changes with the order parameter q . Roughly speaking, smaller q values emphasize the presence or absence of species, while larger values prioritize population sizes, i.e., the proportion of dominant species population in extreme cases. Specifically, 0 D equates to species richness, independent of population sizes. 1 D corresponds to the exponential of the Shannon–Wiener index, where population sizes are considered. 2 D equals the Simpson index, focusing more on the population sizes and highlighting the prevalence of dominant species. The results show that the highest coefficient of determination was observed at a low q value of 0.1 ( Figure 5 E), indicating that species composition was robust and species abundance was flexible. They also satisfy, respectively, not functioning directly and functioning directly. Therefore, the information carrier and functional units of ecosystems were speculated as the species composition and species abundance, respectively. It is important to note that this speculation is based on a comparison between only these two aspects: species composition and species abundance. In reality, ecosystems comprise many other parameters. This nature of flexibility of populations and robustness of species composition would be consistent with the characteristics of the human gut microbiota [ 96 ]. Moreover, this q profile is akin to the predictability of the mean NDVI in natural wetland ecosystems, maximum at q = 0.2 [ 95 ]. Therefore, our finding that the information carrier of ecosystems is species composition might be universally applicable to other ecosystems as well. In our experimental system, the low explanatory power of the dominant species can be readily explained by the presence of predation. For instance, in a system consisting only of Ecoli and CyanoS, as represented in E0, both species exhibit small biomasses, leading to exceedingly high population numbers. When this ecosystem is mixed with one containing the predator, Tetra, the population of these smaller organisms diminishes rapidly. In the same sense, the predator is the largest in biomass; thus, their population is always small, but the outcome changes greatly depending on whether the predator is present or not, just like a keystone species [ 97 ]. Consequently, population size scarcely contributes as an explanatory variable. On the contrary, the rapid decrease in the prey population does not equate to extinction, and species often persist at low population levels, thereby maintaining species richness. Thus, emphasizing the functions of each species can effectively elucidate the results, indicating the advantages of considering functional diversity. In our system, the species variety is too limited to be worth evaluating quantitatively, but conceptually, only ecosystem E0 lacks predators and may be deemed to have reduced functional diversity. Consequently, it might be inferred that ecosystems with greater functional diversity are more stable. Note that microbial experimental ecosystems in the previous study [ 73 ], where the dominant population shows high explanatory power, do not contain any predators, which may affect our results or the natural wetland ecosystem [ 95 ]. These ecosystems are, therefore, perceived as having low functional diversity, with competition occurring exclusively among them. Should a functional diversity index be uniformly applicable across all ecosystems, it might enable comparisons of markedly disparate ecosystems on an equal footing. While functional diversity presents various challenges due to its inherent complexity, ongoing enhancements aimed at ensuring universality and mathematical robustness are promising, positioning it as a potential comprehensive indicator [ 98 ]. Simultaneously, if our synthetic ecosystem were to be developed to include more species, it would become possible to experimentally demonstrate the advantages of functional diversity. As mentioned above, the predator species Tetra plays an important role as a keystone species in this ecosystem. This keystone species is small in number and has a slow maximum rate of proliferation. The population was also robust for this predator species. These characteristics of robust, small in number, and slow are appropriate for an information carrier. For example, if the characteristics of a single individual of this keystone species change due to genetic variation, the characteristics of the whole ecosystem can change rapidly because the population size is small, and this species is influential. Although we did not compare this specific population with other parameters in this analysis, the keystone species itself might be the information carriers of ecosystems. Our coalescence experiments consistently showed that species richness generally demonstrated robustness, thereby serving as an information carrier or a stable macroscopic parameter inherent to the systems. However, it is imperative to acknowledge that this finding does not universally apply to all ecosystems. The ecosystems utilized in the coalescence experiments here represent a recombination of divergent ecosystems that originated from the same source. Systematic investigations are essential to discern under what conditions certain parameters prove most useful or possibly appropriate as information carriers. In our synthetic ecosystems, this investigation is feasible, and further elucidation is expected from future research. 3.3. Ecosystem Constraints for Investigating the Dominant Mode Hypothesis In this study, we experimentally investigated the DMH, which suggests that living systems possess a small degree of freedom by strong constraints, with changes predominantly confined to lower dimensions. Specifically, we examined the two types of ecosystem changes: (i) the rapid response of ecosystems in 7 days due to temperature changes and (ii) the gradual alterations of ecosystems observed in approximately 18 months without any induced environmental changes. Before explaining our results, we describe the inherent limitations of these experiments below. Firstly, the measurements lack microscopic observation and rely solely on fluorometry using a plate reader (see Materials and Methods for details). While the precision of fluorometry is higher than that of data obtained from the microscopic observation, the dimensionality is limited, presenting a problem for the study of dimensionality reduction. Moreover, the low number of species of the synthetic ecosystem is also a significant problem. Nevertheless, we believe that presenting these results is beneficial as a trial demonstration for predicting ecosystem changes. For instance, the reduction from two dimensions to one can also be considered a kind of constraint. We first tested environmental temperature changes. Specifically, for ecosystems initially at 23 °C, we varied the temperature to 25, 28, and 33 °C and observed the changes after seven days. The comparison of responses was not between the initial values and those after seven days, but between the responses at 23 °C after seven days and those at the varied temperatures after the same period because our ecosystems have a kind of stable state in the circumstances of subculturing every two weeks. We used three ecosystems: E0, the simplest ecosystem comprising only bacteria, and E6 and E7, ecosystems with the two largest richness among the eight types of ecosystems depicted in Figure 4 . Figure 6 A presents the results of PCA for the logarithm of the fluorescence intensity, projecting the results in two dimensions. In the case of ecosystem E6, the direction of fluctuation in the standard environment (23 °C, blue dots) appears to align with the response to temperature changes. E7 may adhere as well, suggesting changes within certain constraints. The simplest ecosystem, E0, exhibits little fluctuation and response change. This co-absence of fluctuation and response is also consistent with the implications of the DMH. The reason why such constraints were observed was simple. First, examining the contribution fractions in PCA (as seen in the inset of Figure 6 A), it is evident that only the two dimensions corresponding to cyanobacteria (Cyano) and green algae (Alga) are contributing, indicating that the PCA does not actually compress dimensions, unfortunately. Thus, the utilization of PCA here was merely for demonstration purposes, serving as an example for analyzing higher-dimensional ecosystems in future research. Nevertheless, the constraint from two dimensions to one was indeed present. Second, Figure 6 B illustrates the relationship between the fluorescence intensities representing the populations of cyanobacteria and green algae. These results suggest that the sum of both populations reaches a constant number as a carrying capacity, likely due to a trade-off resulting from competition for a resource such as carbon dioxide. Although this is a simplistic observation, it could be considered a typical constraint anticipated within ecosystems. Next, we observed the long-term changes in the state of ecosystems. We branched each of the three aforementioned ecosystems into 32 replicates, continuing independent cultivation for 18 months. Although the fresh medium is supplied at each subculturing transfer, the biological elements are not supplied from outside each ecosystem, like the situation shown in Figure 2 C(i). The results of the PCA, conducted in the same manner as in Figure 6 A, are presented in Figure 6 C. For both E6 and E7, the state transitions again roughly appear to align along a singular curve, a phenomenon explainable by the constraints in the dominant mode hypothesis. E0 again exhibited little changes. We also tested the long-term changes when the 32 dispensed ecosystems were merged at every subculturing transfer ( Figure 6 D). This experiment tested a situation similar to the one shown in Figure 2 C(ii). The results show the constraints as well. All these results shown in Figure 6 suggest that the DMH is also applicable for ecosystems, which highlights the homeorhesis and adaptability of ecosystems. Although the results were poor, compression from two dimensions to one dimension was visible. However, as mentioned above, there are many problems with this experiment, and it is necessary to set better conditions and confirm it properly. At the same time, it is expected that similar analyses will be attempted in other experimental systems. In this study, replicate experiments were employed to account for the fluctuation of the ecosystem, but in natural ecosystems, utilizing daily fluctuations, for instance, could also be used. Our findings, exemplified by the trajectories in Figure 6 C,D, suggest that the permissible direction of daily variations is constrained to a lower dimensionality. Additionally, the results in Figure 6 C,D, i.e., when closed and completely open situations, respectively, also show interesting results consistent with the scenario shown in Figure 2 C. In Figure 6 C, as closed systems, 32 replicates are scattered on the right edge, left edge, and center. The right and left edges indicated that producers were almost exclusively Cyano or Alga, respectively. One of them might actually be extinct. The plots between them indicate the states in which both Cyano and Alga coexist. Therefore, information on some of the 32 ecosystems decreased, as depicted in Figure 2 C(i). In Figure 6 D, the final results were almost entirely at the right edge. Thus, information on all 32 ecosystems, or more precisely, one large ecosystem, decreased, as depicted in Figure 2 C(ii). These two results suggest that it is impossible for ecosystems to sustain or increase the information if they are completely closed or completely open, as shown in Figure 2 C(i) and Figure 2 C(ii), respectively, despite the fact that ecosystems with higher richness were more competitively stable as above."
} | 17,507 |
23975157 | PMC3819121 | pmc | 4,947 | {
"abstract": "Profiling phylogenetic marker genes, such as the 16S rRNA gene, is a key tool for studies of microbial communities but does not provide direct evidence of a community’s functional capabilities. Here we describe PICRUSt (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States), a computational approach to predict the functional composition of a metagenome using marker gene data and a database of reference genomes. PICRUSt uses an extended ancestral-state reconstruction algorithm to predict which gene families are present and then combines gene families to estimate the composite metagenome. Using 16S information, PICRUSt recaptures key findings from the Human Microbiome Project and accurately predicts the abundance of gene families in host-associated and environmental communities, with quantifiable uncertainty. Our results demonstrate that phylogeny and function are sufficiently linked that this ‘predictive metagenomic’ approach should provide useful insights into the thousands of uncultivated microbial communities for which only marker gene surveys are currently available.",
"introduction": "Introduction High-throughput sequencing has facilitated major advances in our understanding of microbial ecology and is now widespread in biotechnological applications from personalized medicine 1 to bioenergy 2 . Markers such as the 16S ribosomal RNA gene (16S) of bacteria and archaea are frequently used to characterize the taxonomic composition and phylogenetic diversity of environmental samples. Because marker gene studies focus on one or a few universal genes, they cannot directly identify metabolic or other functional capabilities of the microorganisms under study. Conversely, metagenomic sequencing aims to sample all genes from a community and can produce detailed metabolic and functional profiles. Although relatively little sequencing is needed to characterize the diversity of a sample 3 , 4 , deep, and therefore costly, metagenomic sequencing is required to access rare organisms and genes 5 . Thus, marker gene profiling of large sample collections is now routine, but deep metagenomic sequencing across many samples is prohibitively expensive. Although marker gene and shotgun sequencing strategies differ in the type of information produced, phylogeny and biomolecular function are strongly, if imperfectly, correlated. Phylogenetic trees based on 16S closely resemble clusters obtained based on shared gene content 6 - 9 , and researchers often infer properties of uncultured organisms from cultured relatives. For example, the genome of a Bacteroides spp. might reasonably be inferred to contain many genes encoding glycoside hydrolase activity, based on the commonality of these activities in sequenced Bacteroides isolates 10 . This association is in turn closely related to the pan- and core-genomes of each phylogenetic subtree 11 , in that larger and more strongly conserved core genomes result in more confident linkages of genes with clades. Conversely, a clade’s core genome consists of genes its genomes can be expected to carry with high probability. The correlation between phylogeny and functional attributes depends on factors including the complexity of the trait 12 , but the overall degree of correlation suggests that it may be fruitful to predict the functions encoded in an organism’s genome on the basis of functions encoded in closely related genomes. Recently, some 16S studies have extended these intuitions to infer the functional contribution of particular community members by mapping a subset of abundant 16S sequences to their nearest sequenced reference genome 13 - 15 . The accuracy of such approaches has not been characterized, but the correlation between gene content and phylogeny 8 , 9 , 16 (excepting special cases such as laterally transferred elements and intracellular endosymbionts with reduced genomes) suggests that it may be possible to approximately predict the functional potential of microbial communities from phylogeny. Widespread and reproducible application of such a strategy requires an automated method that formalizes the relationship between evolutionary distance and functional potential across the entire metagenome, accounts for variation in marker gene copy number 17 , and accurately recaptures insights from shotgun metagenomic sequencing. Here we describe PICRUSt (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States), a technique that uses evolutionary modeling to predict metagenomes from 16S data and a reference genome database. We investigated the accuracy of this approach as a function of the phylogenetic proximity of reference genomes to sampled environmental strains and the rate of decay of the phylogeny-function correlation owing to a variety of factors including gene duplication, loss, and lateral gene transfer. Lateral gene transfer is particularly relevant because it allows distantly related genomes to share functions that are missing from closer relatives and appears to be particularly widespread in microbes sharing a common environment, including constituents of the human microbiome 18 , 19 as well as extreme and contaminated environments 20 , 21 Quantitative predictions also depend on accurate modeling of community member abundance, which can be affected by 16S copy-number variation 17 ( Supplementary Results ). Although these caveats could theoretically limit the accuracy of any inference of microbial function from 16S sequence data, their quantitative effects on this relationship have not previously been explored in detail. Our results using published data show that PICRUSt recaptures key findings from the Human Microbiome Project and predicts metagenomes across a broad range of host-associated and environmental samples. We applied PICRUSt to a range of datasets from humans 22 , soils 23 , other mammalian guts 24 and the hyper-diverse and underexplored Guerrero Negro microbial mat 23 , 24 , which allowed us to model how the accuracy of PICRUSt varies based on the availability of reference genomes for organisms in each environment. In the best cases, correlations between inferred and metagenomically measured gene content approached 0.9 and averaged approximately 0.8. PICRUSt recaptured most of the variation in gene content obtained by metagenomic sequencing using only a few hundred 16S sequences and in some cases outperforms the metagenomes measured at particularly shallow sampling depths Additionally, we quantified the effects of several other factors on PICRUSt’s accuracy, including reference database coverage, phylogenetic error, gene functional category (a potential surrogate for the effects of lateral gene transfer), ancestral state reconstruction method, microbial taxonomy and 16S sequencing depth. Finally, we applied PICRUSt to several 16S-only datasets to identify previously undescribed patterns in gene content in oral, vaginal and coral mucus samples. Our implementation of these techniques, associated documentation and example datasets are made freely available via the PICRUSt software package at http://picrust.github.com .",
"discussion": "Discussion The application of PICRUSt to diverse metagenomic data sets shows that the phylogenetic information contained in 16S marker gene sequences is sufficiently well correlated with genomic content to yield accurate predictions when related reference genomes are available. Our validation results support widespread application of PICRUSt to 16S datasets containing as few as a few hundred sequences per sample, provided that NSTI or a similar measure is used to quantify the expected prediction accuracy. Although PICRUSt’s predictive approach neither precludes nor outperforms deep metagenomic sequencing, it can predict and compare probable functions across many samples from a wide range of habitats at a small fraction of the cost of such sequencing. This approach thus opens up new avenues for tiered, more cost-effective study designs and provides functional insights into the tens of thousands of existing samples for which only 16S data is available. To best leverage the strengths both of (meta)genomic sequencing and of PICRUSt, we recommend its incorporation into marker gene studies using a deliberate, tiered approach. Because phylogenetic dissimilarity among environmental organisms and sequenced genomes (as captured by NSTI) affects PICRUSt accuracy, NSTI values can be calculated from preliminary 16S rRNA data to assess whether reference genome coverage is sufficiently dense to allow for accurate PICRUSt prediction. If adequate reference genomes are not available, additional genome sequences can be collected to fill in phylogenetic “gaps” in the reference database and allow for accurate prediction. This can be performed either through traditional culture-based techniques, single-cell genomic approaches or deep metagenomic sequencing of samples targeted based on 16S data. If NSTI appears sufficient but additional controls are desired, a preliminary set of paired 16S rRNA and shotgun metagenomic samples can be compared using PICRUSt’s built-in tools to empirically test prediction accuracy on the sample types of interest. On the basis of such validations from select samples, PICRUSt can then be used to extend approximate functional information from a few costly metagenomes to much larger accompanying 16S rRNA gene sequence collections. However, the limitations of this approach must be considered in interpreting PICRUSt predictions. For example, only 16S marker gene sequences corresponding to bacterial and archaeal genomes are currently included; thus this version of the system does not infer viral or eukaryotic components of a metagenome. PICRUSt’s ability to detect patterns also depends on the input data used: the software cannot distinguish variation at the strain level if the marker gene sequence used is identical among strains, and it cannot detect genes families (or summarize them into pathways) if those genes are not included in the input genomic data used, or if pathway annotations are currently poor (e.g. for acetogenesis genes). However, because PICRUSt can accept trees produced by alternative marker genes or gene/pathway annotations, users have the flexibility to customize the tool to meet the needs of their system. Although high overall accuracy was obtained despite microbial lateral gene transfer and other processes of gene gain and loss, gene families or pathways (e.g. methane oxidation) with highly variable distribution throughout the tree of life can still lead to incorrect predictions in individual cases. PICRUSt thus provides confidence intervals for each functional abundance prediction that reflect the degree of variation in that function among sequenced phylogenetic neighbors of predicted (meta)genomes, with wide confidence intervals indicating a high degree of uncertainty ( Supplemental Fig. 7 ). If individual gene abundances (rather than aggregate patterns) are of interest, users can choose to either discard predictions with low confidence, or confirm them experimentally. We anticipate several experimental and computational improvements that will further refine the predictive accuracy of PICRUSt. In addition to extending genome coverage and metagenome calibration as above, PICRUSt predictions could also likely be improved by including habitat information in a predictive model. This may provide additional predictive power in that some genes might correlate strongly with environmental parameters as well as phylogenetic similarity to reference organisms 9 , 16 . Modification of prediction methods that incorporate information from partial genome sequences could expand the sensitivity of predictions in under-studied environments by including additional reference gene content information. Finally, as reference genome sequence databases continue to expand and incorporate isolates from ever more diverse environments, the prediction accuracy of PICRUSt will improve by default over time. Predictive metagenomics thus holds the promise of uniting completed genome sequences, 16S rRNA gene studies and shotgun metagenomes into a single quantitative approach for assessing community function."
} | 3,057 |
28680480 | PMC5496137 | pmc | 4,950 | {
"abstract": "Background Crude glycerol in the waste stream of the biodiesel production process is an abundant and renewable resource. However, the glycerol-based industry is usually afflicted by the cost for refinement of crude glycerol. This issue can be addressed by developing a microbial process to convert crude glycerol to value-added chemicals. In this study, Escherichia coli was implemented for the production of n-butanol based on the reduced nature of glycerol. Results The central metabolism of E. coli was rewired to improve the efficiency of glycerol metabolism and provide the reductive need for n-butanol in E. coli . This was carried out in several steps by (1) forcing the glycolytic flux through the oxidation pathway of pyruvate, (2) directing the gluconeogenic flux into the oxidative pentose phosphate pathway, (3) enhancing the anaerobic catabolism for glycerol, and (4) moderately suppressing the tricarboxylic acid cycle. Under the microaerobic condition, the engineered strain enabled the production of 6.9 g/L n-butanol from 20 g/L crude glycerol. The conversion yield and the productivity reach 87% of the theoretical yield and 0.18 g/L/h, respectively. Conclusions The approach by rational rewiring of metabolic pathways enables E. coli to synthesize n-butanol from glycerol in an efficient way. Our proposed strategies illustrate the feasibility of manipulating key metabolic nodes at the junction of the central catabolism. As a result, it renders the intracellular redox state adjustable for various purposes. Overall, the developed technology platform may be useful for the economic viability of the glycerol-related industry. Electronic supplementary material The online version of this article (doi:10.1186/s13068-017-0857-2) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusions The intracellular redox state in microbes is manifested by the interplay of the carbon flux distributed in the central metabolism. In this study, the catabolic flux of glycerol was modulated by manipulating the fueling pathways in the central metabolism. n-Butanol was highly synthesized as a result of the flux redistribution. It suggests that the DHAP, pyruvate, and acetyl-CoA nodes at the junction of the central metabolic pathways play a vital role in the glycerol-based synthesis of n-butanol. In conclusion, our current study and others propose an appealing way to produce a value-added chemical from crude glycerol [ 27 ]. Continued efforts towards the advance of the technology platform may provide a solution to the economic viability of the glycerol-related industry.",
"discussion": "Results and discussion Microaerobic production of n-butanol from glycerol Strain BuT-8 harbors a functional pathway for the synthesis of n-butanol (Table 1 ) [ 12 ]. This heterologous CoA-dependent pathway consists of hbd , crt, and adhE2 from Clostridium acetobutylicum , phaA from Cupriavidus necator , and ter from Treponema denticola (Fig. 1 ). Moreover, it lacks the endogenous adhE , ldhA , pta , and frdA genes responsible for the production of mixed acids. This helps to reduce carbon waste and increase NADH availability. The microaerobic utilization of glycerol in E. coli is far superior to the fermentative metabolism [ 16 ]. Therefore, the microaerobic production of n-butanol from crude glycerol was investigated in this study. A producer strain was developed starting with strain BuT-8. It is well recognized that the efficient production of n-butanol requires more available NADH [ 10 ]. According to the previous study [ 16 ], pflB plays a main role in the oxidation of pyruvate to acetyl-CoA during the microaerobic utilization of glycerol. In contrast to the PflB counterpart, PDH complex (encoded by aceEF - lpdA* ) mediates the pyruvate oxidation with concurrent generation of NADH. Therefore, PDH in the strain was enhanced to compete with PflB for more NADH production. In addition, the gluconeogensis involving fba , fbp , and pgi occurs in E. coli during the microaerobic growth on glycerol [ 16 ]. Accordingly, zwf and pgl were augmented in the strain to direct the gluconeogenic carbon flux into the PP pathway for generation of NADPH. NADPH was then converted to NADH in the strain equipped with the elevated udhA [ 17 ]. Finally, the construction gave rise to strain BuT-12A. Table 1 The E. coli strains applied in this study Strain Characteristic Source BuT-8 Δ frdA ɸ80 attB ::PλP L - ter λ attB ::PλP L - crt \n Δ adhE ::ɸ80 attB ::PλP L - pha - hbd \n Δ ldhA ::λ attB ::PλP L - adhE2 \n 12 BuT-12A as BuT-8 ∆ lpdA λ attB ::PλP L - lpdA \n * \n PλP L - aceEF PλP L - zwf Δ atoD ::PλP L - pgl \n PλP L - UdhA \n This study BuT-12-2 as BuT-12A PλP L - gldA PλP L - dhaKLM \n This study BuT-12-3 as BuT-12-2 Δ zwf \n This study BuT-16 as BuT-12A lacO - gltA \n This study \n lpdA* the mutant, lpdA exhibiting insensitivity to NADH \n Fig. 1 The central metabolic pathways of E. coli connecting glycerol catabolism to n-butanol synthesis. The catabolic route of glycerol includes the glpK - glpD and the gldA - dhaKLM pathways. The heterologous pathway for the synthesis of n-butanol is composed of phaA , hbd , crt , ter , and adhE2 genes ( dotted line ). The genes involved in the metabolic pathways: aceEF - lpdA *: pyruvate dehydrogenase complex; adhE , aldehyde–alcohol dehydrogenase; adhE2 , butyraldehyde–butanol dehydrogenase; crt , crotonse; hbd , 3-hydroxybutyryl-CoA dehydrogenase; ldhA , lactate dehydrogenase; fba , fructose bisphosphate aldolase; fbp , fructose 1,6-bisphosphatase; frdABCD , fumarate reductase; pflB , pyruvate-formate lyase; gltA , citrate synthase; glpF , glycerol facilitator; gldA , glycerol dehydrogenase; dhaKLM , dihydroxyacetone kinase; glpK , glycerol kinase; glpD , glycerol 3-phosphate dehydrogenase; pgi , phosphoglucose isomerase; pgl , lactonase; phaA , acetoacetyl-CoA thiolase; pta , phosphate acetyltransferase; ter , trans-enoyl-CoA reductase; udhA , transhydrogenase; zwf , glucose-6-phosphate dehydrogenase. The undesired genes in the pathways are deleted as marked with “X.” ACE acetate; EtOH ethanol; DHAP dihydroxyacetone phosphate; FDP fructose 1,6-bisphosphate; F6P fructose-6-phosphate; LAC lactate; FOM formate; G6P glucose-6-phosphate; CIT citrate; OAA oxaloacetate; PEP phosphoenolpyruvate; 3-PGA 3-phosphoglyceraldehyde; PYR pyruvate; SUC succinate \n Pure glycerol was first used for illustration. The microaerobic production of n-butanol was carried out using the shake-flask culture of strain BuT-12A while strain BuT-8 served as a control. As shown in Fig. 2 a, b, strain BuT-12A enabled production of 60% more n-butanol than the control strain (2.1 vs. 1.3 g/L) at 24 h of fermentation. The result leads to productivity of 0.09 g/L-h and the conversion yield of 0.23 g/g (Table 2 ), which accounts for 57.2% of the theoretical yield (ca. 0.40 g/g glycerol). Fig. 2 Microaerobic production of n-butanol in strains with the amplification of the fueling pathways. The E. coli strains were grown in M9Y medium containing 20 g/L pure glycerol and the fermentations were carried out for 24 h. The experiments were conducted in triplicate. Keys \n a the fermentation of strain BuT-8; b the fermentation of strain BuT-12A \n Table 2 Summary of the fermentation kinetics for producer strains Strain \n P \n B \n \n Y \n B/G \n Gene product targeted for manipulation PDH Zwf Pgl GldA DhaKLM GltA BuT-8 0.054 0.16 W W W W W W BuT-12A 0.09 0.23 + + + W W W BuT-12-2 0.13 0.29 + + + + + + BuT-12-3 0.14 0.32 + − + + + W BuT-16 0.18 0.35 + + + + + < 0.24* 0.34* The fermentation was carried out with the cell density at OD 550 of 0.2. Strain BuT-16 was grown in M9Y medium containing crude glycerol of 20 g/L for 40 h. The others were cultured on pure glycerol of 20 g/L for 24 h. The development course of producer strains for the production of n-butanol was shown in Additional file 1 : Fig. S1. Note: W, wild type; +, enhancement; −, absence; <, suppression; P \n B , n-butanol productivity (g/L/h); Y \n B/G , conversion yield of n-butanol on glycerol (g/g) * The fermentation was conducted with the cell density at OD 550 of 5 and crude glycerol of 30 g/L for 36 h"
} | 2,065 |
28878236 | PMC5587736 | pmc | 4,951 | {
"abstract": "Excavating sponges are prominent bioeroders on coral reefs that in comparison to other benthic organisms may suffer less or may even benefit from warmer, more acidic and more eutrophic waters. Here, the photosymbiotic excavating sponge Cliona orientalis from the Great Barrier Reef was subjected to a prolonged simulation of both global and local environmental change: future seawater temperature, partial pressure of carbon dioxide (as for 2100 summer conditions under “business-as-usual” emissions), and diet supplementation with particulate organics. The individual and combined effects of the three factors on the bioerosion rates, metabolic oxygen and carbon flux, biomass change and survival of the sponge were monitored over the height of summer. Diet supplementation accelerated bioerosion rates. Acidification alone did not have a strong effect on total bioerosion or survival rates, yet it co-occurred with reduced heterotrophy. Warming above 30 °C (+2.7 °C above the local maximum monthly mean) caused extensive bleaching, lower bioerosion, and prevailing mortality, overriding the other factors and suggesting a strong metabolic dependence of the sponge on its resident symbionts. The growth, bioerosion capacity and likelihood of survival of C . orientalis and similar photosymbiotic excavating sponges could be substantially reduced rather than increased on end-of-the-century reefs under “business-as-usual” emission profiles.",
"introduction": "Introduction To date approximately 30% of the anthropogenic carbon dioxide (CO 2 ) emissions have been absorbed by the oceans 1 . Since the beginning of the Industrial Revolution, ocean pH has decreased by 0.1 units 2 . In combination with ocean acidification, ocean warming caused by the CO 2 -driven enhancement of the greenhouse effect, and eutrophication are threatening the distribution and abundance of coral reefs worldwide 3 – 5 . Projections conclude that these changes will reduce calcium carbonate (CaCO 3 ) accretion on reefs due to increased mortality and decreased calcification potential of reef-building organisms 6 – 8 . Compared to calcification, responses of decalcification and especially biological erosion (bioerosion) to environmental change remain less well studied 9 , even though bioerosion is of equal significance to the carbonate balance on coral reefs 10 , 11 . Bioeroding taxa on coral reefs include internal bioeroders that excavate and inhabit CaCO 3 materials (e.g. certain poriferan, molluscan, annelid, algal, fungal and cyanobacterial genera) and external bioeroders (e.g. certain echinoderm, crustacean, molluscan and fish genera) 12 . As it is predominantly the internal bioeroders that employ chemical means to rework the substrate, and as they are mostly sessile, changes to seawater chemistry may be directly reflected in their bioerosion capacity 11 . Attention on internal bioerosion has focused more specifically on coral-excavating sponges 9 , which often account for 40–70% and up to >90% of macroborer activity on coral reefs 10 (references therein). Excavating sponges influence seawater carbonate cycling and they play important ecological roles by breaking down and sculpting the reef framework, thereby changing the heterogeneity and availability of space 10 . In contrast to many calcifiers, the abundance, activity and competitive vigour of certain excavating sponges has been observed to increase on perturbed reefs (e.g. refs 13 – 16 ). As excavating sponges are demosponges that have a siliceous skeleton, their skeletogenesis is unlikely to be as strongly impacted by carbonate saturation changes as that of calcifiers 17 . However, ocean warming, acidification and eutrophication may still affect these sponges for several reasons. Sponge bioerosion proceeds through chemical etching of CaCO 3 chips, which purportedly involves acid regulation, followed by mechanical removal of the chips from the substrate 18 , 19 . The energetic cost of chemical bioerosion may be reduced in more acidified oceans, as the CaCO 3 dissolution threshold will be more easily met 20 , 21 . As a result, bioerosion may be enhanced on future reefs 22 , as has been suggested after observing bioerosion patterns in CaCO 3 materials from naturally low-pH waters 23 – 26 . Some dominant bioeroding sponges are aggressive space competitors capable of overgrowing living corals 27 , 28 , and reduced competitive pressure caused by increased weakness and mortality of corals may further elevate their abundances 13 , 14 , 29 , 30 . Moreover, bioeroding sponges are filter feeders with an efficient pumping system 31 , 32 and they have been observed to thrive in eutrophic waters 33 – 35 . Eutrophication may result in greater access to food and thereby increase energetic availability, which may not only affect sponge abundances and growth, but may also increase their bioerosion rates 36 . Apart from their heterotrophic filter feeding, certain bioeroding sponges also benefit from photoautotrophic inputs provided by symbiotic dinoflagellates of the genus Symbiodinium \n 37 . The Cliona viridis species complex consists of such species, which are very competitive and destructive to the CaCO 3 framework 10 . Presumably, this is due to the symbiosis providing greater access to energy and hence promoting greater sponge growth, survival and bioerosion 38 – 40 . In comparison to corals, these sponges are also thought to be relatively resilient to bleaching 10 , which suggests that they are well-positioned to dominate newly available space should corals decline 13 – 15 . The potential of increased bioerosion by excavating sponges in changing environments implies a growing threat to the three-dimensional framework of future reefs and the organisms that inhabit them 41 . However, the physiological limits to the enhanced performance of excavating sponges remain to be explored. Research on photosymbiotic bioeroding sponges under experimentally elevated partial pressure of CO 2 ( p CO 2 ) displayed only little adverse response or accelerated bioerosion (reviewed in ref. 11 ). Experiments observing effects of elevated temperature have either not shown a strong response 42 , 43 , or led to bleaching 44 (also BD Ramsby, pers. comm.) or partial necrosis or mortality 45 . Combined exposure of the Indo-Pacific photosymbiotic sponge Cliona orientalis to temperature and p CO 2 anomalies under two future scenarios over eight weeks (Austral spring to summer transition) dramatically enhanced both the growth and bioerosion of the sponge 44 . However, under spring business-as-usual conditions C . orientalis bleached and showed energetic deficiencies 44 , 46 that are likely to lead to mortality in the longer term over summer (JKH Fang, pers. comm.). The current study assessed the capacity of C . orientalis to erode and survive over the summer on future reefs through a 10-week simulation of both independent and concurrent warming and acidification predicted for the year 2100. To explore the influence of local nutrient availability on the outcome of globally changing climate conditions, supplementation of the diet of the sponges with nitrogen-rich particulate organics was also included as a factor. The diurnally and seasonally variable simulation was based on a “business-as-usual” greenhouse gas concentration trajectory called the Representative Concentration Pathway 8.5 (RCP8.5) 1 . As opposed to previous experiments, the individual and combined effects of temperature, p CO 2 and diet supplementation on bioerosion rates, biomass, oxygen flux, carbon flux and survivorship were studied in an orthogonal design. Such designs permit the unravelling of independent effects, which is crucial to fully understand the physiological mechanisms underpinning an organism’s response. The key questions of the current study were: (1) Do simulated warming and/or acidification and/or diet supplementation accelerate or decelerate bioerosion rates and growth of C . orientalis ? (2) What changes to the carbon budget of the sponge drive the observed responses? (3) Which of the three factors or which of the combinations is most likely to impact the sponge’s survival in future oceans under RCP8.5 emissions?",
"discussion": "Discussion Exposure of the photosymbiotic sponge Cliona orientalis to RCP8.5 summer projections for the year 2100 at Heron Island led to decreases in biomass and rates of bioerosion, autotrophy, heterotrophy and survival. These responses were driven by resource deprivation of both the host and the symbiont and were primarily caused by impacts of simulated warming, but also acidification, even when additional heterotrophic food sources were provided. Excavating sponges are often regarded as “winners” on disturbed coral reefs under projected future conditions 11 . However, based on the combined results of the current experiment, under “business-as-usual” CO 2 emissions C . orientalis and possibly other similar photosymbiotic excavating sponges are expected to suffer losses at the end of the current century comparable to those projected for scleractinian corals 22 . Extrapolating measured present-day summer rates of substrate removal by C . orientalis to an annual mean resulted in a total bioerosion rate of 12.4 kg CaCO 3 m −2 year −1 when the diet was not supplemented, which is consistent with earlier results from similarly dense substrates 56 . Our study demonstrated a 35% stimulation of total sponge bioerosion by the availability of food, purportedly also allowing faster expansion as would be expected in nutrient-rich environments such as inshore reefs 35 . Our results do not imply a sustained increase in bioerosion performance by photosymbiotic sponges in future oceans however, mainly due to observed adverse impacts of simulated warming. We suggest that bioerosion rates were mainly reduced by the failure of the photosynthesis of the symbionts, highlighting the importance of the symbiosis to supply energy for bioerosion 38 – 40 . Previously documented temperature effects on bioerosion of Cliona species have shown variable responses 42 , 45 , 71 , 72 . Our results are in accordance with the reduction of bioerosion rates under warming reported from a short-term experiment using the same species 45 . Simulated acidification co-occurred with heterotrophic carbon losses in our study, which may have resulted in the loss of a clear response through bioerosion as opposed to previous studies: Bioerosion by C . orientalis significantly increased under similar acidification in a 3-day closed system experiment (chemical bioerosion measured through alkalinity changes) 45 , in a 10-day flow-through experiment (total bioerosion measured through buoyant mass changes) 20 and in a 8-week flow-through experiment over Austral spring (both of the above measures) 44 . We hypothesize that in our study over Austral summer, the longer-term losses to the sponge’s carbon budget under acidification may have lessened the energy invested into bioerosion, thereby decoupling acidification effects from erosion enhancement. We measured bioerosion through buoyant mass changes, but identifying chemical and mechanical rates in future studies could better elucidate potential acidification impacts. Bleaching of C . orientalis tissues has previously been shown under concurrent simulation of RCP8.5 warming and acidification 44 . Here, we provide evidence that our observation of bleaching was caused by warming above the mean summer water temperature. Whether the bleaching is a result of oxidative stress and mechanisms similar to those established for the cnidarian- Symbiodinium partnership 73 remains to be explored, but our results show that the bleached sponges experience reduced holobiont productivity and increased mortality, as is known for bleached corals 73 . The decrease of photosynthetic activity translates into reduced access to resources, yet even though bleached specimens had lower biomass and no symbiont population, they maintained a considerable respiratory demand (metabolic needs increase inherently with temperature rise), which would have represented an additional resource drain. To date, bleaching of Cliona spp. has been considered a rare event in or ex situ \n 10 . In October 2015 a natural bleaching event was reported for the first time for a clionaid sponge in the lower Florida Keys 74 , but the sponges were only partially bleached and appeared to survive and recover (M Hill, pers. comm.). An unknown encrusting Cliona sp. exhibited impaired photosynthesis during a heating event in March 2013 75 (MMM + 1 °C or more) in depths down to 15 m near Onslow, NW Australia (CHL Schönberg, pers.comm.). In our experiment, the sponges began to pale prior to the full establishment of the 2100 climate scenarios, indicating that at least partial or occasional bleaching could become more common in natural populations under similar warming events before the end of the current century. Acidification stimulated higher photosynthetic rates in unbleached specimens that were not diet-supplemented, possibly because less resources had to be invested into the conversion of HCO 3 \n − to CO 2 for use by the Rubisco enzyme 76 . No such effect was found when the sponges received a supplemented diet since they may have been less dependent on autotrophic inputs, despite stable respiration that could otherwise provide an alternative source of CO 2 to the symbiont 76 . With regards to organic carbon uptake, an increase in heterotrophic feeding of the bleached sponges could have served to compensate for the loss of autotrophic carbon. However, sponges in the heated treatments had reduced rather than increased carbon uptake rates. Filter feeding is an energetically costly process 77 that could be rendered unsustainable in the bleached sponges. Negative effects of temperature on sponge feeding have been reported for the Great Barrier Reef (GBR) sponge Rhopaloeides odorabile , with filtration efficiency, pumping rate and choanocyte chamber density and size reduced at 31 °C (MMM + 2 °C) 78 . POC uptake in our sponges was only insignificantly reduced, yet choanocyte functioning and water pumping may also facilitate DOC uptake 79 , which could explain the losses observed here under simulated warming. Acidification similarly reduced uptake of DOC, which may point towards a trade-off between autotrophy (due to stimulation of photosynthesis under elevated p CO 2 ) and heterotrophy. Alternatively, prolonged high levels of H + may have had deleterious effects on the filter-feeding capacity of the sponge, for example by affecting mitochondrial ATP recycling which is crucial to the flagellar beating 80 . Further investigations are necessary to confirm what caused the observed decrease in carbon uptake under acidification. In treatments with either warming or acidification, dietary supplementation slightly ameliorated the effect of the other factors; the sponges took advantage of the elevated concentration of organics and more organic carbon was incorporated. Sponges in present-day conditions did not respond in the same way. Compared to supplemented sponges that were not in need of food, unsupplemented sponges may have been relatively starved (only particles up to 10 μm in diameter were retained in unsupplemented tanks) and were therefore filtering more actively (i.e. taking up more carbon from the seawater) when assessed over a 24h-period at the height of summer. The assumption of resource deficiency is also supported by the partial mortality found towards the end of the experiment in unbleached sponges that did not receive a supplemented diet. All bleached sponges had lower organic biomass, which will in part explain the reduction in total bioerosion rates and respiration in the heated treatments. Bleached C . orientalis in previous experiments increased biomass 44 , 46 , but this may have been due to a lag effect after an initial stimulation of growth earlier in spring when the symbionts were still present (JKH Fang, pers. comm.). The duration and the severity of our experimental warming went beyond physiological thresholds, and bleached sponges suffered biomass losses due to reduced autotrophy and heterotrophy. Supplementary feeding with protein-rich organics was expected to stimulate growth, since the extent to which the sponge holobiont can utilize photosynthetically transferred or heterotrophically attained carbon beyond its respiratory needs in the oligotrophic reef waters may depend on the availability of commonly limiting nutrients such as nitrogen 81 . No such effect was observed, but instead the resources gained from the extra nutrition may have been redirected towards bioerosion activity. Alternatively, the sponges may have not been nitrogen-limited in the first place, yet the partial mortality observed in the unsupplemented treatments is contraindicative. Even though spiculogenesis is considered energy-demanding 82 , 83 , the mass of siliceous spicule was not affected by any of the treatments, which confirms previous results 17 , 44 . The formation of one demosponge spicule may take approximately one week and an even longer lag period needs to be considered when looking for environmental effects on spicules 84 . A daily net carbon surplus was only realized under present-day temperatures, and that surplus was significantly reduced with acidification, since organic carbon uptake decreased. Overall, the physiological measurements revealed carbon deprivation in the bleached sponges, which explains their subsequent mortality. Heterotrophy by itself does not appear to be sufficient for the energetic needs of C . orientalis (also shown by ref. 46 ). We conclude that the symbiosis between the sponge and its dinoflagellates is neither facultative nor merely beneficial to bioerosion performance alone, but it is vital to the survival of the holobiont in several ways. For marine sponges, interactive effects of environmental factors such as warming, acidification and food availability remain largely unknown and vary between species depending on the natural conditions that they are adapted to 45 , 85 . The lack of and the need for related studies has recently been highlighted in order to quantify present trends of carbonate budgets and to provide modelled trajectories for different scenarios to facilitate management 11 . Our experiment assessed the effects of simulated warming, acidification and diet supplementation on the bioerosion efficiency and survival of C . orientalis separately and in combination. Overall, warming appears to be a more important factor in determining the physiological thresholds of the sponge, and increased bioerosion rates observed in previous short-term acidification-only experiments 11 may not realistically reflect future developments under coexisting ocean warming and acidification. In non-bioeroding sponges from the GBR, the effects of RCP8.5 warming were also shown to be physiologically more important than acidification effects 86 , but Caribbean sponges exposed to similar warming and acidification for 24 days remained largely unaltered 85 . Our study suggests that future climate conditions may temporarily incur increased bioerosion rates at intermediate levels of environmental change, but that ultimately escalating environmental conditions under presently predicted “business-as-usual” fossil fuel usage will cause photosymbiotic clionaids to fail along with other benthic organisms. This supports a parabolic rather than a linear response of bioeroding sponges to future change 11 . Bioerosion and biomass maintenance are not the only energetically-costly activities of C . orientalis ; other costly processes such as reproduction or competition for space 39 could also be impacted by climate change, thereby possibly further reducing the likelihood of survival in this species. Climate simulations on coral reef assemblages from the southern GBR have provided evidence that firstly, there has been little adjustment of corals to changes over the past 100 years, and secondly, 100 years from now corals are unlikely to calcify or survive over RCP8.5 summers (when temperatures and seawater p CO 2 concentrations are at their seasonal highest) and beyond 47 . To date, excavating sponges were often considered more tolerant to future changes than scleractian corals 10 . However, our study stresses that there are limits especially to the temperature but also the p CO 2 conditions that C . orientalis can tolerate, and that, despite subtle benefits from higher food availability, future summers can be expected to have adverse effects on the bioerosion capacity, general physiology and ultimately survival of the sponge. Keeping in mind the limitations of extrapolating from simulations to the field, we nevertheless cannot support that excavating sponges such as C . orientalis are still likely to be “winners” in 2100. We acknowledge that the present results may only be relevant for excavating sponges that are symbiotic with Symbiodinium and that other excavating sponges might show different response patterns. This remains an essential area to explore if we are to gain a complete understanding of the implications of a warmer, more acidic and more eutrophic ocean for these key reef organisms. It is also important to note that our study did not identify the temperature threshold for the loss of C . orientalis from coral reefs like those of Heron Island and the southern Great Barrier Reef, or how well C . orientalis is performing as climate change intensifies in the interim, causing mass coral mortality events 87 and thus increasing the availability of bioerosion substrate."
} | 5,442 |
31459827 | PMC6648901 | pmc | 4,954 | {
"abstract": "Graphene (GE) has attracted significant\nattention on account of\nits unique structure and superior performance, arousing a new research\nfield for materials science. Herein, a novel GE-coated poly(ethylene\nterephthalate) nonwoven (PGNW) hollow tube (PGNW-T) was fabricated\nfor continuous and highly effective oil collection from the water\nsurface. The PGNW was prepared via a dip-spray coating method, which\npossessed superhydrophobicity–superoleophilicity and could\nabsorb a variety of oils or organic solvents with the absorption capacity\n( Q ) value of 18–34 times its own weight. Then,\nPGNW-T was obtained through winding the PGNW on the surface of a porous\npolypropylene hollow tube. As-prepared PGNW-T was competent for dynamic\noil collection with high flux (18 799.94 L/m 2 h),\noutstanding separation efficiency (97.14%), and excellent recyclability\n(>96% after 10 cycles) from the oil/water mixture. In particular,\na miniature device based on as-prepared PGNW-T was developed for continuous\nthin oil film collection, which could dynamically “catch up”\nfloated oils or organic solvents from the water surface. Finally,\nour strategy is extremely facile to scale up, showing its huge potential\napplication in practical oil-spill remediation.",
"introduction": "1 Introduction Over\nthe past few decades, along with the rapid development of\nindustry, oil spill and chemical leakage accidents have a catastrophic\nimpact on the aquatic system and pose a severe threat to human lives. 1 , 2 Consequently, the development of oil-spill removal technology has\nattracted tremendous attention from both research community and general\npublic. There are many traditional methods, such as skimmers, 3 in situ burning, 4 solidifiers, 5 bioremediation, 6 dispersants, 7 and using absorbents, 8 to deal with large amounts of oil spill. Among these technologies,\nporous absorbents have been frequently used for oil-spill remediation\nowing to their low cost and eco-friendly treating process. Unfortunately,\nthese kinds of absorbents, including bentonite, zeolite, corn stalk,\nand activated carbon, 9 − 11 still have limitations for practical oil-spill cleanup\nincluding low selectivity (absorb both oil and water), inferior absorption\ncapacity, and poor recyclability. To overcome the drawbacks\nof traditional absorbents, numerous efforts\nhave been carried out to develop new suoperhydrophobic–superoleophilic\nporous absorbents, including metal oxides, 12 porous boron nitride, 13 carbon based\nsponge, 14 and swellable oleophilic polymer\nabsorbents. 15 In general, they possessed\nexcellent oil absorption capacity ( Q ) and could selectively\nabsorb oil while repelling water, thus showing good performance for\noil collection. However, they could not absorb oil anymore when the\nabsorbents reached the Q value and have to stop and\ndesorb inner oils by squeezing or distillation for reuse. Consequently,\nthis noncontinuous process is still a relatively ineffective method\nalong with the time-consuming and cost-expensive process. More recently,\nfiltration materials with superhydrophobic–superoleophilic\nproperties, such as metallic mesh 16 , 17 and polymeric\nmembranes, 18 , 19 have aroused considerable attention\nfor oil/water separation owing to their continuous separation ability\nand outstanding separation efficiency, but they often suffered low\nflux, high energy consuming, and easy surface fouling due to their\nsmall pore size (<0.3 μm). 20 Furthermore,\nthe oil/water mixture must be collected before filtration. In fact,\nthe spilled oil usually spread rapidly on the water surface and need\nto collect immediately to alleviate further detriment. Hence, these\nfiltration materials have much limitation to deal with large-area\noil spills. Graphene (GE), a two-dimensional structure consisting\nof sp 2 -hybridized carbon, has aroused a lot of attention\nowing to\nits unique properties, such as unique electronic/thermal conductivity,\nultrahigh surface area, and highlighted hydrophobic–superoleophilic\nproperty. 21 − 23 Recently, owing to its outstanding hydrophobicity–superoleophilicity,\na variety of GE-based (GB) materials have been developed for oil removal,\nincluding spongy GE, 24 GE-polymer composite\nthree-dimensional (3D) porous materials, 25 and GE-carbon nanotube foams. 26 For instance,\nZhang et al. 27 fabricated superhydrophobic–superoleophilic\nGE-coated polyurethane (GN@PU) sponge through a dip-coating method.\nThe resultant GN@PU could selectively absorb various oils with the Q value being up to 31 times its weight. Ge et al. 28 prepared GE-coated cotton using the self-assembly\napproach, where GE oxide was coated on cotton, followed by in situ\nreduction by hydrazine hydrate. The GE-coated cotton exhibited high\noil/water selectivity with a Q value of 30 g/g. The\nin situ reduction method to prepare hydrophobic–oleophilic\nGB materials was also researched by Liu et al., 29 Shi et al., 30 and Wang et al., 31 respectively. Generally, these GB materials\npossessed of outstanding absorption and separation properties, but\nuntil now, excess attention has been paid to the Q value, but ignore their cost-expensive and time-consuming recovery\nprocess by mechanical squeezing, burning, or distillation. The high\ncost, complicated fabrication process, and poor interfacial bonding\ntogether with the noncontinuous oil collection process of GB materials\nblock their further application in practical oil spill collection. To deal with the limitations mentioned above, a novel GE-coated\npoly(ethylene terephthalate) (PET) nonwoven hollow tube was prepared\nfor continuous and highly effective oil collection in this study.\nThe poly(ethylene terephthalate) nonwoven (PGNW) with superhydrophobicity–superoleophilicity\nwas prepared through a dip-spray coating method, which can absorb\na variety of oils and organic solvents. PGNW hollow tube (PGNW-T)\nwas formed through winding the PGNW on the surface of the inner support\nof the porous polypropylene (PP) hollow tube, which was competent\nto dynamic oils collected from the oil/water mixture along with high\nflux, outstanding separation efficiency, and excellent recyclability,\neven under corrosive conditions. More importantly, a miniature device\nbased on PGNW-T was innovatively designed for continuous thin oil\nfilm collection, which shows excellent collection ability for various\nfloated oils or organic solvents. Our strategy shows its huge potential\napplication for industrial scaleup in oil-spill remediation.",
"discussion": "2 Results and Discussion 2.1 Characterization of the\nPGNW GE-coated\nnonwovens with a superhydrophobic–superoleophilic surface become\na key factor to make sure that the as-prepared sample could fulfill\noil sorption while repelling water, a rational loading amount of GE\non nonwovens should be defined at first. Hence, the relationship between\nGE loading (measured by W GE / W nonwoven × 100%, where W GE is the weight of GE on the as-prepared sample and W nonwoven is the weight of original nonwoven) and hydrophobicity\nof the sample was investigated. As shown in Figure 1 , when the sample was treated by a successive\nthree times dipping and drying process, the superhydrophobic surface\ncould be acquired with the water contact angle (WCA) being up to 153.2°.\nThe mechanism of the relationship between GE loading and hydrophobicity\nof the sample could be explained as follows: a superhydrophobic solid\nsurface could be obtained by introducing hydrophobic composition and\ncombining with a proper roughed structure. 32 Commercial PET nonwovens are macroscopically roughed surface materials\nwith 3D network pore structure but their surface could be easily wetted\nby water ( Figure S1 ). So the coating of\nthe GE sheet on PET fibers was expected to build a micro–nano\nroughed structure, thus altering the surface wettability of the nonwoven\ndue to the intrinsically hydrophobic property of the GE sheet. Therefore,\ninsufficient GE loading could not afford the requirement of a superhydrophobic\nsurface, and the 4.73 wt % GE loading was enough to form a superhydrophobic\nsurface. In addition, more GE loading may cause some problems such\nas being easily peeled off or block the pores of the sample. Thus,\nthe sample which suffered the three times dipping and drying process\n(GNW) with 4.73 wt % GE loading was chosen for further study and application. Figure 1 Variation\nof the WCA and graphene loading of the as-prepared sample\nwith different dipping times. The morphologies of pristine PET nonwoven, GNW, and PGNW\nare shown\nin Figure 2 . The 3D\nfibrous network structure could be clearly observed for all three\nsamples, which confirmed that the coating of GE sheets and poly(vinylidene\ndifluoride) (PVDF) had little influence on the structure of PET nonwovens.\nHowever, in the high-magnification images, the uncoated PET nonwoven\ncomposed of microfibers exhibited a smooth surface, and after the\ndipping and drying process, a random-oriented GE sheet was uniformly\ncoated on the PET microfibers for forming a dense layer. The surface\nroughness of the fibers dramatically increased, and the micro–nano\nscale roughness of the fibers combined with the macroscopically roughed\nsurface of the PET nonwoven formed a hierarchical structure. This\nhierarchical roughed structure was adequate to acquire a superhydrophobic\nsurface due to the hydrophobicity of the GE sheet. To immobilize the\nGE sheet on the PET fiber, a diluted PVDF solution was further sprayed\nonto the GNW, and the microstructure of the fibers are shown in Figure 2 c1,c2. A thin PVDF\nfilm was covered on the surface of fibers and could be legibly recognized.\nHowever, this thin film did not cover the GE sheet entirely, and some\nrandom-oriented GE sheets were partly exposed on the surface of fibers,\nthus maintaining the roughness of the fibers while ensuring the interfacial\nbonding between the GE sheet and PET fibers. Figure 2 Field-emission scanning\nelectron microscopy (FESEM) images of PET\nnonwoven (a1, a2), GNW (b1, b2), and PGNW (c1, c2) at different magnifications. The hydrophobicity of PGNW was\ndetermined by WCA measurement, and\nthe processes were recorded by a high-speed camera. As shown in Figure 3 a, when a 2 μL\nwater droplet was dropped onto the surface of PGNW, the droplet held\nan almost spheroidal shape and WCA could reach 154.4°, showing\na superhydrophobic surface while further indicating that the PVDF\nfilm did not cover the GE sheet entirely and a hierarchical roughed\nstructure was formed after PVDF coating. The hierarchical rough structure\nof PGNW dramatically reduced the water–solid contact area,\nand an “air cushion” was formed between the surface\nof PGNW and the water droplet, which suggests the formation of a nonwetting\nCassie–Baxter surface on the surface of PGNW. 33 Besides, the water sliding angle (WSA) is another important\nwetting property for selective absorption, as shown in Figure S2 , and the WSA of PGNW was about 8.3°\n(measured by dropping a water droplet on the surface of PGNW, then\ngradually tilting the PGNW until the water droplet began to roll,\nfinally calculating WSA by acrtan (H/L)), 34 further confirming the superhydrophobic surface of PGNW. Meanwhile,\nwhen a kerosene droplet contacts the surface of PGNW, it immediately\npermeated into the fibrous network of PGNW within 250 ms due to the\noleophilicity of the GE sheet on PET fibers, thus showing the superoleophilicity\nof PGNW with the exceedingly low oil contact angle (OCA) of 0°.\nThe surface wettability of pristine nonwoven and PGNW could be seen\nvisually in Figure S1 . Figure 3 Dynamic process of the\nWCA (a1–a3) and OCA (b1–b3)\nof PGNW. X-ray photoelectron spectroscopy\n(XPS) was employed to analyze\nthe surface modification of nonwovens. As shown in Figure 4 , the signals of carbon (C\n1s) and oxygen (O 1s) at about 285 and 532 eV were observed for pristine\nPET nonwoven, which was similar to a previous report. 35 The main elements of GNW were nearly unchanged after dipping\nin GE suspension, but the intensity of the C signal obviously increased\nand the content of carbon increased from 74.18 to 87.24 atom %, which\ncould be attributed to the GE coating on the surface of PET fibers.\nHigh concentration of carbon resulted in low surface energy, which\nwas beneficial for enhancing the surface hydrophobicity. 28 For PGNW, a new F 1s peak emerged, which could\nbe attributed to −CF 2 from PVDF coating. These results\nfurther confirm the GE sheet and PVDF coating layer on the surface\nof PET fibers. Figure 4 XPS spectra of PET nonwoven, GNW, and PGNW. Q value is a key index for an\nabsorbent for oil\ncollection; also, the Q value of PGNW (including\nengine oil, pump oil, soybean oil, palm oil, butyl acrylate, N , N -dimethylformamide (DMF), diesel, kerosene,\nand toluene) was measured and the result was shown in Figure 5 a. The as-prepared PGNW showed\nexcellent absorption capacity (ranging from 18 to 34 times its own\nweight) for various tested oils or organic solvents, particularly\nfor toluene, which is a well-known toxic organic contaminant on the\nwater surface. Previous studies showed that the Q value of the absorbent was related to the density of oils. 14 , 28 During the sorption process, the oils were penetrated into the 3D\nnetwork pores of PGNW, and when the porosity of PGNW was saturated\nby oils, the density of penetrated oils determined the Q value of PGNW, and no swelling phenomenon was found during the whole\nsorption process. Figure 5 (a) Q values of PGNW for various oils\nand organic\nsolvents, (b) Q values of PGNW with different contact\ntimes with a variety of oils. The absorption rate of PGNW for a variety of oils was also\nstudied\nto further evaluate the oleophilicity of as-prepared PGNW. As shown\nin Figure 5 b, it has\nbeen reported that the absorption rate was related to the viscosity\nof oils. 29 For example, it takes more time\nto reach the Q value for high viscosity engine oil than low viscosity\nkerosene, nevertheless, our PGNW could reach the Q value within 20 s for all tested oils ( Figure 5 b), thus showing ultrafast permeation speed\nfor the cleanup of various oils. To further understand the wetting\nbehavior and absorption capacity\nof as-prepared PGNW, an oil–water selective absorption experiment\nwas necessary. As shown in Figure 6 , when a piece of PGNW was placed on the surface of\nthe kerosene/water mixture, kerosene on the water surface was immediately\nabsorbed into the 3D network pores of PGNW and completely removed\nby PGNW within few seconds; then, pure water was left in the beaker,\nwhile no red oil could be observed through the naked eye ( Figure 6 a–d). Similarly,\nwhen a piece of PGNW was immersed into water by external force, a\nmirror-like appearance on the surface of PGNW could be clearly seen,\nwhich was attributed to the air cushion between water and the superhydrophobic\nsurface of PGNW. Once PGNW was forced into contact with chloroform\n(dyed with Sudan III) under water, red chloroform was also immediately\nremoved by PGNW while repelling water ( Figure 6 e–h). From these images, both in air\nor under water superhydrophobic–superoleophilic property of\nas-prepared PGNW can be observed. Combined with prominent oil selectivity,\nPGNW shows a great promising application for highly effective oil\ncollection. Figure 6 Photographs of the selective absorption process of kerosene on\nthe water surface (a–d) and chloroform under water (e–h)\n(both of them were dyed with Sudan III) by PGNW. 2.2 Continuous Oil/Water Separation Testing of\nPGNW-T As mentioned above, the intermittent oil collection\nprocess as an absorbent was still a less-effective way to oil/water\nmixture treatment. Therefore, PGNW-T was designed and prepared for\nachieving continuous oil collection, and the oil/water (include kerosene,\ndiesel, palm oil, and engine oil/water) separation process is shown\nin Figure 7 . When PGNW-T\nwas immersed into an oil/water mixture, the oil immediately permeated\ninto the 3D network pores of PGNW on PGNW-T. The absorbed oil could\nbe continuously collected into a suction flask along the pipes under\nvacuum, which promoted that the oil around PGNW kept continuously\npermeating into PGNW. In this case, continuous oil collection was\nachieved while water could not enter into the 3D network pores of\nPGNW on PGNW-T due to PGNW’s superhydrophobic property. As\nshown in Figure 7 ,\na variety of oils could be continuously collected by PGNW-T, including\nlow viscosity kerosene, diesel, and, especially, high viscosity palm\noil and engine oil. Figure 7 Photographs of continuous oil/water ((a)–(d) represent\nkerosene,\ndiesel, palm oil, and engine oil/water mixture, respectively, kerosene\nwas dyed with Sudan III) separation process using PGNW-T. To further access the oil collection process using\nPGNW-T, oil\nflux and oil/water separation efficiency were measured and the results\nare shown in Figure 8 . The separation efficiencies of PGNW-T for all tested oils were\nall above 90% and could reach up to 97.14% for kerosene; meanwhile,\nthe maximum oil flux was 18 799.94 L/m 2 h. Movie S1 recorded the kerosene/water separation\nprocess using PGNW-T, and it took only 15 s to accomplish 500 mL kerosene/water\nmixture separation, showing ultrafast oil collection ability and excellent\noil/water separation efficiency. Figure 8 Oil/water separation efficiency and flux\nof PGNW-T for various\noils. In the practical oil/water treatment\nprocess, the oil spills or\nindustrial oily wastewater generally contain highly corrosive components\nsuch as acids, alkalies, or salts, which may destroy the microstructures\nof samples and further weaken their oil/water separation ability.\nTo study the corrosion resistance of as-prepared PGNW-T, the oil/water\nseparation test that involves 250 mL of kerosene and 250 mL of 1 mol/L\nHCl (dyed with methylene blue), and NaOH (dyed with rhodamine B) or\nsalt solution was carried out. As shown in Figure 9 , PGNW-T was immersed into kerosene/acid,\nkerosene/alkali, or kerosene/salt solution mixture, respectively.\nOnce the vacuum system starts up, kerosene was continuously collected\nin the suction flask, whereas acid, alkali, or salt solution was repelled\nand remained in the cylinder. Finally, the oil/water separation efficiency\ncould reach up to 97, 97.02, and 97.12%, respectively, and no significant\ndecline of oil/water separation efficiency occurred. Figure 9 Photographs of the (a)\nkerosene/1 mol/L HCl solution (dyed with\nmethylene blue) mixture separation process, (b) kerosene/ 1 mol/L\nNaOH solution (dyed with rhodamine B) mixture separation process,\nand (c) kerosene/1 mol/L NaCl solution mixture separation process,\nwhere the kerosene was dyed with Sudan III. Reusability under different environment conditions (including\nacid,\nalkali, or salt conditions) was also extremely important in practical\napplication. To investigate the recyclability of PGNW-T, the oil/water\nseparation test for kerosene/distilled water, kerosene/1 mol/L HCl\nsolution, kerosene/1 mol/L NaOH solution, kerosene/1 mol/L NaCl solution\nwas repeated 10 times, respectively, and the oil/water separation\nefficiency was calculated for each group of test. As shown in Figure 10 , there is no significant\ndecrease of oil/water separation efficiency during 10 cycles even\nunder strong acid, strong alkali, and salt conditions. The separation\nefficiencies of PGNW-T all remained above 96% after 10 cycles, and\nthe stable oil collection ability of PGNW-T was contributed by the\nstrong interfacial bonding between the GE sheet and PET nonwoven.\nThe result showed not only the excellent recyclability but also superior\ncorrosion resistance ability of PGNW-T. Figure 10 Reusability of PGNW-T\nunder different environment conditions. The above results show the outstanding oil/water separation\nability\nof PGNW-T, but in practical oil-spill accidents, the oil spill generally\nspread around very fast and a thin oil film was easily formed on the\nwater surface. In particular, this thin oil film needs to be collected\nimmediately to alleviate further detriment. Apparently, the vertical\ninsertion of PGNW-T to treat oil spills is a relatively less-effective\nmethod because of its limited contact area with oils. If PGNW-T could\nfloat on the water surface horizontally and move freely on the water\nsurface for oil spill treatment, its valid contact area with oils\ncould be dramatically improved, and that must make our PGNW-T more\nefficacious and more flexible to deal with the thin oil film on the\nwater surface. Herein, a miniature device based on PGNW-T was developed\nto continuously collect thin oil films on the water surface, and its\nstructure diagram is shown in Figure 11 . A water tank (500 mm × 200 mm × 150 mm)\nwhich contained 9000 mL of water and 250 mL of kerosene spreading\non its surface was employed to simulate actual oil spill on the water\nsurface. Two PGNW-Ts were floated on the kerosene/water interface,\nwhich were combined with a joint and connected with a vacuum system,\nand driven by a drive motor. Figure 11 Structure diagram of a continuous thin oil\nfilm collection device. Once startup, the continuous thin oil film collection process\ncan\nbe achieved ( Figure 12 ). Specifically, when turning on the switch, the floated PGNW-T could\ncontinuously collect a thin kerosene film into a suction flask through\nvacuum pressure while repelling water and automatically move on the\nsurface of the kerosene/water mixture to “catch up”\nthe floated oil, at the same time. It only took 150 s to collect 250\nmL of kerosene into the suction flask thoroughly, and finally no red\nkerosene could be observed on the water surface, showing excellent\nthin oil film collection ability. Figure 12 Continuous thin oil film collection process\nby the miniature device. To further simulate the practical oil-spill collection process\nby the miniature device, a variety of oils (including palm oil, kerosene,\nand diesel) and organic solvents (including xylene, butyl acrylate,\nmethyl methacrylate, benzene, toluene, and nonanol), which were commonly\nused in industrial processes and frequently transported through the\nmarine system, or the well-known toxic organic contaminants were all\nchosen to the spilled-oil collection ability testing of our device.\nFirst, 250 mL of the above-mentioned oils were poured onto 9000 mL\nof the water surface and kept stationary for 120 s to form a thin\noil film on the water surface; then, the system was turned on and\nthe thin oil films were collected by the miniature device. As shown\nin Figure 13 , all\nof the tested oils were successfully collected by our device with\nthe separation efficiency of above 83%, and the maximum was up to\n95.3%, confirming the outstanding spilled-oil collection ability of\nour miniature device towards various oils and organic solvents. In\naddition, the miniature device was extremely facile to scale up to\ndeal with practical oil-spill accidents by increased PGNW-Ts to be\nsuitable for different occasions, showing its huge potential application\nin practical oil-spill remediation. Figure 13 Separation efficiency of various oil\nand organic solvents by the\nminiature device. In summary, a novel\nGE-coated PET nonwoven hollow tube was fabricated\nfor continuous and highly effective oil collection from the water\nsurface. The PGNW was obtained via a dip-spray coating method, which\npossessed superhydrophobicity–superoleophilicity and could\nabsorb a variety of oils or organic solvents with the Q value of 18–34 times its own weight, then the PGNW-T was\nformed through winding the PGNW on the surface of the porous PP hollow\ntube. As-prepared PGNW-T was competent for dynamic oil collection\nfrom the oil/water mixture with high flux (up to 18 799.94\nL/m 2 h), outstanding separation efficiency (up to 97.14%),\nand excellent recyclability (>96% after 10 cycles), the exceptional\ncorrosion resistance ability was also provided for practical application.\nFurthermore, a miniature device based on PGNW-T was developed for\ncontinuous thin oil film collection, which shows excellent collection\nability for various floated oils and organic solvents. Our strategy\nis easy to scale up and may provide a new solution for dealing with\nlarge-area oil spill remediation."
} | 6,050 |
30718738 | PMC6361979 | pmc | 4,955 | {
"abstract": "Cyanobacteria are among only a few organisms that naturally synthesize long-chain alkane and alkene hydrocarbons. Cyanobacteria use one of two pathways to synthesize alka/enes, either acyl-ACP reductase (Aar) and aldehyde deformylating oxygenase (Ado) or olefin synthase (Ols). The genomes of cyanobacteria encode one of these pathways but never both, suggesting a mutual exclusivity. We studied hydrocarbon pathway compatibility using the model cyanobacterium Synechococcus sp. PCC 7002 (S7002) by co-expressing Ado/Aar and Ols and by entirely replacing Ols with three other types of hydrocarbon biosynthetic pathways. We find that Ado/Aar and Ols can co-exist and that slower growth occurs only when Ado/Aar are overexpressed at 38 °C. Furthermore, Ado/Aar and the non-cyanobacterial enzymes UndA and fatty acid photodecarboxylase are able to substitute for Ols in a knockout strain and conditionally rescue slow growth. Production of hydrocarbons by UndA in S7002 required a rational mutation to increase substrate range. Expression of the non-native enzymes in S7002 afforded unique hydrocarbon profiles and alka/enes not naturally produced by cyanobacteria. This suggests that the biosynthetic enzyme and the resulting types of hydrocarbons are not critical to supporting growth. Exchanging or mixing hydrocarbon pathways could enable production of novel types of CO 2 -derived hydrocarbons in cyanobacteria.",
"introduction": "Introduction Cyanobacteria are an ancient and highly diverse phylum of photosynthetic prokaryotes with the capability to reproduce using (sun)light and carbon dioxide (CO 2 ) as their primary energy and carbon sources, respectively. Extant cyanobacteria are key players in the global carbon and nitrogen cycles and as primary producers in the open ocean 1 , 2 . It has been known since the late 1960s that cyanobacteria naturally synthesize branched and linear C 15 to C 19 alkanes and alkenes 3 , 4 . The total amount of hydrocarbon produced per cell is around 0.02–0.35 percent of dry cell weight (% DCW) 5 , 6 . Importantly, the carbon chain lengths of most cyanobacterial hydrocarbons are within the range of fuels used currently in jet (C 8 –C 16 ) and diesel engines (C 9 –C 36 ) 7 , 8 , and there is, accordingly, interest in developing these organisms as bio-production platforms for CO 2 -derived biofuels 9 – 13 . In cyanobacteria, alka/ene hydrocarbons most likely localize to the thylakoid and plasma membranes 14 where they serve numerous important roles. Knock-out strains that are unable to produce alka/enes exhibit a variety of phenotypes including greatly decreased growth rates, increased cyclic electron flow in the photosystems, changes in cell morphology, and salt sensitivity 14 – 16 . Typically, the slower growth of the knockout strains is more pronounced at lower growth temperatures 15 , 17 . Although many of the impacts of hydrocarbons on cells are known, the exact mechanism by which they produce these effects in vivo remains enigmatic. Two distinct biosynthetic pathways for alka/enes have been characterized in cyanobacteria 18 , 19 (Fig. 1 ). The substrates for these native pathways are fatty acids (FAs) covalently linked to acyl carrier protein (fatty acyl-ACPs) produced by the FA synthase pathway 20 . One type of hydrocarbon biosynthesis pathway relies on two enzymes to convert acyl-ACP to alkane (Fig. 1 , blue path). The first enzyme, acyl-ACP reductase (Aar), reduces substrate to the corresponding free fatty aldehyde using NAD(P)H as a cofactor 18 , 21 . The second enzyme, aldehyde deformylating oxygenase (Ado), uses molecular dioxygen and NAD(P)H to deformylate the Aar-produced fatty aldehyde to the corresponding C n-1 alkane 18 , 22 , 23 . Depending on the substrate captured by Aar, this pathway can result in the production of methyl-branched alkanes or alkenes with an internal desaturation 5 . Figure 1 The two major alka/ene hydrocarbon biosynthetic pathways in cyanobacteria are the Ols (red) and Ado/Aar (blue) pathways. Fatty acid (FA) biosynthesis or activation from free fatty acids (FFA) provides acyl-ACP substrates for both pathways. The second type utilizes a large, modular multi-domain enzyme called the olefin synthase (Ols) that is related to polyketide synthases 19 (Fig. 1 , red path). Ols captures fatty acyl-ACP substrates and extends the carbon chain via a decarboxylative condensation reaction with malonyl-CoA to generate a covalently bound C n+2 acyl chain. This intermediate is decarboxylated to generate the C n+1 alkene product with a terminal double bond, an α-olefin 19 , 24 . Enzymatic desaturation of acyl-ACP prior to capture by Ols results in production of doubly-desaturated 1,14-nonadecadiene 17 . The ratio of 1,14-nonadecadiene to 1-nonadecene is dramatically higher at lower growth temperatures due to increased production of singly desaturated acyl-ACP precursor 17 , 19 . Genomic comparisons of cyanobacteria have shown that Ols or Ado/Aar genes are present in nearly all characterized members, though the Ado/Aar pathway is much more prevalent than Ols 5 . Interestingly, it was found during these studies that no cyanobacterium encodes genes for both of the pathways 5 . The reason for this apparent mutual exclusivity remains unknown but it could be due to a negative selection mechanism that precludes the co-existence of Ado/Aar and Ols in a cell. In order to gain insights into the compatibility between the different hydrocarbon pathways, we first studied the effects of co-expressing the Ols and Ado/Aar pathways in the model unicellular cyanobacterium Synechococcus sp. PCC 7002 (S7002). We then tested whether non-native hydrocarbon biosynthetic pathways could functionally substitute for Ols in a knockout strain of S7002. Our data show that Ado/Aar from other model cyanobacteria as well as disparate hydrocarbon biosynthetic enzymes from non-cyanobacteria including Pseudomonas and eukaryotic green algae are able to replace Ols and rescue the slow growth of alkene-deficient S7002. This suggests that these diverse pathways are interchangeable to some extent. Our results also show that the biosynthetic source of alka/enes is not critical to serving their role and various types of alka/ene hydrocarbons are able to support growth of S7002. This work implies malleability in bacterial hydrocarbon synthesis and could enable strategies for expanding the biosynthetic range of accessible alka/ene biofuels from CO 2 using cyanobacteria.\n\nIntroduction of the Ado/Aar genes into S7002 To express the Ado/Aar pathway in S7002, we constructed a series of autonomously replicating plasmids derived from the broad host-range vector pVZ321 25 . The Materials and Methods section provides details on plasmid assembly and Supplementary Figure S1 diagrams the construction workflow. Briefly, we first generated a derivative vector pSL3067 from the broad-host range plasmid pVZ321 that contains a kanamycin resistance cassette and the mobilization/replication genes needed for passage through Escherichia coli and for conjugation into S7002. Alkane expression vectors were then generated from pSL3067 using Gibson Assembly 26 . Genotypes for strains characterized in this study are presented in Supplementary Table S1 and oligos in Supplementary Table S2 . To construct the initial set of vectors, we cloned the alkane biosynthetic operons from the two model cyanobacteria Synechococcus elongatus PCC 7942 (S7942) and Synechocystis sp. PCC 6803 (S6803) for introduction into S7002 via tri-parental mating. The empty plasmid pSL3067 was introduced into S7002 to be used as a control (strain 7002::control). The S6803 ado and aar genes were introduced into S7002 on plasmid pSL3071 to make strain 7002::6803alk. The alkane operon from S7942 containing genes ado , aar , and accA was initially introduced in vector pSL3070 under control of the native S7942 promoter. However, this strain did not produce detectable amounts of non-native hydrocarbons under any growth conditions. We determined that this was due to the S7942 alkane gene promoters being inactive in S7002 (Supplementary Fig. S2 ) and re-designed the expression vector by placing the genes under control of the synthetic promoter P trc1O 27 . This vector, pSL3072, was introduced into S7002 to make strain 7002::trc7942alk.",
"discussion": "Discussion The ability of disparate hydrocarbon biosynthetic pathways from bacteria and eukaryotes to conditionally substitute for Ols in S7002 has several important implications. The effect of the non-native pathways on S7002 growth rates and hydrocarbon pools and growth rate is critically dependent on growth temperature. With the exception of FAP, growth at 27 °C dramatically reduced the production of alka/enes by the non-native pathways in S7002 by 84–98%. It is likely that there is a threshold quantity of hydrocarbons needed per cell in order to facilitate growth and the low quantities produced by most of the ΔOls strains at 27 °C are insufficient to support growth. There are several plausible explanations for the reduction in alka/ene production. The first is that Ado/Aar preferentially react with saturated acyl-ACP substrates that are in increased supply at 38 °C. This preference is evidenced, in part, by the ratio of 5-heptadecene to heptadecane (0.07, 0.05) vs. the ratio of 1,14-nonadecadiene to 1-nonadecene (0.24, 0.37) produced in 7002::6803alk and 7002::trc7942alk at 38 °C. Due to this preference, the increased desaturation of acyl-ACPs in S7002 at lower temperature reduced the impact of Ado/Aar due to a shift to less favorable desaturated substrates. This essentially has the effect sequestering a pool of desaturated acyl-ACPs for reaction with Ols and not Ado/Aar. The production of fatty acids for UndA could be similarly affected by the increased acyl-ACP desaturation. We observed mainly production of 1-tridecene and 1-pentadecene in ΔOls::UndA, suggesting that the enzyme is reactive towards tetradecanoic acid and hexadecanoic acid and not their internally desaturated analogues which would have resulted in production of alkadienes. Since myristic acid is the major substrate for UndA at 38 °C, increased desaturation of C 14 -ACP to C 14:1(Δ9) -ACP at 27 °C would reduce production of the corresponding saturated fatty acids and thereby limit production of UndA-derived alkenes. In contrast, FAP is able to produce predominantly 5-heptadecene at 27 °C by reaction with C 18:1(Δ9) fatty acid, giving it an advantage over UndA since it is more compatible with the internally desaturated substrates. The second reason for the reduction may be tied to a differential availability of reducing equivalents for the enzymes at 27 °C vs. 38 °C. Previous reports show that S7002 still contains quantities of palmitic and stearic acid at 22 °C 33 which would be viable substrates for FAP and, if activated to acyl-ACPs, for Ado/Aar. The delivery of reducing equivalents to Ado and/or Aar could be compromised at the lower temperature due to photoinhibition and/or redox imbalance at 27 °C. We see some evidence for this. At 27 °C, production of 1-nonadecene in 7002::trc7942alk is reduced relative to the control, possibly due to capture by Aar. However, there is little production of Ado/Aar -derived alka/enes (Fig. 3H ). If electron delivery to Ado is compromised at 27 °C, the coupling between Aar acyl-ACP reduction and Ado decarbonylation would result in incomplete turnover of acyl-ACPs into alka/enes. Furthermore, when 7002::trc7942alk was cultured under low light conditions in shake flasks at 25 °C, it contained eight- to ten-fold more non-native hydrocarbons. The decreased light intensity and overall slower growth rate of the strains could have resulted in less photoinhibition and higher Ado/Aar activity. While the electron-donating co-substrate for the UndA reaction is still in question 30 , a similar inefficient delivery of reducing equivalents to the enzyme at 27 °C could contribute to its low activity at this temperature. Interestingly, the use of light as a substrate in the decarboxylation reaction by FAP obviates the need for electron delivery by ferredoxin or other redox partners 31 . This could provide another explanation for this enzyme’s better relative performance at 27 °C. The story is quite different at the higher growth temperature. The improved growth exhibited by the Ols-replacement strains compared to the knockout indicates that these disparate biosynthetic pathways are able to support healthy growth of S7002 at 38 °C. This observation has several implications. First, the biological function of hydrocarbons is not necessarily tied into the specific structure of the alka/ene molecules. Expression of UndA and FAP in S7002 afforded unique hydrocarbon profiles not observed in any other wild cyanobacteria 5 . Hydrocarbons with various chain lengths from C 13 to C 17 containing different desaturation profiles are able to serve the same role as the native C 19 α-olefins in S7002. Secondly, it means the producing enzyme and the source of the substrates are not critical. The biosynthetic mechanisms of these enzymes are distinct, as are their preferred substrates. The Ols and Ado/Aar enzymes use acyl-ACP substrates whereas UndA and FAP react with fatty acids. Since free fatty acids are activated to acyl-ACPs in cyanobacteria, the source for the UndA and FAP substrates is likely to derive from turnover of membrane lipids 34 . Capture and consumption of these molecules apparently does not greatly impact growth of S7002 compared to Ado/Aar and both types of enzymes are able to support growth of S7002 at 38 °C. One of prevailing hypotheses for the role of alka/enes is that they partition into the interior of membrane bilayers and thereby afford cells a means to modulate membrane fluidity and structure in response to a variety of stimuli including light and temperature changes, during cell division, and fluctuating thylakoid lumen pressure resulting from proton pumping 14 , 15 , 35 – 37 . The ability of different types of hydrocarbons to support growth of S7002 in consistent with the hypothesis that these molecules primarily serve basal structural roles in cells. Although different kinds of hydrocarbons can serve this role, regulated and steady production of alka/enes in response to changing growth conditions is critical. In a less controlled growth environment, the native Ols pathway is likely to afford an advantage over Ado/Aar, UndA, and FAP for S7002 since it is able to continue producing hydrocarbons in response to changing temperature. This provides a plausible explanation for why Ols is maintained over Ado/Aar in S7002 and that enzyme compatibility with endogenous acyl-ACP and fatty acid profiles may be an important determinant of which pathway is preferred in a cyanobacterial cell. In conclusion, we found that Ado/Aar and Ols pathways are not mutually exclusive and S7002 is able to support the co-existence of both pathways. Far from being incompatible, alka/ene biosynthetic pathways from disparate organisms and domains of life are able to support growth of S7002. While the native Ols pathway in S7002 does display an advantage over the exogenous pathways, we hypothesize that this has more to do with a compatibility of Ols with the native acyl-ACP substrates over a variety of growth conditions rather than a strict requirement of α-olefin production in cells. These data suggest that the specific biosynthetic mechanism and enzyme pathway for hydrocarbon production is not critical as long as sufficient hydrocarbon production is maintained in response to changing environmental conditions. The unique hydrocarbon profiles of ΔOls::UndA and ΔOls::FAP evidence an opportunity to expand the hydrocarbon production range of cyanobacteria to chain lengths not accessed by WT strains. Our experiments only tested a subset of the hydrocarbon biosynthetic pathways found in nature 38 , 39 , and it will be interesting to determine to what extent these diverse pathways are fundamentally interchangeable in vivo ."
} | 4,017 |
21358041 | null | s2 | 4,956 | {
"abstract": "Bacterial colonies often exhibit complex spatio-temporal organization. This collective behavior is affected by a multitude of factors ranging from the properties of individual cells (shape, motility, membrane structure) to chemotaxis and other means of cell-cell communication. One of the important but often overlooked mechanisms of spatio-temporal organization is direct mechanical contact among cells in dense colonies such as biofilms. While in natural habitats all these different mechanisms and factors act in concert, one can use laboratory cell cultures to study certain mechanisms in isolation. Recent work demonstrated that growth and ensuing expansion flow of rod-like bacteria Escherichia coli in confined environments leads to orientation of cells along the flow direction and thus to ordering of cells. However, the cell orientational ordering remained imperfect. In this paper we study one mechanism responsible for the persistence of disorder in growing cell populations. We demonstrate experimentally that a growing colony of nematically ordered cells is prone to the buckling instability. Our theoretical analysis and discrete-element simulations suggest that the nature of this instability is related to the anisotropy of the stress tensor in the ordered cell colony."
} | 321 |
25469170 | PMC4211721 | pmc | 4,958 | {
"abstract": "Evolution has resulted in thousands of species possessing similar metabolic enzymes with identical functions that are, however, regulated by different mechanisms. It is thus difficult to select optimal gene to engineer novel or manipulated metabolic pathways. Here, we tested the ability of molecular evolutionary analysis to identify appropriate genes from various species. We calculated the fraction of synonymous substitution and the effective number of codons (ENC) for nine genes stemming from glycolysis. Our research indicated that an enzyme gene with a stronger selective constraint in synonymous sites would mainly regulate corresponding reaction flux through altering the concentration of the protein, whereas those with a more relaxed selective constraint would primarily affect corresponding reaction flux by changing kinetic properties of the enzyme. Further, molecular evolutionary analysis was investigated for three types of genes involved in succinate precursor supply by catalysis of pyruvate. In this model, overexpression of Corynebacterium glutamicum pyc should result in greater conversion of pyruvate. Succinate yields in two Escherichia coli strains that overexpressed each of the three types of genes supported the molecular evolutionary analysis. This approach may thus provide an alternative strategy for selecting genes from different species for metabolic engineering and synthetic biology.",
"introduction": "Introduction The synonymous codon usage reflects a balance between selective and neutral evolutionary forces. In microorganisms, for a given set of synonymous codons, the relevant tRNAs are not equally abundant (Ikemura 1985 ). Therefore, there must be a preferred set of codons to match the most abundant tRNAs improving translational efficiency (Kanaya et al. 1999 , 2001 ; Novoa and Ribas de Pouplana 2012 ). Moreover, synonymous sites may affect the secondary structure of mRNA conferring resistance to premature degradation and therefore selection might against synonymous substitutions that disrupt base pairing (Shen et al. 1999 ; Duan et al. 2003 ; Capon et al. 2004 ; Chamary and Hurst 2005 ; Novoa and Ribas de Pouplana 2012 ; Shabalina et al. 2013 ). Thus, particular codons may be selected to optimize structure or stability. And the stability of mRNA also influences the concentration of protein. The aforementioned factors have presented a level of convergent evidence that, in microorganisms, most synonymous substitution may slightly alter the quantity of proteins by altering the translational process. Recently, Ma et al. ( 2010 ) reported that synonymous substitution may be under more relaxed selective pressure when synonymous substitution occur in genes encoding enzymes compared with that occur in genes encoding nonenzyme genes, as regulatory mechanisms may differ between enzyme and nonenzyme genes. The function of enzymes can be regulated by altering their kinetic properties, whereas the function of nonenzyme genes is regulated primarily by altering their expression levels. However, by comparing nearly 70 000 genes, Zhang et al. (X. H. Ma, X. B. Zhang, B. Y. Wang, , Y. F. Mao, Z. W. Wang, T. Chen and X. M. Zhao, personal communication) found that the selective constraint on synonymous sites of some enzyme-encoding genes differs from that of other enzyme-encoding genes, suggesting that some enzymes are primarily regulated by their concentration, whereas other enzymes are regulated by altering their kinetic properties (Wright and Rausher 2010 ). The different pattern in enzyme regulation may inform the choice of the optimum enzymes for engineering of microorganisms for target chemical production. In many cases, enhanced production of a target chemical requires the introduction/activation of additional enzymes and/or pathways in the host strain. However, many metabolic genes are highly conserved across many species and there is functional redundancy of metabolic genes within species, it can be difficult to select the most suitable donor species and genes to maximize production. The in vitro kinetic properties of enzymes from different species provide some selection criteria (Scheer et al. 2011 ), but differences between in vivo and in vitro kinetics are not yet well understood because of the intrinsic complexity of the intracellular environment. Consequently, time-consuming and costly trial-and-error approaches are still widely used (Zheng et al. 2012 ; Meiswinkel et al. 2013 ). Due to increases in global energy consumption and supply concerns, much attention has focused on engineering microorganisms for the production of biofuels, pharmaceuticals, plastics, food products, and more (Balzer et al. 2013 ). Concerning maximizing yields of bio-based productions by microorganism, as a case study, evolutionary analyses were attempted to assist in engineering Escherichia coli for efficient conversion of glucose to succinate. Succinate, a C4-dicarboxylic acid, has a wide range of applications in fields as diverse as agriculture, medicine, polymer synthesis, and chemistry (Ma et al. 2013 ). Escherichia coli is one of the most promising succinate producers because of its well-studied genetics and easy manipulation. Succinate is not the primary product of pyruvate conversion in E. coli under aerobic or anaerobic conditions. Thus, it is necessary to redirect metabolic resources to increase succinic acid production and to reduce the formation of other by-products. Toward this goal, a number of metabolic engineering approaches have been developed to increase succinate production in E. coli (Jantama et al. 2008a ; Jantama et al. 2008b ; Cao et al. 2011 ; Yu et al. 2011 ; Balzer et al. 2013 ; Ma et al. 2013 ; Zhu et al. 2013 ). These approaches mainly rely either on blocking competitive or succinate degradation pathways (Chatterjee et al. 2001 ; Zhang et al. 2009 ; Singh et al. 2010 ; Balzer et al. 2013 ), or on activating endogenous or heterologous enzymes (Singh et al. 2010 ; Ma et al. 2013 ; Zhu et al. 2013 ) to direct the carbon flow to oxaloacetate (OAA) or malate, from which succinic acid can be produced. In E. coli , the formation of succinate mainly occurs via the carboxylation of phosphoenolpyruvate (PEP) to form oxaloacetate. Half of the glucose-derived PEP is consumed by the PEP: carbohydrate phosphotransferase system (PTS) to transport glucose across the cell membrane. This metabolic rigidity can be overcome by overexpressing PEP carboxylase (PEPC) and/or PEP carboxykinase (PEPCK) or by inactivating genes of the PTS, but the consequent reduction in glucose absorption results in a slower growth rate and, therefore, less overall succinate productivity or production. Alternatively, succinate production can be significantly enhanced by the overexpression of native E. coli malic enzyme or non-native pyruvate carboxylase (encoded by pyc ), both of which convert pyruvate to 4-carbon succinate precursors (Fig. S1). However, because both malic enzyme and pyruvate carboxylase genes have been identified in many microorganisms, such as Lactococcus lactis , Rhizobium etli, and Bacillus subtilis , it is difficult to determine the appropriate gene donor species. Overexpression of pyc gene from L. lactis in SBS110MG, an E. coli strain whose adhE and ldhA genes were removed, increased succinate yield from 0.2 to 1.3 mol/mol (more than six times) (Sánchez et al. 2005 ). But, overexpression of pyc gene from R. etli resulted in a 2.7-fold enhancement in succinate production (Gokarn et al. 2001 ). Overexpression of malic enzyme gene from E. coli in NZN110 led its succinate yield increased two to three times (Stols and Donnelly 1997 ; Hong and Lee 2001 ). These different results may derive from the differences in genotype of host strains, modification of overexpressed genes, and process of cultivation. What is more possible, however, the difference in the aspects of succinate yield or production stems from the enzyme activity regulation of overexpressed genes from various donor species. In the present work, we tested the ability of evolutionary genetics information to inform the choice of the optimum enzymes for converting pyruvate to succinate from three species. We analyzed the evolutionary selection on synonymous sites of thirteen genes from three microorganisms, including nine glycolysis genes form B. subtilis , Corynebacterium glutamicum , and E. coli , three types of genes converting pyruvate to OAA or malate, that is, pyc from B. subtilis and from C. glutamicum , the NAD-dependent malic enzyme gene ( maeA) from E. coli , and the NADP-dependent malic enzyme gene ( maeB) from E. coli . Based on our hypothesis, overexpression of pyc from C. glutamicum would be most effective in converting pyruvate to the succinate precursor 4-carbon metabolites. This prediction was supported by heterologous overexpression of these genes for succinate yield in two E. coli strains.",
"discussion": "Discussion Large-scale genome sequencing of microorganisms now allows researchers to apply sequence-based evolutionary analysis approaches to microbial ecology questions. To our knowledge, this is the report of the use of evolutionary analyses to assist in the engineering of microorganisms for chemical production. Our Ks and ENC analyses suggested that synonymous sites within pfkA and pyk genes from three microorganisms are under more relaxed selective constraint than those within genes for the other glycolysis enzymes that we examined. A possible explanation for this difference in selective constraint is that pfkA and pyk are primarily regulated by altering their kinetic properties, whereas the activities of other glycolysis genes are mainly regulated at the level of protein concentration. Further investigation of four enzymes that catalyze the conversion of pyruvate to malic acid and OAA and eventually to succinate demonstrate that pyc of C. glutamicum is under the highest selective constraint, followed by maeB of E. coli , maeA of E. coli , and pyc from B. subtilis . This result implies that C. glutamicum PYC is primarily regulated at the level of protein concentration, whereas the activity of B. subtilis PYC is mainly regulated by altering its kinetic properties. In the Michaelis-Menten model, if the function of an enzyme is primarily regulated by altering its kinetic properties, it will have a wider K m range, and thus, synonymous substitution will have little impact on the flux of the reaction. In contrast, if the function of an enzyme is primarily regulated by its concentration, it will have a narrower K m range, and synonymous substitutions that affect intracellular concentration of the enzyme will alter the flux of the reaction. Our rare-codon analysis, SDS-PAGE experiments, and results with E. coli strains that overexpressed pyc from C. glutamicum and B. subtilis support the hypothesis that the activities of these two enzymes are regulated by different mechanisms. The activity of pyc from C. glutamicum appears to be primarily regulated by changing the concentration of the enzyme, whereas the activity of pyc from B. subtilis appears to be regulated by altering its kinetic properties. In this study, the two pyc genes and two malic enzyme genes that we overexpressed in E. coli were under the same heterologous transcriptional and translational control. Therefore, differences in succinate production from the four overexpressing strains are likely to be due to the different kinetic properties of these enzymes."
} | 2,900 |
35362167 | PMC9287070 | pmc | 4,961 | {
"abstract": "Abstract Historical contingency has long figured prominently in the conceptual frameworks of evolutionary biology and community ecology. Evolutionary biologists typically consider the effects of chance mutation and historical contingency in driving divergence and convergence of traits in populations, whereas ecologists instead are often interested in the role of historical contingency in community assembly and succession. Although genetic differences among individuals in populations can influence community interactions, variability among populations of the same species has received relatively little attention for its potential role in community assembly and succession. We used a community‐level study of experimental evolution in two compositionally different assemblages of protists and rotifers to explore whether initial differences in species abundances among communities attributed to differences in evolutionary history, persisted as species that continued to evolve over time. In each assemblage, we observed significant convergence between two invaded treatments initially differing in evolutionary history over an observation period equal to ~40–80 generations for most species. Nonetheless, community structure failed to converge completely across all invaded treatments within an assemblage to a single structure. This suggests that whereas the species in the assemblage represent a common selective regime, differences in populations reflecting their evolutionary history can produce long‐lasting transient alternative community states. In one assemblage, we also observed increasing within‐treatment variability among replicate communities over time, suggesting that ecological drift may be another factor contributing to community change. Although subtle, these transient alternative states, in which communities differed in the abundance of interacting species, could nonetheless have important functional consequences, suggesting that the role of evolution in driving these states deserves greater attention.",
"introduction": "INTRODUCTION Historical contingency is of long‐standing interest to both evolutionary biologists and ecologists (e.g., Fukami, 2015 ; Gillespie, 2004 ; Lenski, 2017 ; Travisano et al., 1995 ), albeit from widely different perspectives. Evolutionary biologists typically consider whether historical contingency, mostly in the form of chance mutations, will influence divergence or convergence of populations of a single species, often under identical conditions imposed by selection experiments (Lenski, 2017 ; Simões et al., 2008 ). Ecologists, conversely, have more frequently focused on the role of historical contingency in community dynamics by considering how chance differences in the timing of species arrivals can have effects on community properties (e.g., Li et al., 2016 ). These approaches diverge in the level of biological organization on which they focus. Ecologists concentrate on the interspecific level, typically ignoring the role of historical contingency at the intraspecific level, whereas evolutionary biologists frequently begin with a single genetic clone of an asexually reproducing organism, but rarely consider the consequences of historical contingency in a larger community context (but see Meyer et al., 2012 for a two‐species community). Interest in historical contingency among evolutionary biologists concerns whether the sequence and identity of mutations over time result in evolutionary changes that are path dependent (and consequently, divergent), or whether mutations interact with a common selective regime to produce largely repeatable (convergent) changes in phenotypes (e.g., Herron & Doebeli, 2013 ). Specifically, even if every sequence of evolutionary change is unique, it is still possible to reach the same peak in a fitness landscape via different underlying mutations, resulting in phenotypic convergence, despite different evolutionary pathways (Lenormand et al., 2009 ). Historical contingency may, however, result in evolutionary trajectories that reach different fitness peaks (Lenski & Travisano, 1994 ); divergence can arise from factors such as different mutations or epistatic interactions among mutations (Meyer et al., 2012 ). Convergence has been shown for traits directly related to fitness (i.e., adaptation) (Colosimo et al., 2005 ; Gilchrist et al., 2004 ), whereas historical contingency may be more important for traits that are less tightly tied to fitness (Travisano et al., 1995 ). In community ecology, historical contingency is primarily of interest for its role in generating priority effects during community assembly (Chase, 2003 ; Fukami, 2015 ) and successional dynamics (Meiners et al., 2015 ), and for its role in determining species coexistence (Grainger et al., 2019 ; Letten et al., 2017 ). Priority effects do not always have long‐lasting effects on community assembly, but can persist for multiple generations in some situations (Toju et al., 2018 ). They can result in the divergence of community structure and function, including alternative transient or stable states, or compositional cycles (Fukami, 2015 ). Similarly, successional dynamics can result in community divergence (e.g., Taylor & Chen, 2011 ), convergence (e.g., Alday et al., 2011 ), or more complex patterns (e.g., del Moral & Lacher, 2005 ). Historical contingency in community ecology has traditionally focused on how the different timing (order) of species arrivals affects community patterns, but not the potential consequences of genetic variation within species (Chase, 2003 ; Fukami, 2015 ; Violle et al., 2012 ). Only more recently have ecologists begun to investigate the importance of intraspecific diversity (e.g., Jung et al., 2010 ; Laughlin et al., 2012 ; Siefert, 2012 ; Zee & Fukami, 2018 ) or eco‐evolutionary dynamics in community assembly (e.g., Knope et al., 2012 ; Kremer & Klausmeier, 2017 ; Lee et al., 2012 ; Urban & De Meester, 2009 ), species coexistence (e.g., Klauschies et al., 2016 ; Kremer & Klausmeier, 2013 ), or community properties such as resilience (e.g., Barabás & D'Andrea, 2016 ). Nonetheless, work with foundation species (i.e., plant species that define habitats sensu Whitham et al., 2012 ) suggests that different genotypes of the same species can be an important determinant of community assembly (e.g., Keith et al., 2017 ; Lamit et al., 2016 ). Few studies have examined the possibility that differing evolutionary histories of interacting species within a community may produce communities that differ in terms of the abundance of species (and therefore, community composition) (but see Zee & Fukami, 2018 for work with competing pairs of Pseudomonas fluorescens strains in which evolutionary history was manipulated). Even if communities containing populations of species that differ in evolutionary history eventually converge on a single community state (again, in terms of the abundance of species within the community as opposed to species richness or the identity of species present), evolutionary history may be one factor that could govern the emergence of transient alternative states. Transient alternative states can persist for many generations (Fukami & Nakajima, 2013 ), frequently for longer than the interval between disturbances that restart successional dynamics (Fukami & Nakajima, 2011 ). The emergence of transient alternative states may consequently constitute an important mechanism that maintains regional species diversity (Fukami & Nakajima, 2013 ). Here we take advantage of an experiment that assessed the importance of evolutionary history in driving the outcomes of biological invasions (Faillace & Morin, 2016 ) to examine whether communities with differences in the recent evolutionary history of populations would exhibit persistent divergence or convergence among communities over time. We constructed communities of two assemblages of protists and rotifers (from this point forwards termed Assemblages A and B) with combinations of species that differed in their exposure to, and evolutionary history with, a designated invading species. To determine if post‐invasion evolutionary history had the ability to alter the trajectory of community development we compared species performance over time (measured as mean abundance) from treatments with naïve (i.e., with no previous history of evolutionary experience with an invader) or evolved (i.e., with a history of potential post‐invasion evolution) populations of invaders and residents. In this way we determined whether abundances of populations from initially divergent treatments became more similar over time, regardless of the initial differences in evolutionary history of invaders and residents. We reasoned that if post‐invasion interactions represent a common selective regime across all invaded treatments within each assemblage, early differences among invaded treatments might become less pronounced over time as initially naïve populations evolved following invasion during our observation period, due to a tendency to converge toward a single community end state under a shared, post‐invasion, selective regime. We predicted that if post‐invasion evolution occurred in a rapid and mostly deterministic, repeatable fashion, then naïve populations of residents and invaders would evolve to effectively become less naïve over time. In this case, we would expect to see increased similarity among invaded communities after tens of generations of interactions, compared with the initial post‐invasion dynamics assessed when species had just experienced different selective regimes (before naïve populations might be expected to undergo much evolution). Stochastic processes (i.e., drift) could also lead to increased variation among replicates within treatments depending on the degree of repeatability in evolutionary outcomes following invasion, such that drift could conceivably lead to the blurring of differences among treatments, as a result of increased within‐treatment variance relative to the among‐treatments variance. Alternatively, the evolutionary history of interacting species could instead dictate divergent trajectories of community development, leading to the formation of alternative transient or stable states differing in the abundance of species. In that case, we would expect communities that were composed of different combinations of evolved and naïve invaders and residents to remain different, or even become more divergent over time.",
"discussion": "DISCUSSION We evaluated the course of community development in two different species assemblages to understand whether populations differing in evolutionary history could produce transient alternative ecological community states. Our results demonstrated that the evolutionary history of interacting species can drive differences in community development following biological invasions. The differences in converging community structure observed in the two assemblages seemed to be related to whether the invader or resident species evolved. For Assemblage A, in which we previously demonstrated ecological effects related to evolution in the invader (Faillace & Morin, 2016 , 2020 ), we observed increasing similarity among communities from the coevolved and evolved invader treatments. In contrast, for Assemblage B, in which we had shown that evolution of resident species drove differences in the community (Faillace & Morin, 2016 , 2020 ), we found that communities with naïve invasions and evolved residents became more similar over time. Here, the evolved resident communities showed little directional change over time from the initial to final community state, whereas the naïve invasion communities appeared to ultimately become more similar to the communities with evolved residents. The increase in similarity between the naïve invasion and evolved residents communities in Assemblage B was consistent with our expectation that post‐invasion evolution occurs in a rapid and mostly repeatable fashion with the species present in each assemblage representing a common selective regime, such that naïve populations would evolve to effectively become less naïve over time. Additionally, across both assemblages, communities from naïve treatments showed some trends of convergence with coevolved treatments, however the convergence was not statistically significant. It seems likely that these effects might have become stronger and statistically significant had we continued our observations beyond the tens of generations that we observed. These results agree with a recent field experiment with native and invasive grass species occurring in sympatry and allopatry, in which evolution in populations of the native species appeared to produce repeatable responses to competition with the invasive species (Germain et al., 2020 ). Nonetheless, in our experiment, the lack of complete convergence to a single uniform community state within each assemblage suggested that the differences in populations that we attributed to evolutionary history could produce long‐lasting transient alternative community states. When considering the patterns of community similarity over time with the population‐level results, we observed that, for Assemblage A, signals of convergence among treatments appeared to be driven by increases in the invader, E. daidaleos , across all except the coevolved treatment, for which it declined, as well as declines in abundance in all treatments of E. patella and S. teres . Nonetheless, we also observed evidence of drift occurring among communities within treatments. Therefore, the observed pattern appears to be driven, at least in part, by increasing within‐treatment dispersion, implying an additional role for stochastic processes, specifically ecological drift, in community development for this assemblage. For Assemblage B, the increase in among‐treatment similarity between the evolved residents and naïve invasion treatments appeared to be driven by a community structure increasingly balanced among the invading species, P. bursaria , as well as residents Monostyla sp., E. daidaleos , and P. caudatum . Additionally, the abundance of S. coeruleus declined across all invaded treatments over time. Here, although we observed changes to the degree of drift over time, the evolved resident treatment exhibited consistently greater within‐treatment dispersion compared with the remaining treatments. This elevated dispersion suggests that the abundances of interacting species in the treatment with evolved residents remained more variable over time compared with those in the other treatments. One possible explanation for this result is that the effects in this treatment are potentially driven by evolution occurring in multiple interacting species, as opposed to evolution in a single species, as seen in Assemblage A, possibly increasing variability in those interactions. Historical contingency is known to influence the properties of both populations and communities (Fukami, 2015 ; Lenski, 2017 ). When considering interactions among species, historical contingency can determine the composition and functioning of communities through priority effects and eco‐evolutionary dynamics to influence community assembly and coexistence, as well as succession (Fukami, 2015 ; Grainger et al., 2019 ; Kremer & Klausmeier, 2017 ; Meiners et al., 2015 ). Transient alternative states can differ from alternative stable states in the conditions under which they occur and the patterns of diversity in which they result; because transient states may be common in many natural communities, they could be particularly important in maintaining biodiversity (Fukami & Nakajima, 2011 , 2013 ). For instance, different genotypes of foundation species can promote the development of measurably different communities of associated dependent species (e.g., Keith et al., 2017 ; Lamit et al., 2016 ). Because chance mutations can cause divergence in genotypes (and therefore phenotypes) under identical selective regimes (Lenski, 2017 ; Meyer et al., 2012 ), it follows that the dominant phenotypes present among allopatrically evolving populations or in metacommunities may differ in important ways (e.g., Urban, 2010 ), even when the populations evolve under similar conditions. In fact, Brockhurst et al. ( 2006 ) demonstrated that independently evolved populations of wrinkly spreader (WS) phenotypes of P. fluorescens can differ in aspects of the WS phenotype with important consequences for communities, despite evolving under identical conditions. Sympatrically coevolved pairs of strains exhibited greater character displacement, yielding both greater productivity and reduced invasibility compared with randomly assembled (allopatric) pairs. Pairs of sympatrically and allopatrically evolved strains of P. fluorescens also differ in the strength of the priority effects that governed competitive outcomes (Zee & Fukami, 2018 ). Taken together, even in cases in which eventual convergence of phenotypes might be expected, historical contingency resulting in phenotypic variability among populations could cause important long‐term differences in community structure or function. The importance of ecological and evolutionary processes interacting at contemporary timescales is increasingly recognized (Ellner, 2013 ; Koch et al., 2014 ; Schoener, 2011 ). Some studies of both laboratory and natural systems have clearly demonstrated ongoing contemporary evolution (e.g., Bassar et al., 2012 ; Farkas et al., 2016 ; Hiltunen & Becks, 2014 ), highlighting the need to consider ecological and evolutionary dynamics as simultaneous and interacting processes that drive community dynamics. Such eco‐evolutionary dynamics (Fussmann et al., 2007 ; Kinnison & Hairston, 2007 ) can have effects on the phenotypic traits of species (Grant & Grant, 2002 ; Stuart et al., 2014 ), population and community dynamics (Becks et al., 2010 , 2012 ; Faillace & Morin, 2016 , 2020 ; Yoshida et al., 2003 ), and even ecosystem functioning (Palkovacs et al., 2009 ). Eco‐evolutionary feedbacks (such as the ecology → evolution → ecology feedback observed here) may be of particular importance in governing the divergence of populations through their ability to amplify intraspecific phenotypic trait variation (Bailey et al., 2013 ), and may be crucial for our understanding of how species diversity arises (Post & Palkovacs, 2009 ) and is maintained. Despite their potential importance, few studies have attempted to identify the community‐level effects of eco‐evolutionary dynamics in complex communities. Nonetheless, we clearly need to improve our understanding of evolution in a community context (terHorst et al., 2018 ). We did not identify specific molecular genetic targets of evolution, but have previously argued that our design explicitly disentangled possible induced plastic phenotypic responses from heritable changes (Faillace & Morin, 2020 ). Not only did the identity of species in each community remain constant with all other conditions maintained under common garden conditions, but also an additional 3 weeks passed before we began our observation period, representing the passage of ~21–42 generations for the protists. This is well after the period of maximum induction of plastic changes for protists (i.e., plasticity in protists is reasonably well characterized in the literature with phenotypically plastic responses typically fully induced in a population within the first 72 h, corresponding to about two or three generations after exposure to novel conditions) (Duquette et al., 2005 ; Fyda & Wiackowski, 1998 ; Wiackowski & Staronska, 1999 ). Finally, tracking abundances over multiple generations and turnovers of the individuals in populations ensured that any observed differences represent the ecological manifestations of heritable differences among lines of evolved and naïve populations of our species, rather than transient dynamics due to plasticity. The observed convergence among some treatments in each assemblage is consistent with selection acting on standing genetic variation, rather than on the appearance of new mutations (i.e., as for the populations of yeast studied by Burke et al., 2014 ). Nonetheless, similar to both the scale and time dependency of convergence and divergence observed in Escherichia coli (Lenski, 2017 ), new mutations could greatly alter performance and community composition, especially over longer time scales. We argue that evolutionary history can contribute to the emergence of long‐lasting alternative community states in which the abundances of constituent species differ significantly, but it remains to be seen how important it is in natural communities. Common and rare species can differ dramatically in their respective roles in biological communities (Gaston & Fuller, 2008 ; Jain et al., 2014 ). This implies that when the relative contribution of rare and common species is shifted, these kinds of transient alternative states could be expected to have important functional consequences. In our experiment, demonstrably different alternative community states persisted for tens of generations in both assemblages, with incomplete convergence observed during the course of our observations, indicating that these transition states would be biologically relevant for community dynamics. The transient alternative states that we observed here, in which communities differed in the abundance but not in the identity of interacting species, are seemingly subtle, but could nonetheless have important functional consequences, suggesting that the role of evolution in driving these states deserves greater attention."
} | 5,453 |
32760534 | PMC7391324 | pmc | 4,962 | {
"abstract": "Abstract Species occurrence in a site can be limited by both the abiotic environment and biotic interactions. These two factors operate in concert, but their relative importance is often unclear. By experimentally introducing seeds or plants into competition‐free gaps or into the intact vegetation, we can disentangle the biotic and abiotic effects on plant establishment. We established a seed‐sowing/transplant experiment in three different meadows. Species were introduced, as seeds and pregrown transplants, into competition‐free gaps and the intact vegetation. They included 12 resident plants from the locality and 18 species typical for different habitats. Last two years, gaps were overgrown with vegetation from surrounding plants and we observed the competitive exclusion of our focal plants. We compared plant survival with the expected occurrence in target locality (Beals index). Many of the species with habitat preferences different from our localities were able to successfully establish from seeds and grow in the focal habitat if competition was removed. They included species typical for much drier conditions. These species were thus not limited by the abiotic conditions, but by competition. Pregrown transplants were less sensitive to competition, when compared to seedlings germinated from seeds. Beals index significantly predicted both species success in gaps and the ability to withstand competition. Survival in a community is dependent on the adaptation to both the abiotic environment and biotic interactions. Statistically significant correlation coefficients of the ratio of seedling survival in vegetation and gaps with Beals index suggest the importance of biotic interactions as a determinant of plant community composition. To disentangle the importance of abiotic and biotic effect on plant establishment, it is important to distinguish between species pool as a set of species typically found in given community type (determined by Beals index) and a set of species for which the abiotic conditions are suitable.",
"conclusion": "5 CONCLUSIONS Many nonresident species very improbable to occur in the target habitats (i.e., with low Beals index) were able to perform well in competition‐free gaps, but were unable to survive in intact vegetation. These species were thus not limited by the abiotic conditions, but by competition with neighboring plants. Although the appropriate abiotic conditions are important for seedling survival, our experiment suggests that biotic interactions are likely the most important determinants of plant species community composition and operate mainly through prevention of establishment of the “unsuitable” species. Although Beals index is a good predictor of species survival in plant communities, we should be careful to use it as species pool determinant, especially in disentangling the effect of abiotic and biotic filter on species community composition. If we define the community species pool as a set of species able to survive and reproduce in given abiotic environment (Butaye et al., 2001 ), the set of species will be much wider than predicted by Beals index (and generally any comparative method) because we extend the species pool about species otherwise excluded by biotic filter. Comparative methods generally exclude species which are not able to withstand the competition from species pool. If we compare the actual community composition with this species pool with the aim to disentangle the importance of biotic and abiotic factors, we would underestimate the effect of competition because species affected by competition are already excluded from this species pool.",
"introduction": "1 INTRODUCTION Each plant community is formed by a subset of the species pool, that is, a subset of all species available to colonize a given site (Cornell & Harrison, 2014 ). The basic question is then which mechanisms decide which species from the species pool will finally form the community. Dispersal limitation is an important factor for species occurring in the region. For example, the successful establishment of a single individual often requires the arrival of hundreds or thousands of seeds (Vítová & Lepš, 2011 ). Interestingly, low favorability of a particular habitat can be overcome by massive numbers of propagules (Fibich, Vítová, & Lepš, 2018 ). Nevertheless, the main processes limiting species occurrence in a local scale are abiotic environment and biotic interactions (HilleRisLambers, Adler, Harpole, Levine, & Mayfield, 2012 ). Abiotic environment is influenced by many factors such as temperature and precipitations, availability of nutrients and other resources which plants need for their survival. Biotic interactions include the relationships among living organisms in a community. Although other biotic interactions (e.g., mycorrhiza, facilitation, pollination, herbivory) play an important role in plant communities, competition is considered a significant factor that limits co‐occurrence among species (Grubb, 1977 ; Li, Poisot, Waller, & Baiser, 2018 ; Palmer, 1994 ; Wellstein et al., 2014 ). Furthermore, studies typically use competition as biotic filter in community assembly studies (HilleRisLambers et al., 2012 ). In local communities, abiotic environment and biotic interactions operate simultaneously, but their relative importance in structuring local community composition is often unknown and difficult to disentangle on the basis of observational data only (Araújo & Rozenfeld, 2014 ; Cadotte & Tucker, 2017 ; Kraft et al., 2015 ). Although many studies based on observational data use the concept of environmental filtering as the effect of abiotic environment only, they in fact reflect environmental filtering which includes not only the species ability to survive under specific environmental condition of the given site but also withstand under the competition of other species present in a given site (Cadotte & Tucker, 2017 ). By this approach, the effect of biotic interactions on local community structuring could be significantly underestimated. Very probably, only experimental approach manipulating biotic interactions in species communities can reliably distinguish the effect of abiotic environment and biotic interactions (Kraft et al., 2015 ). Nevertheless, some studies (e.g., D'Amen, Mod, Gotelli, & Guisan, 2018 ) claim that the analysis based on combination of observational data and null models is able to separate the effect of biotic filter from the environmental filtering. Sowing and transplant experiments are excellent approaches to disentangle the effects of various “filters” on community composition (Švamberková, Vítová, & Lepš, 2017 ; Turnbull, Crawley, & Rees, 2000 ; Zobel & Kalamees, 2005 ). Excluding dispersal limitation, failure to establish after sowing or transplanting can be attributed to habitat limitation. There are many examples of species that are able to grow in given abiotic conditions, but are excluded by the biotic filter. These species are present within a regional species pool, but are representative for very different habitats. In order to examine the ability of these species to withstand the abiotic conditions of a given habitat, seed/transplant introduction experiments, where biotic filters (especially competition) are experimentally removed, are required (Cornell & Harrison, 2014 ; Švamberková et al., 2017 ). Species that successfully establish in competition‐free experimental plots should be considered a part of the species pool defined as species able to pass only through abiotic filters (Butaye, Jacquemyn, Honnay, & Hermy, 2001 ) while they cannot be a part of usually used species pool defined as species able to pass through the both abiotic and biotic filters (Zobel, 1997 ). Comparing plant performance across artificial competition‐free gaps and intact vegetation (where the biotic and abiotic filters work in concert) can separate the importance of biotic and abiotic effects on plant establishment (HilleRisLambers et al., 2012 ; Kraft et al., 2015 ). Many species require some type of gap (i.e., plot with reduced competition) in natural settings (Puerta‐Piñero, Muller‐Landau, Calderón, & Wright, 2013 ). In nature, gaps are the result of various disturbances, which create competition‐free microhabitats and enable species to germinate and subsequently establish. When studying species establishment in seed/transplant introduction experiments, competition can be artificially excluded (or substantially reduced) using experimentally generated gaps (Kotorová & Lepš, 1999 ; Lemke, Janßen, & Porembski, 2015 ; Tofts & Silvertown, 2002 ; Vítová, Macek, & Lepš, 2017 ). In gaps, competition for light, nutrients, and water is reduced (Frei, Scheepens, & Stöcklin, 2012 ; Lemke et al., 2015 ). On the other hand, species present in gaps are more exposed to extreme environmental conditions, such as desiccation (Kotorová & Lepš, 1999 ; Vítová & Lepš, 2011 ). Seedlings growing in gaps are also more apparent to herbivores than seedlings occurring within intact vegetation (Gustafsson, Ehrlén, & Eriksson, 2002 ; Lemke et al., 2015 ). Both gap size and the time of their formation play a crucial role in the establishment of new seedling species, affecting which species is first to colonize this gap. Even so, the establishment of seedlings in a community is unlikely and seedling survival does not always assure the long‐term persistence of the species (Gustafsson et al., 2002 ; Vítová & Lepš, 2011 ; Zobel, 1997 ). Most species are filtered out of a community during the germination phase and subsequent establishment of individuals (Kotorová & Lepš, 1999 ). The importance of factors (both abiotic and biotic) affecting species survival in a community can differ in different life stages of plants because their regeneration and realized niches are often quite distinct (Grubb, 1977 ). One of the primary reasons for the absence of some species in a community is their inability to establish in the presence of competition from other species. Although biotic interactions affect plants in later stages of their life span, the effect is not as strong as in their early phases of seedling development because older individuals are more biotic resistant than small seedlings (Bennett et al., 2016 ; Tofts & Silvertown, 2002 ). It suggests that competitive exclusion of well‐established individuals in a community may be rather slow (Adler, Ellner, & Levine, 2010 ). The studying of different life stages is thus necessary to get a complete insight into local processes influencing a whole life cycle of species. When comparing the effect of abiotic and biotic filter on species composition of a local community, we need to define a local species pool, ideally as the ability of a given species to establish based on the abiotic environment alone without the effect of competition filter (Butaye et al., 2001 ; Švamberková et al., 2017 ). There are various methods to help determine the species pool: Ellenberg indicator values (Pärtel, Zobel, Zobel, van der Maarel, & Partel, 1996 ; Zobel, 1997 ; Zobel, van der Maarel, & Dupré, 1998 ), functional traits (de Bello et al., 2012 ; Moor, Hylander, & Norberg, 2015 ; Sonnier, Shipley, & Navas, 2010 ), phytosociological knowledge from local experts (Sádlo, Chytrý, & Pyšek, 2007 ), Beals index (Botta‐Dukát, 2012 ; Ewald, 2002 ; Münzbergová & Herben, 2004 ), or ordination methods (Brown et al., 2019 ). Nevertheless, with exception of experimental approach, all other methods of species pool determination reflect the influence of the both biotic and abiotic filters. Nevertheless, because experimental approach is very time consuming, Beals index can be quite invaluable approach to species pool assessment. While most of the above‐mentioned approaches for determination of species pool size depend on either expert's phytosociological experience or models corresponding with environmental gradients, methods related to Beals index employ information based on multivariate structure of real data. It compares species co‐occurrence of examined species with other species of the appropriate habitat from a database of many phytosociological relevés (Chytrý & Rafajová, 2003 ), reflecting thus concerted effect of biotic and abiotic filters. Although Beals index is, in fact, also one of the phytosociological methods, neither any classification nor any environmental gradients determined in advance are employed. It transforms a species pool definition from a strictly determined set of species into species occurrence probability (Botta‐Dukát, 2012 ). We conducted a seed/transplant introduction experiment across three different meadow habitats (Appendix S1 ). Species, both resident in the locality and typical for different habitats (not expected to be part of the species pool), were introduced as either seeds or pregrown transplants into either competition‐free gaps or the intact vegetation. Subsequently, we computed the expected occurrence of species from our experiment on target habitats using Beals index derived from the species co‐occurrence pattern in the National Phytosociological Database (Chytrý & Rafajová, 2003 ) and compared these results with the real plant survival from our experiment. During the last two years of the experiment, surrounding vegetation was left to overgrow into gaps and we observed the competitive exclusion of our focal plants. Our study aimed to (a) compare the species pool determined by seed/transplant introduction experiment with the species pool delimited using Beals index; (b) disentangle the importance of the biotic and abiotic effects on plant establishment via the removal of competition; and (c) compare the survival of target species in different life stages (i.e., sown as seeds and planted as pregrown transplants) and their competitive exclusion. We expect that (a) some species determined by Beals index as improbable to occur in target habitats will be able to establish experimentally in competition‐free gaps. (b) Both abiotic and biotic effects will influence the species establishment, but competition will be the most important determinant. We suggest that if survival is affected by both intrinsic characteristics of individual species and their interaction with the environment, the more an environment discriminates among species, correlations of species successes across ecologically different habitats should be weaker. In this way, we can identify, whether the discrimination among species is more pronounced in gaps (suggesting mainly effect of abiotic environment), or in controls (discrimination by the whole habitat including competition by extant vegetation). (c) Competitive exclusion will be more important for seedlings growing from seeds in the field than for pregrown transplants.\n\n2.2 Seed introduction experiment To assess species establishment and survival in the presence and absence of competition, we introduced seeds and pregrown young individuals (transplants) of both resident and nonresident plant species to our three habitats (Appendix S1 ). We selected species with good germination rate (knowledge from previous studies, e.g., Švamberková et al., 2017 ) from species typical for the region of our target locality. A species residence was determined for individual habitats based on whether a species was present in at least one of the five phytosociological relevés (5 × 5 m) of given habitat type recorded in June 2014 (i.e., “habitat residency,” Table S5 ). We also used an additional classification, where any species present in at least one habitat type (according to phytosociological relevés from June 2014) or found within the study site during the nature conservation‐screening inventory by Jan Horník et al. (unpublished data) were considered residents for the entire locality (i.e., “whole locality residence,” Table S5 ). Nonresident species include species typical for both drier and wetter conditions than target locality. Nevertheless, all the nonresident species can be considered part of the regional species pool, because they are found in close surrounding (see maps of species distribution at www.pladias.cz/en/ , accessed on May 8, 2019) and their propagules are thus able to reach the target locality. Seeds and transplants were placed into either control plots, with the intact vegetation, or artificially created gaps. We created 30 artificial gaps (40 × 40 cm) in two replications in each habitat type, each by digging a hole 20 cm deep, and refilling with soil from the target habitat. To prevent competition from surrounding vegetation, gaps were weeded regularly two times a year (in spring and autumn) until 2016 when gaps were weeded once during spring for the last time. In 2017 and 2018, we observed the potential competitive exclusion of established individuals in gaps from the neighboring vegetation. Control plots of the same size were established without any manipulation of extant vegetation. Seeds from 30 species, 12 residents and 18 nonresidents (Table S5 ), were sowed to the center of 20 × 20 cm plots within gap and control treatments in spring 2013. We used seeds from a commercial supplier (Planta Naturalis, Markvartice, Czech Republic). Each species was sowed separately in its own plot. Within a plot, 200 seeds of species, which had a seed weight of one seed 1 mg or more, were sown for each plant species. We sowed more than 200 seeds for plant species with seeds lighter than 1 mg because small seeds are expected to have reduced probability of establishment (Cornelissen et al., 2003 ). We used an ad hoc formula to increase the amount of seeds lighter than 1 mg: x = 200 (1 − log m ), where x was a weight of seeds required for sowing and m a weight of one seed in mg. This process helped provide enough individuals for the assessment of mortality. The success of seedling establishment and survival was subsequently expressed as the number of survivors out of the number of the sown seeds. The proportion of seedling recruitment and survival was monitored from 2013 to 2018 several times per year.",
"discussion": "4 DISCUSSION 4.1 Seed germination and survival of seedlings in contrast to transplants Across all habitat types, sown species, both resident and nonresident, germinated and subsequently survived better in gaps than in intact vegetation. This result corresponds to many other studies where most species persisted significantly better in plots without competition (Kotorová & Lepš, 1999 ; Švamberková et al., 2017 ; Tofts & Silvertown, 2002 ). Zobel et al. ( 1998 ) suggested that one of the most important factors affecting species survival is the surrounding vegetation. Frei et al. ( 2012 ) highlighted the positive effect that disturbances have on the establishment of Campanula thyrsoides seedlings, which responded positively to cutting the surrounding vegetation and disturbing the turf. Also, in our experiment, many nonresident species with habitat preferences different from our habitats were able to establish from seeds and grow when competition was removed (similarly as in Tofts & Silvertown, 2002 ), but not in the intact community. Also, transplants survived better in gaps than in intact vegetation. However, the difference between transplant survival in gaps and intact vegetation was smaller than when seeds were introduced. In the Sesleria uliginosa‐Briza media habitat, there were no differences between gaps and intact vegetation in the case of transplants in contrast to sown species. Aboveground biomass was there the lowest of the three habitats (Table S2 ), and thus, we can expect least amount of competition for light. While also this small competition was crucial for seedlings growing from seeds in the field, it was not so important problem for transplants, which are generally more resistant than seedlings (Bennett et al., 2016 ). There were many species that were unable to establish from seeds in intact vegetation, but survived as transplants. The biotic filter had thus a more pronounced effect on establishment from seeds, than on transplant establishment (even though they were still young individuals). In concordance with Kotorová and Lepš ( 1999 ), it seems that very early phases of seedling establishment are the most sensitive stages of many plant species and their suppression is an important filtering mechanism in the community. Species survival was dependent on the regular weeding within gaps because both artificially created gaps and other types of naturally disturbed plots tend to become overgrown with surrounding vegetation (Puerta‐Piñero et al., 2013 ). Accordingly, during the last two years of our experiment (i.e., 2017 and 2018) when weeding ceased, the differences between gaps and vegetation started to decrease. Nevertheless, many nonresident species with habitat preferences different from our habitats (i.e., also species with very low Beals index and thus species very improbable to occur in target habitats) successfully established in gaps and survived also after weeding ceased and even reached their reproductive stage; confirming that competitive exclusion can be a slow process (Adler, Fajardo, Kleinhesselink, & Kraft, 2013 ). However, once weeding was stopped, plant mortality increased considerably, especially for seedlings. This supports the results in Gustafsson et al. ( 2002 ), which suggest that initial seedling establishment does not guarantee long‐term species survival and it is important to monitor the complete vegetation cycle of target species because sudden changes can occur in late stages of seedling establishment (Münzbergová & Herben, 2004 ). Also, other studies (Ehrlén, Münzbergová, Diekmann, & Eriksson, 2006 ; Frei et al., 2012 ; Houseman & Gross, 2006 ; Pärtel, Szava‐Kovats, & Zobel, 2013 ) highlight the importance of long‐term monitoring in seed addition experiments because it is possible that seeds of many species germinate and survive as seedlings for several years, but never establish a viable population (Vítová & Lepš, 2011 ). 4.2 Seedling/transplant survival compared with their respective Beals index values and among different habitat types While the effect of species residence is a rather crude binary variable (resident/nonresident), the Beals index is based on individual species performance within an extensive set of phytosociological records from the whole region of the Czech Republic. This metric is able to distinguish between resident species regularly found within a given vegetation type and nonresident species found in similar and dissimilar habitats. In all habitat types and during the entirety of the experiment, seedling survival was positively correlated with Beals index in gaps and intact vegetation. This suggests that species are adapted to both the abiotic (correlation of survival in gaps with Beals index) and biotic conditions (correlation of survival in intact vegetation with Beals index) of particular habitats (HilleRisLambers et al., 2012 ). Positive correlations of species survival with Beals index was also reported by Mudrák et al. ( 2014 ), which sowed Rhinanthus species into a wide range of habitat types and by Milden, Münzbergová, Herben, and Ehrlén ( 2006 ) for Succisa pratensis . On the other hand, Münzbergová and Plačková ( 2010 ) and Frei et al. ( 2012 ) did not observe a positive relationship between Beals index and seedling survival of sown species. For transplants, the correlation of survival with Beals index was weaker than for seedlings. This again confirms that transplants are less sensitive to competition than seedlings. This supports previous observation that the primary reason for the absence of some species in a community is their inability to establish as seedlings from seeds (Tofts & Silvertown, 2002 ; Vítová & Lepš, 2011 ). Higher correlation coefficients between Beals index and survival in intact vegetation compared to gaps and the positive correlations between the ratio of seedling survival in intact vegetation and gaps suggest that competition was the most important determinant of species community composition. These dependences were generally similar also for transplants although they were rather weak. Higher correlation coefficients of survival across habitats in gaps compared to control plots (especially in case of pair Carex acuta‐Carex panicea and Deschampsia caespitosa‐Carex tomentosa habitats) also revealed that differences in species survival within these two habitats are caused more by biotic interactions than by environmental conditions (i.e., the competition is more discriminating among species than the effect of the abiotic environment). Bar‐Massada ( 2015 ) suggested that biotic interactions are the most important drivers of species co‐occurrence, although their effect could be influenced by environmental heterogeneity. Many other studies highlight the importance of biotic interactions in determining species community composition and the necessity to incorporate them into models (Boulangeat, Gravel, & Thuiller, 2012 ; Morales‐Castilla, Matias, Gravel, & Araújo, 2015 ; Myers & Harms, 2011 ; Pollock et al., 2014 ; Wisz et al., 2013 ). Conversely, D’Amen et al. ( 2018 ) suggested that environmental filtering and dispersal limitation are more important drivers of species co‐occurrence than biotic interactions, but this conclusion was based on the analyses of observational data and the use of null models. In our view, without direct experimental manipulation of biotic interactions, it is difficult to distinguish the direct effect of environment from environmentally modified biotic interactions (Cadotte & Tucker, 2017 )."
} | 6,431 |
23024637 | PMC3441193 | pmc | 4,963 | {
"abstract": "Prochlorococcus and Synechococcus , which numerically dominate vast oceanic areas, are the two most abundant oxygenic phototrophs on Earth. Although they require solar energy for photosynthesis, excess light and associated high UV radiations can induce high levels of oxidative stress that may have deleterious effects on their growth and productivity. Here, we compared the photophysiologies of the model strains Prochlorococcus marinus PCC 9511 and Synechococcus sp. WH7803 grown under a bell-shaped light/dark cycle of high visible light supplemented or not with UV. Prochlorococcus exhibited a higher sensitivity to photoinactivation than Synechococcus under both conditions, as shown by a larger drop of photosystem II (PSII) quantum yield at noon and different diel patterns of the D1 protein pool. In the presence of UV, the PSII repair rate was significantly depressed at noon in Prochlorococcus compared to Synechococcus . Additionally, Prochlorococcus was more sensitive than Synechococcus to oxidative stress, as shown by the different degrees of PSII photoinactivation after addition of hydrogen peroxide. A transcriptional analysis also revealed dramatic discrepancies between the two organisms in the diel expression patterns of several genes involved notably in the biosynthesis and/or repair of photosystems, light-harvesting complexes, CO 2 fixation as well as protection mechanisms against light, UV, and oxidative stress, which likely translate profound differences in their light-controlled regulation. Altogether our results suggest that while Synechococcus has developed efficient ways to cope with light and UV stress, Prochlorococcus cells seemingly survive stressful hours of the day by launching a minimal set of protection mechanisms and by temporarily bringing down several key metabolic processes. This study provides unprecedented insights into understanding the distinct depth distributions and dynamics of these two picocyanobacteria in the field.",
"conclusion": "Conclusion The comparison of Synechococcus and Prochlorococcus cultures acclimated to VL supplemented or not with UVR revealed relatively few physiological responses specific to UV. This notably includes a shift of the DNA synthesis phase (Figure 1 ; see also Kolowrat et al., 2010 ), an increase of the Zea:Chl a ratio (Figure 2 ) and an enhanced PSII repair rate (Figure 4 ). Accordingly, UVR seemingly also had limited effects at the transcriptomic level, as shown by the globally similar diel expression patterns between VL and VL + UV in both strains (Figures 7 and 8 ). It is worth noting however that the relative expression of a few genes was either enhanced (e.g., psbD in both strains, D1:2 encoding genes and crtR in Synechococcus only) or reduced (e.g., Synechococcus D1:1 encoding gene or Prochlorococcus pcbA ) in UV-acclimated cultures. A handful of genes, including rpoD4 /8 in Prochlorococcus and kaiABC in Synechococcus , also exhibited a delayed expression peak by about 3 h. However, this surprisingly did not translate into any conspicuous changes in the diel patterns of most of the other genes examined here, as could have been expected from the known regulatory role of the circadian clock and sigma factors on gene transcription (see e.g., Summerfield and Sherman, 2007 ; Ito et al., 2009 ). The most striking result of the present study is likely the markedly distinct response to diurnal light variations between Prochlorococcus and Synechococcus , despite their close phylogenetic relatedness (Scanlan et al., 2009 ). These two picocyanobacteria indeed exhibited very different degrees of PSII photoinactivation at noon that can be partly explained by their distinct (i) PSII repair capacity (Figure 4 ; see also Six et al., 2007a ), (ii) ability to modulate photoprotective pigments (Zea and Car, Figure 2 ), and (iii) resistance capacity against oxidative stress (Figure 6 ). Comparative transcriptomic analyses also revealed that, in Synechococcus , genes coding for a number of protective systems were maximally expressed during hours of highest irradiance, i.e., when these mechanisms are most critical to cope with transitory stressful conditions. This includes several genes involved in ROS detoxification enzymes, DNA repair genes as well as genes involved in photoprotection and/or dissipation of excess energy, such as crtR and ocp (Figure 7 ). In contrast, in Prochlorococcus , very few genes of these pathways, including ptox and psbA , were maximally expressed around midday, while others, such as phrA and crtR , had already reached their maximal (saturating) expression at mid-morning. Furthermore, many photosynthetic genes that were upregulated during the day in Synechococcus were in contrast downregulated in Prochlorococcus , including the PSI core gene psaB as well as genes involved in light-harvesting, ATP synthase and CO 2 fixation (Figure 8 ). Similarly, glgA (encoding glycogen synthase) mRNA abundance was recently shown to exhibit a maximum diel expression at midday in Synechococcus sp. WH8103 grown under a 16/8-h L/D cycle, while it peaked in concert with rbcLS in Prochlorococcus \n marinus MED4 during the LDT (Wyman and Thom, 2012 ). Thus altogether, it seems that while Synechococcus , as many other cyanobacteria (Kucho et al., 2004 , 2005 ; Stockel et al., 2008 ; Toepel et al., 2008 , 2009 ; Shi et al., 2010 ), efficiently copes with the diurnal changes in photon fluxes, Prochlorococcus rather displays a stress-like response at midday. The latter response is most likely related to the high irradiance level used in the present study (reaching 870 μmol photons m −2 s −1 at noon), which are typically found in the upper mixed layer of tropical oligotrophic oceans (Holtzendorff et al., 2001 ). Indeed, several genes that were found here to be downregulated during the light period (e.g., pcb , rbcL , chlG ) were in contrast upregulated in Prochlorococcus cultures grown at lower irradiances, provided as continuous (Berg et al., 2011 ) or cyclic light (Zinser et al., 2009 ). Despite the very atypical transcriptomic response observed here, Prochlorococcus was able to recover high PSII quantum yield at night and to maintain an optimal growth rate under these conditions. Thus, we hypothesize that, in contrast to Synechococcus sp. WH7803, P. marinus PCC 9511 cells manage to cope with harmful light conditions by bringing down temporarily some of the main metabolic processes and by launching a minimal set of protection mechanisms during stressful hours. Whether the absence of a true circadian clock in Prochlorococcus (Holtzendorff et al., 2008 ; Axmann et al., 2009 ) is involved in this differential management of excess light compared to Synechococcus still remains to be investigated. Altogether, our study reinforces previous studies depicting Prochlorococcus as a very specialized organism restricted to a narrow environmental niche, while Synechococcus has adopted a generalist strategy enabling it to cope with more variable environmental conditions, a difference consistent with the distinct habitats in which these two organisms predominate (Scanlan, 2003 ; Kettler et al., 2007 ; Dufresne et al., 2008 ; Scanlan et al., 2009 ).",
"introduction": "Introduction Phytoplanktonic cells, and in particular cyanobacteria, experience dramatic daily fluctuations of solar radiations, which can become suboptimal for photosynthetic processes around midday. Photosystem II (PSII) is particularly sensitive to these changes in photon fluxes and under unfavorable or stressful conditions its activity can decline more rapidly than most other physiological processes (Berry and Björkman, 1980 ; Demmig-Adams and Adams, 1992 ; Aro et al., 1993 ; Andersson and Aro, 2001 ). Photodamages to PSII are thought to start by the inactivation of the oxygen-evolving complex, which is caused by the dissociation of the Mn 4 Ca 2+ cluster. This process leads to the production of long-lived P 680 + , the oxidized form of the reaction center chlorophyll (Chl) pair, a particularly strong oxidant which in turn provokes the destruction of the PSII core protein D1 (Hakala et al., 2006 ; Nishiyama, 2006 ). At low irradiances, the rate of photosynthetic electron transport is proportional to the photon flux density and damaged D1 polypeptides can be removed from the PSII reaction center and rapidly replaced by newly synthesized D1 proteins (Park et al., 1995 ; Tyystjarvi and Aro, 1996 ; Nixon et al., 2005 ; Ohnishi et al., 2005 ). However, at higher irradiances, the rate at which the PSII reaction center is damaged can exceed its repair rate, which results in an increase of inactivated PSII centers and a subsequent decline of the quantum yield of photosynthesis, resulting from photoinhibitory fluorescence quenching (Powles, 1984 ; Prásil et al., 1992 ; Aro et al., 1993 ; Andersson and Aro, 2001 ). Although the visible part of the solar spectrum (400–700 nm), also called photosynthetically active radiations (PAR), is responsible for most photoinhibitory effects, the contribution of UV-B (280–315 nm) and, to a least extent, UV-A (315–400 nm) is also notable in the uppermost layer of the ocean (Dring et al., 2001 ; van de Poll et al., 2001 ; He and Häder, 2002b ). UV-B can indeed damage the photosynthetic apparatus about 100-fold more efficiently than visible light and these radiations might directly affect PSII proteins and the Mn 4 Ca 2+ cluster (Sarvikas et al., 2006 ; Caldwell et al., 2007 ). UV and high visible radiations can also cause indirect photoinhibitory effects via the production of reactive oxygen species (ROS; He and Häder, 2002a , b ; Rastogi et al., 2010 ), mainly formed within reaction centers (Asada, 1999 ) and light-harvesting complexes (Knox and Dodge, 1985 ; Zolla and Rinalducci, 2002 ). ROS are powerful oxidizing agents which can react with DNA, lipids, and proteins. Although these compounds are inevitably produced by cell metabolism, even under optimal growth conditions, their production is drastically enhanced when cells are exposed to a variety of stresses, including excess visible light and UV radiations (UVR; Latifi et al., 2005 ; Ross et al., 2006 ; Houot et al., 2007 ; Allakhverdiev and Murata, 2008 ). The effect of ROS on PSII photoinhibition is thought to act primarily by inhibiting the de novo synthesis of proteins, including those required for the repair of PSII (Nishiyama et al., 2004 ; Nishiyama, 2006 ; Takahashi and Murata, 2008 ). A direct effect of ROS on the inactivation of PSII reaction center has also been suggested through triggering D1 degradation (Vass et al., 1992 ; Aro et al., 1993 ; Miyao et al., 1995 ; Keren et al., 1997 ; Lupinkova and Komenda, 2004 ). In any case, ROS clearly have a major role in light-mediated photoinhibition as well as in other environmental stresses (Nishiyama, 2006 ; Allakhverdiev and Murata, 2008 ; Latifi et al., 2009 ). Thus, survival of phototrophic organisms depends upon the amount of ROS produced and their efficiency in scavenging these oxygen species. In this context, marine picocyanobacteria belonging to the genera Synechococcus and Prochlorococcus constitute two relevant and complementary models to study acclimation processes to high light and UVR and their interrelationships with oxidative stress. In oceanic ecosystems, these two organisms numerically dominate the phytoplanktonic community (Partensky et al., 1999a ; Scanlan, 2003 ) and are considered to be the two most abundant photosynthetic organisms on Earth, with a substantial contribution to Chl biomass and primary production (Liu et al., 1997 ; Partensky et al., 1999a ; Agawin et al., 2000 ; Garcia-Pichel et al., 2003 ). Members of the marine Synechococcus genus are ubiquitously distributed and are most abundant in coastal regions and mesotrophic open ocean surface waters (Partensky et al., 1999a ; Zwirglmaier et al., 2008 ), whereas Prochlorococcus preferentially thrives in warm, stratified, oligotrophic tropical, and subtropical marine areas (Partensky et al., 1999b ; Zubkov et al., 2000 ; Johnson et al., 2006 ). In the field, these organisms experience large variations in irradiance, linked to the combination of the light/dark (L/D) cycle, water mixing, and a variable cloudiness (MacIntyre et al., 2000 ). Moreover, their tiny size (0.5–0.8 and 0.8–1.2 μm diameter for Prochlorococcus and Synechococcus , respectively) confers them a high surface to volume ratio, optimizing their photon capture, and making them particularly sensitive to UVR (Llabres and Agusti, 2006 , 2010 ). Like other photosynthetic organisms, marine cyanobacteria have evolved a variety of protection mechanisms to ensure their growth and survival in highly illuminated habitats. These mechanisms include thermal dissipation of excess light excitation, structural changes of the photosynthetic machinery as well as enzymatic and non-enzymatic scavenging systems to eliminate ROS, in particular those produced in photosynthetic membranes (for reviews, see Bailey and Grossman, 2008 ; Latifi et al., 2009 ). However, several pieces of evidence suggest that Prochlorococcus and Synechococcus lineages could deal differently with light stress. Indeed, two P. marinus strains (PCC 9511 and SS120, a high light- and a low-light-adapted ecotype, respectively) were found to be more sensitive to a transient exposure to high irradiances than three Synechococcus spp. strains representative of various trophic environments and exhibiting different pigmentation (RS9917, RCC307, and WH8102; Six et al., 2007b ). Similarly, measurements of cell abundances and/or mortality rates of field populations of picocyanobacteria exposed to different levels of natural solar radiations showed that Prochlorococcus exhibited a lower resistance to UVR than Synechococcus in surface waters of the central Atlantic Ocean (Llabres and Agusti, 2006 ; Agusti and Llabres, 2007 ) and the Mediterranean Sea (Sommaruga et al., 2005 ; Llabres and Agusti, 2010 ). In order to reveal potential differences in circadian metabolic rhythms between these two genera, the photophysiology of the model strains P. marinus PCC 9511 and Synechococcus sp. WH7803 was examined at different times of a modulated L/D cycle of visible light (hereafter VL) with or without UV. Additionally, the diel variability of the sensitivity of Prochlorococcus and Synechococcus to oxidative stress, as triggered by different H 2 O 2 concentrations was investigated. Expression of key genes involved in photosynthesis, light, and oxidative stress response and a number of other processes were also monitored in order to get insights about the molecular bases of the observed physiological differences.",
"discussion": "Discussion Differences in photosystem activity and regulation between Prochlorococcus and Synechococcus The alternation of light and darkness is one of the most predictive events that cyanobacteria have to deal with in the field. Strong variations of PAR occur over a daily timescale and are associated, in near surface waters, with concomitant changes of UVR fluxes. Here, we compare PCC9511, a strain representative of the Prochlorococcus HLI clade found in near surface oligotrophic waters, to WH7803, a Synechococcus strain characteristic of mesotrophic areas. Although these model strains do not represent the whole physiological diversity existing within these two genera, our data clearly show that both Prochlorococcus and Synechococcus cells are able to tune their photosynthetic apparatus to diurnal irradiance fluctuations. However, Prochlorococcus proved to be more sensitive than Synechococcus to photoinhibition by high photon fluxes, as suggested by a marked drop of the cellular pool of PSII core proteins (Figure 5 ) and a larger decrease of the PSII quantum yield (Figure 3 ) around noontime. It is noteworthy however that the latter phenomenon might also partly be due to NPQ of PSII fluorescence associated with photoprotective dissipation of light energy as heat (Bailey et al., 2005 ; Boulay et al., 2008 ). The diel changes of the photosynthetic activity observed here for P. marinus PCC 9511 (Figure 3 ) are quite comparable with those previously described for this strain grown in similar light conditions (Bruyant et al., 2005 ), except that our cultures exhibited a higher F V / F M during the night (∼0.7 vs. ∼0.55), likely translating a slightly better physiological status. However, they contrast with those obtained on the closely related MED4 strain by Zinser et al. ( 2009 ), who did not observed any significant diel variation of the F V / F M , likely because it was grown at lower irradiance (∼230 vs. 870 μmol photon m −2 s −1 here) and over a different L/D cycle (14/10 vs. 12/12 h). Another noticeable observation from our study was that UVR did not cause any appreciably stronger photoinhibitory effect than VL in both genera, likely due to an increased repair rate under VL + UV compared to VL. However, this UV-induced repair was much more important for Synechococcus than Prochlorococcus (∼twofold vs. ∼fivefold at noon, respectively). Accordingly, in the former organism, the relative D1 content was enhanced in the presence of UVR, while in Prochlorococcus the noontime drop of D1 was somewhat extended for UV-acclimated cells (see also transcriptomic analyses of the psbA genes below). These results are consistent with those obtained by Six et al. ( 2007a ), who observed that in response to a transient high light exposure, P. marinus PCC 9511 exhibited a lower PSII repair rate (0.9 PSII gained per second) than a range of marine Synechococcus strains (1.1–1.6 PSII s −1 ). In both strains in VL, this rate was more or less proportional to the instantaneous irradiance. However for cells grown under VL + UV, while the repair rate was comparable between the two strains at 9:00 and 15:00, it was significantly depressed at noon in Prochlorococcus compared to Synechococcus . This suggests that the PSII repair capacity of Prochlorococcus was already maximum around 400 μmol photons m −2 s −1 (Figure 4 B). The occurrence of different D1 encoding gene copies in Prochlorococcus and Synechococcus could, at least partially, explain such a distinct behavior. Indeed, while Prochlorococcus strains have one to three identical psbA gene copies (one in PCC 9511), coding for a single D1:1-like isoform (Hess et al., 1995 ; Partensky and Garczarek, 2003 ), Synechococcus strains possess three to six psbA genes, with only one copy coding for a D1:1 isoform and two to five copies (three in WH7803), coding for D1:2 isoforms (Garczarek et al., 2008 ). The respective role of these isoforms has been widely studied in the literature, both in freshwater cyanobacteria (Bustos et al., 1990 ; Clarke et al., 1993 ; Campbell et al., 1995 , 1998a ; Sass et al., 1997 ; Kos et al., 2008 ) and in marine picocyanobacteria (Garcia-Fernandez et al., 1998 ; Garczarek et al., 2008 ). Although some variations among cyanobacteria have been observed, it is generally accepted that the D1:1 isoform would confer a higher PSII activity (Campbell et al., 1996 ), while D1:2 would provide a lower quantum yield but a higher PSII resistance to photoinhibition (Krupa et al., 1991 ;Campbell et al., 1995 , 1998a ; Tichy et al., 2003 ). A variety of environmental cues, including UV exposure, can induce the exchange of these isoforms (Sicora et al., 2006 , 2008 ; Garczarek et al., 2008 ; for a review, see Bouchard et al., 2006 ). Here, we indeed noticed in Synechococcus sp. WH7803 an opposite expression pattern of D1:1 ( SynWH7803_ 0784) and D1:2 isoforms encoding genes ( SynWH7803_0790 , 0366 , and 2084 ) during daytime (Figure 7 ). It is worth noting that while there were only slight discrepancies in the expression levels of the different D1:2 encoding genes between VL and VL + UV, the D1:1 encoding gene was about fivefold more repressed at noon under the latter condition, suggesting a complete replacement of the D1:1 isoform by D1:2 isoform(s) that may contribute to the slight midday decrease in PSII quantum yield (Figure 3 ). Interestingly, although the single isoform in Prochlorococcus is phylogenetically a D1:1 isoform, as confirmed by the occurrence of a Gln residue at position 130 of the amino acid sequence (instead of Glu in D1:2; Clarke et al., 1993 ; Giorgi et al., 1996 ), its transcriptomic pattern was clearly closer from that of D1:2 isoforms, at least in our culture conditions. It is likely however that it presents a lower resistance to PSII photoinactivation compared to a true D1:2, as suggested by the larger drop of F V / F M at noon (Figure 3 ) and the concomitant lower PSII repair rate (Figure 4 ), compared to Synechococcus . Interestingly, the D2 subunit of the PSII core is also encoded by a single psbD gene in all Prochlorococcus genomes, whereas all Synechococcus sequenced so far possess two nearly identical psbD genes, as in most other cyanobacteria (such as e.g., Synechococcus sp. PCC 7942; Golden et al., 1989 ), with one co-transcribed with psbC that encodes the internal PSII antenna protein CP43 (Garczarek et al., 2001 ) and the other isolated in the genome. It is likely that as for psbA , the two copies are differently regulated in response to light and/or UV stress as previously reported in freshwater model cyanobacteria (Bustos and Golden, 1992 ; Kos et al., 2008 ), although this was not checked in the present study. Alternatively, this may simply contribute to a higher expression level of this key photosynthetic gene, possibly enabling a higher turnover of the corresponding protein. Another strategy used by cyanobacteria to cope with excess light energy is to decrease the relative amount of PSI reaction center complexes, an adjustment that was shown to decelerate the rate of photosynthetic electron transport (Murakami and Fujita, 1991 ; Hihara et al., 1998 ; Muramatsu and Hihara, 2003 ). Indeed, experiments on Synechocystis sp. PCC 6803 mutants impaired in their ability to modulate photosystem stoichiometry showed that this capacity is indispensable for growth under continuous high irradiance (Hihara et al., 1998 ; Fujimori et al., 2005 ). Accordingly, in Synechococcus sp. WH7803 the relative psaB levels were low before dawn (Figure 7 ), likely reflecting a lower PSI cell content at this time of the day, as previously observed in Crocosphaera \n watsonii (Saito et al., 2011 ). In contrast, this was not the case in Prochlorococcus , in which PSI core transcripts were maximal over most of the dark period. Differential regulation of light-harvesting systems in response to high light and UV radiations The striking structural differences between the major PSII antenna complexes of Prochlorococcus and Synechococcus may also be partially responsible for the different sensitivity of these picocyanobacteria to UV stress. Indeed, whilst Synechococcus , as most cyanobacteria, possess a large membrane-extrinsic antenna, the PBS (Sidler, 1994 ; Six et al., 2007c ), Prochlorococcus , like the other two green oxyphotobacteria, Prochloron and Prochlorothrix , use a transmembrane Chl a/b -binding Pcb antenna (LaRoche et al., 1996 ; Garczarek et al., 2003 ; Partensky and Garczarek, 2003 ). These dissimilar antenna structures may have important consequences on the way excitation energy is funneled downhill toward the reaction centers, on the regulation of this process as well as on the amount and nature of damages caused by UVR on Prochlorococcus and Synechococcus cells, since both organisms show some absorption capacities in the 300- to 400-nm band, which are likely related to their antenna (see e.g., Ong and Glazer, 1991 ; Claustre et al., 2002 ). Comparing and monitoring UV cross-sections for both picocyanobacteria would help answering this question. Here, we indeed observed an opposite expression pattern between the genes encoding antenna systems from these two organisms, with a strong downregulation of the pcb gene in Prochlorococcus during day time under VL (see also Garczarek et al., 2001 ) that was even more dramatic under VL + UV, while all examined PBS genes were upregulated in the afternoon in both light conditions (Figure 8 ). A similar result was previously obtained in L/D-entrained Cyanothece sp. ATCC 51142 cultures, though in the latter case maximum expression was centered at midday (Toepel et al., 2008 ). These observations suggest that the biosynthesis of antenna complexes occurs at different times of the day in the two picocyanobacteria and is under different light regulation controls. The fact that PBS genes remained upregulated under VL + UV suggests that the buildup of PBS complexes was only moderately affected by these radiations. This contrasts with a previous study where cultures of Synechococcus sp. WH8102 grown in continuous low-light were subjected to a sudden shift to UV (Six et al., 2007b ). This stress provoked a strong decrease in the relative expression of all PBS genes, associated with a disconnection of the PBS complexes from the thylakoid membrane as well as a dissociation of the distal PEII disks of PBS rods. In the present study, where cells were acclimated to either VL or VL + UV for several weeks, the PE to PC fluorescence emission ratio did not exhibit a peak at noon in either light condition, as would be expected if terminal PE subunits (i.e., PEII) were disrupted (Figure A1 in Appendix). The higher values of this ratio in UV-acclimated Synechococcus cells are likely not related to a higher PE content of the PBS, but rather to permanently decoupled PEII subunits, which would dissipate incident light as fluorescence. Consistently, UV also induced a slight drop of the expression of most PEII genes at noon and a 3-h delay in the timing of the expression peak of mpeE , which encodes the linker binding the terminal PEII disk in Synechococcus sp. WH7803 (Six et al., 2007c ). Figure 8 Scheme of the daily patterns of gene upregulation for several important functional for picocyanobacterial cells acclimated to a modulated 12/12 h L/D cycle of high visible light with or without UV radiations . This figure, derived from Figure 7 , shows for each individual the time intervals during which they were significantly upregulated [|log 2 (Fold Change)| > 1.0], with regard to the expression level measured at 6:00 in VL. Differential photoprotection mechanisms Prochlorococcus and Synechococcus have also developed specific mechanisms to dissipate excess light energy. Most Synechococcus strains possess one gene encoding the orange carotenoid protein (OCP; for a recent review, see Kirilovsky and Kerfeld, 2012 ), which is thought to mediate energy dissipation as heat through interaction with the PBS core, thus inducing a NPQ of PSII fluorescence. Here, we observed that the ocp gene was upregulated during daytime in VL and that UVR dramatically enhanced its expression in the early light period that may trigger a temporary increase of the OCP cellular pool and/or activity. To our knowledge, this is a first time that UV is shown to control the expression of this gene since it is usually believed that OCP is only blue light sensitive (Wilson et al., 2008 ; Kirilovsky and Kerfeld, 2012 ). While all Prochlorococcus strains lack OCP, all high light-adapted strains (including PCC 9511) and some low-light-adapted strains possess an homolog of PTOX , which is thought to extract electrons between PSII and PSI and to combine them with protons and oxygen to generate water (Bailey et al., 2005 , 2008 ). The ptox gene was (with ftsZ ) one of the two most differentially expressed genes among those analyzed here in PCC 9511, with a maximal relative fold change over the day of ∼73-fold [or log2(FC) = 6.2] at 15:00. These high values are consistent with previous results on the closely related strain P. marinus MED4 (Zinser et al., 2009 ; Berg et al., 2011 ). So the alternative electron flow to oxygen triggered by PTOX might be an important mechanism used by Prochlorococcus to struggle against excess light energy arising to PSII (Bailey et al., 2008 ; Berg et al., 2011 ), although high light induced proteins (HLIPs), which were not analyzed in the present study, are also likely important contributors in this process (He et al., 2001 ). Other key actors in photoprotection mechanisms are carotenoids. While the cellular localization and precise action mechanism of Zea is unclear yet, Car are known to be mostly bound to PS reaction centers, two of them in close vicinity of Chls, and help mitigate oxidative damages, mainly by quenching the 1 O 2 species resulting from the de-excitation of Chl triplets (Telfer, 2005 ). Both Zea:Chl a and Car:Chl a ratios appeared to be tightly coupled to the L/D cycle in Synechococcus , but less so in Prochlorococcus , for which we observed a continuous increase of both ratios during most of the day and a symmetrical drop during the night concomitant with cell division, as previously noticed by Claustre et al. ( 2002 ). In Synechococcus , the sharp drop of the β-Car:Chl a ratio from noon to dusk, which was more pronounced in UV-exposed cells (Figure 2 C), likely reflects the progressive destruction of β-Car molecules by oxidative stress (Telfer, 2005 ). β-Car was then seemingly regenerated at low rate during the night, then much more rapidly during the first hours of the day, suggesting a light-dependency of the β-Car biosynthesis process, as previously reported in other cyanobacteria (Steiger et al., 2005 ; Ryu et al., 2010 ). Similarly, the light-modulated variations of the Zea:Chl a ratio during the day in Synechococcus in both light conditions could be due either to a higher degradation rate of Chl a relative to Zea or to a higher synthesis rate of the latter pigment, a likely photoprotective mechanism against high midday photon fluxes. The observed strong upregulation of crtR at high irradiances tend to favor the second hypothesis (Figure 7 ). A further increase of the relative crtR transcript levels at noon in VL + UV, which was not translated into a significantly higher midday increase in the Zea:Chl a ratio than in VL (Figure 3 A), suggests that Zea msolecules might have a particularly high turnover under UV. In both genera, the systematically higher values of Zea:Chl a ratio under VL + UV than VL only suggest that UV exposure induces a comparable response as a long-term acclimation to high irradiance, as previously reported in other marine cyanobacterial strains (Kana and Glibert, 1987 ; Moore et al., 1995 ; Six et al., 2004 ). Altogether, our results suggest the occurrence of a light-controlled anabolism/catabolism cycle of both photoprotectants (Car and Zea) in Synechococcus , while diel changes in the pigment content of Prochlorococcus cells were rather related to the cell division cycle or other factors not analyzed here such as PS stoichiometry and/or antenna size. Differential protection mechanisms against reactive oxygen species-induced damages Light-harvesting complexes are not only sources but also major targets of ROS (Sies and Menck, 1992 ; Krieger-Liszkay, 2005 ). Because of their localization within the thylakoid membrane around PSII complexes (Bibby et al., 2003 ), the divinyl-Chl a / b -binding Pcb antennae of P. marinus PCC 9511 likely produce ROS, particularly noxious for this photosystem. Indeed, it has been shown in other Chl b -containing organisms that excited singlet Chls lead to the synthesis of Chl triplet states that can react with 3 O 2 to produce the very reactive species 1 O 2 (Sies and Menck, 1992 ; Krieger-Liszkay, 2005 ). Assays performed here by directly adding H 2 O 2 to sub-cultures at different time points of the modulated L/D cycle strongly suggest that the effect of ROS increased in a light-dependent manner for both organisms. Consistently, a synergistic effect of light and oxidative stress on PSII photoinactivation was previously demonstrated by Blot et al. ( 2011 ) who showed in Synechococcus sp. WH7803 that ROS may induce PSII inactivation through both direct damages to the reaction center II and inhibition of the PSII repair cycle. The latter phenomenon resulted in faster PSII inactivation in high light- than in low-light-acclimated cultures, due to their higher D1 turnover. Furthermore, Prochlorococcus was also more sensitive to H 2 O 2 -triggered stress than Synechococcus during the night, which might be related to a lower intrinsic resistance of its PSII. The difference in ROS sensitivity between Prochlorococcus and Synechococcus strains might be explained, at least partially, by a more complete set of genes involved in ROS protection and detoxification in Synechococcus . Indeed, most Prochlorococcus lineages have experienced an extensive genome streamlining during evolution, resulting in the loss of a large number of non-essential but potentially useful genes in this context (Figure 7 ; Dufresne et al., 2005 ; Scanlan et al., 2009 ; Partensky and Garczarek, 2010 ). For instance, while Synechococcus sp. WH7803 synthesizes one catalase/peroxidase (KatG), Prochlorococcus has neither catalases nor peroxidases. Prochlorococcus also lack the ftrC and ftrV genes, encoding the two subunits of ferredoxin-thioredoxin reductase (FTR), as well as one ferredoxin and one thioredoxin, potentially associated to this complex (Dufresne et al., 2008 ). Furthermore, while Synechococcus possesses two superoxide dismutase (SOD), one Fe-type and one Cu-Zn-type, Prochlorococcus has only one, Ni-type SOD (SodN; Scanlan et al., 2009 ). In the present study, the expression levels of most ROS genes showed strikingly higher amplitudes of variation over the day in Synechococcus than in Prochlorococcus . Even though transcriptomic data are not necessarily synchronized with the activity of the corresponding enzymes, the lower resistance of Prochlorococcus to high VL and UVR could at least partially be due to a higher sensitivity to light-driven oxidative stress."
} | 8,594 |
37777794 | PMC10541700 | pmc | 4,964 | {
"abstract": "Background Microbiome recruitment is influenced by plant host, but how host plant impacts the assembly, functions, and interactions of perennial plant root microbiomes is poorly understood. Here we examined prokaryotic and fungal communities between rhizosphere soils and the root endophytic compartment in two native Miscanthus species ( Miscanthus sinensis and Miscanthus floridulus ) of Taiwan and further explored the roles of host plant on root-associated microbiomes. Results Our results suggest that host plant genetic variation, edaphic factors, and site had effects on the root endophytic and rhizosphere soil microbial community compositions in both Miscanthus sinensis and Miscanthus floridulus , with a greater effect of plant genetic variation observed for the root endophytic communities. Host plant genetic variation also exerted a stronger effect on core prokaryotic communities than on non-core prokaryotic communities in each microhabitat of two Miscanthus species. From rhizosphere soils to root endophytes, prokaryotic co-occurrence network stability increased, but fungal co-occurrence network stability decreased. Furthermore, we found root endophytic microbial communities in two Miscanthus species were more strongly driven by deterministic processes rather than stochastic processes. Root-enriched prokaryotic OTUs belong to Gammaproteobacteria, Alphaproteobacteria, Betaproteobacteria, Sphingobacteriia, and [Saprospirae] both in two Miscanthus species, while prokaryotic taxa enriched in the rhizosphere soil are widely distributed among different phyla. Conclusions We provide empirical evidence that host genetic variation plays important roles in root-associated microbiome in Miscanthus . The results of this study have implications for future bioenergy crop management by providing baseline data to inform translational research to harness the plant microbiome to sustainably increase agriculture productivity. \n Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40168-023-01646-3.",
"conclusion": "Conclusions In this study, we provide comprehensive and empirical evidence on the relative contribution of host and environmental factors to microbiome assembly in M. sinensis and M. floridulus . Our results demonstrate that microbiome assembly is shaped predominantly by host genetic variation, environmental factors, and biogeography. Furthermore, we revealed that host selection reduced root microbial diversity and network complexity compared to the rhizosphere soil. In addition, we also demonstrated that host genetic variation influenced fungal communities more than prokaryotic communities in both roots and rhizosphere soil. These findings significantly advance our current understanding of microbial community assembly in bioenergy crops such as Miscanthus under different environmental selection pressures and highlight the importance of the host selection effect for endophytic functions. Moreover, we provide empirical evidence of ecological filtering from rhizosphere soil to root endophyte compartment and selective enrichment of specific microbial taxa. Miscanthus root appears to select some taxa related to nitrogen fixation, which might contribute to native Miscanthus plant fitness and adaptation to diverse environmental conditions and signal desirable sustainability traits for Miscanthus as a bioenergy feedstock. We further revealed that the variation in the core microbial community was highly associated with host genetic variation in each Miscanthus species. The results of this study have implications for future bioenergy crop management by providing baseline data to inform translational research to harness the plant microbiome to sustainably increase agriculture productivity.",
"discussion": "Discussion Enriched microbial OTUs and core microbial taxa in Miscanthus species We found that members within Gammaproteobacteria, Alphaproteobacteria, Betaproteobacteria, Sphingobacteriia, and [Saprospirae] were significantly enriched in the root for both Miscanthus species, whereas a broad range of taxa were enriched in the rhizosphere soil. The enriched endophytic bacteria identified in miscanthus in this study have also been detected in other plants such as rice, barley, and Arabidopsis thaliana [ 50 – 52 ]. These root-enriched microbial taxa may play key roles in modulating host nitrogen uptake and host fitness [ 53 ]. For example, we found members of enriched taxa belonged to the genera of Azotobacter , Azospirillum , Enterobacter , Herbaspirillum , and Rhizobium that are known diazotrophs [ 54 ], which could have contributed to the nitrogen fixation in the M. sinensis and M. floridulus root. Future research is needed to understand the contribution of N fixation to the N budget of miscanthus, and the ability to recruit endophytic diazotrophs would do much to promote the sustainability of this candidate bioenergy crop [ 55 , 56 ]. The higher number of prokaryotic phyla enriched in the rhizosphere soil for both Miscanthus species suggests that rhizosphere soil may have greater functional diversity and/or redundancy for biogeochemical cycling functions compared with root endophytic communities. We also identified a group of Miscanthus core prokaryotic taxa shared between the rhizosphere soil and the root (Fig. 4 A–D). Most members of the Miscanthus core prokaryotes (e.g. Acidobacteriales, Xanthomonadales, Rhizobiales, Burkholderiales, and Enterobacteriales) overlapped with those identified in other plant species such as Arabidopsis thaliana [ 51 , 57 ] and sugarcane [ 58 ], suggesting that the presence of some core microbial taxa may be common across plant species. Multiple members affiliated with these core species have been verified to exert different types of positive functions on plant health and growth [ 13 ]. The existence of common core microbiota members in various host plants implies that a highly conserved, coevolutionary, and host-independent core plant microbiota may exist that maintains plant holobiont fitness [ 53 , 57 , 59 ]. Host plant genetic variation differentially affected root endophytic and rhizosphere soil microbial community composition We found that the influence of host genetic variation on microbial community composition increased from rhizosphere soil to root endophyte, which is consistent with a study on M. truncatula [ 60 ]. This might be due to an increase in host plant selection for microbial communities in the root compartment relative to the rhizosphere soil [ 52 , 61 ]. We also demonstrated that host genetic variation imposed stronger selection on the fungal community than on the prokaryotic community for both Miscanthus species (Fig. 3 ). The different effects of host genetic variation on fungal and prokaryotic community compositions may be explained by close fungal associations with plants (compared to prokaryotes) since some fungi can form biotrophic interactions with plants, and take the form of root symbionts, endophytes, and pathogens [ 62 ]. Host genetic variation differentially affected core and non-core prokaryotic community composition We demonstrated that host genetic variation has a stronger effect on shaping core prokaryotic community assemblages than non-core prokaryotic community composition in roots and rhizosphere soil of the two Miscanthus species (Fig. 4 E). Potential explanations include core microbiota traits for efficient colonization, nutrient acquisition, and stress tolerance [ 58 , 63 ], which are likely to be particularly important for the host fitness and could result in more close associations with host plants. Although we found that host genetic variation was strongly related to the core prokaryotic community variation in Miscanthus , a previous study on common bean demonstrated no correlation between plant genotypes and core bacterial community [ 16 ]. This difference may be because our study employed a microsatellite approach, which is ideal for characterizing plant genetic variation at individual or population level. In comparison, sampling soil simply from different plant genotypes might suffer from inseparable effects between plant genetics and soil environments. We did not observe core fungal taxa in root or rhizosphere soil in either M. sinensis or M. floridulus . This could be explained by the fact that fungi possess traits with a higher degree of resource specialization and host plant specification compared with prokaryotes [ 64 ]. In addition, fungi have relatively lower dispersal ability compared with prokaryotes [ 65 ], which leads to difficulty in the identification of common core fungal taxa among broader geographic sites as in this study. Microbial co-occurrence network stability changed from rhizosphere soil to root endophyte We found prokaryotic networks in the root compartment were highly modular and dominated by negative interactions compared to rhizosphere soil. On the contrary, fungal network in root compartment had low modularity and was dominated by positive interactions compared to rhizosphere soil (Fig. 5 ). An increase in network modularity and the ratio of negative to positive cohesion from rhizosphere soil to root endosphere support the evidence that the host stabilizes prokaryotic communities by restricting species’ responses within small network modules, thereby avoiding propagation of the effect to the remaining network [ 66 , 67 ]. The differences in the co-occurrence patterns between prokaryotes and fungi suggest that hosts might face contrasting tradeoffs between network stability and metabolic efficiency based on Coyte et al. (2015) [ 66 ]: for prokaryotic communities, we observed increased negative interactions in the host-associated microbial communities that could improve ecological stability, but at the cost of decreasing overall metabolic efficiency. However, for fungal communities, hosts might benefit from fungal cooperation to improve metabolic efficiency such that host plants prioritize the positivity of network interactions over ecological stability."
} | 2,532 |
26601203 | PMC4640615 | pmc | 4,965 | {
"abstract": "Coral reef community responses vary but bioerosion increases under natural ocean acidification.",
"introduction": "INTRODUCTION Shifts in ocean chemistry are likely occurring more rapidly now than in the past 300 million years ( 1 ). Excess CO 2 released from fossil fuel emissions and deforestation is absorbed by the surface oceans, driving down seawater pH and calcium carbonate (CaCO 3 ) saturation state (Ω), a process termed ocean acidification (OA) ( 2 ). Coral reefs are considered especially vulnerable to OA. Reefs are made of CaCO 3 produced by calcifying organisms, including corals and coralline algae, and laboratory experiments have shown that biogenic calcification is slowed and its destruction is accelerated at levels of OA projected for the end of this century ( 2 , 3 ). Some experiments have raised key questions regarding the potential for coral reef organisms to adapt to OA or for covarying environmental factors, such as light, water flow, and nutrient availability, to modulate the impacts of OA ( 4 – 8 ). However, most studies of naturally low-pH reefs, including CO 2 vents in Papua New Guinea (PNG) and Japan, freshwater seeps in Mexico, and upwelling regions of the eastern tropical Pacific, have yielded no evidence to support either scenario ( 9 , 10 ). In our previous report ( 11 ), we presented our initial data showing that coral communities of the Palau Rock Islands, formed by a labyrinthine maze of uplifted karst, appear healthy despite the relatively extreme pH conditions in which they live. As water flows from the open ocean over the barrier reefs and into the Rock Island bays, its carbonate system chemistry is altered by a combination of biological and hydrographic processes that elevate p CO 2 and drive down pH and Ω, a natural form of acidification ( 11 ). The long residence time of seawater within the Rock Islands exacerbates this process, and as a result, the benthic communities in Palau’s most acidified reefs live in conditions with pH and Ω levels equivalent to those predicted for the western tropical Pacific open ocean by 2100 ( Fig. 1 ) ( 12 ). Recent continuous pH data collected in situ over multiple, consecutive diel cycles ( Fig. 1 , inset) reveal that whereas natural acidification decreases the mean pH within the Rock Islands, the diel range in pH is maintained across Palau’s natural OA gradient. The downward shift in pH and Ω without change in amplitude or frequency of variability contrasts the extreme and highly variable conditions at CO 2 vent sites and freshwater seeps ( 13 , 14 ), and is more consistent with the predicted nature of the progression of OA in the marine environment ( 12 ). Fig. 1 Natural acidification gradient across Palau mirrors projected anthropogenic CO 2 -driven changes in ocean chemistry. Shown are the mean (±1 SD) dawn-to-dusk pH and Ω ar for our 11 reef study sites and diurnal pH variability at three of those sites over 4 days (inset) (fig. S1: sites 1, 2, and 10). Shaded regions indicate the range of western tropical Pacific open ocean pH and Ω ar levels in 2000 (blue) and open ocean values predicted for 2050 (orange) and 2100 (red) ( 12 , 52 ). Colored points correspond to pH time series in inset. Here, we assess coral reef benthic community structure and key ecosystem processes across a natural gradient in seawater pH and Ω in Palau. We evaluate these response variables against a comprehensive characterization of carbon chemistry to investigate whether measureable changes in benthic community structure; coral community composition; declines in skeletal extension, density, and calcification; and/or increases in the prevalence and rates of bioerosion seen in laboratory experiments and analog sites can be detected and attributed to OA in Palau. Finally, we compare our data with those collected at other naturally low-pH reefs to identify response variables common across all sites, independent of the mechanism of acidification, biogeography, frequencies of variability, and presence of other environmental factors that may exacerbate or mask the OA impact.",
"discussion": "DISCUSSION Despite the pH and Ω ar conditions already at predicted end-of-century open ocean levels and p CO 2 up to 720 μatm, the Rock Islands support high coral cover, richness, and diversity and very low macroalgae cover. This observation counters expectations based on some laboratory CO 2 manipulation experiments and studies of other naturally low-pH reefs in which severe declines in coral richness and coralline algal cover and increases in macroalgae are signature impacts of OA ( 13 , 14 ). In general, coral cover on Palau’s high-pH barrier reefs (28 to 37%) was lower than that of the low-pH bay reefs (32 to 63%), a trend likely exacerbated by a bleaching event in 1998 that caused declines in coral cover on the barrier reef but not the bays. Nevertheless, the relatively high cover and diversity in the low-pH reefs cannot be solely attributed to differential bleaching-induced mortality: the pre-1998 cover on the barrier (50% in 1992) was already lower than the current coral cover on Palau’s lowest-pH reefs (63%) ( 18 ). The skeletal extension, density, and calcification rates of Porites and Favia corals did not change significantly with declining pH and Ω ar , indicating that the rates of CaCO 3 production, a physiological process considered one of the most sensitive to OA, are maintained across Palau’s OA gradient. In contrast, many laboratory CO 2 manipulation experiments with Porites and Favia corals have shown significant declines in calcification of these genera with declining pH/Ω ar ( 7 , 14 , 19 – 21 ). We observed significant changes in coral community structure that track changes in pH and Ω ar . However, the compositional heterogeneity among low-pH, low-Ω ar sites on Palau suggests that coral community structure under OA conditions is not deterministic, and there is no single community of acidification “winners” within Palau’s low-pH reefs. Indeed, other local environmental and/or ecological factors, including changes in wave energy, temperature, and/or light, all of which covary as pH decreases from the barrier reefs to the Rock Island bays, may play a larger role in shaping community composition. Several observations support this interpretation. First, the shift from offshore Acropora -abundant to inshore Porites -abundant communities is consistent with worldwide reef zonation patterns ( 22 ). Second, the presence and abundance of other genera (including Favia and Favites ) did not change with decreasing pH. Finally, a number of branching and foliose genera (for example, Pachyseris , Anacropora , Mycedium , Merulina , and Lobophyllia ) typically associated with lower wave energy and/or light levels, and not considered insensitive to pH ( 23 ), were more abundant on Palau’s lowest-pH reefs than on the high-pH barrier reefs. We also found that coral macrobioerosion increases significantly as pH decreases in Palau. In contrast to the change in community composition, factors other than pH are unlikely to explain this trend. Across multiple studies of coral bioerosion, there is no consistent correlation between macrobioerosion rates and degrees of wave exposure or intensity of flow ( 24 – 28 ). Elevated nutrient concentrations have been shown to correlate with increased coral macrobioerosion ( 15 , 29 ), but nutrient concentrations do not change with decreasing seawater pH or with increased bioerosion in Palau. Our hypothesis that low pH causes the observed increase in coral macrobioerosion is supported by multiple laboratory experiments that show elevated bioerosion at low pH ( 30 – 32 ). Furthermore, in a recent field study, pH/Ω ar emerged as a consistent factor in Porites macrobioerosion on 11 Pacific reef systems ( 15 ) spanning a wide range of pH, wave energy, and flow conditions ( 33 ). There are several potential mechanisms by which OA can cause increased macrobioerosion of live corals. One hypothesis is that skeletons accreted under OA are less dense, making them easier for bioeroders to penetrate ( 34 ). Although low-pH conditions in Palau do not necessarily produce lower-density skeletons, colonies with less dense skeletons are, in general, more likely to show evidence of bioerosion. In addition, lower pH may facilitate bioerosion by increasing the efficiency of biochemical dissolution ( 30 , 35 ), one of the methods Lithophaga employs to excavate coral skeletons ( 36 ). We compared our results with those obtained from similar analyses of naturally low-pH coral reef ecosystems near volcanic CO 2 vents in Milne Bay, PNG ( 13 ), submarine freshwater springs ( ojos ) in Puerto Morelos, Mexico ( 14 , 37 ), and low-pH upwelling zones in the eastern tropical Pacific ( 38 , 39 ) in an effort to identify common response variables solely attributable to changes in pH ( Table 1 ). Despite the paucity of naturally more acidified reefs identified to date, across-site comparisons are necessary because none are perfect analogs for coral reefs under future OA. CO 2 accumulation is not always the primary or only driver of low pH/Ω ar , and pH variability can be extreme relative to projected future values, as well as spatially and temporally heterogeneous. Covariability among pH and other variables including nutrients, salinity, light, water flow, and/or temperature can make it difficult to attribute specific ecological changes solely to acidification. Furthermore, scales of larval connectivity and recruitment sources vary between sites and may influence the adaptive potential of coral communities. Table 1 Diverse reef responses to natural acidification in the Palau Rock Islands, PNG CO 2 vents, Mexico ojos , and eastern tropical Pacific (ETP) upwelling regions. For each site, response variables are reported as the ratio of each variable in low/high pH sites. + indicates a significant increase in the response variable from high to low pH, and − indicates a significant decrease. n.d., no data available. Palau * PNG † Mexico ‡ ETP § Hard coral cover 1.9 1.1 0.5 (−) 0.0 (−) Macroalgae cover 0.7 2.1 (+) n.d. n.d. Coralline algae cover 1.1 0.2 (−) n.d. n.d. Hard coral richness 1.6 0.6 (−) 0.3 (−) 0.2 (−) Porites cover 16.0 (+) 2.3 (+) 0.8 n.d. Porites extension 1.0 1.1 1.0 0.6 Porites density 0.8 1.0 0.8 (−) 0.8 (−) Porites calcification 0.8 1.1 0.7 (−) 0.5 (−) Bioerosion ¶ 11.3 (+) 1.9 (+) 1.8 (+) 1.9 (+) *For Palau, ratios are calculated for the two lowest and the two highest Ω ar reefs, and the indicated significance is for the trend across all sites (Ω ar = 3.7 to 2.3). †For PNG, ratios and trends are reported for Ω ar = 3.5 to 2.9 ( 13 ). ‡Community data for Mexico are reported for Ω ar > 2.5 to Ω ar < 2.5 ( 37 ), and skeletal growth parameters are reported for Ω ar > 2 to Ω ar < 2 ( 14 ). §For the ETP, hard coral cover, hard coral richness, and Porites extension, density, and calcification data are reported for four reef sites within the Galapagos (Ω ar = 3.3 to 2.4) ( 39 ). Porites macrobioerosion rates are compared across the Galapagos, the Gulf of Panama, and Gulf of Chiriquí (Ω ar = 3.5 to 2.5) ( 38 ). ¶Trends in bioerosion are estimated by the percent volume of macrobioerosion of Porites skeletal cores (Palau and Mexico), Porites bioeroder density (PNG), or bioerosion rate (ETP). Our across-site comparison reveals few commonalities among low-pH reefs studied to date ( Table 1 ). Despite comparable natural gradients in pH/Ω ar , trends in coral cover and richness and cover of macroalgae and coralline algae are inconsistent, with the Rock Islands unique in showing no sensitivity to pH in any of these response variables. Coral richness declined steeply with declining pH in Mexico, PNG, and the eastern tropical Pacific, but did not change in Palau. Macroalgae cover increased at low pH in PNG only. On a CO 2 reef vent site in Japan (not included in Table 1 ), hard coral–dominated communities gave way to soft coral–dominated reef communities as pH declined from the fringing reef to the shallow back reef pools cut off from the ocean at low tide ( 40 ). This did not occur at the PNG CO 2 vent site. Porites abundance increased with pH decline in Palau and PNG, but it was not affected by low pH in Mexico. Porites calcification in Palau and PNG was insensitive to decreasing pH, whereas in Mexico, the eastern tropical Pacific, and in laboratory OA experiments, Porites calcification declined with low pH ( 7 , 19 , 21 , 41 ). The inconsistencies in community responses to acidification across naturally low-pH reef systems, and between reefs and laboratory experiments, may be due to a number of different factors. Although ultimately producing similar average pH and Ω ar conditions, distinct mechanisms of acidification at each site lead to differences in extremes and frequencies of variability. For example, whereas the maximum seawater p CO 2 levels in the Palau Rock Islands ( p CO 2 ~720 μatm) are close to the 2100 AD projections for the open ocean, the maximum concentrations at the Yucatan ( p CO 2 ~5120 μatm) ( 14 ) and PNG ( p CO 2 ~5740 μatm) ( 13 ) sites are about seven to eight times higher. pH variability in Palau is dominated by the diurnal and tidal cycles as it will be on future reefs, whereas PNG and Mexico are characterized by high-frequency spikes associated with pulses of CO 2 and groundwater discharge ( 13 , 14 , 42 ). Yet, the hydrographic and biological processes that lower pH/Ω ar in Palau also result in low total alkalinity (TA) and dissolved inorganic carbon (DIC) conditions, which are not expected under future OA. Furthermore, in the eastern tropical Pacific, temperature and nutrient concentrations covary with pH, making it difficult to attribute patterns in reef communities solely to upwelling-driven acidification ( 39 ). Across the handful of naturally low-pH sites studied to date, Palau appears unique in showing no obvious sensitivity to OA across a range of key ecological indices. Laboratory experiments suggest that covarying environmental factors including light, flow, nutrients, and food availability can modulate the negative impact of OA on calcification. None of these factors can explain the apparent OA tolerance of Palau’s benthic communities. Most factors that alleviate the impacts of pH in laboratory experiments are absent from Palau’s Rock Island bays, where high temperature, high shade, low flow, and low nutrient concentrations accompany OA conditions ( 4 , 5 , 8 , 18 ) (table S1). The low-pH communities in Palau differ from other naturally low-pH sites studied to date in their relative isolation within bays and inlets and in OA conditions that are chronic and of less extreme ranges. It is possible that selective pressure has driven local, community-wide adaptation to low pH over long time scales (that is, thousands of years). In contrast, in PNG and Mexico, low-pH communities are more exposed, scales of connectivity may be larger than in low-pH areas, and the possibility that larvae are recruited to low-pH reefs from populations under less acidification pressure may preclude adaptation ( 11 ). Increases in macrobioerosion with declining pH are the only consistent coral reef community response to natural acidification across low-pH reefs. In Palau, bioerosion increased significantly at low Ω ar . The same patterns have been reported at low-pH reef sites in PNG (93% increase in bioeroder density from high pH to low pH) and Mexico (78% increase in Porites coral skeleton eroded) ( 13 , 14 ). In the eastern tropical Pacific, elevated rates of reef bioerosion (87% increase from Ω ar = 3.5 to Ω ar = 2.5) and decreased reef cementation have also been observed in low-pH, nutrient-rich upwelling zones ( 38 ), a component of which is attributable to low pH ( 15 ). Enhanced bioerosion and lack of cementation threaten the structural integrity of corals and reef systems, increasing the impact of predation and the risk of physical destruction by storms ( 30 , 43 ). Moreover, if macrobioerosion is indicative of larger-scale reef erosion and dissolution ( 15 , 25 , 44 ), increases in CaCO 3 erosion under OA will have significant implications for the persistence of reef structures. Structural fragility is incompatible with the ability of barrier and fringing reefs to absorb and dissipate the energy associated with the constant day-to-day pounding of waves, seasonal storms, and the less frequent but more catastrophic tsunamis. Indeed, that bioerosion rates are elevated in all naturally low-pH coral reef systems studied to date suggests that the signature of 21st century OA will emerge most strongly and most universally through its impact on reef structural integrity."
} | 4,197 |
33328495 | PMC7744511 | pmc | 4,966 | {
"abstract": "Developing an efficient deconstruction step of woody biomass for biorefinery has been drawing considerable attention since its xylem cell walls display highly recalcitrance nature. Here, we explored transcriptional factors (TFs) that reduce wood recalcitrance and improve saccharification efficiency in Populus species. First, 33 TF genes up-regulated during poplar wood formation were selected as potential regulators of xylem cell wall structure. The transgenic hybrid aspens ( Populus tremula × Populus tremuloides ) overexpressing each selected TF gene were screened for in vitro enzymatic saccharification. Of these, four transgenic seedlings overexpressing previously uncharacterized TF genes increased total glucan hydrolysis on average compared to control. The best performing lines overexpressing Pt × tERF123 and Pt × tZHD14 were further grown to form mature xylem in the greenhouse. Notably, the xylem cell walls exhibited significantly increased total xylan hydrolysis as well as initial hydrolysis rates of glucan. The increased saccharification of Pt × tERF123 -overexpressing lines could reflect the improved balance of cell wall components, i.e., high cellulose and low xylan and lignin content, which could be caused by upregulation of cellulose synthase genes upon the expression of Pt × tERF123 . Overall, we successfully identified Pt × tERF123 and Pt × tZHD14 as effective targets for reducing cell wall recalcitrance and improving the enzymatic degradation of woody plant biomass.",
"introduction": "Introduction Woody plants are the most abundant source of terrestrial biomass and have been industrially used for construction, paper, energy and many materials and chemicals. Recently, wood has also been considered as a promising sustainable resource for the production of biofuels and other high-value products 1 . However, the efficient conversion of lignocellulose into simple sugars (glucose and xylose) has been stymied by its complex and recalcitrant structures. Therefore, recent studies have focused on optimizing wood cell wall characteristics and traits for conversion processes including more efficient enzymatic deconstruction using gene modification and metabolic engineering 2 . Populus species (poplars, aspens and cottonwoods) are excellent models of woody plant biomass because of their worldwide distribution, fast growth, genomic resources, and technical advances such as the well-established transformation system 3 . Populus stems include many different cell types which differentiate from the cambium into vessel elements, fibers, and axial and radial parenchyma cells. The dominant material in woody biomass is the secondary xylem cell wall of fiber cells. In these cells, a thin and stretchy primary cell wall (PCW) layer is first deposited during cell expansion. Afterwards, a thick and rigid secondary cell wall (SCW) layer is deposited inside the PCW layer. The PCW is mainly constituted of cellulose, xyloglucan, and pectin with negligible amount of lignin, whereas the SCW consists of cellulose, xylan, mannan and lignin 4 , 5 . Genes encoding enzymes involved in the biosynthesis of cell wall components such as cellulose, hemicelluloses and lignin have been relatively well characterized in Populus species (reviewed by Ye and Zhong 6 ). Expression of the cell wall biosynthetic enzymes is controlled by a number of transcriptional factors (TFs). NO APICAL MERISTEM/ARABIDOPSIS TRANSCRIPTION ACTIVATION FACTOR/CUP-SHAPED COTYLEDON (NAC) and MYB families genes play critical roles in regulating xylem cell differentiation and cell wall thickening 7 , 8 . Homologous genes of Arabidopsis thaliana VASCULAR-RELATED NAC DOMAIN (VND) 9 , NAC SECONDARY WALL THICKENING PROMOTING FACTOR/SECONDARY WALL-ASSOCIATED NAC DOMAIN (NST/SND) 10 , 11 and SOMBRERO (SMB) proteins designated as VNS (VND, NST/SND,and SMB RELATED) proteins 12 are key regulators of secondary cell wall formation in xylem fibers, phloem fibers, and xylem ray parenchyma cells in Populus 12 – 16 . Among these, VNS09, VNS10, VNS11 and VNS12 corporately play important roles as a master switch regulators of cell wall formation in Populus 17 , 18 , but their downstream genes including several TFs are functionally yet uncharacterized. In addition to these studies of A. thaliana TF homologs in Populus species, more comprehensive analyses have focused on TFs expressed during the wood formation processes including secondary xylem differentiation, cell expansion along with PCW deposition, SCW deposition, and programmed cell death along with further lignification 19 – 21 . Transcriptomic analyses revealed more than 1000 TF genes that were potentially involved in the development of secondary xylem cells 19 , 20 . However, most of TFs are uncharacterized, with the exception of major TF families such as NAC and MYB 22 , 23 . Furthermore, yet only a few studies have succeeded in improving biomass quality by manipulating Populus TF genes 24 , 25 . Thus, our understanding of the TFs involved in the regulatory network system of Populus xylem cell wall formation and the potential for application in a bioengineering context is still limited. Toward reducing wood recalcitrance and improving its enzymatic degradability, we focused on TF genes associated with wood formation as xylem cell wall modification targets, since most previous studies have focused on genes involved in cell wall biosynthesis or cell wall-degrading/-modifying enzymes in Populus 26 , 27 . In comparison, TFs can affect the synthesis of multiple cell wall components and thereby have greater effects. We selected 33 candidate TF genes that are highly up-regulated during wood formation, over-expressed them in hybrid aspen, Populus tremula × Populus tremuloides ( Pt × t , wild-type clone T89), and analyzed their saccharification properties.",
"discussion": "Discussion In this study, we successfully generated and screened transgenic hybrid aspen overexpressing 33 TFs with a potential role in regulating xylem cell wall biosynthesis (Table 1 ; Fig. 1 A). Four of the previously uncharacterized TFs (Pt × tERF123, Pt × tZHD14, Pt × tTCL1 and Pt × tWLIM2B) had increased saccharification of seedling tissue on average when overexpressed (Fig. 1 B). Among them, all lines of Pt × tERF123ox and Pt × tZHD14ox which showed the excellent saccharification performance at the seedling screening stage were further grown to form mature xylem in the greenhouse. Consequently, most of the tested Pt × tERF123ox and Pt × tZHD14ox lines displayed significantly increased initial glucan hydrolysis rates (glucose release after 3 h-treatment of enzyme cocktail) albeit with less significant enhancement in the total glucan degradation (glucose release after 48 h-treatment of enzyme cocktail) by enzymatic saccharification (Fig. 3 ). Notably, all the Pt × tERF123ox and Pt × tZHD14ox lines displayed increased xylan degradations (xylose release after 48 h-treatment of enzyme cocktail) with 1.5–2.0 times higher xylose yields compared to the controls. Xylan is thought to be one of the limiting factors in enzymatic hydrolysis of cellulose possibly because xylan physically covers cellulose surface in lignocellulose and thereby hinders the access of cellulolytic enzymes to cellulose substrate 43 . Therefore, easier removal of xylan may have enhanced cellulose hydrolysis of the Pt × tERF123ox and Pt × tZHD14ox cell walls. Many studies have investigated the relationship between lignocellulose structure and enzymatic saccharification performance of cell wall materials. Major factors that affect cell wall saccharification performance are thought to be glucan contents as well as the recalcitrant lignin content and structure which substantially prevents the access of hydrolytic enzymes to cellulose and hemicellulose substrates and subsequent hydrolysis reactions 44 , 45 . Indeed, our chemical analysis data showed that Pt × tERF123ox displayed relatively increased glucan and significantly reduced lignin content, albeit with no apparent change in lignin aromatic composition, compared to the control (Table 2 ). Thus, it is plausible that the improved saccharification performance of Pt × tERF123ox is due to the reduced lignin recalcitrance, besides the reduced xylan barrier. The alteration of cell wall composition in Pt × tERF123ox, i.e., reduced lignin and increased cellulosic glucan levels, could be attributed to the up-regulation of a series of cellulose synthase genes upon overexpression of Pt × tERF123 (Fig. 4 ). Previously, several Populus ERFs were reported to be involved in the formation of tension wood (TW) which typically produce cell walls with reduced lignin and increased cellulose compared to normal wood cell walls 34 . In addition, a member of ERF gene family in A. thaliana , AtERF035 , was recently identified as an active regulator of PCW formation; it was demonstrated that overexpression of AtERF035 substantially enrich pectin and cellulose proportionally over lignin in the cell walls through induction of PCW formation 41 . Intriguingly, Pt × tERF123 was significantly up-regulated during the cell expansion together with PCW-associated enzyme genes (Figure S2 ), and CESA1-B, CESA3-A and CESA6-A associated with PCW formation were significantly upregulated (Fig. 4 ). Therefore, it is possible that the altered lignocellulose composition detected in the Pt × tERF123ox cell walls (Table 2 ) might be associated with cell wall alteration analogous to TW and/or PCW formation(s), although our current histochemical and cell wall chemical analysis data do not entirely corroborate this hypothesis (Fig. 2 and Table 2 ). Further rigorous analyses on the transcription network associated with Pt × tERF123 as well as cell wall structures of Pt × tERF123 -overexpressing and/or -downregulated transgenic aspens are needed to clarify this aspect. In contrast to Pt × tERF123ox, the improved saccharification of Pt × tZHD14ox cell walls is unlikely to be associated with compositional change of lignin, xylan and cellulose in the cell walls (Table 2 ). Besides lignin content and structure, several other factors, such as cellulose crystallinity, hemicellulose modifications, and covalent and non-covalent linkages between lignin and polysaccharides, have been proposed as potential factors that may affect cell wall saccharification performance 46 , 47 . Therefore, future study will need to focus on further structural analyses on the Pt × tZHD14ox cell walls to investigate these factors. In this context, we note that a homologous gene of Pt × tZHD14 in A. thaliana ( AtZHD9/AtHB34 ) was expressed in various tissues including inflorescence stems 35 , 48 , but its function remains unclear. In this study, we observed that cell wall thickness was fluctuated along with the introduced expression level of Pt × tZHD14 (Fig. 2 E), and contents of some hemicellulosic sugars in cell walls were also affected in the Pt × tZHD14ox poplar lines (Table 2 ). Given that Pt × tZHD14 is substantially up-regulated during SCW formation together with SCW biosynthetic genes (Figure S2 , Fig. 4 ), it is plausible that Pt × tZHD14 is involved in xylem development in aspen. Nevertheless, as also noted for Pt × tERF123, further detailed analysis on the transcription network associated with Pt × tZHD14 is needed to elucidate its role in xylem cell wall development in Populus . While we found 4 promising TF-overexpressing hybrid aspens with improved cell wall saccharification performance, the other 29 TF-overexpressing lines showing decreased saccharification performance (Fig. 1 B) are also of interest for further investigation of their potential role in xylem development in Populus . As previously reported, overexpression of both strong transcriptional repressors and/or activators of SCW are expected to cause a decrease of enzymatic digestibility along with striking loss or accumulation of secondary cell wall in the xylem. For example, overexpression of the strong repressors of SCW formation, KNAT7 and MYB199 , were reported to result in remarkably thin xylem Populus cell walls 31 , 32 . Also, in the case of strong activators, overexpression of MYB003 and MYB152 in planta deposited ectopic and/or thick cell walls 28 , 29 . Overall, our results suggested that only approximately 10% of the selected TF up-regulated during wood formation increased enzymatic saccharification, therefore, the first seedling screening method in this study was effective to find transgenic lines with the improve enzymatic digestibility. In Arabidopsis , a similar approach successfully identified several cell-wall-related TF genes that increase cell wall saccharification efficiency 49 . The cultivation and the application of genetically modified organisms are controversial. Recent technology of genome editing has drawn attention to solve this problem since it will enables modulations of gene expression and function without leaving foreign gene in a cell 50 . For example, gene expression can be enhanced by altering cis sequence in the upstream region of associated TF genes, or by silencing negative effectors, and also protein function can be modulated by amino acid substitution. The biomass-recalcitrance-associated TFs identified in this study can be considered as promising targets to improve poplar cell wall properties using such genome editing strategies. Moreover, given that biomass conversion often needs pre-treatment before enzymatic saccharification 1 , future study may further investigate the saccharification performance of the TF-overexpressing aspen lines identified in this study using various pretreatment strategies. Collectively, our strategy to target and screen an array of TFs up-regulated during wood formation successfully constructed transgenic hybrid aspens with reduced xylem cell wall recalcitrance of xylem cell walls without apparent biomass loss. The observation that all cell wall components were coordinately changed in the transgenic aspens in this study is likely due to our target TFs regulating xylem cell wall formation at a global level, not a single cell wall component. Furthermore, the identified TF genes such as Pt × tERF123 and Pt × tZHD14 are widely distributed in various angiosperms 34 , 35 , therefore, these may be new molecular breeding targets for the improved enzymatic digestibility of various biomass, especially other woody biomass crops, such as eucalyptus and willow."
} | 3,628 |
39275109 | PMC11397719 | pmc | 4,969 | {
"abstract": "Among polymer wastes, poly(ethylene terephthalate) (PET) is the most important commercial thermoplastic polyester. Less than 30% of total PET production is recycled into new products. Therefore, large amounts of waste PET need to be recycled. We describe a feasible approach for the direct application of the glycolysis products of PET (GP-PET), without further purification, for the synthesis of value-added products. It was established that GP-PET is valorized via phosphorylation with phenylphosphonic dichloride (PPD), as well as with trimethyl phosphate (TMP). When PPD is used, a condensation reaction takes place with the evolution of hydrogen chloride. During the interaction between GP-PET and TMP, the following reactions take place simultaneously: a transesterification with the participation of the hydroxyl group of GP-PET and the methoxy group of TMP and an exchange reaction between the ester group of GP-PET and the methyl ester group of TMP. The occurrence of the exchange reaction was confirmed by 1 H, 31 P, 13 C NMR, and GPC analysis. Thermogravimetric analysis (TGA) revealed that the percentage of a carbon residual (CR) implies the possibility of using the end products as flame retardant (FR) additives, especially for polyurethanes as well as thermal stabilizers of polymer materials or Li-ion cells.",
"conclusion": "4. Conclusions In this study, the possibility of valorizing GP-PET was successfully proven, yielding high-value products. It was found that GP-PET can be directly used as a source for phosphorus-containing oligomers and monomers via a polycondensation reaction with phenylphosphonic dichloride or transesterification and an exchange reaction with trimethyl phosphate. The polycondensation reaction between GP-PET and phenylphosphonic dichloride proceeds in the presence of triethylamine at a molar ratio of 1:1 to obtain a soft, wax-like oligomer after heating at 50 °C for 8 h. In the repeating unit, the oligomer contains a phosphorus atom and an aromatic group, which are expected to increase fire retardancy. The thermogravimetric analysis revealed that the char residue of GP-PET/PPD is around 17%, which suggests the possibility of its application as a flame retardant of polymers (polyurethanes, PET). The interaction between GP-PET and trimethyl phosphate at molar ratio of 1:2 at 190 °C resulted in the formation of triesters of phosphoric acid, possessing reactive groups, which allows new phosphorus-containing products to be obtained on their basis. In addition, triesters of phosphoric acid have been used as thermal stabilizers of polymer materials or Li-ion cells [ 40 , 41 , 42 ].",
"introduction": "1. Introduction In today’s modern society, plastic is an integral part of our daily lives. It has been shown that the most commonly used industrial polymers are not obtained from sustainable sources such as recycling, reuse processes, or renewable sources, and they are not bio-based or biodegradable [ 1 ]. Therefore, problems related to environmental pollution are inevitable. According to statistics, by 2022, about 10.5 Gt of plastics was produced worldwide, of which 6.5 Gt were scattered as waste [ 2 ]. Among these wastes, poly(ethylene terephthalate) (PET) is the most significant commercial thermoplastic polyester. Thanks to its excellent properties, such as thermal stability, mechanical strength, low gas permeability, and nontoxic nature, PET is a widely used polymer in the textile industry and in the production of food and beverage packaging. This polyester has been used in many electronic and electrical applications, especially those requiring high-temperature performance. By 2025, the global demand for the material is expected to reach 22.36 million tons [ 3 ]. On the other hand, 95% of the produced PET is disposed of as waste within a year, which is due to its short lifespan, as it is mainly used for packaging, and in most countries, <30% of discarded PET is recycled [ 4 ]. To solve this problem, researchers and chemical engineers have focused their efforts on recycling PET waste into high-value-added products. Among the possible recycling techniques, the most acceptable method that follows the principle of sustainable development is chemical recycling, mainly because it can lead to the formation of monomers/oligomers from which the polymer was made [ 5 ]. Bis(2-hydroxyethyl) terephthalate (BHET) obtained from the glycolysis process may be utilized as a starting material for a new synthesis of PET, rigid polyurethane [ 6 ], and bio-resorbable polyester [ 7 ], while the oligomers can be used to make polymers [ 8 , 9 ], hydrophobic dyestuffs [ 10 ], textile auxiliaries [ 11 ], water-soluble polyester coatings [ 12 ], and polymeric plasticizer [ 13 ]. Polymer materials are characterized by increased flammability. This can seriously endanger human life, cause much property damage, and limit the application of polymers in many fields. This problem has drawn the attention of researchers, producers, and government regulatory bodies to the creation of polymers with flame resistance [ 14 ]. Major developments for imparting flame resistance to polymers have been reviewed in the literature [ 15 , 16 ]. These methods include the physical addition of halogen/phosphorus-containing additives to the polymer or the chemical incorporation of FR monomers into the polymer chain. Typical FR additives can be classified into two categories—inorganic (metal hydroxides) and halogen-containing compounds [ 17 , 18 ]. Currently, they should meet requirements such as being non-toxic and environmentally friendly [ 19 ]. During combustion, halogen-containing additives emit toxic gases, making them not environmentally friendly, which is why phosphorus-based additives are broadly applicable [ 20 ]. From this class of FRs, polyphosphonates [ 21 ] and TMP [ 22 ] are often used. PPD is used as a starting monomer in the synthesis of poly(arylphosphonates). These polymers have good FR, as indicated by their high values of limited oxygen index (LOI) of 50–60 [ 23 ]. Polyphosphonates synthesized by the polycondensation of PPD with poly (ethylene glycol) 12,000 with and without bisphenol A show good LOI values (in the range 28–38) [ 24 ]. TMP is used in a variety of industrial processes, including as an FR additive, solvent, and methylating agent for chemical reactions, as a fiber color inhibitor, as an intermediate for pesticides, and as a polymerization catalyst in industry and pharmaceuticals [ 25 ]. TMP is also used as a thermal stabilizer for the production of PET [ 26 ]. It has been shown that the thermal stability of Li-ion cells can be improved by using TMP-containing electrolytes [ 27 ]. The aim of the present research is to find new fields of application for the products of the glycolysis of PET, namely, additives for polymers, giving them new properties or improving their existing ones. In the present experiment, in order to develop a circular production model, reduce the cost of product separation and the drawbacks of plastic waste treatment, and increase the value of recycled products, we describe the results of a feasible approach for the direct application of the product of glycolysis of PET (GP-PET). GP-PET [ 28 ] is a well-defined mixture of monomers, dimers, trimers (and other oligomers), and ethylene glycol (EG) and can be used without further purification for the synthesis of phosphorus-containing compounds by polycondensation with PPD and TMP. The resulting products are phosphorus-containing oligomers. The use of oligomeric analogs of a polymer as an additive could be a good option because it would impart good compatibility between the oligomeric additive and the polymer. To the best of our knowledge, there are no previous works in the literature describing the valorization of the products of the glycolysis of PET via phosphorylation with PPD and TMP.",
"discussion": "2. Results and Discussion 2.1. Interaction between GP-PET and PPD The literature does not contain 31 P NMR data for phosphorus atoms with surroundings, as in the expected reaction products obtained from the interaction between PPD and the products of PET glycolysis (GP-PET). An analysis of GP-PET [ 28 ] revealed that approximately 50% of its composition is BHET. Based on this, and for the purpose of the signal assignment, two model reactions between BHET (commercial product) and PPD were performed at molar ratio BHET/PPD = 1:1 and = 2:1 ( Scheme 1 ) (details about the experimental procedure of the model reactions are reported in the Supplementary Materials ). 2.1.1. The Model Reactions—Interaction between BHET and PPD In the 31 P{H} NMR spectrum of the reaction product (BPClTEA) obtained at a molar ratio of 1:1 ( Figure 1 a), there are signals at 20.69 ppm, 20.20 ppm, 11.03 ppm, 10.85 ppm, and −5.45 ppm with integral intensities 0.03, 1.00, 0.08, and 0.02, respectively. The signals at 20.69 ppm and 20.20 ppm are characteristic of the mono and diesters of phenylphosphonic acid [ 29 , 30 ]. The signals at 11.03 ppm and −5.45 ppm should be attributed to phenylphosphonic acid [ 31 ] and pyrophosphonate structures [ 31 ], respectively. The 31 P{H} NMR spectrum of the reaction product (2BPClTEA) obtained at a molar ratio of 2:1 ( Figure 1 b) showed a signal at 20.02 ppm. The signals at 20.20 ppm (molar ratio 1:1) and 20.02 ppm (molar ratio of 2:1) have a significantly stronger integral intensity, which gives us reason to attribute them to the phosphorus atom in the repeating unit of I and in di[bis(2-hydroxyethylterephthalate)] phenylphosphonate II . The phosphorus atoms in both products have the same surroundings. In the 1 H NMR spectrum of BPClTEA ( Figure S1 ), the signal at 3.89 ppm, a triplet with a coupling constant 3 J(H, H) = 4 Hz, refers to HOC H 2 CH 2 -; the triplet at 4.44 ppm relates to methylene protons of HOCH 2 C H 2 -OC(O). In the range of 7.29–8.02 ppm, aromatic protons should be attributed to the hydrogen atoms of BHET and the aromatic nucleus of PPD. Signals in the region 4.34–4.24 ppm, representing multiplets, should refer to -CH 2 CH 2 O-P(O)-OCH 2 CH 2 - protons. The signal at 3.03 ppm for the proton of the end hydroxyl group H O-CH 2 CH 2 - was also observed. The same characteristic signals occur in the 1 H NMR spectrum of product 2BPClTEA ( Figure S2 ). The 13 C NMR data of BPClTEA ( Figure S3 ) indicate characteristic signals at 59.86 ppm (HO C H 2 CH 2 -), 66.95 ppm (HOCH 2 C H 2 -), and aromatic carbon atoms of BHET and PPD in the range of 125.01–132.52 ppm. Resonances at 164.84 ppm and 164.25 ppm for the carbonyl group were also observed. New signals appeared at 62.98 ppm, a doublet with a coupling constant 3 J(P, C) = 6 Hz, typical for -CH 2 C H 2 O-P(O)- carbon atom, and at 66.05 ppm for - C H 2 CH 2 O-P(O)-. Based on the NMR ( 1 H, 13 C, 31 P{H}) data, we assume that the product of the interaction between BHET and PPD at a molar ratio 1:1 has a structure that coincides with the one presented in Scheme 1 ( Scheme 1 , product I). The above results give us reason to assume that if the signal at 20.20 ppm (integral intensity 0.96) in the 31 P NMR spectrum is for the phosphorus atom in the repeating units, and that at 20.69 ppm (integral intensity 0.03) is for the phosphorus atom in the end unit, then the molecular mass (M n ) of polyphenylphosphonate is 12,032 g/mol (n = 32, molecular weight of the repeating unit 376). 2.1.2. Interaction between GP-PET and PPD at a Molar Ratio of 1:1 A total of 10.000 g of PG-PET, including 48.78% BHET; dimer, 20.89%; trimer, 10.96%; and EG, 19.37%, reacts with 11.075 g (0.0568 mol) of PPD. EG, 0.0312 mol, and BHET, 0.0192 mol, have the highest molar concentrations, while the molar concentrations of the dimer and trimer are less by one order of magnitude. This gives us reason to assume that the following two main reactions occur simultaneously ( Scheme 2 ). The interaction between III and IV leads to the formation of the copolymer VI . The total molar concentration of EG (0.0312 mol) and BHET (0.0192 mol) is 88.7% from the molar concentration of GP-PET (0.0568 mol). Products III and IV , which are based on EG and BHET, are approximately 90% of the weight of the reaction product, i.e., these are the primary products of interaction. The 1 H NMR spectrum of the reaction product ( Figure S4 ) showed a signal at 10.97 ppm, which is characteristic of P-OH protons. The signal at 8.02 ppm is characteristic of the aromatic protons of PET. The aromatic protons of PPD are in the range of 7.33–7.94 ppm. The additional signals at 4.63 ppm are attributed to the methylene protons in the segments -C(O)O-CH 2 -CH 2 -OCO-, which are due to the presence of dimers and trimers in GP-PET. The signals at 3.88 ppm and 4.43 ppm are assigned to the methylene protons adjacent to the hydroxyl group in the BHET unit (-CH 2 -OH) and C(O)O-CH 2 . The multiplets at 4.38–4.19 ppm refer to the methylene protons of P(O)O-C H 2 CH 2 and P(O)O-CH 2 C H 2 . At 3.00 ppm, there is a signal for the proton of the H OCH 2 -CH 2 structure. In the 31 P{H}NMR spectrum ( Figure 2 ) of the reaction product, there are signals at 19.73, 20.28, and 21.10 ppm with integral intensities of 5.11, 1.43, and 1.00, respectively. Three types of phosphorus atoms must exist, namely, the phosphorus atom in the repeating unit of product III , the repeating unit of product IV , and the terminal unit. Based on the literature data [ 29 , 30 ], the signals at 19.73 ppm and 20.28 ppm can be assigned to the phosphorus atom in the repeating units, while the one at 21.10 ppm can be assigned to a phosphorus atom in the terminal unit. The quantitative composition of the degradation product suggests that the signal at 19.73 ppm can be attributed to the phosphorus atom in the repeating units of product III and that at 20.28 ppm can be attributed to the phosphorus atom in the repeating unit of product IV . The number average molecular mass of the phosphorylated product, calculated based on the data from the 31 P{H}NMR spectrum, is 1447 g/mol. In the 13 C NMR spectrum of the reaction product ( Figure S5 ), there are signals at 67.97 ppm for HO-CH 2 C H 2 O- and 60.90 ppm for HO- C H 2 CH 2 O-. There are signals at 63.00 ppm for -C(O)O-CH 2 - C H 2 -OCO-, which are due to the presence of dimers and trimers in the product; at 133.99–128.40 ppm for the aromatic carbon atoms of BHET and PPD residues; and at 165.55 ppm and 166.03 ppm for C=O carbon atoms. New signals appear at 66.90 ppm for -P(O)O-CH 2 C H 2 -O(O)C- and a doublet at 63.57 ppm with 2 J(P,C) = 5.7 Hz, characteristic of the -CH 2 C H 2 O(O)P- carbon atom. The NMR data for reaction products of the dimer and trimer with PPD, V ( Scheme 3 ), will be the same as those for products III and IV ( Scheme 4 ) because the substituents attached to the phosphorus atom are the same. The data from NMR spectroscopy confirm the proposed structures. 2.2. Interaction between GP-PET and TMP 2.2.1. Transesterification of TMP with Commercial BHET The model reaction of BHET and TMP at a molar ratio of 1:2 was carried out at 190 °C for 5 and 9 h (details about the experimental procedure of the model reaction are reported in the Supplementary Materials ). In the 31 P{H} NMR spectrum of the reaction product obtained after 5 h of heating, BTMP5 ( Figure S6 ), there are signals (δ, ppm/integral intensity) at 2.63 (1.00), 2.30 (2.48), 1.48 (0.48), 1.17 (0.75), and 0.05 (0.30), which are characteristic of phosphate structures. The main signals are at 2.63 ppm and 2.30 ppm in a ratio of 1:2.48 (28.70%, 71.30%). In the 31 P NMR spectrum of the product ( Figure S7 ), the signals represent multiplets of nine lines ( 3 J (P, H) = 11.38 Hz), which gives information about the phosphorus surrounding atoms. In the 31 P{H} NMR spectrum of the reaction product obtained after 9 h of heating, BTMP9 ( Figure S8 ), the signals are at 2.70 ppm and 2.29 ppm in a ratio of 1.00:2.32 (30.10%, 69.90%). From the 31 P NMR analysis of the same product ( Figure S9 ), it is clear that the signals are multiplets of nine lines with a coupling constant 3 J (P, H) = 11.74 Hz. The data from the 31 P{H} NMR analysis show that the additional increase in the reaction time does not lead to significant changes in the content of the reaction products. The intensity of the signal at 2.70 ppm increases from 28.70% up to 30.10%. The presence of two signals gives us reason to assume that in the reaction mixture, there are two phosphorus-containing compounds with different amounts but with a very similar structure of the substituents at the phosphorus atom. In the 1 H NMR spectrum of BTMP5 ( Figure S10 ), the signal at 3.34 ppm should be attributed to the proton of the OH group ( H OCH 2 -CH 2 - structure). The doublets at 3.66 ppm and 3.69 ppm with 3 J(P,H) = 12 Hz are characteristic of POC H 3 protons and display integral intensities of 1.42 and 3.50, respectively, in a ratio of 1.00:2.47, which is the same as the ratio of the integral intensities of the signals for the phosphorus atoms, i.e., 1.00:2.48. The signals in the region 4.29–4.42 ppm should be attributed to -C(O)OCH 2 - and P(O)OCH 2 - protons. The signals at 8.01 ppm and 8.03 ppm are assigned to the aromatic protons. In the 1 H NMR spectrum, there is a new signal at 3.86 ppm, a singlet, which is characteristic of the methyl protons of the ester group CH 3 OC(O)-Ar-. The reason for assigning this signal to these protons is the fact that in the starting compounds, BHET and TMP, there are no protons whose signals are singlets in this region. Additionally, the signal for the methyl protons of dimethyl terephthalate is at 3.94 ppm [ 32 ]. The 1 H NMR spectrum of BTMP9 ( Figure S11 ) contains the same signals as the 5 h heating product. The ratio of the integral intensities of the signals for POCH 3 protons is almost the same, from 1:2.48 to 1:2.45. In the 13 C NMR spectrum of BTMP9 ( Figure S12 ), there are signals at 52.42 ppm, a singlet; 54.12 ppm, d, 2 J(P,C) = 6.0 Hz, and 54.46 ppm, d, 3 J(P,C) = 6.0 Hz, which should be attributed to P-O C H 3 carbon atoms; 59.04 ppm for HO C H 2 carbon atoms; 63.81 ppm, d, 2 J(P,C) = 5.7 Hz, which should be assigned to P(O)O C H 2 carbon atoms, and 70.41 ppm for C (O)OCH 2 - carbon atoms; and at 133.70 ppm and 129.65 ppm for aromatic carbon atoms. There are also two signals for the carbonyl carbon atom at 165.43 and 166.16 ppm ( C =O carbon atoms). The 13 C NMR spectrum shows the presence of two types of P-OCH 3 carbon atoms, which is in agreement with the 1 H NMR and 31 P NMR spectroscopy results. The singlet at 52.42 ppm can be attributed to the carbon atom of the methyl ester group ( C H 3 OC(O)-) since the signal for this carbon atom of dimethyl terephthalate is at 52.39 ppm [ 32 ]. A phosphorus atom whose signal in the 31 P NMR spectrum is a multiplet of nine lines can be obtained as a result of a transesterification reaction of TMP and BHET, and it is also as a result of an exchange reaction between the ester group of BHET and the methoxy group of TMP. The signals for the phosphorus atoms of the products of transesterification (compounds I and II , Scheme 5 ) should be at the same shift in the spectrum because their surroundings are the same. The signal for the phosphorus atom of dimethyl(2-hydroxyethyl) phosphate III ( Scheme 5 ) should not coincide with that of compounds I and II , since there is a difference in the substituents—in dimethyl(2-hydroxyethyl) phosphate III , the substituent is OCH 2 CH 2 OH, while in I and II , it is OCH 2 CH 2 OC(O)-Ar-C(O)OCH 2 CH 2 OH. A methyl ester group -CH 3 OC(O)- is formed as a result of the exchange reaction. It is known that the alkoxy groups of H-phosphonic and the phosphoric acids participate in exchange reactions with amide [ 33 ], urethane [ 34 ], and carbonate [ 35 ] groups. We assume that the reaction between BHET and TMP proceeds according to the following reaction scheme ( Scheme 5 ). Under these reaction conditions, two reactions take place simultaneously as follows: transesterification between TMP and BHET and an exchange reaction between TMP and BHET. In the first stage of transesterification, product I is formed, which, in the second stage, is converted into product II . Since the exchange reaction proceeds at a lower rate compared with the transesterification reaction [ 33 , 34 , 35 ], it can be assumed that the signal at 2.29 ppm ( Figure S8 ) should be related to the phosphorus atom in product II and that at 2.70 ppm should be related to the phosphorus atom in dimethyl(2-hydroxyethyl) phosphate III . Its content based on 31 P{H} NMR is 30.10%. The content of methyl (2-hydroxyethyl) terephthalate IV is the same. Transesterification of TMP with dimethyl(2-hydroxyethyl) phosphate III and methyl (2-hydroxyethyl) terephthalate IV leads to the formation of compounds V and VI ( Scheme 5 ). The NMR data suggest that the main products of the reaction of BHET (commercial product) with TMP are II , with a content of 70%, and V and VI with a content of 30%. 2.2.2. Interaction between GP-PET and TMP at a Molar Ratio of 1:2 It was found that the content of GP-PET is BHET, 48.78%; dimer, 20.89%; trimer, 10.96%; and EG, 19.37% [ 28 ]. A reaction between GP-PET and TMP was carried out at a temperature of 190 °C for 3 h at a molar ratio of 1:2. The reaction product was characterized by 1 H, 31 P, and 13 C NMR techniques. The data from the 1 H and 13 C NMR analysis ( Figure 3 and Figure 4 ) are similar to those of the reaction product from the interaction between commercial BHET and TMP. The presence of signals at 3.88 ppm in the 1 H NMR spectrum and 52.42 ppm in the 13 C NMR spectrum confirms the assumption that an exchange reaction also takes place in the reaction mixture of GP-PET /TMP. In the 31 P{H} NMR spectrum of GP-PET/TMP ( Figure 5 ), the main signals (δ, ppm/integral intensity) are at 2.48 (1.00) and 2.16 (0.59). In the 31 P NMR spectrum of GP-PET/TMP ( Figure 6 ), the signals represent multiplets of nine lines with a coupling constant 3 J(P,H) = 11.34 Hz. The content of 10.000 g GP-PET is BHET (0.0192 mol), dimer (0.0047 mol), trimer (0.0017 mol), and EG (0.0312 mol). The molar concentration of EG is the highest—1.6 times higher than that of BHET, 6.6 times higher than that of the dimer, and 18 times higher than that of the trimer. The GPC analysis ( Figure S13 ) shows that the reaction mixture contains products with a molecular weight (Mw) of 221, 297, 433, 628, and 835. Based on the ratio of the molar concentrations of the components of the glycolysis product and TMP, and the data from the GPC analysis, we propose the following reaction scheme for the interaction between GP-PET and TMP ( Scheme 6 ). According to the proposed reaction scheme, two reactions occur simultaneously in this interaction as follows: a transesterification reaction, whose products are I , II , V , and VI , and an exchange reaction, whose products are III and IV . The GPC analysis indicates that GP-PET contained products with molecular weights (Mw) of 221, 297, 433, 628, and 835. The molecular mass of 221 should be attributed to product IV (Mw = 224), which confirms the progress of the exchange reaction. The molecular weight of 297 should be assigned to product I (Mw = 278), and the molecular weight of 433 should be assigned to product II (Mw = 470). The molecular weights of 628 and 835 should be attributed to products VIa (x = 2) (Mw = 662) and VIb (x = 3) (Mw = 856). Table 1 summarizes the data for the molecular weights (Mn and Mw) and polydispersity index (PDI) of the products from the interaction between GP-PET and TMP. 2.3. Evaluation of the Thermal Characteristics Molecules containing phosphorus atoms in their architectures (inorganic phosphate and organophosphorus compounds) are used as flame retardants. These substances support inhibitory layer formation on the surface of a polymeric matrix during combustion and reduce the contact area between the polymer and oxidizing agents. During the combustion reaction, a carbon structure is formed, which is a solid layer. Decomposition of the phosphorus-containing mixtures generates radicals of PO·, P 2 , and P that have the ability to capture the radicals of H·, O·, and HO·. These facts result in the formation of high quantities of CR after thermal degradation at high temperatures even in an inert atmosphere [ 36 ]. It is well known that CR is a very important characteristic for determining their abilities as flame retardant additives since CR can promote an intumescent effect in the polymeric matrix and form a physical limitation to oxygen therein. In summary, the CR mass produced after thermal decomposition of phenylphosphonate compounds is directly related to the amounts of phosphorus elements in the polymer chain [ 37 ]. From the TGA curve of product BPClTEA ( Figure S14 ), it can be seen that the decomposition of the sample proceeds in the following three steps: during the first stage of degradation from about 90 to 200 °C, the weight lost is about 9%; in the second stage from 200 to 390 °C, the decomposition rate is higher—31%; and in the third stage from 390 to 500 °C, the weight lost is about 43%. At 800 °C, the carbonized residue is about 17%. From the TGA analyses of BTMP9 ( Figure S15 ), it is evident that the degradation of the material takes place in three stages as follows: at 350 °C, which refers to a weight loss of around 57%; in the second (at 450 °C) and third (at 700 °C) stages, the losses of the material are around 28%. The CR at 800 °C is nearly 15%. From the TGA data of the reaction product from the interaction between GP-PET and TMP ( Figure 7 ), it is obvious that the decomposition of the sample proceeds in three stages as follows: the first stage is at 360 °C, which refers to the weight loss of 62%; in the second (at 500 °C) and third (at 700 °C) stages, the losses are around 25%. The remaining residue at 800 °C is about 13%. From the TGA curve of the reaction product of the interaction between GP-PET and phenylphosphonic dichloride (GP-PET/PPD) ( Figure 8 ), it is established that the thermal degradation of the product takes place in three stages. In the first stage, insignificant losses (~3.5%) are observed at a temperature of about 200 °C. In the next two phases from 250 to 400 °C and from 400 to 600 °C, the losses of material (nearly 80%) indicate the thermal decomposition of the phosphor-containing product. The thermogram also shows that after heating to 800 °C, the amount of CR is approximately 17%. The quantity of the remaining residue of the phosphorylated products is similar to others flame retardants applied as additives to polymers. The CR values of the products synthesized in the present study are comparable to data reported for other phosphorylated materials [ 38 , 39 ] and show the possibility for their potential use as flame retardants."
} | 6,745 |
38167850 | PMC10762202 | pmc | 4,974 | {
"abstract": "In terrestrial ecosystems, plant leaves provide the largest biological habitat for highly diverse microbial communities, known as the phyllosphere microbiota. However, the underlying mechanisms of host-driven assembly of these ubiquitous communities remain largely elusive. Here, we conduct a large-scale and in-depth assessment of the rice phyllosphere microbiome aimed at identifying specific host-microbe links. A genome-wide association study reveals a strong association between the plant genotype and members of four bacterial orders, Pseudomonadales, Burkholderiales, Enterobacterales and Xanthomonadales. Some of the associations are specific to a distinct host genomic locus, pathway or even gene. The compound 4-hydroxycinnamic acid (4-HCA) is identified as the main driver for enrichment of bacteria belonging to Pseudomonadales. 4-HCA can be synthesized by the host plant’s OsPAL02 from the phenylpropanoid biosynthesis pathway. A knockout mutant of OsPAL02 results in reduced Pseudomonadales abundance, dysbiosis of the phyllosphere microbiota and consequently higher susceptibility of rice plants to disease. Our study provides a direct link between a specific plant metabolite and rice phyllosphere homeostasis opening possibilities for new breeding strategies.",
"introduction": "Introduction In nature, plants and their associated microbes, collectively known as the microbiota, form functional entities that rely on each other 1 . The microbiota contributes to aspects such as disease resistance 2 – 5 , stress tolerance 6 , 7 , and nutrient acquisition 8 , 9 . Successful recruitment and maintenance of a sufficient abundance of specific microbial members determines the outcome of plant-microbiota interactions 10 , 11 . Thus, understanding the principles driving microbiome assembly in crop plants has become one of the main pursuits of present studies in order to integrate microbiome functioning into sustainable crop production. The plant microbiota can be compartmentalized into the rhizosphere (root-soil interface), phyllosphere (leaf surface) and endosphere (internal tissues) 12 . Studies revealed that host genetics 13 and metabolic processes 14 participate in the recruitment of specific microbial taxa in the rhizosphere microbiome, and consequentially shape microbiome assembly. In barley, a specific genomic locus was shown to be associated with the recruitment of specific bacterial taxa 15 . Deepening analysis revealed three genes in this locus, including a Nucleotide-Binding-Leucine-Rich-Repeat (NLR) receptor as primary candidates involved in shaping the microbiome 15 . In tomato plants, the identification of specific quantitative trait loci (QTLs) provided further evidence for the importance of host genetics in microbiome assembly 16 . In addition, specific plant root exudates were demonstrated to steer rhizosphere microbiome assembly toward the host’s needs for defense and nutrition 17 . The plant phyllosphere represents the largest biological surface on earth and widespread interface for interactions between plants and their microbiota 18 . Identification of specific genetic components shaping the phyllosphere microbiome is still sparsely understood, although the plant genotype has been repeatedly demonstrated to impact the phyllosphere microbiome in consistent ways across geographically separated sites 19 , 20 . Model plants like Arabidopsis 21 and non-model plants like perennial wild mustard 22 as well as cereal crops such as wheat and barley 23 were implemented to increase our understanding of the phyllosphere microbiome. It is known that external factors such as microbial inocula and climate conditions can alter phyllosphere microbiomes 24 , however, the plant genotype has become the central focus in terms of seeking approaches to engineer plant microbiome for agricultural purposes 25 . Recent studies have revealed that defense responses and cell wall integrity are likely involved in phyllosphere microbiome assembly 19 . Arabidopsis cuticle mutants and ethylene signaling mutants were shown to have a different phyllosphere microbiome compared to wild-type plants 20 . The phosphate starvation response pathway 26 and pattern-triggered immunity 27 were also shown to take part in shaping the Arabidopsis phyllosphere microbiome. A common strategy to investigate host genetic effects on microbiome assembly is based on amplicon sequencing of microbial marker genes coupled with genome-wide association studies (GWAS) 28 . Such studies have resulted in the identification of many putatively causal genes with effects on the relative abundance of plant-associated microbes 19 , 21 , 22 . In the case of the Arabidopsis phyllosphere microbiome, specific microbial hubs were successfully linked to host genomic loci that are related to specialized metabolite biosynthesis 21 . Genes involved in the synthesis of sinapoyl glucose, sinapoyl malate, glucosinolates are assumed to participate in shaping the Arabidopsis phyllosphere microbiome. Nevertheless, there is still lack of direct evidence bridging the host genetic background with the recruitment of specific microbial members. So far, phenotyping phyllosphere microbiome compositions via GWAS was exclusively conducted on the basis of amplicon sequencing of microbial marker genes 17 . However, metagenomic data provides better means to more accurately conduct such analyses, mainly because it is less prone to over- or underestimate the abundance of certain microbial taxa. In this work, with the employment of phyllosphere metagenomes from 110 rice accessions of the Rice Diversity Panel II core collection (C-RDP-II) 29 , we perform GWAS experiments to link bacterial abundances with single nucleotide polymorphisms (SNPs) in the rice genome. Rice genetic variations are shown to be significantly associated with members of four predominant phyllosphere bacterial orders, Pseudomonadales, Burkholderiales, Enterobacterales, and Xanthomonadales. To unravel a prevailing mechanism of how host genetics affect phyllosphere microbiome assembly, we implement mutants and over-expression constructs of a candidate gene associated with Pseudomonadales and assess the resulting microbiome shifts. Furthermore, we analyze rice metabolites in rice leaves that are regulated by the candidate gene. We discover that the compound 4-hydroxycinnamic acid, also known as p-coumaric acid and a precursor in lignin biosynthesis, is required for the assembly and homeostasis of the rice phyllosphere microbiome. Overall, our study provides clear evidence for host metabolite-driven phyllosphere microbiome assembly.",
"discussion": "Discussion Plants are colonized by highly diverse, tissue-specific microbial communities, the plant microbiota. The plant microbiota is known to complement host functioning and to be an important factor for enhancing resilience under abiotic as well as biotic stress. Plants can assemble and maintain certain microbiota structures, which is vital for plant health. A dynamically balanced state of the microbiota is known as ‘homeostasis’ 32 . Disrupting homeostasis often results in a prevalence of detrimental microbiota members, leading to shifted microbiota structures and plant disease occurrence, also known as dysbiosis 33 . It is currently sparsely understood how plants assemble and maintain a functional (homeostatic) microbiota, especially in regards to their phyllosphere 33 . Previous research has mostly focused on microbe-microbe interactions that result in homeostasis and can explain certain colonization patterns 34 – 36 . Here, we show that specific host-microbe interplay is involved in shaping phyllosphere bacteria prevalence in rice plants. Research conducted in the past has provided indications for active host regulation of the phyllosphere microbiota. The plant immune system was demonstrated to be required for maintaining microbiota homeostasis 37 – 40 . Mutations of immunity-related genes in Arabidopsis have been shown to result in a dysbiotic phyllosphere microbiota 27 , 41 . In the present study, we show that rice plants deploy a potentially conserved mechanism to shape the phyllosphere microbiome in a wide range of genotypes. Except for immunity-related mechanisms, plants can also employ metabolites as a selective force to assemble candidate microbiota members to serve their interests 17 . Various plant-specific metabolites can fulfill the purpose of maintaining microbiota homeostasis on leaves 42 , or suppress bacterial pathogens 43 . Nevertheless, fundamental questions remain unaddressed related to which plant genetic components are devoted to control microbiota assembly, and how this takes place on metabolite level. Genome-wide association studies (GWAS) have proven to be a highly useful tool to shed further light on plant-microbe interactions 28 , 44 . In a recent study, researchers showed that the abundance of phyllosphere microbial hubs in Arabidopsis correlated with multiple specialized metabolic pathways affecting both bacterial and fungal communities 19 , 21 . It should be highlighted the previous studies were solely based on amplicon sequencing of microbial marker genes in the phyllosphere. While such approaches are valuable, this method is also known to be prone to specific biases, because marker genes of certain microbial groups are more likely to be amplified than those of others 17 . Therefore, microbiome structures are not always accurately captured. In the present study, a large-scale phyllosphere metagenome dataset was combined with GWAS and facilitated more precise insights into associations between rice plants and their phyllosphere microbes. For one, it provided further evidence for the role of the plant immune system in shaping its microbiota, such as the MAPK signaling pathway. More importantly, it allowed us to pinpoint specific metabolic pathways and genes which we identified as suitable targets to dissect phyllosphere microbiome assembly at molecular level. Furthermore, the high resolution of taxonomic data obtained here was advantageous to connect host genetic traits to specific bacterial taxa, which was critical to validate the effects of a distinct gene and metabolite on microbiome assembly. The present study revealed that primarily bacteria belonging to the order Pseudomonadales are enriched by 4-HCA in the rice phyllosphere. A similar observation was obtained with poplar tree endophytes and down-regulation of cinnamoyl-CoA reductase; it caused the enrichment of 4-HCA in plant tissues and consequently increased the abundance of Pseudomonas 45 . This genus is a ubiquitous member of the plant microbiota 46 , 47 . Previous studies have shown that it is an important component of a functional microbiota that can shield off pathogens and prevent diseases in plants. Pseudomonas were often identified as core microorganisms associated with both the plant rhizosphere and phyllosphere 35 , 36 , 48 , 49 , and as key players in many disease-suppressive soils 50 , 51 . Plants benefit from Pseudomonas as they can produce a wide range of antimicrobial compounds 52 , induce plant systemic disease resistance 53 , and establish an array of chemical dialogues with plants 37 , 54 , 55 . Our identification of plant genetic components controlling Pseudomonas could provide an ideal target for engineering the phyllosphere microbiota of various crop plants. We also showed that Xanthomonas , a highly prevalent phytopathogen, is negatively affected by 4-HCA. It is therefore very likely that plants protect themselves against pathogens by relying on their immune system as well as specific metabolites that not only allow them to directly antagonize pathogens but also to maintain homeostasis in the phyllosphere microbiota. In recent years, plant microbiome studies have facilitated the identification of various disease-reducing or even disease-preventing microorganisms 56 . Leaves are common entry points of various, highly devastating pathogens. The present study shows that metabolite-driven regulation of the phyllosphere microbiota is a hidden mechanism to support plant hosts against pathogen attacks. Previous studies that assessed the rhizosphere microbiome have resulted in similar findings. For example, iron-mobilizing coumarins in Arabidopsis roots were linked to shaping the rhizosphere microbiota by inhibiting proliferation of certain Pseudomonas species via a redox-mediated mechanism 54 . Re-shaping of the rhizosphere microbiome by the coumarin scopoletin was shown to have consequences for plant health 57 . Similarly, the benzoxazinoid breakdown product 6-methoxy-benzoxazolin-2-one (MBOA) was found to regulate rhizosphere microbiota assembly of cereal crops 58 . The present findings related to plant metabolite-driven phyllosphere homeostasis and their implications for host health will serve as a basis for targeted breeding strategies. Previous studies have provided certain indications, mostly via correlation analyses, that plant metabolites play an important role in maintaining a functional microbiota 42 . However, it remained unresolved if the identified compounds are the main drivers of the observed structural changes of the microbial communities. In the present study, knock-out as well as over-expression constructs allowed us to specifically attribute microbiome changes to 4-HCA in the rice phyllosphere. A lack of 4-HCA was linked to dysbiosis of the phyllosphere microbiome which results in higher susceptibility of rice plants to disease. Current resistance breeding strategies known as ‘ R gene breeding strategies’ are increasingly challenged by rapidly evolving phytopathogens. This problem could be addressed by identifying and engineering microbiome-shaping genes, which may become known as ‘ M gene breeding’."
} | 3,450 |
34493930 | PMC7611625 | pmc | 4,980 | {
"abstract": "Highlights • Research on tree diversity and antagonists often neglects plant complementarity. • We studied species richness/mycorrhizal type effects on leaf herbivory/pathogens. • Mycorrhizal type had different effects on herbivory and pathogen infestation. • Elemental not metabolite concentrations determined leaf damage.ld.",
"conclusion": "Conclusions We identified drivers of herbivory and pathogen infestation across tree monocultures and mixtures with different mycorrhizal associations. Including mycorrhizal type revealed a comparably minor role of tree species richness and functional richness in driving antagonist leaf damage in saplings, with tree species identity and AM dominance explaining most of the variation in herbivory and pathogen infestation rates. However, it has to be taken into consideration that the effects of mycorrhizal type are an interplay of the identity of the symbiotic partner and specific associated plant characteristics, mostly plant economics traits, that may co-determine and even contradict or countermand each other. Our study gives a first mechanistic insight into how those functionally distinct associations affect damage by herbivores and pathogens. However, our study also showed that species identity is a determinant of leaf damage by antagonists in tree saplings that have co-evolved with the multitude of plant strategies. Those effects cannot be attributed to a single process based on common ecological theories, but instead depend on a complex interplay of mechanisms. They involve host- as well as antagonist-associated processes that relate to the multifaceted characteristics of plant diversity. Furthermore, our study points to the importance of other mechanisms that are mediated via elemental concentrations, other than shifts in the measured defence-related metabolites. We speculate that shifts in physical defence and traits related to AM per se play a more crucial role in this regard than expected before. Furthermore, volatile organic compounds, that were not assessed in this study, may be more important in this context than the set of secondary metabolites measured. However, our study also showed relationships between the measured variables to be opposed to the ones commonly found. This may trigger a multitude of potential follow-up studies addressing such relationships in the light of different contexts. Our study further reveals how distinct the drivers of herbivory vs. pathogen infestation are in tree saplings with different mycorrhizal associations.",
"introduction": "Introduction The positive effects of plant diversity on the functioning of ecosystems are substantially driven by the complementary use of resources. However, the underlying mechanisms of biodiversity-ecosystem functioning (BEF) relationships are still elusive ( Cardinale et al., 2012 ; Eisenhauer et al., 2016 ; Grossman et al., 2018 ). Indeed, most of the studies focus on plants and plant-related ecosystem functions, where empirical evidence for resource-use complementarity is inconsistent ( Barry et al., 2019 ). This calls for the consideration of the trophic complexity of ecosystems. It includes interactions of plants with higher trophic levels, such as insect herbivores and pathogens as major determinants of plant fitness ( Hines et al., 2015 ; Weisser & Siemann, 2008 ). In forest BEF experiments, there have been few studies and inconsistent findings ( Grossman et al., 2018 ; Schuldt et al., 2017 ), despite the significance of damage caused by insect herbivores (e.g., Franceschi, Krokene, Christiansen & Krekling, 2005 ) and pathogens (e.g., Graniti, 1998 ). Higher tree diversity can affect the fitness of an individual tree species, e.g., through a decrease in leaf herbivory and pathogen infestation rates ( Al-Alouni, Brandl, Auge & Schädler, 2014 ; Hantsch et al., 2014 ; Jactel & Brockerhoff, 2007 ). However, numerous studies have also found neutral and even positive effects of biodiversity on herbivory and pathogen infestation rates ( Schuldt et al., 2010 , 2017 ; Vehviläinen, Koricheva & Ruohomäki, 2007 ). The net effect of tree diversity on herbivory can be regarded as the result of opposing, mutually non-exclusive mechanisms. Negative effects can be the result of an increased diversity and efficiency of natural enemies of plant antagonists in more diverse tree stands ( Andow, 1991 ; Jouveau et al., 2020 ) as well as a consequence of resource dilution especially for specialised antagonists (resource concentration hypothesis; Barbosa et al., 2009 ; Castagneyrol, Giffard, Péré & Jactel, 2013 ; Root, 1973 ). Positive effects of plant diversity on insect herbivore performance may be driven by positive effects on generalists, in particular through diet mixing for example, ( Unsicker, Oswald, Köhler & Weisser, 2008 ) or herbivore spill-over from preferred to other plants (e.g., Jactel & Brockerhoff 2007 ; White & Whitham 2000 ). In the same line, some pathogens depend on the presence of multiple plant species to complete their life cycle ( Nguyen et al., 2016 ). Whilst these hypotheses focus on antagonist-associated processes, there is an increasing recognition of host-associated processes as underlying mechanisms of plant diversity effects on plant antagonists. The complementary use of resources in more diverse plant communities ( Loreau & Hector 2001 ; Oelmann et al., 2007 ), may lead to higher leaf quality and, consequently, higher rates of herbivory. In contrast, the enhanced resource supply may also be invested in defence strategies and reduce the palatability of plant material for herbivores ( Kostenko, Mulder, Courbois & Bezemer, 2017 ; Mraja, Unsicker, Reichelt, Gershenzon & Roscher, 2011 ). The elemental composition of plants determines the plant biomass consumed by herbivores ( Cebrian & Lartigue 2004 ; Sterner, Elser, Fee, Guildford & Chrzanowski, 1997 ). A better nutrient supply of plants, e.g., within a diverse plant community, can lead to an elevated fitness of the antagonists. At the same time, it may strengthen the community's ability to defend itself against the antagonists. As these mechanisms act antagonistically, predictions on the effects of plant diversity on leaf herbivory and pathogen infestation remain difficult. Furthermore, allocation of elements to plant defence may depend on the probability of antagonist attack and the general benefit of defence ( Cipollini, Walters & Voelckel, 2018 ; Stamp, 2003 ). To ultimately unravel the underlying relationship between plant diversity, leaf herbivory, and pathogen infestation rates, it is thus crucial to concomitantly study plant elemental concentrations and the main compounds involved in plant defence against herbivores and pathogens. Mycorrhizal fungi play critical roles in the competitive capabilities of plants ( Smith & Read 2010 ; van der Heijden, Martin, Selosse & Sanders, 2015 ). The two most common mycorrhizal types are arbuscular mycorrhiza (AM) and ectomycorrhiza (EM). They have distinct foraging strategies and life styles as well as different mechanisms for resource exchange ( Bonfante & Genre 2010 ; Soudzilovskaia et al., 2019 ). AM fungi completely depend on their host as the sole carbon supplier and, in turn, provide the plant host with soil phosphorus that is often limiting to the plant ( Brundrett, 2009 ; Smith & Read 2010 ). EM fungi can be obligate or also have a saprotrophic phase, taking up organic and mineral plant resources from various substrates for nutrient exchange with the host. Due to the large hyphae system, EM fungi scavenge more effectively and at further distances from the host roots compared with AM fungi. The mutual interaction between mycorrhizal fungi and plants may lead to an increased nutritional value of leaves and synthesis of defence-related compounds ( Fernández et al., 2014 ; Kaling et al., 2018 ). Moreover, it has been found that, in the initial phase of mycorrhizal fungal colonisation, plants recognise fungi as invaders triggering similar responses like pathogens ( Ferlian et al., 2018a ). In this way, mycorrhizal fungi are able to enhance plant immunity by increasing the levels of defence-related metabolites and foster defence-priming ( Gange & West 1994 ; Kempel, Schmidt, Brandl & Schädler, 2010 ; Martinez-Medina et al., 2016 ). The functional diversity of plants in terms of mycorrhizal association may potentially increase resource partitioning amongst plant species ( Klironomos, McCune, Hart & Neville, 2000 ; Wagg, Barendregt, Jansa & van der Heijden, 2015 ); and mycorrhizal fungi have been proposed to play a critical role in positive BEF relationships ( Ferlian et al., 2018a , 2018b ). Thus, the effects of plant diversity on antagonists may be co-determined by the effects of mycorrhizal diversity via changes in resource acquisition, plant defences, and the nutritive value of plant tissue. We investigated the effects of tree species identity and richness as well as mycorrhizal type on leaf herbivory and pathogen infestation rates. We used a tree sapling diversity experiment that manipulates the two most common mycorrhizal types (AM and EM; via suitable tree species selection) and tree species richness (monocultures and two-species mixtures). We measured the concentrations of defence-related plant metabolites (a set of sugars, amino acids, and phenolics) to shed light on the underlying mechanisms between plant diversity and leaf damage. Furthermore, we measured leaf elemental (carbon [C], nitrogen [N], and phosphorus [P]) concentrations reflecting general plant nutrient uptake and leaf palatability. (1) We hypothesised that herbivory and pathogen infestation rates are lower in the tree species mixtures compared to monocultures. Similarly, herbivory and pathogen infestation rates are lower in tree communities being associated with both mycorrhizal types compared to communities with only one dominant mycorrhizal type. This is attributable to the resource concentration hypothesis and the higher defence ability due to partitioning and better exploitation of resources. (2) We further hypothesised that concentrations of defence-related plant metabolites explain a higher proportion of the effects on herbivory and pathogen infestation rates than element concentrations. (3) Due to the different life strategies and specialisation of leaf herbivores and pathogens, the mechanisms behind the effects of plant diversity differ between leaf herbivory and pathogen infestation rates.",
"discussion": "Discussion Overall, our study showed that tree species identity was most important in driving leaf damage and chemical characteristics in tree saplings, especially in AM communities. Leaf elemental concentrations (C and N) and the presence of AM were significant predictors of leaf damage, whereas the concentrations of the assessed metabolites were not. Moreover, herbivory and pathogen infestation were explained by distinct relationships. These results indicate that aboveground multitrophic interactions depend on belowground associations of tree saplings with mycorrhiza and their related traits as well as their consequences for nutrient uptake. Effects of plant diversity on leaf damage In contrast to the strong effects of tree species identity, the effects of tree species richness on total herbivory and pathogen infestation were not significant, but showed trends towards decreasing rates from monocultures to mixtures. Various underlying mechanisms have been proposed to explain this pattern, such as an increase of natural enemies of antagonists ( Andow, 1991 ) and resource dilution ( Castagneyrol et al., 2013 ; Hambäck, Inouye, Andersson & Underwood, 2014 ). Further, an enhanced resource-use complementarity and, thus, higher nutrient supply, that fosters synthesis of plant defence-related compounds, ( Kostenko et al., 2017 ) may be responsible for the pattern. The lack of significant effects in our study might be explained by the short study duration. It may not have allowed for shifts in the abundance of natural enemies of antagonists, or shifts in resource uptake strategies among the tree species. Further, many parasitoids and predators are highly mobile, thereby making such effects less likely at the spatial scale of the experiment. Leaf damage differed between tree communities of different mycorrhizal types. Specifically, the SEM showed that AM trees, i.e., in AM communities as well as in communities with both mycorrhizal types, increased total herbivory rates. This suggests that tree species associated with AM fungi share certain traits that facilitate attack by antagonists ( Koele, Dickie, Oleksyn, Richardson & Reich, 2012 ; Koricheva, Gange & Jones, 2009 ). As indicated by the colonisation rates of AM and EM, in the AM trees, other traits than the degree of AM colonisation may have dominated the effects. Furthermore, we found that pathogen infestation rates were not directly affected by any of the mycorrhizal types and did not differ among tree species in EM communities. Thus pathogen-related traits seem to be more uniform across species than in communities that are associated with AM or both mycorrhizal types. In contrast to our expectation, tree communities with both mycorrhizal types experienced an intermediate rate of total herbivory and pathogen infestation that was in between that of AM and EM communities. We hypothesised that the effect of functionally diverse tree communities, e.g., communities that associate with different mycorrhizal types or occupy different ends of the leaf economics spectrum, adds to the complementarity in resource uptake of a species-diverse community ( Barry et al., 2019 ; Eisenhauer, 2012 ; Wright et al., 2004 ). Such communities should therefore be better defended against antagonists ( Eisenhauer et al., 2019 ; Meyer et al., 2016 ) by benefiting from the situation that interspecific competition is lower than intraspecific competition ( Loreau & Hector 2001 ). However, our study indicates a pure additive effect, where the rate of leaf damage reflects the mixture of mycorrhizal types in the community. Moreover, with increasing mycorrhizal diversity, competitive relationships diverged between the two mycorrhizal types in the community, as did leaf damage. For example, leaf damage increased in AM trees and decreased in EM trees in communities with both mycorrhizal types (data not shown) as compared to communities with only one. Indeed, the SEM underpins the different roles and strategies of AM and EM trees in terms of resource supply. Drivers of leaf damage C concentration was positively correlated with N and P concentration, which was unexpected as they are typically negatively or not correlated ( Díaz et al., 2016 ; Bruelheide et al., 2018 ). Reasons for this may be related to the particularly high soil nutrient concentrations at the site ( Ferlian et al., 2018b ). Surprisingly, leaf elemental concentrations were not correlated with the profiles of the assessed leaf metabolites, as indicated by the SEM. Nutrients are commonly allocated to growth, defence, maintenance, reproduction, and storage within a plant ( Chapin, Schulze & Mooney, 1990 ; Feng et al., 2009 ). However, investment in plant growth and defence are often suggested to be subject to a trade-off ( McKey, 1974 , 1979 ). Theories suggest that the resource quality of a habitat determines whether a plant invests predominantly into defence (lower antagonist infestation) or growth (lower competition; Cipollini et al., 2018 ; Eichenberg, Purschke, Ristok, Wessjohann & Bruelheide, 2015 ; Holopainen, Rikala, Kainulainen & Oksanen, 1995 ). Recent studies, however, indicate that plant growth and defence are not necessarily alternative strategies ( Kempel, Schädler, Chrobock, Fischer & van Kleunen, 2011 ). In resource-rich habitats, such as our study site ( Ferlian et al., 2018b ), trees may have allocated nutrients preferably to other functions than chemical defence, such as growth. Therefore, plant nutrient status was mostly decoupled from defence-related metabolite concentrations. A further potential explanation could be that nutrients were allocated to secondary metabolites that were not measured in this study, such as volatile organic compounds. Variation within the metabolite groups and profiles was mainly driven by tree species identity, whereas tree species richness and mycorrhizal type contributed similarly and only marginally after accounting for tree species. Metabolite profiles in monocultures generally only tended to differ from that of mixtures. Presence and concentrations of metabolites are highly species-specific representing defence strategies that species have evolved ( Levin, 1976 ). In contrast, the high resource availability at the site may have decoupled metabolite concentrations from sources of nutrient input, such as mycorrhizal types. The different metabolite groups had mostly negative relationships with herbivory and pathogen infestation rates but effects were overall small suggesting a minor role of the secondary metabolites measured for leaf damage. The effects were even weaker for EM communities. Accordingly, in the SEM, metabolite profiles were not correlated with leaf damage. As mentioned above, due to the resource-rich habitat of the experimental site, herbivory and pathogen infestation rates may have been driven by other plant characteristics than by the assessed chemical defence compounds, especially in EM communities ( Cornelissen, Aerts, Cerabolini, Werger & van der Heijden, 2001 ). Especially for leaf damage by pathogens, such characteristics may be related to nutrient status of the plant. Surprisingly, the dominance of AM decreased leaf N status. This was unexpected as AM trees are typically characterised by a lower C-to-N ratio compared to EM trees ( Lin, McCormack, Ma & Guo, 2017 ; Plett & Martin 2011 ; van der Heijden et al., 2015 ). Again, this may be attributed to the unique nutrient dynamics at the site and the young age of the trees presumably leading to weak or even detrimental interactions between the tree host and its symbiotic partner ( Johnson, Graham & Smith, 1997 ). Leaf N concentrations, further, negatively affected pathogen infestation rates. This points to higher attraction of pathogens by AM trees due to additional relevant mechanisms driving pathogen infestation that involve N input but neither increased plant defence (at least not via the metabolites assessed in this study) nor leaf quality ( Rabin & Pacovsky 1985 ). In contrast, the dominance of AM trees positively affected herbivory rates in a direct way suggesting that the effects on herbivory are mediated by other mechanisms than nutrient inputs all well. Such mechanisms may be related to enhanced water uptake, certain morphological traits, or antagonist-associated drivers. Path coefficients within the SEM suggest these positive direct effects to be much stronger than the positive effects through leaf N status (0.30 vs. 0.08). Overall, the effects of AM dominance on the two antagonist groups were positive, whereas EM dominance did not determine them. This suggests that tree species associated with different mycorrhizal types may have evolved different functional characteristics that, to a different extent, influence insect herbivores and pathogens ( Connell & Lowman 1989 ; Gehring, Cobb & Whitham, 1997 ). Mechanisms in herbivory vs. pathogen infestation Our investigated variables and relationships explained a similar part of the variation in herbivory and pathogen infestation rates (15% and 16%, respectively). However, they were affected by different drivers (AM dominance and element concentrations, respectively), suggesting that underlying mechanisms of tree leaf damage by antagonists differ between insect herbivores and pathogens. The SEM revealed that specific leaf elements determine pathogen infestation, whereas AM dominance and, thus, AM tree-associated characteristics determine herbivory (see above). Pathogen infestation was decreased by leaf N and C. The diverse relationships between leaf elemental concentrations and damage include a range of mechanisms, such as effects of chemical defences (that were not part of our set of assessed metabolites), and physical defences. Insect herbivores and pathogens substantially differ in characteristics like dietary needs, mobility, specialisation, enemy taxa, specificity to defence compounds, and susceptibility to physical barriers ( Barrett & Heil 2012 ; Schuldt et al., 2017 ; Thaler, Agrawal & Halitschke, 2010 ). Thus, it is not surprising that they react differently to shifts in elemental concentrations. Phenolic acid concentrations were negatively correlated with total herbivory and herbivore mining rates and positively correlated with rust infestation rates in AM communities, suggesting that here again opposing processes may have contributed to the relationships between phenolic acids and each total damage rate. However, most studies report negative relationships pointing to the important role of phenolic acids in constitutive and induced defence against herbivore and pathogens ( Summers & Felton 1994 ). Our study further showed no correlation between herbivory and pathogen infestation rates, which contrasts with previous studies ( Schuldt et al., 2017 ; Stout, Thaler & Thomma, 2006 ). A multitude of studies has reported antagonistic as well as facilitative relationships between the two ( Biere & Bennett 2013 ; Fernandez‐Conradi, Jactel, Robin, Tack & Castagneyrol, 2018 ; Schuldt et al., 2017 ). The strength of relationships between the two in trees also depends on the context, such as plant diversity and the dominating feeding guilds amongst the antagonist groups, with the latter being a potential explanation in our study ( Schuldt et al., 2017 ; Stout et al., 2006 ). Moreover, the signalling pathways of anti-herbivore and anti-pathogen defences can interact antagonistically leading to opposing damage patterns in both groups ( Pieterse, Schaller, Mauch-Mani & Conrath, 2006 )."
} | 5,508 |
30404178 | PMC6266758 | pmc | 4,981 | {
"abstract": "Microbes influence a wide range of host social behaviors and vice versa. So far, however, the mechanisms underpinning these complex interactions remain poorly understood. In social animals, where individuals share microbes and interact around foods, the gut microbiota may have considerable consequences on host social interactions by acting upon the nutritional behavior of individual animals. Here we illustrate how conceptual advances in nutritional ecology can help the study of these processes and allow the formulation of new empirically testable predictions. First, we review key evidence showing that gut microbes influence the nutrition of individual animals, through modifications of their nutritional state and feeding decisions. Next, we describe how these microbial influences and their social consequences can be studied by modelling populations of hosts and their gut microbiota into a single conceptual framework derived from nutritional geometry. Our approach raises new perspectives for the study of holobiont nutrition and will facilitate theoretical and experimental research on the role of the gut microbiota in the mechanisms and evolution of social behavior.",
"conclusion": "4. Conclusions Nutritional geometry is a powerful approach in the study of nutritional interactions across organizational levels from molecules to individuals, groups and communities [ 28 , 73 , 74 ]. As we have described above, concepts of nutritional geometry can help in modelling how gut microbiota composition, host feeding behaviors and host social interactions integrate within a single theoretical framework. By identifying nutritional interactions and feedback loops between these three different components, this approach allows the formulation of new hypotheses to explore the role of the gut microbiota in the mechanisms and evolution of host social behaviors ( Figure 4 ). While nutritional studies have typically considered animal food intake and performance responses as a proxy of the intake target of an individual, growing evidence shows that these preferences vary among individuals, presumably because animal feeding decisions are influenced by several additional factors, including (but not only) its microbiota and social interactions [ 25 ]. Host nutritional choices can modify the relative abundance of microbe types in their gut by varying nutritional input available for microbial development [ 22 ] or through direct ingestion [ 18 ]. In turn, the gut microbiota composition can affect host nutrition by expanding [ 15 ], digesting and assimilating key nutrients for the host [ 17 ], or by dragging the host towards suboptimal diets [ 19 ]. Microbiota can intervene in gene-expression involved in social propensity [ 75 ], guiding mating choices [ 8 , 9 ], promoting aggregations [ 6 , 7 ], and mediating kin discrimination [ 39 ]. Conversely, social interactions favor the homogenization of the gut microbiota by increasing microbial transmission across individuals in the population [ 13 , 56 ]. Social interactions can also indirectly affect microbiota composition via the impact of collective behaviors during foraging [ 76 ] or differential access to resources [ 77 ]. The interplay between gut microbiota, host nutrition, and social interactions ( Figure 4 ) emphasizes the need to study host-microbiota interactions using a more integrative approach, by considering populations of hosts and their feeding decisions. Using nutritional geometry, we derived speculative scenarios about how gut microbes may affect host social interactions and how host social interactions might constrain feeding decisions, and as a result, microbiota composition. The challenge for future studies will be to carry out experiments on specific components in order to examine the mechanisms underlying co-evolutionary feedback acting between hosts and their microbiota [ 55 ]. Deconstructing the nutritional needs and performance responses of the different components of host-microbe associations will help tackle important ongoing discussions about the relevance of the concept of the holobiont, whilst looking for potential evolutionary conflicts between microbes, host nutrition and social behaviors. Computational models of nutritional geometry have already begun to explore nutritional interactions in populations of agents [ 73 , 78 ] and integrate multiple levels of interactions (e.g., host-pathogen interactions [ 37 ]). From an experimental point of view, it is possible to design manipulative experiments on host and microbiota nutrition, using behavioral assays or physiological measures on organisms provided with artificial diets [ 19 ]. Although it remains difficult to measure the intake target of microbe types in their natural host environment (where several microbes interact within assemblages), manipulating the microbiota composition under controlled conditions, using axenic [ 79 , 80 ] and gnotobiotic [ 19 ] animals, can help estimate the intake target of the animal host and how it is influenced by the microbiota. In the field, it is possible to reconstruct the gut microbiota composition of an animal and its nutrient intake by monitoring the animals’ food choices, and analyzing food samples and feces composition (see [ 81 , 82 ] for details). At the collective level, social network analyses constitute powerful tools to assess how gut microbiota influence social interactions and structures [ 83 , 84 ]. By quantifying social interactions and characterizing the nutritional needs of experimentally manipulated individuals, it is possible to picture the associations between microbiota-induced phenotypic changes (e.g., variation in holobiont nutritional needs) and the temporal dynamics of social interactions. This experimental approach might also be applied in species reared in their natural environments. For instance, using animal models that naturally show flexibility in their social behavior, either at different developmental stages or in different environments (i.e., social tipping points; see Reference [ 85 ] for details), it is possible to evaluate the variation of the onset of changes induced by the microbiota. Additionally, experimental manipulation of host nutrition and control of their social interactions, for instance using model organisms [ 19 , 86 , 87 ], can also help clarify the role of diet and social behavior in defining the effects of gut microbiota in hosts [ 44 , 60 , 88 ]. Since the basic principles of nutrition are shared by virtually all animal species [ 25 ], results from this approach may bring insights into the influence of the microbiota in the mechanisms and evolution of social life across a wide range of organisms. Microbiota composition can be a key component shaping animal behaviors [ 5 ]. Thus, the interplay between microbiota and the host is central to understand multiple host interactions. The behavioral plasticity observed in the host can be seen as a combination of the interactions between environmental factors, and the genomic repertoire of the microbiota (i.e., microbiome) and the one of its host [ 89 ]. It is thus the holobiont itself that is plastic and able to cope with environmental and biological stressors (e.g., parasite infections [ 61 , 90 ]), possibly driving selection of both hosts and microbiota communities. Nutritional choice is one possible evolutionary channel of host diversity [ 91 ], shaping gut microbiota specialization [ 92 ], and driving the emergence of diverse social organizations across multiple taxa [ 65 ]. We hope that quantitative analyses of these interactions, as facilitated with nutritional geometry, will encourage future research in this direction.",
"introduction": "1. Introduction Interactions between hosts and their microbiota, that together form the “holobiont” [ 1 , 2 , 3 ], influence various aspects of an animal’s biology, including nutrition [ 4 ] and behavior [ 5 ]. Growing evidence indicates that the microbiota can also have important consequences on the ways hosts interact with each other, for instance by triggering aggregations [ 6 , 7 ], guiding mate choice [ 8 , 9 ] or allowing kin discrimination [ 10 ]. How these complex host-microbe interactions are mediated is an open question. Recent studies point towards an effect of microbes on brain areas involved in the control of social interactions [ 11 ], or in the production of chemical signals mediating communication [ 12 , 13 ]. Here we argue that microbes that colonize the gut of animals may also influence a wide range of social behaviors through their impact on the nutritional needs and foraging decisions of animals. Nutrition is central to all host-gut microbiota interactions [ 14 ]. On the one hand, gut microbes can influence the nutritional strategies of their hosts by expanding the host’s capability to digest and assimilate key nutrients [ 15 , 16 ], or by supplementing the host with nutrients that are difficult to find in the environment [ 17 ]. Termites, for instance, depend on complex microbial communities to exploit wood and soil nutrients that are otherwise not digestible by insects [ 18 ]. Microbes can also modify the feeding preferences of their hosts, for instance by attracting animals towards a diet that is suboptimal for the hosts but beneficial for the microbes [ 19 ]. On the other hand, nutrient intake by the hosts shapes the composition of their gut microbiota [ 18 , 20 , 21 ]. Variations in protein and carbohydrate intake modify the relative abundance of microbes in an animal’s gut [ 22 ], a change that can be observed within a few hours in mice [ 23 ], or a few days in humans [ 4 ]. Over recent years, state space models of nutritional geometry [ 24 , 25 ] ( Figure 1 ) have been increasingly used to study the nutritional interactions between gut microbes and their hosts, an approach known as “nutritional immunity” [ 14 ]. In these models, animals and foods are represented in a nutrient space defined by two or more key food components for the animal (typically, but not exclusively, the macronutrients carbohydrates, proteins and fats; for a recent review see [ 26 ]). Foods are radials defined by the ratio of the food components under consideration (nutritional rails). The animal is characterized by a nutritional state (NS), a snapshot of its nutritional condition, and by an intake target (IT), an optimal state at which fitness is maximized. The challenge for the animal is to consume available foods in amounts and balances allowing it to reach and maintain its IT [ 25 ]. Knowing the nutritional composition of foods, and the NS and the IT of the animal, it is thus possible to predict the feeding strategy that will most efficiently enable the animal to meet its nutritional needs (see examples in Figure 1 A). Nutritional immunology studies have extended this framework by considering microbes as trophic analogues of animals [ 27 ]. In this approach, an animal’s gut microbiota can be modelled in a nutrient space and characterized by a specific NS and IT [ 26 ]. Here, however, the NS of the gut microbiota varies with the nutritional behavior of its host. Using nutritional geometry designs, the performance responses of the host and the gut microbiota can be derived by experimentally mapping the expression of fitness traits of both holobiont components in separate nutrient spaces (see examples in Figure 1 B). Therefore, in principle, it is possible to predict how the nutrient intake of a host will affect its fitness and the development of different microbe types (strains or species) in its gut [ 26 ]. Concepts of nutritional geometry have also been used to explore the effects of nutrition on social interactions, an approach called “social nutrition” [ 28 , 29 , 30 ]. Here, nutrition models are used to study how nutrition affects many social behaviors and structures, such as collective movements [ 31 ], cooperative foraging [ 32 ], or reproductive division of labor [ 33 ]. In social species, individuals must often trade-off between choosing foods that address their own nutritional needs and following others’ choices to maintain social cohesion, which can generate a variety of social responses [ 28 ]. These interacting effects between nutrition and social behaviors can be modelled by considering all individuals constituting a social group in a common nutrient space (Reference [ 28 ]; Figure 1 C). Each individual is defined by its own NS and attempts to reach its IT. The ability of the different group members to do so depends on the nature and the frequency of their social interactions (e.g., attraction [ 34 ], competition [ 35 , 36 ], cooperation [ 37 ]). In social insects, for example, nutrient balance is achieved collectively by the foragers that collect nutrients to reach a colony IT composed of the different ITs of all colony members, including the other workers that need carbohydrates for energy, and the larvae that need proteins for growth [ 38 ]. In this review, we argue that concepts of nutritional geometry can bring new fundamental insights into emerging research on microbiota and social behaviors, by integrating nutritional interactions at different levels (between hosts and their gut microbes, and among hosts) in a common theoretical framework. We hypothesize that, by acting on the nutritional decisions of their hosts, gut microbes can influence a wide range of social behaviors, which in turn impact the composition of the gut microbiota communities. We develop this hypothesis, by first reviewing research demonstrating the links between the gut microbiota and the feeding behaviors of their hosts. We then describe how these multi-level interactions can be modelled in nutritional geometry to generate new empirical predictions about the influence of the gut microbiota on social behaviour. External microbes living on the surface of the animals may also influence a range of social behaviors [ 10 , 39 , 40 , 41 , 42 , 43 ]. However, it is unlikely that nutritional requirements of these microbes influence host nutritional decisions; at least before they are ingested by individuals, for instance through self-cleaning behaviors [ 44 ]. For these reasons, here we only focus on microbes that colonize the gut."
} | 3,553 |
29464208 | PMC5815863 | pmc | 4,982 | {
"abstract": "Widely distributed outer-membrane cytochromes enable sulfur-reducing bacteria to obtain energy from solid electron donors.",
"introduction": "INTRODUCTION Microbial reduction of oxidized sulfur species (OSS), such as SO 4 2− , SO 3 2− , S 2 O 3 2− , S 0 , and S n 2− , coupled with the oxidation of inorganic substrates is one of the main mechanisms for energy production in organic-poor anaerobic marine subsurfaces ( 1 – 3 ). Although hydrogen (H 2 ) formed via water radiolysis or mineral-water reactions is considered to be the key energy source in organic-depleted environments that extend from shallow sediments to the deeper Earth crust of nearly half of the ocean ( 4 – 6 ), it remains unclear whether H 2 is sufficient to sustain the subsurface population of OSS-respiring bacteria ( 7 , 8 ). In addition, microbial energy requirements for the oligotrophic bacteria found in subsurface environments have remained a challenge ( 9 ). Given the ubiquity and abundance of OSS-respiring bacteria in subsurface environments, determining whether and how they use abundant solid substrates such as reduced sulfide minerals (MnS, FeS, and FeS 2 ) as energy sources in addition to scarce H 2 or organic substrates will enhance our understanding of ecosystems isolated from solar radiation, particularly considering the marked imbalance between H 2 (nanomolar to micromolar) and sulfate (a few millimolar) in shallow sediments. A model of direct electron uptake by marine sulfate-reducing bacteria (SRB) from elemental iron was first proposed by Dinh et al . ( 10 ), who isolated SRB strains from marine sediments supplied with iron granules as the sole electron donor. In addition, recent studies implied that direct electron uptake by SRB via conductive filaments is critical for anaerobic oxidation of methane in a consortium composed of methanotrophic archaea and uncultivable SRB in marine sediments ( 11 – 13 ). Because this extracellular electron transport potentially allows the use of solid minerals as electron sources, microbial OSS-reduction reactions coupled with electron uptake from minerals may occur in sediments limited for H 2 and organics. However, because the biological signatures and energetics of the proposed electron uptake process have not been determined, the distribution and importance of this electron uptake mechanism in energy-limited marine sediments remain unclear. In the present study, we therefore examined the electron uptake mechanism in the SRB Desulfovibrio ferrophilus IS5, using bioinformatic, biochemical, electrochemical, and microscopy methods.",
"discussion": "RESULTS AND DISCUSSION The electron uptake process in SRB has been speculated to involve integral outer-membrane (OM) heme proteins or nanofilaments ( 14 , 15 ), as occurs during extracellular electron transfer by iron-reducing bacteria ( 16 – 18 ). We sequenced the entire 3.7-Mbp (million base pair) circular genome of D. ferrophilus IS5 (table S1) and identified 26 genes encoding multi-heme cytochromes containing at least four heme-binding motifs (table S2), which are essential to form a direct electron transport pathway, that is, cytochrome complexes across the OM [OM cytochromes (OMCs)], and to reduce extracellular solids in model iron-reducing bacteria, such as Shewanella oneidensis and Geobacter sulfurreducens . Although genes encoding conductive nanofilaments in G. sulfurreducens were not found in the genome of strain IS5 (table S3), the multi-heme cytochromes encoded by the genes DFE_450 and DFE_464 were predicted to be extracellular and may therefore be critical components of the extracellular electron uptake in D. ferrophilus IS5 (table S4). A search of the National Center for Biotechnology Information (NCBI) nonredundant protein database revealed that OMCs homologous to those encoded by DFE_450 and DFE_464 were widely distributed among cultured and uncultured sedimentary bacteria of the phyla Proteobacteria, Thermodesulfobacteria, and Aquificales, which reduce SO 4 2− , SO 3 2− , S 2 O 3 2− , S 0 , and S n 2− ( Fig. 1A ). Notably, the homologs of the D. ferrophilus IS5 OMCs displayed clear sequence divergence from OMCs of S. oneidensis , G. sulfurreducens , and Acidithiobacillus ferrooxidans ( Fig. 1B ). The presence of these conserved OMCs among OSS-respiring bacteria suggests that this poorly characterized clade of microbes can extract electrons from reduced sulfide minerals via OMCs under soluble-electron-donor-limited conditions. Fig. 1 Distribution of homologous extracellular cytochromes of D. ferrophilus IS5 in OSS-respiring bacteria. ( A ) Phylogenetic distance tree derived from 16 S ribosomal RNA (rRNA) sequence alignments. ( B ) Protein phylogenetic tree derived from amino acid sequences of OMCs, including those encoded by the genes DFE_450 and DFE_464 of D. ferrophilus IS5 and their homologs (identified using NCBI blastp, max score > 200); OmcB, OmcE, OmcS, and OmcZ of G. sulfurreducens PCA; OmcA, and MtrC of S. oneidensis MR-1; and Cyc2 of A. ferrooxidans ATCC 23270. b, bacterium . Scale bars represent estimated sequence divergence or amino acid changes. Consistent with this speculation, we confirmed that OMCs were highly expressed in D. ferrophilus IS5 under soluble-electron-donor-limited conditions, by both OM extraction and transcriptome analyses of cells cultured in a medium containing limited lactate. Lactate was completely consumed after 5 days of inoculation (table S5), and crude membrane fractions were clearly separated into the inner membrane (IM) and OM fractions as confirmed by electrophoresis ( Fig. 2A and fig. S1). The ultraviolet-visible (UV-vis) absorption spectra of the extracted OM fractions showed characteristic Soret and Q band absorptions of oxidized and reduced c-type cytochromes ( Fig. 2B ). Analysis of the spectral data demonstrated that the lactate-starved IS5 cells had at least 10 times more OMCs than the cells cultured under lactate-sufficient conditions ( Fig. 2B , inset, and fig. S2), in which more than 30 mM lactate was left after the 5-day cultivation. To elucidate the genes coding for the D. ferrophilus IS5 OMCs, the electrophoresed OM fraction of lactate-starved cells was treated with the heme stain (3,3′,5,5′-tetramethylbenzidine–H 2 O 2 ) to visualize OMCs, and the amino acid sequences of proteins in the heme-positive bands were then determined. Using this approach, the products of genes DFE_449 and DFE_461 , which encode cytochromes with 12 and 14 heme-binding sites, respectively, were detected ( Fig. 2A and table S4). Although no other cytochromes were detected in the heme-positive bands, transcriptome analysis revealed the overexpression of other cytochromes including the two extracellular cytochromes encoded by DFE_450 and DFE_464 under lactate-starved conditions ( Fig. 2C ), suggesting that starvation for electron donor promotes the expression of the OMCs in IS5 cells for electron uptake. Fig. 2 Identification of OMCs of D. ferrophilus IS5. ( A ) Protein profiles of IM and OM fractions isolated from lactate-starved cells and stained with Coomassie brilliant blue (CBB) and heme-reactive 3,3ʹ,5,5ʹ-tetramethylbenzidine-H 2 O 2 , respectively. In heme-positive bands (indicated by # and *), transcripts of DFE_461 and DFE_449 were detected by liquid chromatography–mass spectrometry (LC-MS). A minor band corresponding to ferroxidase ( DFE_1154 , 14.1 kDa, indicated by Δ) was detected due to nonspecific heme staining. ( B ) Absorption (abs.) spectra of OM fractions extracted from lactate-starved (Lact-starv.) (red) and lactate-sufficient (Lact-suffi.) (blue) D. ferrophilus IS5 cells normalized by the OM protein concentration. Solid and dotted lines represent spectra measured under reduced and oxidized conditions, respectively. Inset: Comparison of Soret peak absorption at 419 nm (see analysis details in fig. S2). ( C ) Expression of seven multi-heme cytochromes and two β-propeller proteins (hatched bars), which are located in close proximity in the D. ferrophilus IS5 genome, by lactate-starved and lactate-sufficient cells. We further confirmed by electrochemical measurements that lactate-starved D. ferrophilus IS5 cells were able to extract electrons from an electrode surface with markedly higher efficiency compared to lactate-sufficient cells. Using three-electrode anaerobic reactors equipped with indium tin–doped oxide (ITO) electrodes poised at −0.4 V [versus standard hydrogen electrode (SHE)] to prevent the evolution of H 2 ( 19 ), the electrical current generated by extracellular electron uptake coupled with sulfate reduction was monitored. Upon the introduction of lactate-starved D. ferrophilus IS5 cells into the reactor, in which the electrode served as the sole electron donor, the cathodic current density immediately increased to >0.2 μA cm −2 within 1 hour ( Fig. 3A ). In contrast, upon the introduction of H 2 -consuming SRB Desulfobacterium vacuolatum , or sterile medium into the reactor, negligible current increase was detected. The addition of lactate-sufficient D. ferrophilus IS5 cells did not cause immediate cathodic current increase but caused only a small current increase and an alteration in background non-faradaic current, demonstrating that lactate concentration in the cell culture changed the electron uptake capability of D. ferrophilus IS5 cells. Although the cathodic current production of lactate-starved IS5 cells continued, the number of IS5 cells on the ITO electrode did not significantly increase after 3 days, whereas the doubling time of IS5 cells was approximately 13.5 hours under anaerobic growth conditions with lactate (fig. S3), suggesting that the electrode provides a much smaller amount of energy for IS5 cells than lactate. The potential dependency of the sulfate-reduction current measured by linear sweep (LS) voltammetry after cathodic current measurements demonstrated that the sulfate reduction was initiated from −0.3 V and maximized at a potential of −0.42 V in lactate-starved IS5 cells ( Fig. 3B ). In contrast, no significant current for the sulfate reduction was observed in the LS voltammograms of lactate-sufficient IS5, H 2 -consuming D. vacuolatum , or sterile medium. The observed onset potential for sulfate reduction at −0.3 V was consistent with the redox profile of lactate-starved D. ferrophilus IS5 determined from a differential pulse voltammogram (fig. S4). Notably, another peak observed in the LS voltammogram of lactate-starved cells at −0.1 V may be assignable to the reduction of biosynthesized iron sulfide, which also explains the slight current decrease after the peak current generation in lactate-starved IS5 cells after t = 2 hours ( Fig. 3A ), because the formation of iron sulfide may inhibit the direct contact between the cells and the electrode surface. The threshold electrode potential for sulfate reduction of −0.3 V is similar to the redox potential of NADH but is more positive than that of ferredoxin, suggesting that NADH may serve as the primary electron transport component that receives electrons from periplasmic cytochromes to cytoplasmic adenosine phosphosulfate reductase ( Fig. 3C ). These results further suggest that the identified electron uptake mechanism requires almost the minimum amount of energy in the form of electrons to drive the sulfate-reduction process. Because the generation of most key agents (for example, menaquinone) that donate electrons to other OSS (SO 3 2− , S 2 O 3 2− , S 0 , and S n 2− ) also occurs at a redox potential equal to or more positive than that of NADH, the new clade of OMCs identified here may be an important component of the microbial energy acquisition processes of OSS-respiring bacteria, particularly in electron-donor-limited environments, including some marine sediments. In addition, given that the growth of IS5 coupled with the electron uptake was negligible on the electrode (fig. S3), the requirement for the minimum but sufficient energy to drive metabolism may be related to (and provide an explanation for) the extremely slow growth rates observed in subsurface oligotrophic environments ( 9 ). Fig. 3 In vivo electrochemical measurements with intact D. ferrophilus IS5 cells. ( A ) Cathodic current versus time measurements conducted in anaerobic reactors equipped with ITO electrodes poised at −0.4 V (versus SHE) as the sole electron donor. Reactors into which lactate-starved or lactate-sufficient D. ferrophilus IS5 cells, H 2 -consuming D. vacuolatum , and sterile medium were introduced are depicted in red, blue, pink, and black lines, respectively. Arrows indicate the timing of addition of cells or sterile medium into the reactor. ( B ) LS voltammograms measured after current generation in (A). ( C ) Energy diagram of the extracellular electron uptake model of strain IS5, in which the redox potential of OMCs capable of generating the reduced form of nicotinamide adenine dinucleotide (NADH) and/or reduced menaquinone (MQH 2 ) drives the reductions of OSS and intermediates, such as adenosine phosphosulfate (APS). Cyts represents periplasmic c-type cytochromes. ATP, adenosine 5ʹ-triphosphate. Scanning electron microscopic observation of the electrode surface after electrochemical measurements revealed that lactate-starved IS5 cells formed a single-layer biofilm on the electrode ( Fig. 4A ), which is similar to many bacterial strains capable of extracellular electron transfer and consistent with the notion that electron uptake occurs via OMCs. Furthermore, nanofilamentous structures were observed between cells and extended from cells to the electrode surface ( Fig. 4B and fig. S5). Similar nanowires produced by S. oneidensis MR-1 have been shown to be extensions of the OM ( 20 , 21 ), implying that cytochromes detected in the OM fraction of strain IS5 are also localized to the observed nanowires. Fig. 4 Microscopy on lactate-starved D. ferrophilus IS5 cells and their nanowires. ( A and B ) Scanning electron microscopy images of cells attached to the electrode surface. Scale bars,10 μm (A) and 500 nm (B). ( C to E ) Transmission electron microscopy images of cells stained with cytochrome-reactive DAB-H 2 O 2 . (C) Negative DAB staining in the absence of H 2 O 2 . (D and E) Positive DAB staining with the addition of H 2 O 2 . Scale bars, 500 nm (C and D) and 50 nm (E). Note that bamboo-like nanowire structures were clearly visible only with positive DAB staining. ( F and G ) Fluorescence microscopy images of cells stained with protein-specific NanoOrange (green) and membrane-specific FM 4-64FX (red). Scale bars, 5 μm. Arrowheads indicate nanowires. The formation of nanowire-like structures was induced in D. ferrophilus IS5 cells in response to lactate limitation, as was observed for OMCs. To examine the distribution of OMCs on the nanowires, transmission electron microscopy was performed on thin sections of cells subjected to cytochrome-reactive 3,3′-diaminobenzidine (DAB)–H 2 O 2 staining, in which heme iron centers catalyze the formation of a DAB polymer with high binding affinity to OsO 4 . As positive staining controls, S. oneidensis and G. sulfurreducens cells grown under conditions that promote nanowire formation were used ( 21 , 22 ). Consistent with the expression analysis of OMCs, the outer edges of lactate-starved IS5 cells ( Fig. 4 , C to E) were stained with nearly twofold greater intensity than lactate-sufficient cells (figs. S6 and S7). Similar staining intensities were observed for S. oneidensis (fig. S8A) and G. sulfurreducens (fig. S8B) cells (see analysis details in fig. S9 and Supplementary Methods), whereas Escherichia coli cells (fig. S8C), which lack OMCs and were used as a negative staining control, exhibited more intense staining within the cell interior compared to the membrane region. Nanowires of D. ferrophilus IS5 resembled bamboo-like structures with diameters ranging from 30 to 50 nm, which were clearly visible only upon positive DAB staining, suggesting high coverage of cytochromes on the surface of nanowires. Although IS5 nanowires were slightly thinner than those of S. oneidensis , both showed identical segmented structures and strong cytochrome positive staining, suggesting that IS5 nanowires also have a role in electron transfer as proposed in S. oneidensis . In contrast, G. sulfurreducens nanowires had diameters of approximately 7 nm and displayed both weakly positive or negative DAB staining. These differences are consistent with previous structural models that propose that the nanowires of S. oneidensis are extensions of the OM ( 20 ), whereas those of G. sulfurreducens are predominantly composed of the type IV pilin PilA ( 22 ). To examine the composition of the nanowires of D. ferrophilus IS5 in more detail, lactate-starved cells were stained with the protein- and membrane-specific fluorescent dyes NanoOrange and FM 4-64FX, respectively ( Fig. 4 , F and G, and fig. S10). Fluoresence microscopy of the nanowires of strain IS5 that were stained for both protein and membrane revealed that the nanowires are the extensions of the cell membrane, which is identical to the nanowires of the S. oneidensis ( 20 ). Combined with the bamboo-like structures observed by transmission electron microscopy, nanowires of strain IS5 and S . oneidensis may be the ordered alignments of OM vesicles of a few tens of nanometers in size. In accordance with this idea, we observed the secretion of membrane vesicles and their alignments in lactate-starved IS5 cells by combinatorial heavy metal staining, which emphasizes the contrast of membranes in transmission electron microscopy (fig. S11). Therefore, the extracted OM fraction of lactate-starved IS5 cells also likely includes nanowires. These data further support the speculation that OMCs and nanowires of D. ferrophilus IS5 are able to mediate electron uptake from extracellular solids (see the schematic image in fig. S12). As verified in some iron-reducing bacterial strains, the identified multi-heme cytochromes may form an integral protein complex across the OM in D. ferrophilus IS5. Because the cytochromes encoded by DFE_449 and DFE_461 detected in the extracted OM fraction were predicted to be at the periplasm; and these genes are located in close genomic proximity to DFE_450 and DFE_464 , which are estimated to encode the extracellular cytochromes (table S4); DFE_449 and DFE_461 may encode OM-bound periplasmic cytochromes that perform a similar functional role to the S. oneidensis MtrA, a periplasmic decaheme cytochrome in the OM MtrCAB cytochrome complex that is important for metal reduction in Shewanella species ( 23 , 24 ). Transcriptome analysis also revealed comparable expression levels for the genes DFE_448 to DFE_451 and DFE_461 to DFE_465 ( Fig. 2C ), encoding proteins including three other cytochromes and two NHL-repeat β-propeller proteins, which may function as protein-protein interaction sites and stabilize the OMC complex, as occurs with β-barrel proteins in iron-reducing bacteria ( 25 ). Notably, homologous β-propeller proteins were also identified in a number of sedimentary OSS-respiring bacteria (table S6), suggesting that the potential integral OMC complex may be broadly conserved among this microbial group. Similar electron uptake has also been reported with Fe 2+ -oxidizing A. ferrooxidans , which have a distinct OMC (Cyc2) from IS5 ( Fig. 1B ) ( 26 , 27 ); however, the bacterial oxidation of sulfide minerals under anaerobic conditions was proposed to be an indirect oxidation of Fe 2+ ions released from iron sulfides ( 28 , 29 ), which is different from the direct electron uptake model proposed in this study. Because genes encoding OMCs homologous to those of D. ferrophilus IS5 were identified in a variety of uncultivable SRB strains that form consortia with methanotrophic archaea ( Fig. 1 ), which perform the anaerobic oxidation of methane, the observed electron uptake capability of D. ferrophilus IS5 OMCs supports the recently proposed model that SRB can directly receive electrons from methanotrophic archaea cells via OMCs or nanowires ( 11 – 13 ). According to the present electrochemical data, the SRB partner in the methane-oxidizing consortia requires electrons with potentials that are more negative than −0.3 V, which is also consistent with the previous finding that anthraquinone-2,7-disulfonate (redox potential, −0.185 V) can be reduced by electrons from methanotrophic archaea ( 12 ). The findings related to the identified clade of OMCs and the energetics of the electron transfer processes mediated by these heme proteins are expected to accelerate our understanding of not only microbial energy production for sedimentary OSS-respiring bacteria in soluble-energy-source-limited conditions but also interspecies electron transfer in syntrophic consortia containing SRB."
} | 5,246 |
30013839 | PMC6046197 | pmc | 4,983 | {
"abstract": "Black Band Disease (BBD) is a widely distributed and destructive coral disease that has been studied on a global scale, but baseline data on coral diseases is missing from many areas of the Arabian Seas. Here we report on the broad distribution and prevalence of BBD in the Red Sea in addition to documenting a bleaching-associated outbreak of BBD with subsequent microbial community characterization of BBD microbial mats at this reef site in the southern central Red Sea. Coral colonies with BBD were found at roughly a third of our 22 survey sites with an overall prevalence of 0.04%. Nine coral genera were infected including Astreopora , Coelastrea , Dipsastraea , Gardineroseris, Goniopora , Montipora , Pavona , Platygyra , and Psammocora. For a southern central Red Sea outbreak site, overall prevalence was 40 times higher than baseline (1.7%). Differential susceptibility to BBD was apparent among coral genera with Dipsastraea (prevalence 6.1%), having more diseased colonies than was expected based on its abundance within transects. Analysis of the microbial community associated with the BBD mat showed that it is dominated by a consortium of cyanobacteria and heterotrophic bacteria. We detected the three main indicators for BBD (filamentous cyanobacteria, sulfate-reducing bacteria (SRB), and sulfide-oxidizing bacteria (SOB)), with high similarity to BBD-associated microbes found worldwide. More specifically, the microbial consortium of BBD-diseased coral colonies in the Red Sea consisted of Oscillatoria sp. (cyanobacteria), Desulfovibrio sp. (SRB), and Arcobacter sp. (SOB). Given the similarity of associated bacteria worldwide, our data suggest that BBD represents a global coral disease with predictable etiology. Furthermore, we provide a baseline assessment of BBD disease prevalence in the Red Sea, a still understudied region.",
"conclusion": "Conclusions Our study represents the first comprehensive assessment of Black Band Disease in the central Red Sea. Elucidation of the bacteria associated with BBD microbial mats of corals at a southern reef site confirms that BBD represents a disease with predictable etiology where the three main bacterial players are globally distributed with regional differences. Notably, our reef survey data, in line with data from other regions, identify BBD as a widespread disease, but as one with low prevalence in comparison to other coral diseases. Additional surveys including other coral diseases as well as pathogen infection experiments with Red Sea corals could further increase our understanding of coral stress tolerance in this understudied coral reef region. Importantly, the prevalence of BBD might increase with ongoing ocean warming and thermal anomalies, as supported by the here-documented disease outbreak coinciding with a thermal anomaly and widespread coral bleaching. The collection of long-term monitoring disease data in the Arabian Seas is important in order to establish baselines, which can then assist in more accurate prediction of disease prevalence and potential impact of climate change on coral communities in this region.",
"introduction": "Introduction The rise of coral disease outbreaks contributes to the decline of coral reefs globally ( Cróquer & Weil, 2009 ; Harvell et al., 2009 ; Hoegh-Guldberg, 2012 ; McLeod et al., 2010 ; Randall & Van Woesik, 2015 ) and coral disease appears to be the most destructive factor on many reefs. For instances, the Caribbean has been named a “disease hot spot” due to the fast emergence, high prevalence, and virulence of coral diseases in this region ( Rosenberg & Loya, 2013 ). Coral disease outbreaks in the last decades in the Caribbean have resulted in significant losses in coral cover, diversity, and habitat ( Aronson & Precht, 2001 ; Bruckner, 2002 ; Hughes, 1994 ; Precht et al., 2016 ; Weil, 2002 ). Following the mass-bleaching event in 2005 in the US Virgin islands, coral disease outbreaks reduced coral cover by more than 50% ( Cróquer & Weil, 2009 ; Miller et al., 2009 ). Coral diseases were first reported in the Caribbean in the 1970s, including black band disease (BBD), which is considered the most studied coral disease due to its widespread occurrence on reefs around the world ( Bourne, Muirhead & Sato, 2011 ; Richardson, 2004 ). Black band disease has been reported from reefs throughout the Caribbean, the Indo-Pacific regions, the Red Sea, and the Great Barrier Reef ( Al-Moghrabi, 2001 ; Dinsdale, 2002 ; Green & Bruckner, 2000 ; Kaczmarsky, 2006 ; Lewis et al., 2017 ; Montano et al., 2012 ; Page & Willis, 2006 ; Sutherland, Porter & Torres, 2004 ; Weil et al., 2012 ). BBD is the first described coral disease ( Antonius, 1973 ), affecting scleractinian and gorgonian corals ( Green & Bruckner, 2000 ; Sutherland, Porter & Torres, 2004 ; Weil, 2004 ). BBD prevalence generally is considered low ( Dinsdale, 2002 ; Edmunds, 1991 ; Weil, 2002 ); however, this disease is a serious threat to coral reef ecosystems worldwide due to its persistence, leading to coral mortality in the long-term ( Bruckner & Bruckner, 1997 ; Green & Bruckner, 2000 ; Kaczmarsky, 2006 ; Kuta & Richardson, 1996 ; Page & Willis, 2006 ; Sutherland, Porter & Torres, 2004 ; Zvuloni et al., 2009 ). Susceptibility to BBD differs between coral taxa and may result in long-term changes to coral community structure ( Bruckner & Bruckner, 1997 ; Page & Willis, 2006 ). The abundance of BBD is affected by several environmental factors, including seawater temperature, water depth, solar irradiance, host population diversity, and anthropogenic nutrients ( Al-Moghrabi, 2001 ; Kaczmarsky, 2006 ; Kuta & Richardson, 2002 ; Montano et al., 2013 ). Interestingly, seasonal temperatures influence BBD prevalence, with increased virulence during warmer summer months ( Richardson & Kuta, 2003 ; Rützler & Santavy, 1983 ; Willis, Page & Dinsdale, 2004 ), as for example in the Maldives where sea surface temperatures above 28 °C promoted BBD infections ( Montano et al., 2013 ). BBD manifests as a dark band that migrates across the coral colony at a rate of >1 cm/day ( Richardson, 1998 ) leaving behind bare skeleton. The base of the BBD mat is anoxic and high in sulfide levels, causing damage and necrosis to coral tissue ( Ainsworth et al., 2007 ; Carlton & Richardson, 1995 ; Richardson et al., 1997 ). The BBD mat is composed of a polymicrobial consortium, dominated by filamentous cyanobacteria, sulfate-reducing bacteria (SRB), including members of Desulfovibrio spp., sulfide-oxidizing bacteria (SOB) ( Beggiatoa spp.), and other heterotrophic bacteria ( Cooney et al., 2002 ; Miller & Richardson, 2011 ; Sato, Willis & Bourne, 2010 ). As a result of diel light changes, the microbial members of the BBD mat undergo vertical migrations, which causes the harmful microenvironment on top of the coral tissue ( Carlton & Richardson, 1995 ; Miller & Richardson, 2011 ; Richardson, 1996 ). Oxygen depletion and high sulfide concentrations are produced by SRB, which is lethal to the coral tissues and considered the most important factor in BBD pathogenicity ( Glas et al., 2012 ; Richardson, 1996 ; Richardson et al., 1997 ; Richardson et al., 2009 ). Although the functional composition of the BBD mat is conserved, the diversity of the microbial consortium in BBD differs according to geographic location and coral species ( Cooney et al., 2002 ; Frias-Lopez et al., 2004 ; Sekar et al., 2006 ). The occurrence of BBD in the Red Sea was first recorded by Antonius (1988) where the severity of BBD was measured from rare to moderate and mostly correllated with elevated temperatures and seawater pollution. However, baseline data on BBD prevalence in the Red Sea is still lacking. To fill this gap, we conducted surveys to determine the distribution and prevalence of BBD across central Red Sea reefs spanning 4 degrees of latitude. We also detected a bleaching-associated outbreak of BBD on a coral reef in the southern central Red Sea and characterized the microbial community of BBD microbial mats from Coelastrea sp., Dipsastraea sp., Goniastrea sp., and Platygra sp. using high-throughput sequencing. We compared the microbial consortium to that reported from other regions of the world in order to identify biogeographic patterns in the main BBD consortium members.",
"discussion": "Discussion In this study, we report on the distribution and prevalence of coral black band disease in the Red Sea. Our surveys ranged from 19.9 to 24.1 degrees of latitude and confirm the presence of BBD across the central Red Sea. Molecular characterization of the bacterial community identified the three main bacterial members of the disease consortium across coral species at a BBD outbreak site in the southern central Red Sea. Black band disease distribution and prevalence in the Red Sea in comparison to other global sites BBD is a global disease found in numerous regions, but its prevalence on coral reefs is generally low compared to other diseases such as white syndrome (WS) ( Dinsdale, 2002 ; Edmunds, 1991 ; Page & Willis, 2006 ; Willis, Page & Dinsdale, 2004 ). The low prevalence recorded in this study is similar to levels reported elsewhere across the globe ( Sutherland, Porter & Torres, 2004 ) with localized outbreaks of BBD also reported in the GBR ( Sato, Bourne & Willis, 2009 ), Hawaii ( Aeby et al., 2015b ), Jamaica ( Bruckner & Bruckner, 1997 ), Venezuela ( Rodríguez & Cróquer, 2008 ), and the Red Sea ( Al-Moghrabi, 2001 ). In the Red Sea, BBD was first discovered in the 1980s ( Antonius, 1981 ) and our study confirms that BBD is a chronic threat to coral reefs in the Red Sea with localized outbreaks continuing to occur. BBD is not a selective disease; multiple species and various levels of severity can affect colonies within and between coral species and across reefs ( Bruckner, Bruckner & Williams, 1997 ; Dinsdale, 2002 ; Green & Bruckner, 2000 ; Peters, 1993 ). This was also observed in our study, where multiple species were infected, but with differences in prevalence among coral taxa. At the outbreak site, we found BBD prevalence to be highest in the genus Dipsastraea , which suggests that this genus may be an important host for BBD in the Red Sea. Our observations match previous reports and shows that this pattern is consistent through time ( Antonius, 1985 ). Interestingly, although differential susceptibility to BBD among coral taxa has been found globally, the most vulnerable taxa differ by region. For example, in the Caribbean Montastraea/Orbicella are commonly infected ( Bruckner & Bruckner, 1997 ; Porter et al., 2001 ), Montipora in Hawaii ( Aeby et al., 2015b ), and Acropora on the GBR ( Page & Willis, 2006 ). It would be fruitful to examine the underlying defense mechanisms in the different coral taxa that lead to these differences in BBD occurrence. BBD, climate change, and coral bleaching The occurrence of BBD has been linked to elevated seawater temperatures ( Boyett, Bourne & Willis, 2007 ; Kuta & Richardson, 2002 ; Muller & Van Woesik, 2011 ). The occurrence of a BBD outbreak during a bleaching event in the present study reflects previous reports from the Caribbean, where the positive correlation between bleaching events and BBD incidence was proposed first ( Brandt & McManus, 2009 ; Cróquer & Weil, 2009 ). For instance, in the Florida Reef Tract, the prevalence of BBD increased from 0 to 6.7% following bleaching events in 2014 and 2015 ( Lewis et al., 2017 ). Also, Cróquer & Weil (2009) found a significant linear correlation between coral bleaching and the prevalence of two other virulent diseases (yellow band disease and white plague) affecting Montastraea / Orbicella species. This further supports a strong relationship between bleaching events and the emergence of some coral diseases on a global scale. Understanding how climate change-related thermal anomalies and coral bleaching drive the emergence and virulence of coral diseases is essential for future research. It has further been suggested that other anthropogenic activities, such as coastal pollution or ocean acidification, contribute to the increase of coral disease incidents ( Jackson et al., 2001 ; Muller et al., 2017 ; Rosenberg & Ben-Haim, 2002 ). The surveyed outbreak area was adjacent to the outflow of a large aquaculture facility, which might have further aggravated the effects of the bleaching event due to increased nutrient availability ( Roder et al., 2015 ; Ziegler et al., 2016 ). In comparison, other reefs in the Al-Lith area that were farther away from the coast displayed similar levels of bleaching, but BBD prevalence stayed at baseline levels in these locations. This suggests that bleaching alone was not the only factor that could have contributed to the BBD outbreak. The synergistic effects of high temperatures and nutrient pollution find further support in the Caribbean where BBD prevalence increased in reef sites with direct sewage input compared to control sites ( Sekar, Kaczmarsky & Richardson, 2008 ) and in the Bahamas where BBD migration was faster in nutrient-enriched areas ( Voss & Richardson, 2006 ). Further work is needed to directly examine the relationship between bleaching, nutrient stress, and BBD susceptibility. Bacterial community composition of BBD microbial mats from the southern central Red Sea reflects global microbial patterns with local characteristics Our results verify the presence of the three main consortium members in BBD microbial mats (Cyanobacteria, SOB, SRB) of corals from the southern central Red Sea. We identified Oscillatoria sp. as a BBD-associated cyanobacterium, which is similar to the BBD-associated cyanobacteria in other regions of the world ( Aeby et al., 2015b ; Arotsker et al., 2015 ; Buerger et al., 2016 ; Casamatta et al., 2012 ; Cooney et al., 2002 ; Frias-Lopez et al., 2003 ; Gantar, Sekar & Richardson, 2009 ; Glas et al., 2010 ; Meyer et al., 2016 ; Miller & Richardson, 2011 ; Rasoulouniriana et al., 2009 ; Sato, Willis & Bourne, 2010 ; Sussman, Bourne & Willis, 2006 ). However, we retrieved only a low number of cyanobacterial sequences, although cyanobacterial filaments were visually abundant in the sampled microbial mats, which could possibly be related to primer amplification bias. In addition, members of the SOB and SRB functional groups ( Arcobacter sp. and Desulfovibrio sp., respectively) from BBD microbial mats in the southern central Red Sea were similar to those found in other BBD-affected corals worldwide ( Barneah et al., 2007 ; Cooney et al., 2002 ; Klaus, Janse & Fouke, 2011 ; Sekar, Kaczmarsky & Richardson, 2008 ). This confirms that BBD-associated bacteria are not restricted to a specific coral species or region ( Barneah et al., 2007 ; Cooney et al., 2002 ; Dinsdale, 2002 ; Frias-Lopez et al., 2003 ). Interestingly, we did observe white filaments within lesions that were morphologically similar to Beggiatoa , a sulfide-oxidizing bacterium associated with BBD in other regions ( Cooney et al., 2002 ; Miller & Richardson, 2011 ; Sato, Willis & Bourne, 2010 ). However, we found no sequences aligning with Beggiatoa in our study. This suggests that either the white filaments were not Beggiatoa or that the methods used were not adequate to extract and identify Beggiatoa . Aeby et al. (2015b) sequenced Beggiatoa from BBD lesions in Hawaii by first culturing the white filaments from lesions and then using universal bacterial primers 8F and 1513R for sequencing. However, they found that no DNA sequences were available for Beggiatoa found in BBD from other regions even though numerous studies using molecular techniques have been published. Further work is needed to clarify these discrepancies. Besides the three main bacterial consortium members that dominate BBD microbial mats, we detected other bacterial families as part of the BBD consortium. Members of the Firmicutes were abundant in BBD microbial mats, which is consistent with other studies ( Arotsker et al., 2016 ; Arotsker et al., 2009 ; Barneah et al., 2007 ; Cooney et al., 2002 ; Frias-Lopez et al., 2002 ; Miller & Richardson, 2011 ; Richardson, 2004 ; Sekar, Kaczmarsky & Richardson, 2008 ). In addition, we detected the presence of Vibrio species. The pathogenicity of this genus has been documented previously in corals and other marine organisms ( Ben-Haim, Zicherman-Keren & Rosenberg, 2003 ; Harvell et al., 1999 ; Kushmaro et al., 1996 ), and more broadly Vibrios have been characterized as opportunistic taxa ( Cervino et al., 2004 ; Rosenberg & Falkovitz, 2004 ; Thompson et al., 2004 ; Ziegler et al., 2016 ). To date it is unknown whether this group plays a role in the etiology of BBD ( Arotsker et al., 2009 ; Barneah et al., 2007 ) ( Meyer et al., 2016 ), or whether the high number of Vibrios could be related to seasonal increases in the coral microbiome and coral bleaching (reviewed in Rosenberg & Koren, 2006 ; Tout et al., 2015 )."
} | 4,259 |
32676588 | PMC7334800 | pmc | 4,984 | {
"abstract": "In this work, we present a versatile surface engineering strategy by the combination of mussel adhesive peptide mimicking and bioorthogonal click chemistry. The main idea reflected in this work derived from a novel mussel-inspired peptide mimic with a bioclickable azide group (i.e., DOPA 4 -azide). Similar to the adhesion mechanism of the mussel foot protein (i.e., covalent/noncovalent comediated surface adhesion), the bioinspired and bioclickable peptide mimic DOPA 4 -azide enables stable binding on a broad range of materials, such as metallic, inorganic, and organic polymer substrates. In addition to the material universality, the azide residues of DOPA 4 -azide are also capable of a specific conjugation of dibenzylcyclooctyne- (DBCO-) modified bioactive ligands through bioorthogonal click reaction in a second step. To demonstrate the applicability of this strategy for diversified biofunctionalization, we bioorthogonally conjugated several typical bioactive molecules with DBCO functionalization on different substrates to fabricate functional surfaces which fulfil essential requirements of biomedically used implants. For instance, antibiofouling, antibacterial, and antithrombogenic properties could be easily applied to the relevant biomaterial surfaces, by grafting antifouling polymer, antibacterial peptide, and NO-generating catalyst, respectively. Overall, the novel surface bioengineering strategy has shown broad applicability for both the types of substrate materials and the expected biofunctionalities. Conceivably, the “clean” molecular modification of bioorthogonal chemistry and the universality of mussel-inspired surface adhesion may synergically provide a versatile surface bioengineering strategy for a wide range of biomedical materials.",
"conclusion": "3. Conclusion In summary, we upgraded current mussel-inspired surface engineering strategies by the combination of mussel adhesive peptide mimicking and bioorthogonal click chemistry. The mainline of this work is a novel mussel adhesive peptide mimic capped with a bioclickable azide group (i.e., DOPA 4 -azide). Similar to the mussel adhesion mechanism, the peptide mimic DOPA 4 -azide could stably bind onto a broad range of materials via covalent/noncovalent comediated molecular adhesion. In addition, the azide residues on the DOPA 4 -azide-bound surfaces enabled a second-step specific grafting of DBCO-modified bioactive ligands through click reaction. To demonstrate the applicability of our strategy for diversified biofunctionalization, we bioorthogonally conjugated three typical biomolecules on different substrates. The results verified the feasibility to fabricate functional surfaces that matched highly with some essential requirements of medical implants, for instance, the antifouling, antibacterial, and antithrombogenic activity. Overall, this novel surface bioengineering strategy has shown broad applicability in both the types of substrate materials and the intended bioactivities. The molecular modification of bioorthogonal chemistry without hazardous side products and the universality of mussel-inspired molecular adhesion synergistically provide a versatile surface bioengineering strategy for a wide range of biomedical materials.",
"introduction": "1. Introduction Advanced biomedical implants should have the abilities to actively integrate the surrounding tissue, communicate with surrounding cells, trigger cell responses, maintain tissue and organ functions, combat hostile microorganisms, etc. [ 1 , 2 ]. In this regard, surface biofunctionalization represents one of the most straightforward ways to endow biomaterials with such “vitality” [ 3 – 5 ]. Physical adsorption or chemical conjugation is a typical method for surface modification with bioactive ligands, which enables inherently bioinert materials to modulate cell-material interactions, induce specific cell behaviors, and subsequently generate relevant biological effects [ 6 – 8 ]. Common physical means for surface biofunctionalization, such as surface layer-by-layer assembly [ 9 ] and Langmuir-Blodgett deposition [ 10 ], depend on noncovalent molecular bindings. These weak interactions inevitably result in biomolecular desorption and subsequently the lack of long-term activity. Although chemical conjugations show stronger molecular anchoring, current chemical means frequently still suffer from tedious reactions as well as complex surface treatment technologies [ 11 , 12 ]. Moreover, these traditional methods for surface biofunctionalization (i.e., physical binding or chemical conjugation) are not equally applicable on a wide range of material surfaces but require specific adaptation. In this context, a novel surface engineering method, inspired by the marine mussel adhesion, was developed in 2007 [ 13 ]. The molecular mechanism of this method derived from mussel foot proteins (e.g., Mytilus edulis foot proteins, Mfps), in which the repetitive catechol residues of DOPA (3,4-dihydroxy-L-phenylalanine) can produce covalent and noncovalent comediated molecular adhesion [ 14 ]. A great deal of studies indicates that Mfps-mimics (e.g., polydopamine [ 15 – 17 ], DOPA-rich peptides [ 18 , 19 ], and catecholic polymers [ 20 , 21 ]) with catechol groups can adhere stably to virtually all kinds of substrates under wet conditions [ 22 ]. In addition, a second-step conjugation with bioactive molecules through amino- or thiol-mediated Michael addition allows a variety of biofunctionalizations. Undoubtedly, mussel-inspired molecular adhesion can provide a potentially universal strategy for surface bioengineering [ 23 , 24 ]. Despite the simplicity and generality for diversified materials, current mussel-inspired surface strategies still are critically limited with respect to biomolecular modification. First, the second-step chemical conjugation through Michael addition or Schiff base potentially impedes the function of the biomolecule by consumption of essential amino and thiol groups [ 25 ]. Second, the Michael addition or Schiff base has only low specificity and efficiency, taking a toll on the reproducibility and controllability (e.g., heterogeneous molecular conjugation and random molecular orientation) [ 26 ]. Therefore, advanced modification technologies of current Mfps mimics are still demanded for improved surface bioengineering with easy operability, good controllability, and high reproducibility. Herein, we report an advanced surface bioengineering strategy by the combination of mussel-inspired molecular adhesion and bioorthogonal click chemistry ( Scheme 1 ). In contrast to classical chemistry, bioorthogonal click reaction (e.g., the dibenzylcyclooctyne-azide (DBCO-azide) cycloaddition chemistry) shows advantages like specificity, rapidity, thoroughness, and biocompatibility [ 27 , 28 ]. Thus, we considered designing an azide-bearing peptide with multiple catechol groups, mimicking the molecular properties of Mfps. Similar to the Mfps adhesion mechanism, the azide-bearing mussel adhesive peptide can stably bind onto a wide range of material surfaces via the covalent and noncovalent comediated molecular adhesion. Subsequently, the surface anchored azide groups enable a specific grafting of DBCO-modified bioactive ligands through DBCO-azide click reaction in a second step. Since DBCO modification is industrially mature and commercially available for biomolecules, we anticipate that the bioclickable mussel-inspired peptide might provide a flexible and more precise strategy for surface biofunctionalization. As a proof of principle, we synthesized several typical DBCO-modified biomolecules with abilities to modulate cell-material interactions and induce specific biological effects. The basic and essential requirements of biomedical implants, such as antibiofouling [ 29 , 30 ], antibacterial [ 31 ], and antithrombotic activity [ 32 ], were separately introduced onto different substrate materials corresponding to clinically applied medical devices. We demonstrated that the surface bioengineering strategy based on bioclickable and mussel adhesive peptide mimic had broad applicability in both the types of substrate materials and the intended functions. The “clean” molecular modification of bioorthogonal click chemistry and universal surface adhesion of mussel-inspired chemistry may synergically provide a versatile surface bioengineering strategy for a wide range of biomedical materials.",
"discussion": "2. Results and Discussion 2.1. Bioclickable, Mussel Adhesive Peptide Mimic The azide-bearing mussel adhesive peptide mimic was designed based on published sequences and prepared by standard Fmoc-mediated solid-phase peptide synthesis [ 33 – 35 ]. To mimic the multiple catechol structure in Mfps [ 36 ], acetonide-protected DOPA (i.e., Fmoc-DOPA (acetone)-OH) was programmatically linked into the main chain of peptide with one glycine (G) or lysine (K) spacer, leading to a mussel-inspired peptide with tetravalent DOPA sequence (i.e., DOPA-G-DOPA-K-DOPA-G-DOPA). Glycine and lysine act as the spacers to improve molecular twisting and facilitate the Mfps-like molecular adhesion. The gamma amino group of lysine was linked with an azide-terminated poly(ethylene glycol) (PEG), finally obtaining a clickable mussel-inspired peptide mimic DOPA-G-DOPA-K(PEG-azide)-DOPA-G-DOPA (i.e., DOPA 4 -azide, Figure 1(a) ). The peptide was then cleaved from resin and purified through high-performance liquid chromatography (HPLC) (purity: 98.1%). Electrospray ionization mass spectrometry (ESI-MS) and nuclear magnetic resonance (NMR) spectroscopy were used to confirm the success of molecular synthesis. As shown in Figure 1(b) , the monoisotopic mass [M+H] + of DOPA 4 -azide was found at 1336.8 Da, which is in line with its theoretical molecular weight (1335.6 Da) of the chemical structure. The spectrum of 1 H NMR indicated the presence of several diagnostic peaks, including the catecholic and aromatic hydrogens of DOPA, the amide hydrogens, and the hydrogens of ethylene glycol repeating units ( Figure 1(c) ). These results jointly confirmed the successful synthesis of the bioclickable mussel adhesive peptide mimic. 2.2. Diversity of Surface Adhesion We then investigated the applicability of DOPA 4 -azide for surface modification of diversified materials. The materials are widely used for biomedical implants and are commonly very demanding for surface biofunctionalization, such as metals, inorganic materials, and organic polymers. The coating process was carried out by incubating the clean substrates in PBS solutions containing 0.1 mg·mL −1 of DOPA 4 -azide for 1 h ( Figure 1(d) ). As shown in Figure 1(e) , the substrates exhibited significant changes in surface wettability. All the DOPA 4 -azide-coated surfaces showed water contact angles with a rough regression value around 45° (dashed line), attributed to the high hydrophilicity of the PEG chain in DOPA 4 -azide. The typical chemical composition of the DOPA 4 -azide-coating on Au substrate was then characterized by grazing incidence attenuated total reflection Fourier transform infrared (GATR-FTIR) spectroscopy ( Figure 1(f) ). Besides the characteristic peaks of peptide bonds, such as the carboxylic acid (1728 cm −1 , C=O stretching), the amide I band (1629 cm −1 , C=O stretching), and amide II band (1519 cm −1 , N-H stretching), a characteristic peak of azide at 2112 cm −1 was also found. X-ray photoelectron spectroscopy (XPS) analysis was further performed to examine the surface elemental compositions ( Figure 1(g) and Figure S1 ). After DOPA 4 -azide-coating, a significant N1s signal and efficient shielding of the substrate signal were observed for the metallic and inorganic materials. The signals of polymeric materials (e.g., the carbon-based PVC, PET, PU, and PS) were hard to be distinguished, but there was a remarkably enhanced N1s signal after coating. These results demonstrated the versatility of DOPA 4 -azide for surface modification of different classes of materials. 2.3. Antibiofouling Surface Since the bioinspired and bioclickable peptide mimic could be coated on various substrates via mussel-inspired adhesion and lead to azide-functionalized surfaces, a second-step bioorthogonal grafting process in solution with DBCO-capped molecules was further investigated ( Figure 2(a) ). DBCO is a bulky cycloalkyne which reacts specifically with azides through copper-free (i.e., catalyst-free), strain-promoted azide-alkyne cycloaddition (SPAAC) [ 27 ]. Owing to the high specificity, efficient kinetics, and high compatibility in biosystems, bioorthogonal DBCO-azide click chemistry has been widely used for molecular conjugations both in vitro and in vivo [ 28 ]. As a proof of concept, we first employed a DBCO-terminated PEG molecule (Mw = 5000) to form a PEGylated antibiofouling surface. Biofouling, adsorption of biomolecules and cells and subsequent loss of function, is an ongoing problem in the field of biochips and biosensors in contact with biological fluids in vitro or in vivo [ 37 , 38 ]. Herein, the bioorthogonally PEGylated antifouling surface was fabricated on a TiO 2 -deposited quartz slide ( Figure 2(b) ), because surface TiO 2 deposition is widely used on biomedical devices (e.g., vascular stents). GATR-FTIR analysis was used to confirm the success of PEGylation. As shown in Figure 2(c) , the azide peak in FTIR spectra disappeared, accompanied by the appearance of a group of triazole peaks after bioorthogonal PEGylation, indicating the efficient bioorthogonal reaction between DBCO-PEG and azide residues. In addition, XPS analysis revealed a significant decrease of the N1s signal, probably due to the shielding effect of PEG chains on the N-element-rich DOPA 4 -azide layer ( Figure 2(d) ). These results demonstrated the efficient PEGylation on the TiO 2 surface via DOPA 4 -azide adhesion and DBCO-azide conjugation. The antifouling property of the PEGylated surface was further examined by checking the inhibition of nonspecific cell adhesion. It is well known for vascular implants that excessive smooth muscle cell (SMC) growth and platelet adhesion will result in intimal hyperplasia and thrombosis [ 39 ], which are the main causes of device failure. As an example for illustration, here, we investigated the inhibitory effect of our PEGylated surfaces on SMCs and platelet adhesion. Human umbilical artery SMCs were first seeded on the TiO 2 , DOPA 4 -azide-coated, and PEGylated surfaces and cultured for 2, 24, and 72 h. The SMC adhesion and proliferation behaviors were investigated by fluorescence microscopy and CCK-8 (cell counting kit 8) assay. As shown in Figure 2(e) , remarkable inhibition of SMC adhesion and growth on the PEGylated surface could be observed, and the inhibitory effect did persist for 3 days. In contrast, the original TiO 2 surface and DOPA 4 -azide-coated surface showed strong SMC adhesion (nonspecific) and proliferation. Quantitative results by the CCK-8 assay and cell counting further confirmed the strong inhibitory effects of our PEGylated surface for SMC growth, in particular, the continuous inhibition of SMC adhesion (Figures 2(f) and 2(g) ). In addition to SMCs, we also applied the PEGylated surface to test blood platelet adhesion. Likewise, bioorthogonal PEGylation on DOPA 4 -azide-coated TiO 2 significantly inhibited the adhesion of platelets (Figures 2(h) – 2(j) ). Besides the significant reduction of platelet adhesion and spreading, the PEGylated surface also showed a low degree of fibrinogen adsorption and activation. Obviously, the bioorthogonal PEGylation assisted by mussel-inspired peptide adhesion would be a promising strategy for the fabrication of antifouling coatings on implanted biomaterials like stents, biosensors, and biochips. 2.4. Antibacterial Surface Apart from the engineering of an antifouling surface, the bioinspired peptide mimic DOPA 4 -azide could also be used for the fabrication of an antimicrobial surface. For medical implants and devices (e.g., urinary catheters and orthopedic and dental implants), bacterial infections after implantation are associated with increased frequency and length of hospitalization as well as the risk of implant failure [ 40 , 41 ]. Antibacterial functionalization of implants thus is highly demanded in the field of surface bioengineering. In this context, we designed a DBCO-modified antibacterial peptide (ABP), which was then used to engineer an antibacterial surface with the assistance of DOPA 4 -azide. Currently, there are only limited strategies for surface bioengineering of polymeric implants compared to metal implants, probably due to the chemical inertness of biomedically used polymer materials. Thus, we chose polyvinyl chloride (PVC) substrate (which is commonly used for medical tubes) to demonstrate the possibility of our method for biomodification of polymer implants. In this part, a representative ABP (HOOC-WFWKWWRRRRR-NH 2 ) [ 42 ] was employed as the antibacterial backbone, which was linked with a PEG spacer and the DBCO group to obtain a DBCO-modified ABP (DBCO-ABP, Figure 3(a) ). The ESI mass spectrum of the DBCO-ABP indicated the monoisotopic mass of [M+3H] 3+ and [M+4H] 4+ at 871 and 654 Da, respectively ( Figure 3(b) ). The result was in line with its theoretical molecular weight (2612.0 Da). The DBCO-ABP was then incubated with DOPA 4 -azide-coated PVC substrates to fabricate antibacterial surfaces ( Figure 3(c) ). As shown in Figure 3(d) , there is a remarkable decrease of azide groups in the FTIR spectra, accompanied by the appearance of triazole after bioorthogonal conjugation. In addition, a significant increase of N1s signal and an appearance of S2p signal were found in the XPS spectrum of the ABP-modified surface ( Figure 3(e) and Figure S2 ), due to the N-element-rich chemical composite of the DBCO-ABP layer. These results jointly confirmed the successful fabrication of ABP-modified PVC substrates by the combined use of DOPA 4 -azide and DBCO-ABP. The antibacterial properties of ABP-modified PVC were further examined by using E. coli and S. aureus . A drop of bacterial suspension was distributed on the bare, DOPA 4 -azide-coated, and ABP-modified PVC substrates for solid culture tests, or the samples were fully immersed in bacterium suspensions for liquid culture. The ABP-modified surface showed potent bacterial inhibition in both solid and liquid media. As shown in Figure 3(f) , no bacterial colony, regardless of Gram-positive or Gram-negative strains, was found on the ABP-modified surfaces after 24 h of culture. In contrast, the two control groups, including the bare and DOPA 4 -azide-coated surfaces, were both covered with a high density of bacterial colonies. A similar result was also observed in liquid media. The bacterial solutions incubated with ABP-modified PVC show a clear state, while the others appeared distinctly turbid in the bacterial suspension, implying the efficient inhibition of bacterial growth. According to the optical density at 600 nm (OD 600 ), we found that more than 99% of the bacteria could be killed by the ABP layer in 12 h (Figures 3(g) and 3(h) ). In addition, such potent inhibitory effect on bacteria could last for 1 month and more (Figure S3 ), indicating the durable antibacterial activity and also the high stability of the ABP layer. The above study confirmed the successful fabrication of an antibacterial surface, indicating the high applicability of our bioclickable mussel-inspired peptide mimic for surface engineering antibacterial coatings on the medically used, in particular, polymer-based implants. 2.5. Antithrombogenic Surface As one of the most common physiological and pathological phenomena, thrombosis occurs as a host defense mechanism to preserve the integrity of the closed circulatory system after vascular damages [ 43 ]. However, the development of clots in circulation after therapeutic intervention is the most frequent cause of morbidity and mortality. Particularly, in the field of cardiovascular stents, chronic and acute interfacial thrombogenesis inevitably happens due to the vascular injury caused by stent expansion [ 44 ]. Thus, biofunctionalized stents with antithrombotic activity are highly desired. In this context, we further demonstrated the potential of our methods for the fabrication of an antithrombotic surface. The NO (nitric oxide, a gaseous signaling molecule)-generating compounds have been well studied for surface engineering of vascular stents to prevent platelet activation and aggregation, inhibit thrombogenesis, suppress SMC proliferation, promote EC growth, etc. [ 45 , 46 ]. Accordingly, we designed a DBCO-capped NO-generating compound for surface engineering. As is well studied in previous work, the transition metal ion Cu(II) has excellent glutathione peroxidase- (GPx-) like activity [ 47 , 48 ], which can catalytically generate NO from both endogenous and synthetic S-nitrosothiols (RSNOs) by decomposing them in the presence of reduced glutathione (GSH). In order to immobilize Cu ions, a cyclen DOTA (1,4,7,10-tetraazacyclododecane- N , N ′, N ″ , N ‴ -tetraacetic acid) was conjugated with a DBCO group ( Figure 4(a) and Figure S4 ) [ 49 ]. The Cu(II)-cyclen complex (DOTA@Cu) thus could be bioorthogonally conjugated on a DOPA 4 -azide-coated substrate to obtain a NO-generating surface (Figures 4(b) and 4(c) ). In this study, 316L stainless steel (SS) foil was used as the model substrate since the material is widely used for vascular stents. After DOTA@Cu modification (Figure S5 ), the in vitro NO-releasing property was first determined by a real-time chemiluminescent assay. PBS solution containing 10 μ M reducing agent GSH and 10 μΜ S-nitrosoglutathione (GSNO, an endogenous NO donor) [ 50 ] was used to simulate the blood environment. Real-time monitoring of the NO flux revealed a steady NO generation from the DOTA@Cu-modified surface (Figure S6 ). Ageing studies showed that efficient NO release could last for more than 2 weeks ( Figure 4(d) ), indicating the suitability for long-term use. Since thrombogenesis involves a series of biochemical processes like platelet aggregation, coagulation, and fibrinolysis [ 43 ], we then checked the in vitro antiplatelet property. Without donor supply, all surfaces induced substantial platelet adhesion and activation in 30 min, and the DOTA@Cu-modified 316L SS foil showed almost no inhibition in the amount and activation rates of adherent platelets (Figures 4(e) and 4(f) ). Upon the addition of the NO donor, significant changes were observed on the DOTA@Cu-modified surface. The controls (i.e., the bare and DOPA 4 -azide-coated 316L SS substrates) had evident platelet adhesion, and the spread morphology of platelets indicated a high degree of activation and aggregation. In contrast, the DOTA@Cu-modified 316L SS foil showed substantially reduced platelet adhesions with an inactive spherical state. With the positive result in vitro , we then investigated the antithrombogenic property using ex vivo perfusion experiments. The control and DOTA@Cu-modified 316L SS foils were curled up and placed onto the inner walls of commercially available cardiopulmonary perfusion tubes, which were then connected to a rabbit arteriovenous (AV) shunt ( Figure 4(g) ) [ 45 ]. The ability of different groups to support blood flow was evaluated in the presence of the NO donor. After 2 h of ex vivo circulation, the sizes of occlusive thrombosis, thrombus weight, and blood flow rates in the circuit were evaluated (Figures 4(h) – 4(j) ). Optical microscope photos and SEM images both showed serious thrombus formation on the two control groups (i.e., the bare and DOPA 4 -azide-coated 316L SS foils). In contrast, only a small number of cruor were observed on the NO-releasing DOTA@Cu-modified foil. These results jointly confirmed the perfect hemocompatibility and antithrombogenic property of the DOTA@Cu-modified surface. It can be concluded that this study demonstrated the potential of our clickable peptide mimic for surface bioengineering of vascular implants with high antithrombogenic activity."
} | 6,045 |
34699370 | null | s2 | 4,985 | {
"abstract": "Various nonclassical approaches of distributed information processing, such as neural networks, reservoir computing (RC), vector symbolic architectures (VSAs), and others, employ the principle of collective-state computing. In this type of computing, the variables relevant in computation are superimposed into a single high-dimensional state vector, the collective state. The variable encoding uses a fixed set of random patterns, which has to be stored and kept available during the computation. In this article, we show that an elementary cellular automaton with rule 90 (CA90) enables the space-time tradeoff for collective-state computing models that use random dense binary representations, i.e., memory requirements can be traded off with computation running CA90. We investigate the randomization behavior of CA90, in particular, the relation between the length of the randomization period and the size of the grid, and how CA90 preserves similarity in the presence of the initialization noise. Based on these analyses, we discuss how to optimize a collective-state computing model, in which CA90 expands representations on the fly from short seed patterns-rather than storing the full set of random patterns. The CA90 expansion is applied and tested in concrete scenarios using RC and VSAs. Our experimental results show that collective-state computing with CA90 expansion performs similarly compared to traditional collective-state models, in which random patterns are generated initially by a pseudorandom number generator and then stored in a large memory."
} | 392 |
25437804 | PMC4243450 | pmc | 4,986 | {
"abstract": "The global clustering of gene families through network analysis has been demonstrated in whole genome, plasmid, and microbiome analyses. In this study, we carried out a plasmidome network analysis of all available complete bacterial plasmids to determine plasmid associations. A blastp clustering search at 100% aa identity cut-off and sharing at least one gene between plasmids, followed by a multilevel community network analysis revealed that a surprisingly large number of the plasmids were connected by one largest connected component (LCC), with dozens of community sub-groupings. The LCC consisted mainly of Bacilli and Gammaproteobacteria plasmids. Intriguingly, horizontal gene transfer (HGT) was noted between different phyla ( i.e. , Staphylococcus and Pasteurellaceae ), suggesting that Pasteurellaceae can acquire antimicrobial resistance (AMR) genes from closely contacting Staphylococcus spp., which produce the external supplement of V-factor (NAD). Such community network analysis facilitate displaying possible recent HGTs like a class 1 integron, str and tet resistance markers between communities. Furthermore, the distribution of the Inc replicon type and AMR genes, such as the extended-spectrum ß-lactamase (ESBL) CTX-M or the carbapenemases KPC NDM-1, implies that such genes generally circulate within limited communities belonging to typical bacterial genera. Thus, plasmidome network analysis provides a remarkable discriminatory power for plasmid-related HGT and evolution.",
"conclusion": "4. Conclusions We conducted a plasmidome network analysis for all of the available complete bacterial plasmids. Most of the plasmids were connected to a single LCC, which mainly consisted of plasmids from Bacilli and Gammaproteobacteria . The LCC was subdivided into dozens of communities, which consisted of plasmids from a mostly single class of bacteria. Such community network analysis facilitate displaying possible recent HGTs like a class 1 integron, str and tet resistance markers between communities. As well as the class 1 integron, the CTX-M ß-lactamase can be identified in multiple communities. Therefore further dissemination could be generated either by plasmid transfer itself or transposition of the CTX-M ß-lactamase into other broad-host range plasmids “en bloc”, suggesting that this community network analysis could propose crucial information for the manner of dissemination in AMR. Intriguingly, HGT was noted between different phyla ( i.e. , Staphylococcus and Pasteurellaceae ( P. multocida and H. parasuis )), suggesting that both organisms can acquire AMR genes from closely contacting each other. The distributions of the Inc-type and AMR genes implied that such genes generally circulate within communities composed of typical bacterial taxa. This plasmidome network analysis under very strict parameter settings provides remarkable discrimination power for plasmid-related recent HGT. It also provides a large-scale analysis of plasmid associations, improving our understanding of current plasmid dissemination and evolution among bacterial communities.",
"introduction": "1. Introduction Bacterial plasmids are self-replicating, extrachromosomal replicons that are key agents of change in microbial populations [ 1 ]. It is well known that plasmid vectors are frequently transmitted from one bacterium to another [ 2 , 3 , 4 , 5 ]. Plasmids, as well as conjugative transposons, are crucial mediators of horizontal gene transfer (HGT), a process that significantly influences bacterial activity and evolution. Most antimicrobial resistance (AMR) genes can be acquired through plasmid-mediated HGT [ 2 ]. The increased number of AMR genes acquired by pathogenic bacteria threatens to return us to the pre-antibiotics era. The misuse of antimicrobial medications accelerates this natural phenomenon. Moreover, poor infection control practices have encouraged the spread of AMR. Thus far, it has been challenging to reveal the global dissemination of AMR plasmids by molecular genotyping. Molecular genotyping techniques, such as MLST, MLVA, and IS-typing, are available to hospitals, communities, and global organisations for tracing clonal bacteria dissemination. Furthermore, next-generation sequencing (NGS) can identify bacterial strain-specific genetic markers including single nucleotide polymorphisms, facilitating the identification of disseminated pathogenic bacteria clones, such as Staphylococcus aureus [ 6 ], Clostridium difficile [ 7 ], Acinetobacter baumannii [ 8 ], and Mycobacterium tuberculosis [ 9 ]. AMR can be widely transferred among different strains and species from closely related taxonomic classes [ 2 , 3 , 10 ]. Indeed, the extended-spectrum ß-lactamase (ESBL) CTX-M and several carbapenemases have been widely identified in many species of Enterobacteriaceae [ 11 ], indicating that a molecular characterisation of plasmid sequences would be more beneficial for tracing the transmission of AMR than would characterising whole genome sequences, because most AMR genes can be acquired through plasmid-mediated HGT [ 2 ]. As whole genome sequencing of a strain harbouring multiple plasmids is not able to characterize single source of plasmid genetic feature, sequencing only plasmid is rather simple and cost effective. Previous research has investigated global clustering of gene families using network analysis, as well as global network analysis of DNA families shared among cellular, plasmid, and phage genomes [ 12 ]. Similar studies focusing on plasmids showed frequent HGT over geographical habitat or taxonomical barrier [ 13 , 14 ]. The evolutionary dynamics of plasmids in Acinetobacter spp. (pan-plasmidome analysis) have also been examined [ 14 ]. Furthermore, both microbiome and plasmidome analyses have been performed for the bovine rumen [ 15 ]. Plasmidome analysis in particular has been carried out for ESBL in Escherichia coli [ 16 ] and various plasmids in Enterococcus faecalis [ 17 ]. Module-based phage network analysis has been demonstrated to suggest a reticulate representation of evolutionary and functional relationships between phage genomes [ 18 ]. These intensive studies focused on comprehensive network analysis using relatively lower gene similarity (~95%), thus, we performed a network analysis at 100% aa identity cut-off and sharing at least one gene between plasmids, leading to characterize the most recently occurred HGT or plasmid transfer. Such strict parameters will facilitate a better understanding of the current and future prospects of AMR dissemination based on the plasmid transfer.",
"discussion": "2. Results and Discussion In total, 3793 complete plasmid sequences were selected from the NCBI database. Accordingly, 275,954 protein sequences were extracted (downloaded on 10 June, 2013). 2.1. Number of Edges and Nodes between Each Plasmid Plasmids that share at least one homologous sequence were extracted as connected nodes based on either a blastp or uclust homology search of coding sequences in each plasmid. As shown in Table 1 , the number of connected-nodes decreases as the identity cut-off value increases, indicating that lower thresholds generate more connected nodes. The single connected components (CCs) are composed of all of the plasmids (nodes) with at least a single connection to others. The number of CCs is always slightly larger for uclust than it is for blastp, suggesting that blastp is a more rigorous clustering search method for finding the homologous lineages of ORFs following CC assignment. Thus, we performed blastp searches for the subsequent plasmidome network analysis. We obtained an image of the plasmidome network using a 100% identity cut-off value ( Figure 1 A). pathogens-03-00356-t001_Table 1 Table 1 Number of connected component and community according to identity cutoff values. Clustering program Identity cutoff (%) Edges Connected nodes Connected components Nodes in LCC LCC % in total Number of communities in the LCC using multilevel method BLASTP 50 251,961 3529 57 3265 86.1% 19 BLASTP 60 177,932 3444 70 2793 73.6% 25 BLASTP 70 129,545 3358 90 2613 68.9% 31 BLASTP 80 104,815 3229 127 2265 59.7% 26 BLASTP 90 81,752 3060 162 1995 52.6% 34 BLASTP 99 52,029 2633 236 1389 36.6% 36 BLASTP 100 41,956 2496 259 1241 32.7% 26 UCLUST 50 188,844 3524 71 3233 85.2% ND UCLUST 60 147,899 3435 75 2762 72.8% ND UCLUST 70 121,841 3349 98 2558 67.4% ND UCLUST 80 101,902 3213 133 2241 59.1% ND UCLUST 90 82,046 3057 174 1949 51.4% ND UCLUST 99 52,578 2626 240 1369 36.1% ND UCLUST 100 43,275 2501 263 1232 32.5% ND CC: connected component; LCC: largest connected component; ND: not determined. 2.2. Connected Components (CCs) The largest CC (LCC) contains 32.7% of all the investigated plasmids, even after increasing the identity cut-off value to 100% ( Figure 1 A, Table 1 ). Although this network may contain identical plasmid pairs by means of vertical inheritance or plasmid transfer, the existence of such a large LCC from vast variety of hosts (will be discussed in the next section) clearly indicates that most of the plasmids share genes by HGT or plasmid transfer, suggesting that a frequent and ongoing dissemination of genes occurs between plasmids. To detect most recently occurred HGTs (not just phylogenetic relationships) we selected a threshold with identity cut-off value of 100% for further analysis. Under this condition 1297 plasmids from many kinds of bacterial species remain unconnected. As well as the previous study, the LCC showed power law distribution of the degree measures of nodes, which comes from the “scale-free” network feature [ 19 , 20 ]. Accordingly, the LCC mostly is composed of Enterobacteriaceae, such as Escherichia, Klebsiella, and Salmonella, corresponding to the result in Tamminen et al. [ 13 ]. Figure 1 Plasmidome network analysis. ( A ) The filled circles (nodes) represent the respective plasmids. The grey lines (edges) represent the connections among the plasmids. The thickness of the edges represents the number of genes shared among the plasmids. The edges denoting large numbers of shared genes appear as grey ellipses. The colours of the nodes are automatically set according to the plasmid communities inferred from the multilevel method using the R igraph package. The nodes within the dotted circle represent the largest connected component (LCC). ( B ) Close-up of the connecting points between Gammaproteobacteria and Bacilli . 2.3. Community Network Analysis The obtained LCC contains large number of nodes for further representing the characteristic network structure, thus, we subdivided the LCC into several “communities”, using available clustering methods in the R igraph package [ 21 ] including the fastgreedy, multilevel, edge betweenness, walktrap, label propagation, and infomap methods. To estimate the best method for community analysis, we employed the Shannon index [ 22 ], which estimates the diversity index of the host bacterial lineage. We selected the method that provided the set of communities with the lowest Shannon index. As shown in Figure 2 , the fastgreedy, multilevel and betweenness methods generated relatively better results, with lower diversity index values at every identity cut-off level for distinct taxonomic levels ( Figure 2 A: phylum, 2B: class, 2C: order). Accordingly, the fastgreedy and multilevel methods generated significantly fewer communities in the LCC ( Figure 2 D). The multilevel method, however, was considered a better choice at every identity cut-off. Thus, we decided to employ the multilevel method in the subsequent analysis because it obtained the fewest communities. Based on the multilevel method, a total of 291 communities (e.g., co68 …) were extracted from all CCs at an identity cut-off of 100%, while the LCC can be well subdivided into 26 communities ( Table 1 and Figure 3 ). Figure 2 Evaluation of community connections using several methods. The Shannon index values for the LCC using different identity cut-off values and community estimation methods at the phylum ( A ), class ( B ), and order ( C ) levels. ( D ) The numbers of communities in the LCC using different identity cut-off values. Figure 3 A schematic representation of the communities from the whole image shown in Figure 1 . The communities mainly included Alphaproteobacteria , Betaproteobacteria , Gammaproteobacteria , Bacilli , and Borrelia , which are indicated by purple, green, pink, grey, and black circles, respectively. The other communities are indicated by blue circles. The number of the top three species in the community is shown beside the circle. The circle with the light-blue rim represents the communities without AMR genes. The red lines represent the AMR gene related connections among the communities. Most of the communities in the LCC consisted of plasmids from the Gammaproteobacteria, Alphaproteobacteria, Betaproteobacteria and Bacilli classes ( Figure 4 A) or Enterobacteriales and Bacillales orders ( Figure 4 B). For instance, co68 was mostly composed of Enterobacteriales, such as Klebsiella , Escherichia , and Salmonella spp. ( Figure 3 and Figure 4 B). In addition, co22 and co29 were composed of Enterobacteriales such as Escherichia , Salmonella and Haemophilus spp. ( Figure 3 and Figure 4 B). Some specific bacterial plasmids were predominant in smaller communities (e.g., Yersinia in co63, Acinetobacter in co82, Staphylococcus in co8 and co258, and Xanthomonas in co53). Figure 4 The host bacterial components of the communities. The upper and lower graphs show the numbers and ratios of the host bacterial components, respectively, for the class ( A ) and order ( B ) levels. Clustering coefficient values in any communities in LCC was estimated as median 0.681 (SD: 0.125), indicating that no community consisted of homogenous plasmids. The community network analysis suggests that bacterial plasmids can be grouped into communities according to biologically relevant features. 2.4. Connections between Communities 2.4.1. Genes Involved in the Connections between Communities The abovementioned multilevel community analysis suggests a crucial linkage among the 26 communities in LCC. Thus, we searched for genes involved in the unique connections between communities. Blast search against the COG database revealed that the “[L] Replication, recombination and repair” subgroup was the predominantly detected at 60.0% (2283 of 3806) as transposases on the edges ( Figure 5 ), indicating that multiple IS-related transposases could be involved in HGT between communities. We found that most of the edges were related to antimicrobial resistance genes (see the red line edges between the communities in Figure 3 ), suggesting that the “[V] defence mechanism” category could play a crucial role in the connection for HGT due to antimicrobial selective pressure ( Figure 5 ). This result suggests that genetic exchange events between communities could be driven by genes related to replication/recombination/repair functions (COG [L]), as well as genes related to AMR. Figure 5 COG annotation for genes on the edges between communities. ( A ) COG categories on all of the edges in the LCC. ( B ) The COG ID for the category [L] of replication, recombination and repair. ( C ) The COG ID for the category [V] of defence mechanisms. 2.4.2. Frequently Identified Gene Combination among the Edges between Communities Since shared genes on the edge between communities could be a possible candidate for recent HGT genes, frequently shared gene was extracted from the total non-redundant 1506 genes on the 66 edges in LCC. The combination of ybaA, sul1, intI1 and qacEΔ1 were most frequently identified on 21 edges ( Figure 6 ). This combination has been well characterized as class 1 integron, which is known as a major player of indiscriminate dissemination of AMR [ 23 ]. In addition, the combinations of “insB, strA, and strB” (streptomycin resistance marker) and “tetR and tetA” (tetracycline resistance marker) were also frequently found on 15 edges, respectively. The str and tet combinations were not found on the co193 and co 64-related edges, respectively. Figure 6 Frequently hit sequence on the 66 edges in LCC. Heatmap of sequence pairs of 1506 sequences on the 66 edges in LCC. Close-up of top 9 frequently hit sequences. Class 1 integron (21 edges), str (15 edges) and tet (15 edges) genes on the edges were shown in bold red line. In addition, we confirmed whether class 1 integron could be transmitted en bloc. The basic components of class 1 integrons consist of IntI1, QacEdelta1, Sul1 and YbaA ( Table 2 : 01-02/03-04-05). Forty plasmids from the 3793 complete plasmid sequences in this study have been identified to carry the contiguous components of a class 1 integron. Further investigation suggested that additional AMR genes are integrated into the class 1 integron as cassette gene ( Table 2 ). A class 1 integron group including additional aminoglycoside resistance (01-12-02-04-05) was found in communities co18 (plasmid: DQ364638 and AP000342), co29 (plasmid: AB605179), and co68 (plasmid: HQ201416). Furthermore, class 1 integrons with integrated trimethoprim-resistant dihydrofolate reductase (01-36-12-02-04-05) was found in communities co29 (FN432031) and co64 (JX566770). These results strongly suggest that complete class 1 integrons together with additional integrated AMR genes could be transmitted among communities . Such AMR markers could play a crucial role for recent HGT, leading to contribute to the vast linkages among variable plasmid communities as the global dissemination of AMR. pathogens-03-00356-t002_Table 2 Table 2 The class 1 integron mediated en bloc transmission of AMR genes among communities. Order of gene groups Community Total co18 co22 co29 co44 co64 co68 co193 1 1 1 1 1 1 2 1 1 4 1 1 1 1 1 1 2 2 1 1 2 1 1 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 3 1 1 1 1 1 1 1 1 1 1 Total 3 9 2 2 1 22 1 40 Gene group Products 01 integron integrase IntI1. class 1 integron component 02, 03 quaternary ammonium compound-resistance protein QacEdelta1. 04 sulphonamide resistant dihydropteroate synthase Sul1. 05 protein YbaA. 06 aminoglycoside 3'-\n N -acetyltransferase protein aacC . aminoglycoside resistance 07 aminoglycoside 3'-\n N -acetyltransferase. 08 aminoglycoside acetyltransferase. 09 aminoglycoside adenylyltransferase. 10 aminoglycoside\n N (6')-acetyltransferase. 11–14 aminoglycoside resistance protein\n aadA . 15 aminoglycoside-2'-adenylyltransferase. 16 acetyltransferases\n aac(3)-Ia . 17 putative aminoglycoside 6'-\n N -acetyltransferase. 18 putative aminoglycoside adenyltransferase. 19 class D β-lactamase OXA-10. β-lactamase 20 class D β-lactamase OXA-21. 21 class D β-lactamase OXA-2. 22 class D β-lactamase OXA. 23 metallo-β-lactamase GIM-1. 24 metallo-β-lactamase IMP-6. 25 metallo-β-lactamase VIM-1. 26 chloramphenicol acetyltransferase catB2. chloramphenicol resistance 27 chloramphenicol acetyltransferase catB8. 28, 29 chloramphenicol acetyltransferase. 30 chloramphenicol aminotransferase. 31 chloramphenicol resistance protein CmlA1. 32 dihydrofolate reductase DfrA5. dihydrofolate reductase 33-35 dihydrofolate reductase. 36 trimethoprim-resistant dihydrofolate reductase type I DhfrA1. 37–40 transposase. \n 41 molecular chaperone GroEL. \n 42 molecular chaperone GroES. \n 43 fluoroquinolone resistance protein QnrB2. \n 44 quinolone-resistance determinant Qnr. \n 45–52 Others \n 53-60 hypothetical protein. \n 2.4.3. Connections between Plasmid Communities from Different Divisions The community network analysis revealed unique linkages between Bacilli (co42 and co60) and Gammaproteobacteria (co22) ( Figure 1 B and Figure 3 ). The plasmids from Staphylococcus and Pasteurellaceae were connected by two plasmids from Pasteurella multocida (pCCK411, GenBank_ID: FR798946) and Haemophilus parasuis (pQY431, GenBank_ID: KC405065) ( Figure 1 B). The pQY431 in H. parasuis appeared to have acquired a bifunctional ACC/APH gene ( aacA-aphD , AGK85216.1) from a plasmid in Staphylococcus spp., because GC-content of ACC/APH gene corresponds to the average of S. aureus plasmid, while it is lower than the average of H. parasusis plasmid. This plasmid also have a ß-lactamase ( bla ROB-1 , AGK85217) from another plasmid in Pasteurellaceae species ( Figure 7 A). In addition, the pCCK411 plasmid in P. multocida shares a kanamycin resistance gene ( aphA3 , CBZ06037) with Staphylococcus plasmids and the bla ROB-1 gene from another Pasteurellaceae plasmids. Interestingly, higher GC content in this region in Staphylococcus suggests that this gene might be acquired from other plasmids such as Pasteurella multocida ( Figure 7 B). An additional blastp search with the NCBI nt database indicated that the identical sequences of the aacA-aphD and aphA3 genes were detected in plasmids from the bacteria from the Bacilli class, such as Staphylococcus , or Enterococcus plasmids. Figure 7 Pairwise comparisons of the ( A ) Haemophilus parasuis and ( B ) Pasteurella multocida plasmids carrying the aacA-aphD and aphA3 genes compared to Staphylococcus aureus plasmids. The light-blue boxes represent the genes on the plasmids. The red and blue bars represent the homologous regions in the forward and inverted directions, respectively. The plasmids carrying the homologous genes are shown below the pairwise comparison. ( C ) Average GC-content of the host chromosomes. Of particular note, Haemophilus / Pasteurella spp. required an external supplement of V-factor (NAD) from staphylococci for growth support [ 24 ]. In addition, the Pasteurellaceae Hin subclade ( H. influenzae, P. multocida ) possessed the natural competence and transformation ability [ 25 ]. These observations support the hypothesis that the aacA-aphD gene in H. parasuis (pQY431) was introduced through close contact with Firmicutes such as staphylococci. Contrary, there has been no evidence how Staphylococcus could acquire the aphA3 gene, but the close relationship may have promoted gene exchange between plasmids originating from different phyla. This study revealed high relationship between Pasteurellales and Staphylococcus ( Figure 7 ), although previous report characterized the gene transfer between Actinobacteridae and Gammaproteobacteria [ 13 ]. Such difference generated from the difference of parameter threshold, i.e. , Tamminen et al. (2012) required at least five shared homologous sequencess and 95% aa identity to connect plasmids, while this study picked up plasmid connections from at least one shared homologous sequences at 100% aa identity, indicating more stringent threshold could display a potential recent HGT in this study. 2.5. Distribution of Inc Types among Plasmid Communities Plasmids can be categorised by the plasmid replicon incompatibility group (Inc-group), using PCR-based methods to identify the replicons of the major plasmid families found in Enterobacteriaceae [ 26 ]. We performed in silico replicon typing for all of the plasmids against 22 known Inc groups ( Table 3 ). The Inc types could be assigned to some plasmids in the LCC, and showed strong correlations between the community and the Inc type (e.g., IncFII: co18, FIIS: co18 and co29, FIIk: co68, N: co68, A/C: co22, I1: co64, and P: co193, in Table 3 ). IncFII-co18 and IncFIIS-co18 were found predominantly in Escherichia , while IncFIIk-co68 was found predominantly in Klebsiella and IncFIIS-co29 was found predominantly in Salmonella ( Figure 8 ) . In contrast, IncA/C-co22 was composed of plasmids from various bacterial genera ( Figure 8 ), suggesting that IncA/C has greater dissemination potential among Enterobacteriacea . IncN-co68 was closely related between Escherichia and Klebsiella . In addition, IncI1-co64 was closely related between Escherichia and Salmonella , suggesting that frequent potential transmissions can occur between these two bacterial genera. Interestingly, significantly larger amount of plasmids in both co22 and co68 could not be assigned for Inc type, implying the variety of Inc types remain to be elucidated. pathogens-03-00356-t003_Table 3 Table 3 Distribution of Inc types and communities. Inc type Community co11 co18 co22 co29 co64 co68 co192 co193 A/C - - 25 - - 2 - - B/O - 3 - - 1 1 - - FIA - 5 - - - 3 - - FIB - 5 - - - - - - FIB-M - - - - - 1 - - FII - 60 - - 1 3 - - FIIk - - - - - 17 - - FIIS - 23 - 27 - 3 1 - HI1 - 1 - - - 1 - - HI2 - - - - - 5 - - HIB-M 3 - - - - - - - I1 - - - - 28 - - - K - 1 - - 3 - - - L/M - - - - - 7 - - N - - - - - 26 - - P - - - - - - - 7 R - - 2 - - 3 - - T - 1 - - - 1 - - U - - - - - 3 - - W - - 2 - - 3 - - X1 - 1 - 7 - - - - X2 - - - - - 1 - - Not Assigned 1 13 83 37 3 144 2 10 Total 4 113 112 71 36 224 3 17 Shaded communities showed significant over-representation with p -value < 0.001 by chi-square test. Figure 8 Components of the major Inc-type plasmids. The components of the communities of the major Inc-type plasmids and the bacterial components for each community are shown ( Table 3 ). 2.6. Distribution of AMR among Plasmid Communities Apart from the Inc-type related communities, the associations of AMR genes within communities were also investigated. CTX-M ß-lactamase is a broadly disseminated ESBL in Enterobacteriacea [ 27 ]. The CTX-M-1 and CTX-M-9 subgroups were broadly classified into multiple communities (co18, co22, co29, co64, and co68) ( Figure 9 A), while the CTX-M-2 subgroup was classified only as co68. More than 80% (40/49) of plasmids carrying the CTX-M ß-lactamase were found in co18 or co68. In addition, CTX-M-1 and CTX-M-9 were the most frequently found in the IncFII replicon plasmid ( Table 4 ), suggesting that the dissemination of the CTX-M genes is linked to Inc replicon-dependent plasmid communities between co18 and co68 ( Figure 3 ). Intriguingly, the abovementioned communities (co18, co22, co29, co64, and co68) were tightly connected to each other according to the community network analysis ( Figure 3 ). Most of the plasmids carrying CTX-M-1 and CTX-M-9 were derived from Escherichia or Klebsiella , while most of the plasmids carrying CTX-M-2 were from Salmonella and Klebsiella . As well as the class 1 integron, CTX-M ß-lactamases have been acquired through insertion sequence mediated transposition into a certain Enterobacterial plasmid with a possible natural competence from Kluyvera species [ 23 , 27 ]. The CTX-M-1 and -9 can be identified in multiple communities in this study ( Figure 9 A). Therefore, further dissemination of CTX-M-1 and -9 could be generated by plasmid transfer itself. In addition, the CTX-M-1 and -9 positive mobile element could transpose into other broad-host range plasmids, leading to wide dissemination of ESBL in Enterobacteriacea. On the other hand, CTX-M-2 subgroup is identified in only co68, suggesting that CTX-M-2 positive plasmid disseminates between Salmonella and Klebsiella species ( Figure 9 A). The carbapenemases K. pneumoniae carbapenemase (KPC) and New Delhi metallo-ß-lactamase (NDM-1) were unique to the co22 and co68 communities ( Figure 9 B). KPC was only found in co68, mainly in IncFIIk and IncN plasmids. NDM-1 was found in co22 and co68. Moreover, the NDM-1 in co22 was related only to the IncA/C plasmids. Meanwhile, the NDM-1 in co68 was not related to any Inc plasmids ( Table 4 and Figure 9 C). NDM-1 was closely related to co18/co22/co68 communities including plasmids carrying CTX-M and KPC. Thus, co22/co68 have the most potential for the dissemination of AMR genes in Enterobacteriacea. Figure 9 Components of the communities with AMR genes. ( A ) The CTX-M-1, -2 and -9 subgroups and the community and bacterial components. ( B ) KPC carbapenemase and the bacterial components. The plasmids carrying KPC are only found in co68. ( C ) The communities and bacterial components of the plasmids carrying the NDM-1 gene ( Table 4 ). pathogens-03-00356-t004_Table 4 Table 4 Distribution of Inc types, antimicrobial resistance genes, and communities. Inc type CTX-M KPC NDM-1 M-1 M-2 M-9 2 3 4 10 co22 co68 FII 7 8 1 FIIk 4 5 2 1 N 2 1 3 1 1 1 A/C 1 6 L/M 2 1 2 I1 3 FIB-M 1 1 FIA 1 FIIS 1 HI1 1 HI2 1 K 1 R 1 Not Assigned 5 4 4 4 1 1 12 Total 26 5 18 11 4 2 1 6 18"
} | 7,098 |
24149708 | PMC3917226 | pmc | 4,987 | {
"abstract": "Biofilm formation by the Gram-positive bacterium Bacillus subtilis is tightly controlled at the level of transcription. The biofilm contains specialized cell types that arise from controlled differentiation of the resident isogenic bacteria. DegU is a response regulator that controls several social behaviours exhibited by B. subtilis including swarming motility, biofilm formation and extracellular protease (exoprotease) production. Here, for the first time, we examine the prevalence and origin of exoprotease-producing cells within the biofilm. This was accomplished using single-cell analysis techniques including flow cytometry and fluorescence microscopy. We established that the number of exoprotease-producing cells increases as the biofilm matures. This is reflected by both an increase at the level of transcription and an increase in exoprotease activity over time. We go on to demonstrate that exoprotease-producing cells arise from more than one cell type, namely matrix-producing and non-matrix-producing cells. In toto these findings allow us to add exoprotease-producing cells to the list of specialized cell types that are derived during B. subtilis biofilm formation and furthermore the data highlight the plasticity in the origin of differentiated cells.",
"conclusion": "Concluding Remarks Here we have studied the prevalence and origin of exoprotease-producing cells in the developing B. subtilis biofilm. We have determined that the production of extracellular proteases is correlated with later stages of biofilm formation. This is perhaps due to a role in biofilm dispersal or nutrient acquisition and is consistent with the in situ localization of the exoprotease-producing cells at the biofilm–air interface. Note that this is the same region of the mature biofilm where developing spores are located and are therefore perhaps dispersed into the environment ( Branda et al. , 2001 ; Vlamakis et al. , 2008 ). Through the use of live single-cell fluorescence microscopy we have defined the origin of exoprotease-producing cells and established that there is not a strict dependence on the phenotype of the parental cell. We noted the development of the exoprotease-producing state from both a matrix OFF state and matrix ON state. Indeed, a matrix- and exoprotease-producing cell state can exist for an extended period of time, demonstrating that they are not mutually exclusive cell states ( Fig. 8 ). The biological significance of having a group of cells that can contribute to the production of the biofilm extracellular matrix and extracellular proteases remains to be elucidated. However, co-production of these molecules may indicate a need to increase nutrient acquisition from the extracellular environment when a sessile lifestyle is adopted. The single-cell analyses techniques used in this study clearly demonstrate the diversity of cell differentiation processes in the biofilm and indicate that, unlike matrix-producing cells that arise only from motile cells ( Vlamakis et al. , 2008 ), the origin of exoprotease-producing cells in the population is more flexible. In addition, as matrix-producing cells can transition into exoprotease-producing cells, it will be of interest in the future to determine if exoprotease-positive cells subsequently transition into spore formers. If correct, this would add an additional step to the cell fate lineage previously observed during biofilm development ( Vlamakis et al. , 2008 ).",
"introduction": "Introduction The formation of sessile communities of microbial cells called biofilms is a process common to many bacterial strains ( Costerton et al. , 1995 ). The resultant biofilm communities can have both beneficial and detrimental impacts on human society and are linked with processes as diverse as bioremediation and chronic infections ( Costerton et al. , 1987 ). Biofilm formation is underpinned by the production of an extracellular matrix that is commonly composed of DNA, proteins and exopolysaccharides ( Flemming & Wingender, 2010 ). The extracellular matrix generates and stabilizes the 3D structure and provides protection to the resident bacteria ( Branda et al. , 2005 ). Bacillus subtilis is a Gram-positive soil-dwelling bacterium used as a model for biofilm formation ( Vlamakis et al. , 2013 ). The biofilm matrix is composed of TasA amyloid-like fibres ( Branda et al. , 2006 ; Romero et al. , 2010 ), a secreted exopolysaccharide ( Branda et al. , 2001 ; Chai et al. , 2012 ) and a bacterial hydrophobin called BslA that forms a hydrophobic coat over the biofilm ( Hobley et al. , 2013 ; Kobayashi & Iwano, 2012 ; Ostrowski et al. , 2011 ). Synthesis of the components within the biofilm extracellular matrix is tightly regulated at the level of transcription ( Vlamakis et al. , 2013 ). In B. subtilis , one key regulator that is required for biofilm formation, due to its role in controlling the biosynthesis of the BslA coat protein, is DegU ( Kobayashi, 2007 ; Ostrowski et al. , 2011 ). DegU is the response regulator of the DegS–DegU two-component regulatory system ( Dahl et al. , 1991 ). Phosphorylated DegU (hereafter DegU-P)-regulated processes are upregulated in response to several environmental signals (for a review see Murray et al. , 2009a ) but of particular relevance to biofilm formation the system is activated when rotation of the flagella is impeded ( Cairns et al. , 2013 ). B. subtilis biofilm formation is hallmarked by the differentiation of genetically identical cells within the population into specialist subtypes ( Branda et al. , 2001 ; Vlamakis et al. , 2008 ). To date, cells specialized towards motility, biofilm matrix production and sporulation have been identified ( Vlamakis et al. , 2008 ). However, the occurrence and origin of cells that produce extracellular proteases (hereafter exoproteases) within the biofilm have not been examined. The production of exoproteases by B. subtilis occurs heterogeneously in planktonic culture ( Veening et al. , 2008 ) and transcription is dependent on DegU-P ( Dahl et al. , 1992 ; Tsukahara & Ogura, 2008 ). Moreover, exoprotease activity has been shown to be required for pellicle formation in laboratory isolates of B. subtilis ( Connelly et al. , 2004 ). Therefore, given this knowledge, and the known role for DegU-P in controlling biofilm formation ( Kobayashi, 2007 ; Stanley & Lazazzera, 2005 ; Verhamme et al. , 2007 ), here we define the impact of changing DegU-P levels on the proportion of cells in the biofilm population that transcribe the genes required for exoprotease synthesis. We identify that as biofilm formation progresses, exoprotease production increases at the level of both transcription and activity. Using live cell microscopy analysis of microcolony formation, we assess the origin of the exoprotease-producing cells and identify that they arise from both matrix-producing cells and non-matrix-producing cells. These findings shed light on the diversity of specialized cell types contained within the biofilm and highlight plasticity in their origin.",
"discussion": "Results and Discussion An increase in transcription of the exoprotease-encoding bpr gene is observed at the single-cell level in the presence of high DegU-P The two major exoproteases synthesized by B. subtilis are subtilisin and bacillopeptidase, which are encoded by the aprE and bpr genes, respectively ( Msadek, 1999 ). We followed exoprotease production using a P bpr – gfp reporter construct that was integrated at the native location on the chromosome ( Veening et al. , 2008 ). The P bpr – gfp construct was introduced into NCIB3610 wild-type and degU mutant strains ( Table 1 ) and transcription was monitored using flow cytometry and single-cell microscopy based on detection of GFP. Transcription from the bpr reporter fusion in the wild-type biofilm colony was assessed after 17 h of incubation. All the cells exhibited a low and homogeneous level of expression ( Fig. 1a ). Analysis confirmed that expression from the bpr promoter was DegU-dependent as the level of fluorescence decreased to the background basal level in the presence of a mutation in degU ( Fig. 1b ). These findings demonstrate that the transcriptional reporter behaved as expected during biofilm formation and in the NCIB3610 isolate of B. subtilis used here. Fig. 1. Transcription of the exoprotease-encoding bpr gene observed at the single cell level is dependent on DegU-P. Flow cytometry analysis of 3610 P bpr – gfp (NRS2315) (black line) (a) and 3610 P bpr – gfp degU (NRS2769) (black line) (b) using the parental 3610 strain as a negative control (grey-shaded area). Cells were grown under biofilm formation conditions for 17 h at 37 °C. A representative example is shown from three independent experiments. We next investigated the impact of increasing levels of DegU-P on transcription from the bpr promoter element to establish if this would promote heterogeneity in bpr transcription. To control the level of DegU-P in the cell, the degU32 hy mutant allele of the degU coding region was introduced into the chromosome under control of the IPTG-inducible promoter (P hy-spank using pDR111) at the heterologous amyE locus ( Verhamme et al. , 2007 ). This variant of DegU carries a histidine to leucine mutation in amino acid 12 and exhibits a lower level of dephosphorylation than wild-type DegU ( Dahl et al. , 1991 ). The strain additionally contained a deletion of the native copy of degU and carried a P bpr – gfp reporter fusion at the native bpr locus on the chromosome ( Table 1 ). Note that as transcription of degU32 hy is increased, biofilm formation itself is inhibited ( Verhamme et al. , 2007 ). Both flow cytometry and fluorescence microscopy analysis demonstrated that consistent with DegU-P activating transcription from the bpr promoter region the number of GFP-positive cells increased with the addition of up to 25 µM IPTG to the growth medium ( Fig. 2 ). The number of GFP-positive cells increased from 7 % in the absence of IPTG ( Fig. 2a, f, k ) to 84 % in the presence of 7.5 µM IPTG ( Fig. 2e, j, o ). The GFP-positive cells could be divided into low expression (10 1 –10 2 AU) and high expression (10 2 –10 4 AU) cells with the number of cells within the high expression category increasing alongside the level of DegU-P. Thus, it is evident that transcription from the bpr promoter can be highly heterogeneous in the biofilm population and that this is correlated with increases in the level of DegU-P. Fig. 2. An increase in transcription of the exoprotease-encoding bpr gene is observed at the single-cell level in the presence of high DegU-P. (a–e) Flow cytometry analysis. The grey-shaded area represents the parental 3610 strain as a negative control and the black lines the experimental sample. (f–o) Microscopy of P bpr – gfp degU , amyE : : P hy-spank-degU32 hy cells (NRS2771). (f–j) Fluorescence was imaged in the FITC channel to detect GFP production. (k–o) The same cells analysed by phase-contrast fluorescence microscopy showing the overlay with the GFP expression with the cells. The cells were grown under biofilm formation conditions for 17 h at 37 °C in the presence of 0 µM IPTG (a, f, k), 2.5 µM IPTG (b, g, l), 7.5 µM IPTG (c, h, m), 10 µM IPTG (d, i, n) and 25 µM IPTG (e, j, o) prior to collection. In each case, one representative example is presented from three independent experiments. The level of exoprotease production increases during biofilm formation We next assessed if changes in the frequency of exoprotease transcription occurred over time during biofilm formation. Increases in exoprotease production would indicate an increase in the level of DegU-P during biofilm formation. To test this, transcription from the P bpr – gfp reporter construct in the wild-type strain was assessed at the single-cell level from samples isolated from complex colony and pellicle biofilms over a time-course of development, namely 17, 24, 48 and 72 h for colony biofilms ( Fig. 3 ) and 24, 36, 48, 60, 72 and 96 h for pellicle biofilms ( Fig. 4 ). Representative pellicle biofilm images at the point of collection are presented as inserts within Fig. 4 . Flow cytometry and single-cell microscopy analysis of the disrupted biofilms revealed that transcription from the bpr promoter region increased during biofilm development for both biofilm types ( Figs 3 and 4 ); moreover, transcription became highly heterogeneous as the biofilm matured [compare Fig. 3a (i) with 3a (iv) and Fig. 4c with 4d, e and f ]. As reported above ( Fig. 2 ), the GFP-positive cells in the biofilm could be subdivided into low (10 1 –10 2 AU) and high (10 2 –10 3 AU) expressing cells with the number of individual cells within the high expression category increasing over time. In fact, the peak in exoprotease transcription occurred at 72 h for the colony biofilm and 96 h for the pellicle biofilm. These findings demonstrate that transcription from the bpr promoter increases over time during biofilm formation. Fig. 3. Transcription of bpr increases over time during biofilm formation. (a) Transcription of bpr from cells extracted from complex colonies was monitored over a 72 h period using a bpr – gfp transcriptional fusion strain (NRS2315). Colonies were grown at 30 °C and collected for flow cytometry analysis (grey-shaded area, non-fluorescent control 3610 strain; black line, bpr – gfp ) after 17 h (i), 24 h (ii), 48 h (iii) and 72 h (iv). (b) The same cells were then analysed by phase-contrast fluorescence microscopy after 17 h (i), 24 h (ii), 48 h (iii) and 72 h (iv). Shown are the phase-contrast (top) and FITC (GFP) channel (middle) and an overlay of both channels (bottom). A representative example is presented from three independent experiments. Fig. 4. An increase in transcription of the exoprotease-encoding bpr gene occurs during pellicle formation. Transcription of bpr from cells extracted from pellicle biofilms was monitored over a 96 h period using a bpr – gfp transcriptional fusion strain (NRS2315). Transcription of bpr from cells extracted from pellicles grown at 23 °C was monitored over a 96 h period using a bpr – gfp transcriptional fusion. Flow cytometry data are shown from cells extracted from pellicles at 24 h (a), 36 h (b), 48 h (c), 60 h (d), 72 h (e) and 96 h (f). The non-fluorescent NCIB3610 control line is shown as the grey-shaded area and the experimental sample as a black line. A representative example of both the expression analysis and pellicle formation is presented from three independent experiments. To correlate bpr transcription with exoprotease production, the level of active extracellular proteases in the pellicle biofilm supernatant was quantified biochemically ( Fig. 5 ). Consistent with the increase in bpr transcription that was observed in more mature biofilms, the level of extracellular proteases in the pellicle supernatant fraction increased during biofilm formation as measured using caesin digestion assays (see Methods; compare Fig. 3 with Fig. 5 ). At 96 h, exoprotease activity in the extracellular environment was four fold higher than that quantified for the 24 h biofilm ( Fig. 5 ). These biochemical analyses link increases in transcription from the P bpr – gfp reporter with increased exoprotease activity levels in the biofilm community. Fig. 5. Protease secretion was analysed from supernatant collected from pellicles during biofilm formation. Protease secretion was assessed using an azocasein assay from pellicle supernatants. Pellicles were grown at 25 °C for up to 96 h and samples were collected at 24 h (day 1), 48 h (day 2), 72 h (day 3) and 96 h (day 4). Enzyme activity was normalized against wet pellet weight of the pellicle. Data are presented as the mean of four independent experiments and the error bars represent sem . Exoprotease-producing cells are located at the surface of the mature biofilm Flow cytometry and single-cell microscopy analysis of disrupted biofilms allow quantification of the population that express the P bpr – gfp reporter ( Figs 1 – 4 ). However, these techniques do not allow analysis of the spatial localization of transcription in the biofilm ( Vlamakis et al. , 2008 ). To determine the spatial localization of the exoprotease-producing cells in the biofilm, 48 h colony biofilms formed by strain NRS3921 were cryosectioned as described previously [see Methods and Vlamakis et al. (2008) ]. Following fixation, the thin layer cross-sections of the biofilm were imaged by confocal microscopy and expression from the P bpr – gfp reporter was detected. These microscopy analyses confirmed the flow cytometry analysis presented in Figs 3 and 4 , which demonstrated that a subpopulation of the cells was highly fluorescent (indicative of high levels of P bpr – gfp transcription). Retention of the biofilm structure allowed us to determine that this group of cells was found towards the top of the colony biofilm near the air–biofilm interface ( Fig. 6 ). The subpopulation of cells with low levels of GFP was found towards the centre of the biofilm section. It was highly apparent that transcription of the reporter fusion was heterogeneous within the biofilm and that there was structure in the transcription profile with respect to the organization of the mature biofilm. Fig. 6. Spatial analysis of bpr transcription within the mature biofilm. (a) Bright-field and FITC merged image and (b) FITC image of a 9 µm vertical cross-section of a 48 h colony biofilm harbouring the P bpr – gfp reporter fusion (NRS3921) detected by confocal microscopy. The top (air interface) and bottom (agar interface) of the colony biofilm are indicated for reference purposes and shown as white lines. In (a), a 48 h colony biofilm is shown and the white box represents the approximate region of the biofilm that was imaged in cross-section. Bars, 10 µm. The images shown are representative of at least three independent experiments during which multiple fields of view were examined. The protease-producing cell population overlaps with the matrix-producing cell population The analysis reported above allows us to add exoprotease-producing cells to the list of specialized cell types that are found in the developing B. subtilis biofilm ( Branda et al. , 2001 ; Vlamakis et al. , 2008 ). Previously characterized cell types include cells that are motile, cells that transcribe the eps and tapA operons needed for biofilm matrix assembly (hereafter matrix-producing cells), and cells that are sporulating ( Vlamakis et al. , 2008 ). It has previously been established that motile cells transition into matrix cells and that the matrix-producing cells progress to form endospores at late biofilm stages ( Vlamakis et al. , 2008 ). In addition, it has been proposed that matrix production and protease production are mutually exclusive events and that both cell types arise directly from motile cells in response to different environmental signals ( Lopez et al. , 2009 ). This has not been tested experimentally and is somewhat at odds with the knowledge that matrix production decreases during biofilm formation ( Vlamakis et al. , 2008 ) while exoprotease production increases ( Figs 3 and 4 ) at a time during biofilm formation when motile cells are absent from the biofilm ( Vlamakis et al. , 2008 ). Therefore, to define the origin of the exoprotease-producing cells and investigate the relationship between exoprotease production and matrix production, we constructed a dual reporter strain which carried the P bpr – gfp fusion at the native locus and a P tapA – mCherry transcriptional fusion at the heterologous lacA locus (NRS3378). We examined the prevalence of cells that co-expressed both fusions as indicated by fluorescence in both the FITC (P bpr – gfp ) and TRITC (P tapA–mCherry ) channels in cells extracted from 24 h pellicle biofilms. As indicated in Fig. 7(a) (asterisks), co-expression from the tapA and bpr promoter regions was clearly observed within the cells that are false coloured yellow. We next examined expression from each promoter by fluorescence microscopy over a time-course of biofilm formation using cell samples that were extracted from complex colonies ( Fig. 7b ). The parental strain NCIB3610 was used as a control for microscopy (data not shown). Our analysis demonstrated, as expected, that the proportion of matrix-producing cells was high at early time points of biofilm formation and was lower in the later stages of biofilm development ( Fig. 7b , compare 14 h with 72 h) ( Vlamakis et al. , 2008 ). Moreover, as described above, the proportion of cells in the biofilm that had transcribed the P bpr–gfp reporter fusion increased over time ( Fig. 7b ). Thus, each transcriptional fusion behaved as expected in the dual reporter fusion strain NRS3378 and the findings suggest that matrix production and exoprotease production are not necessarily mutually exclusive cell states. However, note that the co-expressing cells could represent a transition between one cell state and another or possibly apparent co-expression that is a reflection of the stability of the fluorescent reporter fusions. Fig. 7. Co-expression of the bpr and tapA genes. (a, b) Static microscopy of NRS3378 cells harbouring the P tapA – mCherry and P bpr – gfp transcriptional reporter fusions extracted from a pellicle biofilm after 24 h growth at 37 °C (a), where the asterisks indicate selected cells for which fluorescence was detected in both the TRITC (false coloured red) and the FITC (false coloured green) channels, for colony biofilms grown at 37 °C for the time (hours) indicated in the upper left-hand corner (b). Bars, 5 µm; the images are representative of multiple fields of view. (c) Microscopy analysis of NRS3921 harbouring P tapA – mKate2 and P bpr – gfp transcriptional reporter constructs in real-time during microcolony development at 30 °C. Strain NRS3921 was imaged every 15 min. Images from the DIC, FITC (false coloured green) and TRITC (false coloured red) channels are shown above. The time (minutes) is indicated in the upper left-hand corner. Bars, 10 µm. Matrix-producing cells can transition into exoprotease-producing cells To trace the origin of exoprotease-producing cells and to investigate the relationship between exoprotease production and matrix production in greater detail, we performed real-time fluorescence single-cell microscopy analysis in developing microcolonies ( de Jong et al. , 2011 ; Young et al. , 2012 ). Our initial analysis highlighted that the mCherry fluorescent protein was not a suitable reporter protein for the live cell microscopy analysis. Live cell microscopy demands multiple images to be taken and the fluorescence from mCherry was found to be susceptible to rapid photo-bleaching (data not shown). Therefore, strain NRS3921 was constructed where the mCherry reporter was replaced with mKate2, yielding a strain that carried the P tapA – mKate2 and P bpr – gfp reporter fusions ( Table 1 ). The strain was grown in microscope chambers for up to 13 h, with images acquired every 15 min (see Methods). As expected, P tapA – mKate2 matrix gene expression was bimodal in the developing microcolony and P bpr – gfp exoprotease gene expression was heterogeneous in the population ( Fig. 7c ). Moreover, consistent with microscopy and flow cytometry analysis from the time-course of biofilm formation ( Figs 3 and 4 ), transcription from the P bpr – gfp reporter fusion was observed more frequently at later time points in microcolony development (compare 570 with 750 min time points in Fig. 7c ). The data collected from the live cell imaging were used to trace the origin of matrix-positive cells over several cell cycles. To achieve this we followed multiple cells during division, noting the phenotype as indicated by expression from the reporter fusions. We established that the majority of exoprotease-producing cells arose from cells that had persisted in a non-matrix-expressing state for more than one generation ( Figs 7c and 8 , cell highlighted by the green arrowheads, and data not shown). However, we established that exoprotease-producing cells were not precluded from arising directly from matrix-producing cells as the transition of a matrix-producing cell into an exoprotease producer was frequently detected ( Fig. 8 ). This is exemplified in Fig. 8 where the white and yellow arrowheads on the micrographs highlight two cells that transition directly from matrix production to exoprotease production over time. In addition, it was observed that exoprotease-producing cells were (infrequently) capable of transitioning back to matrix-transcribing cells. This is demonstrated in Fig. 8 by the blue arrowheads. These findings demonstrate that exoprotease production and tapA matrix gene expression are not incompatible events and can exist for a sustained period of time within one cell. Fig. 8. Origin of exoprotease-producing cells. Microscopy analysis of NRS3921 harbouring P tapA – mKate2 and P bpr – gfp transcriptional reporter constructs during microcolony development at 30 °C. Strain NRS3921 was imaged every 15 min. Images from the DIC, FITC (false coloured green) and TRITC (false coloured red) channels are shown. The white and yellow arrowheads indicate cells that have transitioned directly from matrix production to exoprotease production. The green arrowheads indicate cells that have transitioned directly from a non-fluorescent state to exoprotease production. The blue arrowheads indicate cells that have transitioned from matrix-producing cells to exoprotease production and back to matrix production. The time (minutes) is indicated in the upper right-hand corners. Bars, 5 µm. Concluding Remarks Here we have studied the prevalence and origin of exoprotease-producing cells in the developing B. subtilis biofilm. We have determined that the production of extracellular proteases is correlated with later stages of biofilm formation. This is perhaps due to a role in biofilm dispersal or nutrient acquisition and is consistent with the in situ localization of the exoprotease-producing cells at the biofilm–air interface. Note that this is the same region of the mature biofilm where developing spores are located and are therefore perhaps dispersed into the environment ( Branda et al. , 2001 ; Vlamakis et al. , 2008 ). Through the use of live single-cell fluorescence microscopy we have defined the origin of exoprotease-producing cells and established that there is not a strict dependence on the phenotype of the parental cell. We noted the development of the exoprotease-producing state from both a matrix OFF state and matrix ON state. Indeed, a matrix- and exoprotease-producing cell state can exist for an extended period of time, demonstrating that they are not mutually exclusive cell states ( Fig. 8 ). The biological significance of having a group of cells that can contribute to the production of the biofilm extracellular matrix and extracellular proteases remains to be elucidated. However, co-production of these molecules may indicate a need to increase nutrient acquisition from the extracellular environment when a sessile lifestyle is adopted. The single-cell analyses techniques used in this study clearly demonstrate the diversity of cell differentiation processes in the biofilm and indicate that, unlike matrix-producing cells that arise only from motile cells ( Vlamakis et al. , 2008 ), the origin of exoprotease-producing cells in the population is more flexible. In addition, as matrix-producing cells can transition into exoprotease-producing cells, it will be of interest in the future to determine if exoprotease-positive cells subsequently transition into spore formers. If correct, this would add an additional step to the cell fate lineage previously observed during biofilm development ( Vlamakis et al. , 2008 )."
} | 7,015 |
23087494 | null | s2 | 4,991 | {
"abstract": "Many biological and physical systems exhibit population-density dependent transitions to synchronized oscillations in a process often termed \"dynamical quorum sensing\". Synchronization frequently arises through chemical communication via signaling molecules distributed through an external medium. We study a simple theoretical model for dynamical quorum sensing: a heterogenous population of limit-cycle oscillators diffusively coupled through a common medium. We show that this model exhibits a rich phase diagram with four qualitatively distinct physical mechanisms that can lead to a loss of coherent population-level oscillations, including a novel mechanism arising from effective time-delays introduced by the external medium. We derive a single pair of analytic equations that allow us to calculate phase boundaries as a function of population density and show that the model reproduces many of the qualitative features of recent experiments on BZ catalytic particles as well as synthetically engineered bacteria."
} | 255 |
29751652 | PMC5982004 | pmc | 4,993 | {
"abstract": "Long-term unregulated mining of ion-adsorption clays (IAC) in China has resulted in severe ecological destruction and created large areas of wasteland in dire need of rehabilitation. Soil amendment and revegetation are two important means of rehabilitation of IAC mining wasteland. In this study, we used sludge biochar prepared by pyrolysis of municipal sewage sludge as a soil ameliorant, selected alfalfa as a revegetation plant, and conducted pot trials in a climate-controlled chamber. We investigated the effects of alfalfa revegetation, sludge biochar amendment, and their combined amendment on soil physicochemical properties in soil from an IAC mining wasteland as well as the impact of sludge biochar on plant growth. At the same time, we also assessed the impacts of these amendments on the soil microbial community by means of the Illumina Miseq sequences method. Results showed that alfalfa revegetation and sludge biochar both improved soil physicochemical properties and microbial community structure. When alfalfa revegetation and sludge biochar amendment were combined, we detected additive effects on the improvement of soil physicochemical properties as well as increases in the richness and diversity of bacterial and fungal communities. Redundancy analyses suggested that alfalfa revegetation and sludge biochar amendment significantly affected soil microbial community structure. Critical environmental factors consisted of soil available K, pH, organic matter, carbon–nitrogen ratio, bulk density, and total porosity. Sludge biochar amendment significantly promoted the growth of alfalfa and changed its root morphology. Combining alfalfa the revegetation with sludge biochar amendment may serve to not only achieve the revegetation of IAC mining wasteland, but also address the challenge of municipal sludge disposal by making the waste profitable.",
"conclusion": "4. Conclusions Alfalfa revegetation and sludge biochar amendment both improved soil physicochemical properties and enhanced the diversity and richness of the microbial community. In addition, the combined treatment (soil amended with alfalfa revegetation and biochar) resulted in the greatest improvement of soil physicochemical properties, the enhancement of diversity and richness of microbial community, and the promotion of plant growth. Redundancy analyses showed that soil physicochemical properties could explain 98.9% and 99% of the variation in bacterial and fungal community structure, respectively, and soil available potassium, pH, organic matter, C/N ratio, bulk density, and total porosity were the critical environmental factors affecting soil microbiota. Moreover, sludge biochar could be used to promote the growth of alfalfa and change their root morphology, which in turn accelerated the soil rehabilitation process of IAC mining wastelands. In this way, the combined amendment of alfalfa revegetation and sludge biochar amendment not only serve as soil remediation for IAC mining wastelands but also resolve the difficult problem of municipal sludge disposal by making the waste profitable. Thus, a combined strategy is recommended to achieve sustainable soil restoration for IAC mining wastelands.",
"introduction": "1. Introduction Ion-adsorption clay (IAC) contains rare earth elements (REEs). IAC mines are widely distributed throughout several adjacent provinces of southern China, including Jiangxi, Guangdong, Fujian, Guangxi, Hunan, Yunnan, Zhejiang provinces [ 1 ]. In the early days of mining, IAC was extracted via tank or heap leaching [ 2 ]. These methods cause serious ecological destruction in IAC mining wastelands [ 3 ], including loss of vegetation, pollution of water and soil, and geological disasters (e.g., landslides). As a result, the soil in IAC mining wastelands has a loose texture, poor aggregation, low water-holding capacity and fertility, and decreased microbial diversity, all of which makes it hard for plants to colonize these soils. Rehabilitation is urgently needed for IAC mining wastelands and two potential measures are soil amendment and revegetation. A key step for soil amendment is the selection of an appropriate soil ameliorant. Biochar, which is prepared by slow pyrolysis of biomass under oxygen-limited conditions, has been a focus of research on soil amendments [ 4 , 5 , 6 , 7 , 8 , 9 , 10 ]. The characteristics of biochar determine how it could improve soil properties. Its porous structure can increase soil porosity [ 11 , 12 ], reduce soil bulk density [ 13 ], and provide a habitat for microorganisms [ 14 ]. Furthermore, the huge surface area [ 15 ] and abundant functional groups (e.g., carboxyls and phenolic hydroxyls) [ 16 ] could enhance soil cation exchange capacity (CEC), increase water-holding capacity and decrease fertilizer leaching [ 17 ]. The aromatic hydrocarbon structure contributes to the long-term retention of biochar in soil [ 18 ]. Biochar can be prepared from a wide range of raw materials, such as agricultural waste [ 5 ], animal manure [ 19 ], and municipal sludge [ 20 ]. It has been proposed that municipal sewage sludge may be an important raw material for biochar preparation because it is rich in mineral nutrients (N, P and K) and organic matter [ 21 ]. China’s yield of municipal sludge has been growing rapidly, with an average increase of more than 10% each year from 2008 to 2014 and an estimated 34 million tons produced in 2015 [ 22 ]. More than 20% of municipal sludge, which contains pathogens, heavy metals and other pollutants, remains stacked on land on the outskirts of cities [ 23 ] where it could threaten human health by entering food or water supplies [ 24 ]. Currently, the disposal of municipal sludge is an urgent problem for the government and society. Traditional methods of municipal sludge disposal include agricultural applications, incineration and landfill disposal, all of which have drawbacks, including the fact that these methods require a lot of land resources and result in air pollution, and heavy-metal pollution in soil and water. Pyrolysis is an effective way to dispose of sludge, which is transformed into sludge biochar and used as a soil-amending resource. In addition, sludge biochar can prevent the leaching of heavy metals in raw sewage sludge [ 25 ]. Many studies have reported the application of sludge biochar in the remediation of different kinds of sites. Méndez et al. [ 26 ] evaluated the effects of sludge biochar derived from sewage sludge on heavy metals solubility and bioavailability in a Mediterranean agricultural soil; the result showed that the risk of leaching of Cu, Ni and Zn were lower in the soil treated with sludge biochar, which also reduced plant availability of Ni Zn, Cd and Pb when compared with soil treated with raw sewage sludge. Sardar Khan. et al. [ 18 ] investigated the impact of sludge biochar upon rice ( Oryza sativa L.) yield, metal bioaccumulation and greenhouse gas emissions from acidic paddy soil, and concluded that sludge biochar increased soil pH, total nitrogen, soil organic carbon and available nutrients and decreased bioavailable As, Cr, Co, Ni, and Pb in soil as well as significantly ( p ≤ 0.01) increasing shoot biomass, grain yield and the bioaccumulation of phosphorus and sodium. Méndez et al. [ 27 ] assessed the influence of sewage sludge and sewage sludge biochar on peat properties as growing media and on lettuce ( Lactuca sativa ) growth, and they confirmed that sewage sludge transformation into biochar proved to be a sustainable waste management approach in order to promote their future use as growing media components. To our knowledge, little is known about the effect of the use of sludge biochar on the soil properties of IAC mining wastelands. Revegetation is an important measure used for rehabilitation and a critical step in the use of this measure is the best choice of revegetation plants for a particular site. Alfalfa is an attractive revegetation plant, an important gramineous forage legume, because of its adaptability to climates and soil environments, rapid growth, high yield, and ability to fix nitrogen [ 28 , 29 ]. Alfalfa is widely used in the remediation of heavy metals, oil, and other contaminated soils as well as to ameliorate the effects of degraded soil [ 30 , 31 ]. Alfalfa revegetation has not been tested for soil restoration of IAC mining wastelands. Microorganisms in soil play a critical role in cycling materials. The sensitivity of the response of a microbial community, specifically bacterial and fungal communities, reflects a change in soil quality [ 32 , 33 ]. For instance, a significant alteration of microbial community in IAC mining wasteland was demonstrated in our previous study [ 34 ] and some other reports [ 35 , 36 , 37 ]. In this study, we used sludge biochar as a soil ameliorant and alfalfa as the revegetation plant and conducted pot trials in a climate-controlled chamber, and we test the hypothesis that sludge biochar amendment and alfalfa plantation can ameliorate physicochemical properties and increase the microbial diversity of soil from IAC mining wastelands. Our objectives were to determine the effects of alfalfa revegetation and municipal sludge biochar amendment, and both in combination, on the physicochemical properties and microbial communities of soil from IAC mining wastelands. We aimed to not only explore a new way to utilize municipal sludge as a resource, but also to test a new rehabilitation method for IAC mining wastelands.",
"discussion": "3. Results and Discussion 3.1. Response of Soil Physicochemical Properties to Alfalfa Revegetation and Sludge Biochar Amendment 3.1.1. Effects on Soil Physical Properties The soil physical properties of four treatments are shown in Figure 1 . There were significant differences among four treatments. The specific gravity values of G (2.55 ± 0.02), SBC (2.62 ± 0.02), and SBCG (2.44 ± 0.02) treatments were all significantly lower than that of the CK treatment (2.70 ± 0.03) ( p < 0.05). A similar trend was also observed in the bulk density of four treatments, with those of G, SBC, and SBCG significantly lower than that of CK ( p < 0.05) by 9.1%, 5.6% and 14.7%, respectively. The lower levels of specific gravity and bulk density in treated soils (G, SBC, and SBCG) suggested that sludge biochar application and alfalfa revegetation both significantly reduced soil compaction and enhanced soil porosity and ventilation, which are attributed to the porous structure of the sludge biochar [ 18 ] and the strong root system of alfalfa [ 48 , 49 ]. Total porosity only showed significant differences between CK and SBCG ( p < 0.05). The soil water-holding capacity values of G, SBC, and SBCG were all significantly higher, by 20.2%, 10.2%, and 31.4% ( p < 0.05), respectively, than that of CK (21.59 ± 1.52%). A previous study [ 37 ] showed that long-term exploitation of rare earth resources had destroyed soil structure (poor porosity and ventilation) in an IAC mining wasteland. The significant increases of total porosity and water-holding capacity in SBCG treatment suggested that the combination of alfalfa revegetation and sludge biochar amendment not only improved soil porosity but also helped maintain water content. Many previous studies have demonstrated that the application of biochar can improve the soil physical properties [ 50 ], such as soil aggregation [ 51 ], water-retention capacity [ 52 , 53 ], pore-size distribution [ 54 ], and bulk density [ 55 ]. The application of sludge biochar, by means of providing nutrients (N, P, K) [ 56 ], can benefit the growth of alfalfa, which in turn helps ameliorate the soil’s physical properties [ 57 ]. An additive effect between alfalfa revegetation and sludge biochar amendment was detected first in this study, suggesting the combination of these two variables improved soil physical properties was more intense than either in isolation, which was in accordance with other previous study [ 58 ]. 3.1.2. Effect on Soil Chemical Properties The soil chemical properties of the four treatments, including soil pH, electrical conductivity, organic matter, total nitrogen, C/N, available nitrogen, available phosphorus and available potassium, are shown in Figure 2 . The SBC and SBCG treatments had significantly higher pH than CK ( p < 0.05), indicating the significantly acidic decrease in soils from IAC mining wastelands for sludge biochar application ( p < 0.05). The effect of sludge biochar on soil pH was attributed to the relatively high pH (6.17 ± 0.03), which is higher than that of municipal sludge (5.64 ± 0.18) but lower than in most biochars (>8.0) [ 59 , 60 ]. In the slow pyrolysis process, many cations (such as Ca 2+ , Mg 2+ , K + and Na + ) form carbonates or oxides, and can reduce soil acidity [ 61 ] when sludge biochar is applied to soil, which is consistent with the effects of biochar on strongly acidic soils [ 62 , 63 ]. EC, organic matter, total nitrogen, available nitrogen, and available phosphorus in treated soil (G, SBC, and SBCG) increased significantly, by 0.8–2.6 times, 78.6%–6.82 times, 1.9–6.8 times, 4.3–13.8 times, and 0.5–27.8 times, respectively, over CK values, and the highest values were in the SBCG treatment. The difference in C/N ratio between SBCG (11.03 ± 0.81) and CK (3.77 ± 1.17) was significant ( p < 0.05) but that between G and SBC was not ( p > 0.05). Available potassium increased from 9.91 ± 0.42 in CK to 86.30 ± 0.86, 79.55 ± 1.17 and 110.79 ± 0.44 in the G, SBC, and SBCG treatments, respectively. This suggested that alfalfa revegetation and sludge biochar amendment significantly increased the contents of soil organic matter, nutrients and their availability (available nitrogen, available phosphorus and available potassium levels) ( p < 0.05). Our results are consistent with the effects of biochar use and alfalfa revegetation on soil chemical properties from other reports. Many studies have demonstrated that the use of biochar can significantly increase soil organic C [ 64 , 65 ], mineral nutrient content (e.g., N, P, K) [ 66 ] and the cation exchange capacity of soil [ 18 ] for its carbon-rich and porous structure [ 56 ] and huge specific surface area [ 67 ]. Increasing quantities of data have shown that alfalfa revegetation can significantly increase soil soil organic carbon and total nitrogen concentrations present in the root ball [ 68 ], C/N ratio [ 57 ], porosity, aeration conditions, and water-holding capacity [ 49 ], all of which could accelerate the mineralization of organic carbon and the release of mineral nutrients from sludge biochar, and in turn promote the alfalfa growth. In the present study, an additive effect on soil chemical properties, like that of soil physical properties, was also detected when combined sludge biochar amendment and alfalfa revegetation ( Figure 2 ). 3.2. Response of Soil Microbial Community to Alfalfa Revegetation and Sludge Biochar Amendment 3.2.1. Soil Microbial Alpha Diversity After de-multiplexing and quality filtering, 468,000 and 360,000 high-quality bacterial and fungal sequences for each treatment were obtained for further analysis, and the results were shown in Table 3 . The mean length of bacterial and fungal sequences of all samples was 229.99 and 208.67 bp, respectively. The number of bacterial and fungal OTUs as well as ACE, Chao1 and Shannon indices among the four treatments were CK < SBC < G < SBCG, whereas the Simpson index showed an opposite trend: CK > SBC > G > SBCG. This indicated that alfalfa revegetation and sludge biochar amendment significantly increased soil microbial diversity in IAC mining wastelands ( p < 0.05), with an additive effect generated by their combined amendment. The results were in accordance with previous studies in biochar-enriched Terra preta soils [ 69 , 70 ] and alfalfa amended soils [ 71 , 72 ]. This increased microbial diversity might be due to the supply of a more comfortable habitats, many nutrients or mineralized organic matter from sludge biochar [ 6 , 73 ] and root exudates of alfalfa [ 74 ]. Across all soil samples, the gene coverage was more than 0.98, indicating that the sequencing results were representative. 3.2.2. Soil Microbial Community Structure We detected 5927 OTUs in bacterial communities across all four treatments, for a total of 24 phyla, 48 classes, 97 orders, 236 families, and 737 genera. As shown in Figure 3 a, the bacterial communities at the phylum level varied among the four treatments. The relative abundances of dominant bacteria ranked as follows: Proteobacteria > Actinobacteria > Firmicutes > Bacteroidetes > Acidobacteria > Verrucomicrobia > Planctomycetes . These seven phyla accounted for more than 97% gene sequences of soil in each treatment, with the respective proportion of 58.96%, 14.10%, 10.31%, 8.73%, 2.53%, 2.03%, and 0.88% in total. Alfalfa revegetation (G), sludge biochar amendment (SBC), and the combined amendment (SBCG) significantly reduced the relative abundances of Proteobacteria , and Firmicutes in soil of IAC mining wastelands ( p < 0.05), whereas significantly increased abundances of Actinobacteria , Bacteroidetes , and Verrucomicrobia ( p < 0.05). The results were in agreement with other reports showing that Proteobacteria , Actinobacteria and Firmicutes were the main phyla in biochar [ 75 , 76 , 77 ] and alfalfa [ 78 , 79 ] amended soil, with the relative abundance of Proteobacteria decreased in biochar treated soil [ 77 ]. In addition, Actinobacteria was demonstrated to be the representative specie in recalcitrant carbon-rich soils like Terra preta [ 70 ] and pyrogenic carbon-treated soils [ 80 ], which explained the increased abundance of Actinobacteria in our study. We detected 2974 OTUs in fungal community across all four treatments, for a total of 6 phyla, 24 classes, 72 orders, 154 families, 280 genera, and 368 species. The relative abundance of fungal community at the phylum level varied among the four treatments ( Figure 3 b). The relative abundances of dominant fungi ranked as follows: Ascomycota > Basidiomycota > Zygomycota > Chytridiomycota > Glomeromycota. The five phyla accounted for more than 50% gene sequences of soil in each treatment, with the respective proportion of 50.14%, 1.72%, 0.22%, 0.05% and 0.005% in total. The results were quite similar to the findings in a previous study [ 81 ], which showed that Ascomycota , Basidiomycota , and Zygomycota were the dominant fungal phyla across all soil samples while Chytridiomycota and Glomeromycota were minor phyla. Alfalfa revegetation (G), sludge biochar amendment (SBC), and the combined amendment (SBCG) significantly elevated the relative abundance of Ascomycota in the soil of IAC mining wastelands ( p < 0.05), and significantly reduced that of Basidiomycota ( p < 0.05). At the genus level, there was a total of 737 bacterial genera ( Figure 4 a) detected in four treatments, with different patterns of dominance. Alfalfa revegetation, sludge biochar amendment, and the combined amendment significantly elevated the relative abundances of Arthrobacter, Burkholderia. Devosia, Edaphobacter, Leifsonia, Massilia, Mucilaginibacter, Sinomonas, Sphingomonas and Stenotrophomonas in soil ( p < 0.05), whereas significantly reduced abundances of Exiguobacterium , Citrobacter , Pseudomonas, and Bradyhizobium ( p < 0.05). Arthrobacter, an gram-negative bacterium, was demonstrated to have the ability to degrade hydrocarbons and its’ higher relative abundance was detected in biochar [ 82 , 83 ] and alfalfa-amended soil [ 84 ]. Sphingobium was reported to increase in soil treated with biochar, and can degrade recalcitrant compounds [ 85 ]. A total of 280 fungal genera, including 368 species, were detected across four treatments, with different patterns of dominance ( Figure 4 b). Alfalfa revegetation, sludge biochar amendment, and the combined amendment significantly enhanced the relative abundances of unclassified _Nectriaceae , unclassified _Sordariomycetes , Penicillium , Humicola, unclassified _Chaetomiaceae , and Myrothecium ( p < 0.05), whereas significantly decreased that of Aspergillus , unclassified _Capnodiales , unclassified _Agaricostilbaceae , Clonostachys , unclassified_ Ascomycota , unclassified_ Pleosporaceae, and unclassified_ Davidiellaceae ( p < 0.05). Our results were in accord with the relative abundance change of fungal community in biochar [ 76 , 86 , 87 ] and alfalfa [ 79 , 88 ] amended soil. PCoA ( Figure 5 ) and hierarchical cluster heat-map analysis ( Figure 6 ) revealed the similarity and differences of bacterial and fungal communities in the four treatments at the genus level. The results of PCoA showed that PCoA 1, PCoA 2, and PCoA 3 explained 82.9%, 10.6%, and 2.9%, respectively, as well as 57.9%, 26.7%, and 9%, respectively, differences of bacterial ( Figure 5 a) and fungal ( Figure 5 b) community structure in soil. Heat-map analysis suggested that for bacterial communities ( Figure 6 a), CK treatment alone formed its own cluster and the three amendment treatments (G, SBC, and SBCG) formed a separate cluster, indicating the similarity of amendment treatments (G, SBC, and SBCG) and the differences between amended (G, SBC, and SBCG) and un-amended (CK) treatments. For fungal communities ( Figure 6 b), CK and G were clustered together while SBC and SBCG were clustered together, indicating that sludge biochar amendment had a greater effect on soil fungal community than alfalfa revegetation. Our results suggested the significant change of bacterial and fungal communities’ compositions for sludge biochar application and alfalfa revegetation, which were in agreement with previous studies. Organic amendments were demonstrated to be the most important means of managing soil biodiversity, and their quantity, quality, and distribution each affected the trophic structure of the soil food web [ 89 , 90 ]. Sludge biochar and root exudates of alfalfa both provided organic matter for soil microbes’ growth. Generally, the amount of soil organic matter from sludge biochar was larger than that from root exudates of alfalfa, which were in agreement with the much higher organic matter value of SBC than of G ( Figure 2 ). Additionally, fungi were known to be saprophytes, associated with degradable soil organic matter [ 91 ]. Normally, soil organic matter was successively utilized by bacteria and fungi. All the aforementioned could explain the bacterial difference between control (CK) and treated soils (G, SBC, SBCG) as well as the fungal difference between sludge biochar amended (SBC and SBCG) and un-amended soils (CK and G). 3.3. The Complex Relationship between Soil Physicochemical Properties and Microbial Communities, and the Response of Plant Growth to Remediation 3.3.1. Redundancy Analyses of Soil Physicochemical Properties and Microbial Community Redundancy analyses between soil properties and bacterial and fungal community structure were conducted and a bi-plot is shown in Figure 7 . The investigated soil physicochemical properties could explain 98.9% and 99.0% of the variation of bacterial and fungal community structure ( Table 4 ), respectively. All soil physicochemical properties (except available phosphorus) had a significant impact on bacterial community structure ( p < 0.05). The effects of available potassium, pH, C/N, bulk density, water-holding capacity, specific gravity, EC, total nitrogen, and total porosity were significant ( p < 0.01). All soil physicochemical properties had a significant influence on fungal community ( p < 0.05). The effects of pH, available potassium, C/N, EC, available nitrogen, total nitrogen, organic matter, and bulk density were significant ( p < 0.01). The variation in the bacterial community explained by the soil physicochemical properties decreased as follows: available potassium > pH > C/N > bulk density > water-holding capacity > specific gravity > EC > available nitrogen > total nitrogen > organic matter > total porosity > available phosphorus. Furthermore, available potassium, pH, C/N, organic matter, bulk density, and total porosity could be used to explain 93.6% variation of the bacterial community data, and were confirmed to be the key environmental factors. The variation in fungal community explained by soil physicochemical properties decreased as follows: pH >available potassium > C/N > EC > available nitrogen > total nitrogen > organic matter > bulk density > water-holding capacity > specific gravity > total porosity > available phosphorus. Moreover, available potassium, pH, organic matter, C/N, bulk density, and total porosity could be used to explain 88.0% variation of fungal community data, and were confirmed to be the critical environmental factors. In general, the key environmental factors affecting bacterial community were similar to those affecting the fungal community. Our results suggested that alfalfa revegetation could enhance soil microbial community diversity and richness of IAC mining wastelands, which are similar to the results of Chen [ 92 ]. Alfalfa revegetation significantly improved soil physicochemical properties, including enhancement of soil porosity, water-holding capacity and content of organic matter and nutrients (e.g., N, P, and K), decrease of soil bulk density, and amelioration of soil structure, all of which were helpful to the proliferation of soil microorganisms (bacteria and fungi). Sludge biochar amendment also enhanced the diversity and richness of the microbial community in IAC mining wasteland soil. This is because biochar not only provides several types of nutrients (C, N, and other trace elements) for soil microbial growth [ 93 ], but will also ameliorate the soil environment for microbial proliferation, by for example reducing soil acidity, elevating soil porosity, and improving soil aeration conditions [ 94 ]. These claims are consistent with the results of biochar on microbiota in acidic soil [ 76 , 95 ]. Additionally, due to its porous structure, large specific surface area and cation exchange capacity, biochar has the ability to retain nutrients [ 96 ] and would provide them to the soil after oxidation. The combination of alfalfa revegetation and sludge biochar amendment further enhanced the diversity and richness of the soil microbial community in IAC mining wastelands. This is because sludge biochar not only directly increases soil microbial diversity and richness, but also benefits alfalfa growth to affect soil microbiota. The aromatic hydrocarbon structure contributes to the long-term retention of sludge biochar in soils [ 18 ], and thus provides nutrients continuously to soil after its oxidation. 3.3.2. Responses of Plant Growth and Root Morphology Alfalfa growth ( Figure 8 a) and its root morphology ( Figure 8 b) in the G and SBCG treatments differed. Alfalfa growth in the sludge biochar amendment treatment (SBCG) was significantly better than the alfalfa-only treatment (G) ( Figure 8 a). Plant height, shoot biomass, root biomass, and total biomass were significantly higher in SBCG than in G ( Table 5 ). The root segment of alfalfa had similar results, and TRL, RSA, RV (root volume, see Part 2.5), RAD, RTN, and RFN of SBCG were significantly larger than those of G ( Table 6 , p < 0.05). Sludge biochar application could continuously ameliorate the soil environment (physicochemical properties) of IAC mining wastelands. It may reduce soil acidity, specific gravity, and bulk density, improve soil texture and ventilation conditions, enhance soil porosity, water-holding capacity, the content and effectiveness of nutrients, and change the microbial community structure to promote the growth of alfalfa [ 97 , 98 ]."
} | 6,944 |
27148240 | PMC4840303 | pmc | 4,994 | {
"abstract": "The methanogenic biodegradation of crude oil involves the conversion of hydrocarbons to methanogenic substrates by syntrophic bacteria and subsequent methane production by methanogens. Assessing the metabolic roles played by various microbial species in syntrophic communities remains a challenge, but such information has important implications for bioremediation and microbial enhanced energy recovery technologies. Many factors such as changing environmental conditions or substrate variations can influence the composition and biodegradation capabilities of syntrophic microbial communities in hydrocarbon-impacted environments. In this study, a methanogenic crude oil-degrading enrichment culture was successively transferred onto the single long chain fatty acids palmitate or stearate followed by their parent alkanes, hexadecane or octadecane, respectively, in order to assess the impact of different substrates on microbial community composition and retention of hydrocarbon biodegradation genes. 16S rRNA gene sequencing showed that a reduction in substrate diversity resulted in a corresponding loss of microbial diversity, but that hydrocarbon biodegradation genes (such as assA/masD encoding alkylsuccinate synthase) could be retained within a community even in the absence of hydrocarbon substrates. Despite substrate-related diversity changes, all communities were dominated by hydrogenotrophic and acetotrophic methanogens along with bacteria including Clostridium sp., members of the Deltaproteobacteria, and a number of other phyla. Microbial co-occurrence network analysis revealed a dense network of interactions amongst syntrophic bacteria and methanogens that were maintained despite changes in the substrates for methanogenesis. Our results reveal the effect of substrate diversity loss on microbial community diversity, indicate that many syntrophic interactions are stable over time despite changes in substrate pressure, and show that syntrophic interactions amongst bacteria themselves are as important as interactions between bacteria and methanogens in complex methanogenic communities.",
"introduction": "Introduction Since the dawn of the industrial age, widespread use and processing of petroleum products has led to an increase in the hydrocarbon contamination of a wide range of environments. Despite increasing environmental awareness and improved remediation technologies, contamination of the subsurface with hydrocarbon mixtures remains a problem, as the fate of hydrocarbons in the subsurface is not fully understood especially under anoxic conditions. The exposure of subsurface environments to heavy organic loads such as hydrocarbons leads to the rapid development of anoxic conditions in which the majority of hydrocarbon biodegradation is thought to proceed via methanogenesis ( Jones et al., 2008 ). This process is also important in many fossil energy reservoirs, wherein hydrocarbon metabolism over geologic time has led to the accumulation of biogenic methane in gas caps overlying oil legs ( Jones et al., 2008 ). Many studies have now demonstrated that diverse hydrocarbon substrates can be biodegraded under methanogenic conditions (e.g., as reviewed in Foght, 2008 ; Gray et al., 2010 ; Gieg et al., 2014 ). Methanogenic hydrocarbon metabolism requires the presence of at least two groups of organisms in order to proceed in a thermodynamically favorable manner: the syntrophic bacteria that catalyze the activation and subsequent degradation of hydrocarbons to methanogenic substrates (e.g., acetate, formate, CO 2 , and H 2 ), and methanogenic archaea that bioconvert these simpler substrates to CH 4 (plus CO 2 or H 2 O). Methanogenic communities degrading hydrocarbon mixtures are typically diverse ( Gray et al., 2010 ; An et al., 2013 ; Tan et al., 2015a ), but how these microorganisms coordinate their metabolisms to utilize diverse hydrocarbons as carbon and energy sources and to conserve sufficient energy to support life is poorly understood ( Gieg et al., 2014 ). Furthermore, the mechanisms involved in hydrocarbon activation are not fully understood, though fumarate addition has emerged as a key mechanism for the activation of aliphatic, substituted monoaromatic hydrocarbons, and substituted polycyclic aromatic hydrocarbons under various anaerobic electron-accepting conditions ( Foght, 2008 ; Widdel and Musat, 2010 ; Callaghan, 2013 ). Alkylsuccinate synthase (encoded by the assA / masD gene; assA will be the designate name used in this study) is the key enzyme responsible for addition of alkanes to fumarate ( Callaghan et al., 2008 ; Grundmann et al., 2008 ), while benzylsuccinate synthase ( bssA ) adds fumarate to substituted aromatic hydrocarbons ( Heider, 2007 ). In this study, we describe four new methanogenic enrichment cultures that were used to assess community changes as a result of decreased substrate diversity and that were probed for the presence of fumarate addition genes. Two cultures degrading the long-chain fatty acids (LCFA) palmitate and stearate were established from a whole crude oil-degrading methanogenic culture ( Gieg et al., 2008 ) as the inoculum. The LCFA-degrading cultures were subsequently transferred to their parent alkanes, hexadecane and octadecane, in order to see if these cultures maintained the ability to degrade the hydrocarbon substrates present in the original oil degrading culture after long-term incubation on LCFA. All of these cultures (including the whole crude oil-degrading culture) were subjected to pyrotag sequencing of the 16S rRNA gene. We hypothesized that variations in the microbial community composition would be related to the specific carbon substrate supplied, which could provide clues to the identity of hydrocarbon-degraders in the cultures. We expected that the crude oil-degrading culture, which is exposed to a diverse hydrocarbon mixture and is the original parent culture, would be the most biodiverse of the cultures. LCFA- and n -alkane-amended cultures were expected to exhibit less diversity due to the restriction of carbon and energy sources within the culture, and the dilution effects of successive transfers. We further postulated that community members that were maintained across the majority of cultures over time likely play fundamental roles in the syntrophic degradation of shared metabolic products such as fatty acids, acetate, and formate. In light of this, we conducted a co-occurrence network analysis including community members that were retained across the different cultures in an attempt to establish an understanding of the syntrophic interactions occurring in the cultures. A better understanding of methanogenic hydrocarbon metabolism could lead to the improvement of biotechnological applications for in situ bioremediation and for the bioconversion of residual oil to methane as a tertiary energy recovery strategy from fossil-energy reservoirs. Furthermore, insight into the syntrophic lifestyle can help shed light on novel mechanisms for interspecies communication or coordination, interspecies electron and metabolite transfer, and energy conservation in low energy-yielding environments.",
"discussion": "Discussion Reports of methane generation from n -alkanes, and the description of the communities mediating these transformations have become increasingly widespread in recent years (e.g., Zengler et al., 1999 ; Gieg et al., 2008 ; Jones et al., 2008 ; Callaghan et al., 2010 ; Sherry et al., 2013 ; Berdugo-Clavijo and Gieg, 2014 ; Abu Laban et al., 2015 ; Bian et al., 2015 ; Tan et al., 2015b ). Alkanes comprise an abundant fraction of many crude oils, thus their biodegradation under anaerobic conditions is of practical relevance to biotechnological applications in fossil energy reservoirs and fuel-contaminated sites. There remains much to be learned with regards to the pathways, enzymes, and genes involved in strictly anaerobic alkane degradation, as well as the organisms and interactions amongst organisms that methanogenically metabolize hydrocarbons. In this study, we describe new methanogenic LCFA- and n -alkane-degrading cultures, including analysis of their community structure and amplification of known biodegradative genes. Co-occurrence network analysis of the microbial communities of the five related cultures was conducted in order to make a first step in unraveling syntrophic interactions in methanogenic hydrocarbon-degrading systems. As the downstream degradation of both alkanes and fatty acids proceed via a common pathway – β-oxidation, followed by conversion to methanogenic substrates and methane production ( Callaghan, 2013 ) – syntrophic interactions are expected to be similar regardless of the hydrocarbon or fatty acid substrate being degraded. We propose that over time, stable and efficient syntrophic interactions have evolved within the microbial community and that these interactions are fairly resilient to the substrate being degraded. Degradation of crude oil components was previously demonstrated by the source inoculum for the cultures described herein ( Gieg et al., 2008 ). This culture was subsequently transferred to the LCFAs palmitate and stearate, and then these cultures were transferred to their respective parent alkanes, hexadecane and octadecane, to examine the effects of different substrates on microbial community structure and to determine whether the ability to degrade hydrocarbons was maintained following long-term incubation on LCFAs. Over several years of routine culture transfer and substrate amendment, the LCFA-degrading cultures typically converted approximately 84–98% of their fatty acid substrates to methane (plus CO 2 or H 2 O; Figure 1A ). Alkane cultures were subject to a single culture transfer, after which methane production was monitored for close to 900 days. After an extended lag period, methane generation from octadecane became apparent ( Figure 1C ), while the degradation of hexadecane could not be confirmed ( Figure 1B ) because the methane production did not exceed the maximum amount of methane that could be derived from the reductant, cysteine sulfide. Lag periods exceeding several weeks or months have been reported for other methanogenic hydrocarbon-degrading cultures ( Edwards and Grbić-Galić, 1994 ; Townsend et al., 2003 ; Berdugo-Clavijo and Gieg, 2014 ). This delay may be related to a toxic effect as inhibition of microbes by hydrocarbon substrates has been well documented and is thought to be related to interference with biological membranes ( Sikkema et al., 1995 ). The methane production observations aligned with the results of fumarate addition gene amplification efforts. Alkylsuccinate synthase, the enzyme involved in anaerobic n -alkane activation via addition to fumarate, was previously detected in the sediments from which these cultures were initially derived ( Callaghan et al., 2010 ). In the present study, we detected assA amplicons, all presumably derived from a single species, in the residual oil-, stearate-, and octadecane-amended cultures. These results show that the potential ability to biodegrade hydrocarbons can be retained within a syntrophic microbial community even following the long-term absence of hydrocarbon exposure. However, gene detection does not indicate actual expression, thus further studies will be required to confirm that this gene is actually expressed during biodegradation under these different substrate conditions. In contrast, the assA gene was not detected in the palmitate- or hexadecane-amended cultures ( Figure 4 ), for reasons that are not clear given its detection in the stearate and octadecane enrichments. Palmitate metabolism does not require assA , thus a simple explanation is that the gene was lost (i.e., the species harboring this gene was lost) upon transfer from the residual oil culture to palmitate. This explains why a subsequent transfer of the palmitate-degrading culture onto hexadecane did not lead to the biodegradation of this n -alkane ( Figure 1B ). Another possibility is that the specific assA -containing organism in these cultures is involved in the degradation of longer chain alkane substrates, and was thus not capable of hexadecane degradation. If this were true, it would indicate a distinct difference between the fumarate addition genes involved in the degradation of octadecane and higher alkanes, and hexadecane and shorter alkanes. While the necessary evidence to fully test this hypothesis is not yet available due to a shortage of assA gene sequences with known substrate range, this idea was previously postulated for the assA genes involved in the degradation of short chain alkanes ( n -C 3 -C 10 ; Callaghan et al., 2010 ; Tan et al., 2015a ). A distinction between hexadecane and octadecane would not be particularly surprising, as hexadecane is a liquid at ambient temperature, while octadecane is a solid, making the bioavailability of each different in an aqueous environment under mesophilic conditions. A similar hypothesis was made for gaseous alkane assA being phylogenetically distinct from non-gaseous alkane fumarate addition enzymes ( Musat, 2015 ). Phylogenetic analysis in the present study of the amplified assA genes with known references and environmental samples did not reliably pinpoint the organism(s) harboring this assA gene. While the phylogeny of bssA (encoding the alpha subunit of benzylsuccinate synthase) is now generally well enough resolved to infer the clade involved in aromatics activation ( von Netzer et al., 2013 ), this is not yet the case for assA ( Figure 4 ). In our phylogenetic analysis, as in others ( Callaghan et al., 2010 ), members of the Deltaproteobacteria ( Desulfoglaeba sp.) grouped more closely with Betaproteobacteria alkane degraders ( Azoarcus sp., Aromatoleum sp.), than with other Deltaproteobacteria ( Desulfatibacillum sp.) which may indicate that genes encoding alkylsuccinate synthase are more closely related based on the specific alkane substrates being degraded rather than phylogeny or that they are subject to a high degree of horizontal gene transfer (this is currently unknown). It remains to be seen if the purification of additional strictly anaerobic alkane degraders, or the improved description of alkane degraders from the environment or enriched cultures will eventually result in the ability to predict either the taxonomic affiliation of alkane degraders based on phylogenetic analysis of the assA gene and/or the substrate range of the degraders. Members of the Deltaproteobacteria are often cited as key alkane and/or fatty acid degraders in methanogenic cultures (e.g., members of the Syntrophaceae such as Syntrophus / Smithella spp., Gray et al., 2011 ; Cheng et al., 2013 ; Embree et al., 2014 ; Tan et al., 2014 ; Mathai et al., 2015 ) along with other sulfate-reducing alkane degrading Deltaproteobacterial isolates ( Cravo-Laureau et al., 2005 ; Davidova et al., 2006 ; Callaghan et al., 2008 ). While Smithella sp. was abundant (16%) in the residual oil culture, it was present at <2% abundance in the stearate- and octadecane-degrading cultures ( Table 2 ) and the recovered assA gene fragments did not cluster with the assA of known Smithella sp. ( Figure 4 ). These findings suggest that this taxon is not the main stearate- or octadecane-degrading organism in these enrichments. No Deltaproteobacterial OTUs previously associated with hydrocarbon biodegradation were particularly enriched in the octadecane-amended culture ( Table 2 ), suggesting that an as of yet unidentified alternate organism(s) catalyzes the activation of this n -alkane. The bssA gene was not detected in any of the enrichments, which was expected, as the residual oil-amended microcosms contained negligible concentrations of substituted monoaromatic hydrocarbon substrates such as toluene ( Gieg et al., 2008 ). In comparing the microbial communities of the five different cultures, we found that the microbial richness observed was related to the order in which the cultures were enriched, with the most highly enriched ( n -alkane-incubated) cultures containing the lowest species richness ( Table 1 ). A much greater richness and evenness was observed in the presence of more diverse hydrocarbon substrates as found in the residual oil-containing culture ( Tables 1 and 2 ). In addition, there were substrate-specific variations in the microbial communities, with a particular enrichment of methanogens in the n -alkane-incubated cultures ( Table 2 ). While observed and predicted richness decreased with degree of enrichment, evenness actually increased in both of the alkane-incubated cultures relative to the LCFA-degrading cultures ( Table 1 ). Nonetheless, all cultures shared a similar microbial community structure and were dominated by members of the Firmicutes ( Clostridium sp.), Deltaproteobacteria, and Euryarchaeota (mainly hydrogenotrophic and acetotrophic methanogens; Figure 2 , Table 2 ). These findings are consistent with the previous clone library analysis of the residual oil culture ( Gieg et al., 2008 ), and also of a toluene-degrading methanogenic culture derived from the same contaminated sediments ( Fowler et al., 2012 ). This is not particularly surprising as the majority of known syntrophic bacteria are members of the Firmicutes or Deltaproteobacteria ( Sieber et al., 2012 ), and the methanogenic archaea are members of the Euryarchaeota. The extremely high abundance of Clostridium sp. is also consistent with the community from the toluene degrading enrichment from the same sediments in which 30.7% of the culture was found to consist of Clostridium sp. ( Fowler et al., 2012 ). While Clostridium sp. is an abundant organism in these new enrichment cultures, it is possible that its extremely high abundance is partly an artifact of PCR as Clostridium spp. often have multiple copies of 16S rRNA genes with 14 copies having been observed in a single genome ( Větrovský and Baldrian, 2013 ). The presence of a large Clostridium sp. OTU and several Clostridiaceae sp. OTUs in the network analysis, and their connectivity to methanogens in this analysis suggests that, despite their high abundance, this clade is not involved in hydrocarbon activation, but is involved in the downstream conversion of smaller molecules to methanogenic intermediates ( Figure 3 ). This is also in agreement with the results from the aforementioned toluene-degrading culture, in which the highly abundant Clostridium sp. did not incorporate 13 C label from toluene during a 7-day time course experiment ( Fowler et al., 2014 ). While it can not be ruled out that Clostridium spp. might be directly involved in hydrocarbon activation, these results collectively point toward a general role for Clostridium sp./ Clostridiaceae in the downstream degradation of hydrocarbons to methanogenic substrates in oil-associated environments, rather than directly activating hydrocarbons in these cultures. Co-occurrence network analysis revealed the presence of three distinct networks within the cultures ( Figure 3 ). The first consisted of diverse methanogenic archaea including hydrogenotrophs, acetotrophs, and methylotrophs. These methanogens were found to co-occur with syntrophic bacteria including Geobacter sp., Ruminococcaceae , Synergistetes , and in particular, two Clostridiaceae -affiliated OTUs with a high degree of connectivity that linked the two sections of this network. This network is likely reflecting a number of direct interactions between syntrophic bacteria and methanogenic archaea involving interspecies metabolite transfer, and possibly even direct interspecies electron transfer (DIET) between Geobacter sp. and methanogens ( Rotaru et al., 2014 ). The second network consisted of 14 OTUs of diverse syntrophic bacteria and a single Methanolinea sp. OTU. The syntrophic bacteria within this network were densely interconnected with a high mean degree of connectivity. Whether these OTUs are all connected due to a highly interactive syntrophic network, or due to the presence of a small number of organisms that interact with a large number of partners (which has previously been observed in these densely connected networks; Berry and Widder, 2014 ) is unknown. However, this network indicates that it is not only interactions between syntrophic bacteria and methanogens that are important in these communities, but also that interactions amongst syntrophic bacteria are substantial. The third network also emphasizes the importance of interactions amongst syntrophic bacteria as it consists solely of four bacterial OTUs related to obligate anaerobic fermenters (three Clostridiales , one Anaerolineaceae ) that all co-occur. Lachnospiraceae and Ruminococcaceae are common inhabitants of GI tracts, anaerobic digesters, and other methanogenic environments ( Nelson et al., 2011 ; Biddle et al., 2013 ). Anaerolineaceae are also commonly found in syntrophic environments including the GI tract and anaerobic digestors, where they are typically characterized as secondary fermenters that sequentially degrade fatty acids and/or carbohydrates to methanogenic or other syntrophic intermediates, and are known to associate with hydrogenotrophic methanogens ( Yamada et al., 2006 ; Biddle et al., 2013 ; St-Pierre and Wright, 2013 ). Anaerolineaceae have also been previously detected in high abundance in alkane-degrading cultures and anaerobic oil-impacted environments ( Savage et al., 2010 ; Sherry et al., 2013 ; Liang et al., 2015 ) including methanogenic and non-methanogenic syntrophic hydrocarbon-degrading cultures ( Kleinsteuber et al., 2012 ). Further, Anaerolineaceae have previously been postulated to be involved in alkane activation under methanogenic conditions ( Sherry et al., 2013 ; Liang et al., 2015 ). Thus, an alternative possibility is that Anaerolineaceae is involved in alkane activation, and subsequent fatty acid degradation is catalyzed by the Clostridiales OTUs, though additional evidence to support Anaerolineaceae as alkane degraders in these cultures is currently lacking. Overall, network analysis indicates that there are strong interactions between syntrophic bacteria and methanogens, but the strongest and most abundant interactions we observed in these cultures occurred amongst the bacteria themselves. This suggests the existence of numerous cooperative interactions between groups of bacteria as well as between bacteria and methanogens within syntrophic methanogenic ecosystems. While the use of co-occurrence networks can provide clues as to how organisms interact in syntrophic cultures, they must also be interpreted with caution. Co-occurrence in this analysis merely indicates that the organisms were observed to co-occur repeatedly, but does not preclude the possibility that co-occurring organisms merely share similar niches within these enrichment cultures and do not interact. However, due to the difficulty in elucidating syntrophic interactions in mixed cultures, we believe that co-occurrence network analysis provides a method that can be used to predict syntrophic relationships in complex communities when applied to multiple related communities. In summary, we demonstrated the methanogenic bio degradation of palmitate, stearate, and octadecane in cultures derived from a whole crude oil-degrading enrichment culture ( Gieg et al., 2008 ). The fact that octadecane degradation occurred following 3 years of pre-incubation on a non-hydrocarbon substrate (stearate) showed that alkane degraders can persist in environments despite the absence of hydrocarbons. In addition, we described the microbial communities of each of these cultures and a hexadecane-amended culture, and observed an expected diversity reduction when whole crude oil-amended cultures were successively transferred onto single carbon substrates. Confirmation of syntrophic interactions between individual OTUs ultimately requires physiological evidence. However, given the complexity of methanogenic communities, and the difficulty in culturing syntrophic bacteria as individuals or in co-culture, applying microbial co-occurrence network analysis provides a means to predict microbial interactions, enabling insight into potential interspecies interactions and the microbial foodwebs that exist in complex communities. By examining microbial co-occurrence in these cultures, we were able to identify organisms that were insensitive to the carbon substrate being metabolized, and examine their degree of co-occurrence with other community members. While these co-occurrences likely do not all represent syntrophic interactions, this is a first step toward identifying organisms that form associations within this stable syntrophic community. Our analysis reveals not only stable interactions between syntrophs and methanogens, including possible DIET interactions, but also strong interactions amongst the syntrophic bacteria themselves. These findings emphasize the complex foodwebs existing in methanogenic communities. Furthermore, these predictions can provide preliminary evidence for further hypothesis testing using metagenomic and/or metatranscriptomic data and/or physiological investigations."
} | 6,344 |
22355311 | PMC3280245 | pmc | 4,995 | {
"abstract": "Raspy crickets (Orthoptera: Gryllacrididae) are unique among the orthopterans in producing silk, which is used to build shelters. This work studied the material composition and the fabrication of cricket silk for the first time. We examined silk-webs produced in captivity, which comprised cylindrical fibers and flat films. Spectra obtained from micro-Raman experiments indicated that the silk is composed of protein, primarily in a beta-sheet conformation, and that fibers and films are almost identical in terms of amino acid composition and secondary structure. The primary sequences of four silk proteins were identified through a mass spectrometry/cDNA library approach. The most abundant silk protein was large in size (300 and 220 kDa variants), rich in alanine, glycine and serine, and contained repetitive sequence motifs; these are features which are shared with several known beta-sheet forming silk proteins. Convergent evolution at the molecular level contrasts with development by crickets of a novel mechanism for silk fabrication. After secretion of cricket silk proteins by the labial glands they are fabricated into mature silk by the labium-hypopharynx, which is modified to allow the controlled formation of either fibers or films. Protein folding into beta-sheet structure during silk fabrication is not driven by shear forces, as is reported for other silks.",
"introduction": "Introduction The ability to produce silk has evolved in at least 23 groups of insects [1] , in spiders [2] and in several other arthropods [3] , [4] . Silk research has focused on silkworm cocoon and spider dragline silks, which have independently evolved a number of convergent features. Spider and silkworm silks consist of long, repetitive proteins that fold predominantly into beta-sheets, with the protein backbone parallel to the fiber axis [2] . Highly ordered nanocrystals are embedded in regions of less order and confer high tensile strength to the fibers [5] . The molecular arrangement in spider and silkworm silks is the result of shear forces and controlled dehydration acting on highly concentrated silk protein solutions as they pass through a hardened aperture known as a spinneret [6] , [7] . Although less characterised, other silks are dramatically different. For example, protein backbones in silks made by glow-worms and adult lacewings are orientated perpendicular instead of parallel to the fiber axis [8] ; the silks of fleas, bees and lacewing larvae contain proteins arranged in alpha-helices instead of beta-sheets [8] , [9] ; and the fibrous proteins in some silks are an order of magnitude smaller than spider dragline and silkworm cocoon silk proteins [10] . Further characterisation of silks in addition to spider and silkworm silks will allow a comparative approach to understanding the complex molecular arrangements found in silk. Crickets in the family Gryllacrididae (raspy crickets) produce silk, while only one other insect in the order Orthoptera does so [11] , [12] . Raspy crickets use silk fibers to build shelters into which they retreat during the day [12] , [13] . The fibers are used variously to sew leaves together, to stabilise burrows in earth or sand, or to restrict access to tree hollows depending on species [12] , [14] . The shelters are generally presumed to be a defense against predation, though it has also been suggested that they may limit desiccation in drier environments [12] . Both sexes are capable of producing fibers within hours of hatching and continue to produce shelters throughout their lives [15] . Shelters are highly valued and individuals may label their own shelters with a chemical cue [16] allowing them to return to the same shelter many times. Very little is known about the method of fabrication of silk fibers by raspy crickets. Rentz and John [12] observed silk production from cricket mouthparts, but the origin of the material is unknown and the internal anatomy of raspy crickets is poorly described. Other insects that generate silk from their mouthparts do so using protein solutions produced in modified labial glands [17] . Wetas and king crickets (Anostostomatidae), the closest relatives of raspy crickets [18] , use their labial glands to produce saliva [19] . Anostostomatid labial glands are arranged in grape-like clusters called acini [20] . Acinar cells secrete into the lumen of a branching series of ductules joined to the common duct on each side of the body. The left and right common ducts join at the labium, where they empty into a cavity between the labium and hypopharynx, called the salivarium. An additional organ, the reservoir, is formed by a sack-like outgrowth of the common duct on each side [19] , [20] . Nothing is known about the material composition of raspy cricket silk fibers or how they are produced. We investigated the biochemistry and physical structure of raspy cricket fibers and the method of their production. Our motivation in this work was to enhance understanding of which features of different silks that have evolved independently in different arthropod groups are convergent and functional, and which features are historical and accidental.",
"discussion": "Discussion The raspy cricket species used in this study produced shelters by joining leaves together with silk-webs. Given that the silk-webs were not air-tight, they are unlikely to be effective in preventing desiccation or the ingress of parasites in the wild, and the most likely function is to reduce predation. Silk-webs were found to be made of protein, and visual observations of crickets fabricating silk suggested that the labial glands might function as silk glands. Our identification of transcripts encoding silk proteins in labial glands confirms this directly. The silk gland consists of acini connected by a network of ducts to the insect's salivarium. Silk glands from species that make cocoons have large lumens that store silk proteins in preparation for a short period of intense silk production [27] . In contrast, raspy cricket gland lumens are small, the amount of silk required for shelters is low, and animals probably produce silk as required. A reservoir attached to the common duct immediately before it joins the salivarium does not contain silk proteins but does contain amylase, suggesting that it has a salivary function. The silk-webs consisted of fibers and films, with the fibers providing the mechanical backbone of the webs and the films serving to glue the fibers to other building materials and to each other. Micro-Raman spectra indicated that the proteins present in fibers and films are indistinguishable at the levels of amino acid composition and secondary protein structure, suggesting they are made from the same protein solution. We propose that the anatomical arrangement in raspy crickets is specialised to be able to produce fibers and films interchangeably. Instead of having an external, tubular spinneret like lepidopterans and spiders [28] , [29] the labial ducts of raspy crickets end in an aperture that is too large to act as a draw-down taper. Instead, silk dope probably exists in a liquid state in the salivarium chamber created by tucking the hypopharynx under the raised margins of the labial paraglossae. Single fibers are formed by drawing through the taper at the extremity of the labium, between the tips of the two paraglossae and the hypopharynx ( figure 5 ). Opening the labium and hypopharynx by muscular control allows the insect to deposit silk dope in globules, which dry into films. Both films and fibers are found in the silks of other insects, including sawflies, honeybees, hornets and some beetles, e.g. [30] . 10.1371/journal.pone.0030408.g005 Figure 5 Fabrication of fibers and films by raspy crickets. ( a ) Indicative position of labial gland acini (aci) and reservoirs (res) and associated ducts. ( b ) Breakdown of raspy cricket mouthparts, including labium (lb), maxillae (mx), mandibles (md) and labrum (lr). ( c ) Expanded view of the labium with the hypopharynx (h) in closed position, forming a draw-down taper with the labial paraglossae (pg) to allow fiber fabrication. Liquid silk in the salivarium chamber between paraglossae and hypopharynx is shown in purple. lp = labial palp; rd = reservoir duct; ad = acinar duct. (d) Expanded view of the labium with the hypopharynx in open position, allowing silk dope to flow over paraglossae, allowing fabrication of films. The major silk protein Ail SP1 has evolved convergent features to those of the dominant fibrous proteins from silkworm cocoon and spider dragline silks: it is large, contains a high proportion of small amino acids, and adopts primarily an extended beta-sheet conformation in the mature silk. The spacing between beta-sheets in the side group direction is 1.27 nm, within the range reported for lepidopteran beta-sheet silks (0.93–1.57 nm) and slightly greater than that of polyalanine (1.06 nm) [31] , [32] . If crystallites are formed from the repeat units of Ail SP1 this is a reasonable estimate, since 27% of residues in repeat regions are glycine, 27% are alanine, and 18% are serine, with the remainder larger residues. The presence of a cysteine residue in the C-terminus of Ail SP1 and the high frequency of cysteines in Ail SP3, as well as the observation that reducing agents are required for silk solubilization, suggest that disulfide bonding plays a structurally important role in cricket silk. If raspy crickets produce silk films by depositing globules of liquid silk dope which are subsequently allowed to dry, then proteins in films are exposed to minimal shear and compression forces. Since the secondary structure of the proteins in fibres and films is essentially the same, shear forces cannot be a primary mechanism driving beta-sheet formation, as it is for the proteins in silkworm cocoon fibers [33] . Some insect species produce silk consisting of proteins with alpha-helical [34] and cross-beta [35] molecular structures, which shear forces are likely to disrupt rather than favour. In these species, the final molecular structure must form due to some combination of dehydration and the ordering of proteins within the gland prior to silk fabrication. Similar mechanisms may drive beta-sheet formation by raspy cricket silk proteins. Although not required for beta-sheet formation, shear force most likely accounts for the long range alignment of proteins within the fibers demonstrated by WAXS patterns and by birefringence measurements. A comparison between raspy cricket silk and two unrelated insect silks, produced by silkworms and webspinners, is shown in Table 2 . The three types of insects have been faced with a similar problem, the need to produce an insoluble and stable building material, and have independently evolved solutions with some convergent features. All three insect families produce silk proteins, with primary fibroin sizes ranging from moderately large to enormous, and effective size further increased due to cysteine cross-linking. The fibroins are rich in particular amino acids and in repetitive motifs that foster formation of extensive intermolecular beta-sheets, which assemble into insoluble and stable beta-sheet crystallites. On the other hand, there is wide variation between the three silks in regard to the gland of production, form of spinneret, and morphology of the mature silk product. Convergent evolution of silks has occurred at the molecular level; however the silk fabrication process and its anatomical substrates are highly flexible and idiosyncratic. 10.1371/journal.pone.0030408.t002 Table 2 Comparison of silks composed of beta-sheet forming proteins. Silkworm Webspinner Raspy cricket Major silk fibroin size (kDa) 390 [39] \n ∼67 [40] \n 220, 300 Fibroin composition 46% Gly, 30% Ala, 12% Ser [39] \n 44% Gly, 31% Ser [41] \n 19% Ala, 19% Gly, 17% Ser Repetitive sequence motifs Yes [39] \n Yes [42] \n Yes Dominant protein structure beta-sheet [42] \n beta-sheet [41] \n beta-sheet Role for cysteine cross-links Yes [43] \n Yes [40] \n Yes Silk gland Labial gland [17] \n Dermal glands on tarsi [44] \n Labial gland Spinneret Rigid tube on labium [45] \n Tubular bristles [44] \n Movable labium-hypopharynx Form of mature silk Double fiber Multiple fine fibers [41] \n Fiber or film"
} | 3,082 |
26831705 | null | s2 | 4,996 | {
"abstract": "Heterologous biosynthesis of natural products is meant to enable access to the vast array of valuable properties associated with these compounds. Often motivated by limitations inherent in native production hosts, the heterologous biosynthetic process begins with a candidate host regarded as technically advanced relative to original producing organisms. Given this requirement, E. coli has been a top choice for heterologous biosynthesis attempts as associated recombinant tools emerged and continue to develop. However, success requires overcoming challenges associated with natural product formation, including complex biosynthetic pathways and the need for metabolic support. These two challenges have been heavily featured in cellular engineering efforts completed to position E. coli as a viable surrogate host. This chapter outlines steps taken to engineer E. coli with an emphasis on genetic manipulations designed to support the heterologous production of polyketide, nonribosomal peptide, and similarly complex natural products."
} | 259 |
30883020 | PMC6922529 | pmc | 4,998 | {
"abstract": "Summary The use of renewable waste feedstocks is an environment‐friendly choice contributing to the reduction of waste treatment costs and increasing the economic value of industrial by‐products. Glycerol (1,2,3‐propanetriol), a simple polyol compound widely distributed in biological systems, constitutes a prime example of a relatively cheap and readily available substrate to be used in bioprocesses. Extensively exploited as an ingredient in the food and pharmaceutical industries, glycerol is also the main by‐product of biodiesel production, which has resulted in a progressive drop in substrate price over the years. Consequently, glycerol has become an attractive substrate in biotechnology, and several chemical commodities currently produced from petroleum have been shown to be obtained from this polyol using whole‐cell biocatalysts with both wild‐type and engineered bacterial strains. Pseudomonas species, endowed with a versatile and rich metabolism, have been adopted for the conversion of glycerol into value‐added products (ranging from simple molecules to structurally complex biopolymers, e.g. polyhydroxyalkanoates), and a number of metabolic engineering strategies have been deployed to increase the number of applications of glycerol as a cost‐effective substrate. The unique genetic and metabolic features of glycerol‐grown Pseudomonas are presented in this review, along with relevant examples of bioprocesses based on this substrate – and the synthetic biology and metabolic engineering strategies implemented in bacteria of this genus aimed at glycerol valorization.",
"conclusion": "Conclusions and outlook Over the last few years, glycerol has become an appealing choice for bioproduction, especially in processes designed for the synthesis of reduced chemicals. Pseudomonas species display a unique combination of genetic and metabolic architectures when growing on glycerol as the main carbon substrate, in particular, P . putida and P . aeruginosa , where the issue has been examined to some extent. As indicated in the first part of this article, multi‐omic strategies have strongly helped to elucidate the regulatory networks that rule glycerol utilization in P . putida KT2440 (including stochastic activation of genes encoding key enzymes needed for glycerol processing). In silico ‐guided metabolic engineering strategies have also been implemented to increase the production of PHAs from this substrate. Admittedly, the full potential of glycerol as a biotechnological substrate for Pseudomonas has not been fully realized yet, but promising avenues can be envisioned in the near future – including novel strategies merging synthetic biology designs and laboratory evolution of engineered strains (Nørholm, 2019 ). First, once regulatory constraints for expression are overcome (e.g. by eliminating the GlpR repressor, as discussed above), core metabolic reactions linked to glycerol can be manipulated to foster synthesis of value‐added C3 compounds. For example, the connection of glycerol to biomass formation could be severed, and enzymatic sub‐networks could be set up for generating molecules of biotechnological interest, e.g. DHAP and derivatives thereof in resting cells [or, in any case, uncoupled from growth (Durante‐Rodríguez et al., \n 2018 ; Volke et al., \n 2019 )]. Along the same lines, glycerol metabolism could be refactored to reduce carbon loss as CO 2 , while either concomitantly or separately, adjusting the redox balance to provide a better intracellular environment for hosting transformations of interest on other substrates. Furthermore, several metabolic routes could be rewired for fuelling the EDEMP cycle bottom‐up by means of a synthetic C3 neogenesis, empowering NADPH overproduction from glycerol processing, particularly useful for biosynthesis of reduced bioproducts. To this end, genetic editing of the Pseudomonas metabolism will benefit from systems biology approaches for simulating and predicting the effects of given mutations on specific carbon fluxes and pathways (Cho and Palsson, 2009 ; Gray et al., \n 2015 ; Galardini et al., \n 2017 ). A second, considerable challenge is the further adaptation to meet the composition of industrial‐grade crude glycerol from biodiesel production. Typically, the glycerol stream has a polyol content in the range of 30‐65% (wt/vol), with the rest of the stream being CH 3 OH, fatty acid methyl esters, free fatty acids and glycerides together with ashes (Hu et al., \n 2012 ). The waste also has a high pH due to the residual KOH or NaOH carried on from upstream transesterification of oils and fats that generate biodiesel. Industrial‐grade glycerol, often available as a non‐homogeneous oily mixture, is obviously quite different to what one can have in the controlled and pure‐substrate conditions of a shaken‐flask cultivation in the laboratory. While pre‐treatment (i.e. purification) may help improving the physical characteristics of this carbon source, bacteria have to ultimately face a mixture of compounds – some of them toxic and others not easily metabolizable. This offers again an opportunity to genetically knock in heterologous traits for the whole‐cell catalyst to endure the stressful conditions imposed by the use of crude glycerol. The issue here includes both endurance to the toxic effect of the non‐glycerol compounds of the mixture and the introduction of additional pathways for degrading or even growing on the additional carbon sources present in the medium. Some partial successes using industrial glycerol waste in bioproduction have been reported (see Table 1 ), yet the room for improvement in this field is still considerable. Glycerol‐based bioprocesses have to be run in bioreactors, with a very large liquid‐to‐biomass ratio and sterile culture media that, after the operation takes place, need to be processed for purification of the molecules of interest. This scenario makes the production of such compounds costly and only appealing when the price tag of the thereby‐generated chemical is sufficiently high. The third avenue for improving glycerol valorization is therefore (re)designing the industrial engineering part of the bioprocesses better, and easing the downstream operations for reducing fermentation costs. This challenge not only applied to fermentations using this particular substrate, but to Microbial Biotechnology as a whole (de Lorenzo and Couto, 2019 ). Yet, even marginal improvement in bioprocess performance can make a considerable difference in the choice of substrates for feeding industrial‐scale production. In each of these fronts, synthetic biology and metabolic engineering are bound to contribute to the overarching goals of sustainable production from renewable resources, zero waste, and circular management of feedstocks and products, in the frame of the so‐called 4th Industrial Revolution (Schwab, 2017 ).",
"introduction": "Introduction Contemporary synthetic biology and metabolic engineering offer the possibility of expanding the substrate range of microbial cell factories beyond the sugars typically used as carbon sources (Calero and Nikel, 2019 ; Prather, 2019 ). Examples of this sort of metabolic manipulation for broadening substrate ‘palatability’ of bacteria include several chemical species, ranging from simple C1 compounds such as CO 2 or HCOOH (Antonovsky et al., \n 2016 ; Yishai et al., \n 2018 ) to structurally complex substrates such as lignocellulosic materials derived from biomass (Beckham et al., \n 2016 ; Barton et al., \n 2018 ; Kim and Woo, 2018 ). Alcohols conform a special category of alternative substrates for biotechnology, and they are currently being discussed as promising renewables for sustainable bioproduction (Stowell et al., \n 1987 ; Smith, 2004 ; Dahod et al., \n 2010 ; Hoffmann et al., \n 2018 ). Glycerol (1,2,3‐propanetriol, C 3 H 8 O 3 ), for instance, is a widely available, versatile and structurally simple compound that can be used as a carbon source or as a precursor in a variety of chemical and biological conversions. This polyol has been traditionally used in multiple industrially relevant areas, e.g. as an ingredient in foods and beverages (by exploiting its sweetening properties; in fact, the name glycerol is derived from the Greek γλυκερός, ‘sweet’), as well as pharmaceuticals and cosmetic products, both as solvent and humectant (Pagliaro and Rossi et al., \n 2008b ). Biodiesel is a fuel comprised of monoalkyl (methyl, ethyl or propyl) esters of long‐chain fatty acids derived from vegetable oils or animal fats (Hollinshead et al., \n 2014 ). Its value as a fuel has been recognized as early as the 19 th century: the transesterification of a vegetable oil catalysed by a base was conducted four decades before the first diesel engine became functional (Henriques, 1898 ). Biodiesel has promising lubricating properties and cetane ratings compared to low sulfur diesel fuels, with a calorific value of about 37 MJ kg −1 . The current transesterification process used for biodiesel production involves the treatment of yellow grease (recycled vegetable oil), virgin vegetable oil or tallow with a mixture of NaOH or KOH and CH 3 OH (van Gerpen and Knothe, 2010 ). The main by‐product of this production process is glycerol: ca. 10 kg of crude glycerol is generated for every 100 kg of biodiesel produced. The fast development of the biofuel industry in several countries over the last three decades (with a global production volume of 3.8 million tons in 2005) has generated a considerable amount of crude glycerol (Suppes, 2010 ). Approximately 85% of all the biodiesel production over the last decade came from the European Union (Ntziachristos et al., \n 2014 ). In addition, the bioethanol process (using Saccharomyces cerevisiae as biocatalyst) generates glycerol up to 10% of the total sugar (usually sucrose) consumed in the fermentation (Hasunuma and Kondo, 2012 ; Mohd Azhar et al., \n 2017 ). As a consequence of this global situation, the last 10 years have witnessed the rise of glycerol as a very attractive substrate for bacterial fermentations (Mota et al., \n 2017 ). The excess of crude glycerol produced in the biofuel industry led to a decrease in glycerol price, and some years ago, it was even considered a waste (with an associated disposal cost) by many biodiesel‐production plants. Converting crude glycerol into value‐added products thus became a relevant need to improve the viability of the biofuel economy (Pagliaro and Rossi et al., \n 2008a ), and both chemical and biological approaches have been explored to convert glycerol into more valuable products. Considering that crude glycerol is a non‐edible renewable, its use has also advantages in terms of sustainability as it does not compete with other substrates that could be otherwise used in the food industry (Stichnothe, 2019 ). Compared to chemical routes for transformation of the polyol, biological transformation offers several advantages, ranging from less energy use (thus making the process more environment‐friendly) to higher specificity, and increased tolerance to impurities such as salts and CH 3 OH, both of which occur at high levels in crude glycerol (Katryniok et al., \n 2009 ). Over the last few years, however, global markets have changed and oil prices have stabilized – which has directly impacted biodiesel production (Pagliaro, 2017 ). Nevertheless, glycerol continues to attract attention as a substrate for biotechnology as it can be used by a myriad of microorganisms for the synthesis of a wide range of bioproducts (da Silva et al., \n 2009 ; Dobson et al., \n 2012 ; Pettinari et al., \n 2012 ; Mattam et al., \n 2013 ; Mitrea et al., \n 2017 ). Moreover, current trends indicate that biodiesel will become the clean liquid fuel of choice in many countries, especially in those that have legal requirements to use alternatives to petrochemical fuels (Guo and Song, 2019 ). The United States Environmental Protection Agency, for instance, established a fuel standard volume requirement for biodiesel of 8 millon litres for 2019 (Weaver, 2018 ) – requirements that will inevitably result in an increasing availability of raw glycerol. Interestingly, the biotechnological value of glycerol as a substrate has been recognized since the early times of industrial microbiology (Johnson, 1947 ; Gunsalus et al., \n 1955 ). In fact, some of the oldest examples of technical‐scale bioreactor fermentations include the transformation of glycerol into biomass and reduced biochemical products. Nakas et al . ( 1983 ), for instance, described the fermentation of glycerol by Clostridium pasteurianum in an attempt to obtain a marketable product [a mixture of n ‐butanol, 1,3‐propanediol (1,3‐PDO) and ethanol] from glycerol photosynthetically formed by halophilic Dunaliella algae. Due to the more reduced nature of the carbon atoms in glycerol as compared to sugars (e.g. glucose and xylose, customary substrates in bioprocesses), the polyol is mostly processed via oxidative metabolism in aerobic processes. There are, however, several bacteria that can ferment this substrate anoxically, e.g. some clostridia and a few enterobacteria (Hatti‐Kaul and Mattiasson, 2016 ) – a circumstance that has been also exploited for the design of industrial bioprocesses. Until the last decade, for instance, it was widely accepted that Escherichia coli was unable to use glycerol as a substrate in the absence of external electron acceptors (Booth, 2005 ). Since then, several studies describing the fermentation of glycerol by different wild‐type or mutant E . coli strains have paved the way for the efficient use of this low‐cost, readily available substrate to synthesize a variety of biotechnologically relevant products under different oxygen availability conditions (Yazdani and González, 2007 ; Murarka et al., \n 2008 ; Nikel et al ., 2006 , 2008a , 2010a ) – thus increasing the sustainability of fermentation processes using this polyol as the substrate. The higher degree of reduction of glycerol (γ = 4.7) over glucose (γ = 4) facilitates the synthesis of reduced bioproducts as demonstrated in E . coli strains (Nikel et al., \n 2008a , 2008b , 2010b ). Since less carbon has to be oxidized into CO 2 to generate reducing power, the use of glycerol potentially offers higher yields on substrate than when using sugars. Yet, what are the biotechnological uses of glycerol beyond the so‐called model bacterial species? The last decade has witnessed an exponential increase in the number of studies exploiting Pseudomonas species as biocatalysts. In particular, P . putida KT2440, a non‐pathogenic soil bacterium that has been adapted to laboratory conditions (Nelson et al., \n 2002 ; Belda et al., \n 2016 ), has emerged as the chassis of choice for engineering biochemical pathways while exploiting its intrinsically high tolerance to different types of physicochemical stresses (Poblete‐Castro et al., \n 2012a , 2017 ; Nikel et al., \n 2014a , 2014b ; Nikel and de Lorenzo, 2018a , 2018b ; Abram and Udaondo, 2019 ). Several studies have described the use of glycerol by Pseudomonas species, and biochemical and genetic studies have disclosed a rather different metabolic operation, genetic regulation and physiological responses as compared to other bacteria. Against this background, in this article, we review our current knowledge on the use of glycerol by Pseudomonas species either via natural or engineered pathways – with an emphasis on the physiology and metabolism of P . putida and the many opportunities that this substrate brings forth for biotechnological applications."
} | 3,942 |
36753538 | PMC9908019 | pmc | 4,999 | {
"abstract": "Neuromorphic computing is expected to achieve human-brain performance by reproducing the structure of biological neural systems. However, previous neuromorphic designs based on synapse devices are all unsatisfying for their hardwired network structure and limited connection density, far from their biological counterpart, which has high connection density and the ability of meta-learning. Here, we propose a neural network based on magnon scattering modulated by an omnidirectional mobile hopfion in antiferromagnets. The states of neurons are encoded in the frequency distribution of magnons, and the connections between them are related to the frequency dependence of magnon scattering. Last, by controlling the hopfion’s state, we can modulate hyperparameters in our network and realize the first meta-learning device that is verified to be well functioning. It not only breaks the connection density bottleneck but also provides a guideline for future designs of neuromorphic devices.",
"introduction": "INTRODUCTION Topological solitons have been extensively studied in field theory ( 1 , 2 ). It was not until recently that their magnetic versions showed great potential for future electronic devices ( 3 – 8 ). Magnetic solitons are robust due to their topological nature and are easy to manipulate using various techniques based on fields, currents, photons, etc. ( 9 – 12 ). One of the successful examples is skyrmions. Observed in a vast diversity of magnetic materials, they have been developed into devices and circuits for a range of applications operating even at room temperature ( 13 , 14 ). Among the applications of magnetic solitons, nonconventional computing attracts us the most for being unexplored in traditional electronics, especially neuromorphic computing ( 15 – 18 ). Various synaptic devices based on magnetic solitons are proposed as the building blocks for neural networks, and most of them are designed in the light of the leaky-integrate-fire model. In some of those proposals, the position of the domain wall or the size of the skyrmion plays the role of the electric potential in a neuron ( 16 , 17 ). The domain or the skyrmion is expanded in the presence of an input current or shrinks to its original size as a result of a certain energy descent when the current is withdrawn. In another design, skyrmions function as neurotransmitters, which would trigger a spiking signal once their concentration in the detecting area has reached a certain level ( 18 ). However, these practices are confronted with an emergent requirement called meta-learning. In addition to modulating weights and biases in a neural network, meta-learning requires taking the network structure or other metadata under training as well. It reduces manual labor in model training, helps us escape the overfit trap, improves training efficiency, and, lastly, makes machine learning more available in various real-life scenarios ( 19 – 23 ). Despite all these advantages, this requirement would be difficult to realize at a hardware level without adding extra circuit complexity. During circuit design, a neural network is usually decomposed into the basic units that can be directly replaced by electronic devices. Under current practice, the basic unit is a synapse, and the network structure is implemented by physical connections between them. As a result, the network structure is hardwired, and it entails an additional controller to program these connections at a software level, forcing the circuit designer to choose between inflexibility and unnecessary complexity. Besides, since all the synaptic devices are inevitably distributed and manufactured in a plane, their connection density becomes so limited that it is almost impossible for the circuit to achieve performance as high as that of a human brain, in which neurons are densely connected. To tackle the above problem, we introduce the concept of magnon scattering. It is based on a reflection on two customary ways of thinking prevailing in current practice. First, instead of concentrating on the implementation of synapses, we should pay more attention to seeking a physical process equivalent to neural network structure. Second, following this idea, instead of regarding magnetic solitons merely as signal carriers, we should develop their potential value for representing network structure. Here, we propose magnons as the signal carriers and their scattering by a magnetic soliton as the physical process for substituting the network structure. Since magnons are not exclusive particles like magnetic solitons, it is therefore possible to break the connection density bottleneck by building multiple channels of different frequencies in parallel. These advantages make them a more ideal carrier than magnetic solitons. The frequency spectrum of magnons will be used to encode multidimensional signals, and in this view, the magnon scattering can be seen as a nonlinear map between these signals, which implicitly has the structure of a neural network. To obtain an adjustable neural network, the scattering pattern should be easily modulated by controlling the state of the soliton. The flexibility required by meta-learning is lastly settled by selecting hopfions as the kind of magnetic solitons responsible for causing the magnon scattering. Skyrmions are excluded because they cause relatively constant scattering results due to their simple spin configuration. Hopfions, instead, are fully qualified to satisfy our needs. Discovered in 1975 as a group of three-dimensional (3D) soliton solutions of the Skyrme-Faddeev model ( 24 , 25 ), they have been demonstrated either theoretically or experimentally in several magnetic systems such as chiral magnets ( 26 – 29 ), frustrated magnets ( 30 , 31 ), magnetic multilayers ( 32 ), and ferromagnets with high-order exchange interactions ( 33 ). As knot-shaped solitons, hopfions host such a rich variety of 3D configurations that even their topological classification becomes a complicated question. This complexity provides a scattering characteristic pathological enough to serve as a neural network. The most important point is that, as an anisotropic object in 3D space, a hopfion has all the degrees of freedom of a 3D rigid body, a set of state parameters that are simple but complete enough for modulating the neural network through the pathological scattering characteristic. However, up until now, the intricate dynamics of hopfions has not been fully explored ( 34 – 37 ). Recent research on hopfions driven by spin transfer torque has revealed the “exotic dynamics” of hopfions, including translation, rotation, and dilation ( 35 ). However, none of them has reached the level of omnidirectional motion in 3D or programmable velocity control for all degrees of freedom in 3D. The constrained motion of a hopfion places a restriction on the number of its state parameters, represses the “gene expression” of hopfions’ scattering properties and, therefore, should be settled before the meta-learning framework is constructed. Spin-wave polarization, an additional degree of freedom emerging from antiferromagnetism, is used to achieve omnidirectionality and to simplify the velocity control mechanism. Here, we implement a meta-learning framework using magnon scattering modulated by the omnidirectional motion of a hopfion. The omnidirectional motion is realized on the basis of a first study into the hopfion-driving effect of spin-wave polarization in antiferromagnets. An empirical relationship is displayed geometrically between the velocity of the hopfion and the spin-wave polarization, and it results in a simple scheme where the velocity of a hopfion is encoded in two spin-wave sources. Four 3D trajectories of the hopfion are shown in ascending order of complexity to demonstrate the power of the encoding scheme. The frequency dependence of hopfions’ motion is also studied for the later use of the frequency spectrum of magnons. On the basis of all these preparatory investigations, the meta-learning framework is realized. Signals are carried by the frequency spectrum of magnons, and the neural network is built into the scattering process, which can be modulated by the position and altitude of the hopfion. The framework is applied to predicting periodic and chaotic signals, and the prediction turns out to be of high accuracy, indicating the well functionality of our approach.",
"discussion": "DISCUSSION Here, we first study the polarization-dependent motion of a hopfion and the frequency-dependent scattering of magnons by a hopfion. We unveil the hopfion-driving effect of spin-wave polarization by associating it with the Hall angle of the hopfion. The relationship is so simple that it enables efficient programming of a hopfion’s trajectory. In light of this, we devised a way to achieve omnidirectional motion in 3D space. A series of trajectories are constructed step by step in ascending order of complexity: first, a circle with its tangent velocity limited in a coordinate plane; second, a helix with an extra velocity perpendicular to that plane; third, a Chinese knot as a natural extension to the helix; and lastly, a trefoil knot involving omnidirectional motion. We further propose a linear model for predicting magnon-driven velocities for different solitons, and it has been confirmed by the frequency dependence of the hopfion’s motion. Under the eikonal approximation, spin waves are dealt with as particles moving in the emergent magnetic field, and the motion of a soliton is understood as a phenomenon of momentum exchange between these particles and the soliton. Our model highlights the wave-particle duality of spin waves and can explain the results of frequency dependence well. On the basis of these mechanisms, we propose a method for neuromorphic computing using the omnidirectional motion of a hopfion and nonlinear magnon-hopfion scattering. The states of neuron nodes in the neural network, including the input signals, are encoded into the frequency distribution of magnons, and the connections between these neurons are built into the scattering of magnons. The frequency distribution has intrinsically infinite degrees of freedom, permitting a high connection density. The nonlinearity of the scattering process meets the requirements of neuromorphic computing. The meta-learning is realized by using the 3D omnidirectional motion of the hopfion. During the network training, the position of the hopfion is tuned to generate the proper scattering pattern, which results in the optimal network hyperparameters. In this way, our design circumvents the problems in the way of conventional neuromorphic devices, such as the connection density bottleneck and hardwired neural network structure. The whole neural network is now implemented with a single device instead of an array of elementary units. As the device is scaled down greatly, the power consumption is reduced to a large extent as well. Last, the structures of hopfions are reminiscent of proteins or DNA in structural biology. Hopfions have knot-like structures, and it can be very complicated for them to consist of more than one loop. Their knot-like structure reminds us of peptides in a protein or DNA for their topological equivalence to intertwining strings. The exchange interaction can be used to glue spins from different parts of a hopfion together, playing the same role as hydrogen bonds play in forming the secondary structure of a biomacromolecule. In some situations, the magnetic interaction becomes as strong as the bonds of amino acid side chains in a protein, on which a complex 3D shape or, in terms of structural biology, “tertiary structure” is built. It has been confirmed decades ago that the functional diversity of proteins is closely connected with their structural complexity, implying a great potential of hopfions in achieving complex logic operations. Our work on precise manipulation of hopfions is the first step to fulfilling this potential, and it opens the door to the possibility of biological-like information processing in the future."
} | 3,023 |
26387825 | PMC4585669 | pmc | 5,000 | {
"abstract": "We developed a book-shaped triboelectric nanogenerator (TENG) that consists of electrospun polyvinylidene fluoride (PVDF) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) nanofibers to effectively harvest mechanical energy. The dispersed graphene oxide in the PVDF nanofibers acts as charge trapping sites, which increased the interface for charge storage as well as the output performance of the TENG. The book-shaped TENG was used as a direct power source to drive small electronics such as LED bulbs. This study proved that it is possible to improve the performance of TENGs using composite materials.",
"discussion": "Results and Discussion This book-shaped TENG is based on the contact electrification between PVDF as the electronegative material and PHBV as the electropositive material. PVDF and PHBV were carefully chosen according to the triboelectric series ( supplementary information, Fig. S2 ) 39 to possess a very large difference in their ability to attract and retain electrons, and more importantly, their ease of fabrication into nanofibers. The book-shaped structure of the TENG helps to realize the action of effective charge separation and contact using the natural elasticity of the bent PET substrate ( Fig. 1a ). The working mode of the book-shaped TENG is carried out by applying a cycled compressive force to the whole area of the device during the periodic pressing and release motion of the pressing stage ( Fig. 1b ) so that the top and bottom of the PET substrate coated with PVDF and PHBV nanofibers are periodically pressed into close contact with each other. Once released, the two sides separate automatically because of the stored elastic energy of the PET substrate. The output voltage and current of the fabricated book-shaped TENG under the external force of the pressing stage operating at a frequency of 1.8 Hz were investigated. As displayed in Fig. 2a , under continual pressing and release cycles, the TENG composed of PHBV and PVDF nanofibers without GO repeatedly generated a peak-to-peak voltage of 190 V and peak-to-peak current of 27 μA. Interestingly, the voltage and current increased sharply when a small amount of GO was included in the PVDF nanofibers. The generated peak-to-peak voltage and current reached 260 V and 55 μA, respectively, for the sample containing 0.1% GO in PVDF. The performance of the TENG was further improved by increasing the content of GO in the PVDF nanofibers. The maximum output peak-to-peak voltage and current reached 340 V and 78 μA, respectively, when the GO concentration was 0.7%, which are increases of 78.9% and 188.9%, respectively, compared with those of the TENG containing bare PVDF nanofibers ( Fig. S3 and S4 ). The durability of the book-shaped TENG containing PVDF nanofibers doped with 0.7% GO was assessed, as presented in Fig. 2b . The TENG showed little change in performance after 18000 cycles, indicating that the nanofiber structures were not damaged after thousands of contact-and-release motions, unlike TENGs fabricated by traditional complex or expensive methods such as lithography, etching, and microimprinting 26 . The surface structure of triboelectric materials strongly affects the output voltage of the corresponding TENG 11 15 . Thus, to find out the reason for the enhanced output voltage of the book-like TENG, the surface morphology of the triboelectric materials were studied ( Fig. 3 ). The morphology of the PHBV nanofibers was kept the same in our devices to simplify the analysis, as presented in Fig. 3a . Interestingly, as shown in Fig. 3b–f , the average fiber diameter of the PVDF/GO nanofibers decreased from 650 nm for pure PVDF to 400 nm for PVDF containing 0.7% GO ( Fig. 4a ). Therefore, the enhanced output voltage of the TENG with increased GO content was possibly caused by the diameter of the PVDF/GO fibers decreasing as their content of GO increased. Thus, a control experiment was conducted to verify this conjecture. A TENG containing bare PVDF nanofibers with an average diameter of 250 nm was prepared to examine the potential effect of fiber diameter ( Fig. 4b ). If the performance of the TENGs had nothing to do with GO, the generated peak-to-peak voltage of this device should have been over 340 V according to the curve in Fig. 4a . However, the actual output peak-to-peak voltage of the TENG containing 250-nm PVDF nanofibers was 245 V. That is, the peak-to-peak voltage generated by bare PVDF nanofibers only increased by 28.9% (from 190 to 245 V) when the diameter was decreased by 61.5% (from 650 to 250 nm) ( Fig. 4c,d ). This result indicates that the decrease in diameter of the bare PVDF nanofibers did improve the output voltage of the TENG to some extent, but the decrease in diameter was not the main origin of the improved performance. To further identify the contribution of GO to the performance of the TENGs, a series of PVDF/GO cast films without any nanostructures on their surfaces were fabricated ( Fig. S5 ) and the output voltages of the TENGs containing these film were measured ( Fig. S6 ). The increase of voltage of these devices was in accordance with that of the PVDF/GO nanofiber-based TENGs. This indicated that the voltage increase was not caused by GO changing the morphology of the samples. Therefore, it was the GO in the PVDF nanofibers or cast films that played an important role in the performance enhancement of the TENGs. The mechanisms of TENGs are based on triboelectrification and electrostatic induction, as shown in Fig. S7 . Electrons were induced to flow back and forth through the external circuit under the driving force of the change in triboelectric potential between the PHBV and PVDF/GO nanofibers, which produces the observed voltage–current behavior ( Fig. S3c ) until the TENG reaches electrical equilibrium ( Fig. S7d and b ). Niu et al. 40 reported that the output voltage and current of a TENG are positively correlated with surface charge densities. Therefore, we can deduce that for a specific TENG, the greater the charge distributed on the surface of the triboelectric materials, the higher the surface potential will be, which leads to a larger driving force for the transfer of electrons as well as the higher observed voltage and current. The surface potential of the PVDF/GO nanofibers that resulted from the triboelectric effect between PVDF/GO and PHBV nanofibers was characterized to confirm this deduction. Fig. 5(a,b ) reveal that the surface potential of the PVDF/GO nanofibers becomes larger as the concentration of GO increases, which is in accordance with the output voltage. In addition, the decay of the surface potential is slowed down when the PVDF nanofibers contain more GO. This phenomenon indicates that the improved output voltage of the TENG is caused by the increased surface charge of PVDF/GO nanofibers that resulted from the dispersed GO in the PVDF nanofibers. The improved surface potential (charge) of PVDF/GO compared with that of PVDF alone can be explained as follows. GO is one of the most important derivatives of graphene and possesses a layered structure composed of a carbon network of hexagonal rings and oxygen functional groups on the basal planes and edges 41 42 . The dispersed GO in the PVDF nanofibers acts as charge trapping sites, which increased the interface for charge storage. Electrons attracted from PHBV nanofibers through the friction process were stored either in the discrete, quantized levels of these nanosized graphene particles, or trapped in the amorphous GO dielectric 43 44 , as shown in Fig. 5c–e . This both increased the surface charge on the PVDF/GO nanofibers, and slowed the dissipation of surface charge. The charge storage property of GO was another reason for the decrease of diameter of the PVDF/GO fibers. The factors that influence the diameter of electrospun nanofibers have been studied extensively 34 45 46 47 , and include: (1) the intrinsic properties of the polymer and solvent such as the type and molecular weight of the polymer, and the polarity and boiling point of the solvent; (2) solution parameters such as its concentration, viscosity, and electrical conductivity; (3) operating conditions including the strength of the applied electric field, feed rate, and the collection distance; (4) environmental conditions like temperature and relative humidity. However, in our case, the only factors influencing the fibers were the solution parameters; all other parameters were fixed in the spinning process. Thus, we characterized the viscosity and electrical conductivity of the PVDF spinning solutions containing different weight ratios of GO. However, the changes of electrical conductivity ( Table S1 ) and viscosity ( Fig. S8 ) of the PVDF/GO solutions with different contents were so small that these factors cannot markedly affect the fiber diameter. Instead, as discussed above, the dispersed GO sheet could be easily charged when a high voltage was applied to the spinning solution, leading to increases of electrostatic repulsion and Coulombic force on the Taylor cone, which caused the fiber diameters to decrease. For self-powered systems, the energy generated by a TENG needs to be managed. Fig. 6a shows a full-wave rectifier bridge and 2.2-μF capacitor to store the energy generated by a TENG. Fig. 6(b,c ) depicts the accumulated charges and energy across the capacitor over 5 min when powered by different TENGs. As the GO content of the PVDF nanofibers increased, so did the charging power of the TENG, which is in accordance with the output voltage. The charge and energy generated by the bare PVDF nanofibers after 5 min were 174.3 μC and 6.9 mJ, respectively, giving an average charging current of 0.581 μA and average power of 0.023 mW, respectively. In contrast, the charge and energy of PVDF nanofibers containing 0.7% GO were 431.3 μC and 42.3 mJ, respectively. These values gave an average charging current of 1.438 μA and average power of 0.141 mW, which are increases of 147.5% and 513.0%, respectively. In practical use, the output power for a load depends on the resistance of the load itself. Therefore, we characterized the variations in output voltage and current of a working TENG composed of PVDF containing 0.7% GO under different external loads from 0.1–40 MΩ, as shown in Fig. 6d . As load resistance was increased, the voltage rose and was saturated at about 200 V, which is consistent with the output voltage of the device, while the current decreased because of the ohmic loss. As displayed in Fig. 6e , the instantaneous power on the load reached a maximum value of 4.5 mW at a load resistance of 8 MΩ, corresponding to a power density of 2.3 W/m 2 . To demonstrate the feasibility of the book-shaped TENG as a direct power source for electronics, a total of 113 commercial LEDs connected in series to form a character sequence of DHU-MSE were used as an external load ( Supplementary Video 1 ). As illustrated in Fig. 6f , a bridge rectifier was used to convert the AC output signals into DC signals to ensure that the LEDs could be lighted by the TENG in both the pressing or release state. The photograph of the flashing LEDs was captured while the TENG was being pressed at a frequency of 1.8 Hz by the pressing stage. In summary, a simple book-shaped TENG composed of PVDF and PHBV nanofibers was developed to harvest mechanical energy. The TENG was fabricated by electrospinning using a facile and cost-effective process. GO sheets were introduced to increase the charge storage ability of the PVDF nanofibers and enhance the output performance of the TENG. The maximum output peak-to-peak voltage and current were 340 V and 78 μA, respectively, and the average charging current and power for charging a 2.2-μF capacitor were 1.438 μA and 0.141 mW, respectively, which are 78.9%, 188.9%, 147.5% and 513.0%, respectively, higher than those of a TENG with neat PVDF nanofibers. The energy generated could either be stored or directly used to drive small electronics and the performance of the TENG was quite stable over 18000 cycles. Importantly, this study demonstrated groundbreaking progress in the enhancement of the output power of TENGs by using inorganic/organic hybrid materials and a simple electrospinning method, which may open new avenues of research for triboelectric materials."
} | 3,084 |
38486283 | PMC10941469 | pmc | 5,004 | {
"abstract": "Glycolate is produced by microalgae under photorespiratory conditions and has the potential for sustainable organic carbon production in biotechnology. This study explores the glycolate production balance in Chlamydomonas reinhardtii , using a custom-built 10-L flat panel bioreactor with sophisticated measurements of process factors such as nutrient supply, gassing, light absorption and mass balances. As a result, detailed information regarding carbon and energy balance is obtained to support techno-economic analyses. It is shown how nitrogen is a crucial element in the biotechnological process and monitoring nitrogen content is vital for optimum performance. Moreover, the suitable reactor design is advantageous to efficiently adjust the gas composition. The oxygen content has to be slightly above 30% to induce photorespiration while maintaining photosynthetic efficiency. The final volume productivity reached 27.7 mg of glycolate per litre per hour, thus, the total process capacity can be calculated to 13 tonnes of glycolate per hectare per annum. The exceptional volume productivity of both biomass and glycolate production is demonstrated, and consequently can achieve a yearly CO 2 sequestration rate of 35 tonnes per hectare. Although the system shows such high productivity, there are still opportunities to enhance the achieved volume productivity and thus exploit the biotechnological potential of glycolate production from microalgae. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-024-02479-4.",
"conclusion": "Conclusions In summary, the presented system can sequester 15 t ha −1 y −1 CO 2 in glycolate and about 20 t ha −1 y −1 CO 2 in biomass, resulting in a total annual CO 2 sequestration rate of 35 t ha −1 y −1 of CO 2 . It can be further concluded that photosynthesis in the biomass phase is at its maximum and cannot consume all the CO 2 supplied, whereas during the glycolate production the added CO 2 is completely consumed. Both biomass and glycolate production exhibit exceptional volume productivities. The total capacity of the process can be calculated to 13 t ha −1 y −1 of glycolate per year. Although the system is already set up very efficiently using artificial light with high efficiency, there are still possibilities to enhance the achieved volume productivity.",
"discussion": "Discussion The role of nitrogen in the glycolate production system Before the glycolate production was started, a high-density biomass culture was grown to a maximum biomass of 3.57 g dry weight l −1 . The rapid reduction of nitrogen in the medium during this growth phase can be attributed to culture growth, since nitrogen is essential for the increase of cell numbers and Chl a . To prevent N-limitation, additional N was supplied, resulting in a total of 14 mM NH 4 + . Assuming that the carbon content of the biomass is 50% [ 21 ], this results in a maximum C:N ratio of 10.6, if the supplied nitrogen is completely taken up by the cells. Under adequate nitrogen supply, the C:N ratio in Chlamydomonas should be about 6, although it can increase to 14 under nitrogen deficiency [ 23 ]. It can, therefore, be assumed that the cells are already N-limited despite the addition of nitrogen. This conclusion is supported by the observed steady decrease in photosynthesis with increasing nitrogen deficiency, which has been reported previously [ 24 , 25 ]. While sufficient nitrogen availability during biomass growth is essential for the formation of proteins, pigments and nucleic acids, a nitrogen deficiency in the glycolate phase may have certain advantages. Nitrogen deficiency should inhibit the C2 cycle, in which glyoxylate, the downstream product of glycolate, is further converted to the amino acid glycine. This reaction requires an NH 2 group from alanine or glutamate as a donor [ 26 , 27 ]. Nitrogen deficiency, therefore, limits the availability of donors and favours the accumulation of large amounts of glycolate in the cell. These elevated glycolate concentrations must be detoxified by excretion and consequently lead to accelerated glycolate accumulation in the medium. In this study, however, additional nitrogen supply during the glycolate phase in run 1 had no significant effect on glycolate production (Fig. 3 F). In contrast to the biomass phase, where nitrogen was quickly depleted, the nitrogen added during the glycolate production phase declined more slowly and was still detectable after 24 h. Along with this, no changes in biomass growth were observed with the addition of N. In this study, NH 4 + was used as the preferred nitrogen source for Chlamydomonas due to its reduced state and, therefore, energetically favourable assimilation. Thus, the electrons obtained from the light reaction, which are normally required for the reduction of other nitrogen species (i.e., NO 3 − ), are available for the photosynthetic reduction of CO 2 . This increased CO 2 fixation usually leads to higher growth rates during the growth phase [ 28 ]. During the glycolate phase, however, only small amounts of nitrogen are required because biomass growth is strongly inhibited since the assimilated carbon is invested in glycolate excretion rather than cell proliferation. The amount of nitrogen, therefore, plays only a marginal role. In Chlamydomonas , ammonium is the preferred N source for growth [ 29 ] and was, therefore, used for this study. However, it is well known that nitrate is also an efficient N donor in Chlamydomonas reinhardtii, with important impacts on physiological performance. First, NO 3 − is an additional electron sink and can reduce carbon assimilation. Second, nitrate assimilation depends on a complex regulatory network of transporters and enzymes located in different cellular compartments [ 30 ]. Third, nitrate assimilation increases the pH during growth, in contrast to the decreasing pH during ammonium assimilation [ 31 ], so the pH control needs to be adjusted. Such an increase in pH may also be beneficial as it counteracts the acidification of the medium caused by glycolate production, thus reducing the amount of added KOH in the glycolate phase. As an alternative, the use of NH 4 NO 3 as nitrogen source, which combines the different properties of NH 4 + and NO 3 − , may be advantageous [ 32 ]. Further studies on the cellular degree of reduction and an optimal stoichiometry between the different nitrogen species could help to stabilize the system. In conclusion, nitrogen is a critical component in biotechnological batch processes for glycolate production. In the future, pathway analysis can help to further elucidate the importance of nitrogen in glycolate metabolism, for example, using mutants of N-related pathways together with sophisticated metabolomics analysis. However, it is clear that monitoring nitrogen levels in batch processes is essential and that the amount of nitrogen should be carefully calculated to ensure optimum performance. Dissolved oxygen pattern during whole cultivation The oxygen concentration in the reactor is influenced by various overlapping biological and physical processes. During the biomass phase, the concentration of externally added oxygen accounts for 19%, corresponding to the atmospheric oxygen content. However, the growing biomass with high photosynthetic activity causes the oxygen concentrations in the medium to increase up to 40%. The drop in oxygen that follows the addition of nitrogen is not pH dependent, as the O 2 decrease was observed in both runs, but only in one run the pH dropped (in the other run it was counteracted with KOH) (Fig. 2 ). Thus, the conclusion is obvious that the oxygen drop is caused by the short-term high NH 4 + concentrations. Indeed, studies on chlororespiration have shown that 5 mM NH 4 + uncouples the proton gradient across the thylakoid membrane [ 33 ]. This uncoupling very quickly leads to an ATP bottleneck, consequently inhibiting the photosynthetic electron transport and O 2 production. The large nitrogen deficit and the high biomass loading apparently result in a rapid removal of excess NH 4 + within a short time (30 min), and the original oxygen rates of 40% can be restored (Fig. 2 a). It is assumed that oxygen inhibition starts to occur above 40%, which easily can be reached in dense and fast growing cultures, especially in closed PBRs [ 34 ]. Such high oxygen is not desirable, as they can accompanied by a decrease in biomass productivity and microalgal growth [ 35 ]. In this study, however, such high oxygen concentrations do not seem to influence growth and photosynthesis. After the initial rise to 40%, oxygen concentrations subsequently dropped rapidly to 33%. This decrease occurred at the same time as nitrogen in the medium was almost completely depleted. This suggests that the decrease in oxygen production is primarily due to reduced photosynthetic activity caused by nitrogen deficiency and probably not due to the harmful oxidative stress. Oxidative stress can lead to processes such as photoinhibition, since it is accompanied by the production of reactive oxygen species (ROS) such as singlet oxygen ( 1 O 2 ), superoxide radicals (O2 •– ) or hydrogen peroxide (H 2 O 2 ). Due to their highly reactive nature, they can damage cellular components i.e., membrane lipids, DNA and photosystem II (PSII) [ 36 ]. ROS are formed when the excited PSII is no longer able to discharge electrons via the electron transport chain to a suitable acceptor, and instead the electrons are transferred to oxygen. To prevent this, Chlamydomonas has developed several mechanisms to distribute or dissipate the absorbed light energy. This can be achieved, for example, through state transitions, where a fraction of the PSII outer antenna is transferred to PSI, balancing the excitation energy in both photosystems [ 37 ]. Other mechanisms for the redistribution and dissipation of excessive energy within the PSII include zeaxanthin-dependent heat dissipation (NPQ) via the xanthophyll cycle, or the use of alternative electron transport pathways [ 38 , 39 ]. However, the maximum efficiency of PSII is unchanged, as no significant decrease in F v /F m levels is observed during this time. This indicates that the structure of the PSII is still largely preserved, which is consistent with the finding that short-term oxidative stress with low levels of ROS has no detectable effect on photosynthetic efficiency in Chlamydomonas . However, higher levels of ROS would decrease F v /F m levels and inhibit photosynthetic efficiency [ 40 ]. Chlamydomonas possesses an efficient scavenging system for ROS and uses the increasing presence of ROS as a signal to adapt to high oxygen stress [ 41 ]. This means that Chlamydomonas can tolerate these short-term elevated oxygen levels caused by its own photosynthetic activity. Despite the high concentration of dissolved O 2 during the biomass phase, no glycolate excretion could be detected. There are two main reasons for this: First, the addition of CO 2 increases the C:O ratio and thus counteracts photorespiration. Under such optimal CO 2 supply the rate of photorespiration is drastically reduced. Secondly, any glycolate produced is quickly metabolized in the C2 cycle. The released carbon is recovered and fed to the carboxylation reaction at the Rubisco. This scenario changed with the change in gassing and the addition of the inhibitor. The O 2 partial pressure was kept high above 30%. However, the C:O ratio changes drastically, making photorespiration much more pronounced, and because CCMs are inhibited, an increase in glycolate is observed. However, F v /F m decreases only slightly, indicating that there is no massive accumulation of ROS and that the effects of photoinhibition on PSII function and performance are limited. Despite the high O 2 concentrations in the medium, the cells are still able to produce glycolate with high efficiency over a long period of time. As the O 2 level is critical to the performance of the cells, it is advantageous to measure and monitor the O 2 concentration inside the reactor. While a high O 2 supply promotes photorespiration and glycolate production during the glycolate phase, a maximum dissolved O 2 level of 40% should not be exceeded during the growth phase. However, in closed PBRs, high dissolved O 2 concentrations can be reached quickly by the photosynthetic activity alone. Outgassing of O 2 is slower than O 2 production by photosynthesis in the growth phase. In contrast, gassing and photosynthesis during glycolate production lead to constant O 2 levels. Furthermore, the gas inflow rates are not high enough to promote increased outgassing. Efficient reactor design and novel design concepts can facilitate the targeted outgassing [ 42 ]. Integrated airlift systems and advanced degassing units can increase oxygen mass transfer and help evacuate oxygen from the cultivation system. Here, a good reactor design with optimal control of gas flow is advantageous. Oxygen levels slightly above 30% proved to be a good value for glycolate production without influencing photosynthetic efficiency. Balancing with absolute measured values enables structured modelling of upscaling Photosynthesis-based biotechnological processes can be used to fix and reduce CO 2 from the atmosphere, which can help to fight climate change. In this study, we have successfully evaluated carbon fixation in terms of carbon use efficiency for both the initial biomass growth and the subsequent carbon fixation into glycolate as the biotechnological product. The technical conditions of the reactor design enable the determination of carbon utilisation in individual reactor units, which in themselves can be regarded as individual parts of a large-scale plant. The glycolate production rate observed in this reactor was 7.5 mM d −1 , or 667 mg l −1 d −1 , which is similar to the efficiency previously measured in small shallow reactors, where 500 mg l −1 d −1 was measured [ 11 , 12 ]. These rates, together with the technical equipment of the reactor, allowed a correlation between the amount of CO 2 inflow and the amount of carbon fixed in the product at the 1m2 reactor scale. This correlation was then extrapolated to larger scales, up to the ha range. In general, the carbon mass balances obtained showed high carbon fixation of the total CO 2 supplied during the biomass (25%) and glycolate (97%) production phases. Our data clearly show that carbon sequestration is much more efficient during glycolate production than during biomass formation. The low CO 2 use efficiency of 25% during the biomass production phase is mainly due to two reasons: the changes in biomass and the CO 2 concentration of the aeration. On one hand, the low initial biomass loading combined with the high CO 2 influx leads to low total CO 2 fixation rates relative to the amount of CO 2 available. On the other hand, gassing the system with 2% CO 2 to ensure high growth rates leads to CO 2 surplus and thus high CO 2 efflux rates. Thus, 75% of the CO 2 input is lost through emission during the biomass production phase. Further enhancement of the already high growth rates by increasing the CO 2 supply would only result in higher growth if the cells were carbon limited at the current 2% CO 2 supply. Considering that the Q Phar at high biomass concentrations is close to 100% (total absorption of incident light), it can be concluded that no CO 2 limitation occurred during the biomass phase. As a consequence, the CO 2 supply can be adjusted to lower concentrations, e.g., as a dynamic function with increasing biomass, to optimise the process. In the future, a fully realized circular economy regarding aeration is, therefore, mandatory to effectively capture and recycle CO 2 . In contrast to the biomass phase, during the glycolate production phase, almost all of the supplied CO 2 is fixed and incorporated into the product glycolate. Compared to the frequently used biomass approaches, we have used a highly sophisticated alternative method to direct the fixed carbon directly into the product. To the best of our knowledge such high rates of CO 2 fixation in the product as achieved with the presented system have not been demonstrated before with other approaches. However, the total photosynthetic carbon fixation during the glycolate phase (8.8 mg C l −1 h −1 ) was only half of that in the biomass phase (14.6 mg C l −1 h −1 ). This is mainly due to the lower CO 2 influx (0.2% CO 2 compared to 2% CO 2 ) required to induce glycolate production by altering the CO 2 :O 2 ratio at Rubisco. As the fixed carbon is completely incorporated into glycolate as the product, a shortage of C in the cell metabolism is to be expected. At the same time, the overall photosynthetic capacity did not change much. Since Q Phar was largely unchanged during the glycolate production phase, it is likely that the reduced CO 2 fixation resulted in a massive electron excess in cell metabolism accompanied by a reduced consumption of NADPH and ATP. Such a high electron excess triggers two responses: the induction of photoprotective mechanisms on the one hand and photoinhibition on the other. Due to the unchanged photosynthetic capacity, it can be concluded that the protective mechanisms such as alternative electron transport pathways and increased energy dissipation are sufficient to prevent photoinhibition. This raises the question of how the cell maintains its physiological health as measured by F V /F M (Fig. 3 E). We observed a steady decrease in the previously built biomass and consequently a loss of fixed carbon of up to 5.7% (Table 1 ). Due to the lack of CO 2 , it is very likely that the regeneration of ribulose-1,5-bisphosphate in the Calvin cycle is strongly limited. Thus, the loss of biomass during glycolate also indicates the reductive pentose phosphate pathway is used to ensure the supply of carbon skeletons in the form of ribulose-1,5-bisphosphate at the Rubisco. Fine tuning of the CO 2 supply could minimise this biomass loss and lead to a more stable process. Besides environmental compatibility, the aim of intensive research for biotechnological applications has been to maximize the time-per-space yield (volumetric productivity) of biomass production or a product [ 11 ]. High volumetric productivity of biomass production in photoautotrophic reactors ranges from 15 up to 21 mg DW l −1 h −1 in Chlamydomonas [ 11 , 43 ], depending on the setup. In contrast, with mixotrophic cultivation of Chlamydomonas , slightly higher biomass time-per-space yields have been reported, reaching values between 5 and 29 mg DW l −1 h −1 [ 44 ]. These biomass productivities correspond to a maximum carbon fixation of 2.5 up to 14.5 mg C l −1 h −1 , assuming 50% carbon content in the biomass. In comparison, the setup used here showed a biomass volume productivity of VP C = 14.6 mg C l −1 h −1 (29.2 mg DW l −1 h −1 ) in photoautotrophic conditions during biomass growth Table 3 . Up-scaling towards an illuminated reactor area of 1 ha would result in VP C = 452 kmol C ha −1 y −1 (10.9 t biomass ha −1 y −1 ). Assuming a continuous production cycle without glycolate production phases, where these biomass rates can be maintained for one year with artificial illumination, an annual yield of 25.5 t of biomass can be achieved. In comparison, under natural light conditions in Tuscany (Italy), the cells receive > 2 times more photons per m2 and day for eight months of operation. Within this time period, the productivity of a green algae Tetraselmis suecica was calculated to 36 t, assuming an average productivity of 15 g biomass m −2 d −1 [ 45 ]. Additionally, assuming very high photosynthetic efficiencies, up to 64 t ha −1 of microalgae biomass production have been calculated [ 46 ]. This means that real biomass production reported here is relatively close compared to the theoretically calculated rates. It can be concluded from this that an increase in the amount of light in the system presented has the potential to significantly increase the rates. For a practical industrial application, however, the crucial parameter is the volume productivity of the actual end product. The observed high volume productivity for the product of VP Glycolate is about 27.7 mg glycolate l −1 h −1 , (0.66 g glycolate l −1 d −1 ) during the glycolate phase. This resulted in a VP Glycolate of 0.36 g glycolate l −1 d −1 if calculated for an entire year, taking into account the time required for biomass phase, harvesting and cleaning. Compared to Chlamydomonas mass cultivation for lipid production with a volume productivity of about 0.035 up to 0.11 g lipid l −1 d −1 , glycolate production showed a 3 to tenfold higher carbon fixation efficiency for the product [ 43 , 47 ]. Other fast-growing green algae have been calculated to reach higher lipid volume activities of about 0.15 – 0.25 g lipid l −1 d −1 [ 48 ], however, these values are still below the glycolate production observed in this study. In addition, the volume productivity of glycolate production would also be much higher compared to other high-value products, such as astaxanthin, phycocyanin and other pharmaceuticals, since these products generally represent a much lower mass proportion of the dry weight of the biomass [ 49 ]. Taken together, glycolate production is a very efficient system, outperforming lipid production in terms of productivity per time and space. The recorded rate of glycolate production is inherently comparable to a high rate of carbon fixation for glycolate, with VP C = 8.7 mg C l −1 h −1 . As CO 2 is an important and expensive resource and will become even more so in the future, this efficient use of carbon in glycolate production reduces the economic burden of CO 2 supply. This is even more important if the excessive CO 2 is not recycled but emitted into the atmosphere. The high carbon utilisation rate during the glycolate production phase also contributes to the environmental sustainability of this approach, as the injected CO 2 does not escape from the system, preventing the need for refixation. This is not the case with many other biomass-based applications that use flue gas or high levels of CO 2 for aeration. A more comprehensive understanding of the presented system will enable further assumptions for improvements that need to be considered in future research. In terms of carbon fixation rates, increasing CO 2 gassing during glycolate production could improve product formation, but may lead to increased CO 2 :O 2 ratios at the Rubisco and, therefore, reduced photorespiration rates. This can possibly be counteracted by adjusting the culture to higher light intensities so that more CO 2 is fixed and subsequently directed into glycolate. Since the light intensity used is close to the inclination point (Ik) of photosynthesis, algae cells would face more photoinhibition and subsequent photodamage, which would be a disadvantage. As mentioned earlier, higher nitrogen supply can be used as option to increase biomass loading of the reactor. Together with increased light and CO 2 gassing, this could stabilize the rate of photorespiration. Furthermore, higher nitrogen supply may be followed by higher protein content, especially if the amount of Rubisco is increased. This would also lead to a higher throughput of the carboxylation/oxygenation reactions, thereby increasing the productivity of the process. Future work will focus on this aspect. Finally, the overall productivity can also be controlled by the length of the individual phases within one run. Previous work has shown that the glycolate production phase can be extended to 16 days ([ 11 ]). If the current run time for the glycolate production phase can be doubled, the annual volume productivity VP Glycolate will increase from 0.36 g glycolate l −1 d −1 to 0.56 g glycolate l −1 d −1 (Additional file 1 : Table S2), which impressively underpins the potential of the system. Energetic balance—balance is based on Q Phar data The determination of the quantum use efficiencies φC in microalgae cultivation is crucial for optimizing growth, improving cultivation systems and quantifying the photosynthetic performance. It allows for a more efficient use of resources, an increase in productivity, and the development of sustainable solutions. Using the flat panel reactor, it is possible to measure the absorbed radiation (Q Phar ), since its specific geometric design has a defined layer thickness across its entire surface, ensuring the homogeneous distribution of both the cells and the irradiated radiation. The determination of Q Phar is necessary to subsequently measure the quantum efficiency φC as photons per C fixed. A theoretical value of 8 photons per C can be calculated for φC, but it has been shown that the actual φC measured in biotechnological microalgae applications is generally much higher (Table 3 ) [ 50 ]. With 30 mol photons per mol C, the φC in the biomass phase is extraordinarily efficient and exactly corresponds to the values calculated in [ 50 ]. However, glycolate production shows significantly lower efficiencies, with 113 mol photons mol C −1 . This shows that under the conditions of glycolate production, a significantly higher amount of the absorbed photons is not used for the accumulation of C in organic compounds (biomass or glycolate). Instead, the energy is dissipated, e.g., as heat or fluorescence, used for alternative electron transport pathways (e.g., Mehler reaction), for regeneration of ribulose-1,5-bisphosphate via reductive pentose phosphate pathway, or consumed by respiration. From an energetic point of view, an increase in efficiency can best be achieved by a targeted reduction of these losses. The overall good φC together with the high C fixation rates (chapter 4.3) can only occur if sufficient energy is provided to meet the demand for CO 2 fixation. Within the presented system, the energy ultimately required to illuminate the reactor with a surface area of 1 m 2 was measured to be 3.4 kWh. If this scale is increased to 1 ha, the annual energy demand would reach 10.3 GWh (Table 3 ). The average incident solar flux is between 100 and 270 W m −2 [ 51 ]. This corresponds to an annual available energy supply of 8.8 – 23.7 GWh, of which only 45% account for PAR [ 52 ]. These numbers show that, from an energetic perspective, a glycolate plant with the current volumetric productivity is physically limited. Only small parts of the world qualify as suitable location due to sufficient light availability. Increasing the volumetric productivity is, therefore, rarely feasible when using incident solar radiation as the light source. In addition, under natural sunlight, fluctuating light intensities significantly affect the photosynthetic activity and electron usage of different physiological pathways of microalgae [ 38 ]. Therefore, artificial illumination is a viable option to set up a glycolate production facility. Moreover, it enables a constant process of glycolate production without fluctuations in light intensity or temperature. In addition, it offers the possibility of 24-h illumination, without constraints on the orientation of reactors and installation sites. Economic feasibility Taking into account the data presented, the calculation shows that even with this improved technological path, the critical level for economic production of low-cost products has not yet been reached. Additional energy would be required for various reactor operations, such as aeration and heating, and the subsequent harvesting and processing of the product. To conduct a comprehensive techno-economic analysis, it is necessary to incorporate these energy inputs. Real-world data on the operation and economic feasibility of various microalgal production systems for high-value microalgal products are available [ 53 ], but cannot be directly applied to the glycolate technology. The different production process, which focuses on continuous production while using microalgae as a catalyst to very efficiently convert CO 2 and light into the product by minimizing biomass growth, comes with low CO 2 aeration and unique requirements for harvesting and processing. These characteristics outperform traditional biomass production approaches and must be taken into account. However, in particular for high-volume, low-value products such as glycolate, estimated production costs can vary by a factor of 100 depending on the calculations and production process [ 54 ]. Therefore, the next step will be to evaluate the technology at the pilot-scale production level. In general, maximizing the value of a biomass producing approach needs extracting of a diverse product portfolio to facilitate economic feasibility [ 55 , 56 ]. In contrast, the microalgal glycolate production system yields a very high product excretion rate in relation to the residual usable biomass and, therefore, has significant advantages in terms of feasibility."
} | 7,291 |
38313033 | PMC10836209 | pmc | 5,005 | {
"abstract": "Anthropogenic activities increase sediment suspended in the water column and deposition on reefs can be largely dependent on colony morphology. Massive and plating corals have a high capacity to trap sediments, and active removal mechanisms can be energetically costly. Branching corals trap less sediment but are more susceptible to light limitation caused by suspended sediment. Despite deleterious effects of sediments on corals, few studies have examined the molecular response of corals with different morphological characteristics to sediment stress. To address this knowledge gap, this study assessed the transcriptomic responses of branching and massive corals in Florida and Hawai‘i to varying levels of sediment exposure. Gene expression analysis revealed a molecular responsiveness to sediments across species and sites. Differential Gene Expression followed by Gene Ontology (GO) enrichment analysis identified that branching corals had the largest transcriptomic response to sediments, in developmental processes and metabolism, while significantly enriched GO terms were highly variable between massive corals, despite similar morphologies. Comparison of DEGs within orthogroups revealed that while all corals had DEGs in response to sediment, there was not a concerted gene set response by morphology or location. These findings illuminate the species specificity and genetic basis underlying coral susceptibility to sediments.",
"conclusion": "Conclusion This study incorporated data from two separate experiments to more fully characterize the molecular mechanisms induced by sedimentation in corals. We found that developmental processes are primarily impacted in branching corals across location, highlighting potential future research avenues with regards to sediment stress and reproductive potential and output. While few specific genes were shared across morphology and location, orthogroup analysis uncovered potential overlap in generalized sediment stress responses. Direct comparisons across species are necessary to further elucidate the genetic basis of coral susceptibility to sediment stress.",
"introduction": "Introduction Coral reefs are incredibly diverse marine ecosystems, providing numerous ecological and economic services such as biodiversity, cultural value, coastal protection, fisheries, and tourism ( Reaka-Kudla, 1997 ; Sumaila & Cisneros-Montemayor, 2010 ; Costanza et al., 2014 ). Reef-building corals form a critical nutritional symbiotic relationship with unicellular photosynthetic algal endosymbionts in the family Symbiodiniaceae ( LaJeunesse et al., 2018 ). The carbohydrates produced by the algal photosynthesis are translocated to the coral to be used as its primary energy source, supporting the daily respiratory carbon demand of tropical corals ( Muscatine & Porter, 1977 ; Muscatine et al., 1984 ). This coral-algal symbiosis fuels reef productivity and accretion ( Roth, 2014 ), but is sensitive to changing environmental conditions that can impact the symbiosis such as light, nutrients, temperature, pH, and sediment ( Hoegh-Guldberg et al., 2007 ; Davy, Allemand & Weis, 2012 ). For example, under exposure to sedimentation, or the downward fall of sediment from the water column toward the benthos ( Schlaefer, Tebbett & Bellwood, 2021 ), corals display reduced photosynthetic efficiency ( Weber, Lott & Fabricius, 2006 ; Rushmore, Ross & Fogarty, 2021 ), increased respiration rates ( Riegl & Branch, 1995 ; Browne et al., 2014 ), decreased calcification ( Bak, 1978 ), and rapid consumption of energy reserves ( Sheridan et al., 2014 ). While sediment transport naturally occurs on reefs, suspended sediment caused by anthropogenic activities such as dredging, runoff, and coastal development have increased ( Rogers, 1990 ; Fabricius, 2005 ; Erftemeijer et al., 2012 ; Miller et al., 2016 ; Cunning et al., 2019 ). Deposited sediment and suspended sediment are the two primary ways that sediment interacts with corals ( Rogers, 1990 ; Fabricius, 2005 ; Erftemeijer et al., 2012 ). Deposited sediment occurs when sediment particles settle directly on the coral surface, making physical contact with the tissue. Passive removal of sediment includes gravity or flow removing it from the coral surface ( Lasker, 1980 ; Jones, Fisher & Bessell-Browne, 2019 ). In response to deposited sediment, corals can also initiate an acute response to attempt to move the sediment using active mechanisms. Active sediment removal mechanisms include ciliary and tentacle movement, increased mucus production, and hydrostatic inflation ( Rogers, 1990 ; Stafford-Smith & Ormond, 1992 ; Stafford-Smith, 1993 ; Bessell-Browne et al., 2017 ). However, these active mechanisms are often very energetically costly, and thus cannot be sustained for long periods of time ( Riegl & Branch, 1995 ; Erftemeijer et al., 2012 ). If the sediment deposition rate exceeds the coral’s sediment clearance rate, sediment will accumulate on the coral, reducing heterotrophic feeding and light transmission to algal endosymbionts and creating hypoxic conditions near the coral tissue, which often leads to tissue necrosis and coral mortality ( Philipp & Fabricius, 2003 ; Weber, Lott & Fabricius, 2006 ; Weber et al., 2012 ). Corals can also, or alternatively, interact with suspended sediment, which occurs when particles such as clay, silt, and sand are moved into the water column by some natural or anthropogenic disturbance and remain in the water column for a period of time ( Rogers, 1990 ; Fabricius, 2005 ; Erftemeijer et al., 2012 ). Suspended sediment reduces the amount of light that reaches the coral, impeding the ability of the algal endosymbionts to photosynthesize and provide the coral host with sufficient energy for metabolism and growth ( Rogers, 1990 ; Fabricius, 2005 ; Erftemeijer et al., 2012 ; Bessell-Browne et al., 2017 ). Reduced photosynthetic efficiency can induce corals to switch to heterotrophic feeding, a much less efficient way to obtain carbon than through its endosymbionts ( Muscatine & Porter, 1977 ; Anthony & Fabricius, 2000 ; Houlbrèque & Ferrier-Pagès, 2009 ). Additionally, heterotrophic feeding in the presence of sediments may lead the coral to ingest sediment particles, disrupting its nutritional intake and potentially acting as a vector for harmful bacteria and toxins ( Erftemeijer et al., 2012 ; Studivan et al., 2022 ). Suspended sediment has also been observed to induce immune responses and increase disease prevalence in corals ( Pollock et al., 2014 ; Sheridan et al., 2014 ). In addition to sediment type, the morphology of the coral can modulate its interaction with sediments. For example, massive, plating, and encrusting corals have a higher planar surface area and thus higher capacity to trap sediments in comparison to branching corals with high three-dimensional and more vertical structure. Sediment removal from massive corals often requires active removal mechanisms ( Dallmeyer, Porter & Smith, 1982 ; Rogers, 1990 ; Stafford-Smith, 1993 ). However, massive corals may also be more resilient to high suspended sediment concentrations because their greater surface area allows for increased opportunities to capture light, maximizing the photosynthetic efficiency of their algal endosymbionts ( Fabricius, 2005 ; Erftemeijer et al., 2012 ). In contrast, the relatively small surface area and vertical branches of branching coral species means sediment is minimally trapped and can be more easily removed by gravity or currents ( Lasker, 1980 ; Rogers, 1983 ; Stafford-Smith, 1993 ). Branching coral species typically have faster clearance rates than non-branching species, and active removal mechanisms are required less frequently, allowing branching corals to devote that energy towards other functions such as reproduction and growth ( Stafford-Smith, 1993 ; Jones, Fisher & Bessell-Browne, 2019 ). Collectively, these studies support that morphology plays a role in response to sediment stress. While most research has focused primarily on physiological responses to sediment stress, there is a small but growing body of work on gene expression of corals exposed to sediment stress. Early microarray studies found that the upregulation of heat shock proteins (HSPs) occurred in response to sediment stress ( Wiens et al., 2000 ; Hashimoto et al., 2004 ). As a generalized stress response protein, HSPs have also been implicated in coral response to other stressors, such as deoxygenation, temperature and ocean acidification ( DeSalvo et al., 2008 ; DeSalvo et al., 2010 ; Kaniewska et al., 2012 ; Alderdice et al., 2022 ). Another general response to sediment stress is upregulation of biomarkers of oxidative stress ( Morgan, Edge & Snell, 2005 ), more specifically, thioredoxin, a protein that modulates redox and cell-to-cell signaling ( Tomanek, 2015 ). A potentially more specific gene responding to sediment stress is indicated by the differential expression of urokinase plasminogen activator surface receptor (uPAR) transcripts in the coral Diploria strigosa along a sedimentation/pollution gradient in Castle Harbor, Bermuda ( Morgan, Edge & Snell, 2005 ). uPAR is associated with proteolysis, wound healing and inflammation, and is hypothesized to contribute to coral tissue remodeling in response to elevated levels of sedimentation ( Morgan, Edge & Snell, 2005 ). Genes related to immunity, as well as energy metabolism, were also implicated in the transcriptomic response to sediment stress in corals from two locations, Singapore ( Goniastrea pectinata and Mycedium elephantotus ) and Eilat, Israel ( G. pectinata only) using RNASeq ( Bollati et al., 2021 ). While there were some methodological differences between their experiments, shared mechanisms were identified across different species and populations, demonstrating a conserved response to sediment stress across species and sediment types. However, the corals evaluated in that study, G. pectinata and M. elephantotus , both have similar morphological characteristics (massive and encrusting), indicating that this response may only be relevant to massive and encrusting corals ( Bollati et al., 2021 ). While these initial molecular analyses have provided important insights into the coral transcriptomic response to sediment stress, less is known about shared molecular responses by morphology and location/sediment type. To this end, our study aims to fill these knowledge gaps by examining gene expression across different coral morphologies and the use of multiple locations/types of sediment. Here we quantified the transcriptomic responses of corals with different colony morphologies in response to different types of sediment stress. Floridian corals ( Acropora cervicornis , Montastraea cavernosa and Orbicella faveolata ) were exposed to sterilized white carbonate sediment for 18 days, whereas Hawaiian corals ( Montipora capitata, Pocillopora acuta (formerly Pocillopora damicornis ) and Porites lobata ) were exposed to unsterilized terrigenous red soil for up to 7 days. In this study, A. cervicornis and P. acuta were categorized as branching corals, while M. cavernosa, O. faveolata and P. lobata were categorized as massive corals. The morphology of M. capitata was considered as intermediate between branching and plating, as M. capitata tends to form plates growing horizontally with branches sprouting upward ( Veron, 2002 ). The methodological differences prevent us from making direct statistical comparisons between the experiments. However, it is still relevant to highlight general biological processes and mechanisms related to sediment stress responses across morphology and location.",
"discussion": "Discussion In this study, we conducted two separate experiments to characterize the molecular underpinnings of corals with differing morphological characteristics responding to sediment stressors in two locations, Florida and Hawai‘i. The methodological differences prevent us from making direct statistical comparisons between the experiments. However, it is still relevant to highlight general biological processes and mechanisms related to sediment stress responses across morphology and location. Responses to unsterilized red sediment in Hawai‘i It is well established that morphology can play a role in modulating a coral’s response to sediment stress, but it is unknown if gene expression varies in corals with differing morphological traits. Our study found that, across morphologies in Hawaiian corals, developmental processes were primarily affected by unsterilized sediment. Previous work has demonstrated that sediment deposition, high turbidity, and low light have been shown to adversely affect development and reproductive output across morphologies ( Kojis & Quinn, 1984 ; Gilmour, 1999 ; Humphrey et al., 2008 ), though M. capitata development and fecundity are not always negatively affected by high sedimentation rates ( Padilla-Gamiño et al., 2014 ). In the present study, the branching coral, P. acuta , shared a relatively high number of overlapping developmental processes-related responses with corals with intermediate ( M. capitata ; 20 shared terms) and massive ( P. lobata ; 11 shared terms) morphology. Shared responses corresponded to reproduction and developmental processes like embryonic axis specifications, embryonic organ development, eye development, neural crest cell fate specification, neural plate anterior/posterior regionalization, embryonic hindlimb morphogenesis and others. Unexpectedly, these terms primarily correspond to vertebrate developmental processes, complicating our ability to interpret these results. Given our use of specific annotation databases ( i.e ., NCBI, SwissProt, InterPro), it is possible that these databases are all dominated by vertebrate-related annotations and thus, invertebrate protein sequences are assigned vertebrate-centric annotations. While it is clear that developmental processes are affected by short term sediment stress across morphologies, it is unclear how vertebrate specific gene functions may translate to developmental processes in basal metazoans. More work is necessary to evaluate equivalency across annotation softwares. During the year, M. capitata , P. acuta , and P. lobata develop their gametes and then release when conditions are optimal, typically June and July in Hawai‘i ( Stimson, 1978 ; Richmond & Jokiel, 1984 ; Padilla-Gamiño & Gates, 2012 ; Brown et al., 2020 ). Because the corals in our study were exposed during this time period, it is possible that we captured gene expression signatures unique to corals at or near the peak of their reproductive phenotype. Although the corals were not sampled at different times of year under the same experimental conditions, these results suggest that because the corals were sampled during their reproductive period, signals of developmental processes may be higher than they might have been at other points in the year. Given the high number of shared and unique developmental process GO terms in the Hawai‘i corals, however, it is likely that sediment has the potential to have reproductive and developmental consequences, which can have subsequent impacts on population growth and dynamics. The only specific shared response across the three Hawaiian species was the downregulation of microtubule-based processes. Microtubules, tubulin polymers that maintain structure and shape to eukaryotic cells, are major components of cilia, which coral utilize for activities such as feeding and clearing sediment ( Westbroek, Yanagida & Isa, 1980 ; Rogers, 1990 ; Erftemeijer et al., 2012 ). Downregulation of processes relating to cilia biogenesis/degradation and motility was also observed in the P. acuta host and its endosymbionts in response to combined acute heat and sediment stress ( Poquita-Du et al., 2019 , 2020 ). The authors of these studies hypothesized that the corals were likely diverting energy resources away from feeding and active sediment clearing in order to prioritize energy for basic homeostasis. While our experiment did not include heat stress, it is probable that downregulation of microtubule or cilia-related processes is a generally conserved response to sediment stress. Exhaustion of sediment-clearing activity of corals and eventual loss/breakdown of cilia cells can also be a consequence of continued exposure to high levels of sediment stress ( Stafford-Smith & Ormond, 1992 ; Stafford-Smith, 1993 ; Erftemeijer et al., 2012 ). Given the acute nature of the stress in the Hawai‘i experiment ( i.e ., sediment was added at day 0 and day 4), the corals here may have exhausted their ciliary action abilities, thus leading to downregulation of microtubule-based processes. Responses to sterilized white sediment in Florida No specific responses were shared among A. cervicornis , M. cavernosa , and O. faveolata , despite being exposed to the same sediment for the same amount of time. Given the length of time of this experiment, it is possible that we only captured the longer-term stress responses of these corals and that their responses may have been more similar at the beginning of the exposures. Morphology also may have contributed to the divergent responses between these species. In previous sediment stress studies, morphology contributed to physiological response variability ( Stafford-Smith, 1993 ; Fabricius, 2005 ; Erftemeijer et al., 2012 ). For example, Rogers (1979) evaluated the effect of shading (as a proxy for turbidity) for 5 weeks on several coral species from San Cristobal Reef, Puerto Rico. The branching coral, A. cervicornis , had entirely bleached, while the massive coral, M. cavernosa , was visibly unaffected and appeared to have little response. On the other hand, M. annularis had substantial bleaching after 5 weeks, despite being a close relative of M. cavernosa ( Rogers, 1990 ). Different branching coral species exhibit a wide range of sediment tolerances; after 12 weeks of exposure, the lowest sediment treatments that caused full colony mortality were 30 mg/L −1 for M. aequituberculata and 100 mg/L −1 for A. millepora ( Flores et al., 2012 ). Stafford-Smith (1993) examined sediment rejection efficiency in 22 species of Australian corals from a range of morphologies, finding that there was a wide range of active-rejection efficiencies between species. For instance, Gardineroseries planulata is a competent rejector of a variety of sediment sizes, but only for a short period of time. Favia stelligera and Leptoria phrygia had moderate clearance rates, but tissue mortality occurred within one to two days. M. aequituberculata and Porites spp. had low rejection efficiency and had bleached tissues, but no tissue mortality for up to 8 days. Morphology was a driving factor in those results, as branching corals had faster clearing rates than massive corals ( Stafford-Smith, 1993 ). Thus, it is likely that morphology played a role in driving differences in long-term response in the Floridian corals, highlighting the challenge of predicting responses to sedimentation across species. Despite differing morphologies, the branching coral, A. cervicornis , shared responses related to developmental processes and signal transduction with both massive corals, M. cavernosa and O. faveolata , independently. Interestingly, no specific responses were shared between M. cavernosa and O. faveolata , despite similar morphologies. Downregulation of Rho protein signal transduction was observed in A. cervicornis and M. cavernosa (with the exception of upregulation in the M. cavernosa T2vT3 treatment comparison). Rho proteins are part of a superfamily of signaling GTPase proteins, which typically control the assembly and organization of the cytoskeleton, as well as participate in functions such as cell adhesion, contraction, migration, morphogenesis, and phagocytosis ( Mackay & Hall, 1998 ; Moon & Zheng, 2003 ; Phuyal & Farhan, 2019 ). In corals, Rho GTPases participate in cytoskeleton remodeling during phagocytosis, as well as cell division of endosymbionts within symbiotic gastrodermal cells ( Li et al., 2014 ). Rho GTPase pathways have been identified in coral polyp bailout responses to heat stress and hyperosmosis, as well as in bleached corals in response to low flow environments ( Chuang & Mitarai, 2020 ; Fifer et al., 2021 ; Gösser et al., 2021 ). In this study, downregulation of Rho protein pathways suggests that minimal cytoskeleton maintenance, assembly and organization is occurring and that the corals may not be able to properly maintain their cellular structures under sedimentation. A. cervicornis and O. faveolata both downregulated chondrocyte development, a developmental response. Chondrocytes are cells in cartilage that make up the cytoskeletal matrix in humans and other animals, including some marine invertebrates ( Philpott & Person, 1970 ; Cowden & Fitzharris, 1975 ; Libbin et al., 1976 ; Archer & Francis-West, 2003 ; Kamisan et al., 2013 ). In corals, the cytoskeleton matrix is made up of calcium carbonate, as opposed to cartilage, which forms through rapid accretion of protein rich skeletal organic matrix and extracellular calcium carbonate crystals to form a stony skeleton ( Vandermeulen & Watabe, 1973 ; Akiva et al., 2018 ). Skeletal matrix formation begins when planktonic coral larvae settle and begin to secrete calcium carbonate, which helps to anchor the coral to the substrate ( Akiva et al., 2018 ). The skeleton grows as the coral animal continues to secrete calcium carbonate, building up a large and intricate 3D skeletal structure ( Tambutté et al., 2011 ). Sedimentation can hinder coral skeletal growth by depositing sediment on the tissue and diverting energy away from growth, as well as decreasing the amount of light that reaches the coral, thus affecting the possibility for light-enhanced calcification, which is responsible for most of the skeletal growth in corals ( Dodge, Aller & Thomson, 1974 ; Erftemeijer et al., 2012 ). Downregulation of chondrocyte development may be related to the disruption of skeletal matrix formation, which may influence skeletal density and growth. Decreased skeletal density was observed in corals from inshore reefs which experience higher levels of sedimentation as compared to corals from offshore reefs ( Lough & Barnes, 1992 ). Therefore, the downregulation of genes relating to chondrocyte development across morphologies, suggests that corals with a range of morphological characteristics may have decreased skeletal growth in response to sediment stress. Differences in response based on morphological characteristics This study combined data from two independent experiments in order to characterize the molecular mechanisms that corals use to respond to different sedimentation stressors. The experiments used different sediment types (unsterilized red clay sediment in Hawai‘i, sterilized carbonate sand sediment in Florida) and lengths of exposure (up to 7 days in Hawai‘i, 18 days in Florida). Differences in experimental methodology prevent us from making direct statistical comparisons between the two experiments; however, the shared morphologies between experiments enables us to identify broad generalizations about conserved gene regulation in response to sediment stress. These comparisons are relevant, as knowledge of what genes and biological processes are broadly affected by sediment stress can help coral reef management. We did not identify a generalized response across morphology nor gene expression patterns across taxa. There were two groups of three species ( M. capitata , M. cavernosa , P. acuta and A. cervicornis , M. capitata , M. cavernosa ) that shared specific responses. However, no group contained a single morphology and gene expression patterns varied among species. For example, M. capitata (intermediate), M. cavernosa (massive), and P. acuta (branching) all expressed genes relating to riboflavin transport, the transport of certain vitamins in cells, though they had very different expression patterns. Riboflavin transport genes were upregulated in P. acuta , but downregulated in M. cavernosa ; M. capitata , on the other hand, differentially expressed two genes relating to riboflavin transport, one of which was upregulated and the other downregulated. Thus, even though a shared response was identified, the direction of the response varied greatly. This result demonstrates that responses may be shared across morphologies, locations and sediment types, but it may be difficult to predict the directionality of the response. This response may also be due a level of acclimation attained by M. cavernosa during the longer exposure in the Florida experiment. The other group, which contained A. cervicornis (branching), M. capitata (intermediate), and M. cavernosa (massive), all downregulated genes relating to positive regulation of skeletal muscle tissue development. This term refers to the activation, maintenance, or increase of the rate of skeletal muscle tissue development ( Buckingham et al., 2003 ; Grefte et al., 2007 ). Given that this term is downregulated, it means that there is little to no activation, maintenance, or increase of the rate of muscle tissue development in these corals. These gene expression patterns highlight the complexity of characterizing responses to different kinds of sedimentation stress in species with different morphotypes. Some terms were shared across both morphology and location, indicating a generalized sediment stress response. For example, DNA methylation-dependent heterochromatin assembly was shared between A. cervicornis (branching) and P. lobata (massive). Opposite expression patterns were again observed, in which upregulation occurred in A. cervicornis and downregulation in P. lobata . DNA methylation-dependent heterochromatin assembly refers to repression of transcription by DNA methylation leading to the formation of heterochromatin ( Jones & Wolffe, 1999 ; Grewal & Moazed, 2003 ). In the case of A. cervicornis , upregulation suggests that repression of transcription by DNA methylation and subsequent heterochromatin formation is occurring. Therefore, certain portions of DNA cannot be accessed, giving A. cervicornis more stringent control on gene expression. On the other hand, the downregulation of these genes in P. lobata means less repression of transcription by DNA methylation occurring and heterochromatin is not being fully formed, leaving much of the DNA accessible to transcription machinery and ultimately, reducing the amount of control that P. lobata has on gene expression. To date, no work has examined how sediment stress affects epigenetic mechanisms, such as DNA methylation, in coral. However, previous studies have found changes to DNA methylation in response to stress and environmental change ( Putnam, Davidson & Gates, 2016 ; Liew et al., 2018 ; Dimond & Roberts, 2020 ; Rodríguez-Casariego, Mercado-Molina & Garcia-Souto, 2020 ). Additionally, it has been observed in corals that genes with weak methylation signals are more likely to demonstrate differential gene expression ( Dixon, Bay & Matz, 2014 ; Entrambasaguas et al., 2021 ). Epigenetic modifications and their regulation of gene transcription are highly species and context dependent. Furthermore, the directionality of epigenetic regulation on gene expression or repression can vary depending on the underlying genetic machinery and the environment. This is exemplified in the two species with overlapping response terms. Namely, A. cervicornis exhibits a more regulated control on gene expression in contrast to P. lobata , which exhibits a less regulated profile of gene regulation. Thus, it is likely that sedimentation stress from each location impacted DNA methylation and heterochromatin in different ways, causing opposing expression patterns. More general responses were shared over morphology and location, as identified by the orthogroup analysis. The orthogroup analysis grouped homologous gene sequences in different species related to one another by linear descent. The resulting ‘orthogroup’ represents a group of similar gene sequences across multiple species ( Emms & Kelly, 2015 , 2019 ). Orthogroups were shared across morphology and location, though in relatively low numbers (sharing between one and three orthogroups). This sharing may represent a group of orthogroups that form a core group of genes in response to sediment stress. Although we cannot directly compare the genes or orthogroups between experiments, the shared orthogroups represent potential overlap in sedimentation response over location and morphology. The branching corals, A. cervicornis from the Florida experiment and P. acuta from the Hawai‘i experiment, shared a metabolic response, ‘Mo-molybdopterin cofactor biosynthetic process’, which describes the creation of the Mo-molybdopterin cofactor, an essential component for catalytic activity of certain enzymes ( Kisker, Schindelin & Rees, 1997 ; Mendel, 2013 ). Molybdenum (Mo) is a trace metal synthesized de novo through GTP-based processes; cyclic pyranopterun monophosphate (cPMP) is initially formed, which is then converted to the molybdopterin cofactor ( Mendel, 2013 ). This essential cofactor catalyzes the oxidation and reduction of molecules in enzymatic processes regulating nitrogen, sulfur, and carbon ( Daniels et al., 2008 ; Iobbi-Nivol & Leimkühler, 2013 ). Similar to other results in the present study, both species demonstrated opposing differential gene expression for this term. Mo-molybdopterin cofactor biosynthetic process was downregulated in A. cervicornis , but upregulated in P. acuta , suggesting that while different sediment types and exposure durations can induce similar differentially expressed genes, it can produce different expression patterns for those genes. Molybdopterin are co-factors for oxidoreductases, a family of enzymes that catalyze the transfer of electrons between molecules ( Kisker, Schindelin & Rees, 1997 ). Upregulation of molybdopterin synthesis may suggest that these types of enzymes are more metabolically active. Stressful conditions have made these kinds of enzymes more active in plants and corals ( Bouchard & Yamasaki, 2008 ; Zdunek-Zastocka & Sobczak, 2013 ). Upregulation of metabolism related genes was also observed in a study that examined transcriptomic responses of corals in response to two different sediment experiments ( Bollati et al., 2021 ). Downregulation of Mo-molybdopterin cofactor synthesis may indicate that the coral does not have enough energy to synthesize molybdopterin which in turn makes it so the activity of these specific enzymes is decreased or stopped altogether, suggesting a decrease in metabolism for A. cervicornis . It is also possible that the unsterilized sediment was providing molybdopterin to P. acuta , making it necessary for P. acuta to upregulate genes relating to molybdopterin processing to manage the influx ( Fujimoto & Sherman, 1951 ; Siebert et al., 2015 ). Because the sediment was sterilized in the Florida experiment, no molybdenum would be present in the sediment, indicating that molybdenum-related enzymes may not have been functioning at a high level, leading to downregulation."
} | 7,859 |
23226544 | PMC3513314 | pmc | 5,006 | {
"abstract": "Disease is increasingly viewed as a major factor in the ecology of marine communities and its impact appears to be increasing with environmental change, such as global warming. The temperate macroalga Delisea pulchra bleaches in Southeast Australia during warm summer periods, a phenomenon which previous studies have indicated is caused by a temperature induced bacterial disease. In order to better understand the ecology of this disease, the bacterial communities associated with threes type of samples was investigated using 16S rRNA gene and environmental shotgun sequencing: 1) unbleached (healthy) D. pulchra 2) bleached parts of D. pulchra and 3) apparently healthy tissue adjacent to bleached regions. Phylogenetic differences between healthy and bleached communities mainly reflected relative changes in the taxa Colwelliaceae , Rhodobacteraceae , Thalassomonas and Parvularcula . Comparative metagenomics showed clear difference in the communities of healthy and diseased D. pulchra as reflected by changes in functions associated with transcriptional regulation, cation/multidrug efflux and non-ribosomal peptide synthesis. Importantly, the phylogenetic and functional composition of apparently healthy tissue adjacent to bleached sections of the thalli indicated that changes in the microbial communities already occur in the absence of visible tissue damage. This shift in unbleached sections might be due to the decrease in furanones, algal metabolites which are antagonists of bacterial quorum sensing. This study reveals the complex shift in the community composition associated with bleaching of Delisea pulchra and together with previous studies is consistent with a model in which elevated temperatures reduce levels of chemical defenses in stressed thalli, leading to colonization or proliferation by opportunistic pathogens or scavengers.",
"conclusion": "Conclusion Marine macroalgae, including D. pulchra , are major habitat formers in temperate ecosystems, and diseases of these organisms have the potential to cause significant impacts on community structure, population levels and biodiversity in marine ecosystems [1] . Campbell et al [9] have proposed a model for bleaching in D. pulchra , in which environmental stress (particularly elevated seawater temperatures) leads to a decrease in chemical defenses (furanones), which in turn leads to a shift in bacterial community composition, which can include opportunistic pathogens or scavangers. This shift might be further exacerbated by the loss of certain key members of the “normal” microbiota of D. pulchra, which might have probiotic functions. Our observations are consistent with such processes and we have identified bacterial species that might be involved in either the disease initiation or secondary scavenging. Functional gene analysis of communities of bleached sample revealed further an abundance of activities that are likely to play key roles in algal colonization and competiton. However the role of these groups of bacteria and genes in the bleaching disease of D. pulchra needs further experimental confirmation. Further advances in understanding the bleaching disease in D. pulchra would require tracking temporal changes in key species, and in the expression of these key functional genes, at different stages of disease.",
"introduction": "Introduction Disease in natural communities is increasingly seen as a major ecological factor. Moreover, a number of studies have argued that the frequency and impact of disease on natural communities is on the rise, likely due to the increasing impact of environmental stressors, such as global warming or other anthropogenic effects [1] , [2] . The impact of disease is arguably felt most strongly when the affected hosts are biogenic habitat formers, or so-called “ecosystem engineers”, because decline in these organisms results in a fundamental change in the physical structure of the habitat, and the loss of not just the hosts, but of the substantial biodiversity associated with habitat forming species. In marine systems, to date the most prominent example of disease impacting habitat-forming organisms are tropical reef-building corals [3] , [4] . However, on temperate and boreal rocky shorelines, macroalgae (i.e. kelps and other seaweeds) dominate, and there they form the basis for extensive and highly diverse communities [5] . There is now evidence that these macroalgal forests are in decline globally, and one suggested mechanism is that of an increased impact of disease [6] . The red macroalga Delisea pulchra \n [7] is common and in places dominant in the shallow sub-tidal zone in southern Australia, New Zealand, Japan and Antarctica [8] . We have recently described a bleaching phenomenon in this alga, in which portions of the thallus lose their pigmentation and decay [9] , [10] . The consequences of this bleaching disease are severe; because of the pattern of bleaching on the thallus relative to the production of reproductive tissue, bleached D. pulchra are essentially neutered, with the amount of reproductive tissue an order of magnitude less than that of healthy individuals [11] . Bleaching is most common in summer, but rather than being a direct effect of light or temperature or other environmental stressors, it appears to be due to bacterial infection of (in particular) temperature stressed plants [10] , [11] . Two bacteria from the Rosebacter clade, namely Ruegeria sp. R11 and Phaeobacter sp. LSS9, have been identified from the surface of D. pulchra , which can cause bleaching in the laboratory [11] , [12] , and strain R11 has also been shown experimentally to cause bleaching in the field [11] . Further evidence for the role of bacteria in causing bleaching comes relates to the alga’s antibacterial chemical defenses. D. pulchra produces halogenated furanones at its surface [13] , [14] . These compounds are strong antibacterials [14] – [20] and one important mechanism for their antibacterial activity is the inhibition of N-acyl homoserine lactone (AHL) based quorum sensing (QS) [13] , [16] \n . Furanones in D. pulchra are typically at their lowest in summer [21] , corresponding to the peak in the incidence of bleaching. A decrease in the furanone concentration of the thallus is also correlated with tissue bleaching [9] . Most compellingly, direct experimental manipulation of furanones results in rapid bleaching of D. pulchra in the laboratory [10] , [11] and in the field [11] . A number of significant questions remain about the nature of this bacteria-alga interaction, but a critical one is: What is the nature of changes in the microbial community associated with bleaching? To address these questions we performed an in-depth microbial community analysis using 16S rRNA gene sequencing and metagenomics, expanding upon previous community studies based on relatively low-resolution techniques (tRFLP), which showed consistent differences in the community composition of healthy and bleached D. pulchra across sampling years, location and depth [9] . We compared bacterial communities from unbleached algae, from bleached tissue, and from apparently healthy (pigmented) tissue adjacent to bleached tissue. This later category is of particular interest, because such tissue, although visibly unaffected, contains furanone levels comparable to bleached tissue [9] . Thus by understanding the microbial communities on chemically poorly defended, but apparently otherwise undamaged tissue, we may gain insight into the progression of the infection process.",
"discussion": "Discussion Bacterial Community of Bleached and Adjacent Tissue Bleaching of D. pulchra is associated with clear changes in the community composition, not only directly at the affected part of the thallus, but also in the surrounding areas. Community richness of bleached and adjacent tissue was in general similar, but greatly increased (>3-fold; see Table 1 ) compared to healthy samples. This increased diversity could be the result of changes in the furanone concentrations, which are high in healthy D. pulchra \n [19] , but greatly reduced in bleached algae and their apparently healthy tissue [9] . The loss of selective and inhibitory effect of the alga’s chemical defense could thus allow for a greater variety of bacteria to colonize and these might potentially include opportunistic pathogens and scavengers (see below). Increases in bacterial diversity have also been observed when the Caribbean coral Montastraea faveolata was afflicted by White Plague Disease [36] and when the common reef-building coral, Porites cylindrica was exposed to stressful environmental events [37] . Communities of diseased thalli are characterized by an abundance of OTUs belonging to the genera Cellulophaga and Thalassomonas as well as the family Colwelliaceae , some of which are also found in adjacent tissue, but absent on healthy algae. Algal diseases are often associated with cell-wall degradation and members of the genus Cellulophaga have been shown to exhibit potent extra-cellular alginolytic or agarolytic activity [38] . OTUs within the Colwelliaceae account for most of the difference between healthy and bleached tissue and this family contains several genera with known or putative pathogenic and anti-algal representatives. For example, Thalassomona loyana has been implicated as a pathogen for ‘white plague-like disease’ in corals in the Red Sea [39] . Bacteria within the genera Alteromonas and Pseudoalteromonas (family Colwelliaceae ) have also been shown to exhibit algicidal activity against eukaryotic microalgae [40] and have been found associated with decaying matter of algal blooms [41] . These observations are consistent with two alternative models for the changes in microbial communities associated with bleaching in D. pulchra. Firstly, changes could be a result of opportunistic pathogens that either colonise or proliferate on relatively chemically undefended tissue and subsequently cause the bleaching disease. The absence of previously identified pathogens of D. pulchra (i.e. Ruegeria sp. R11 and Phaeobacter sp. LSS9) in the samples taken in the current study would also indicate that other bacterial taxa might have the ability to induce disease. Secondly, an alternative model would be that disease occurs independently of pathogens and that some of the taxa enriched on bleached thallus might be involved in tissue decay or nutrient scavenging post disease initiation. These taxa would hence be considered to be opportunistic, secondary colonizers. Further time-series experiments under controlled conditions and tracking of potential pathogens identified here (e.g. Thalassomona ) and in our previous work [11] , [12] across a larger number of disease events are required to distinguish between these possibilities. Loss of Microbiota from Healthy Samples Taxonomically, healthy tissue was characterized by the abundance of bacteria belonging to the genera Parvularcula and Haliscomenobacter, and the family Rhodobacteraceae . The most abundant Rhodobacteraceae phylotype (19% of the sequences found) could also be assigned with moderate confidence (70%) to the genus Thioclava. The genus Thioclava as well as the second most abundant group, the genus Parvularcula , harbor genes for the degradation of the abundant algal osmolyte dimethylsulfoniopropionate (DMSP) [42] and are known to oxidize inorganic sulfur metabolites [43] , such as the byproducts of DMSP degradation [44] . Bacteria involved in the biogeochemical cycling of sulfur are also abundant in coral-associated communities and are thought to play a significant role in structuring the community and to be important for coral health [45] , [46] . Members of the family Saprospiraceae (order Bacteroidetes ), to which the genus Haliscomenobacter also belongs (see Figure 2 ), have also been isolated from marine sponges and algae from the southern coastline of Thailand [47] . While little is known about the ecophysiology of other Saprospiraceae , members of the Bacteroidetes are generally associated with the degradation of complex organic materials and Haliscomenobacter hydrossis is abundant in activated sludge flocs and thought to be involved in the hydrolysis of polysaccharides [48] . While the ecological role of these bacterial associates that dominate the community of healthy D. pulchra is not clear, they belong to taxonomic groups that are commonly found on marine algae [49] – [52] . These groups might thus represent epiphytes that are generally not harmful to the host or are important for structuring a stable microbial community. Importantly, these community members are also in lower abundance on bleached tissue as well as on adjacent parts, where there are no visible changes. This observation would be consistent with a probiotic model, which was proposed for corals [53] and posits that the normal microbiota associated with living surfaces offers protection against microbial infection and disease. According to this concept, the community of the adjacent tissue would have already lost key members of the “normal” microbiota, a shift that may allow for subsequent proliferation of harmful bacteria in the community. This community shift might be due to the loss of algal defense (furanone) and could result in a dysbiosis of the epiphytic bacterial community prior to disease development. Thus it might be important to consider the state of the microbial community prior to disease development and define its effect on disease initiation or outcome. Taxonomic Shifts are Accompanied by Changes in the Functional Gene Composition Accompanying the taxonomic changes, shotgun metagenomic analysis revealed clear differences also in the functional gene composition across communities from healthy, diseased and adjacent tissue. In particular, a series of functional genes were enriched in communities of diseased tissue. These include non-ribosomal peptide synthetases (NRPSs; COG1020), which synthesize molecules with a range of biological functions [54] , including antibiotics [55] , siderophores [56] and other virulence determinants [57] \n [58] , [59] . Either of these functions could allow bacteria on diseased D. pulchra to outcompete other epiphytes, support growth or act directly as a virulence factor in the bleaching disease. Abundant in communities of diseased tissue were also cation/multidrug efflux pumps (COG0841), which belong to a larger group of multidrug resistance pumps (MDRs) that bacteria utilize to extrude toxins [60] . Recently, the MDRs of plant pathogenic bacteria, including Erwinia amylovora , Dickeya spp. and Agrobacterium tumefaciens , have been shown to play a crucial role in the resistance to plant antimicrobials and their inactivation compromised initial colonization and virulence [61] . MDRs are also important for bacterial competition and colonization within surface-associated bacterial communities. For example, efflux pump mutants of the plant pathogens E. chrysanthemi and E. amylovora are less infectious than the wild-type strain when co-inoculated with other endophytic bacteria producing antimicrobial compounds [62] , [63] . Two-component signal transduction (TCST) systems (COG 0642) are used by many bacteria to sense environmental signals and modulate expression of genes [64] , including ones involved in virulence, antibiotic resistance responses, and colonization in many pathogenic bacteria [65] , [66] . For example, in the plant pathogenic bacteria Xanthomonas campestris , Pectobacterium carotovorum and Ralstonia solanacearum , TCST systems have been shown to provide environmental signal inputs to global virulence regulation [67] . The abundance of TCST systems found in communities of bleached tissue could thus enable bacteria to modulate gene expression in response to environmental changes and regulate virulence genes responsible for bleaching. The potential importance of regulation in disease was also recently highlighted in the analysis of the genomes for the D. pulchra pathogens Nautella sp. R11 and Phaeobacter sp. LSS9 [12] . These two pathogens contained no unique set of putative virulence factors when compared to other closely related non-pathogenic strains, but had a unique LuxR-type regulator that was hypothesized to regulated virulence. Conclusion Marine macroalgae, including D. pulchra , are major habitat formers in temperate ecosystems, and diseases of these organisms have the potential to cause significant impacts on community structure, population levels and biodiversity in marine ecosystems [1] . Campbell et al [9] have proposed a model for bleaching in D. pulchra , in which environmental stress (particularly elevated seawater temperatures) leads to a decrease in chemical defenses (furanones), which in turn leads to a shift in bacterial community composition, which can include opportunistic pathogens or scavangers. This shift might be further exacerbated by the loss of certain key members of the “normal” microbiota of D. pulchra, which might have probiotic functions. Our observations are consistent with such processes and we have identified bacterial species that might be involved in either the disease initiation or secondary scavenging. Functional gene analysis of communities of bleached sample revealed further an abundance of activities that are likely to play key roles in algal colonization and competiton. However the role of these groups of bacteria and genes in the bleaching disease of D. pulchra needs further experimental confirmation. Further advances in understanding the bleaching disease in D. pulchra would require tracking temporal changes in key species, and in the expression of these key functional genes, at different stages of disease."
} | 4,476 |
30333383 | PMC6308002 | pmc | 5,007 | {
"abstract": "The hydrothermal vent squat lobster Shinkaia crosnieri Baba & Williams harbors an epibiotic bacterial community, which is numerically and functionally dominated by methanotrophs affiliated with Methylococcaceae and thioautotrophs affiliated with Sulfurovum and Thiotrichaceae . In the present study, shifts in the phylogenetic composition and metabolic function of the epibiont community were investigated using S. crosnieri individuals, which were reared for one year in a tank fed with methane as the energy and carbon source. The results obtained indicated that indigenous predominant thioautotrophic populations, such as Sulfurovum and Thiotrichaceae members, became absent, possibly due to the lack of an energy source, and epibiotic communities were dominated by indigenous Methylococcaceae and betaproteobacterial methylotrophic members that adapted to the conditions present during rearing for 12 months with a supply of methane. Furthermore, the overall phylogenetic composition of the epibiotic community markedly changed from a composition dominated by chemolithotrophs to one enriched with cross-feeding heterotrophs in addition to methanotrophs and methylotrophs. Thus, the composition and function of the S. crosnieri epibiotic bacterial community were strongly affected by the balance between the energy and carbon sources supplied for chemosynthetic production as well as that between the production and consumption of organic compounds.",
"discussion": "Discussion As shown in previous studies ( 40 , 41 , 43 ), a 16S rRNA gene clone analysis of epibiotic communities before rearing showed that they primarily comprised typical bacterial thioautotrophic and methanotrophic phylotypes belonging to the genus Sulfurovum and the families Thiotrichaceae and Methylococcaceae ( Table S1 and Fig. S2 ). However, after rearing for 3 and 12 months in a methane-fed tank, 16S rRNA gene clone and microscopic analyses showed that Sulfurovum -affiliated populations disappeared from epibiotic communities ( Table S1 and Fig. 2 ). The incorporation of [ 13 C]bicarbonate in the epibiotic communities of S. crosnieri individuals was markedly less after rearing for 3 and 12 months than before rearing, and was not enhanced by the presence of sulfide ( Table 1 ). These results indicate that S. crosnieri individuals lost their Sulfurovum -affiliated populations and their thioautotrophic function during methane-fed rearing for 12 months and even for 3 months. The abundances and functions of Thiotrichaceae -affiliated thioautotrophic epibionts as well as Sulfurovum decreased during methane-fed rearing. We confirmed that the S. crosnieri epibiotic community assimilated inorganic carbon without the addition of an energy source before rearing ( Table 1 ). A previous study suggested that Thiotrichaceae epibionts potentially assimilate inorganic carbon using sulfur stored within their cells and without an external supply of reduced sulfur compounds ( 41 ). However, inorganic carbon was not assimilated by S. crosnieri epibionts after rearing for 3 months, even in the presence of sulfide utilized by thioautotrophs as a potential energy source for chemosynthesis ( Table 1 ). These results indicate that the active thioautotrophic function from the original epibiotic community was strongly affected by methane-fed rearing for 3 months. The 16S rRNA gene clone analysis also showed that the abundances of most of the Thiotrichaceae phylotypes in the epibiotic community were lost during rearing, whereas the clonal abundance of one phylotype (12 methane_07) of Thiotrichaceae increased during rearing ( Table S1 and Fig. S2 ). The heterotrophic strain KP105 isolated from the epibiotic community of an S. crosnieri individual after rearing for 12 months was closely related to the phylotype 12 methane_07 (97% identity). The cultivation test indicated that this strain did not grow with methane (methanotrophy) and methanol (methylotrophy), or even with reduced sulfur compounds and inorganic carbons (thioautotrophy) ( Table 4 ). The 97% 16S rRNA gene sequence identity between the heterotrophic strain KP105 and the phylotype 12 methane_07 does not necessarily represent similarities in their metabolic and physiological functions; however, the Thiotrichaceae populations detected in epibiotic communities during methane-fed rearing for 12 months appeared to function as heterotrophic consumers rather than as thioautotrophic primary producers. In addition, the weak but detectable incorporation of [ 13 C] bicarbonate independently of sulfide and slight sulfide consumption were observed in epibiotic communities after methane-fed rearing for 12 months ( Tables 1 and 3 ). Rhodobacteraceae -affiliated phylotypes were detected in the 16S rRNA gene clone libraries of epibionts obtained from S. crosnieri individuals after rearing for 12 months ( Table S1 and Fig. S2 ). Several members of Rhodobacteraceae are known to exhibit mixotrophic carbon metabolism ( 30 , 33 , 39 ). Thus, the potentially anaplerotic inorganic carbon fixation of potentially mixotrophic Rhodobacteraceae -affiliated and other heterotrophic epibionts may have contributed to the incorporation of [ 13 C]bicarbonate independently of sulfide in the epibiotic communities after methane-fed rearing for 12 months. In addition, hydrogen sulfide (sulfide) is known to be toxic to animals because it may bind to cellular iron and disrupt the functions of mitochondria ( 8 ). The brachyuran crab, Bythograea thermydron Williams, which is endemic to deep-sea hydrothermal environments, may detoxify hydrogen sulfide (sulfide) via its oxidation to thiosulfate and sulfate using its own detoxification enzymes ( 37 ). Although we lack any physiological and genetic evidence, slight sulfide consumption by S. crosnieri individuals during methane-fed rearing may have been catalyzed by the function of the host S. crosnieri rather than by its epibiotic bacterial community. [ 13 C]methane assimilation and methane consumption experiments clearly indicated that S. crosnieri individuals reared in the methane-fed tank harbored active methanotrophs in their epibiotic communities even after 12 months ( Tables 1 and 2 ). Previous studies demonstrated that the oval cells of Methylococcaceae members represented methanotrophic populations in the epibiotic communities of naturally living S. crosnieri individuals ( 40 , 42 ). The 16S rRNA gene clone analysis demonstrated that the phylotypes affiliated with Methylococcaceae were preserved in the S. crosnieri epibiotic communities during methane-fed rearing for 12 months and one Methylococcaceae phylotype (12 methane_1-14) remained abundant throughout the rearing period ( Table S1 and Fig. S2 ). The phylogenetic analysis of pmoA gene sequences showed that all of the potential methanotrophic populations present in S. crosnieri epibiotic communities throughout the rearing period were related to members of Methylococcaceae ( Fig. 1 ). pMMO is present in almost all known methanotrophs, except for the genera Methylocella and Methyloferula , which are acidophilic methanotrophs within Alphaproteobacteria ( 5 , 38 ), and the primer set used in the present study covered most of the known pmoA diversity ( 22 ). In addition, our microscopic observations using FISH and TEM analyses verified the abundant occurrence of Methylococcaceae -like oval cells in S. crosnieri epibionts even after rearing for 12 months ( Fig. 2 and S3 ). Overall, these results strongly suggest that active Methylococcaceae methanotrophs were maintained as one of the predominant populations in S. crosnieri epibiotic communities when reared in the methane-fed tank for 12 months. The net methane consumption rates by S. crosnieri individuals after rearing for 3 and 12 months decreased to approximately half of those before rearing ( Table 2 ), whereas the rate of [ 13 C]methane incorporation by the epibiotic community was greater after than before rearing ( Table 1 ). These results may be explained by the effects of decreases in the biomass abundance and increases in the relative biomass proportion of methanotrophs in epibiotic communities during methane-fed rearing for 12 months. Although the setae of S. crosnieri immediately after capture are covered with more than 100 μm of filamentous microbes and have an abundant microbial biomass ( 36 , 40 ), these very long filaments were not observed on the setae after rearing for 12 months ( Fig. S3 ). In addition, typical long and thick filamentous Sulfurovum -affiliated epibionts were not observed after rearing for 3 and 12 months ( Fig. 2 ). Thus, the overall biomass of the epibiotic community, including methanotrophic populations, continued to decrease whereas the relative abundances of the methanotrophic populations and their activities in the epibiotic communities increased with the methane-fed rearing period. Stable isotope probing (SIP) with [ 13 C]methane has been employed to investigate active methanotrophs in a number of environments ( 23 , 24 ). However, the use of SIP techniques with a relatively long incubation time may lead to cross-feeding, through which non-targeted microorganisms incorporate the labeled substrates via secondary carbon flow from the targeted microorganisms ( 12 ). A methane-SIP study of sediments from an Arctic lake with active methane seepage indicated that [ 13 C]methane incorporation was observed in members of various bacterial taxa, such as Proteobacteria (including Methylococcaceae and Methylophilaceae ), Bacteroidetes , Acidobacteria , Planctomycetes , Verrucomicrobia , and Actinobacteria ( 12 ). After rearing for 3 and 12 months, the predominant phylotype components of the S. crosnieri epibiotic bacterial communities were similar to those found in the aforementioned [ 13 C]methane-incorporating microbial community in Arctic lake sediments ( Table S1 and Fig. S2 ). In addition, the bacterial strains isolated from the epibiotic community after rearing for 12 months, which were closely related to the phylotypes found in the 16S rRNA gene clone library of epibionts after rearing, were not methanotrophic, but instead were methylotrophic and heterotrophic ( Table 4 ). Therefore, we conclude that the epibiotic communities of S. crosnieri individuals reared in the methane-fed tank changed from the original state dominated by chemolithotrophs (mainly thioautotrophic and methanotrophic populations) to a rearing-adapted state that mainly comprised residual methanotrophs and heterotrophs, which grow by cross-feeding on methanotrophically produced organic carbon. Differences in hydrothermal fluid chemistry may have an effect on the epibiotic community composition of hydrothermal vent animals because epibiotic community compositions differed among the same species of animals from geographically distant and geologically different fields ( 23 , 44 ). However, the relationship between the epibiotic community composition and the chemical environment of host animals’ habitats has not been clearly justified due to the difficulties associated with elucidating the chemical conditions of the habitat in detail. In the present study, S. crosnieri individuals were reared in a tank, the chemical environment of which was different from the natural habitat and was artificially controlled. The most prominent difference in the chemical environment was the availability of reduced sulfur compounds under in situ and methane-fed rearing conditions. In a previous study, the laboratory rearing of several S. crosnieri individuals was conducted in a tank fed with certain amounts of H 2 S and CO 2 as the energy and carbon sources for 3 months ( 20 ). A preliminary 16S rRNA gene clone analysis of the epibiotic community suggested that after rearing for 3 months, S. crosnieri individuals hosted the Sulfurovum and Thiotrichaceae phylotypes, but lost the Methylococcaceae phylotypes ( 20 ). This preliminary rearing experiment did not fully explain the adaptive compositional and functional shifts in the epibiotic community of S. crosnieri during sulfide and CO 2 -fed rearing, but highlighted the effective impact of the energy source supply on the composition and function of the epibiotic community in S. crosnieri. The present study showed that the methane supply during rearing was a powerful environmental factor that induced compositional and functional shifts in the epibiotic community of S. crosnieri , which strongly suggests that the development of the S. crosnieri epibiotic community was affected by the balance between the energy and carbon sources supplied for primary production, and even by the balance between the production and consumption of organic compounds. Previous studies provided molecular insights into the biological interactions between deep-sea vent-endemic chemosynthetic animals and their epibionts, in which S. crosnieri and the hydrothermal worm Alvinella pompejana Desbruyères & Laubier use an antimicrobial non-lectin polysaccharide and peptide to select particular bacterial members in the hydrothermal vent ecosystem ( 10 , 34 ). The interaction between the functions of these environmental and biological factors may act as a selective force to control the compositional and functional development of epibiotic bacterial communities in the natural habitats of the deep-sea vent-endemic S. crosnieri , as well as other crustaceans and polychaete annelids. Rearing systems for deep-sea vent-endemic animals with artificial control over physical and chemical conditions in the habitat are effective for investigating the relationships between epibiotic microbial communities, host animals, and environmental conditions. S. crosnieri from the Okinawa Trough deep-sea hydrothermal systems may be reared for a relatively long time (>1 year) and may be useful as a model organism for future rearing-based chemosynthetic symbiosis research."
} | 3,510 |
29214058 | null | s2 | 5,010 | {
"abstract": "The extracellular matrix (ECM) is a dynamic environment that constantly provides physical and chemical cues to embedded cells. Much progress has been made in engineering hydrogels that can mimic the ECM, but hydrogel properties are, in general, static. To recapitulate the dynamic nature of the ECM, many reversible chemistries have been incorporated into hydrogels to regulate cell spreading, biochemical ligand presentation and matrix mechanics. For example, emerging trends include the use of molecular photoswitches or biomolecule hybridization to control polymer chain conformation, thereby enabling the modulation of the hydrogel between two states on demand. In addition, many non-covalent, dynamic chemical bonds have found increasing use as hydrogel crosslinkers or tethers for cell signalling molecules. These reversible chemistries will provide greater temporal control of adhered cell behaviour, and they allow for more advanced "
} | 235 |
35558927 | PMC9088807 | pmc | 5,011 | {
"abstract": "Long-term operation of microbial fuel cells (MFCs) results in an electrochemical activity decline by the degradation of the anodic biofilm. In this work, an alkaline soaking treatment is proposed as an efficient and simple method for anode regeneration. The alkaline treatment was employed in a used carbon-brush anode, and its performance was compared with those of two other traditional treatment methods, i.e. air drying and carbonization. Among all the treated MFC anodes, the one treated by alkaline soaking exhibited the highest recovery rate. A series of tests including a start-up process, scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS) and MFC performance were performed. The results show that alkaline soaking can modify the carbon fiber by introducing carboxyl groups onto the carbon surface and completely remove the aged biofilm, demonstrating that the alkaline treatment of used anodes is a practically effective method for the performance recovery of MFCs.",
"conclusion": "4 Conclusions In this work, the performance recovery of MFCs equipped with the used anodes treated by different methods was investigated. The proposed alkaline treatment for the used anode induces the increase in the electrochemical activity and the kinetic feature of biofilm, leading to the high-recovery of MFC performance. However, the used anode with air drying treatment cannot be reused to generate power in MFC. This is because the limited diffusion of substrate and product within biofilm decreases MFC performance recovery. The carbonization treatment of the used anode caused the weak electrochemical activity of biofilm and the reduction of MFC performance recovery.",
"introduction": "1 Introduction Over the past decades, research on microbial fuel cells (MFCs) has attracted wide attention for its unique advantages to degrade organic matter in wastewater and simultaneously produce electricity by the electrochemically active bacteria (EAB). 1–3 However, the power output of MFCs is still limited. The anode is a critical component of MFCs, a serves as a matrix of bacterial immobilization and a place of bioelectrochemical reaction. 1,4,5 Among previously reported anode materials for MFCs, carbon-brush with a high specific surface area, low electrode resistance and strong biocompatibility has been considered as one of the most suitable electrodes for simultaneous power generation and wastewater treatment. 3,6,7 The carbon-brush electrode was firstly used as an anode material in a single-chamber MFC, which achieved a high power-density of 73 W m −3 . 8 A 90 liter MFC stack by linking individual carbon-brush anodes in series produces enough energy to power a pumping system. 9 When an MFC equipped with carbon-brush anodes is fed with wastewater, a high chemical oxygen demand (COD) removal efficiency is observed. 10,11 Therefore, carbon-brush anodes exhibit attractive advantages in practical sewage treatment systems due to their high power-output and COD removal. Unfortunately, during the long-term operation of MFCs, the degradation of the anodic biofilm leads to a decrease of its electrochemical activity. 12 More importantly, inside a carbon-brush anode, a naturally metabolism-induced aged or dead biofilm is difficult to discharge, thereby increasing the transfer resistance of substrates between the interior space of the anode and the bulk solution, and further impairing the performance of MFCs. 13 This fact may lead to the decline in electric energy recuperation of an individual MFC. What's worse, in MFC stacks, if the oxidation current of MFC with the degraded anodes could not satisfy the needs of the higher uniform current from other MFCs, voltage reversal will occur and seriously limit energy recovery of MFC stacks. In this case, a number of deteriorated anodes need to be replaced, resulting in relatively high costs. Therefore, a cost-effective method for restoring power generation of used anodes is highly desired. In recent years, some methods have been proposed to detach biofilm from the anode surface and achieve the performance recovery, like high shear stress. 14–16 However, such physical method could only exfoliate parts of the biofilm and the remaining biofilm would hinder the formation of fresh biofilm. High-temperature carbonization technology is an effective strategy to decompose protein-rich biomass, e.g. , such as microalgae, macroalgae and biofilm, thus removing the biofilm. 17 Yang et al. 18 reported that the total mass loss of the marine microalgae is more than 80% during the pyrolysis process because of the degradation of crude protein. Thermogravimetric analysis shows that the main weight loss of the marine macroalgae is attributed to the decomposition of the protein and carbohydrate content. 19 However, the carbonization treatment of the used anodes causes the weak electrochemical activity of biofilm and the reduction of MFC performance recovery, leading to an infeasible operation in practice. Complete removal of the aged biofilm is a crucial step towards the regrowth of electrochemically activated biofilm, which determines the recovery of the used anode capacity. In some food processing plants, the biofilm attached on plastic media is thoroughly detached by soaking it in extreme pH conditions, suggesting that the treatment of extreme pH solution may be an effective way to eliminate the aged biofilm on the substrate. 20,21 In line with this idea, this work proposes an alkaline treatment method of the MFC anode regeneration by soaking the used carbon-brush in alkaline solution. The used carbon-brush anodes are treated by three different ways including air drying, high-temperature carbonization and alkaline soaking to highlight the advantage of the alkaline treatment method. A series of tests including start-up process, scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), cyclic voltammogram (CV), electrochemical impedance spectroscopy (EIS) and MFC performance are conducted to study the mechanism of alkaline treatment on the anode. The proposed alkaline treatment is expected to be a simple and effective method to restore power generation of used MFCs in practical applications.",
"discussion": "3 Results and discussion 3.1 Surface characteristics of the used carbon-brush anodes After a 90 h inoculation, the four MFCs exhibited a similar steady voltage output. For clarity's sake, only one representative MFC start-up is shown in Fig. S1. † The MFC showed a lag time of about 35 h and a stable voltage of 0.37 V. Fig. 1 displays the surface morphology of mature anode samples dealt with different treatments. Since air drying was used, the entire anode surface was covered by the microorganism ( Fig. 1a ). As shown in Fig. 1d , a thin and compact biofilm was formed on each carbon fiber surface. Notably, the carbonization-treated anode exhibited much rougher surface and more agglomeration than the original carbon-brush anode. From high magnification observations in Fig. 1e , agglomeration particles were tightly attached to anode surface because of carbonization treatment. The Energy-Dispersive X-ray Spectroscopy (EDS) results of those particles further showed that C (80.5 wt%), O (18.4 wt%) and slight amount of P (0.7 wt%), Na (0.3 wt%) and K (0.1 wt%) were detected (Fig. S2 † ). It has been reported that the protein and carbohydrate components of biomass is prone to convert to a mixture of char and mineral salts during high-temperature carbonization at 900 °C, indicating that the attached agglomeration probably corresponds to char and mineral salts. 19,22,23 However, after alkaline treatment, no microbes were observed on the smooth anode surface. This is because extreme alkaline conditions would lead to irreversible biofilm degradation and then biofilm detachment from the electrode surface. 24,25 Fig. 1 SEM images of the mature carbon-brush anode treated by air drying (a), carbonization (b), alkaline soaking (c) and higher magnification of a (d), b (e), c (f). XPS measurements were conducted to investigate chemical compositional information of the used carbon-brush anodes with different treatments. The EDS spectrum of the original carbon-brush anode (MFC-CB) demonstrated the presence of C and O. As shown in Fig. S4a–c, † the high-resolution spectrum of C1s exhibited three main peaks at 284.6, 285.8 and 288.5 eV, which could be assigned to C–C, C–O, and 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, respectively. Fig. S4d † shows a high-resolution O1s scan, which indicated the existence of C O or P O (531.7 eV) and C–O–C (533.1 eV) bonds in MFC-CB anode. The high-resolution O1s patterns of MFC-C and MFC-A anodes revealed a new C–OH (532.8 eV) bond ( cf. Fig. S4e and f † ). These data suggest that, compared with the original carbon-brush, both carbonization- and alkaline-treatments increase the number of carboxylic acid groups. In addition, the P2p spectrum of the MFC-C anode showed a peak at 133.3 eV associated with P O bond, confirming phosphate functional groups on electrode surface (Fig. S4g † ). 3.2 MFC restart-up and performance The start-up processes of the three treated MFCs are shown in Fig. 2 . Obviously, the different treatments for the used anodes significantly affected the MFC start-up. The lag time of the three treated MFCs was shorter than that of MFC-CB (thirty-five h). Interestingly, MFC-D showed the shortest lag time of fifteen h among all the treated MFCs. Because of the formation of agglomeration by carbonization, the anode surface of MFC-C can provide more attaching sites for biofilm development and thus reduce the lag time (about twenty-five h). The carbon-brush anode soaking in alkaline solution (MFC-A) may increase the affinity of microbes, and the lag time decreased to 30 h. Furthermore, after ninety h inoculation, the cell voltages reached a steady state. Notably, MFC-A exhibited a high voltage output, while MFC-D and MFC-C showed a low voltage output in comparison with MFC-CB. Fig. 2 Voltage evolution of MFC equipped with the used anode treated by different methods during restart-up. \n Fig. 3 displays the power density curves of the three treated MFCs and the control sample MFC-CB. The discrepant recovery degree was observed in these MFCs processed by different treatments. The maximum power densities of MFC-D (1.04 ± 0.03 W m −2 ), MFC-C (1.21 ± 0.01 W m −2 ) and MFC-A (1.42 ± 0.02 W m −2 ) were restored to ∼85%, 99% and 116% of MFC-CB (1.22 ± 0.01 W m −2 ), respectively. The maximum current densities of MFC-D (3.66 ± 0.08 A m −2 ), MFC-C (4.36 ± 0.07 A m −2 ) and MFC-A (7.62 ± 0.44 A m −2 ) were restored to ∼59%, 71% and 123% of MFC-CB (6.17 ± 0.22 A m −2 ), respectively. In terms of the recovery degree of both maximum power and current densities, soaking the used anodes in alkaline solution is beneficial for the performance restoration of MFC. Fig. 3 Power density curves of MFCs with different anodes. 3.3 Cyclic voltammetry analyses of anodic biofilm CV tests were carried out to evaluate the bio-electrocatalytic activity of biofilm on the anode surface. Fig. 4a displays CVs of biofilm under turnover condition. The CV curves of the anodic biofilm exhibited similar sigmoidal catalytic waves but different levels of bio-catalytic current, which is regarded as an indicator of the bio-electrocatalytic activity of biofilm. Characterized by voltammograms in Fig. 4a , the peak current values of the MFC anodes increased in the following order: MFC-D < MFC-C < MFC-CB < MFC-A. Therefore, the electrochemical activity of the biofilm was enhanced after the used anode was treated by alkaline soaking, whereas the air drying and carbonization treatment for the used anode showed a decreased performance compared to the anode biofilm of MFC-CB. Fig. 4 CV responses on anodic biofilms of MFC-CB, -D, -C and -A at a scan rate of 1 mV s −1 under turnover (a) and non-turnover (b) condition. With acetate serving as electron donors, the typical sigmoidal CV resulted from multiple turnovers of each redox system. 26,27 The CVs were performed on the anodic biofilm of MFCs under non-turnover condition at a slow scan rate of 1 mV s −1 to confirm dominant individual redox system. As illustrated in Fig. 4b , all MFC anodes revealed several similar reversible peaks in the absence of substrate. Among them, redox system E (marked by arrow) was the dominance of electricity generation regarding the height of peak currents. It has been reported that the redox system E with a formal potential of −0.17 V vs. SHE may belong to OmcB (outer membrane c-type cytochrome B), which mediates the electron transfer of bacteria to solid electrodes by direct electron transfer. 28,29 Moreover, the formal potential (−0.17 V) of the redox system E was consistent with the acetate-oxidation formal potential of biofilm under turnover condition, implying that the bio-electrocatalytic activity of all anodes predominantly depends on the electron transfer capability of OmcB. Therefore, the CV results under both turnover and non-turnover conditions illuminated that the bio-electrocatalytic activity of the biofilm changed using different methods to treat the used anode, whereas the anodic electron transfer mechanism of all MFCs was unchanged, corresponding to the direct electron transfer by OmcB. 3.4 Electrochemical kinetics of anodic biofilm Although electron transfer path of all MFC anodes was identical, their electrochemical kinetics was diverged according to the CV analysis with different scan rates. As seen in Fig. 5a–d , increasing the CV scan rate ( v ) from 10 to 80 mV s −1 , there was a good linear correlation between the scan rates and peak currents of the anodic biofilm for all MFCs except MFC-D (inset in Fig. 5a–d ). The linear dependence of the peak heights on the scan rates indicates typical surface-controlled electrochemical processes for anodes. In this case, the current generation of anodes is limited by the slow interfacial electron transfer between the bacterial outer membrane proteins and the anode surface. Considering OmcB mentioned above as the final electron carrier of the redox proteins in MFC anodes, it can be concluded that the performance of MFC-CB, MFC-C and MFC-A strongly depends on the electron transfer properties between OmcB and anode surfaces. The dependency of peak currents on scan rates ( v ) and their square roots ( v 1/2 ) was analyzed in Fig. 5e and f to further investigate the electron transfer of the MFC-D anode. The peak current was proportional to v (slow-scan rate range: 1–10 mV s −1 ) while it was linear to v 1/2 (high-scan rate range: 10–100 mV s −1 ), implying typical diffusion-controlled electrochemical processes for the MFC-D anode. It has been reported that diffusion-control results from the confined diffusion of charge-compensating ions in the biofilm in the absence of redox mediators. 30 This indicates that the MFC-D performance is limited by the diffusion of charge-compensating ions within the biofilm (such as acetate substrate and proton product) at high current densities. The existence of the diffusion resistance inside the biofilm of MFC-D probably results from the hindrance of parts of dead bacteria on the anode surface, which was formed via air drying treatment for the mature biofilm. This indicates that the shortest start-up times of MFC-D is partly caused by bacterial resurgence rather than the fresh biofilm formation. Fig. 5 CVs of anodic biofilms under non-turnover condition with different scan rate. The biofilm was formed on original anode (a) and the used anode treated by air drying (b), carbonization (c), alkaline soaking (d). Plot of peak currents of the air drying-treated anode as a function of scan rates (e) and square root of scan rate (f). As shown in Fig. 6 , insights were performed to further investigate the influence of the treatments for the used anodes on the electrochemical kinetics of biofilm. On the basis of the redox peak potential of the MFC anodes with scan rates ranging from 1 to 100 mV s −1 , the curves of versus ln( v ) ( E pa is the anodic peak potential and is the anodic formal potential) was plotted in Fig. 6a to analyze the electron transfer rate constant from the following equation: 31 1 where R is the ideal gas constant, T stands for the temperature in kelvin degree, F is Faraday constant, n represents the electron transfer number, ν is the scan rate, α is the charge transfer coefficient and κ app is the apparent electron transfer rate constant. When the separation of anodic to cathodic peak potentials was larger than 200 mV/ n , the κ app values for the redox system E could be gained according to eqn (1) and are summarized in Table 1 . As shown in Table 1 , the κ app values of the anodes of MFC-CB, MFC-D, MFC-C and MFC-A were 5.59, 1.20, 1.27 and 27.31/s, respectively. A high κ app of the alkaline-treated anode implies a fast electron transfer rate, while a decreased κ app value suggests hindered electron transfer at air drying- and carbonization-treated anodes. Fig. 6 Electrochemical kinetics characterizations on anodic biofilms of MFC-CB, -D, -C and -A. Anodic peak potentials as a function of the Napierian logarithm of CV scan rates (a) and Nyquist plots of the anodic biofilms (b). Summary of P max , I max , electron transfer rate constant κ app and charge-transfer resistance R ct on anodic biofilms of MFC-CB, -D, -C and -A MFC samples \n P \n max (W m −2 ) \n I \n max (A m −2 ) \n κ \n app (s) \n R \n ct (Ω) MFC-CB 1.22 ± 0.01 6.17 ± 0.22 5.59 9.9 MFC-D 1.04 ± 0.03 3.66 ± 0.08 1.20 18.8 MFC-C 1.21 ± 0.01 4.36 ± 0.07 1.27 11.6 MFC-A 1.42 ± 0.02 7.62 ± 0.44 27.31 8.0 In addition, EIS was carried out to evaluate the internal resistances of the MFC anodes. From Nyquist plot in Fig. 6b , the impedance spectra of four MFC anodes was successfully fitted with the equivalent circuit as described previously, in which R 1 is the resistance of oxidation or reduction processes of metal salts and R ct represents the interfacial charge-transfer resistance. 32 The detailed resistances of the MFC anodes are summarized in Table 1 . The differences in the charge transfer resistances were observed. The R ct values of the anode of MFC-CB, MFC-D, MFC-C and MFC-A were 9.9, 18.8, 11.6 and 8.0 Ω, respectively. In general, a low charge transfer resistance represented a fast electron transfer rate. 32,33 Thus, the lowest charge transfer resistance of the MFC-A anode suggests the fastest electron transfer rate among the four cases. The EIS result was consistent with the aforementioned electron transfer rate result. The above results demonstrate that compared with the original carbon-brush anode, the air drying, and carbonization treatments of the used anodes are detrimental to the electron transfer from bacteria to electrodes. However, the reason for decreasing electron-transfer rate was discrepant. In the case of air drying treatment, parts of dead bacteria on the anode surface might inhibit the metabolite diffusion from biofilm to solution and then the aggregate metabolite within biofilm caused an unfavorable microenvironment ( i.e. low pH) for bacterial electricity generation, thus leading to a low electron-transfer rate. By comparison, after carbonization, the electrode surface might be connected with active functional group and partly covered by mineral salts (such as sodium/potassium phosphate) produced by the pyrolysis of biofilm growing in a Na + /K + -rich phosphate buffer solution. 19,22 In this case, on the anode surface, the weakly conductive materials resulted from phosphate coverage is probably responsible for the decrease in the electrochemical activity area and the conductive capability of the electrode, finally leading to the reduction of the electron-transfer rate. On the other hand, the alkaline treatment for the used anode accelerated the electron transfer rate between the regenerative biofilm and the anode surface. XPS results mentioned above demonstrate that alkaline treatment could increase the numbers of carboxylic acid groups on the carbon-brush surface. 34,35 Between carboxyl groups and bacterial cytochromes, the hydrogen bonding that facilitated the electron transfer is prone to be formed. 36–38 Thus, the alkaline-treated anode showed the fastest electron-transfer from the electrode surface to OmcB among all the cases. In conclusion, the performance of the MFC treated by air drying was limited by the diffusion of substrate through biofilm. A slow diffusion of products might decrease the electrochemical activity of biofilm. Thus, both the slow diffusion rate and the low electrochemical activity probably resulted in the low-recovery power generation of MFC-D. However, considering that the performance of MFC-CB, MFC-C and MFC-A mainly depended on the electron transfer between anodes and OmcB, the surface property of the electrode had significant effects. The carbonization treatment of the used anode resulted in the decrease of the electrochemical activity of the anode and thus decreased the power production of MFC-C at high current densities. On the contrary, alkaline treatment achieved a complete removal of the aging biofilm and enhanced the electrochemical activity of the regenerative biofilm, making this treatment a simple and effective method for the high-recovery of MFC performance in practical application."
} | 5,490 |
19275941 | null | s2 | 5,013 | {
"abstract": "The byssal threads of the California mussel, Mytilus californianus, are highly hysteretic, elastomeric fibers that collectively perform a holdfast function in wave-swept rocky seashore habitats. Following cyclic loading past the mechanical yield point, threads exhibit a damage-dependent reduction in mechanical performance. However, the distal portion of the byssal thread is capable of recovering initial material properties through a time-dependent healing process in the absence of active cellular metabolism. Byssal threads are composed almost exclusively of multi-domain hybrid collagens known as preCols, which largely determine the mechanical properties of the thread. Here, the structure-property relationships that govern thread mechanical performance are further probed. The molecular rearrangements that occur during yield and damage repair were investigated using time-resolved in situ wide-angle X-ray diffraction (WAXD) coupled with cyclic tensile loading of threads and through thermally enhanced damage-repair studies. Results indicate that the collagen domains in byssal preCols are mechanically protected by the unfolding of sacrificial non-collagenous domains that refold on a slower time-scale. Time-dependent healing is primarily attributed to stochastic recoupling of broken histidine-metal coordination complexes."
} | 334 |
33997787 | PMC8067885 | pmc | 5,014 | {
"abstract": "The construction of ionic conductive hydrogels with high transparency, excellent mechanical robustness, high toughness, and rapid self-recovery is highly desired yet challenging. Herein, a hydrogen-bonding network densification strategy is presented for preparing a highly stretchable and transparent poly(ionic liquid) hydrogel (PAM-r-MVIC) from the perspective of random copolymerization of 1-methyl-3-(4-vinylbenzyl) imidazolium chloride and acrylamide in water. Ascribing to the formation of a dense hydrogen-bonding network, the resultant PAM-r-MVIC exhibited an intrinsically high stretchability (>1000%) and compressibility (90%), fast self-recovery with high toughness (2950 kJ m −3 ), and excellent fatigue resistance with no deviation for 100 cycles. Dissipative particle dynamics simulations revealed that the orientation of hydrogen bonds along the stretching direction boosted mechanical strength and toughness, which were further proved by the restriction of molecular chain movements ascribing to the formation of a dense hydrogen-bonding network from mean square displacement calculations. Combining with high ionic conductivity over a wide temperature range and autonomous adhesion on various surfaces with tailored adhesive strength, the PAM-r-MVIC can readily work as a highly stretchable and healable ionic conductor for a capacitive/resistive bimodal sensor with self-adhesion, high sensitivity, excellent linearity, and great durability. This study might provide a new path of designing and fabricating ionic conductive hydrogels with high mechanical elasticity, high toughness, and excellent fatigue resilience for skin-inspired ionic sensors in detecting complex human motions.",
"introduction": "1. Introduction A skin-inspired ionic sensor is widely concerned in the next generation of smart wearable electronics for the applications of artificial intelligence, human-machine interfaces, healthcare monitoring, and soft robotics [ 1 – 3 ]. An ionic sensor is capable of sensing external stimulations and transforming them into conductivity signals rapidly and in real time by imitating human skin [ 4 ]. The real-time response of an ionic sensor is realized through the directional migration of ions in an ionic conductor under deformation, which can realize the integrated functions of high elasticity and skin comparable modulus that are difficult to realize in a traditional electronic conductor. Due to the frequent movement of the human body and its rough and complex surface, ion sensors would inevitably be damaged and fall off during long-term wearing, which puts forward high requirements for tailored adhesive performances of ionic sensors [ 5 , 6 ]. However, traditional adhesives are very difficult to meet the requirements of ionic sensors in long-term wearing or multiple-time adhering. Moreover, for their practical applications of human-machine interfaces, intelligent windows, and touchscreens, not only the ability to perceive external stimuli but also a high transmittance to achieve an output of visual information is required for an ionic sensor [ 7 ]. Therefore, the development of an ionic sensor with high transparency, high mechanical robustness, adaptive self-adhesion, and long cycling life is extremely demanded. Poly(ionic liquid)s (PILs) are polymers formed through polymerization of ionic liquid (IL) monomers that feature repeated anionic or cationic groups. PIL hydrogels are capable of combining unique integrations of attractive mechanical characteristics of hydrogels and superior physicochemical properties of ILs [ 8 – 10 ]. PIL hydrogel is more thermally stable with strongly locked counter ions, and these features can help to overcome the leakage and poor environmental tolerance of conventional salt-impregnated polymer hydrogels. However, PIL hydrogel usually has poor mechanical ductility, leading to irreversible mechanical failures and poor cycling stability under large deformation. More importantly, PIL hydrogel is difficult to achieve high and tailored adhesion on various surfaces. Therefore, solving the problems mentioned above, i.e., achieving high mechanical elasticity, excellent fatigue resistance, and self-adhering performance, is necessary for the wide applications of PIL hydrogels for high-performance ionic sensors [ 11 – 13 ]. The human body mainly relies on ion channels inside the neurons to transmit information, and an electrolyte in organisms as an ionic conductor plays an essential role [ 14 , 15 ]. This study attempts to construct a highly stretchable ionic conductive hydrogel to imitate the sensing functions of the skin. Herein, a highly stretchable and transparent poly(1-methyl-3-(4-vinylbenzyl) imidazolium chloride)-random-polyacrylamide copolymer hydrogel (PAM-r-MVIC) is fabricated by a hydrogen-bonding network densification strategy. Ascribing to the formation of a dense hydrogen-bonding network, the resultant PAM-r-MVIC exhibits excellent ductility with a large fracture strain (>1000%), along with high tensile strength of 0.47 MPa, high toughness of 2950 kJ m −3 , and excellent fatigue resistance with no deviation for 100 cycles. Dissipative particle dynamics simulations further reveal that the mechanical properties are enhanced by the orientation of the dense hydrogen-bonding network along the stretching direction. Besides, mean square displacement calculations indicate that the high density of hydrogen bonds displays a large restriction of molecular chain movements, which also leads to dramatic enhancements in the mechanical strength. The positively charged 1-N atoms of imidazole rings among the PAM-r-MVIC are beneficial for uniformly locking counter ions, contributing to excellent ionic conductivity in a wide temperature range. Due to its high mechanical elasticity, high ionic conductivity, good transparency (close to 100% in visible light range), and excellent self-adhering properties, the PAM-r-MVIC can readily work in a resistive/capacitive bimodal sensor, showing high sensitivity, wide response range, and excellent stability in real-time monitoring of large-strain movements (i.e., finger and wrist bending) and small-strain movements (i.e., swallowing) of complex human motions. Therefore, this newly developed hydrogen-bonding network densification strategy for the design and construction of functional ionic conductive hydrogels provides new ideas for the development of ionic skin sensors with high mechanical elasticity, good transparency, self-adhering property, and excellent durability in a wide temperature range.",
"discussion": "3. Discussion A novel PAM-r-MVIC hydrogel was prepared by a hydrogen-bonding network densification strategy, during which a chemically cross-linked random copolymer with an intermolecular dense hydrogen-bonding interaction was in situ formed. Benefitted from the imidazole groups in the hydrogel backbone and dense hydrogen-bonding network, the resultant hydrogels combined outstanding mechanical properties (e.g., high strength, stretchability, compressibility, toughness, fast self-recovery, and fatigue resistance), high transparency of nearly 100% in the visible light range, and excellent ionic conductivity. The DPD simulations and MSD calculations further proposed the mechanical enhancement mechanism for the PAM-r-MVIC hydrogel, manifesting that the orientation of hydrogen bonds during the stretching and the restriction of molecular chains by the dense hydrogen-bonding network synergistically led to dramatically improved mechanical strength and toughness. More importantly, the PAM-r-MVIC hydrogel is capable of being adhered to diverse surfaces, such as metal, wood, plastic, glass, and skin, and the tailored adhesive mechanical strength was highly maintained after repeated adhering and stripping cycles for 100 times. Besides, the PAM-r-MVIC hydrogel is capable of maintaining high ionic conductivity, excellent compressive sensitivity, and high durability at extremely cold temperatures, allowing them to be designed as a capacitive/resistive bimodal sensor for human-motion detections. The PAM-r-MVIC sensors could accurately monitor both large-range human movements and small stress changes, such as the motions of finger bending, wrist flexion, and swallowing. It was envisioned that the hydrogen-bonding network densification strategy provided a new path for the preparation of ionic conductive hydrogels with high mechanical elasticity, excellent fatigue resilience, high sensitivity, and outstanding durability in a wide temperature range for skin-inspired ionic sensors."
} | 2,134 |
30034977 | null | s2 | 5,016 | {
"abstract": "We report a new class of textiles with electrochemical functions which, when moistened by a conductive liquid (saline solution, sweat, wound fluid, etc.), generate DC voltage and current levels capable of powering wearable electronics on the go. Contrary to previously reported power generation techniques, the proposed fabrics are fully flexible, feel and behave like regular clothing, do not include any rigid components, and provide DC power via moistening by readily available liquids. Our approach entails printed battery cells that are composed of silver and zinc electrodes deposited onto a polyester fabric to generate power in the microwatt range. Electrochemical characterization of the discharge of a single printed battery cell in a 10 M NaOH electrolyte shows reproducible results with a sustained power level of ∼80 "
} | 207 |
35478808 | PMC9034150 | pmc | 5,018 | {
"abstract": "In order to improve the waterproof and mildew resistance of electronic equipment, a superhydrophobic coating was prepared on a circuit board. First, hexadecyl trimethoxysilane was used to modify the nano silica and nano zinc oxide particles, and then the modified nanoparticles were mixed with the silica sol. Then the superhydrophobic coating was prepared on the surface of the printed circuit board by a spraying process. The preparation technology and physical and chemical properties of the coating were studied. The contact angle of the final sample can reach 169.47°, the sliding angle can reach 1.2°, it has good acid and alkali corrosion resistance, resistance to NaCl, self-cleaning performance and antimildew performance.",
"conclusion": "4. Conclusions The nano particles were modified by hexadecyl trimethoxysilane. The modified nano particles were compounded with silica sol, and then prepared by one-step spray method SiO 2 & ZnO/silica sol composite superhydrophobic coating. The effects of the mass ratio of SiO 2 to ZnO, the mass ratio of surface modifier to nanoparticles, and the mass ratio of nanoparticles to silica sol on the hydrophobicity of the coating were investigated. The results show that when the mass ratio of SiO 2 to ZnO is 2 : 1, the mass ratio of surface modifier to nanoparticles is 4.5%, and the mass ratio of nanoparticles to silica sol is 5%, the superhydrophobic coating has the best hydrophobicity, the contact angle can reach 169.47° and the sliding angle can reach 1.2°. The superhydrophobic coating has good mildew resistance, corrosion resistance, friction resistance and self-cleaning performance.",
"introduction": "1. Introduction Electronic equipment will be covered by some dust in the process of use, and some special electronic equipment is used in acid–base, seawater and humid environments, so these external factors will seriously affect the heat dissipation and service life of the circuit board. 1–3 At present, although some circuit board protective adhesives can achieve water and corrosion resistance, there are still some problems, for example, the self-cleaning and antimildew performance of the coating is poor, which increases the failure rate of electronic products and reduces the service life of electronic products. A superhydrophobic surface means that the contact angle between the water droplet and the contact surface is greater than 150° and the sliding angle is less than 10° at the same time. The water droplet will remain spherical on the superhydrophobic surface and roll easily, 4,5 therefore, superhydrophobicity has great potential application value in waterproofing, self-cleaning, corrosion resistance and other aspects, 6–16 but rarely reported in electronic products such as circuit boards. There are two key factors to construct superhydrophobic surface: one is that the surface has a certain roughness, the other is that the surface has a low surface energy. 17,18 In nature, the appearance of the muddy lotus and the rolling water droplets on the lotus leaf are the result of the special structure on its surface. 19,20 Based on the above two key factors, scientists have explored a number of methods for preparing super hydrophobic surfaces: sol–gel, layer by layer self-assembly, surface etching, electrospinning, film forming, vapor deposition, etc. 21–27 Sol–gel method has many advantages, such as easy control of reaction process, high uniformity of coating and wide application of substrate. For example, Wang et al. , 3 a fluorinated silica sol based super hydrophobic coating was prepared on the printed circuit board by sol–gel method. The contact angle of the coating was 158°, the sliding angle is 3°. Yang et al. , 28 the superhydrophobic coating of fluorinated functionalized titanium dioxide sol was prepared on the surface of cotton fabric by sol–gel method. The contact angle of samples was 160°. In this paper, silica sol was prepared by sol–gel method, and then the different particle size of SiO 2 & ZnO modification, then the superhydrophobic composite coating was prepared by mixing silica sol and modified nanoparticles. The effects of the mass ratio of silica to zinc oxide, the mass ratio of surface modifier to nanoparticles, and the mass ratio of nanoparticles to silica sol on the hydrophobicity of the coating were investigated, and the microstructure of the coating was characterized. The chemical and mechanical durability of superhydrophobic coatings, such as acid and alkali corrosion resistance, friction resistance and mildew resistance, were studied.",
"discussion": "3. Results and discussion 3.1. Influencing factors of coating wettability 3.1.1. Influence of preparation process There are two key factors to construct superhydrophobic surface, surface roughness and low surface energy. In this paper, the micro–nano rough structure is constructed by using SiO 2 and ZnO with different particle sizes, the influence of mass ratio of SiO 2 to ZnO on the hydrophobicity of the coating was investigated, the test results of contact angle and sliding angle are shown in Fig. 2(A) . At this time, the mass ratio of surface modifier to nano particle is 3%, the mass ratio of nanoparticles to silica sol is 3%. It can be seen from Fig. 2(A) that the contact angle is 164° and the sliding angle is 8.2° when SiO 2 is used alone. The contact angle is 167.5° and the sliding angle is 10.3° when ZnO is used alone. When SiO 2 and ZnO are mixed, with the increase of ZnO content, the SiO 2 content decreases, and the contact angle basically remains stable, while the sliding angle first decreases and then increases, this is because the mixed use of nanoparticles with different sizes can better construct the rough structure. Fig. 3(a) and (c) shows the SEM picture of SiO 2 and ZnO with mass ratio of 1 : 0, Fig. 3(b) and (d) are SEM images of SiO 2 and ZnO with mass ratio of 1 : 1. It can be seen from Fig. 3(a)–(d) that the surface with the mass ratio of SiO 2 to ZnO of 1 : 1 is coarser than the surface with the mass ratio of SiO 2 to ZnO of 1 : 0, and has the micro–nano rough structure required for superhydrophobic surface. When the two nanoparticles are mixed, the hydrophobicity of the coating has little change, this is because when the surface roughness of the coating reaches a certain level, the surface energy of the coating surface plays a very important role. Fig. 2 The effect of SiO 2 and ZnO quality ratio on hydrophobicity of coating (A); the effect of mass ratio of surface modifier and nanoparticles on hydrophobicity of coating (B); the effect of the mass ratio of nano particles and silica sol on hydrophobic properties of coating (C). Fig. 3 SEM images of different samples. SEM with mass ratio of SiO 2 to ZnO 1 : 0 (a and c); SEM with mass ratio of SiO 2 to ZnO 1 : 1 (b and d); SEM with 1% mass ratio of nano particles to silica sol (e and g); SEM with 1% mass ratio of nano particles to silica sol (f and h). The mass ratio of SiO 2 to ZnO is 2 : 1 and the mass ratio of nano particle to silica sol is 3%, other production processes remain unchanged, the effect of the mass ratio of surface modifier to nanoparticles on the hydrophobicity of the coating was investigated, the test results of contact angle and sliding angle are shown in Fig. 2(B) . As can be seen from Fig. 2(B) , when the mass ratio of HDTMS to nanoparticles is 1.5%, the contact angle is 121° and the sliding angle is 58.3° and the superhydrophobic effect cannot be achieved. With the increase of HDTMS content, the contact angle increases and the sliding angle decreases. When the content of HDTMS is 4.5%, the maximum contact angle is 168.3° and the sliding angle is 4.3° lower. With the increase of HDTMS content, the stronger the ability of nanoparticles to resist hydrophilic groups, the smaller the surface energy and the stronger the hydrophobicity of the coating. After adding HDTMS, the contact angle and sliding angle remained stable, this is because the surface energy of the surface is difficult to continue to decrease when it is low to a certain extent, considering the economic and practical effects, the mass ratio of HDTMS to nanoparticles is 4.5%. The mass ratio of SiO 2 to ZnO is 2 : 1, the mass ratio of HDTMS to nanoparticles is 4.5%, and other processes remain unchanged, the influence of the mass ratio of nano particles to silica sol on the hydrophobicity of the coating is explored. The test results of contact angle and sliding angle are shown in Fig. 2(C) . As can be seen from Fig. 2(C) . When the mass ratio of nanoparticles to silica sol is 1%, the contact angle and sliding angle are 168.2° and 3.2°. When the mass ratio of nanoparticles to silica sol is 5%, the contact angle is 169.47° and the sliding angle is 1.2°. At this stage, with the increase of the amount of nanoparticles, the hydrophobicity of the coating becomes better, this is due to the increase of nanoparticles, the structure of a better rough structure. When the mass ratio of nano particles to silica sol increases to 9%, the contact angle is 158.9° and the sliding angle is 9.7° and the hydrophobicity of the coating decreases, the reason is that the content of silica sol decreases, resulting in a large number of nanoparticles cannot adhere to the coating surface, and cannot create a better rough structure, so the hydrophobicity of the coating decreases. Fig. 3(e) and (g) shows SEM with 1% mass ratio of nanoparticles to silica sol, and Fig. 3(f) and (h) shows SEM with a mass ratio of 9% between nanoparticles and silica sol. Compared with the low power mirror (e and f), it can be found that the sample with mass ratio of 1% adheres more coatings on the surface, so it has better hydrophobicity. 3.1.2. Influence of surface chemical composition and morphology The FTIR spectra of HDTMS, SiO 2 , ZnO and their mixed modified nanoparticles are shown in Fig. 4(A) . The absorption peak arising in the spectrum of modified nanoparticles at 3353.05 cm −1 is attributed to stretching and bending vibrations of hydroxyl or bridged hydroxyl groups as a result of dissociation of a large number of water molecules on the surface of nano-SiO 2 and ZnO. 31 In Fig. 4(A) , spectral line a is compared with spectral lines c and d, 2976.93 cm −1 and 2910.86 cm −1 is the stretching vibration peak of –CH 3 and –CH 2 , this indicates that HDTMS is successfully grafted on nanoparticles. The absorption peak at 1646.97 cm −1 is the water content in potassium bromide. The peak at 1381.42 cm −1 is the vibration generated by ester group (–COOR) radicals of a coupling agent. Those at 1086.78 cm −1 and 1043.27 cm −1 are brought by the anti-symmetric and symmetric contraction vibrations of Si–O–Si bonds. 32 Finally, the signature at 879.15 cm −1 refers to the contraction vibration of Zn–O bonds that are formed by the dehydration reaction between HDTMS hydrolyzed to Si–OH and –OH radicals on the SiO 2 and ZnO surfaces, as shown in Fig. 4(B) . 33 Fig. 4 FTIR spectra of different samples (A); possible reaction mechanism of HDTMS modified nanoparticles (B). \n Fig. 5 SEM of pristine PCB surface, it can be seen from the low power Fig. 5(a) that there are some rough structures arranged disorderly on the surface of the pristine PCB, however, in the high power Fig. 5(c) , the surface is smooth and there is no obvious rough structure, which indicates that the surface of the untreated circuit board has no rough structure needed to construct superhydrophobic surface. SEM of superhydrophobic PCB surface (d–f), in the low power Fig. 5(d) , it can be seen that there are many synaptic nanoparticles on the surface of the sample, which are densely and neatly arranged on the surface, this is a micro–nano rough structure composed of hydrophobically modified nanoparticles added into the coating, in the high power Fig. 5(f) , there is still a rough structure of nanometer scale. Compared with the two groups of micrographs, it can be found that the surface of the circuit board treated by superhydrophobic coating has a better rough structure, so it has a better hydrophobicity. It can be found from Table 1 that the content of C, N, Si and Zn in the circuit board treated by superhydrophobic coating is obviously increased, especially Si and Zn, this is corresponding to the element content of HDTMS used in the modification of nanoparticles, which further indicates that the nanoparticles are successfully mixed into the coating and the grafting of HDTMS is successful. Fig. 5 SEM of pristine PCB surface (a–c); SEM of superhydrophobic PCB surface (d–f). EDS comparison between pristine and superhydrophobic PCB Element wt% PCB Pristine PCB Superhydrophobic PCB C 53.49 59.46 N 03.05 04.65 O 09.94 04.37 Si 01.19 02.73 Zn 00.06 00.63 3.2. Stability analysis 3.2.1. Corrosion resistance Electronic equipment in the process of using will encounter some harsh environment, such as acid rain, sea water, etc. , these bad factors will corrode the circuit board, so the circuit board has good corrosion resistance is very necessary. The samples were soaked in aqueous solutions with different pH values or NaCl (3.5 wt%) aqueous solutions, and then washed and dried with water at intervals. The changes of surface wettability of the samples were recorded. The wettability relationship of samples immersed in aqueous solutions with different pH values for 12 h is shown in Fig. 6(A) , the wettability of the sample immersed in NaCl (3.5 wt%) solution for different time is shown in Fig. 6(B) , the Nyquist plots of bare Cu samples, superhydrophobic samples and superhydrophobic samples soaked in NaCl (3.5 wt%) solution for different times are shown in Fig. 6(C) . Fig. 6 Wettability diagram of samples soaked in different pH aqueous solutions for 12 h (A); wettability of samples immersed in NaCl (3.5 wt%) aqueous solutions for different time (B); Nyquist plots of different samples (C). It can be seen from Fig. 6(A) , the dramatic changes in both parameters are observed within the whole pH scale, particularly in the strong acid (1–5) and alkali (9–12) ranges. However, even after 12 h of soaking in the aqueous solutions with pH values of 1 and 12, the contact angle can still reach the values of 153°–154°, whereas the sliding angles may strongly deviate from the references. In the case of pH 5–9, the contact angle is greater than 150° and the sliding angle is less than 10°. Therefore, coatings remain superhydrophobic and exhibit good resistance to acids and alkalis at pH of 5–9, thus allowing the circuit boards to be applied within this pH range. Both wettability parameters were found to undergo drastic changes with increasing soaking time in NaCl aqueous solutions: whereas the contact angle decreased from 169.9° to 149.6°, the sliding angle rose from 2.2° to 34.5°. It can be seen from Fig. 6(B) that the contact angle decreased by about 6° and the sliding angle increased by 6° after 48 h of soaking. After 144 h, the static contact angle was 152.3°, still achieving the superhydrophobic effect, and the sliding angle increased to 29.8°. With the increase of time, Na + and Cl − begin to destroy the waterproof barrier of superhydrophobic coating, invade into the coating, and then destroy the sample. On the whole, the sample has good stability in NaCl aqueous solutions. It can be seen from Fig. 6(C) that the charge transfer impedance of the superhydrophobic sample is about 10 000 times that of the bare Cu sample, the results show that the superhydrophobic coating has good corrosion resistance. With the increase of immersion time in NaCl (3.5 wt%) aqueous solutions, the charge transfer impedance of the sample decreases, but it still has good corrosion resistance compared with the bare Cu sample. The reason why the samples have good corrosion resistance is that the superhydrophobic coating has low surface energy and repels the solution. At the same time, air can be stored in the surface of a nano-rough structure, forming an air layer, thus protecting the circuit board from the exposure of other corrosive media. Since superhydrophobicity itself is the physical barrier against NaCl aqueous solutions immersion. The synergistic effect of the above three aspects enhanced the corrosion resistance of the samples. 3.2.2. Wear resistance test of coating The relationship between friction times and wettability of samples is shown in Fig. 7(A) and (B) shows SEM after 50 times wear of samples on 800 mesh sandpaper. Fig. 7 Relationship between friction times and wettability of samples (A); SEM of the worn surface of the sample for 50 times (B). It can be seen from Fig. 7(A) that both the contact angle and sliding angle noticeably vary within the entire friction time range. The contact angle of the coating decreases with increasing wear time. When the wear time reaches 40, the contact angle is still at the expected level (151.47°), showing the good superhydrophobicity. Fig. 7(B) shows the SEM of samples with different magnification after 50 times of friction. The coating surface friction marks, a portion of the coating was off, but still part of the coating is preserved, can be seen in Fig. 7[B(d)] at high magnification is still has many nanoscale rough structure, sample has good friction resistance. 3.3. Application analysis: mildew resistance and self cleaning performance The results of this sample under GB/T 2423.16-2008. It can be found from Fig. 8 that no obvious long mildew was found under 50 times magnification, and the mould resistance level reached grade 0, which indicated that the sample had good mildew resistance. The surface layer of superhydrophobic coating is composed of SiO 2 and ZnO nanoparticles, when ZnO is in contact with mildew, ZnO can produce Zn 2+ , which can combine with organic matter in mildew, destroy its internal structure, cause cell damage, and achieve antimildew effect. This is the chemical mechanism of antimildew of samples. The surface of superhydrophobic coating is micro–nano rough structure, and the air can be stored in the micro–nano rough structure to form an air layer, which can prevent mildew from adhering to and immersing in the sample. This is the physical mechanism of mildew resistance of the sample. The cooperation of the above two mechanisms makes the samples have good mildew resistance. Fig. 8 Sample magnification 50 times. \n Fig. 9 shows the self-cleaning performance of the sample by simulating dust with cement powder. The sample is tilted at 10° and the water drops slowly from 1 cm away from the sample surface through the dropper at the higher end of the sample. It is found that the water droplets will automatically adhere to the passing dust during the rolling process of the superhydrophobic PCB surface, so as to take away all the dust on the path. The water droplets are dry and clean on the rolling path, and there is neither residual dust nor adhesive water droplets, showing a good self-cleaning effect. However, water droplets will agglomerate with dust on the pristine PCB surface, which cannot achieve the effect of self-cleaning. Fig. 9 Self cleaning properties of the superhydrophobic PCB as compared with the pristine PCB. Cement soils were washed away by the water droplets impacted onto the superhydrophobic PCB (a) and wetted by the water droplets impacted to the pristine PCB (b)."
} | 4,837 |
35410135 | PMC8996662 | pmc | 5,020 | {
"abstract": "Background Arbuscular mycorrhizal (AM) fungi and roots play important roles in plant nutrient acquisition, especially in nutrient poor and heterogeneous soils. However, whether an accumulation strategy of AM fungi and root exists in such soils of karst shrubland ecosystems remains unclear. Root traits related to nutrient acquisition (root biomass, AM colonisation, root acid phosphatase activity and N 2 fixation) were measured in two N 2 -fixing plants (i.e. Albizia odoratissima (Linn. f.) Benth. and Cajanus cajan (Linn.) Millsp.) that were grown in heterogeneous or homogeneous nutrient (ammonium) soil with and without AM fungi inoculation. Results Both of these plants had higher AM colonisation, root biomass and relative growth rate (RGR), but lower N 2 fixation and root acid phosphatase activity in the rhizosphere in the heterogeneous soil environment, than that in the homogeneous soil environment. Plants grown in the AM fungi-inoculated heterogeneous soil environment had increased root biomass and root acid phosphatase activity compared with those grown in soil without inoculation. AM colonisation was negatively correlated with the N 2 fixation rate of A. odoratissima , while it was not significantly correlated with the root phosphatase activity. Conclusions Our results indicated that enhanced AM symbiosis and root biomass increased the absorptive surfaces for nutrient acquisition, highlighting the accumulation strategies of AM and root traits for plant nutrient acquisition in nutrient poor and heterogeneous soils of the karst shrubland ecosystem. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-022-03514-y.",
"conclusion": "Conclusions In this study, accumulation strategies between roots and AM fungi were shown to exist in the belowground nutrient resource foraging of N 2 -fixing plants in a karst shrubland ecosystem. The N 2 -fixing plants grown in nutrient poor and heterogeneous soil environments relied more on AM fungi and an increased absorptive root surface for the acquisition of nutrients. This result suggested that an increase in the AM colonisation of roots and an increase in root biomass beneficially increased the absorptive surfaces for the acquisition of nutrients under conditions of spatially heterogeneous soil nutrient availability. Plants grown in soil inoculated with AM could increase their root-shoot ratio to a higher degree in the heterogeneous soil environment than that in the homogeneous soil environment. AM fungi and N 2 fixing symbionts play important roles in plant nutrient acquirement. However, the relationship between AM colonisation and the N 2 fixation rate differ in the two N 2 -fixing plants, which indicated that host-specificity characteristics influence the nutrient acquisition strategies of plants. Our findings suggested that plants regulate root-mycorrhizal interactions to adapt to the nutrient poor and heterogeneous soil environments of karst shrubland ecosystems. Future studies combining plant functional groups with AM fungal species to measure how soil conditions, mycorrhizal type and root traits (e.g. root length and density) can collectively mediate resource acquisition strategies in belowground.",
"discussion": "Discussion Accumulation strategy of AM fungi and roots in plant nutrient acquisition Increasing evidence supports the idea that root traits (e.g. root biomass) exhibit wide variations in heterogeneous and homogeneous soil environments [ 12 , 13 ], which strongly influence the colonisation of AM with roots [ 10 , 24 , 25 ]. Our results showed that the plants had a higher root biomass and AM colonisation in the heterogeneous soil environment than in the homogeneous soil environment (Figs. 1 c and 4 a). AM colonisation was positively correlated with root biomass in the heterogeneous soil environment (Fig. 2 ). These results indicated that accumulation strategies involved in belowground resource acquisition existed between the roots and their associated AM fungi, which benefitted the nutrient acquisition of plants for growth in the poor and heterogeneous of soil environment of karst shrubland ecosystems. The accumulation strategies of plant nutrient acquisition between root biomass and AM fungi in poor and heterogeneous soil environments can be explained by the root absorption capacity [ 26 ]. For example, during seedling establishment in soil with low and heterogeneous levels of P, as in our study (soil available P content 1.92 mg kg −1 ), plant species increase their root absorption area as a strategy to increase their acquisition of soil nutrients. Larger absorption areas of the roots can be induced in more heterogeneous soil environments. Many previous studies have indicated that plants grown in heterogeneous soil environments preferentially partition more photosynthetic products to underground parts [ 12 , 13 ], and further increase their root biomass. More roots increase the absorption area of the roots, which benefits plant nutrient capture [ 6 ]. Furthermore, a large absorption area of the roots allows the roots to amplify the interface contact with soil hyphae, and thus increase the chance of attracting symbiotic AM fungi. Simultaneously, more roots can exude a sizeable quantity of polysaccharides [ 27 ], which can invest more C for AM colonisation and indirectly promote nutrient acquisition. Thus, a large absorption area of the roots can be gained by AM colonisation. Higher levels of AM colonisation can be found in heterogeneous soil environments compared with homogeneous soil environments [ 14 ], as found in the present study, which potentially enhances the nutrient acquisition for plant root growth. Importantly, the hyphae of AM can extend between rocks to reach areas that are not accessible to plant roots [ 28 ]. This characteristic is a very efficient way to generate absorptive surfaces for plant growth in karst regions with a high rock-soil ratio. Thus, increasing the AM colonisation and root biomass are two important strategies to construct absorptive surface areas for plants to adapt to nutrient poor and heterogeneous soils. Therefore, our results were consistent with the hypothesis that higher levels of AM colonisation and a larger root biomass increases the absorptive surface for plant nutrient acquisition, to maintain a high growth rate of plants in nutrient poor and heterogeneous soil of karst shrubland ecosystems. AM fungi symbiosis with root enables plants better acquirement soil nutrients, and greatly affect their plant growth. Many previous studies reported that plant growth responds different to inoculation AM fungi [ 29 , 30 ]. Our results showed that plant growth responses were negatively correlated with the inoculation of soil with AM fungi, independent of the heterogeneity or homogeneity of the soil environment. These findings were consistent with those of Zhang et al. [ 31 ], but inconsistent with those of Liang et al. [ 29 ] who demonstrated positive effects on plant growth when inoculated AM fungi. Several possible reasons could explain this phenomenon. First, differences in the plant functional group could influence the AM colonisation. For example, plants have coarse root architecture, including short root hairs, large root diameters and low root hair densities, are positively correlated with the plant growth responses to inoculation with mycorrhiza [ 4 , 10 ]. Certain plants have limited intrinsic abilities to acquire nutrients [ 31 ], and they thus obtain nutrients mainly depending on AM fungi. In contrast, AM fungi colonisation is only an alternative for plant species with fine root architecture (e.g. greater root density and hair length) to absorb nutrients [ 10 ]. Thus, plant growth of this kind of plant would be negatively correlated with the inoculation of mycorrhiza. The present study only assessed the growth responses of N 2 -fixing plants to AM fungi inoculation of the soil environment. Thus, the growth responses of plant functional groups, such as N 2 -fixing plants and non-N 2 -fixing plants, to AM fungi inoculation of the soil environment should be assessed and compared in future studies. Second, the plant growth responses to AM fungi inoculation were related to the species-specific interactions between AM fungi and the host plant. Some previous studies focused on inoculation AM fungi influencing plant growth mainly through the inoculation of commercial AM fungi strains [ 32 , 33 ]. Commercial AM fungi strains that are not optimally matched to the host plant lead to lower AM colonisation rates and less benefit from AM fungi. This action reduces the diversity of AM fungal species, which may play key roles in plant growth and even have negative effects on plant growth. The AM fungi colonisation of plants growing in natural soil was determined in the present study, which served as the best representative of the soil biota pool, including the total AM fungi. Therefore, plants were exposed to their natural AM fungi assemblages, which may have increased the actual benefits of the plant-AM fungi symbiosis. However, a negative effect of mycorrhizal inoculation on plant growth was found in the present study, which was related to the soil pathogens. Plants not only show host-specificity in symbiotic relationships with beneficial microbes (i.e. AM fungi) [ 34 , 35 ] but also share the same associations with pathogens [ 36 ]. Pathogens have negative effects on plant growth by inducing higher disease mortality rates in the plants [ 29 , 36 ], or competing with the plants for carbohydrates [ 37 , 38 ]. The treatment soil (e.g. without AM inoculate) was sterilised in the present study, which killed all pathogens and thus would enable the promotion of plant growth independent of soil heterogeneity and homogeneity (Fig. 1 ). The same findings were found by Zhang et al. [ 30 ]. Although plant growth was not significantly improved by the AM inoculation in the present study, it increased the root biomass in the heterogeneous soil environment compared with homogeneous soil environment. These findings suggested that plants would strengthen their symbiosis with AM, and then allocate more C to the roots and for hyphal production in heterogeneous soil environments, with a higher investment in root growth at the expense of shoot growth [ 39 – 41 ]. Similar results have been reported by previous studies, whereby plants increase their root:shoot ratio to enhance nutrient absorption in heterogeneous soil environments [ 42 , 43 ]. Roles of AM fungi in N and P nutrient acquisition in karst shrubland ecosystems AM fungi are well known to improve plant phosphorus nutrients, especially under low phosphorus conditions. For example, AM fungi improve the absorption of soil inorganic phosphorus through hyphae. Simultaneously, fungi produce phosphatase enzymes [ 5 , 44 ], which mineralise more organic phosphorus from ester-bound forms to the orthophosphate form to increase the plant uptake [ 19 – 21 ]. Therefore, AM fungi and root phosphatase enzymes are two vital phosphorus acquisition strategies for plants [ 5 , 45 ]. The results of the present study showed that AM colonisation and root phosphatase enzyme activity did not significantly correlation in heterogeneous and homogeneous soil environments (Fig. 3 a). These results suggested that AM fungi most likely enhanced inorganic phosphorus acquirement in the heterogeneous soil environment of karst shrubland ecosystems, which is consistent with as the previous studies reported [ 46 , 47 ]. The AM colonisation-N 2 fixation tripartite symbionts also played important roles in plant nutrient acquisition. The AM colonisation-N 2 fixation tripartite symbionts were much more beneficial for plant growth in natural nutrient limitation environments (e.g. N and P) [ 48 – 50 ]. A negative correlation between AM colonisation and N 2 fixation of A. odoratissima was found in the heterogeneous soil environment (Fig. 5 c). This relationship can be explained by the complementary strategies between AM fungi and N 2 -fixing symbionts in the nutrient acquisition of nitrogen and phosphorus. Nutrient acquisition strategies, including those of AM fungi and rhizobia, cost a large amount of resource [ 45 , 51 ]. From a cost–benefit perspective, plants select strategies that maximise the benefits while minimizing the costs [ 52 ]. Therefore, when nitrogen was abundantly available in the present study, plants absorbed nitrogen directly through their roots with less resource, and further reduced the N 2 fixation. Thus, plants mainly depend on AM for the acquisition of phosphorus in low phosphorus soils, and more C is invested for AM colonisation. However, no correlation between AM colonisation and N 2 fixation was detected in C. cajan. Therefore, the relationship between AM colonisation and N 2 fixation is more complex, and it is still unclear if fixing N 2 is necessary to acquire phosphorus or vice versa."
} | 3,247 |
24430239 | PMC4049913 | pmc | 5,021 | {
"abstract": "Renewable lignocellulosic plant biomass is a promising feedstock from which to produce biofuels, chemicals, and materials. One approach to cost-effectively exploit this resource is to use consolidating bioprocessing (CBP) microbes that directly convert lignocellulose into valuable end products. Because many promising CBP-enabling microbes are non-cellulolytic, recent work has sought to engineer them to display multi-cellulase containing minicellulosomes that hydrolyze biomass more efficiently than isolated enzymes. In this review, we discuss progress in engineering the surfaces of the model microorganisms: Bacillus subtilis, Escherichia coli, and Saccharomyces cerevisiae . We compare the distinct approaches used to display cellulases and minicellulosomes, as well as their surface enzyme densities and cellulolytic activities. Thus far, minicellulosomes have only been grafted onto the surfaces of B. subtilis and S. cerevisiae , suggesting that the absence of an outer membrane in fungi and Gram-positive bacteria may make their surfaces better suited for displaying the elaborate multi-enzyme complexes needed to efficiently degrade lignocellulose.",
"introduction": "Introduction Dwindling supplies of petroleum and the need to reduce net carbon emissions have driven the search for innovative and cost-effective methods to produce biofuels, chemicals, and materials from lignocellulosic biomass. 1 In the United States alone, it is estimated that over 1 billion tons of non-food lignocellulosic biomass can be produced annually on a sustainable basis at costs of only $40–50 per ton. 2 , 3 However, a major obstacle limiting the use of lignocellulose as feedstock is its recalcitrance to degradation. 2 , 4 While a number of technologies are being explored in industry to degrade lignocellulose, enzyme-based methods predominate, and are currently used to produce cellulosic ethanol ( Fig. 1A ). 4 In this hydrolytic method, plant biomass is degraded in a two-step process in which it is first pretreated with various chemicals (e.g., acids or ionic liquids) to expose and partially degrade the cellulose and hemicellulose sugar polymers, and then hydrolyzed by adding a consortium of purified cellulase enzymes. 4 - 9 Yeast then ferments the sugars into ethanol. To produce biomass-derived commodities cost-effectively, several groups are developing consolidated bioprocessing (CBP) microbes that combine cellulase production, cellulose hydrolysis, and fermentation into a single process ( Fig. 1B ). In principle, their use would significantly lower costs, as it would circumvent the need for adding purified cellulase enzyme cocktails and hydrolysate separation procedures. 10 - 13 Avoiding the use of purified enzyme cocktails would be particularly advantageous as it is currently the single largest contributor to overall costs ($0.68–1.47 per gallon of cellulosic ethanol). 14 An ideal CBP-enabling microbe would catabolize biomass efficiently and completely, utilize all of the sugars released from the biomass, and generate products at good yields, rates, and titers. It would also require minimal nutrient supplementation, be tolerant to low pH and high temperatures, and possess generally regarded as safe (GRAS) status. 2 , 13 Many promising CBP-enabling microbes possess several of these characteristics, but they are unable to degrade and use biomass as a nutrient. To overcome this limitation, several groups have devised methods to create recombinant cellulolytic microbes that deconstruct plant biomass using surface displayed cellulases. Figure 1. Consolidated bioprocessing of lignocellulosic biomass by cellulase displaying microbes. ( A ) The current steps involved in the industrial processing of plant biomass into ethanol using cellulase enzymes. Biomass degradation involves thermochemical pretreatment to expose its cellulose polymers, followed by exposure to purified cellulases to degrade cellulose into its component sugars. This is followed by a fermentation step in which yeast convert the sugars into ethanol. In principle, many other biocommodities can be produced from plant biomass using similar methods. ( B ) Steps in the consolidated bioprocessing (CBP) of biomass. CBP-enabling microbes would produce cellulase enzymes that degrade the cellulose and hemicellulose components of biomass, and then convert the resultant sugars into useful biocommodities. Microbes that naturally or recombinantly display cellulase enzymes are well suited for this process, as they are highly cellulolytic. In order to degrade the complex structure of plant biomass, naturally cellulolytic microbes produce an array of cellulases that have different substrate specificities. Although a variety of plants are being considered as industrial feedstocks (corn stover, straw, Miscanthus , switchgrass, poplar, sugarcane bagasse, etc.), their cell walls all contain lignocellulose which is comprised of varying amounts of cellulose (25–55%), hemicellulose (8–30%), and lignin (18–35%) ( Fig. 2 ). 15 The most abundant component, cellulose, is a homopolymer of β-1,4-linked glucose molecules that are hydrogen bonded with other cellulose polymers to form both amorphous and crystalline regions, the latter of which is particularly recalcitrant to degradation. 16 Naturally cellulolytic microorganisms produce three main types of cellulases that function synergistically: endoglucanases, exoglucanases, and β-glucosidases. 17 Endoglucanases hydrolyze internal β-1,4-glucosidic bonds in the polymer, creating reducing and non-reducing ends that are further hydrolyzed by exoglucanases. 18 Working together, the enzymes create shorter cellodextrins, including the disaccharide cellobiose, which is degraded into its component sugars by β-glucosidases. 18 The hemicellulose component of lignocellulose is a heterogeneous polymer of pentose and hexose sugars. 19 To liberate these sugars, microbes employ a variety of hemicellulases that have distinct substrate specificities, including exoxylanases, endoxylanases, arabinases, and mannanases, among others. 20 Finally, the cellulose and hemicellulose carbohydrate polymers are embedded in lignin, a complex polymer containing a mix of phenolic compounds connected by a variety of linkages. 21 Microbial lignin degradation remains poorly understood, but in white-rot fungi, it is mediated by a combination of extracellular peroxidases and laccases. 22 Figure 2. Schematic showing the structure of lignocellulose. Lignocellulose is composed of three components, including cellulose (solid tan lines), hemicellulose (dotted black lines), and lignin (solid brown lines). Cellulose is composed of β-1,4-linked glucose polymers, while hemicellulose is composed of a variety of pentoses. Lignin, which provides structural support for lignocellulose, coats these polymers. Recent work has engineered microorganisms to display multi-cellulase containing complexes called minicellulosomes ( Fig. 3 ). These complexes are miniaturized versions of the cellulosomes used by naturally cellulolytic anaerobes to degrade plant biomass. Native cellulosomes contain a variety of cellulases that function synergistically to degrade biomass more efficiently than isolated enzymes. The cellulosome from the cellulolytic thermophile Clostridium thermocellum is archetypal ( Fig. 3A ). It contains a central scaffoldin protein, CipA, which coordinates the binding of nine cellulases. 23 Binding is mediated by type-I cohesin modules within CipA that interact with sub-nanomolar affinity with type-I dockerin modules present in the cellulases. 24 CipA also contains a carbohydrate-binding module (CBM) that tethers the cellulosome complex to its substrate, as well as a type-II dockerin module located at its C-terminus that anchors the cellulosome complex to cell wall associated proteins. 25 Other species of anaerobic bacteria also display cellulosomes, which can adopt more elaborate structures that contain as many as 96 enzymes. 24 Figure 3. The prototypical CipA cellulosome and methods used to recombinantly display miniaturized cellulosomes (minicellulosomes). ( A ) Architecture of the prototypical CipA cellulosome produced by C. thermocellum . It houses 9 cellulases enzymes that are bound to the central scaffoldin protein, CipA. 23 Binding is mediated by type-I cohesin modules within CipA that interact with sub-nanomolar affinity with type-I dockerin modules present in the cellulases. CipA also contains a carbohydrate-binding module (CBM) that tethers the cellulosome complex to its substrate, as well as a type-II dockerin module located at its C-terminus that anchors the cellulosome complex to the cohesin module of cell wall associated proteins. ( B ) Ex vivo approach used to display minicellulosomes on the surfaces of B. subtilis or S. cerevisiae . The microbes secrete and display a scaffoldin protein that is displayed on their surface. Cellulase enzymes containing the appropriate type-1 dockerin module are incubated with the cells to construct the minicellulosome. The enzymes that are added to the cells are either purified enzymes or secreted by other microbes as part of a microbial consortium. Distinct colors are used to indicate species-specific type-1 dockerin and cohesin domains that selectively interact with one another to construct the minicellulosome. (C) Self-assembled approach used to display minicellulosomes on the surfaces of B. subtilis or S. cerevisiae . All of the components of the minicellulosome (scaffoldin and enzymes) are produced by the microbe and spontaneously assemble on the cell surface. Surface displayed minicellulosomes exhibit enhanced cellulolytic activity. Studies have shown that co-localizing cellulases with different substrate preferences into a cellulosome facilitates enzyme-enzyme synergism; the enzymes in the complex collectively exhibit greater cellulolytic activity than the sum of the activities of the isolated enzymes. 26 Synergy occurs because the enzymes have complementary activities, and their spacing and relative abundance is presumably optimal. The presence of both hemicellulolytic and cellulolytic activities in the cellulosome is also advantageous, since by working together these enzymes can remove “physical hindrances” blocking substrate access (e.g., hemicellulolytic enzymes degrade hemicellulose polysaccharides that might otherwise block access to cellulose). The displayed cellulosomes also tether the microbe to the biomass, thereby promoting cellulose-enzyme-microbe (CEM) synergistic interactions that increase the rate of hydrolysis. 27 CEM interactions minimize the distance over which the hydrolysis products must diffuse to the cell, facilitating more efficient sugar uptake and preventing the build-up of potential enzyme inhibitors (e.g., cellobiose and glucose). 28 It may also facilitate biomass degradation by promoting favorable substrate channeling of long-chain hydrolysis products to proximally bound cells. Thus, CBP-enabling microbes that display minicellulosomes should degrade biomass more rapidly and thoroughly than microbes that only secrete cellulases. There have been many excellent reviews describing efforts to create cellulolytic and consolidated bioprocessing microorganisms. 2 , 10 , 13 , 29 In this review, we focus solely on recent synthetic biology efforts to construct microbes that display cellulases and miniaturized cellulosomes (minicellulosomes). Specifically, we review progress in engineering three microorganisms: Saccharomyces cerevisiae, Escherichia coli , and Bacillus subtilis. Because they are well studied and robust genetic tools are available to manipulate them, they serve as model organisms for eukaryotes, and Gram-negative and Gram-positive eubacteria, respectively. Here we discuss the distinct approaches used to display cellulase complexes on their structurally unique surfaces, and we compare the cellulolytic activities that have been thus far achieved. This exciting work may lead to the direct use of these microbes in consolidated bioprocessing and it promises to facilitate the engineering of other industrially useful microbes."
} | 3,041 |
33673106 | PMC7918670 | pmc | 5,025 | {
"abstract": "Currently, polymers are competing with metals and ceramics to realize various material characteristics, including mechanical and electrical properties. However, most polymers consist of organic matter, making them vulnerable to flames and high-temperature conditions. In addition, the combustion of polymers consisting of different types of organic matter results in various gaseous hazards. Therefore, to minimize the fire damage, there has been a significant demand for developing polymers that are fire resistant or flame retardant. From this viewpoint, it is crucial to design and synthesize thermally stable polymers that are less likely to decompose into combustible gaseous species under high-temperature conditions. Flame retardants can also be introduced to further reinforce the fire performance of polymers. In this review, the combustion process of organic matter, types of flame retardants, and common flammability testing methods are reviewed. Furthermore, the latest research trends in the use of versatile nanofillers to enhance the fire performance of polymeric materials are discussed with an emphasis on their underlying action, advantages, and disadvantages.",
"conclusion": "6. Conclusions and Outlook In this review, we covered the flame-retardant combustion processes and recent flame-retardant technologies for polymers. Designing flame-retardant polymers with potential for producing new fire-safe materials is essential because large fires result in the loss of human life. Despite the challenges, flame-retardant technology is improving, and the potential fire-resistant polymer materials can decrease fire risks and damage. Understanding the basic processes of fire resistance and researching new flame-retardant processes is crucial as it is directly connected to the safety of citizens and fire prevention. In recent years, flame-retardant performance test methods such as the upgraded cone calorimeter have made it easier to analyze toxic combustion gases in microscale units, which is expected to facilitate research on flame retardants. Active gas-phase flame retardants are being studied more intensively because they can circumvent the environmental issues of halogenated flame retardants. In contrast, along with improving the mechanical properties, the incorporation of nanofillers into polymer matrices has been widely studied to improve the fire performance of polymeric materials. Appropriate nanofillers can result in improved fire performance and significantly elevate the smoke suppression effect of polymers. Some graphitic carbon and inorganic nanoparticles incorporated into polymer matrices in a controlled fashion have been demonstrated as capable of enhancing the polymer fire performance ( Table 9 ). Specifically, one-dimensional SWNTs and MWNTs form network-like protective layers during combustion, which shield the surrounding/underlying polymer matrix from external radiation and heat feedback from flames. Strong nanoparticle–polymer interactions increase the viscosity of polymer melt, which can increase the nanoparticle concentration and improve the fire performance. However, in the case of SWNTs, there are limited studies showing that SWNT/polymer composite materials have effective fire performance. In contrast, several studies have employed MWNTs for enhancing the fire performance of polymers, as the interacting MWNTs result in the formation of a compact layer that protects the substance from flames. The important factor is the dispersion of SWNTs and MWNTs in the polymers because this determines the flammability effect of the polymer nanocomposite. Additionally, 2D graphene affords a strong barrier effect in polymer matrices, resulting in a high thermal stability. Graphene reduces heat release and mass transfer, as well as delays the fire ignition time, because it can promote the formation of a dense, continuous char layer during the decomposition process, which can act as a barrier to prevent heat transfer from the heat source. Furthermore, 2D layered inorganic materials such as layered double hydroxides (LDHs) have exhibited similar barrier properties [ 252 ]. One of the latest trends in developing new flame-retardant systems is combining more than two flame-retardant components. A representative example of combining different types of flame retardants is shown in Figure 40 . Similarly, several flame retardants can be combined in the nanometer regimen to create promotional effects. As mentioned before, nanofillers themselves do not usually show outstanding fire resistance such as self-extinguishing properties, and thus, they are commonly coupled with other flame-retardant additives. For improved fire performance, it is also essential to control the dispersion of nanofillers with other additives in polymer matrices. The efficient dispersion of nanofillers enables a substantial reduction in the loading amount. Moreover, the contribution of each type of nanoparticle to the fire performance of the polymer varies and strictly depends on the chemical structure and geometry [ 72 ]. The large interfacial contact area between the nanofiller and polymer may enhance catalytic effects such as the catalysis of charring reactions or radical trapping processes. It has been found that the formation of a continuous protective layer consisting of a network of nanoparticles is a key flame-retardant process for nanofillers, where the layer appears to act as a physical shield [ 188 ]. The formation of such a network is also important in reinforcing other physical properties of polymeric materials. Furthermore, the polymer composites should meet all the requirements in terms of major properties for practical applications, which is still a challenge. Some existing flame retardants (e.g., typically halogenated compounds) are subject to various regulations, since many environmental side-effects have been reported. The chemicals have the potential to volatilize or leach into the environment, where they can accumulate in fatty tissues and thus enter the food chain. In addition, the fire products from halogenated flame-retardant system are mostly both toxic and acidic, causing post-fire corrosion issues. Therefore, it is desirable that the flame retardants developed in the future have little effect on the environment. Research on reactive flame-retardant approaches that can be incorporated directly into polymer structures or the development of recyclable flame retardants is still actively underway, but more time is expected to be necessary to reach economic feasibility. In the future, more focus should be placed on the development of more environmentally friendly flame retardants. Furthermore, we believe that the development of inherently fire-resistant polymers will also attract attention. It will be highly important to identify the fire-resistant units and properly combine them with other units in developing new fire-resistant polymers.",
"introduction": "1. Introduction According to the statistics from the National Emergency Management Agency of South Korea during the period of 2010 to 2020, the number of large-scale fires (standard: 5 deaths, 10 casualties, and $4 million of property damage) increased six-fold from 3 to 18, and the casualties (from 45 deaths in 2010 to 232 deaths in 2019) and property damage costs (from $5 million to $330 million) increased significantly as well. These fires were mainly caused due to electrical and mechanical faults, with unknown causes also accounting for a significant proportion of these large-scale fires [ 1 , 2 , 3 ]. Based on the fire statistics for 2019, burns, smoke, and inhalation of toxic gases were the main reasons for the casualties [ 4 , 5 , 6 , 7 ]. Thus, fire protection becomes crucial; however, it is significantly challenging. As shown in Figure 1 , a typical fire scenario includes several processes. First, ignition, which is defined as the initiation of combustion, occurs; this is followed by fire growth, which is defined as the fire development stage during which the heat release rate and fire temperature increase. During the initial stage of a fire outbreak, the fire spreads quickly, and within a few minutes, the generated smoke and heat result in “flashover.” Once the fire has reached this stage, it is difficult to control the fire [ 8 , 9 , 10 , 11 , 12 ]. Since polymer materials are used for various applications, the incorporation of functional additives to polymer materials has attracted significant research attention [ 13 , 14 , 15 , 16 ]. In particular, the development of flame-retardant polymer materials has attracted attention toward managing the disadvantages of heat-sensitive polymers [ 17 , 18 , 19 , 20 , 21 , 22 , 23 ]. The poisonous gases released due to combustion, which is the secondary damage caused by fire, increase the harm done to humans; therefore, developing flame retardants and flame-resistant polymer materials is still crucial [ 24 , 25 , 26 ]. The typical characteristics of a fire include the following: i) Flame spread: The size of flame and/or the time it takes for the flame to cover a defined distance from the sample; ii) dripping: The presence of flame droplets that can ignite other objects; iii) heat release: Heat generated by the combustion of a sample in the room; and iv) the opacity and toxicity of the smoke, which are important for the evacuation of people trapped in the fire."
} | 2,345 |
28716030 | PMC5514506 | pmc | 5,026 | {
"abstract": "Background A unique combination of mechanical, physiochemical and biological forces influences granulation during processes of anaerobic digestion. Understanding this process requires a systems biology approach due to the need to consider not just single-cell metabolic processes, but also the multicellular organization and development of the granule. Results In this computational experiment, we address the role that physiochemical and biological processes play in granulation and provide a literature-validated working model of anaerobic granule de novo formation. The agent-based model developed in a cDynoMiCs simulation environment successfully demonstrated a de novo granulation in a glucose fed system, with the average specific methanogenic activity of 1.11 ml C \n H \n 4 /g biomass and formation of a 0.5 mm mature granule in 33 days. The simulated granules exhibit experimental observations of radial stratification: a central dead core surrounded by methanogens then encased in acidogens. Practical application of the granulation model was assessed on the anaerobic digestion of low-strength wastewater by measuring the changes in methane yield as experimental configuration parameters were systematically searched. Conclusions In the model, the emergence of multicellular organization of anaerobic granules from randomly mixed population of methanogens and acidogens was observed and validated. The model of anaerobic de novo granulation can be used to predict the morphology of the anaerobic granules in a alternative substrates of interest and to estimate methane potential of the resulting microbial consortia. The study demonstrates a successful integration of a systems biology approach to model multicellular systems with the engineering of an efficient anaerobic digestion system.",
"conclusion": "Conclusions A model of anaerobic granulation from digestion of glucose to methane has been successfully implemented in an agent-based simulator framework, cDynoMiCs . Simulation studies incorporated modeling of both reactor and single agglomerate scale granule development. Utilized growth mechanisms for generalized glucose-consuming/acetate-producing bacteria and acetate-consuming/methane-producing bacteria resulted in a well-correlated kinetic patterns of substrate conversions and biomass growth (Fig. 3 ). We were able to successfully qualitatively and quantitatively validate the architecture of the developed simulated anaerobic granule with the granule images and cell distribution from experimental literature studies (Figs. 4 and 5 ). The described granulation model has direct applications for designs of experiments, to predict yields of methane gas from substrates of interest. One application of the model was successfully demonstrated in this paper via parameter scan algorithm, searching through different acidogens:methanogens cell ratios and glucose feed that is needed to start anaerobic system to achieve a desired (maximum) methane yield. By changing the parameters of microbial growth to fit bacteria of a specific interest (the bacteria one is targeting to explore in an AD experiment), researchers can apply this model to predict efficiencies of anaerobic digestion in a system. The tested parameter scan is directly applicable to the studies with low-strength feed streams to UASB reactors, such as AD of brewery wastewater (COD =100-800 mg/L) [ 38 ], some municipal and industrial wastewaters (COD =100-400 mg/L) [ 39 , 40 ] and effluents from petroleum refineries (COD from 68 mg/L) [ 41 ]. Further development of the model will include a parameter search to investigate methane production from medium and high strength wastewaters. The current model of anaerobic granulation and methane production from simple feed sources (glucose) can be expanded to accommodate microbial conversion of more substrates, such as a mixture and proteins and carbohydrates. This expansion will make it possible to study granulation and methane potential from a more realistic scenario of wastewater feed, such as dairy and municipal wastewaters. A granulation model from a complex feed should result in a less stratified granule, due to the differential diffusions of the main feed components and a more complex patterns of microbial growth kinetics [ 18 ]. In addition, a model framework ( iDynoMiCs ) can be further modified to simulate detachment of excessive biomass from granular surface (simulating sheer stress described in the UASB reactor environment [ 4 , 42 – 44 ]) and breakage of a granule into daughter clusters, that subsequently give rise to mature granules with a more complex morphology [ 18 , 21 , 45 ]. Since current model assumes spherical types of cells, exploration of filamentous type of methanogenic bacteria influencing de novo granulation based on the “spaghetti theory” is something of future interest [ 32 , 46 ]. Another possible realm to expand development and application of current granulation model is to explore the mechanisms of enhancing anaerobic granulation, such as addition of positively charged ions and particles of polymers into the UASB system [ 47 , 48 ]. To converge granulation model with reactor-like environment, a Biocellion modelling environment can be used [ 49 , 50 ]. Possibility to parallelize computation load in Biocellion would eliminate the main bottleneck of the cDynoMics and allow development of a whole reactor model with simultaneous substrate conversion and anaerobic granule development. The current model of the de novo anaerobic granulation and its immediate applications will aid future discoveries in the field of anaerobic digestion, which is regaining its value and popularity in sustainable energy.",
"discussion": "Results and discussion Simulation experiments were conducted on the computational granulation model to give insights into different stages in the development of granules in aerobic sludge reactors. Where available, literature supported model parameters were employed. Other parameters, such as those that influence particle aggregation and mechanical sorting, were fine tuned based on correspondence between observations made from simulations and comparisons with reported granule images. The resulting granule spatial organization and product production of model simulations are analyzed and compared with values from real biological systems. Another objective of the study was to employ a search engine to find the amount of initial glucose concentration and populations of methanogens and acidogens that lead to optimal methane production. Study I: reactor scale model In the reactor scale phase of modeling, randomly distributed acidogens and methanogens (illustrated in Fig. 1 \n a ) interact with each other in a simulated UASB reactor environment, where upflow velocity and agitation play key roles to promote granulation of sludge. In the simulated environment microbial cells move around the system due to agitation and cells are bound together due to biomechanical adhesive forces, allowing formation of cell agglomerates (illustrated in Fig. 1 \n b ). Study IIa: stages of granule formation To investigate the development of a mature granule and dynamic changes in the cell growth, consumption of glucose, a series of simulator output snapshots were performed (Fig. 2 ). At the initial stage (t=0 h), single cell aggregate appear as a small cluster of acidogens and methanogens (zoomed from Reactor scale model, Fig. 1 ). As time proceeds (t =300, t =480 and t =700 h) cells grow and corresponding solute gradients demonstrate accumulation of acetate and methane in the system. Methane, being a volatile compound, is slowly diffused out of the system and depicted values on the scale of gradient images are not the cumulative values, as in the case of the glucose and acetate. At 480 h of granule development, a black “dead” core of cells start to emerge in the middle of the granule sphere. Appearance of a “dead” core is due to the diffusion boundaries of glucose or acetate inside granular cluster. Thus, cells of both types (acidogens and methanogens) are not getting enough energy supply and are forced to transition into the inert biomass. This transition is set to be irreversible in the model, thus leading to a formation of a “dead core”. A similar core can be seen on the Fig. 4 a of the laboratory-observed granule, which is used as evaluation criterion in current study and is descried later in detail. The final stage of granule development simulation (t =650 h) demonstrates a mature granule with 0.5 mm in diameter.\n Fig. 2 Simultion of 0.5 mm granule formation. Stages of simulated de novo granulation and associated dynamic changes in the solutes concentrations (glucose, acetate and methane). Only the critical time points of simulation are depicted through stages I-IV (t =0 h through t =650 h) \n Study IIb: analysis of granule growth dynamics In addition to visual (qualitative) investigation of de novo granulation, a close up quantitative study was performed on dynamic changes in solute amounts and cell biomass accumulation (both in values of cell numbers and cell biomass numbers). Graphs for dynamic changes are provided in Fig. 3 . Figure 3 a demonstrates changes in the total number of two types of cells (acidogens and methanogens) with regard to the simulation time. Simulation was initiated with 100 cells of each type. Due to the fast growth of the acidogens (see the Table 1 with growth kinetics parameters), we can see an exponential growth of acidogens from t =80 h to t =360. A similar dynamic is depicted in Fig. 3 b. Due to the product inhibition by the produced acetate and lack of diffused glucose, acidogens decrease their relative growth rate and reach the stationary phase of growth at around t =600 h. Dynamics of methanogens growth is slightly different, mainly due to the lack of available acetate from the start-up of the system and a lower growth rate, contrary to acidogens (Table 1 with model parameters). Methanogen growth goes through a long lag phase (t =0 h until t =220 h), where biomass is accumulated at a very slow rate (Fig. 3 b). At this lag phase methanogen cells are waiting for the supply of acetate from acidogens. As soon as enough acetate is accumulated in the system (around t =220 h), methanogens start exponential growth and decrease their relative growth rate at about t =520 h. This decrease is in direct correspondence with the amount of available acetate in the system at the same time period (t =480–500 h), (Fig. 3 c) when acidogens are inhibited by the produced acetate and are not provided with a high flow of glucose (due to the slow diffusion into the center of the granular biomass). Kinetics of acetate accumulation/conversion and methane production are in a good correlation with experimental data reported by Kalyzhnyy et al. and others [ 33 – 36 ].\n Fig. 3 Simulation related changes in solute concentrations and cell biomass. A close-up of the dynamic changes in the a cell number over simulation time, b cell biomass over simulation time and c solutes concentrations over simulation time. All the changes are graphed for each type of the cell (acidogens, methanogens, inert dead type) and each type of the solute (glucose, acetate, methane). Ten simulations with different random seeds were graphed to demonstrate standard deviation in the monitored values \n Fig. 4 Validation of the de novo granulation model via qualitative analysis. a Laboratory image courtesy of Sekiguchi et al. (1999), where green fluorescence label was used for Bacteria (represented by a single group of acidogens in current study), red fluorescence was emitted by Archaea (represented by a single group of methanogens in current study), yellow color correlates with overlapped red and green fluorescence and black color represents absence of fluorescence hybridization, and thus, absence of cell biomass (denoted as dead core here). b An image of granule simulated with current model. Same color labeling of the cell types is applied. c , d and e Distribution of the three solutes defining simulation of granulation (glucose, acetate, methane) at the final time point (t =800 h) of the simulation \n Table 1 Parameters used in model and their correspondent values Parameter summary Model parameter Symbol Value Unit References Solutes Diffusion of glucose in liquid \n D \n g \n 5.8×10 −6 \n \n m \n 2 /day [ 59 ] Diffusion of acetate in liquid \n D \n a \n 1.05×10 −4 \n \n m \n 2 /day [ 59 ] Diffusion of methane in liquid \n D \n m \n 1.29×10 −4 \n \n m \n 2 /day [ 60 ] Biofilm Diffusivity \n γ \n 30 % [ 42 ] Acidogens Cell mass \n B \n a \n 300 fg [ 61 ] Division radius 3 \n μ m [ 62 ] Maximum growth rate \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}$\\hat {\\mu _{a}}$\\end{document} μ a ^ \n 0.208 \n h \n −1 \n [ 61 ], [ 56 , 63 ] Substrate saturation constant Ks 0.26 g/L [ 35 , 56 ] Product inhibition constant Ki 0.1 g/L [ 56 , 63 ] Biomass conversion rate \n α \n bg \n 0.3 \n \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$\\frac {g_{biomass}}{g_{glucose}}$\\end{document} g biomass g glucose \n [ 56 , 57 ] Substrate conversion rate \n α \n ag \n 0.82 \n \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$\\frac {g_{acetate}}{g_{glucose}}$\\end{document} g acetate g glucose \n [ 56 , 63 ] Death delay 48 \n h \n Estimated Death threshold 0.02 g/L Estimated Methanogens Cell mass \n B \n m \n 1500 fg [ 62 ] Mass of EPS capsule 10 fg [ 54 ] Division radius 3 \n μ m [ 62 ] Maximum growth rate \n \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$\\hat {\\mu _{m}}$\\end{document} μ m ^ \n 0.1 \n h \n −1 \n [ 33 , 54 ] Substrate saturation constant Ks 0.005 g/L [ 54 ] Biomass conversion rate \n α \n ba \n 0.15 \n \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$ \\frac {g_{biomass}}{g_{acetate}}$\\end{document} g biomass g acetate \n [ 33 , 35 ] Substrate conversion rate \n α \n ma \n 0.26 \n \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$\\frac {g_{methane}}{g_{acetate}}$\\end{document} g methane g acetate \n [ 33 ] Death delay 48 \n h \n Estimated Death threshold 0.00001 g/L Estimated \n Study III: formation of a mature granule Figure 4 shows images of a 1 mm in diameter granule, obtained from both a laboratory experiment reported by Sekiguchi et al. [ 19 ] (Fig. 4 a) and an image from our simulated model (Fig. 4 b). Simulation of 1 mm in diameter granule formation took 800 h (around 33 days), which corresponds to the published studies observing granulation in UASB reactors [ 20 , 37 ]. Figure 4 c, d and e depict distribution of solutes (glucose, acetate, and methane) at the final stage of simulated granule growth (t =800 h). One can note a sharp decrease in the glucose diffusion inside the granule, with regard to the biofilm diffusivity capacity. Since acetate is consumed by methanogens during their growth and converted to methane, there is a low concentration gradient of both chemicals on the final images (Fig. 4 c, d, e). Overall, solute distributions for 1mm granule follow a similar pattern as for the 0.5 mm granule, described earlier. Key point in conducting simulation of a 1mm granule development is to demonstrate radial growth, without substantial changes in the overall morphology. Thus, initial stages of granule formation are the key factors for granulation per se . Validation of the model Validation of the model performance was conducted both qualitatively (Fig. 4 a, b) and quantitatively (Fig. 5 ). Visual comparison of a published fluorescent-labeled image of granule with simulated granule image demonstrates a striking similarity in spatial distribution of main trophic groups of microorganisms - acidogens, methanogens and “dead” biomass. Irregularities and hollow parts (black color) in the published granule image (Fig. 4 a) are possibly caused by the upflow velocity of the liquid and particulate matter in a UASB reactor, where the granule was developed [ 19 ], which might have damaged spherical shape of the immature granule, causing mature granule to change its shape and grow further with hollow compartments. Another possible explanation might be granule division. It is well documented [ 8 – 10 ] that due to the shear stress in a UASB reactor, granules cannot grow uncontrollably and will eventually split into “daughter” granules. Those “daughter” granules are susceptible to attachments of additional microbial cells, floating in UASB sludge bed. Those newly attached cells might cause irregularities in future mature granules in forms of randomly distributed cell clusters in a presumably inert (“dead”) core (red-labeled cell clusters on Fig. 4 a). To validate our simulated model quantitatively, we conducted image processing of the published data and used an algorithm to count the number of distinctly colored pixels/cells at the different distances from the center of the granule image (Fig. 5 ). We used 4 quarters of a spherical granule in the analysis to provide standard deviations of spatial distribution of three distinct cell groups – acidogens, methanogens and inert (“dead”) biomass. Results of quantitative distribution of three main cell types in both simulated and real images are in a good correlation, accept for the radial section “3”. Such slight discrepancy is due to the possible “division to daughter granules” history of the laboratory granule.\n Fig. 5 Validation of the de novo granulation model via quantitative analysis. Validation was done via analysis of the three cell type radial distribution in the both laboratory ( a ) and simulated granules ( b ). Both granules were divided into four quarters and each quarter was analyzed for cell distribution. Differences in the cell numbers at the same radial distance in four quarters are depicted in a form of standard deviation. Red , green and black colors of the bars on bar chart represent acidogen, methanogen and dead cells respectively \n Parameter scan for optimized methane production Main objective of the parameter scan is to estimate a combination of cell ratio (acidogens:methanogens) and glucose supply needed to start anaerobic system to achieve a desired (maximum) methane yield. The corresponding protocol parameter for glucose value is “SBulk” in world section. The “init area number” for acidogens and methanogens in the species section is used to determine the initial cell ratio for the simulations. The minimum and maximum value of the interval in which the search should be performed is given as an input to the search engine. The methane productivity (calculated from the solute concentration file output from simulator) is given as fitness function for the engine. The search engine simulated granule formation for several combinations of parameter values within the input interval and calculated total methane produced. The result is produced as a heatmap in Fig. 6 .\n Fig. 6 Parameter scan for the methane production in simulated granule. Parameter scan for the methane production in simulated granule with a varying initial number of methanogen cells (constant initial acidogen cell count) and b varying initial number of acidogen cells (constant initial methanogen cell count). Red color of the heatmap section has the highest value of methane produced (in milliliters of methane per gram of biomass), while blue heatmap section has the lowest value of produced methane. Parameter scan was conducted for 0.5 mm granule size and for the period of 650 simulation hours \n Figure 6 depicts amount of methane produced (in milliliters) per gram of biomass with varying amount of glucose supplied initially into the system (0.1 to 0.4 g/l). Figure 6 a has a constant initial acidogen count of 100 cells, and heatmap demonstrates varying amounts of methane produced with different glucose concentrations and different numbers of initial methanogen cells (from 1 to 900 cells). Same scheme is followed on Fig. 6 b, but with varying initial numbers of acidogens (from 1 to 400) and constant initial methanogen count of 100 cells. One can note from both Fig. 6 a and b that increased amount of glucose correlates with increased amount of methane produced in the system. Also, in general increased number of starting cells of acidogens (Fig. 6 b) let to the higher amounts of methane produced. This correlates with the earlier explored kinetics of methanogen/acidogen growth, when methanogens are waiting for acetate supply until they start to grow and produce methane. Parameter scan also helped to identify an important observation that a ratio of methanogen cells to acidogens should not be in a high favor of methanogens (100 acidogens and 900 methanogens on Fig. 6 a), since this leads to a decreased amount of methane production. The reason for such correlation is lack of acetate in the system to support growth of such a big number of methanogenic cells, which are forced to starve and die off."
} | 5,548 |
25725015 | PMC4398279 | pmc | 5,027 | {
"abstract": "We summarize different studies describing mechanisms through which bacteria in a biofilm mode of growth resist mechanical and chemical challenges. Acknowledging previous microscopic work describing voids and channels in biofilms that govern a biofilms response to such challenges, we advocate a more quantitative approach that builds on the relation between structure and composition of materials with their viscoelastic properties. Biofilms possess features of both viscoelastic solids and liquids, like skin or blood, and stress relaxation of biofilms has been found to be a corollary of their structure and composition, including the EPS matrix and bacterial interactions. Review of the literature on viscoelastic properties of biofilms in ancient and modern environments as well as of infectious biofilms reveals that the viscoelastic properties of a biofilm relate with antimicrobial penetration in a biofilm. In addition, also the removal of biofilm from surfaces appears governed by the viscoelasticity of a biofilm. Herewith, it is established that the viscoelasticity of biofilms, as a corollary of structure and composition, performs a role in their protection against mechanical and chemical challenges. Pathways are discussed to make biofilms more susceptible to antimicrobials by intervening with their viscoelasticity, as a quantifiable expression of their structure and composition.",
"conclusion": "CONCLUSIONS Bacterial biofilms provide added protection against environmental detachment forces, antimicrobials and the host immune system, increasing the chance of survival under mechanical and chemical attack. This review identifies the viscoelasticity of biofilms as a corollary of the structure and composition of biofilms. The viscoelasticity of biofilms therewith becomes a quantifiable property of biofilms, pivotal for the way biofilms deal with mechanical and chemical challenges. This identification opens new pathways to prevent or control bacterial biofilms by intervening with their viscoelastic properties that may be particularly relevant in the medical arena where over 60% of all human infections are due to biofilms.",
"introduction": "INTRODUCTION Biofilms appear in many environmental settings and industrial processes where they can appear either beneficial or detrimental (Bos, Van der Mei and Busscher 1999 ). Biofilms are beneficial, for instance, to degrade environmentally hazardous substances in soil, but detrimental on food and slaughterhouse equipment. In many biotechnological processes, it is attempted to maintain biofilms in order to stimulate efficient degradation of chemicals. In the medical arena, it is currently estimated that over 60% of all human infections treated by physicians are due to biofilms (Fux, Costerton and Stewart 2005 ), examples being oral biofilms (‘dental plaque’) and biofilms involved in a variety of pathological conditions like for instance osteomyelitis, chronic otitis media, the infected diabetic foot, chronic bacterial prostatitis or in biomaterial-associated infections (Costerton, Stewart and Greenberg 1999 ; Busscher et al. , 2012 ). Accordingly, research is focused mostly on how to prevent and control formation of an infectious, pathogenic biofilm, and on how to keep the commensal microflora of the skin, urinary and intestinal tract or oral cavity intact and free of potential pathogens (Reid et al. , 2011 ). In their biofilm mode of growth, bacteria adhering to a substratum surface and co-adhering with each other (Kolenbrander et al. , 2010 ) embed themselves in a matrix of extracellular polymeric substances (EPS). This matrix not only yields bacterial phenotypes that can be different from their planktonic counterparts, but also offers physical protection against mechanical and chemical challenges (Flemming and Wingender 2010 ). Biofilms and mechanical challenges Biofilm formation starts with the adhesion of bacteria to a substratum surface that can either be of biological or synthetic origin. This layer of initially adhering bacteria provides a link connecting other bacteria that either grow or adhere on top of it to the substratum surface (Bos, Van der Mei and Busscher 1999 ). Biofilms can be mechanically challenged during growth, for instance by water pressure in marine environments, industrial pipelines or membrane filtration, in the oral cavity during fluid flow arising from powered toothbrushing and tongue movement, from pulsatile blood flow in intravascular catheters or from the movement of tissues, fluid and biomaterial components in an orthopedic joint prosthesis. When mechanical challenges occur (Fig. 1A ) and detachment forces acting on a biofilm exceed the forces acting between different organisms in a biofilm, the biofilm is overloaded and failure occurs in the biofilm (‘cohesive failure’). Alternatively, when detachment forces operate exceeding the forces by which the initially adhering organisms connect with a substratum surface, the entire biofilm dislodges from the substratum surface (‘adhesive failure’) (Towler et al. , 2003 ). Often biofilms go through cycles of fluctuating mechanical challenges, and cohesive or adhesive failure along with growth occur accordingly. Figure 1. Key-properties of biofilms governing biofilm recalcitrance toward mechanical and chemical challenges. (A) Structure and composition govern biofilm resistance to environmental detachment and deformation forces. Resistance to deformation can be a time-dependent process yielding relaxation to an original shape over time. (B) Structure and composition govern the resistance of biofilms against chemical challenge in combination with altered phenotypes that are intrinsically more resistant to antimicrobials. (C) Composition and structure are jointly reflected in the viscoelasticity of a biofilm. Elasticity is generally presented as a spring with spring constant E, while the viscosity η (or its inverse, the fluidity φ) is shown as a dashpot. Springs and dashpots can be arranged in series (named ‘Maxwell’ element) or in parallel (called ‘Kelvin–Voigt’ element). Springs react immediately to an applied force, while dashpots dampen the speed of reaction. Usually, biological materials like biofilms cannot be represented by a single combination of springs and dashpots. Biofilms and chemical challenges The biofilm mode of growth protects individual bacteria from a variety of environmental challenges, including chemically diverse biocides (Mah and O'Toole 2001 ), host immune responses and antimicrobials (Vu et al. , 2009 ). Poor antimicrobial penetration is the major obstacle for treating biofilm infections with antibiotics and this has been known since Van Leeuwenhoek (Van Leeuwenhoek 1684 ) reported in the 17th century that ‘the vinegar with which I washt my teeth, kill'd only those animals which were on the outside of the scurf, but did not pass thro the whole substance of it’. Although it has become clear in the meantime that the protection offered by biofilms to its inhabitants against chemical challenges consists of multiple mechanisms working in tandem , the exact mechanism is still not fully understood (Ito et al. , 2009 ). In Fig. 1B , we identify three key properties of biofilms that govern the mechanisms through which biofilms become recalcitrant to antimicrobials (Thomas and Nakaishi 2006 ; Aslam 2008 ; Lazar and Chifiriuc 2010 ; Eastman et al. , 2011 ; Stewart 2012 ). Biofilm structure determines many transport processes within a biofilm and can lead to micro-environments with specific pH or nutrient availability (Koo, Falsetta and Klein 2013 ). Penetration of antimicrobials and nutrients into a biofilm depends on the degree of channelization of the biofilm and the presence of a suitable medium for molecular transport through the biofilm. Usually transport of antimicrobials and nutrients is limited, based on whether the composition of the EPS matrix and the bacterial cell surfaces adsorb these compounds. Importantly, biofilm structure is dynamic, adapting in both space and time to its environmental conditions (Donlan 2002 ), amongst which are pH, temperature, fluid shear, nutrient availability and host defenses. As a result of nutrient deprivation, bacterial phenotypes in a biofilm can alter, leading to formation of persister cells (Lopez, Vlamakis and Kolter 2010 ; Poole 2012 ) that can remain dormant without causing disease for prolonged periods of time. Moreover, many antimicrobial agents target macromolecule synthesis inside bacteria (Dodds, Grobe and Stewart 2000 ) during active metabolism. Thus, the presence of a slow metabolism contributes to antimicrobial recalcitrance of biofilms in general, and persister cells can tolerate higher concentrations of antimicrobials than nearby recalcitrant biofilm organisms (Spoering and Lewis 2001 ; Keren et al. , 2004 ). Consequently, whereas planktonic organisms have ample access to nutrients and, by the same token, are highly susceptible to antimicrobials, structural and compositional features of a biofilm form the major impediments for nutrient deprivation, the development of altered phenotypes and antimicrobial recalcitrance. Aim of this review In this review, we advocate that through a relation between structure and composition of biofilms with their viscoelasticity (Fig. 1C ), viscoelasticity of biofilms can be recognized as a reflection of their structure and composition, including the EPS matrix and bacterial interactions. Using stress relaxation analysis, macroscopic physical properties of a biofilm can be derived that facilitate explanation of the resistance of biofilms in ancient and modern environments as well as of infectious biofilms to mechanical and chemical challenges on a more quantitative basis than can be obtained by microscopic means. Finally, this review identifies new pathways for the treatment of infectious bacterial biofilms by interfering with their viscoelastic properties."
} | 2,488 |
23940510 | PMC3734138 | pmc | 5,028 | {
"abstract": "Although consumers can strongly influence community recovery from disturbance, few studies have explored the effects of consumer identity and density and how they may vary across abiotic gradients. On rocky shores in Maine, recent experiments suggest that recovery of plant- or animal- dominated community states is governed by rates of water movement and consumer pressure. To further elucidate the mechanisms of consumer control, we examined the species-specific and density-dependent effects of rocky shore consumers (crabs and snails) on community recovery under both high (mussel dominated) and low flow (plant dominated) conditions. By partitioning the direct impacts of predators (crabs) and grazers (snails) on community recovery across a flow gradient, we found that grazers, but not predators, are likely the primary agent of consumer control and that their impact is highly non-linear. Manipulating snail densities revealed that herbivorous and bull-dozing snails ( Littorina littorea ) alone can control recovery of high and low flow communities. After ∼1.5 years of recovery, snail density explained a significant amount of the variation in macroalgal coverage at low flow sites and also mussel recovery at high flow sites. These density-dependent grazer effects were were both non-linear and flow-dependent, with low abundance thresholds needed to suppress plant community recovery, and much higher levels needed to control mussel bed development. Our study suggests that consumer density and identity are key in regulating both plant and animal community recovery and that physical conditions can determine the functional forms of these consumer effects.",
"introduction": "Introduction Understanding factors that regulate the recovery and secondary succession of communities following disturbances is a core focus of ecology and conservation [1] – [7] . In general, the species composition of plant and animal communities is thought to be driven by the combined effects of biotic interactions, the physcial charateristics of habitats, disturbance events and propagule supply rates [8] – [13] . For most systems, however, we know little about how propagule establishment is interacively controlled by resident consumer dynamics (e.g. density-dependence), trophic structure and local phsycial factors and how these interactions in turn determine community composition (e.g., biodiversity, spatial dominance, or the emergence of alernate community states). The recruitment and establishment of plant and animal propagules in local communities can be under strong trophic control because consumers often create unoccupied space for new propagules to exploit (by consuming or disrupting competitors of the settlers) or by consuming or aggravating propagules after they have settled [14] – [20] . The strength of these top-down consumer effects is often a function of habitat type, consumer density and consumer species [21] – [23] . Although numerous studies have demonstrated that trophic structure can impact community development, we still have little appreciation for how the magnitude and direction of these consumer effects vary under different abiotic (i.e., temperature) charactersitics [21] – [23] . Indeed, field manipulations that addess the interactions between multiple biotic and abiotic factors are rare, in part, because of the complex and logistical challenge of such large experimental designs. Consequently, much of what we know about how the effects of comsumer density, identity and phsycial factors interact to impact plant communities has therefore been drawn from untested models [23] . In this study, we experimentally examined the combined effects of consumer assemblage and the physical factors that dictate propagule supply on recovery of macroalgae and invertebrates in a rocky intertidal community after disturbance. We found that consumer identity and density interact with abiotic processes (i.e. flow rate) to regulate recovery and that a keystone consumer can impose strong control over the composition and structure of communities that develop after disturbance.",
"discussion": "Discussion In many systems, stochastic settlement events are thought to be a dominant force regulating the assembly of plant and animal communities following a disturbance [25] , [30] – [32] . However, the integral role consumers can play in driving the outcome of community assembly is receiving increased attention [33] – [36] . It is likely that both processes are playing important roles in most systems, but their relative contributions have often been difficult to disentangle partly because the biotic and environmental drivers of community recovery following disturbance have been confounded, thereby obfuscating pre- and post-colonization processes. Most studies investigating consumer effects on community recovery have employed total consumer exclusions to isloate and quantify the net effects of consumers [25] , [26] , [37] – [41] . This method has been extremely effective in demonstrating the general importance of top-down consumer control to community organization [19] , [42] and recovery after disturbance [26] , [43] , [44] . This experimental approach, however, does not discriminate the relative importance of propagule input rates, the effects of individual consumer species, or the role of density-dependent processes. A better understanding of the role played by variation in recruitment and species-specific and density-dependent consumer effects is critical for identifying key species and mechansims that are regulating community recovery [19] , [45] and for predicting how natural- and anthropogenic-driven fluctuations in species' population densities will affect ecosystem structure and function [46] . Our results provide a unique demonstration that both consumer density and identity can be key regulators of whether plant or animal assembalges recover and dominate after a disturbance, and that both the shape and the magnitude of these density-depedent consumer effects are determined by abiotic conditions. Specifically, after ∼1.5 years, we found that: 1) plant and animal recovery from a disturbance in both low and high flow regimes on rocky intertidal shores in this tidal river are under strong consumer control, 2) grazing snails, more than predators, are the key bitoic agent imposing top-down control, and 3) that snail density and flow rate interact in non-linear ways to affect community composition. At low flow sites, mussels were essentially excluded (likely by low larval delivery and bulldozing by low densities of snails), while the potential for Fucus to dominant these low flow sites (i.e., near 100% Fucus cover in all consumer exclusion cages) decreased dramatically and non-linearly with increasing snail density ( Fig. 2 ). Only low to medium densities of snails were needed to generate the largest and disporprotionate suppression of Fucus establishement ( Fig. 2 and 3 ). At high flow sites, mussels displayed contrastingly higher recruitment and dominated the rock surfaces unless snails were at their highest densities. Only at these highest densities were snails effective at suppressing mussel receuitment, and thus at the high end of the naturally-occuring density spectrum strong top-down control of community (i.e. mussels in this case) recovery can emerge. In these same high flow areas, Fucus did not show up or was extremely rare, likely reflecting the fact the Fucus is a local disperser and adults are not in these areas [47] . Barnacles, in comparison, were able to establish at both low- and high-flow sites, but in contrast to the pattern observed for Fucus and mussels, barnacle abundance increased with snail density. This positive association with snail density ( Fig. 2a ) likely occurrs because snails bulldoze sediment and dislodge settling mussels and Fucus from the surface leaving the space open for settlement by competitively-inferior barnacles [29] , [44] . Another potential explanation for the positive assocaition of snails and barnacle cover is that snail suppression of Fucus removes algal inhibition of barnacle settllement that could occur through physical and/or chemical inhibition. Although snails can also negatively effect barnacle settlement, these inhibitory impacts appeared to have been overwhelmed by the positive effects of reducing sediment, mussels and algae. Consumer Identity- and Density-dependent Effects on Community Recovery In this study, we show that in this marine-river ecosystem and during the time of the study predators (crabs) played a secondary role compared to grazers (snails) in controlling the recovery of disturbed rocky intertidal habitat patches. The most pronounced effects of having crabs in additon to snails on patch recovery occurred at low flow sites where crabs and snails limited barnacle abundance more than snails alone ( Fig. 1 ), which is consistent with other studies showing that crabs can limit barnacle recruitment [48] . Green crabs do not commonly prey on adult barnacles, [48] , but routinely consume recently settled, lightly calcified barnacle recruits. Crabs also slightly reduced the recruitment of mussels, but this effect was small in comparison to the impacts of grazing snails on mussel recruitment ( Fig. 1 and 2 ). Although we know that crabs readily consume mussels [29] , [44] their foraging efficiency may be depressed at the high flow sites because flows can disrupt prey localization (via chemical cues) and green crabs mobility [48] . Experimental manipulation of periwinkle snail abundance demonstrated that in high densities snails alone can influence the composition of the community that assembles after disturbance in both low and high flow habitats types. At low flow sites, snail grazing even at low densities of snails (48–128 snails m −2 ) suppressed percent cover by fucoids, cleared the substrate of sediment, and facilitated barnacle success ( Fig. 2A ). Moreover, green macroalgae (i.e., Ulva and Entermorpha spp.) were only found in cages without snails [29] , [49] , [50] . At moderate and high densities (256–512 snails m −2 ), snails entirely prevented algal establishment at low flow sites, even though adult barnacles were present and are known to facilitate fucoid establishment by increasing refugia from grazing [17] . In addition to limited larval supply at low flow sites, snail grazing also limited mussel establishment, likely through bulldozing and/or the elimination of dense algal canopy, which is known to attract mussel recruits [29] , [49] , [50] . The interaction between flow and consumers, where higher flow environments dampen top-down effects, has been observed before in this [26] , [48] and other intertidal systems [13] , [43] , [51] , [52] . Our study expands this knowledge by showing that these interactions are density-dependent and that increased supply of mussel recruits in high flow habitats likely preempts the consumer suppression of community development observed at low-flow sites. In other words, high mussel recruitment at high flow sites swamps out the suppressing influence of top-down effects. We caution the extrapolation of our species-spefiic results to other similar rocky shore systems without additional experiments at those sites. Because predator diversity was low at our tidal river sites (primarily just Littorina and Carcinus ) compared to more open coast areas where drilling snails, more crab species, seas stars and urchins occcur (e.g., 26) and because we could have conducted this experiment during years when green crabs were at realtively lower densities (we did not measure crab abundance but inferred relative densities based on past studies at these sites which did measure green crab abundance [24] – [26] ), our results showing that snails are more important than crabs in controlling community development are likely to be spatially and temporally variable and depend on consumer diveristy, relative densities, and year of study. Implications for understanding alternative community states Ecologists have long argued whether natural communities of plants and animals are deterministic products of specific environmental conditions or stochastic products of chance recruitment events [53] , [54] . Recently, the debate over the deterministic nature of natural communities has shifted to discussions of whether assemblages of organisms can commonly occur as stochastically generated alternative stable community states [55] – [59] . These debates are not simply academic exercises, because understanding the relative importance of deterministic versus stochastic processes in community development has important implications for the conservation, management and the restoration of natural communities [60] . Our results concur with past studies [26] and show that secondary succession in low and high flow habitats on rocky shores in this Maine tidal estuary are likely the outcome of the combined effects of stochastic events (disturbances), environmental forcing (i.e. flow rate), and consumers. Our results reveal that consumer species and their densities set the context under which top-down control is expected and that the thresholds for these effects are regulated by the abiotic flow regime. Thus, understanding how the effects of species identity and density interact with environmental factors will likely be essential to make robust predictions regarding community recovery from natural- and anthropogenic-driven ecosystem disturbances and should be incorporated into future studies in this and other ecosystems [45] , [46] ."
} | 3,411 |
40013369 | PMC11923407 | pmc | 5,029 | {
"abstract": "Summary \n The increased positive impact of plant diversity on ecosystem functioning is often attributed to the accumulation of mutualists and dilution of antagonists in diverse plant communities. While increased plant diversity alters traits related to resource acquisition, it remains unclear whether it reduces defence allocation, whether this reduction differs between roots and leaves, or varies among species. To answer these questions, we assessed the effect of plant species richness, plant species identity and their interaction on the expression of 23 physical and chemical leaf and fine root defence traits of 16 plant species in a 19‐yr‐old biodiversity experiment. Only leaf mass per area, leaf and root dry matter content and root nitrogen, traits associated with both, resource acquisition and defence, responded consistently to species richness. However, species richness promoted a decoupling of these defences in leaves and fine roots, possibly in response to resource limitations in diverse communities. Species‐specific responses were rare and related to chemical defence and mutualist collaboration, likely responding to species‐specific antagonists' dilution and mutualists' accumulation. Overall, our study suggests that resource limitation in diverse communities might mediate the relationship between plant defence traits and antagonist dilution.",
"conclusion": "Conclusion Overall, our results emphasise the complexity of plant defence strategies and their interactions with antagonists across plant diversity gradients. They highlight that plant responses to resource limitation across SR gradients are probably the main drivers of changes in some defence traits, which may in turn influence antagonist pressure. Conversely, chemical defence traits appear to respond to changes in antagonist pressure and, together with traits related to cooperation with mutualists, show more species‐specific responses. This suggests a stronger bottom‐up effect of leaf and fine root physical defence traits on invertebrate herbivores and a top‐down effect of antagonists on leaf chemical defence traits. Finally, our results show that responses of defence traits to plant SR can differ significantly between above‐ and belowground compartments, highlighting the need to integrate both in future studies.",
"introduction": "Introduction Biodiversity is vital for the functioning of an ecosystem and its services to humanity (Cardinale et al ., 2012 ; Tilman et al ., 2014 ; Isbell et al ., 2017 ). Researchers have demonstrated a strong increase over time in the positive relationship between biodiversity and ecosystem functioning (BEF), both in forests (Guerrero‐Ramírez et al ., 2017 ; Huang et al ., 2018 ) and in grasslands (Cardinale et al ., 2007 ; Reich et al ., 2012 ; Ravenek et al ., 2014 ; Meyer et al ., 2016 ; Wagg et al ., 2022 ), suggesting greater importance of biodiversity than previously presumed from shorter‐term studies (Eisenhauer et al ., 2019 ). Over the past decade, we have accumulated evidence that multitrophic interactions, particularly plant–soil feedback, are important drivers of the strengthening biodiversity‐functioning relationships over time (Kulmatiski et al ., 2012 ; Eisenhauer et al ., 2012b ; van der Putten et al ., 2013 ; Thakur et al ., 2021 ). The two most likely mechanisms through which biotic interactions influence BEF relationships are the dilution of antagonists (Maron et al ., 2011 ; Schnitzer et al ., 2011 ; Wang et al ., 2023 ; Mahon et al ., 2024 ) and the accumulation of plant mutualists (van der Heijden et al ., 1998 ) with increasing diversity. However, it is still unknown whether changes in antagonists and mutualists impose differing selective pressures on plants along diversity gradients (Stamp, 2003 ) and allow plants to reduce resource allocation to defence to promote growth in more diverse communities. Using a trait‐based approach, we aim to test how plant defences change in response to the dilution of antagonists and the accumulation of mutualists along a long‐term plant diversity gradient in experimental grasslands. Despite the consensus that the negative effects of below‐ and aboveground antagonists decrease, while the positive effects of mutualists increase along diversity gradients, studies reveal inconsistencies (Halliday & Rohr, 2019 ). For example, while the dilution of antagonists appears to affect primarily specialists (Mommer et al ., 2018 ; Wang et al ., 2023 ), possibly due to host dilution, (resource concentration hypothesis; Root, 1973 ; Civitello et al ., 2015 ), generalists may benefit from increased plant diversity through expanded feeding opportunities (Keesing et al ., 2006 ). Nonetheless, several studies have consistently shown that antagonist pressure decreases regardless of specialisation. This was shown for soil antagonists, such as soil‐borne fungal pathogens (Mills & Bever, 1998 ; Yang et al ., 2015 ; Wang et al ., 2019 ), root‐feeding nematodes (Cortois et al ., 2017 ; Dietrich et al ., 2021 ) and arthropods (Amyntas et al ., 2025 ), as well as for aboveground antagonists, such as arthropods and pathogens (Mitchell et al ., 2002 ; Rottstock et al ., 2014 ; Muiruri et al ., 2019 ; Strauss et al ., 2024 ). This effect may be attributed to enhanced top‐down predator control on herbivores, which can impact both generalists and specialists (Barnes et al ., 2020 ; Amyntas et al ., 2025 ). Similarly, the positive effect of mutualists, such as mycorrhizal fungi, biocontrol bacteria and decomposers, seems independent of specialisation (van der Heijden et al ., 1998 ; Latz et al ., 2012 ; Eisenhauer et al ., 2012a ). Overall, these findings suggest that antagonistic pressure on plants decreases, while support from mutualists increases with plant diversity. Studies further suggest that this dilution of antagonists and accumulation of mutualists along plant diversity gradients may intensify over time (Eisenhauer et al ., 2019 ). However, recent research found limited support for such temporal changes for belowground antagonists (Amyntas et al ., 2025 ) and mutualists (Albracht et al ., 2024 ). This implies that initial shifts in soil communities in response to plant diversity may stabilise relatively quickly (Eisenhauer et al ., 2012b ), maintaining belowground antagonist pressure relatively constant over the years. On the contrary, Bröcher et al . ( 2024 ) found that the relationship between aboveground herbivores and plant diversity is highly variable over time, suggesting that aboveground antagonists may be more susceptible to seasonal and interannual biotic and abiotic fluctuations than belowground antagonists (De Deyn & Van der Putten, 2005 ). These fluctuations may lead to periodic die‐offs of aboveground antagonists, potentially hampering their dilution and thus selection pressure in diverse plant communities. Furthermore, aboveground antagonists generally exert milder pressure on plants than belowground antagonists (Johnson et al ., 2016b ). Thus, belowground antagonists may be a stronger and more constant driver of plant productivity than aboveground ones. Given that differing selection pressures along plant diversity gradients are known to promote changes in plant phenotypes and genotypes over time (Miehe‐Steier et al ., 2015 ; van Moorsel et al ., 2018 , 2019 ), changes in antagonists and mutualists along plant diversity gradients likely lead to changes in plant resource allocation to defence. Based on the assumption that there may be trade‐offs between growth and defence (Stamp, 2003 ; Lind et al ., 2013 ; Cappelli et al ., 2020 ; Zaret et al ., 2024 ) and that growth is closely related to plant fitness, a dilution of antagonists and an accumulation of mutualists in more diverse plant communities would reduce the need for investment into defence (Fig. 1a ). Considering that Bassi et al . ( 2024 ) identified a stronger influence of belowground mechanisms on long‐term monoculture plant performance, we also postulate that the effect of plant diversity on root defences may be stronger than on leaves (Fig. 1b ). Furthermore, while Mraja et al . ( 2011 ) observed a reduction in some leaf chemical defences along a plant diversity gradient for Plantago lanceolata L., it remains unclear whether this phenomenon can be generalised across multiple species (Fig. 1c ). Fig. 1 Graphical illustration of the hypothesised reduction in defence traits along the plant diversity gradient (a), a stronger response of fine root compared with leaf defence traits (b) and species‐specific responses of three hypothetical species (1, 2 and 3) (c). Plant defence traits, that is functional traits that promote plant fitness in the presence of antagonists relative to the absence of antagonists (Didiano et al ., 2014 ), are often divided into physical and chemical defences (Table 1 ; Moore & Johnson, 2017 ). Physical defences are those that deter herbivores from feeding on plant tissues through morphological or anatomical traits (Hanley et al ., 2007 ), while chemical defences encompass traits related to the tissue's nutritional quality (Mattson, 1980 ; Poorter et al ., 2004 ) and defensive phytochemicals (Raguso et al ., 2015 ). Another strategy to counteract antagonists is collaboration with mutualists. Along the recently defined ‘root economics space’ (Bergmann et al ., 2020 ), this is captured by the ‘collaboration gradient’, represented by a trade‐off between specific root length (SRL) and root diameter (RD), which is related to the root colonisation by mycorrhizal fungi. Despite mycorrhizal fungi enhancing plant physical defence by limiting antagonist access to roots through competition for space and resources (Rasmann et al ., 2011 ), they may also promote plant chemical defences by inducing the production of defensive metabolites (Jung et al ., 2012 ; Cameron et al ., 2013 ; Frew et al ., 2022 ). Table 1 Leaf and fine root defence traits examined in this study. Tissue Traits Defence correlation Defence type Defence mechanisms References Leaf Water repellency + Physical defences Surface barrier: reduced attachment and mobility of antagonists Hanley et al . ( 2007 ), Gorb & Gorb ( 2017 ) Hair density + Toughness + Mechanical strength Poorter et al . ( 2004 ), Hanley et al . ( 2007 ), Johnson et al . ( 2010 ), Loranger et al . ( 2012 ), Schuldt et al . ( 2012 ), Caldwell et al . ( 2016 ), Moore & Johnson ( 2017 ), Bröcher et al . ( 2023 ) LMA + Leaf/fine root DMC + SRL − Fine root Diameter + Protection through AMF \n † \n (root collaboration gradient) Rasmann et al . ( 2011 ), Jung et al . ( 2012 ), Cameron et al . ( 2013 ), Cortois et al . ( 2016 ), Johnson et al . ( 2016a ), Bergmann et al . ( 2020 ), Frew et al . ( 2022 ) MC + Leaf / fine root Nitrogen − Chemical defences Palatability: the nutritional quality of the tissue Mattson ( 1980 ), Poorter et al . ( 2004 ) Alkaloids + Toxicity; Inducing or priming plant resistance or tolerance against plant antagonists (pathogens and insects) Hol et al . ( 2004 ), Steppuhn et al . ( 2004 ), Alves et al . ( 2007 ), Nuringtyas et al . ( 2014 ), Dugé de Bernonville et al . ( 2017 ) Terpenoids + Marak et al . ( 2002 ), Nakashita et al . ( 2003 ), Zhang et al . ( 2015 ), Zahid et al . ( 2017 ), Lackus et al . ( 2018 ), Murata et al . ( 2019 ) Shikimates + Alonso et al . ( 2009 ), Koskimäki et al . ( 2009 ), Hölscher et al . ( 2014 ), Prasannalaxmi & Rani ( 2016 ), Olivier et al . ( 2018 ), Lea et al . ( 2021 ), Grover et al . ( 2022 ), Zhang et al . ( 2022 ), Yang et al . ( 2023 ) The table reports the tissues (leaf or fine root), the traits and the defence correlation (i.e. the expected positive or negative correlation between trait and defence). As we hypothesise defence to be negatively correlated with diversity, the hypothesised trait–diversity relationships are exactly reversed. The table further reports the classification in defence type (physical or chemical), the corresponding defence mechanisms and relevant references. For ‘Alkaloids’, ‘Terpenoids’, and ‘Shikimates and phenylpropanoids’, we refer to the chemical compounds within those chemical pathways (detailed in Table S2 ). It is important to note that this table is based on studies using targeted metabolome quantification of one group of compounds within those pathways, while our study employs a broader untargeted metabolomic approach to assess phytochemical richness or the sum of range‐scaled feature intensity within these pathways. DMC, dry matter content; LMA, leaf mass per area; MC, mycorrhizal colonisation; Shikimates, Shikimates and phenylpropanoid; SRL, specific root length. Part of this table is derived from Bassi et al . ( 2024 ). \n † \n While we have classified mycorrhizal colonisation as a physical defence, we acknowledge that mycorrhizas may also promote plant chemical defences through priming. Plant defence traits provide a framework that enables us to investigate how plant resource allocation to defence changes along plant diversity gradients. Similarly, functional traits, related to collaboration with mutualists, enable us to investigate plant investment in mutualistic interactions. For instance, some leaf and root traits related to physical and chemical defences, such as leaf mass per area (LMA), leaf toughness, nitrogen and phenolics content, and metabolome, as well as traits related to collaboration with mutualists, such as SRL, have been shown to change along plant diversity gradients (Scherling et al ., 2010 ; Mraja et al ., 2011 ; Ristok et al ., 2019 ; Peng & Chen, 2021 ; Weinhold et al ., 2022 ; Felix et al ., 2023 ). Although several of these traits are also related to resource acquisition (Weigelt et al ., 2021 ), their changes often affect plant–antagonist and plant–mutualist interactions (Muiruri et al ., 2019 ; de Vries et al ., 2021 ; Bröcher et al ., 2023 ; Ristok et al ., 2023b ). However, plant species frequently deploy different types of defences: While grasses are often defended through physical defences, forbs are more commonly defended through chemical defences (Eichenberg et al ., 2015 ; Bassi et al ., 2024 ). Furthermore, plant–antagonist interactions are also affected by associational effects, where the herbivory rate experienced by a focal plant species depends on the characteristics, such as defence traits, of the surrounding plant community and herbivores' feeding preferences, leading to either associational susceptibility or resistance (Barbosa et al ., 2009 ; Underwood et al ., 2014 ). For example, the ‘neighbour contrast susceptibility and defence’ hypothesis (Alm Bergvall et al ., 2006 ) suggests that if a focal plant species is better defended than the surrounding plant community, it will experience lower herbivory compared with when it grows among plants that are better defended than itself. Indeed, changes in herbivory rates along plant diversity gradients have been shown to differ between plant species in magnitude and direction (Bröcher et al ., 2023 ). It is thus reasonable to hypothesise that changes in defences along plant diversity gradients are species‐specific (Fig. 1c ). Indeed, despite consistent trait–diversity relationships among plant species (Roscher et al ., 2018 ), species‐specific responses have also been observed (Gubsch et al ., 2011 ; Roscher et al ., 2011b ; Lipowsky et al ., 2015 ). In this study, we measured for the first time a comprehensive set of physical and chemical defence traits in leaves and fine roots (summarised in Table 1 ) of 16 grassland plant species growing along a 19‐yr‐old plant species richness gradient in the Jena Experiment (Roscher et al ., 2004 ). We hypothesised that the observed dilution of antagonists and accumulation of mutualists along plant species richness gradients would trigger:\n a reduction in plant defence traits (Fig. 1a ); a more pronounced defence reduction in fine roots compared with leaves (Fig. 1b ); and species‐specific defence responses (Fig. 1c ).",
"discussion": "Discussion This study investigated intraspecific responses of 23 leaf and fine root physical and chemical defence traits to a 19‐yr‐old plant diversity gradient of sixteen grassland plant species. Our main goal was to test whether the accumulation of mutualists and dilution of antagonists, often observed along plant diversity gradients, would promote a reduction in plant defences in high‐diversity communities. In addition, we tested whether this reduction would be stronger in fine root compared with leaf defences, and if it would differ among species. Our results showed that most plant defence traits do not respond to species richness. Furthermore, while defence traits, which are involved in other plant functions, such as resource uptake and usage, responded to species richness similarly among species, chemical defence traits, such as the production of terpenoids, or traits related to the collaboration with mutualists, showed species‐specific responses. Interestingly, some leaf and root defences responded in opposing directions to species richness, suggesting that while changes in resource availability along species richness gradients have a consistent effect among species, they promote the decoupling of defence traits between leaves and fine roots. On the other hand, the species‐specific responses of chemical defence traits and traits related to collaboration with mutualists suggest that the commonly observed dilution of antagonists and accumulation of mutualists across SR gradients may not affect all species equally. Effect of plant species richness on leaf defence traits The leaf defence traits with the most consistent response to the diversity gradient among species were LMA and LDMC. Similar to other studies, these two traits decreased along the diversity gradient (Gubsch et al ., 2011 ; Roscher et al ., 2011b ; Lipowsky et al ., 2015 ) and are associated with plant defence, against leaf chewers (Caldwell et al ., 2016 ), as well as foliar pathogens (Cappelli et al ., 2020 ). These traits' responses seem to align with our first hypothesis, that plant defence traits decrease along the plant diversity gradient, due to a reduction in antagonist pressures. However, despite being significant and marginally significant only without correcting for FDR, the production of features within the terpenoids and shikimate and phenylpropanoids pathways in leaves tended to increase with SR in line with the findings of Poeydebat et al . ( 2021 ). Several phytochemicals within those pathways are known for their defensive role against a variety of leaf antagonists, including arthropods (Dugé de Bernonville et al ., 2017 ), pathogens (Lackus et al ., 2018 ) and viruses (Zhang et al ., 2015 ). This response contradicts our first hypotheses, as it suggests that leaf chemical defences increase with plant diversity. The opposite responses of these leaf physical and chemical defences to the diversity gradient suggest a trade‐off between physical and chemical defences within species. Despite this general trend, this trade‐off was consistent within all species only for LDMC and leaf terpenoids (Fig. S3 ). However, the PCA revealed a comparable trade‐off between physical and chemical defences across species in leaf and, to a lower extent, in roots, aligning with findings from other studies (Fernandez‐Conradi et al ., 2022 ; Bassi et al ., 2024 ). Overall, the finding that physical and chemical defences exhibit a trade‐off within and across species aligns with the growth‐defence trade‐off hypothesis (Lind et al ., 2013 ; Zaret et al ., 2024 ) as it suggests that plants can simultaneously optimise resource competition and promote defence, as shown in previous studies investigating trade‐offs between constitutive and induced defences (Kempel et al ., 2011 ). In our experiment, this trade‐off may arise due to the differential strengths and effects that multiple stressors, such as resource limitation and antagonist pressure, exert on leaf defence traits. Leaf mass per area and dry matter content have a pivotal role in other plant functions than defence, such as the acquisition of light (Poorter et al ., 2019 ). Thus, the response we observed is most likely related to the changes in light availability, as previously detected in our experiment (Roscher et al ., 2011a ; Bachmann et al ., 2018 ), a major limiting resource along species richness gradients (Hautier et al ., 2009 ), rather than to changes in antagonistic pressure. However, the reduction in LMA and dry matter content, to optimise light‐capturing surface per carbon cost (Poorter et al ., 2009 ), inevitably reduces the defensive benefits conferred by these traits. This poses the leaves of plants growing in highly diverse mixtures under a higher risk of antagonist attack. Indeed, even though arthropod herbivore pressure, measured on a biomass or energy flux basis (herbivore biomass or herbivore energy influx divided by plant biomass) was shown to decrease (Ebeling et al ., 2018 ; Barnes et al ., 2020 ), the proportion of leaf area and leaf biomass eaten by invertebrate herbivores, including gastropods, a group containing many generalists, increases with plant species richness (Meyer et al ., 2017 ; but see Seabloom et al ., 2017 ). Indeed, at the same site, Bröcher et al . ( 2023 ) found that the increase or decrease in leaf area eaten by invertebrate herbivores along the species richness gradient was proportional to the change in LDMC, in line with the ‘neighbour contrast susceptibility and defence' hypothesis (Alm Bergvall et al ., 2006 ). Given that under light limitation in diverse communities, the relative importance of leaves should increase, and at the same time, the probability of leaf attack increases due to a reduction in physical defence, according to optimal defence theory, plants should promote allocation to defence in leaves (Stamp, 2003 ). Increasing leaf chemical defences, such as the production of features within the terpenoids and shikimate and phenylpropanoids pathways, could be a cost‐effective way to counterbalance the loss of physical defence without hampering the light‐capturing capacity. Effect of plant species richness on root defence traits Contrary to our first hypothesis, our study indicated that fine roots in high‐diverse mixtures exhibited greater defence than those in low‐diverse mixtures. While RDMC increased, root nitrogen decreased along the plant diversity gradient, suggesting that fine root defences, particularly against root chewers and root‐feeding nematodes (Johnson et al ., 2010 ; Moore & Johnson, 2017 ; Bassi et al ., 2024 ), increase along the diversity gradient. However, similar to LMA and LDMC, root nitrogen content and RDMC are related to functions other than plant defences, such as nutrient and water uptake. The reduction in root nitrogen content along the species richness gradient is consistent with other studies (Mulder et al ., 2002 ; van Ruijven & Berendse, 2005 ) and is probably due to the lower availability of nitrogen, which was previously found in our experiment (Roscher et al ., 2008 ), as well as increased productivity and thus nitrogen demand in more diverse communities. Although nitrogen access in diverse mixtures may be enhanced through facilitation effects and resource use complementarity (Bessler et al ., 2012 ), our results rather suggest that nitrogen limitation becomes more pronounced in diverse communities. Notably, the experimental plots have not been fertilised since the experiment began, and consistently, higher plant biomass removal from high‐diversity plots (Wagg et al ., 2022 ) may have depleted soil nitrogen. Interestingly, the reduction in leaf nitrogen was smaller and not significant, suggesting that the decreased nitrogen content in roots may partly result from reallocation to leaves (Wang et al ., 2024 ) to maintain sufficient nitrogen for photosynthesis. The increased RDMC along the diversity gradient may be attributed to reduced nutrient or water availability (Fischer et al ., 2019 ). The simultaneous increase in RDMC and decrease in LDMC may suggest a reallocation of water from the roots to the aboveground part of the plant to support higher biomass and construction of shade leaves in high‐diverse mixtures. A similar opposite response of root and LDMC to light, nutrient and water availability gradients was observed by Freschet et al . ( 2013 ). However, contrary to root nitrogen content, the observed increase in RDMC is less consistent among species and inconsistent with other studies (Gould et al ., 2016 ; Chen et al ., 2017 ). Hennecke et al . ( 2025 ) found no changes in the community‐level root tissue density, a trait strongly related to RDMC (Birouste et al ., 2014 ) in the same experiment and at the same time as this study. This suggests that the increase in root dry matter that we observed in our species pool, does not affect the majority or the most dominant species. Nonetheless, the reduction in RDMC of some species is consistent with the findings of Roeder et al . ( 2019 , 2021 ), who found that species with taproots, tend to become older and simultaneously reduce growth rates along the species richness gradient. Thus, in high‐diverse mixtures, some species exhibit a more conservative growth strategy belowground (Bergmann et al ., 2020 ), whether this is related to changes in water and nutrient availability or plant age has yet to be determined. Decoupled response of leaf and fine root physical defence traits Contrary to our second hypothesis that fine root defence traits would decrease more than leaf defence traits along diversity gradients, our study revealed that while leaf chemical defences showed a tendency to increase, the few leaf and fine root defence traits that responded to the diversity gradient did so in the opposing direction. These opposite trends between leaf and root defences were driven by traits associated with resource uptake. This suggests a complex relationship between plant defence traits and antagonist pressure, likely mediated by resource availability. While it might be surprising that plants can simultaneously increase root defence and reduce leaf defence, our results contribute to the ongoing debate about whether leaf and root traits are coordinated (Carmona et al ., 2021 ; Weigelt et al ., 2021 , 2023 ; Bueno et al ., 2023 ), showing that, at least within species, this may not always be the case. In addition, although other studies have shown that plant diversity has a direct effect on aboveground and belowground antagonist pressures (Ristok et al ., 2023a , 2023b ), our study suggests that the observed changes in aboveground and belowground antagonist pressure with species richness may be mostly driven by changes in plant defence traits and not vice versa, as we hypothesised. Combined with the lack of response in several defence traits, our results indicate that the dilution or accumulation of antagonists and mutualists may not occur uniformly (Halliday & Rohr, 2019 ). Instead, these dynamics may be more complex, differing between aboveground and belowground, as well as among antagonist groups. Competition for resources in diverse plant communities seems to promote fine root defences. This increase in root defences may explain the reduction in belowground antagonistic pressure from root‐feeding nematodes and arthropods observed in the same experiment (Cortois et al ., 2017 ; Dietrich et al ., 2021 ; Amyntas et al ., 2025 ). However, this increase in defences might also be a response to increased pressure from other groups of antagonists, which have not yet been documented. Similarly, competition for light reduces leaf physical defences while promoting their chemical defences. This could suggest that under low light availability, plants maintain high chemical defences that, in turn, promote arthropod herbivore and pathogen dilution along the diversity gradient. However, it also raises the possibility that leaf chemical defences may increase due to the accumulation of other antagonist groups, aside from arthropods and leaf pathogens (Ebeling et al ., 2018 ; Barnes et al ., 2020 ). For example, the observed increase in leaf herbivory rates (Meyer et al ., 2017 ; Bröcher et al ., 2023 ) could be attributed to damage inflicted by generalist antagonists, such as gastropods, which were not accounted for in other measurements. This may suggest that certain groups of antagonists, particularly generalists, may accumulate rather than dilute along diversity gradients (Keesing et al ., 2006 ), thereby explaining the increase in certain defence traits we observed. Further studies are needed to disentangle the complex relationship between plant defence traits, antagonists and mutualists accumulation or dilution. Species‐specific response of leaf and fine root defence traits Plant species identity was the main driver of all defence traits assessed in this study, explaining a substantial proportion of the observed variation. However, species‐specific responses to diversity gradients were less common than anticipated under our third hypothesis. While plant defence traits linked to resource acquisition showed consistent responses across species, species‐specific responses were more frequent in traits with tighter links to defence or collaboration with mutualists. Although only a few traits exhibited this response, root chemical defences showed stronger species‐specific patterns than leaf chemical traits, aligning with the findings of Weinhold et al . ( 2022 ). Given the known link between mycorrhiza and the metabolome through the priming of defensive phytochemicals (Frew et al ., 2022 ), as well as the association between MC and root chemical defences observed across species in the PCA, it is surprising that our results did not reveal any consistent trends between MC and chemical defence responses to the diversity gradient. Only T. flavescens showed a consistent response, with increases in both MC and the number of leaf terpenoids, providing some evidence for priming. Overall, this species‐specific response may indicate that the dilution of below‐ and aboveground antagonists and the accumulation of belowground mutualists across plant diversity gradients are also species‐specific. Indeed, at the same site, Bröcher et al . ( 2023 ) found that herbivory rate changes along diversity gradients were species‐specific. These species‐specific responses can be explained by the associational effects, where plant community composition can promote resistance or susceptibility (Barbosa et al ., 2009 ; Underwood et al ., 2014 ) of focal species to herbivores, depending on whether the focal species is more or less attractive or defended than neighbouring species. While these mechanisms are relatively well understood for aboveground herbivores, it remains unclear to what extent belowground antagonists and mutualists follow the same mechanism. Further studies addressing these issues from a belowground perspective are needed. Conclusion Overall, our results emphasise the complexity of plant defence strategies and their interactions with antagonists across plant diversity gradients. They highlight that plant responses to resource limitation across SR gradients are probably the main drivers of changes in some defence traits, which may in turn influence antagonist pressure. Conversely, chemical defence traits appear to respond to changes in antagonist pressure and, together with traits related to cooperation with mutualists, show more species‐specific responses. This suggests a stronger bottom‐up effect of leaf and fine root physical defence traits on invertebrate herbivores and a top‐down effect of antagonists on leaf chemical defence traits. Finally, our results show that responses of defence traits to plant SR can differ significantly between above‐ and belowground compartments, highlighting the need to integrate both in future studies."
} | 8,055 |
35603544 | PMC9786923 | pmc | 5,030 | {
"abstract": "Abstract Growth‐coupling product formation can facilitate strain stability by aligning industrial objectives with biological fitness. Organic acids make up many building block chemicals that can be produced from sugars obtainable from renewable biomass. Issatchenkia orientalis is a yeast strain tolerant to acidic conditions and is thus a promising host for industrial production of organic acids. Here, we use constraint‐based methods to assess the potential of computationally designing growth‐coupled production strains for I. orientalis that produce 22 different organic acids under aerobic or microaerobic conditions. We explore native and engineered pathways using glucose or xylose as the carbon substrates as proxy constituents of hydrolyzed biomass. We identified growth‐coupled production strategies for 37 of the substrate‐product pairs, with 15 pairs achieving production for any growth rate. We systematically assess the strain design solutions and categorize the underlying principles involved.",
"conclusion": "4 CONCLUSION In this work, we detailed the potential of I. orientalis as an organic acid producer. We found that for most of the selected products OptKnock suggested designs involved high yields. As we noted, currently I. orientalis metabolizes xylose slowly after a long delay, and this ability to consume xylose is consistent with the presence in its genome of all genes needed for d ‐xylose conversion to d ‐xylose‐5‐P via the oxido‐reductase pathway of xylose reductase, xylitol dehydrogenase and xylulokinase, which can then funnel into the pentose phosphate pathway. \n 31 \n However, metabolic engineering methods could be applied to improve xylose as an effective substrate. Genetic interventions or serial cultivation could be used to modify directly or indirectly the existing genes needed for the oxido‐reductase pathway along with their regulation. We note that metabolic engineering methods have permitted the yeast Saccharomyces cerevisiae to use xylose via the oxido‐reductase pathway by expressing genes from Scheffersomyces stipites , \n 60 \n which could similarly be implemented in I. orientalis . Alternatively, an isomerase pathway from another organism could be introduced that directly converts xylose into xylulose followed by the xylulokinase of the oxido‐reductase pathway. The oxygen uptake analysis underscores one facet of operational conditions and the impact it can have on production yields. Our results suggest that 3‐hydroxypropionic acid production is well‐suited to require little input of oxygen beyond that required to grow the necessary cell mass in a bioreactor. Similar outlooks were found for succinate, fumarate and malate in the presence of additional functionality added to the cell such as fumarate reductase or pyruvate carboxylase activity. We anticipate future examination to highlight other adjunct heterologous pathways for other products that improve oxygen requirements as well as other pathway augmentations that can increase the in vivo yield of organic acids closer to the carbon and/or available electron maximum theoretical yields. We were able to find growth‐coupled production designs or potentially growth‐coupled designs for almost every product (i.e., 37 of the 43 substrate–product pairs). We found some form of unique GCP designs (i.e., wGCP or sGCP), for 29 substrate–product pairs, of which 9 had at least one sGCP design with yields near the theoretical biological maximum. Work remains, however, to move those with wGCP designs into dGCP or SUCP designs and to identify even pGCP design solutions for some pairs, including 2‐oxobutanoate, 3‐hydroxybutanoate, and 3‐hydroxypentanoate. As seen in Figure 2 , the latter two derive wholly from acetyl‐CoA. The inability to couple an acetyl‐CoA drain to growth was observed in E. coli network analysis, \n 21 \n although there the problem was described as being due to the model's inability to compensate for the CoA drain, which similarly impacted succinyl‐CoA. We were able to find one pGCP design for citramalate, which also derives in part from acetyl‐CoA. In the absence of available wGCP or sGCP designs, it is advantageous to identify, as we have done here through the use of OptKnock, if there are strain designs or conditions under which the product could nevertheless be noncompeting with biomass production (i.e., pGCP) instead of directly antagonistic to biomass production (i.e., ∅GCP). Additional effort is required to determine effective production routes for those products which only had pGCP designs. One alternative approach to growth coupling is to use a method such as OptForce \n 61 \n in combination with labeled substrates in order to determine optimal flux values for high yields and then engineer the system to have such fluxes, by up or down regulating specific fluxes. Such an approach is especially attractive for products without sGCP designs. We provide the caveat that our results do not necessarily preclude higher order RE strategies from forming wGCP or sGCP designs. We furthermore note that additional analyses are required to map RE into implemented gene knockouts and check feasibility. Such analyses include considerations such as GPR associations and identification of major isozymes. \n 62 \n Here, we were able to elucidate effective GCP gene implementations for nearly all the top scoring RE designs for substrate‐product pairs having sGCP designs. We point out that the flux coupling reduction process we used identifies equivalent reaction eliminations and has the potential to avoid reactions that would otherwise be prohibited because of synthetic lethals or other gene implementation difficulties. The somewhat low Jaccard index between solutions for different substrates for the same product suggests that deriving a single strain capable of effectively metabolizing both substrates into product could be challenging, especially if much of the metabolic engineering work is primarily tested and focused at the onset on one of the sugars. Early examinations of strain design with an eye for one that targets both could behoove the process. Another approach could be to design and use mixed cultures, with one strain for each substrate. Organic acid production at low pH can have energetically expensive energy requirements for product export. \n 63 \n Typically GSM models, including the one used in the current work, have been reconstructed with charges on metabolites determined at a single neutral pH value across the model and with constant ATP maintenance values, \n 64 \n although some models have included physicochemical specification differences between pH 5 and 8. \n 65 \n , \n 66 \n We anticipate that accounting for physicochemical specifications of different compartments, allowing pH‐dependence of ATP maintenance, and incorporating regulatory constraints could further enhance predicting the production potential of I. orientalis . The process and methods we apply herein can be readily applied to other product categories.",
"introduction": "1 INTRODUCTION Metabolic engineering approaches modify cellular activities in order to improve the production of metabolite or protein products. \n 1 \n The rise of genomic sequencing tools has enabled the rapid reconstruction of genome‐scale metabolic models for a number of organisms. \n 2 \n These models can be used to inform intervention strategies through the use of constraint‐based reconstruction and analysis (COBRA) approaches. \n 3 \n Current advances in constraint and machine learning‐based metabolic modeling have been recently reviewed. \n 4 \n The use of such approaches has aided the successful commercialization of processes to produce 1,4‐butanediol, \n 5 \n lactic acid, \n 6 \n , \n 7 \n and heterologous proteins for therapeutic use as biopharmaceuticals, \n 8 \n among others. One possible approach for strain design is to couple the production of a desired product to the growth of a microbe. \n 9 \n This alignment of industrial objectives with biological fitness ones can improve pathway stability by reducing the pressure on selection to divert carbon away from the product toward biomass. Such a designed strain can then be acted upon adaptive evolutionary strategies \n 10 \n to indirectly select for improved product yield through growth selections. \n 11 \n The foundational computational tool for rational strain design that generates growth‐coupled production (GCP) strategies is OptKnock. \n 12 \n OptKnock solves a bilevel optimization problem to pinpoint a set of reactions that should be eliminated simultaneously from a metabolic network in order to ensure that the desired product becomes a potential byproduct of biomass formation. Each reaction elimination (RE) design solution can be examined by making the indicated changes on bounds to the reactions and then plotting a metabolic production envelope \n 13 \n that projects the accessible flux space onto the plane of growth rate and the target's production rate. Through the use of integer cuts, alternative solutions can be computed. These alternative solutions can be examined for implementation via the gene‐protein‐reaction (GPR) associations that are part of genome‐scale metabolic models, as well as subjected to additional analyses and rankings, by using Flux Balance Analysis (FBA). \n 14 \n Extensions and improvements to OptKnock has been a fertile area of algorithmic research and has been reviewed. \n 15 \n These tools include examples such as RobustKnock, \n 16 \n FOCAL, \n 17 \n ReacKnock, \n 18 \n and SMET. \n 19 \n \n When examining generated RE design solution sets, a number of classifications and production phenotype metrics can be helpful to describe each solution that produces product p and biomass X from substrate s , as shown by production envelopes in Figure 1 . For simplicity, in this section, we make substrate notations implied. Qualitatively, a solution can be partitioned broadly into various cases based on the production phenotype. \n 12 \n , \n 20 \n , \n 21 \n , \n 22 \n Here, we adopt an existing classification scheme \n 22 \n with some minor extensions and distinguish five cases: null, potentially, weakly, and directionally growth‐coupled production (GCP) and substrate‐uptake‐coupled production; we further use a collective category termed strongly growth‐coupled production comprising directionally growth‐coupled production and substrate‐uptake‐coupled production. The first, null (referred to hereafter as ∅GCP), indicates no growth coupling and occurs when at the maximum biomass production rate, g max ≔ v X , max , the maximum product rate, v p , max g max , is zero. Typically, the starting network will be ∅GCP. For the second, potentially growth‐coupled production (pGCP), the product rate, v p , displays a phenotype consisting of equivalent optimal solutions that does not ensure production of the desired compound. That is, at the maximum biomass production, g max , the maximum product rate v p , max g max is positive whereas the minimal product rate, v p , min g max , is zero. For weakly and strongly growth‐coupled production, conversely, instead the solution displays a phenotype with v p , min g max > 0 . For weakly growth‐coupled production (wGCP), the production envelope allows for zero production of the target (i.e., v p , min = 0 ) until reaching some positive growth rate, g 0 , after which its production is always greater than zero. For strongly growth‐coupled production (sGCP, i.e., both directionally growth‐coupled production and substrate‐uptake‐coupled production) the production envelope displays positive production of the target for all growth rates greater than zero (i.e., v p , min g > 0 for all g > 0 ) and so thus has a maximum growth at production onset, g 0 ≔ v X , max v p , min = 0 , of zero. Directionally growth‐coupled production (dGCP) and substrate‐uptake‐coupled production (SUCP) differ in the minimal production rate at no growth, v p , min 0 ≔ v p , min g = 0 . dGCP has v p , min 0 = 0 , indicating that any growth necessitates product formation, whereas SUCP has v p , min 0 > 0 , indicating that product is always produced, even at no growth. For some wGCP or sGCP cases, instead of a singular product rate v p g max = v p , max g max = v p , min g max there exist equivalent optimal solutions (i.e., v p , max g max ≥ v p g max ≥ v p , min g max > 0 ). The RE design solutions can be ranked by quantitative criteria such as converting the above‐mentioned product and biomass production rates into equivalent product yields ( Y ), substrate‐specific productivity (SSP), \n 20 \n growth‐coupling strength (GCS). \n 21 \n \n FIGURE 1 Illustration of production envelopes for a wild type strain and four mutant strains having different qualities of growth‐coupled production. The accessible solution space is below each curve and the production envelopes are shown stacked with each including any regions to the left. For example, the production envelope of mutant strain C encompasses the regions marked C and D, as mutant strain D is drawn on top of mutant strain C. The wild type strain produces no product at its maximum growth rate and thus has null growth coupling. Key product and growth rates used in defining qualities are indicated. An industrially viable process relying on a microbial production of organic acids needs to be an efficient producer (i.e., have a high yield from sugar). \n 23 \n Although some of the above‐mentioned organic acids have been produced using production hosts such as Escherichia coli (e.g., succinic acid), the industrial processes require pH neutralization and thereby result in byproducts such as gypsum. Thus, ideally a production host tolerates the low pH associated with a high titer \n 23 \n which can enable product separation without extensive neutralization. The yeast Issatchenkia orientalis (also known as Pichia kudriavzevii , Candida glycerinogenes , and Candida krusei \n \n 24 \n ) has been proposed to be one such candidate host since it exhibits tolerance to high levels of succinic acid, itaconic acid, adipic acid, and acetic acid. \n 25 \n Strains can produce ethanol in media containing 5% sodium sulfate at pH 2.0. \n 26 \n Recombinant I. orientalis strains can produce titers of 11.6 g/L succinic acid \n 27 \n and 15–20 g/L lactic acid under anaerobic conditions in an unbuffered medium at a pH of 2. \n 28 \n Indeed, a strain was reported to produce as much as 154 g/L d ‐lactic acid at a pH of 4.7 after genetic modifications and subsequent adaptions to high lactic acid concentrations. \n 29 \n \n I. orientalis is capable of growth on a number of carbon substrates, with growth/no growth assays previously tested on 34 carbon substrates \n 26 \n as well as in a separate study of 26 carbon substrates. \n 30 \n In the former, seven of the carbon substrates scored positive (i.e., glucose, lactose, glycerol, lactic acid, succinic acid, citric acid, and ethanol) and four were delayed positive (i.e., xylose, sucrose, xylitol, and glucuono‐1,5‐lactone). \n 26 \n In the latter, six scored positive (i.e., same as the positives in the former excepting lactose). \n 30 \n Quantitative analysis of growth rates was recently evaluated for glucose, glycerol, fructose, succinic acid, lactic acid, citric acid, and ethanol. \n 31 \n Isolated strains have been found to grow at a pH of 2.5 on hemicellulosic and cellulosic oligosaccharides obtained by two‐step extraction with sulfuric acid from six plant sources. \n 32 \n \n Recent advances in genetic systems for this nonmodel microorganism \n 33 \n , \n 34 \n expedite its domestication as an industrial host, and a recently published genome‐scale metabolic (GSM) model \n 31 \n allows for the application of COBRA \n 3 \n tools. In the current work, we extend our limited‐scope examination of succinic acid production \n 31 \n and employ constraint‐based modeling to exhaustively examine the potential maximum theoretical yields and dependence of yield on oxygen uptake for a bevy of 22 organic acids from three to six carbons, including seven of the organic acid DOE building‐blocks, for individually both glucose and xylose carbon substrates. For non‐native organic acids, we introduce synthetic pathways that enable their production in the GSM model. We then use OptKnock to perform a large‐scale computational study to identify RE solutions that facilitate GCP strain designs. For each solution, we classify the various GCP strain designs with qualitative (i.e., pGCP, wGCP, dGCP, or SUCP) and quantitative (i.e., Y P / S , min g max , Y P / S , max g max , SSP, Y 0 , g 0 , and GCS) characteristics. We examine the distribution of RE solutions that occur for multiple organic acids. We also examine the impact on introducing other reactions into the system that can increase carbon yield for specific targets.",
"discussion": "3 RESULTS AND DISCUSSION 3.1 Selection of substrates and products Although I. orientalis is capable of growth on a range of carbon substrates, for our analysis we focused on the two major sugar monomers derived from components in processed hydrolyzed biomasss \n 47 \n , \n 48 \n —glucose and xylose—which can be >90% of it by weight. \n 49 \n As noted earlier, wild‐type I. orientalis grows well on glucose, and grows on xylose after a delay. \n 26 \n Thus, we selected both glucose and xylose as the carbon substrates as proxy constituents of hydrolyzed biomass in the study. Over a decade ago, the US Department of Energy identified twelve building block chemicals that can be produced from sugars \n 50 \n and these platform chemicals remain relevant to date. Eight of these building block chemicals are organic acids which range in length from three to six carbons. They are 3‐hydroxypropionic (i.e., 3‐hydroxypropanoic), fumaric, malic, succinic, itaconic, levulinic (i.e., 4‐oxopentanoic), 2,5‐furandicarboxylic and muconic acids. This report delineated their subsequent conversion to high‐value bio‐based chemicals and materials. For example, 3‐hydroxypropionic acid can be converted into chemicals such as 1,3‐propanediol, acrylic acid, methyl acrylate, and acrylamide, whereas succinic acid (i.e., butanedioic acid) can readily be converted to polymer precursors such as 1,4‐butanediol, N ‐methyl−2‐pyrrolidone, tetrahydrofuran and γ‐butyrolactone. In addition to these building block chemicals, a number of other small organic acids have important uses and markets. For example, lactic acid is used as a monomer for polymers, \n 28 \n 3‐hydroxy‐3‐methylbutanoic acid (HMB) is used as a human dietary supplement, \n 51 \n and 2,4‐dihydroxybutyric acid is used for chemical synthesis of the methionine analogue 2‐hydroxy‐4‐(methylthio)butyrate used in animal feed. \n 52 \n \n The products examined in the current study were largely drawn from the study from the US Department of Energy \n 50 \n ; in addition to the eight organic acids given above, we expanded the list to include the organic acids present in the report's down selection of the top 30 results, excluding 2,5‐furandicarboxylic acid. We also included a number of other small organic acids that have been examined for production in E. coli , \n 20 \n , \n 53 \n , \n 54 \n or have important uses and markets. The final list of 22 target products is given in Table 1 , arranged by increasing number of carbons, and the location of these products is shown in context of the metabolic network in Figure 2 , which indicates those requiring heterologous enzymes be added to the model as summarized in Table 2 . When multiple potential pathways existed, we chose those with the highest number of identified and characterized enzymes. TABLE 1 List of organic acids examined in the current study and computed maximum theoretical yields with respect to carbon balances on the indicated substrate Glucose Xylose Name ID \n a \n \n Formula Charge \n Y P c / S c \n \n Y P e / S e \n \n Y P / S \n \n Y P c / S c \n \n Y P e / S e \n \n Y P / S \n Malonate malon C 3 H 2 O 4 \n −2 2 3 2 1.667 2.5 1.667 Pyruvate pyr C 3 H 3 O 3 \n −1 2 2.4 2 1.667 2.0 1.667 \n d ‐Lactate lac__D C 3 H 5 O 3 \n −1 2 2 1.876 1.667 1.667 0.520 3‐Hydroxypropanoate 3hpp C 3 H 5 O 3 \n −1 2 2 1.994 1.667 1.667 0.553 Fumarate fum C 4 H 2 O 4 \n −2 1.5 2 1.5 1.25 1.667 1.25 \n l ‐Malate mal__L C 4 H 4 O 5 \n −2 1.5 2 1.5 1.25 1.667 1.25 Succinate succ C 4 H 4 O 4 \n −2 1.5 1.714 1 1.25 1.428 0.833 2‐Oxobutanoae 2obut C 4 H 5 O 3 \n −1 1.5 1.5 1.367 1.25 1.25 1.13 2,4‐Dihydroxybutanoate 24dhbut C 4 H 7 O 4 \n −1 1.5 1.5 1.428 1.25 1.25 1.146 3‐Hydroxybutanoate bhb C 4 H 7 O 3 \n −1 1.5 1.333 1 1.25 1.111 0.827 4‐Hydroxybutanoate ghb C 4 H 7 O 3 \n −1 1.5 1.333 1 1.25 1.111 0.833 2‐Oxoglutarate akg C 5 H 4 O 5 \n −2 1.2 1.5 1 1 1.25 0.833 Itaconate itacon C 5 H 4 O 4 \n −2 1.2 1.333 1 1 1.111 0.833 Citramalate citm C 5 H 6 O 5 \n −2 1.2 1.333 1 1 1.111 0.833 \n d ‐Xylonate dxylnt C 5 H 9 O 6 \n −1 1.2 1.333 0 1 1.111 1 2‐Oxopentanoate 2oxptn C 5 H 7 O 3 \n −1 1.2 1.091 1 1 0.909 0.827 4‐Oxopentanoate 4oxptn C 5 H 7 O 3 \n −1 1.2 1.091 1 1 0.909 0.818 3‐Methyl‐2‐oxobutanoate 3mob C 5 H 7 O 3 \n −1 1.2 1.091 1 1 0.909 0.810 3‐Hydroxy‐3‐methylbutanoate 3hivac C 5 H 9 O 3 \n −1 1.2 1 0.667 1 0.833 0.556 3‐Hydroxypentanoate 3hpt C 5 H 9 O 3 \n −1 1.2 1 0.891 1 0.833 0.706 Citrate cit C 6 H 5 O 7 \n −3 1 1.333 1 0.833 1.111 0.833 Muconate ccmuac C 6 H 4 O 4 \n −4 1 1.091 0.857 0.833 0.909 0.675 \n a \n ID taken from BiGG Models when available. FIGURE 2 Connections of target organic acids to I . orientalis metabolism. Pathways involving heterologous reactions added to the network are indicated in green. Arrows indicate overall reaction pathway reversibility, with multiple reactions condensed into a single line. The targeted organic acids are displayed with their chemical structures, as well as two possible liquid by‐products, glycerol, and ethanol. TABLE 2 Heterologous reactions added to enable the production of the indicated organic acid product Product Added reactions \n a \n \n Malonate Aspartate 1‐decarboxylase (ASP1DC); beta‐alanine‐pyruvate aminotransferase (ALABAT); malonic semialdehyde oxidoreductase (MSADx, MSADy) \n d ‐Lactate Lactate dehydrogenase (LDH_D) 3‐Hydroxypropanoate Aspartate 1‐decarboxylase (ASP1DC); beta‐alanine‐pyruvate aminotransferase (ALABAT); malonic semialdehyde reductase (MSAR) 2‐Oxobutanoate Transporter 2,4‐Dihydroxybutanoate Malate kinase (MALK); Malate semialdehyde dehydrogenase (MASD); malate semialdehyde reductase (MALSARx, MALSARy) 3‐Hydroxybutanoate Acetoacetyl‐CoA hydrolase (AACOAH); 3‐hydroxybutanoate oxidoreductase (BDH) 4‐Hydroxybutanoate Gamma‐hydroxybutyrate dehydrogenase (GHBDHx, GHBDHy) Itaconate \n Cis ‐aconitic acid decarboxylase (CAD) \n d ‐Xylonate \n d ‐Xylose NADP+ 1‐oxidoreductase (XYLOR); xylonolactonase (XYLC) 2‐Oxopentanoate 3‐Hydroxyacyl‐CoA dehydrogenase (acetoacetyl‐CoA) (HACD1); 3‐hydroxyacyl‐CoA dehydratase (3‐hydroxybutanoyl‐CoA) (ECOAH1); acyl‐CoA dehydrogenase (butanoyl‐CoA) (ACOAD1f); 2‐oxopentanoic acid decarboxylase (2OXPTNDH) 4‐Oxopentanoate 4‐Hydroxy‐2‐oxopentanoate aldolase (HOPNTAL); 4‐hydroxy‐2‐oxo‐pentanoate reductase (R4H2OPNTNR); 2,4‐dihydroxy‐pentanoate dehydratase (24DHPNTADH and 4O2PNTNR) 3‐Methyl‐2‐oxobutanoate Transporter 3‐Hydroxy‐3‐methylbutanoate 4‐Hydroxyphenylpyruvate dioxygenase/alpha‐ketoisocaproate dioxygenase (RE1266C) 3‐Hydroxypentanoate Acetyl‐CoA:2‐oxobutanoate C‐acetyltransferase (AC2OBUTAT); 2‐ethyl‐2‐hydroxybutanedioate carboxy‐lyase (2E2HOBTAECBOX) Muconate 3‐Dehydroshikimate dehydratase (DHSKDH); protocatechuic acid (PCA) decarboxylase (PCADC); catechol dioxygenase (CATADOX) \n a \n Transporters are only listed if the sole addition. Exchange fluxes are not listed. Complete details of added pathways are in Supporting Information, Data S1 . 3.2 Production potential When evaluating the potential of forming a product from a substrate, factors from both chemistry and biochemistry weigh in. We performed an initial analysis of the three theoretical yields (two chemical and one biochemical) from conversion of the substrates into each product. The first was carbon balance yield, Y P c / S c , that uses is the number of carbons in the product and sugar carbon substrate. Second was the yield based on available electrons, Y P e / S e , which considers the charges of the products and substrate and computes the amount of oxygen required to balance an oxidation reaction of each product in its charged form. Third was the biological theoretical yield, Y P / S , which requires placing production in context of the metabolic network of an organism. Here, we used the recent GSM model iIsor 850 that accounts for the metabolic capabilities of I. orientalis SD108; the model has a customized biomass reaction determined form experimental data and is mass and charge balanced. \n 31 \n We set the oxygen uptake rate upper limit to the minimal oxygen uptake that does not impinge the maximum growth rate of the wild‐type model, and we note that I. orientalis is incapable of anaerobic growth. \n 25 \n \n 3.3 Chemical theoretical yields We first performed a carbon balance, denoted as Y P c / S c in Table 1 , which was simply the ratio of carbons in the C3–C6 organic acid products and glucose (C6) or xylose (C5). We used these values to order the products in the table and as a metric for comparing the other yield computations. As seen in Table 1 , for the yields computed from available electrons, Y P e / S e , can differ for organic acids with the same number of carbons and generally decreases as the number of carbons in the product increases, reflecting the net charge on these mostly mono or dicarboxylic acids. By depending on the structure and charge of each product, these chemical theoretical yields form the upper bound of what is achievable in the absence of additional reducing power. Because they permit electron‐balanced carbon uptake from CO 2 in the form of HCO − \n 3 as a reactant in the equation, Y P e / S e can be greater than Y P c / S c since this yield is from a sugar substrate carbon standpoint. Notably, the 11 products with Y P e / S e > \n Y P c / S c highlight potential opportunities for engineering carbon fixing strategies. For seven of the products Y P e / S e is less than Y P c / S c , reflecting a lower potential, at least without a source of additional reducing power. For these cases, Y P e / S e instead of Y P c / S c tempers expectations of what could be theoretically achievable for any biological system, without other sources of reducing power. 3.4 Biochemical theoretical yields Examination of the model revealed that six target metabolites could be produced by I. orientalis without modifications to the model and include products such as succinic acid, pyruvic acid, l ‐malic acid, and citric acid. For the remaining product targets, we separately implemented pathways that enable their production by adding mass and charge balanced reactions to the network. The connection of the products to metabolism are highlighted in Figure 2 and the added pathways are outlined in Table 2 . For instance, for itaconic acid we added cis‐aconitic acid decarboxylase (CAD) and associated transporters and exchange fluxes. For 3‐hydroxypropionic acid we added the non‐native reactions in the β‐alanine pathway (i.e., pathway III \n 55 \n ). The full specifics of all these pathways are given in Supporting Information, Data S1 . At this stage, no other modifications were considered such as reaction eliminations or additions to other parts of the network beyond that minimally required to enable metabolite production, and we performed each computation without any additional constraints other than the model's value for nongrowth‐associated ATP maintenance. The product d ‐xylonate was not able to be produced from glucose and was only examined using xylose substrate conditions. Using models with added reactions, as required for product formation, we used flux balance analysis (FBA) \n 14 \n to compute the maximum biological product yields, Y P/S . As seen in Table 1 for glucose six acids can be produced at the maximum carbon balance yield and for xylose, seven can. Four of the targeted organic acids could be produced at biochemical yields close to the carbon balances for both sugars: 3‐hydroxypropanoate, 2,4‐dihydroxybutanoate, d ‐lactate, and 2‐oxobutanoate. Interestingly, these same four are the ones for which the corresponding carbon balance and available electron balance have the same value. The lower yield from the biological analysis is not unexpected, at least in part, for some products because the lower yield stems from the non‐growth associated ATP maintenance that diverts a small amount of carbon through ATP production irrespective of other processes in the model. The remaining products had lower yields, with 3‐hydroxy‐3‐methylbutanoate having a substantially low one. Via FBA, we found that all of the selected organic acids have the potential under some condition to be homofermenting, with the only byproduct being CO 2 . We also found that, as expected, enforcing a minimal specific growth rate, v X,s , of 10% that of the wild‐type maximal specific growth rate corresponding to a carbon source uptake of uptake of 10 mmol gDW −1 h −1 (i.e., growth rate of at least 0.1033 h −1 for growth on glucose or 0.0855 h −1 for growth on xylose) resulted in correspondingly decreased yields. 3.5 Potential from augmentation Noting that many yields could use improvement, we examined the impact of additions to the network that could increase yield. Specifically, for each organic acid we examined if the addition of fumarate reductase (FRD) activity would improve yield, which has been used experimentally \n 27 \n and in silico \n 31 \n to increase succinic acid production in I. orientalis . We also examined if the uptake of carbonate, HCO 3 \n − , facilitated by pyruvate carboxylase (PC) activity which has increased citric acid production in Yarrowia lipolytica . \n 56 \n We found that only succinic acid production was improved by fumarate reductase activity, and only malate and fumarate were improved by carbonate uptake, as was succinic acid but only with simultaneous expression of fumarate reductase activity. Notably, both malate and fumarate were able to reach the available electron theoretical yields for glucose, a 33% increase, as was succinate for both glucose and xylose via simultaneous carbonate uptake and fumarate reductase expression, a 71% increase. 3.6 Oxygen dependence We generated production envelopes for each target product to examine how restricting oxygen uptake rates impacts product exchange flux rates. We illustrate these results in Figure 3 for glucose and xylose using the wild‐type or augmented networks without any imposed reaction eliminations. We varied the oxygen uptake from 0 (i.e., anaerobic, under which conditions I. orientalis does not grow) to 18.18 mmol gDW −1 h −1 (i.e., the minimal oxygen uptake that does not impact the maximum growth rate on a 10 mmol glucose gDW −1 h −1 basis). In general, for the nonaugmented pathways, the products for which Y P c / S c is less than Y P e / S e required higher oxygen uptake rates to achieve the maximum production rate, whereas those for which Y P c / S c equals Y P e / S e had relatively low oxygen uptake requirements (i.e., below 5 mmol gDW −1 h −1 ). In particular, 3‐hydroxypropionic acid had its maximum value at near anaerobic conditions. The impact on oxygen requirements is improved for all the augmented pathways when compared to the native pathways, which is especially pronounced for the uptake of HCO 3 \n − via PC. For xylose, the results are similar but typically shifted to higher oxygen requirements. Of particular note is that some products, including 3‐hydroxypropionic acid, using xylose as the carbon substrate have an oxygen threshold below which they cannot be produced. This situation arises from inability of I. orientalis to grow anaerobically and the resulting incapacity to balance cofactors under those conditions. We also find that the oxygen utilization for succinate production has low oxygen requirements in the context of FRD and PC expression. FIGURE 3 Oxygen utilization using glucose (solid line) and using xylose (dashed line) as the substrate. The products are arranged by number of carbons and by a comparison of carbon and available electron yields. A dotted line separates production envelopes for the organic acids augmented with FRD and PC expression. 3.7 Network analysis and accessible reactions for targeted eliminations At the onset of our analysis, we performed analyses to reduce the reaction space for subsequent analyses to those that are accessible and practical as targets for genetic knockout. The first stage was performed in common to all carbon source inputs, whereby examination of the model's 1832 reactions found 580 have Systems Biology Ontology (SBO) \n 42 \n terms for transporters and 173 have SBO terms for exchange or other boundary reactions. As many as 719 have no defined gene–protein–reaction (GPR) associations or are spontaneous, which can overlap with the previous sets. For aerobic glucose conditions, by using FVA we found 760 blocked reactions and 386 essential reactions. Similarly, for aerobic xylose conditions we found 755 blocked and 391 essential reactions. Because we seek to find implementable interventions, we then examined the set of essential genes and the reactions they encode; some of these reactions might not be essential individually, but knocking out a gene could potentially remove several reactions at once thereby forming a synthetic lethal. Others are not essential but are only associated with a gene that is essential because it catalyzes a different reaction which is essential. We identified 230 essential genes for glucose and 239 for xylose, which led to 98 (glucose) and 97 reactions (xylose) non‐essential reactions encoded by only an essential gene. Combining these results with a subsequent examination of reactions with complex GPR identified a total of 106 (glucose) and 107 (xylose) reactions excluded by the GPR analysis. By combining the results for classification, blocked, and essential analyses, we identified a set of potentially reactions accessible to be eliminated containing 279 (glucose) and 280 (xylose) reactions. Optimal reaction elimination algorithms such as OptKnock are computationally expensive and memory intensive as the number of variables and allowed simultaneous eliminations increase. To further improve computational tractability by reducing the number of reactions targeted for elimination, we turned to flux coupling analysis by using the Flux Coupling Finder (FCF) algorithm. \n 40 \n Doing so allows us to identify equivalent knockouts, and reduce the reaction space for subsequent analyses. We found 81 (glucose) and 82 (xylose) sets of fully coupled reactions. By permitting only one member of each to represent the group, taking into account the previous excluded reactions, we found a set of non‐redundant practical reactions for glucose and xylose to both be 223 reactions, which differ by six reactions: ferrocytochrome‐c:oxygen oxidoreductase (FECOOR_m), ubiquinol:ferricytochrome c reductase (FECRq7_m) and glucose‐6‐phosphate isomerase (PGI) present for glucose and aldehyde dehydrogenase (3‐aminopropanal, NAD) (ALDD22x), hexokinase ( d ‐glucose:ATP) (HEX1) and spermine synthase (SPRMS) present for xylose; these six are essential reactions for the other substrate. These sets are made available in Supporting Information, Data S2 . For each substrate–product pair, we used Flux Variability Analysis (FVA) to identify sets of reactions in the non‐redundant practical reactions that individually permitted zero flux through each when constraining the production of the target to its maximum value obtained when biomass production was constrained to at least 10% of its maximum value, that is, v j , min g = 0.1 g max ≤ 0 ≤ v j , max g = 0.1 g max . For some, zero was the only permitted value (i.e., any nonzero flux through them would negatively impact product yield), whereas others had non‐zero v j , min g = 0.1 g max and/or v j , max g = 0.1 g max . We also used FBA to examine the effect on product yield from eliminating each reaction in the respective non‐redundant practical reactions. Most had no impact and the vast majority of those that did only did so with minimal reductions (i.e., < 1%). Some, such as pyruvate decarboxylase (PYRDC), glucose‐6‐phosphate isomerase (PGI) and NADH:ubiquinone oxidoreductase (complex I) (NADHcplxI_c_m) had larger negative impacts (i.e., 20% or more reduction) on production for some but not all substrate‐product pairs. These sets (and corresponding reductions in yield, as appropriate) are made available in Supporting Information, Data S2 . 3.8 Initial computationally predicted strain designs We used OptKnock to determine RE designs for the target organic acids in order to broadly examine the full range of growth‐coupled production strategies available, including potentially growth‐coupled production. For initial OptKnock simulations, all reactions other than the 223 non‐redundant practical reactions for glucose and xylose were fixed to be on and not allowed to be eliminated. We also excluded reactions for which single elimination resulted in zero product formation. All RE designs were subsequently partitioned into pGCP, wGCP, dGCP, and SUCP categories for each substrate–product pair. Excluding those for the augmented pathways, across all glucose, reaction pairs we found as many as 3674 RE solutions containing of up to five reactions that had a potential product yield of at least 10% carbon yield at maximum growth rate (i.e., Y P / S , max g max / Y P c / S c ≥ 0.1 ) ; these RE solutions contained 105 different reactions. For xylose we found as many as 3318 solutions involving 95 reactions meeting the same criterion. These reactions are provided in Supporting Information, Data S3 . Of these, 13 (glucose) and 3 (xylose) reactions were unique to the one substrate. The number of designs per substrate–product pair varied considerably, with some such as native metabolites pyruvate, fumarate, and succinate having many solutions. For glucose, we found dGCP or SUCP designs for only six of the products (viz., malonate, pyruvate, d ‐lactate, 2‐oxoglutarate, and 2‐oxopentanoate) and for xylose, we found dGCP or SUCP designs for malonate, d ‐lactate, succinate, 2‐oxoglutarate, d ‐xylonate, and 2‐oxopentanoate, with succinate, 2‐oxoglutarate, and d ‐xylonate having numerous solutions. Interestingly, d ‐xylonate had the largest number of SUCP designs, reaching as high as 0.923 of the carbon yield from xylose. We found wGCP designs for 15 and 13 products from glucose and xylose, respectively. Only pGCP designs existed for malate, 4‐hydroxybutanoate, and 4‐oxopentanoate for both sugars, and as many as 10 substrate‐product pairs had no solutions, including 2‐oxobutanoate, 3‐hydroxybutanoate, citramalate, 3‐hydroxypentanoate. As seen in Figure 2 , the later three all derive wholly or in part from acetyl‐CoA. For each strain design we computed our evaluation metrics including Y p , min , s g max , s , Y p , max , s g max , s , g 0 , Y p , min , s g = 0 , SSP p,s , and GCS p,s , as well as the range of oxygen uptake at g max,s . We ranked the RE solutions descending using Y p , max g max for pGCP and Y p , min g max for wGCP, dGCP, and SUCP; the RE solutions and corresponding metrics values are provided in Supporting Information, Data S3 . We compared the similarity of solutions for the two carbon sources by computing the Jaccard index across all products, shown in Table 3 . Comparisons of solutions made within a substrate are in Supporting Information, Data S4 . In general, we found that there was less than 50% overlap of designs for the two sugars. TABLE 3 Jaccard similarity coefficient for reaction elimination solutions using glucose and xylose as substrates Xyl Glc malon pyr lac__D 3hpp fum mal__L succ ghb akg 2oxptn 4oxptn 3mob 3hivac cit ccmuac dxylnt malon 0.451 0.010 0 0 0 0 0 0 0 0 0 0.025 0 0 0.003 0.018 pyr 0.020 0.444 0 0 0 0 0 0 0 0.219 0.337 0 0.017 0 0.001 0.004 lac__D 0 0 0.421 0 0 0 0 0 0 0 0 0 0 0 0 0 fum 0 0 0 0 0.194 0 0.007 0 0 0 0 0 0 0 0.003 0.011 mal__L 0 0 0 0 0.004 0.333 0 0 0 0 0 0 0 0 0 0 succ 0 0 0 0 0.008 0 0.268 0 0.002 0 0 0 0 0 0 0.005 ghb 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 akg 0 0 0 0.022 0 0 0 0.206 0.208 0 0 0 0 0 0 0.009 2oxptn 0 0.217 0 0 0 0 0 0 0 0.372 0.219 0 0.023 0 0.011 0.019 4oxptn 0 0.314 0 0 0 0 0 0 0 0.170 0.352 0 0.014 0 0 0 3mob 0.036 0 0 0 0 0 0 0 0 0 0 0.204 0 0 0 0.010 3hivac 0.011 0.026 0 0 0 0 0 0 0 0.026 0.019 0 0.237 0 0.008 0.021 cit 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 ccmuac 0.004 0 0 0 0 0 0.001 0 0 0.006 0 0 0.002 0 0.249 0.014 3.9 Examination of factors contributing to designs We examined the reactions that occurred within the solutions of each product for the substrate‐product pairs by creating incident matrices encompassing the reactions for each qualitative category of solutions. We found that some reactions were ubiquitous for a given substrate‐product pair, or nearly so. As given in Table 4 , we found that for as many as 13 of the products there were such reactions. Some of these 28 reactions, such as ATP synthase (APTS_m), pyruvate decarboxylase (PYRDC), appeared in more than one product. We also observed differences between the two substrates which was particularly noticeable for 3‐hydroxypropionate, for which had mutually exclusive lists, with ribulose 5‐phosphate 3‐epimerase (RPE) only occurring for xylose and aspartate transaminase (ASPTA), ATPS_m, and NADH dehydrogenase (NADHq7) only occurring for glucose. TABLE 4 Ubiquitous and near‐ubiquitous reactions occurring in GCP solutions Glucose Xylose ID \n a \n \n Total # Ubiquitous \n b \n \n Near‐ubiquitous (#) \n c \n \n Total # Ubiquitous Near‐ubiquitous (#) All GCP malon 311 PYRDC 194 PYRDC (186) pyr 531 ATPS_m (499) 520 ATPS_m (508) lac__D 79 RLFC2O, RLFC2O_m, TPI 56 RLFC2O, RLFC2O_m, TPI 3hpp 10 ASPTA, ATPS_m, NADHq7 68 RPE (67) fum 427 FUM 231 FUM (229) mal__L 2 CRNOAT, MDH 6 MDH G6PDH2i (5) succ 787 SUCDq7_m (775) 727 SUCDq7_m (676) 24dhbut 10 GLUDy, MDH, NADHcplxI_m 0 ghb 14 ASPTAi_m, NADHq7, SUCDq7_m 14 ASPTAi_m, NADHq7, SUCDq7_m akg 298 GLUDC (290) 271 GLUDC G6PDH2i (235) dxylnt N/A 703 G6PDH2i (690) 4oxptn 630 ATPS_m 410 ATPS_m 3mob 338 PDH_m (301) 145 PDH_m (137) wGCP+sGCP malon 306 PYRDC 181 PYRDC pyr 62 PYRDC 50 PYRDC lac__D 63 RLFC2O, RLFC2O_m, TPI 47 RLFC2O, RLFC2O_m, TPI 3hpp 10 ASPTA, ATPS_m, NADHq7 0 fum 427 FUM 229 FUM succ 779 SUCDq7_m (677) 681 SUCDq7_m (676) 24dhbut 10 GLUDy, MDH, NADHcplxI_m 0 akg 290 GLUDC 249 GLUDC G6PDH2i (235) dxylnt N/A 696 G6PDH2i (687) 2oxptn 0 276 NADHcplxI_m (271) 3mob 4 3MOBDC, MDH_m, NADHq7, PDH_m 0 3hivac 38 4MOPDC, NADHcplxI_m 14 4MOPDC, NADHcplxI_m \n a \n ID taken from BiGG Models when available. \n b \n Reaction abbreviations are: 3MOBDC 3‐methyl‐2‐oxobutanoate decarboxylase, 4MOPDC 4‐methyl‐2‐oxopentanoate decarboxylase, ASPTA aspartate transaminase, ASPTAi_m aspartate transaminase, ATPS_m ATP synthase, CRNOAT carnitine O‐acetyltransferase, DDPA 3‐deoxy‐D‐arabino‐heptulosonate 7‐phosphate synthetase, FUM fumarase, G6PDH2i glucose 6‐phosphate dehydrogenase, GLUDC glutamate decarboxylase, GLUDy glutamate dehydrogenase (NADP), ICDHx_m isocitrate dehydrogenase (NAD+), ICDHyi isocitrate dehydrogenase (NADP), ICL_1 isocitrate lyase, MDH malate dehydrogenase, MDH_m malate dehydrogenase, ME1_m malic enzyme (NAD), ME2_m malic enzyme (NADP), NADHcplxI_c_m NADH:ubiquinone oxidoreductase (complex I), NADHq7 NADH dehydrogenase, PDH_m pyruvate dehydrogenase, PGI glucose‐6‐phosphate isomerase, PYRDC pyruvate decarboxylase, RLFC2O (R)‐lactate:ferricytochrome‐c 2‐oxidoreductase, RLFC2O_m (R)‐lactate:ferricytochrome‐c 2‐oxidoreductase, RPE ribulose 5‐phosphate 3‐epimerase, SUCDq7_m succinate dehydrogenase (ubiquinone‐7), TPI triose‐phosphate isomerase. Reaction abbreviations ending with _m signify reactions localized to the mitochondria. \n c \n Numbers in parentheses are the incidence counts of the respective reaction. Analysis for reactions that we found occurring ubiquitously in the solutions revealed enzymes that have been the target of industrial‐based processes or previous strain designs. For instance, (R)‐lactate:ferricytochrome‐c 2‐oxidoreductase has been reported as a target for improving lactic acid production in recombinant yeast. \n 29 \n In E. coli , aspartate transaminase activity has been a target for deletion to increasing the production of β‐alanine, \n 57 \n which is an intermediate in the 3‐hydroxypropionate pathway we used. A strain of S. cerevisiae that oxidatively produces succinate involved the deletion of succinate dehydrogenase. \n 58 \n The deletion of pyruvate decarboxylase, as seen in our malate and pyruvate designs, has also been examined in S. cerevisiae , \n 59 \n although the authors focused on its impact in succinic acid production; it does occur in our designs, albeit not ubiquitously. The deletion of ATP synthase has been observed to be a frequent deletion target in other growth‐coupling strategies in E. coli \n \n 21 \n whereby the ATP yield from the pathway producing the target product is greater than the other residual ATP forming pathways after ATP synthase is deleted. In general, these ubiquitous and near‐ubiquitous reactions are early candidates for implementation in I. orientalis , especially those that occur for several organic acids, such as pyruvate decarboxylase. As expected from our previous limited study on adding fumarate reductase (FRD) activity to improve succinic acid yield in silico, \n 31 \n which agreed with in vivo experiments \n 27 \n and in the current work, we found that its production could be enhanced through the uptake of carbonate. The RE designs for succinate, fumarate and malate augmented with FRD and PC expression were similar to those without. Using the incident matrices, the structure of RE design solutions for a substrate‐product pair can be revealed, for instance, using hypergraphs \n 46 \n and bipartite graphs with nodes representing the reactions in a solution as well as a node for each solution. In general, we found that higher order sets are not necessarily constructed by simply adding on reactions to existing solutions, though such smaller sets can be kernels for some of them. For example, for sGCP designs for xylose‐succincate, the five solutions of size three could be extended into 23 of the 40 solutions of size four upon the addition of one differing extra reaction. Similarly, pGCP designs can be promoted to wGCP or sGCP through the addition of reactions in higher‐order sets, as can wGCP to dGCP or SUCP designs. 3.10 Improving strain designs categorization Using the results from Table 4 and the sets of reactions found in Supporting Information, Data S2 that impact each product's yield and from the initial computationally predicted strain designs, we subsequently examined higher‐order simultaneous reaction eliminations focusing on finding designs with improved qualitative and/or quantitative metrics: either entry into a better category of GCP for those with at best ∅GCP, pGCP, or wGCP designs or improved quantitative metrics for those already with dGCP or SUCP designs. We also continued searches without any additional constrains on selectable reactions. Table 5 summarizes our results for all strain design solutions. We found as many as 3841 solutions involving 106 reactions for glucose and 3393 solutions and 95 reactions for xylose. These reactions are tabulated in Supporting Information, Data S2 . TABLE 5 Classification and quantification of strain design solutions ID \n a \n \n Glucose Xylose pGCP wGCP dGCP SUCP pGCP wGCP dGCP SUCP # \n Y P / S , max g max \n # \n Y P / S , min g max \n # \n Y P / S , min g max \n # \n Y P / S , min g max \n # \n Y P / S , max g max \n # \n Y P / S , min g max \n # \n Y P / S , min g max \n # \n Y P / S , min g max \n malon 5 1.729 288 1.743 0 18 1.809 13 1.443 165 1.268 0 16 1.450 pyr 469 1.842 61 1.803 0 1 1.822 470 1.325 50 1.299 0 0 lac__D 16 0.638 70 0.918 6 0.915 35 0.948 9 0.558 43 0.610 0 14 0.637 3hpp 3 1.058 46 1.189 0 0 68 1.016 14 1.144 0 7 1.028 fum 0 427 0.578 0 0 2 0.200 229 0.589 0 0 mal__L 20 0.235 0 0 0 31 0.229 0 0 0 succ 8 0.288 4 0.727 775 0.728 0 46 0.587 5 0.317 547 0.713 129 0.727 2obut 0 0 0 0 15 0.927 0 0 0 24dhbut 0 36 0.487 0 0 0 0 0 0 bhb 4 0.339 0 0 0 0 0 0 0 ghb 20 0.682 11 0.469 0 0 23 0.569 7 0.392 0 0 akg 8 0.357 278 0.660 12 0.543 0 22 0.481 202 0.716 0 47 0.726 itacon 0 11 0.259 0 0 0 1 0.174 0 0 citm 2 0.365 0 0 0 0 0 0 0 dxylnt N/A N/A N/A N/A 7 0.771 464 0.906 8 0.862 224 0.923 2oxptn 244 0.908 162 0.298 97 0.926 4 0.923 221 0.658 260 0.489 4 0.160 12 0.313 4oxptn 630 0.921 0 0 0 410 0.662 0 0 0 3mob 334 0.591 4 0.443 0 0 145 0.503 0 0 0 3hivac 147 0.505 38 0.505 0 0 99 0.394 14 0.391 0 0 3hpt 0 0 0 0 0 0 0 0 cit 1 0.133 0 4 0.203 0 0 0 3 0.169 0 ccmuac 2 0.131 288 0.408 0 0 4 0.104 298 0.406 0 0 \n a \n ID abbreviations are as in Table 1 . We found that the constrained maximum oxygen bound described above in the oxygen dependence analysis (i.e., 18.18 mmol gDW −1 h −1 ) was reached by 2498 glucose–product pair designs, of which 1037 also had the minimum oxygen uptake at the same limit. In concordance with the oxygen dependence analysis, examination revealed that except for succinate, nearly all sGCP designs had maximum oxygen uptake rates below this constraint. Results for xylose were proportionally smaller, with 1732 designs having maximum oxygen uptake rates constrained by the bound and 364 designs having the minimum oxygen uptake also so constrained. This difference arises because the oxygen uptake constraint was set based on glucose uptake; selecting xylose designs with oxygen uptake rates above 15.42 mmol gDW −1 h −1 (i.e., the minimal oxygen uptake that does not impact the maximum growth rate on a 10 mmol xylose gDW −1 h −1 basis) reveals 2695 designs (maximum) and 1368 designs (minimum). We added 3‐hydroxypropionic and citric acids to the set of acids with sGCP designs, added 4‐oxobutanoic acid to those with wGCP, and identified solutions with increased quantitative metrics for other acids. We were able to find one pGCP design for citramalate from glucose and one for 2‐oxobutanoate from xylose, but were unable to find any solutions for the other substrate‐product pairs previously with only ∅GCP. Figure 4 summarizes the wGCP and sGCP designs for the 15 substrate–product pairs for which we were able to find at least one dGCP or SUCP design. It also shows the corresponding production envelope for the highest‐ranked design of each pair, as determined by GCS. A total of 11 substrate–product pairs had at least one SUCP design. In total, nine pairs are predicted to have yields near the target biological theoretical maximum with seven also having high carbon yields. All design solutions summarized in Table 5 are provided in Supporting Information, Data S3 , and production envelope scatter plots for all organic acids are provided in Supporting Information, Data S5 . FIGURE 4 Production envelopes showing the product rates accessible for each feasible growth rate using (a) glucose and (b) xylose as the substrate. In the main of the paired plots, each design solution is represented by a scatter point at the maximum growth rate and associated minimum production rate. These designs are color coded as in Figure 1 , with blue circles indicating weakly growth‐coupled production designs, orange triangles directionally growth‐coupled production designs, and red squares substrate‐uptake coupled production designs. On the inset of the paired plots, the production envelope for the strongly growth‐coupled production design with the highest GCS metric is filled with the color of its corresponding symbol. For the highest‐ranked RE designs for the 15 pairs in Figure 4 , we used GPR associations to ascertain the feasibility of implementing them via gene knockouts. For each pair, we report the corresponding number of reaction:gene interventions; the differing numbers for each intervention type arose typically because of isozymes associated with some reactions. For glucose, five products had identical production envelopes for both RE and the corresponding gene knockouts (viz., malonate (5:5), d ‐lactate (5:7), succinate (5:7), 2‐oxoglutarate (5:5), and 2‐oxopentanoate (3:4)). For xylose, three had identical production envelopes for both RE and gene knockouts (viz., malonate (5:5), d ‐lactate (5:7), and 3‐hydroxypropanoate (6:7)). Glucose production of pyruvate (5:6) and xylose production of 2‐oxoglutarate (5:5) and d ‐xylonate (5:6) had the same v p , min g max for both RE and gene knockouts and had lower v p , min 0 for each respective gene implementation but remained SUCP. For xylose production of succinate (5:7) the gene implementation revealed a synthetic lethal involving glycerol‐3‐phosphate acyltransferase (G3PAT_l). Use of an equivalent alternative RE design (i.e., one having all the same number of reactions, v p , min g max , and GCS score) corresponded to a 6 gene knockout design with the same v p , min g max and a lower v p , min 0 that remained SUCP. Xylose production of 2‐oxopentanoate (4:4) similarly had the same v p , min g max for both RE and gene knockout designs but the latter's v p , min 0 decreased to zero which relegated it to dGCP. Finally, analysis of citrate produced from glucose (6:7) and xylose (6:7) also uncovered a synthetic lethal at the gene level, arising from the inclusion of the cytosolic reaction 3‐deoxy‐ d ‐arabino‐heptulosonate 7‐phosphate synthetase (DDPA_c). Examination for both substrates of reaction and gene implementations excluding this reaction reveals that all designs remain dGCP but fall below the carbon yield cutoff filter. A more general analysis using FBA to uncover other synthetic lethals in other RE designs also revealed participation by the reactions glycerol‐3‐phosphate acyltransferase (G3PAT_rm), ribonucleotide reductase (UDP) (RNDR4_c) and 3‐deoxy‐ d ‐arabino‐heptulosonate 7‐phosphate synthetase (DDPA_m). These five reactions impact as many as 246 glucose solutions and 314 xylose solutions. In summary, although there can be execution challenges when designing based on RE, we found realizable gene‐level implementations with identical or comparable production envelopes to the RE designs for 12 of the 15 substrate–product pairs that exhibited sGCP, and the remaining three nevertheless retained sGCP."
} | 13,666 |
34427516 | PMC8409731 | pmc | 5,031 | {
"abstract": "ABSTRACT Investigation of microbial communities has led to many advances in our understanding of ecosystem function, whether that ecosystem is a subglacial lake or the human gut. Within these communities, much emphasis has been placed on interspecific variation and between-species relationships. However, with current advances in sequencing technology resulting in both the reduction in sequencing costs and the rise of shotgun metagenomic sequencing, the importance of intraspecific variation and within-species relationships is becoming realized. Our group conducts multi-omic analyses to understand how spatial structure and resource availability influence diversification within a species and the potential for long-term coexistence of multiple ecotypes within a microbial community. Here, we present examples of ecotypic variation observed in the lab and in the wild, current challenges faced when investigating intraspecies diversity, and future developments that we expect to define the field over the next 5 years."
} | 255 |
21524904 | null | s2 | 5,033 | {
"abstract": "Recapitulating the elegant structures formed during development is an extreme synthetic and biological challenge. Great progress has been made in developing materials to support transplanted cells, yet the complexity of tissues is far beyond that found in even the most advanced scaffolds. Self-assembly is a motif used in development and a route for the production of complex materials. Self-assembly of peptides, proteins and other molecules at the nanoscale is promising, but in addition, intriguing ideas are emerging for self-assembly of micron-scale structures. In this brief review, very recent advances in the assembly of micron-scale cell aggregates and microgels will be described and discussed."
} | 176 |
26277623 | null | s2 | 5,034 | {
"abstract": "Flagella propel bacteria during both swimming and swarming, dispersing them widely. However, while swimming bacteria use chemotaxis to find nutrients and avoid toxic environments, swarming bacteria appear to suppress chemotaxis and to use the dynamics of their collective motion to continuously expand and acquire new territory, barrel through lethal chemicals in their path, carry along bacterial and fungal cargo that assists in exploration of new niches, and engage in group warfare for niche dominance. Here, we focus on two aspects of swarming, which, if understood, hold the promise of revealing new insights into microbial signaling and behavior, with ramifications beyond bacterial swarming. These are as follows: how bacteria sense they are on a surface and turn on programs that promote movement and how they override scarcity and adversity as dense packs."
} | 216 |
35767228 | PMC9546210 | pmc | 5,035 | {
"abstract": "Abstract Understanding the role of animal behaviour in linking individuals to ecosystems is central to advancing knowledge surrounding community structure, stability and transition dynamics. Using 22 years of long‐term subtidal monitoring, we show that an abrupt outbreak of purple sea urchins ( Strongylocentrotus purpuratus ), which occurred in 2014 in southern Monterey Bay, California, USA, was primarily driven by a behavioural shift, not by a demographic response (i.e. survival or recruitment). We then tracked the foraging behaviour of sea urchins for 3 years following the 2014 outbreak and found that behaviour is strongly associated with patch state (forest or barren) transition dynamics. Finally, in 2019, we observed a remarkable recovery of kelp forests at a deep rocky reef. We show that this recovery was associated with sea urchin movement from the deep reef to shallow water. These results demonstrate how changes in grazer behaviour can facilitate patch dynamics and dramatically restructure communities and ecosystems.",
"introduction": "INTRODUCTION The importance of behaviour in linking individuals to ecosystems is widely recognised in the ecological literature (Ovadia & Schmitz, 2002 ; Schmitz, 1998 ; Sih et al., 2012 ; Werner & Peacor, 2003 ). Behaviour can facilitate community structure and functioning by altering the relative influence of key species interactions (e.g. competition, predation, mutualisms), changing the distribution of resources and through other non‐consumptive response pathways (Estes et al., 1998 ; Pace et al., 1999 ; Werner & Peacor, 2003 ). Although the debate continues over the relative importance of density versus behaviorally mediated influences of predators and primary consumers, both occur widely in nature and are often associated with trophic cascades (Beckerman et al., 1997 ; Kauffman et al., 2010 ; Schmitz et al., 1997 ; Werner & Peacor, 2003 ). Therefore, understanding how the presence of predators and resource availability reciprocally influence the behaviour of primary consumers is central to advancing knowledge of community structure, functioning, stability, and transition dynamics. Sea urchin grazing in marine ecosystems around the world is often considered a fundamental driver of shifts from algal‐dominated habitats to alternative sea urchin ‘barrens’ that are void of macroalgae (Filbee‐Dexter & Scheibling, 2014 ; Ling et al., 2015 ). These shifts have profound consequences on the structure and functioning of coral reefs, seagrass, kelp forest and rocky intertidal ecosystems (Done, 1992 ; Filbee‐Dexter & Scheibling, 2014 ; Watson & Estes, 2011 ). Resource availability and predator‐driven impacts are perhaps the two most well‐documented factors known to influence patterns in sea urchin grazing behaviour (Burt et al., 2018 ; Cowen, 1983 ; Harrold & Reed, 1985 ; Mann, 1982 ). Cascading effects resulting from the loss of sea urchin predators provide strong evidence of density‐mediated indirect interactions (Burt et al., 2018 ; Estes et al., 1998 ), whereas reductions in the availability of food or risk‐cues have been associated with behaviorally mediated indirect interactions (Harding & Scheibling, 2015 ; Spyksma et al., 2017 ). However, the relative influence of these factors is often context‐dependent and difficult to decouple from other more environmentally driven processes such as how grazers respond to substrate complexity, seasonality, swell and water temperature (Konar et al., 2014 ; Randell et al., 2022 ; Vivian‐Smith, 1997 ). Therefore, the factors that contribute to modifications in sea urchin grazing behaviour can have important implications for the state of communities and ecosystems. In temperate kelp forest ecosystems, sea urchin behaviour can be categorised into two fundamental modalities. In kelp forests, where abundant detrital (i.e. ‘drift’) algae are deposited in crevices, urchins mainly employ a cryptic passive‐grazing modality (Duggins et al., 1989 ; Krumhansl & Scheibling, 2012 ; Sala & Zabala, 1996 ). The presence of predators may also elicit a direct response in sea urchins that influence cryptic behaviour or indirectly by maintaining forests (and therefore abundant drift) through trophic cascades (Cowen, 1983 ; Estes et al., 1998 ). However, when the production of detrital kelp is limited, sea urchins fundamentally shift their behaviour to an active grazing modality, where they emerge from the refuge and roam on an open reef surface in search of live macroalgae (Harrold & Reed, 1985 ; Kriegisch et al., 2019 ). Additionally, because sea urchins have a dispersive larval‐stage life history, kelp‐urchin dynamics can also be strongly driven by spatially explicit and episodic recruitment (Lafferty & Kushner, 2000 ; Okamoto et al., 2020 ). Kelp forests along the west coast of North America recently experienced a rapid and pronounced shift from highly expansive forests to unproductive sea urchin barrens. Starting in late 2013, a coast‐wide sea star epizootic decimated the urchin predator Pycnopodia helianthoides (hereafter, Pycnopodia ), followed by an episodic marine heatwave event that occurred from mid‐2014 into 2016 (Harvell et al., 2019 ; McPherson et al., 2021 ). Shortly after (2014–2016), large‐scale reductions in kelp biomass were recorded along the mainland coasts of California, the United States and Baja California, Mexico (Beas‐Luna et al., 2020 ), with pronounced urchin outbreaks occurring in central and northern California (McPherson et al., 2021 ; Smith et al., 2021 ). In northern California where bull kelp ( Nereocystis luetkeana ) is the dominant structure‐forming foundation species, over a 95% reduction in historical kelp biomass was documented (McPherson et al., 2021 ; Rogers‐Bennett & Catton, 2019 ). Similar large‐scale loss of kelp biomass was recorded at the southern range limit of the giant kelp ( Macrocystis pyrifera ) near Bahía Asunción, Mexico (27.1°N; Arafeh‐Dalmau et al., 2019 , Beas‐Luna et al., 2020 ). However, along the central coast of California, giant kelp‐dominated forests experienced a shift to a patchy mosaic of remnant forests interspersed with sea urchin barrens (Smith et al., 2021 ). As such, whether the observed 2014 sea urchin outbreak resulted from a behavioural shift (i.e. from passive grazing of detrital algae to active grazing on live macroalgae) or from changes in recruitment remains unresolved. In this study, we explore whether an outbreak of purple sea urchins that occurred in 2014 along the Monterey Peninsula, CA, USA was driven by a behavioural shift (i.e. emergence from refuge or redistribution following the regional extirpation of Pycnopodia and reduced productivity of kelp associated with the marine heatwave) or by a demographic response (i.e. changes in survival or recruitment). We then tracked the behaviour of sea urchins in the years following the 2014 outbreak to determine how grazer behaviour shapes alternations between kelp‐dominated (hereafter, ‘forested’) and urchin‐dominated (hereafter, ‘barren’) states. This study was motivated by the following hypotheses: (1) sea urchins emerged from refuge following the regional collapse of Pycnopodia , the 2014–2016 marine heatwave, and a decline in kelp production, (2) sea urchin behaviour (passive or active) explains patch state (forested or barren) transition dynamics and (3) sea urchin migration in search of alternative food sources is associated with macroalgae recovery.",
"discussion": "DISCUSSION This study demonstrates the important role of grazer behaviour in facilitating patch‐state transition dynamics. The kelp forest‐urchin barrens mosaic that developed following the extirpation of Pycnopodia and the marine heatwave revealed how grazer behaviour shapes alternations between kelp‐dominated and urchin‐dominated states. These findings suggest that the initial 2014 sea urchin outbreak along southern Monterey Bay, California was primarily driven by the emergence of adult sea urchins from refuge, not by a demographic response (i.e. recruitment). Behaviorally driven alternations among patch states across the mosaic further demonstrate the role of grazer behaviour in facilitating transition dynamics. In many systems, behaviour is a primary mechanism for the organisation of ecological communities (Karatayev et al., 2021 ; Lima & Zollner, 1996 ; Werner & Peacor, 2003 ). However, behavioural‐driven community patterning often results from demographic (i.e. recruitment) or density‐dependent responses of predators and their prey (Levin, 1976 ). Our study supports how multiple biotic (e.g. recruitment, loss of key predators) and environmental (e.g. marine heatwaves, grazer metabolic responses to warming events) perturbations may interact to influence behavioural switching that can facilitate persistent patterning of community states. The initial sea urchin outbreak observed in this region in 2014 is likely reflective of a shift in grazing modality (from passive to active grazing), potentially in response to several coinciding factors such as reduced food availability, increased metabolic demands from the warming event (e.g. Rasher et al. 2020 ), recruitment leading to adult behavioural switching and from a reduction in the abundance of a benthic mesopredator (Burt et al., 2018 ; Cowen, 1983 ; Harrold & Reed, 1985 ). While we did not find strong evidence of a demographic response coinciding with the 2014 sea urchin outbreak, recruitment facilitation is a known driver of alternative state dynamics (Baskett & Salomon, 2010 ; Karatayev et al., 2020 ). Sea urchin recruitment dynamics are often episodic, with considerable geographic variation (Ebert & Russell, 1988 ; Okamoto et al., 2020 ; Pearse & Hines, 1987 ). Following the initial sea urchin behavioural shift in 2014, it is possible that the formation of barren patches enhanced sea urchin recruitment to barrens within the mosaic. Another alternative hypothesis is that increased sea urchin recruitment may have led to the observed behavioural response in adults, especially after 2014. Additionally, recruitment may have occurred prior to 2012 in this system or with variable timing and magnitude at other locations along the northeastern Pacific Ocean (Okamoto et al., 2020 ; Rogers‐Bennett & Catton, 2019 ). Long‐term monitoring observations along the central coast of California, USA indicated that the 2014 sea urchin outbreak continued for at least 6 years and was potentially reinforced by recruitment after 2014. During this same period, canopy‐forming kelps to the north and south of the study region experienced unprecedented declines resulting from the marine heatwave and even more expansive outbreaks of purple sea urchins (Arafeh‐Dalmau et al., 2019 ; Beas‐Luna et al., 2020 ; McPherson et al., 2021 ; Rogers‐Bennett & Catton, 2019 ). One explanation for the persistence of remnant kelp patches in this system (as opposed to adjacent neighbouring areas) is the presence of trophically redundant predators. The urchin predator guild along the west coast of North America is comprised of six key species: sea otters ( Enhydra lutris nereis ), lobsters ( Panulirus interruptus ), sheephead ( Semicossyphus pulcher ), sunflower sea stars ( Pycnopodia helianthoides ), rock crab ( Cancer spp.) and wolf eels ( Anarrhichthys ocellatus ; Scheibling & Hamm, 1991 , Eisaguirre et al., 2020 ). The abundance of these species varies geographically along the west coast. In northern California, where rock crab and wolf eels are the only alternative predators of urchins, forests were reduced by over 95% with the loss of Pycnopodia (McPherson et al., 2021 ). However, forests in southern California that have a suite of urchin predators (e.g. lobster, sheephead, rock crab) experienced an apparent buffer from kelp decline following the demise of Pycnopodia (Eisaguirre et al., 2020 ). Finally, on the central coast of California, remnant patches of kelp forests were indirectly maintained by sea otters that target energetically profitable sea urchins in patches of forest (Smith et al., 2021 ). This spatially explicit foraging by sea otters is likely the mechanism responsible for the persistence of kelp patches within the mosaic. In this study, switching among patch states within the mosaic was explicable in part by changes in the density of exposed (i.e. active foraging) sea urchins. Behavioural switching within the mosaic across such a short temporal duration may be driven by spatial variability in drift kelp. High levels of drift kelp have been shown to facilitate reef‐scale behavioural feedback in California, Chile and New Zealand (Karatayev et al., 2021 ; Kriegisch et al., 2019 ; Ling et al., 2019 ; Vásquez & Buschmann, 1997 ). We also found evidence of strong discontinuous state shift thresholds, with at least two discontinuous thresholds required to facilitate switching among patch states. A number of studies have suggested a critical threshold of a forcing variable that drives state transitions to less productive configurations (Casini et al., 2009 ; Petraitis & Dudgeon, 2004 ). The strong forward‐ and reverse‐shift thresholds identified in this study provide an empirical demonstration of this phenomenon. Sea urchin movement from deep to shallow water may explain the isolated recovery of kelp forest patches in deep water. The dramatic reduction in medium‐ and large‐sized urchins at deep reefs, simultaneous increase of those size classes inshore and the pronounced reduction of foliose red macroalgae in shallow water all indicate that sea urchin movement is one possible explanation for the observed changes in the cover of macroalgae. Although other studies have documented sea urchin migrations between depth zones (Ling et al., 2016 ; Vadas et al., 1986 ), an alternative explanation in this system is that sea urchins occupying the deep reefs switched to a passive‐grazing modality and those in the shallow zone emerged from the refuge. However, because there was not a reduction of macroalgae in the shallow zone prior to the increase in the density of medium‐ and large‐sized urchins, the movement hypothesis (as opposed to behavioural switching) remains the most parsimonious explanation for observed recovery dynamics. At the locations where kelp recovery was observed, it is important to note the kelp species that repatriated the once barren grounds was the bull kelp ( Nereocystis luetkeana , a predominately annual species), not the giant kelp ( Macrocystis pyrifera , a perennial species). Prior to the 2014 sea urchin outbreak, kelp forests along the Monterey Peninsula were dominated by the giant kelp (Foster & Schiel, 1985 ; Graham et al., 1997 ). It is well established that shading by giant kelp limits algal recruitment and the growth of other non‐calcareous species (Kennelly, 1989 ; Reed & Foster, 1984 ). The removal of long‐standing giant kelp forests by purple sea urchin grazing may have released Nereocystis from light limitation, thereby enabling the rapid recolonisation and growth of Nereocystis following sea urchin movement inshore to shallow water. The results presented in this study highlight the role of behaviorally mediated effects in structuring ecological communities. One of the most unusual aspects of this system is the ability of sea urchins to persist in low‐resource environments for extended periods of time (Ebert, 1967 ; Ebert, 1982 ; Smith & Garcia, 2021 ), which may contribute to the longevity of the alternative barrens state of the ecosystem. Therefore, the behaviour of grazers, especially ecosystem engineers, is fundamental to community and ecosystem dynamics."
} | 3,926 |
29748514 | PMC5983801 | pmc | 5,044 | {
"abstract": "The use of natural products (NPs) as possible alternative biocidal compounds for use in antifouling coatings has been the focus of research over the past decades. Despite the importance of this field, the efficacy of a given NP against biofilm (mainly bacteria and diatoms) formation is tested with the NP being in solution, while almost no studies test the effect of an NP once incorporated into a coating system. The development of a novel bioassay to assess the activity of NP-containing and biocide-containing coatings against marine biofilm formation has been achieved using a high-throughput microplate reader and highly sensitive confocal laser scanning microscopy (CLSM), as well as nucleic acid staining. Juglone, an isolated NP that has previously shown efficacy against bacterial attachment, was incorporated into a simple coating matrix. Biofilm formation over 48 h was assessed and compared against coatings containing the NP and the commonly used booster biocide, cuprous oxide. Leaching of the NP from the coating was quantified at two time points, 24 h and 48 h, showing evidence of both juglone and cuprous oxide being released. Results from the microplate reader showed that the NP coatings exhibited antifouling efficacy, significantly inhibiting biofilm formation when compared to the control coatings, while NP coatings and the cuprous oxide coatings performed equally well. CLSM results and COMSTAT analysis on biofilm 3D morphology showed comparable results when the NP coatings were tested against the controls, with higher biofilm biovolume and maximum thickness being found on the controls. This new method proved to be repeatable and insightful and we believe it is applicable in antifouling and other numerous applications where interactions between biofilm formation and surfaces is of interest.",
"conclusion": "4. Conclusions The use of the fast-tracking plate reader technology combined with nucleic acid staining to assess biofilm formation on coatings has been successful, reproducible, and confirmed by CLSM. CLSM is highly specialised and time consuming for 3D reconstructions (especially if this level of detail is not required), therefore epifluorescence microscopy can be used as a complementary method in future experiments. Model coatings using inert polymer binders were designed containing the NP juglone and the biocide Cu 2 O and compared against controls over 48 h. Coatings containing the NP juglone and the biocide Cu 2 O inhibited biofilm formation after 48 h, when compared to the control coatings that illustrated thick biofilm formation. Juglone has previously shown inhibition of bacterial attachment in short-term experiments [ 15 ]. However, this is the first time that this NP was incorporated into a coating system and biofilm formation was assessed. A successful AF technology against biofouling should be able to last for more than five years in order to be economically viable. The main aim is to test whether an AF technology is effective against initial biofilm formation, which takes place within minutes to hours of immersion (for any submerged surface) in the sea, and as such, the methodology developed in the current study illustrated good AF efficacy within the desired time scales. The experimental times could be extended if this is required. AF technologies are now focusing on no-biocide-release coatings due to acknowledged problems with legislative approvals, shifting attention towards foul release coatings and/or nonleaching (“tethered”) biocide coatings [ 33 ], as well as topographically modified surfaces [ 34 ]. For such technologies, AF activity is aimed to have an immediate response towards fouling. Therefore, testing for initial attachment at short time scales would be appropriate. Thus, the methodology developed in the current study would be ideal for assessing initial biofilm processes, such as bacterial initial attachment and biofilm formation, making this method comparable to realistic scenarios.",
"introduction": "1. Introduction Biofouling, the accumulation of marine organisms on a surface, severely affects manmade structures such as ship hulls, energy systems, environmental sensors, aquaculture settings, etc. A number of technological advances focusing mainly on coating these structures with antifouling agents, have been utilised over the last few decades, with the use of biocides being the most successful. However, the biocide with the best antifouling (AF) efficacy against a wide range of target species, known as tributyltin (TBT), has been found to be detrimental to the marine environment [ 1 , 2 ]. Since TBT’s complete ban in 2008, there has been increased effort to discover new compounds that would perform equally well but with no environmental impact. Screening of natural products (NPs) as potential AF agents is an ongoing interest within the scientific community and industry [ 3 , 4 , 5 , 6 ]. An increasingly important target within the AF community is the inhibition of biofilm growth on manmade surfaces. Biofilms, which are composed mainly of bacteria and diatoms embedded within extracellular polymeric substances (EPS), are the first type of biofouling to colonise a surface, while a growing body of studies supports the position that biofilms attract larger organisms to settle upon them [ 7 ]. Biofilms are tremendously costly to society, as they adversely affect numerous industries ranging from petrochemicals to health care. Currently there are several high-throughput methods that allow the simultaneous evaluation/screening of new agents for their AF efficacy at a range of concentrations and against various marine target species [ 8 , 9 ]. These laboratory methods are necessary to determine whether a compound may be considered for further evaluation. However, there is a lack of studies for determining their effect once incorporated into a coating system. Stafslien et al. [ 10 , 11 ] have developed a rapid and high-throughput laboratory-based coating testing system, which, however, requires highly specified equipment that is uniquely available to their laboratory. The first objective of the current work was to develop a bioassay for the rapid in situ assessment of biofilm formation (model marine species: Cobetia marina ) on coated surfaces using universally available equipment. This was achieved by utilising 24-well plates combined with nucleic acid staining, high-throughput plate reader technology, and confocal laser scanning microscopy (CLSM). Microplate readers have the capacity to rapidly screen bacterial responses (e.g., via spectrophometric and fluorescence measurements) under different conditions and have been routinely used for the discovery of novel antimicrobials. Furthermore, CLSM allows three-dimensional (3D) visualisation of the biofilm, providing new insights into biofilm development. A few recent studies have reported the use of plate readers (microtiter plate assays) in conjunction with biological stains to measure biofilm inhibition of adhesion and/or growth within an antifouling context [ 12 , 13 , 14 , 15 , 16 ], although these studies were not designed to perform tests on antibiofilm coatings. The second objective of the current work was to use this novel bioassay for the evaluation and screening of a terrestrial NP juglone (5-hydroxy-1,4-naphthalenedione), derived from walnut trees, for potential use as a natural AF agent within a coating system. This investigation explored the proof of concept for a rapid assessment method (48 h in this case) that has the potential to be used for a broad range of antibiofilm studies, where the investigation of bacterial/surface interactions is of primary importance (i.e., antifouling industry, medical, and dental).",
"discussion": "2. Results and Discussion 2.1. Determination of Juglone Leaching Juglone was first dissolved in artificial seawater (ASW) in order to establish the wavelength at which the compound has maximum absorption via UV-Vis, which was found to be at 520 nm. A calibration curve was generated at this wavelength. To establish the extent of juglone leaching from the two juglone-containing coatings, tests were performed over 48 h. This was done in order to determine whether juglone actually leached from the coated surfaces, and if so, to quantify the concentrations within the 24-well fluid volumes. During the leaching tests, two samples were taken at 24 h and 48 h. Overall, juglone leaching was detected from both coating formulations at concentrations ranging from 1.70 ppm to 1.99 ppm after 24 h, and 1.96 ppm to 2.35 ppm after 48 h, for the coating containing juglone (JUG)+poly(methyl methacrylate) (PMMA)+Rosin (ROS) and the coating JUG+PMMA, respectively (see Table 1 ). Overall, most of the juglone had leached out at 24 h for both coatings. PMMA is a common acrylic polymer material also used in orthopaedic surgery as part of bone cement. Antibiotics are often incorporated into these cements, where subsequent elution produces high local concentrations of antibiotics while simultaneously minimising systemic toxicity [ 17 , 18 , 19 ]. A major characteristic of currently available antibiotic cements is that an initial burst of antibiotic elution occurs within 24 h to 48 h, with poor subsequent sustained release [ 20 , 21 , 22 , 23 ]. Similarly, in the current study, a burst elution when NPs are incorporated into PMMA cannot be excluded. The 24–48 h leaching window observed for antibiotics would be ideal for AF bioassays, as biofilm growth can easily be supported within that time scale. Biofilm-mediated release rates of biocides have been previously proposed, as the natural bacterial extrapolymeric matrix has shown to bind metals such as copper [ 24 ] and potentially affect the diffusion rate of the compound and consequently control its release rate [ 25 ]. 2.2. MIC, EC 50 , and MBC for Juglone in Solution The growth-inhibiting activity of juglone was examined against planktonic growth of the model species Cobetia marina (ATCC 25374). C. marina has previously been used as model species for marine biofilm related studies [ 7 , 13 , 15 , 26 , 27 ]. It was found to have strong inhibition against C. marina at 16 ppm, while the EC 50 was found to be at 5 ppm, Figure 1 . This is in good agreement with previous work where bacterial attachment for the same species was assessed [ 15 ]. The minimum bactericidal concentration (MBC) for C. marina was found to be at 16 ppm. The measured concentration of released juglone, as shown in the previous section, falls within the same range as the concentration for the EC 50 for this species. 2.3. Assessing Biofilm Growth on Antibiofilm Coatings Using the Microplate Reader The in situ evaluation of biofilm growth on coated surfaces was achieved through the use of a plate reader utilising nucleic acid staining. In Figure 2 , the effect of each coating, i.e., 2 NP, 2 biocidal, and 2 controls, on biofilm attachment and growth is shown (detailed coating formulations can be seen in Materials and Methods, Overall, from the plate reader coating scans (live cells stained with Syto9) ( Figure 2 a), a clear difference in biofilm growth is apparent for juglone- and Cu 2 O-containing coatings when compared to their controls, PMMA and PMMA + Rosin which acquired significantly higher growth ( p < 0.004). For the “dead” (propidium iodide stained cells) wavelength measurements (“dead” = cells with compromised membranes), there was no significant difference between coatings and there was an overall low number of compromised cells for the duration of the experiment. Specifically, coatings with the binder system PMMA + Rosin showed significantly higher biofilm formation when compared to the NP coating (juglones) and the biocidal coating (Cu 2 O) with p < 0.042 and p < 0.003, respectively ( Figure 2 b,c). For the coatings containing PMMA as the sole binder (i.e., rosin was not incorporated), no significant difference in biofilm formation was observed when comparing them against the ones containing juglone and Cu 2 O (i.e., the JUG PMMA vs. Cu 2 O PMMA vs. PMMA). This indicates that PMMA alone as a binder did not facilitate sufficient compound release and therefore did not perform as well as the binder with the rosin additive. The leaching tests ( Table 1 ) have demonstrated that when PMMA was used as the sole binder, most of the juglone was found to be released within the first 24 h. On the other hand, the combination of PMMA + Rosin demonstrated a more controlled release with the juglone concentration accumulating throughout the test duration, i.e., reaching higher values at 48 h. The rosin addition may have allowed greater availability of the compound at the coated surface throughout the 48 h, potentially resulting in a better antibiofilm performance. When the juglone coatings were compared against the Cu 2 O coatings, no significant difference in biofilm growth was found ( Figure 2 b,c). In terms of dead cell wavelengths, there appear to be differences between coatings, however, there are no clear trends. Therefore, it can be concluded that the juglone-containing coatings performed similarly to those containing the biocide (Cu 2 O), indicating a clear antibiofilm activity against C. marina . 2.4. Planktonic Growth The planktonic growth measured at the 48 h end-point within the wells containing the coated coupons can be seen in Figure 3 . Interestingly, differences were also found in planktonic growth within the wells containing the different coatings. Planktonic growth was found to be significantly lower in the wells where coatings contained juglone ( p = 0.039) and Cu 2 O ( p = 0.044 and p = 0.019) when compared to the controls (PMMA ROS and PMMA), with the exception of the Juglone–PMMA combination. As shown earlier, the rosin additive performed better in releasing juglone into the solution, therefore its higher availability could account for lower bacterial growth in these wells. 2.5. Biofilm Morphology Data analysis from the confocal laser scanning microscopy (CLSM) using COMSTAT was undertaken in order to: ( i ) confirm the plate reader results and allow direct comparisons between the two techniques and ( ii ) to utilise the high resolution of the CLSM to assess the biofilm microstructures that formed on the various coatings. Figure 4 , Figure 5 and Figure 6 show examples of CLSM images obtained from 3D stacks associated with the NP, biocidal, and control coatings, respectively, and Table 2 summarises all obtained results from COMSTAT. The NP coatings (juglone) are characterised by single and/or chains of cells with an apparent lack of EPS, Figure 4 . Conversely, on the control coatings (PMMA + Rosin and PMMA), clear biofilm structures can be seen, especially on the PMMA + Rosin, where a thick and relatively homogenous biofilm is formed, while on the PMMA control, evidence of clustered cells and EPS is also apparent, however to a lesser extent (when compared to PMMA + Rosin) (see Figure 6 ). For the biocidal coatings (Cu 2 O), no biofilm formation can be seen and the cell morphology appears to be different, with maybe a few rod-like cells and more undefined shaped particles, leading to doubts as to whether these bacterial cells are intact ( Figure 5 ). Specifically, there was significantly higher biofilm thickness ( Figure 7 ) and biovolume ( p < 0.001 for both parameters) on the control coatings PMMA + Rosin when compared to the juglone and Cu 2 O coatings. These results are in good agreement with the data obtained from the plate reader ( Table 2 ). The same significant effect ( p < 0.002) of the PMMA controls on biofilm morphology was also observed for the JUG PMMA vs. Cu 2 O PMMA vs. PMMA (control), i.e., greater biofilm thickness ( Figure 7 ) and biovolume ( p < 0.001 and p < 0.002, respectively) were found on the PMMA control coatings ( Table 2 ). This effect was not clearly evident from the plate reader data, although a trend could be observed. In the current work, some level of autofluorescence was observed when fluorescence measurements were taken prior to bacterial addition. However, this was accounted for, as blanks were part of the experimental design, allowing for the subtraction of background signals. Indeed, blanks are crucial in this type of experiment, as polymer-based coatings, for instance, can often autofluoresce. Here, we showed that this can be accounted for, although care must be taken to avoid any false readings due to lack of appropriate blanks in the experimental matrix. As observed from the plate reader measurements, the CLSM data also confirmed the almost entire absence of dead cells. This implies a nontoxic effect of the coatings on biofilms, and therefore, only inhibition of attachment and growth. However, the almost complete lack of dead cells still remains to be fully understood. For instance, as described earlier, there were lower OD values accounting for the planktonic cells in the juglone and biocidal coating wells when compared to the controls implying cell death. In future experiments, planktonic cells could also be stained to assess their viability. The overall toxicity mechanism of naphthoquinones, such as juglone, has still not been clearly established, especially towards prokaryotes. However, the potential interference of juglone (which is a strong redox cycler with high potential to react with oxygen and its reactive species) on cell division and membrane transport [ 29 , 30 ] may account for the overall inhibition of biofilm growth on juglone-coated surfaces seen in the current work. The use of CLSM was found to be informative on various additional biofilm characterisation parameters corroborating/adding to the less sensitive but high-throughput analysis by the plate reader. Overall, there was good consistency between the two techniques, underlining a good antibiofilm performance of the juglone-containing coatings. The few discrepancies between the two datasets may be attributed to the nature of the measurement methods, i.e., CLSM uses laser light which provides better controllability since emitted wavelengths can usually be optimised to limit overlapping of the coatings’ potential autofluorescence and the nucleic acid stains."
} | 4,553 |
35544567 | PMC9094664 | pmc | 5,045 | {
"abstract": "Although proteins are considered as nonconductors that transfer electrons only up to 1 to 2 nanometers via tunneling, Geobacter sulfurreducens transports respiratory electrons over micrometers, to insoluble acceptors or syntrophic partner cells, via nanowires composed of polymerized cytochrome OmcS. However, the mechanism enabling this long-range conduction is unclear. Here, we demonstrate that individual nanowires exhibit theoretically predicted hopping conductance, at rate (>10 10 s −1 ) comparable to synthetic molecular wires, with negligible carrier loss over micrometers. Unexpectedly, nanowires show a 300-fold increase in their intrinsic conductance upon cooling, which vanishes upon deuteration. Computations show that cooling causes a massive rearrangement of hydrogen bonding networks in nanowires. Cooling makes hemes more planar, as revealed by Raman spectroscopy and simulations, and lowers their reduction potential. We find that the protein surrounding the hemes acts as a temperature-sensitive switch that controls charge transport by sensing environmental perturbations. Rational engineering of heme environments could enable systematic tuning of extracellular respiration.",
"introduction": "INTRODUCTION Electron flow between cofactors in proteins is central to many life processes such as respiration, photosynthesis, and nitrogen fixation ( 1 ). The protein architecture between cofactors has long been considered a passive tunneling bridge resulting in transport that decays exponentially with distance, thus limiting transport to 1 to 2 nm ( 1 ). Some proteins can conduct over distances up to 20 nm ( 2 ). Long-distance (>20 nm) charge transport is usually achieved via a redox potential gradient generated by distinct redox enzymes, which provides an energetic driving force for the transport of charge toward the terminal acceptor ( 3 ). However, the common soil bacterium Geobacter sulfurreducens transports electrons extracellularly over micrometer distances via “microbial nanowires.” These micrometer-long filaments are homopolymers of cytochromes OmcS ( Fig. 1 ) ( 4 , 5 ) or OmcZ ( 6 ), which contrasts with multiple distinct cytochromes used by other bacteria for extracellular electron transfer up to nanometers ( 7 ). This extracellular electron transfer is important in global biogeochemical cycling of carbon, nutrients, and metals as well as in regulating the bacterial release of methane to the atmosphere ( 7 ). Fig. 1. Closely stacked hemes in OmcS nanowires provide a continuous path for extracellular electron transport over micrometers. ( A ) Transmission electron microscopy image showing a G. sulfurreducens producing OmcS nanowires. Scale bar, 200 nm. Image levels were adjusted selectively in different areas to show the most information. ( B ) Each pair of parallel-stacked hemes is perpendicular to the next pair in a T-shaped geometry. ( C ) Upon polymerization, the hemes align to stack along the entire length of the OmcS filament with a 20-nm helical pitch. Each OmcS monomer and corresponding hemes are highlighted in different colors. The atomic structure of OmcS nanowires revealed seamless stacking of heme cofactors, which provide a continuous path for electron flow ( Fig. 1, B and C ) ( 4 , 5 ). Prior theoretical studies of multiheme cytochromes have assumed that charges hop between heme cofactors ( 8 ), with the heme-to-heme hopping rates determined by the reduction potentials of the individual hemes and the electronic coupling between each pair. However, each protomer in OmcS must have a similar midpoint redox potential. Therefore, while there may exist a redox potential gradient within a protomer, there remain questions regarding how G. sulfurreducens generates high enough current through such a structure to meet its respiratory needs. Prior computational studies on OmcS nanowires suggested multiple mechanisms such as hopping ( 9 ), quantum transport ( 10 ), and coherence-assisted transport ( 11 ) to explain the experimentally measured conductivity ( 4 ). However, experimental evidence supporting a particular mechanism is lacking. In this work, we measure the intrinsic (contact-free) conductivity of individual OmcS nanowires as a function of nanowire length, voltage, temperature, and pH, which reveal an unexpectedly high hopping rate (>10 10 s −1 ), comparable to synthetic molecular wires, with negligible carrier loss over micrometers. We also find a 300-fold increase in nanowire conductance upon cooling, which vanishes upon deuteration. Computations show that cooling causes a massive rearrangement of hydrogen (H) bonding networks, making hemes more planar, thus lowering their reduction potential and increasing conductivity.",
"discussion": "DISCUSSION Electron transfer via c-type cytochrome plays a central role in many chemical and biological processes. However, cytochromes were long-considered monomeric, limiting transfer to over a few nanometers. Here, we show that common soil bacteria Geobacter can transport electrons over micrometers via polymerized cytochrome nanowires with negligible electron loss. Combining theory with measurements of intrinsic, contact-free, conductivity measurements of individual nanowires as a function of nanowire length, voltage, and temperature, we find that a hopping mechanism explains the unexpected length and temperature dependence of nanowire conductivity. This was previously reported only in synthetic nanowires and nonredox proteins where it was limited to distances less than 20 nm ( 2 , 16 ). Alternative models based on quantum transport have introduced the possibility that such a mechanism has a role in OmcS ( 10 , 11 ). To date, all studies of quantum transport in OmcS have not incorporated the effect of the protein environment. Such a heterogeneous environment with thermal fluctuations would greatly affect the stability of delocalized electronic states. Our calculations suggest that the protein has an important role in curating unique environments for each heme. Further studies are required to evaluate quantum transport in OmcS nanowires. Protein environment controls the reduction potentials of c-type hemes ( 43 ). We demonstrate that this protein control of reduction potentials allows nanowire conductivity to be tuned over a remarkable range. Thus, the OmcS nanowires are a previously unknown class of metalloproteins with highly tunable electronic properties. Our studies can help guide systematically fine-tuning the conductivity of nanowires by modulating protein-induced heme distortions. Using a suite of complementary experiments to measure the conductivity of individual nanowires as a function of molecular length and temperature, and combining with computational studies using MD and QM/MM electronic structure calculations, here we show that the thermally activated vibrations induce out-of-plane heme distortions that modulate activation energy to enhance conductivity by 300-fold. We thus present a strategy to induce cooperative and large-scale conformational changes that modulate heme distortions, which were previously thought to have only local influence. Our findings have consequences for the physiology of G. sulfurreducens ( 49 ). OmcS is essential for the reduction of iron oxides, an important electron acceptor in the native environment of G. sulfurreducens , as well during the early growth stages of electricity-producing biofilms in bioelectrochemical systems ( 49 ). In addition, artificially expressing cytochrome OmcS in photosynthetic cyanobacteria increased catalytic performance in a diversity of processes such as an increase in photocurrent by 9-fold ( 50 ), increased nitrogen fixation by 13-fold ( 51 ), and improved photosynthesis by increasing 60% biomass ( 52 ) compared to the wild-type cyanobacteria. These studies highlight the important role of OmcS in light-driven biocatalysis. Therefore, improving the conductivity of OmcS could help to control bacterial physiology and ecology as well as improve biocatalysis efficiency. The increased nanowire conductivity upon cooling could also help bacteria to compensate for the loss of metabolic rate at cold environments. Furthermore, protein structure and heme geometries can significantly differ after binding to substrates ( 43 ) such as minerals or anodes of microbial fuel cells, potentially enhancing the coupling between the nanowires and their electron acceptors. Our discovery also has implications in the rational design of protein nanowires ( 18 , 53 – 55 ). Our mechanistic studies of OmcS nanowires provide a framework for rational design of nanowire structures with tunable conductivity by modulating the redox potential of hemes and heme distortions by leveraging changes in the H-bonding."
} | 2,185 |
36679027 | PMC9864307 | pmc | 5,046 | {
"abstract": "Arbuscular mycorrhizal fungi (AMF) form mutualistic symbiotic relationships with many land plants and play a key role in nitrogen (N) acquisition. NO 3 − -N and NH 4 + -N are the main sources of soil mineral N, but how extraradical mycelial transfer affects the different N forms and levels available to tomato plants is not clear. In the present study, we set up hyphal compartments (HCs) to study the efficiency of N transfer from the extramycelium to tomato plants treated with different N forms and levels of fertilization. Labeled 15 NO 3 − -N or 15 NH 4 + -N was placed in hyphal compartments under high and low N application levels. 15 N accumulation in shoots and the expression of LeNRT2.3 , LeAMT1.1 , and LeAMT1.2 in the roots of tomato were measured. According to our results, both 15 NO 3 − -N and 15 NH 4 + -N were transported via extraradical mycelia to the shoots of plants. 15 N accumulation in shoots was similar, regardless of the N form, while a higher 15 N concentration was found in shoots with low N application. Compared with the control, inoculation with AMF significantly increased the expression of LeAMT1.1 under high N and LeNRT2.3 under low N. The expression of LeAMT1.1 under high N was significantly increased when NO 3 — N was added, while the expression of LeNRT2.3 was significantly increased when NH 4 + -N was added under low N. Taken together, our results suggest that the N transfer by extraradical mycelia is crucial for the acquisition of both NO 3 − -N and NH 4 + -N by the tomato plant; however, partial N accumulation in plant tissue is more important with N deficiency compared with a higher N supply. The expression of N transporters was influenced by both the form and level of N supply.",
"introduction": "1. Introduction Arbuscular mycorrhizal fungi (AMF) can form symbiotic relationships with more than 80% of land plants and play a key role in their nutrition [ 1 , 2 ]. After symbiosis is established, mycorrhizal roots with a “mycorrhizal nutrient absorption pathway” improve the mineral nutrient content of plants via a hyphal network known as the extramycelium (ERM), which is an extension of the plant root system [ 3 , 4 ]. Plants promote this symbiosis, as they are commonly limited by one of the two major nutrients, phosphorus (P) and nitrogen (N) [ 5 ]. Under such conditions, the mycorrhizal roots have two means of nutrient absorption: the plant pathway and the mycorrhizal pathway. The plant pathway involves the absorption of nutrients through high- or low-affinity absorption transporters in the epidermis or root hairs. For nutrients with low mobility in the soil, absorption through the plant pathway is usually limited by their depletion in the zone around the root. In contrast, the mycorrhizal pathway involves high-affinity nutrient transporters in the ERM, which take up nutrients and transport them along the hyphae to the rhizosphere hyphae (IRM) in the root cortex [ 1 ]. As one of the most important macronutrients, N accounts for 1–5% of the dry weight of plants. Over the last two decades, it has also been recognized that AMF plays a crucial role in the uptake by plants [ 6 ], while the soil nitrogen level is one of the factors affecting the inoculation effect of AMF [ 7 ]. The absorption of plant N transporters is induced by mycorrhization [ 1 ]. Isotopic labeling with 15 N directly demonstrates that AMF hyphae can absorb and transport mineral N from the soil to their host plants [ 8 ]. As AMF can obtain enriched N sources, N transfer from hyphae to hosts may be huge [ 9 ]. However, compared with root absorption, AMF-derived N alone may be limited, as the potential N absorption and transport rates of mycelia are only higher than those of roots with low N (both NO 3 − or NH 4 + ) content in their soil [ 10 ]. Both the amounts and forms of N in the cultivation medium can affect the mycorrhizal infection rate, the amount of N transported by AMF to host plants, and mycelial density [ 11 ]. In most soil environments, the main form of mineral N is NO 3 − ; however, in wetlands or highly acidic soils, NH 4 + is dominant [ 12 ]. Although both forms of N (NO 3 − or NH 4 + ) can be absorbed by the external hyphae of AMF ( Rhizophagus intraradices ) and transported to the host plant [ 13 , 14 , 15 , 16 ], the hyphae preferentially absorb NH 4 + [ 14 , 17 , 18 ]. The amount of N transferred to plants is high following an NH 4 + fertilizer application [ 19 ]. However, applying only NH 4 + reduces the activity of mycorrhiza compared with applying only NO 3 − [ 20 , 21 ]. Mycorrhizal formation can directly affect the process of plant nutrient absorption and metabolism; further, it can make the growth and development of plants more advantageous than non-mycorrhizal plants and improve crop yield and fruit quality [ 22 ]. The tomato plant ( Lycopersicum esculentum L.) is the second most important vegetable crop worldwide; however, the effect of AMF on N uptake by tomato plants in relation to N availability and forms has yet to be identified. At the molecular level, both the NH 4 + and NO 3 − transporters of hosts are regulated in AMF symbiotic plants [ 23 , 24 ]. In tomatoes, the NH 4 + transporters of LeAMT1.1 and LeAMT1.2 are expressed in root hairs and leaves under N-deficient conditions, while under hydroponic growing conditions, the transcript level of LeAMT1.2 in the roots increases after NO 3 − or NH 4 + supplementation, whereas that of LeAMT1.1 is induced by N deficiency [ 25 ]. AMF excrete NH 4 + to levels that can be sensed by tomato roots, and this is consistent with the induced expression of LeAMT1.2 by as little as ≥1 µM external NH 4 + with root-associated N2-fixing bacteria [ 26 ]. Rice plants colonized by R. irregularis strongly induce the expression of an NH 4 + transporter ( OsAMT3.1 ) in roots under both low and high rates of N application. As AMF increases NH 4 + , AMT expression could be changed due to colonization. However, as mycorrhizal-inducing N transporters are up-regulated, the expression of nitrate transporter genes changes in host plants, thus changing the ability of plants to obtain NO 3 − [ 24 , 27 , 28 , 29 ]. Hildebrandt et al. (2002) found that inoculation with R. irregularis up-regulated LeNRT2 in tomato roots [ 30 ]; in particular, LeNRT2.3 is related to mycorrhization and is abundantly expressed in root cells containing AMF structures, such as plexus branches and vesicles [ 30 ]. These results indicate that AMF colonization positively affects nitrate uptake from the soil and nitrate allocation to the plant partner, probably preferentially mediated by LeNRT2.3 . So, LeNRT2.3 functions as a low-affinity transporter, whose activity allows higher N-use efficiency in tomatoes [ 31 ]; therefore, AMF colonization positively affects nitrate uptake from the soil and nitrate allocation to the plant partner, probably preferentially mediated by LeNRT2.3 [ 30 ]. How different N levels available and N forms in the mycorrhizal symbiosis system induce the expression of these nitrogen transporters is not fully understood. We hypothesized that N transport via AMF hyphae and LeNRT2.3 , LeAMT1.1, and 1.2 expression might be correlated with N status and N forms in the hyphosphere. In the present study, hyphal compartments were used to explore the effects of two N levels and forms on mycorrhizal tomato plants. The colonization rates, plant nutrition, growth status, and LeNRT2.3 , LeAMT1.1 , and 1.2 expression were monitored.",
"discussion": "4. Discussion 4.1. Nitrogen Transport and Acquisition via AMF with N Levels and Forms in HCs Nitrogen acquisition in plant tissues was significantly correlated with N fertilizer application levels and AMF inoculation under conditions of low N application ( Table 2 ). The concentrations of 15 N binding were from 0.074‰ to 0.138‰ in the shoot tissues of all mycorrhizal plants; with high N application, 15 N binding (0.078‰) was lower than that with low N application (0.118‰) ( Table 1 ). However, the total 15 N transported via the extramycelium to shoot parts showed no significant difference between N levels, even with 20.4 µg per plant with high levels of N, compared with 16.3 µg at low levels of N ( Table 1 ). Under high N application, there were no differences in either the N concentration or N content between mycorrhizal and non-inoculated plants ( Table 2 ). In contrast, with low N application, the N concentration was increased by mycorrhization by 17.7% (from 1.24% to 1.46%), almost to the same level as plants with high N application. The N uptake by plants was 22.5%, increasing from 114.4 mg to 140.1 mg per plant, owing to the double effects of biomass and concentration. Although 15 N binding was not significantly different between NH 4 + and NO 3 − applications in HCs, the actual 15 N transfer was 14.2 µg per plant with NO 3 − application and 18.4 µg with NH 4 + with low N application ( Table 3 ). This difference implies that more 15 NH 4 + was transported from HCs to host plants via the hyphae compared with NO 3 − ( Table 1 ). This difference had no further effects on biomass accumulation in tomato plants; however, the biomass was increased when NO 3 − was added to HCs ( Table 3 ). These results suggest that almost the same amount of N transfer via MP, in the case of NO 3 − , had a greater influence on biomass accumulation as compared to that of NH 4 + supplied to the extramycelium. With NO 3 − in HCs, P uptake was significantly increased by 11.3% as the result of higher biomass at the low-N fertilization level ( Table 3 ). It is reported that AMF contributes substantially to the N nutrition of their host plants [ 6 ]. Hyphae can directly and effectively utilize inorganic compounds and transfer a large amount of N to the roots of host plants [ 19 , 39 ]. The direct labeling of 15 N has shown the flux of N through AM fungal hyphae to plants ( Andropogon gerardii ) [ 40 ]. In the present study, no differences in N uptake were shown with the two levels of N application in the HCs; however, 15 N binding was higher with low N application than with high N application. This demonstrated that N transfer from the fungus to the host plant was similar at high and low levels of N application; however, with a lower N application, the HP becomes more important than with higher N rates. These results indicate that mycorrhization plays a substantial role in the absorption of plants regardless of N availability ( Table 2 and Table 3 ). Under lower N availability, the mycorrhizal pathway becomes more important compared with the root pathway. A similar result has been reported, showing that high amounts of N application can significantly decrease N uptake by mycorrhizal plants from the soil [ 15 ]. When nutrients were insufficient, the advantages of mycorrhizal symbionts were reflected because the nutrients absorbed by plant roots were insufficient to support normal growth, while sufficient nutrients often inhibited the infection of fungi in the root system of host plants [ 5 ]. In summary, it may be concluded that a substantial amount of N can be adsorbed and transported from fungi to their host plants, and only the N uptake by hyphae, i.e., the hyphae pathway related to the root pathway, is influenced by the interaction of the N nutritional status in the environment of both the roots and fungi. This agrees with the previous hypothesis that the hyphae of AMF may absorb NH 4 + preferentially over NO 3 − , but that the export of N from the hyphae to the roots and shoots may depend on the amount of N supplied/available for uptake [ 41 ]. However, in the present study, increased growth was not accompanied by greater concentrations of N and P in the shoots of plants. Taking biomass into account, the total content of P in shoots was increased. 4.2. Transporter Genes LeAMT1.1, LeAMT1.2, and LeNRT2.3 Were Regulated by Inoculation with AMF in the Root Tissue of Tomato Plants As previously reported, the expression of the encoded LeNRT2.3 protein is related to AMF colonization [ 30 ]. In our study, the expression of LeNRT2.3 in roots was significantly increased following inoculation with AMF compared with the control plants at low N levels ( Table 4 ). Although it is a low-affinity transporter, a difference in expression was not detected between the two N levels ( Table 4 ). LeNRT2.3 expression was not correlated with the N form with high N application but had a significantly higher expression level with NH4+ compared with that of NO 3 − with N deficiency ( Table 4 ). As N is a major limiting factor for plant growth and yield, genes affect plant growth through nitrate uptake or remobilization. The higher expression levels of LeNRT2.3 in flowers and leaves indicate that LeNRT2.3 plays a pivotal role in shoot development [ 31 ]. LeNRT2.3 is also suggested to play a key role in the xylem transport of N from roots to shoots and in N uptake by roots [ 31 ]. Taken together, the expression of LeNRT2.3 driven by symbiosis could be important for N-use efficiency in tomatoes, and its induced expression indicates a higher N-use efficiency in tomatoes [ 42 ]. The expression of only LeNRT2 . 3 among the transporters assayed was higher in AMF-colonized tomato roots than in non-colonized controls. AMF colonization caused the significant expression of a nitrate reductase gene of G. intraradices . The results may mean that AMF colonization positively affected nitrate uptake from the soil and nitrate allocation to the plant partner, probably preferentially mediated by LeNRT2 . 3 . In addition, part of the nitrate taken up is reduced by the fungal partner itself and, if in excess, may then be transferred as glutamine to the symbiotic plant partner [ 30 ]. The expression of LeNRT2 . 3 is negatively controlled by ammonium but, remarkably, not by glutamine [ 30 ]. The specific expression of these up-regulated AMT genes in arbuscule-colonized cortical root cells has been shown in M. truncatula [ 29 ], L. japonicus [ 28 ], G. max [ 24 ], and S. bicolor [ 43 ]. In the present study, regarding the two important high-affinity NH 4 + transporters in roots, LeAMT1.1 was up-regulated by inoculation with AMF, especially with NO 3 − feed in HCs with high N application, while there were no significant differences in LeAMT1.2 between treatments ( Table 4 ). Other research work has reported strong inductions of LeAMT1.1 and LeAMT1.2 gene expression in mycorrhizal roots, evidence that host plants had NH 4 + transporters that were up-regulated under AMF colonization, with the specific expression of the up-regulated AMTs genes in arbuscule-colonized cortical root cells shown in M. truncatula [ 29 ], L. japonicus [ 28 ], G. max [ 24 ], and S. bicolor [ 43 ]. In particular, AMF symbiosis down-regulated OsAMT1.1 expression under low-N conditions (1.825 mM NO 3 − ) but not under high-N (7.5 mM NO 3 − ) conditions [ 44 ]. In the present study, LeAMT1.1 was significantly increased by inoculation with AMF and high N application and particularly up-regulated by the addition of NO 3 − in HCs ( Table 4 ). LeAMT1.1 and LeAMT1.2 are differentially regulated by N and contribute to root-hair-mediated NH 4 + acquisition from the rhizosphere; the transcript levels of LeAMT1.2 increased after NH 4 + or NO3± application, while LeAMT1.1 was induced by N deficiency [ 45 ]. LeAMT1.2 , an important high-affinity NH 4 + transporter, was reported to have contrasting responses to LeAMT1.1 and was induced by N application [ 45 ]. By contrast, in the present study, the expression of LeAMT1.2 was affected by neither mycorrhization nor the N level or form ( Table 4 ). LeAMT1.2 mRNA levels are controlled in a concentration-dependent manner by the NH 4 + concentration or the plant N status, and peak expression occurs at 2–5 µM NH 4 + , with a further increase in NH 4 + causing a reduction [ 26 ]. In our previous study, the expression of LeAMT1.2 was significantly induced by AMF inoculation in an isolation-specific manner [ 46 ]. As N is a major factor determining plant growth and yield, it likely influences plant growth by modulating N uptake rates or remobilization activity [ 31 ]. The induction of N transporters varied with the level of N application and the N form in HCs; however, their increasing expression indicated a higher N-use efficiency in tomatoes. This plays a key role in the xylem transport of nitrate from roots to shoots and uptake in roots [ 31 ]. In AMF symbiosis, several studies indicate that plants absorb a large amount of N through the mycorrhizal pathway [ 12 , 19 ]. In the present study, AMF hyphae absorbed and transported both nitrate and ammonium N to the shoots of tomato plants with both high and low levels of N application, while under low N levels, the transported N became more important with a higher N application rate, although almost the same amount of N was transported via extraradical mycelia. Inoculation with AMF significantly increased the expression of LeNRT2.3 and LeAMT1.1 , which was also related to the N level and form in hyphal compartments. In conclusion, substantial amounts of both NO 3 − -N and NH 4 + -N can be transferred via extramycelia to their tomato hosts with the colonization of AMF. Under a low N supply in root environments, the partially transferred N in the plant’s total N uptake is more important than under high N supply. The expression of LeAMT1.1 and LeNRT2.3 were differentially influenced due to N supply levels."
} | 4,401 |
33893232 | PMC8092576 | pmc | 5,047 | {
"abstract": "Significance Human-designed infrastructures and networks relying on centralized or hierarchical control are susceptible to single-point catastrophic failure when disrupted. By contrast, most complex biological systems employ distributed control and can be more robust to perturbations. In field experiments with Eciton burchellii army ants, we show that scaffold structures, self-assembled by living ants, emerge in response to disrupted traffic on inclines, facilitating traffic flow and stemming losses of foragers and prey. Informed by our observations, we present a theoretical model based on proportional control and negative feedback, which may be relevant to many distributed systems in which group-level properties can be modified through individual error sensing and correction. The mechanism is simple, and ants only require information about their individual state.",
"discussion": "Discussion Our experimental results reveal that E. burchellii army ants consistently form scaffold structures when crossing surfaces inclined beyond a threshold of around 40 ° , above which ants begin to lose their footing. Scaffolds are responsive to local environmental geometry and specific traffic conditions: steeper slopes and higher rates of prey delivery and traffic are all associated with the formation of larger structures. When they form, scaffolds exhibit a saturating growth response, growing rapidly at first, then settling to a final size, which is a predictable function of the environmental variables. We also identify some of the functional benefits that scaffolds provide for the colony, including reducing losses of foragers and prey and alleviating traffic disruption. Informed by these results, our theoretical model offers a simple and plausible mechanistic explanation of scaffold formation. It describes scaffold growth as a response at the collective level, akin to a form of distributed control, which results from a process of individual error correction. As observed in our experiments, ants crossing steeper inclines are more likely to slip, and we hypothesize that ants stop to join scaffolds with a conditional probability that depends on slipping, responding to this stimulus by attaching to the surface and remaining stationary. Our experimental observations quantify how scaffold formation reduces the probability of slipping over time, and the model predicts the initial slipperiness of a surface at a given slope and how this is reduced over time as scaffolds grow. These predictions emerge from the model even though none of our experimental observations of ants slipping are used as inputs for the model or for the model-fitting procedure. The model captures the growth dynamics of scaffolds across a range of traffic conditions, providing insight into how particular combinations of traffic and environmental geometry influence their growth. The growth rate and ultimate size of scaffolds are emergent properties that can arise from individual sensing and error correction, without the need for complex communication, enabling a system-level response that is suited to each particular configuration of traffic and environmental geometry. The error in this system is detected at the individual level, and individual error correction leads to the emergent group-level response of scaffold growth. This individually sensed error may be ants responding to their own slipping, experiencing disruption due to other ants slipping at an increased rate, or some combination of such stimuli. In any case, this error is reduced by the formation of scaffolds and is detected at the individual level; therefore, we consider this a distributed form of proportional control. In an early description of army ant self-assembly from 1874 ( 40 ), the naturalist Thomas Belt asked, “Can it not be contended that such insects are able to determine by reasoning powers which is the best way of doing a thing, and that their actions are guided by thought and reflection?” Indeed, many dramatic examples of collective animal behavior were long thought to result from sophisticated reasoning capabilities or even unknown mechanisms involving thought transfer. However, research on collective behavior has revealed common principles that explain complex group-level behaviors as the result of relatively simple individual rules and local interactions ( 41 , 42 ). A key challenge in understanding these principles is ascertaining how inputs from individual-level sensors are combined and filtered to produce group-level outputs ( 43 ), and such insights have come to inform the design of complex engineered systems that rely on distributed forms of control ( 44 ). Thus, while the experiments described here involved ants crossing an inclined surface, our results may be relevant to other systems in which group-level properties can be modified via individual error sensing and correction. As human technological and social systems increase in complexity, the need for simple, robust mechanisms for error correction to rapidly respond to systemic disruption at multiple scales without relying on complex communication or oversight is paramount. Such simple mechanisms can avoid some of the pitfalls that often arise from communication within groups like biases ( 45 , 46 ) that result from the oversharing of correlated information ( 47 , 48 ), and the model presented here may inform approaches to addressing these. Our model and experimental results should provide insights for the field of swarm robotics as well ( 49 – 51 ), in which increasingly complex group behaviors, including self-assembly ( 52 ), are often constrained by technical limitations on both individual sensors and communication capabilities ( 53 ). In this context, our model relates to previous models of self-organized aggregation that have informed the development of robotic systems, such as models of cockroach aggregation in which individuals assess others already in a cluster ( 50 , 54 – 56 ) or corpse piling in ants, where individuals respond to corpses already deposited, and to other ants ( 57 , 58 ). However, the mechanism we have identified requires less complex sensing since the ants need only to respond to their own slipping and do not need to assess the size of a structure in place. This describes scaffold growth as a rapid, on-demand response that alleviates disruption that would otherwise result from the dynamic environmental conditions. With minimal requirements for sensing and information processing, this finding may also be relevant at smaller scales, for the design of self-healing materials ( 59 – 61 ) and future developments in biofabrication ( 62 , 63 )."
} | 1,665 |
23490197 | null | s2 | 5,048 | {
"abstract": "Bacillus subtilis has adopted a bet-hedging strategy to ensure survival in changing environments. From a clonal population, numerous sub-populations can emerge, expressing different sets of genes that govern the developmental processes of sporulation, competence and biofilm formation. The master transcriptional regulator Spo0A controls the entry into all three fates and the production of the phosphorylated active form of Spo0A is precisely regulated via a phosphorelay, involving at least four proteins. Two proteins, YmcA and YlbF were previously shown to play an unidentified role in the regulation of biofilm formation, and in addition, YlbF was shown to regulate competence and sporulation. Using an unbiased proteomics screen, we demonstrate that YmcA and YlbF interact with a third protein, YaaT to form a tripartite complex. We show that all three proteins are required for proper establishment of the three above-mentioned developmental states. We show that the complex regulates the activity of Spo0A in vivo and, using in vitro reconstitution experiments, determine that they stimulate the phosphorelay, probably by interacting with Spo0F and Spo0B. We propose that the YmcA-YlbF-YaaT ternary complex is required to increase Spo0A~P levels above the thresholds needed to induce development."
} | 326 |
35736075 | PMC9224570 | pmc | 5,050 | {
"abstract": "Transporters of the NRAMP family are ubiquitous metal-transition transporters, playing a key role in metal homeostasis, especially in Mn and Fe homeostasis. In this work, we report the characterization of the NRAMP family members ( RiSMF1 , RiSMF2 , RiSMF3.1 and RiSMF3.2 ) of the arbuscular mycorrhizal (AM) fungus Rhizophagus irregularis. Phylogenetic analysis of the NRAMP sequences of different AM fungi showed that they are classified in two groups, which probably diverged early in their evolution. Functional analyses in yeast revealed that RiSMF3.2 encodes a protein mediating Mn and Fe transport from the environment. Gene-expression analyses by RT-qPCR showed that the RiSMF genes are differentially expressed in the extraradical (ERM) and intraradical (IRM) mycelium and differentially regulated by Mn and Fe availability. Mn starvation decreased RiSMF1 transcript levels in the ERM but increased RiSMF3.1 expression in the IRM. In the ERM, RiSMF1 expression was up-regulated by Fe deficiency, suggesting a role for its encoded protein in Fe-deficiency alleviation. Expression of RiSMF3.2 in the ERM was up-regulated at the early stages of Fe toxicity but down-regulated at later stages. These data suggest a role for RiSMF3.2 not only in Fe transport but also as a sensor of high external-Fe concentrations. Both Mn- and Fe-deficient conditions affected ERM development. While Mn deficiency increased hyphal length, Fe deficiency reduced sporulation.",
"conclusion": "5. Conclusions This manuscript describes, for the first time, characterization of the NRAMP family members, the RiSMF genes, in an AM fungus. The R. irregularis SMF genes are expressed both in the ERM and IRM and are differentially regulated by environmental Fe and Mn. RiSMF3.2 encodes a protein mediating Mn and Fe transport from the environment, being the first Mn transporter reported in an AM fungus. These data indicate R. irregularis uses various strategies to increase Fe uptake from the environment: the previously identified plasma-membrane Fe permease RiFTR1 and the RiSMF3.2 NRAMP transporter. Further studies are required to understand the relative contribution of these transporters to Fe uptake by the IRM and the ERM and to elucidate the role of the other members of the R. irregularis NRAMP family.",
"introduction": "1. Introduction Transition metals, such as iron (Fe) and manganese (Mn), are essential micronutrients necessary for the correct development and survival of all organisms. These metals play important roles in multiple biochemical processes, since they have structural roles in many proteins and act as enzyme cofactors. However, at high concentrations, they generate noxious reactive oxygen species (ROS) that are deleterious for growth and development [ 1 , 2 ]. Therefore, metal homeostasis must be strictly balanced at the cell and whole-organism level. To maintain cellular metal homeostasis, all organisms have evolved a series of mechanisms, including metal uptake, chelation, trafficking and storage systems [ 3 ]. Metal transporters play a major role in keeping appropriate metal concentrations in the different cellular compartments [ 4 ]. One of the most ubiquitous classes of metal transporters are the natural-resistance-associated macrophage proteins (NRAMPs) [ 5 ]. The NRAMP family represents an evolutionary-conserved strategy for acquiring and trafficking essential transition metals, such as Mn and Fe. All NRAMP transporters possess a highly conserved core of 10–12 transmembrane domains and two motifs that are essential for their transport function, the DPGN metal-binding domain at the first transmembrane domain and a high-conserved metal-transport motif in the cytoplasmic loop, between transmembrane domains 8 and 9 [ 6 ]. Members of the NRAMP family have been identified and characterised in many organisms, ranging from bacteria to mammals. The model yeast Saccharomyces cerevisiae has three homologs of this family in its genome: SMF1, SMF2 and SMF3 [ 7 ]. Smfp1 and Smfp2 were, initially, identified as extremely hydrophobic proteins that supress a lethal mutation in the yeast-mitochondrial-processing protein. Thereafter, they were shown to be Mn transporters that can, also, transport other divalent metal ions [ 8 ]. Smf1 is located at the plasma membrane and is responsible for Mn uptake, while Smf2p is located in intracellular vesicles and imports Mn for Mn-requiring enzymes. Smf3p localises at the vacuolar membrane and regulates vacuolar Fe transport [ 9 ]. While Smf1p and Smfp2 are induced by Mn deficiency and, to a lesser extent, by Fe deficiency [ 10 ], Smf3p is strongly induced under Fe starvation [ 9 ]. Orthologues of the SMF genes have been characterised in other fungi, including Cryptococcus neoformans [ 11 ], Schizosaccharomyces pombe [ 12 , 13 ] and Aspergillus niger [ 14 ], among others. Most fungal NRAMP family transporters have been involved in Mn homeostasis. For example, the SMF1 transporter of Candida albicans plays a role in Mn assimilation under alkaline conditions [ 15 ], and the high-affinity PsMnt transporter of the white-rot fungus Phanerochaete sordida is involved in cellular Mn accumulation under Mn-deficient conditions [ 16 ]. Other fungal NRAMP transporters, such as the plasma-membrane Fe-transporter EpNrampp of the dark endophyte Exophiala pisciphila [ 17 ] and the Cryptococcus neoformans Mn-transporter smf1 have been, also, involved in Cd tolerance [ 18 ], as their expression is down-regulated by Cd toxicity. In A. niger , deletion of the high-affinity Mn transporter DmtA leads to defects in germination and hyphal morphology [ 14 ]. However, NRAMP transporters in arbuscular mycorrhizal (AM) fungi remain uncharacterised. AM fungi are obligate plant mutualistic microorganisms of the subphylum Glomeromycotina, within the Mucoromycota [ 19 ], that form a symbiotic association with most plant species [ 20 ]. The fungus colonises the root and forms highly branched structures, called arbuscules, inside the cortical cells. Outside the root, the fungus develops an extensive network of hyphae in the soil that can absorb nutrients beyond the depletion zone that develops around the roots. This extraradical mycelium (ERM) provides a new pathway to the plant, for the uptake of low-mobility nutrients in the soil, mainly P, N and some transition metals (Zn, Fe and Cu). In return, the plant provides sugar and lipids to the fungus [ 21 , 22 ]. This nutrient exchange takes place in the arbuscule-colonised cortical root cells. Besides enhancing nutrient uptake to their host plants, AM fungi provide increased tolerance against biotic and abiotic stresses, including drought, salinity or metal toxicity [ 23 , 24 , 25 ]. AM fungi play a crucial role in modulating plant metal acquisition, over a wide range of soil metal concentrations, as they increase plant metal acquisition in soils deficient in these elements but reduce metal uptake in contaminated soils [ 23 , 26 ]. Despite the importance of AM fungi for plant metal homeostasis, very few metal transporters have been characterised in these organisms. Up to now, just three components of the reductive iron uptake pathway (RiFTR1, RiFTR2 and RiFRE) and three copper transporters of the CTR family (RiCTR1-3) have been characterised in the model fungus Rhizophagus irregularis [ 27 , 28 ]. Iron uptake by the ERM starts with the reduction of Fe 3+ to Fe 2+ in the soil solution, by the plasma-membrane ferric reductase RiFre1. Then, Fe 2+ is taken up by the plasma-membrane Fe permease RiFTR1 [ 27 ]. Although RiFTR2 function has not been determined, it has been proposed to play a role in Fe homeostasis, under Fe-limiting conditions. Regarding the R. irregularis Cu transporters, RiCTR1 mediates Cu uptake by the ERM from the environment, RiCTR2 is involved in mobilization of Cu vacuolar stores and RiCTR3a has been suggested to function as a Cu transceptor, involved in Cu tolerance [ 28 ]. Previous bioinformatics analysis of the R. irregularis genome identified four NRAMP family members ( RiSMF1 , RiSMF2 , RiSMF3.1 and RiSMF3.2 ) that remain uncharacterised [ 29 ]. At the plant side, up-regulation of NRAMP transporter genes has been observed in tomato (LeNramp1 and LeNramp3 [ 30 ]) and alfalfa (MsNramp1 [ 31 ]) mycorrhizal roots. However, LeNramp1 and LeNramp3 expression was down-regulated, when the symbiosis was established in a metal-contaminated soil, probably since the heavy-metal content is lower in mycorrhizal than in non-colonised roots [ 30 ]. The aim of this work was to characterise the R. irregularis NRAMP family members, in order to get further insights into the role of the NRAMP transporters in the AM symbiosis and into the mechanisms of the metal homeostasis in AM fungi. Data presented in this manuscript describe, for the first time, a Mn transporter in an AM fungus and show that R. irregularis uses various strategies for Fe uptake from the environment.",
"discussion": "4. Discussion A previous genome-wide analysis of metal transporters in R. irregularis revealed the presence of four gene sequences, RiSMF1 , RiSMF2 , RiSMF3.1 and RiSMF3.2 , putatively encoding transporters of the NRAMP family [ 29 ]. Mining of the more recent R. irregularis genome and transcriptome databases confirms that the R. irregularis NRAMP family is composed of four members. In this work, we functionally characterised RiSMF3.2 and analysed gene expression patterns of the R. irregularis NRAMP family members. Our data indicate that the RiSMF genes are differentially regulated by Mn and Fe availability, and that RiSMF3.2 encodes a functional Mn and Fe transporter. The obligate biotrophic and multinucleate nature of AM fungi prevents the use of the most common strategies to investigate the functionality of a gene of interest. Recent advances have been made to assess the functionality of AM fungal genes, by host-induced gene silencing or virus-induced gene silencing [ 49 , 50 ]. However, these techniques are used to investigate the function of genes highly expressed in the IRM and have not always been successfully applied. To bypass this technical constrain, we tried to assess the function of the RiSMF gene products in yeast. Unfortunately, we could only determine the transport function of RiSMF3.2 in the heterologous system. Despite only RiSMF3.2 presented a metal transport activity in yeast, the four R. irregularis NRAMP sequences contain all the structural features of NRAMP proteins. In fact, they contain the conserved transmembrane motif GQSSTITGTYAGQY(/F)V(/I)MQGFLD(/ E/N) and the DPGN motif characteristic of the NRAMP family [ 6 ]. Numerous mutational studies have shown that both domains are essential for the metal-transport activity of NRAMP transporters [ 51 ]. For example, conservative substitutions of the aspartate (D) and asparagine (N) residues impaired metal binding of the ScaNramp of Staphylococcus capitis [ 52 ] and eliminated metal transport in human NRAMP2 [ 53 ] and Ecoli Nramp [ 54 ]. Thus, this motif in the R. irregularis NRAMP proteins should, also, play a role in metal binding. The phylogenetic analysis revealed that the Glomeromycotina NRAMP sequences are divided in two subfamilies, a Group I belonging to the clade grouping fungal NRAMP sequences, and a Group II that is independent from the clusters formed by animal, plant, fungal and bacterial sequences. Group II is more closely related to known plant NRAMP sequences than to fungal sequences. In addition to the divergence in primary amino acid sequence between genes from Group I and II, they display some differences in gene organization. Members of Group II are more fragmented and have three–four introns, located at conserved positions at the carboxi-terminus. This sequence divergence suggests an early evolutionary separation between the two groups of Glomeromycotina NRAMPs. The observation, that all AM fungal species analysed have NRAMP genes from Group I and Group II, suggests that both groups are required for proper metal homeostasis in AM fungi. Further studies are required, to determine if proteins of the two groups display different transport functions. The finding that RiSMF3.2 reverts the mutant phenotype of the smf1 Δ and fet3 Δ fet4 Δ strains, lacking, respectively, the high-affinity Mn transporter smf1 and the iron uptake systems, indicates that RiSMF3.2 encodes a NRAMP transporter that mediates Mn and Fe transport. Although subcellular localization of RiSMF3.2 could not be demonstrated in yeast, it is the orthologue of the plasma-membrane smf1 transporter. The other members of the R. irregularis NRAMP family could not be characterised in the heterologous system because, as revealed by the yeast localization assays, they were not expressed in the yeast membranes, most likely as a consequence of an artefact of the heterologous system. Although the principles of targeting seem to be conserved between organisms, the heterologous proteins may lack the sequences required for targeting to the correct compartment in the cell, so problems regarding the correct folding and targeting of the heterologous expressed proteins can occur [ 55 ]. RiSMF1 function could not be determined, but it seems to be involved in Mn and Fe homeostasis, as its expression levels in the ERM are regulated by Mn and Fe availability. The contrasting expression patterns of RiSMF1 in the ERM grown in the in vivo and in vitro cultures systems under Mn deficient conditions may be because the ERM grown in monoxenic cultures is not Mn harvested. Since the culture medium of the root compartment contains Mn, it is possible that Mn is transferred from the IRM to the ERM in the monoxenic cultures. Down-regulation of RiSMF1 by Mn deficiency is striking, as high-affinity metal transporters are expected to be up-regulated under metal deficient conditions. However, unexpected regulation patterns of NRAMP genes by Mn have been observed in various organisms. For example, expression of the Aspergillus oryzae AoNramp1 gene increases under Mn toxicity [ 56 ] and transcript levels of the cucumber CsNRAMP1 , CsNRAMP4 and CsNRAMP5 genes decrease under Mn deficiency [ 57 ]. How Mn availability affects NRAMP gene expression remains to be investigated. As it has been described for the yeast smf1p and smf2p transporters, the RiSMFs could, also, be regulated at the post-translational level by Mn [ 10 ]. Up-regulation of RiSMF1 expression in the ERM, under Fe-limiting conditions, suggests a role for its encoded protein in Fe-deficiency alleviation. Since the RiSMF1 subcellular location could not be determined in the heterologous system, its role in Fe homeostasis could either be due to its capacity to increase Fe uptake from the environment or to mobilise the vacuolar Fe stores under Fe-limiting conditions. Although RiSMF3.2 transcript levels were not regulated by Fe deficiency, the yeast-complementation assays revealed that it encodes a plasma-membrane Fe transporter of the NRAMP family involved in Fe uptake. Recent work by Tamayo et al. [ 27 ] has identified two R. irregularis Fe permeases (RiFTR1 and RiFTR2) involved in Fe homeostasis. RiFTR1 is involved in Fe acquisition by the plasma membrane and RiFTR2 in Fe homeostasis under Fe-limiting conditions. These data, altogether, suggest that R. irregularis uses various strategies to increase Fe uptake from the environment, the plasma-membrane RiFTR1 and RiSMF3.2 transporters. Similarly, multiple systems operate at the cell surface for Fe uptake in S. cerevisiae , the high-affinity Fe transporters Ftr1p and smfp1 and the low-affinity transporter Fet4p [ 8 , 46 , 58 ]. Additional studies at the protein level are needed to understand whether RiSMF1 mediates Fe uptake from the environment or mobilises the Fe vacuolar stores and the relative contribution of the different Fe transporters to Fe uptake by the ERM. As expected for an uptake plasma-membrane transporter, RiSMF3.2 expression levels decreased when the ERM was grown in a media containing 45 mM Fe for 20 d. A similar expression pattern was observed by Tamayo et al. [ 27 ], for the Fe transporters RiFTR1 and RiFTR2, under these experimental conditions. Yeast NRAMP Smf1p, the orthologue of RiSMF3.2, is also down-regulated by metal toxicity [ 59 ]. Transcriptional down-regulation of the R. irregularis proteins, mediating Fe transport into the cytosol, will limit uptake of toxic levels of Fe by the ERM. However, the effect of Fe toxicity on RiSMF3.2 expression, 16 h after the addition of 45 mM Fe, resulting in enhanced transcript levels, was striking. This difference implies that the contribution of RiSMF3.2 at the early stages of Fe toxicity would be increased Fe uptake, leading to an increased Fe toxicity. Alternatively, it could be hypothesised that RiSMF3.2 could act as a sensor of high external Fe concentrations, to activate the signalling cascades involved in Fe tolerance. In fact, nutrient sensing in fungi can be mediated by transceptors, which are proteins with both transport and receptor functions [ 60 , 61 ]. In the absence of a methodology to silence AM fungal genes in the ERM, it is not possible to understand the biological significance of this transient increase in RiSMF3.2 mRNA levels. Gene expression of the RiSMF s in the IRM reveals the importance of keeping both Mn and Fe homeostasis during the in planta phase of the fungus. The finding that RiSMF2 was the most highly expressed gene in IRM and arbuscules collected from Medicago roots, while RiSMF1 was the gene displaying the highest expression levels in carrot mycorrhizal roots, indicates that expression of these genes is regulated by the plant genotype and/or the experimental conditions. Host-dependent expression of a subset of R. irregularis secreted proteins has been, also, reported [ 33 ]. Further studies are required to determine the host cues regulating RiSMF1 and RiSMF2 expression in the IRM, and if they are involved in metal uptake from the apoplast of the symbiotic interface. Up-regulation of RiSMF3.1 expression by Mn deficiency in the IRM suggests that, under Mn starvation, the fungus needs to increase its Mn cytosolic content, in order to provide it to the Mn-requiring enzymes, such as the mitochondrial Mn superoxide dismutase and the Golgi-located enzymes involved in the glycosylation of secretory proteins [ 62 ]. In fact, the yeast smf2p transporter has been shown to be a central player in Mn trafficking to the mitochondria and other cellular sites [ 63 ]. Taking into account that expression of the RiSMF genes is not affected by Fe deficiency in the IRM and the high expression levels reported for the high-affinity transporter RiFTR1 in the IRM [ 27 ], it is likely that RiFTR1 is the major player in Fe homeostasis, in the structures the fungus develops in the root. Based on data presented in this manuscript and on previous reports of the R. irregularis Fe uptake systems [ 27 ], we propose a model for Mn and Fe transport in the AM fungus R. irregularis ( Figure 9 ). Our gene-expression and functional analyses in yeast strongly suggest that RiSMF3.2 is involved in Mn and Fe uptake by the ERM from the soil solution and by the IRM from the apoplast of the symbiotic interface. The plasma-membrane Fe transporter RiFTR1 also contributes to Fe uptake in both fungal structures [ 27 ]. In depth analyses of the cellular function of the identified NRAMP transporters RiSMF1, RiSMF2 and RiSMF3.1 in different fungal structures is required, to better understand whether they also contribute to Mn and Fe uptake, or if they are involved in Mn trafficking or in mobilization of the metal vacuolar stores. Numerous studies have shown that AM fungi increase plant acquisition of the essential metals Zn, Cu and Fe; however, information about the role of the symbiosis in plant Mn nutrition and on the underlying mechanisms is scarce. While a few studies have shown that Mn uptake is higher in mycorrhizal plants [ 64 , 65 ], it has been, repeatedly, reported that Mn acquisition decreases in mycorrhizal plants [ 66 ]. AM fungi have been shown to reduce the number of Mn-reducing bacteria [ 67 ] or increase the number of Mn-oxidizing bacteria in the rhizosphere [ 68 ], decreasing indirectly Mn availability. Nevertheless, the observed hyphal length observed under Mn-deficient conditions indicates that, under these conditions, the fungus explores a higher volume of soil, which will increase the nutrient-uptake effectiveness of the mycorrhizal root. Under field conditions, Mn uptake by mycorrhizal roots may depend on which of the two functions (Mn availability in the mycorrhizosphere or volume of soil exploited by the mycorrhizal root) prevails in the soil. Regarding the effect of Mn on the developmental partner of the ERM, the increased hyphal length observed, when the fungus grows in the absence of Mn, agrees with previous observations for other nutrients [ 69 , 70 ]. As proposed by Bago et al. [ 69 ] and Olsson et al. [ 70 ], this growth pattern is, probably, designed to explore and exploit more efficiently the growth medium. An effect of Mn deficiency on hyphal development has been, also, reported in Aspergillus niger [ 14 ]. Although Fe deficiency did not affect hyphal length, it decreased sporulation. Fe is an essential micronutrient that is a cofactor of numerous enzymes, thanks to its ability to easily accept and release electrons [ 71 , 72 ]. Therefore, inhibition of sporulation might be as a consequence of the inhibition of the activity of the enzymes required for spore formation, when the fungus is grown in media lacking Fe. Previous studies have shown that AM fungal-spore formation is affected by nutrient availability [ 64 ]. Detection of Fe in the spores developed under Fe toxicity agrees with previous observations for other metals, such as Cu, Zn and Cd [ 73 , 74 , 75 ], and supports the hypothesis that a survival strategy of AM fungi in metal-contaminated environments is to accumulate the excess metal in some spores of the fungal colony. Iron accumulation in the fungus will reduce plant Fe availability, which will explain, at least partially, the improved performance of mycorrhizal plants, in soils affected by iron mining tailing [ 76 , 77 ]."
} | 5,586 |
34149752 | PMC8210828 | pmc | 5,051 | {
"abstract": "Synthetic microbial communities (SynComs) are a useful tool for a more realistic understanding of the outcomes of multiple biotic interactions where microbes, plants, and the environment are players in time and space of a multidimensional and complex system. Toward a more in-depth overview of the knowledge that has been achieved using SynComs in the rhizosphere, a systematic review of the literature on SynComs was performed to identify the overall rationale, design criteria, experimental procedures, and outcomes of in vitro or in planta tests using this strategy. After an extensive bibliography search and a specific selection process, a total of 30 articles were chosen for further analysis, grouping them by their reported SynCom size. The reported SynComs were constituted with a highly variable number of members, ranging from 3 to 190 strains, with a total of 1,393 bacterial isolates, where the three most represented phyla were Proteobacteria, Actinobacteria, and Firmicutes. Only four articles did not reference experiments with SynCom on plants, as they considered only microbial in vitro studies, whereas the others chose different plant models and plant-growth systems; some of them are described and reviewed in this article. Besides, a discussion on different approaches (bottom-up and top-down) to study the microbiome role in the rhizosphere is provided, highlighting how SynComs are an effective system to connect and fill some knowledge gaps and to have a better understanding of the mechanisms governing these multiple interactions. Although the SynCom approach is already helpful and has a promising future, more systematic and standardized studies are needed to harness its full potential.",
"conclusion": "Conclusion and Future Prospects Not enough is known about plant–microbe interactions and even less about microbe–microbe interactions in the plant rhizosphere. The development and employment of SynComs are giving more insights into the rhizosphere dynamics and structure, and how these dynamics influence plant fitness and behavior. Nevertheless, more systematic and standardized investigations are needed to harness the full potential of this approach or to perform statistical meta-analyses. Over time, experimental approaches may incorporate even more factors and complexity into the system, such as spatiotemporal analyses and ecological dynamics, and may include more taxonomic levels to the interactions. Current research on Archea and Eukarya kingdoms is emerging; and as more individuals are being sequenced, more protists—key microbiome predators—are also being linked to the plant holobiont. Finally, better tracking systems to consider, for instance, the auxotrophy phenomena or the rapid community fluctuations, and the use of computational and mathematical modeling will also bring additional insights and tools to uncover the still-hidden mysteries of the rhizosphere.",
"introduction": "Introduction One of the most relevant discoveries in biological research of the past couple of decades is the role that host-associated microbial communities (microbiomes) have in health (Berendsen et al., 2012 ; Gallo and Hooper, 2012 ; Mendes et al., 2013 ; Haney et al., 2015 ; Pieterse et al., 2016 ), nutrition (Hacquard et al., 2015 ), growth (Lugtenberg and Kamilova, 2009 ), and even behavior throughout the plant and animal kingdoms (Wagner et al., 2014 ; Vuong et al., 2017 ; Lowry et al., 2018 ). Nevertheless, the mechanisms underlying individuals' interactions from different taxonomic domains are generally challenging to assess, and the rhizosphere is not an exception. Indeed, the rhizosphere is a perfect example of an environment where it is possible to find highly complex intradomain and interdomain interactions. The challenges are usually associated with the vast biodiversity that is present in this environment (Durán et al., 2018 ; Xiong et al., 2020 ), the edaphic factors and physical structures that make up countless microniches available for microbial colonization (Allard-Massicotte et al., 2016 ; Robertson-Albertyn et al., 2017 ; Howard et al., 2020 ; Kong et al., 2020 ), and the chemical richness with its spatiotemporal variety as a consequence of the organisms that interact under and above this realm (Chaparro et al., 2014 ; Staley et al., 2017 ; Vives-Peris et al., 2020 ). The interplay between hosts and their associated microbiomes affects the ontogeny of the partners and is also seen as a fundamental basis of microbial ecology and evolution (Rausch et al., 2019 ; Batstone et al., 2020 ). Likewise, the role of microbiomes in the adaptation and evolutionary processes of plants is under analysis (Rosenberg and Zilber-Rosenberg, 2016 ; Hawkes et al., 2020 ). Over the years, many different techniques and procedures have been used to decipher plant–microbe and microbe–microbe interactions in the rhizosphere. In a broad sense, there are two ways to address the environmental–molecular biology topic questions. Reductionist approaches seek to control as many experimental factors as possible, usually analyzing plant–microbe interactions where the players are well-known (Liu et al., 2019 ). Good examples of these approaches are those studying interactions, such as plant–mycorrhizal associations (Krajinski and Frenzel, 2007 ; Nadeem et al., 2014 ), plant–pathogen protection (Pieterse et al., 1998 ; Pascale et al., 2020 ), the interactions between plants with symbiotic nitrogen-fixing (Ferguson et al., 2010 ; Sulieman et al., 2015 ), or plant growth–promoting rhizobacteria (Poupin et al., 2013 ; PGPRs, Pinedo et al., 2015 ; Timmermann et al., 2019 ). In contrast, holistic approaches aim to study the plant microbiome as a whole (holobiont), focusing on diminishing interferences to reduce environmental variation and elucidate how it operates in its natural environment (Fang and Casadevall, 2011 ; Tecon et al., 2019 ). Thus, different technologies and protocols are more or less well-fitted to each of these approaches; while the reductionist usually depends on culture-dependent methodologies, the holistic typically relies on culture-independent techniques such as high-throughput genomic (Raes et al., 2011 ). These two approaches are required to address microbial ecology issues as microbiomes operate as a whole or with subsets of them, but at the same time, each of their members is indeed an individual organism that may exert particular effects in plants. Unfortunately, holistic approaches likely stay in the first step of top-down analyses, and the reductionist approaches remain only in the earlier steps of a bottom-up exploration. Synthetic microbial communities (SynComs) are consortia designed to test hypotheses and to mimic, to some extent, the role of a particular microbiome and have received particular interest in recent years (Vorholt et al., 2017 ; reviewed in de Souza et al., 2020 ). Here, a systematic review of the literature on SynCom was performed to identify the overall rationale, design criteria, experimental procedures, SynCom characteristics, and outcomes of in vitro or in planta tests using SynComs in the rhizosphere. Additionally, this information was used to compare different approaches (bottom-up and top-down) to study the microbiome's role in the rhizosphere. Finally, the need for more systematic and standardized studies with SynCom is discussed. This will allow us to fill and connect some knowledge gaps, perform statistical meta-analyses, and better understand these interactions, where microbes, plants, and the environment are players of a multiscale and complex system.",
"discussion": "Discussion SynComs to Narrow the Gap Between Bottom-Up and Top-Down Strategies As mentioned previously, research and scientific foundations to date have been developed using reductionist or holistic strategies, and as the more information emerges, the more interesting it is to fill the gaps between these two approaches. The use of SynComs arises as a promising “middle-out” point of view that can narrow the gap between the knowledge obtained with single strains and whole microbiomes ( Figure 4 ). In general, experimentation complexity and certainty of the interacting factors are inversely proportional; whereas the first tend to increase as more strains are added to the research, the second one is bigger when fewer strains are studied. Moreover, correlational and causality analyses also increase and decrease inversely ( Figure 4 ). While some researches focus on attaining specific traits within the community (Moronta-Barrios et al., 2018 ) or look for phylogenetic proximity (Burghardt et al., 2018 ; De Vrieze et al., 2018 ; Gadhave et al., 2018 ), others try to design SynComs by mixing different strains that differ either in their phylogenetic identity (Niu et al., 2017 ; Voges et al., 2019 ), origin of isolation (de Souza et al., 2019 ; Zhang et al., 2019 ), synthesis, emission, and/or response to volatile (Schulz-Bohm et al., 2015 ) and nonvolatile metabolites (Lebeis et al., 2015 ), or response patterns to root exudates (Herrera Paredes et al., 2018 ). Figure 4 Different approaches to understand the role of the root microbiome in plants. Schematic representation of information processing strategies (top-down and bottom-up) and the approaches that stand on the edges of them (reductionist and holistic). Experimentation complexity increases when more microorganisms are added to the investigation, whereas traits and member certainty decrease. Reductionist approaches tend to rely on causality analyzes, whereas holistic approaches most likely rely on correlational analyzes. SynCom experimentation lies in the middle ground, allowing a better understanding of microbe–microbe interactions. The Whole of a Small SynCom Is Usually More Than the Sum of Its Parts When the research aimed to evaluate PGPR traits, SynComs tended to be smaller, prioritizing a functional examination rather than an ecological analysis of the communities, looking for a simplified way to add the “community factor” or the interactions that can cause different outcomes. Employing a top-down approach, Zhuang et al. ( 2020 ) explored the wild bacterial community of the garlic rhizosphere to screen PGPR strains that also had effects on radish seedlings ( Raphanus sativus ). They found that increasing Pseudomonas community richness is beneficial to the plant biomass and nutrient content and, compared with single-strain inoculants, multistrain microbial inoculants can promote plant growth more reliably and effectively. Small SynComs have also been used to assess the community impact in other PGPR traits such as biocontrol and host–immune response modulation. Lee et al. ( 2020 ) analyzed the microbial community structure of diseased and healthy tomato plants, detecting that the abundance of Gram-positive Actinobacteria and Firmicutes phyla was different, depending on the infection. They designed a SynCom with four Gram-positive species, which did not directly antagonize the pathogen, but displayed greater immune activation against it. Interestingly, plant protection was longer with the SynCom compared to each of the individual strains (Lee et al., 2020 ). Based on the native community of N. attenuata , Santhanam et al. ( 2019 ) designed a five-strain SynCom with known biocontrol mechanisms. They concluded that the complementary abilities of the SynCom members account for the desired protective effect as only the consortium and not the single isolates or smaller combinations of them were able to significantly reduce the disease incidence caused by Fusarium–Alternaria phytopathogens in field-grown plants (Santhanam et al., 2015 , 2019 ). Moreover, Niu et al. ( 2017 ) assembled a greatly simplified but representative seven-member SynCom that exerted a biocontrol effect over a pathogenic fungus that causes maize seedling blight, where the plants treated with the SynCom displayed the lowest disease severity index. Although each one of the seven community members showed some biocontrol effects against Fusarium verticillioides , the effects exhibited by the entire community were more substantial (Niu et al., 2017 ). Small SynComs have also been used to study communication along the rhizosphere through the release of different chemicals. Schulz-Bohm et al. ( 2015 ) developed a soil model system mimicking more closely the natural context around the rhizosphere to understand the ecological role of volatiles in soil microbial interactions. They detected that a different blend of volatile organic compounds (VOCs; alcohols, ketones, esters, aromatic and organosulfur compounds, among others) was produced by a bacterial mixture of five strains, after comparison with the VOC profile of monocultures in vitro , and that some volatiles were only emitted by the bacterial mixture and not by the monocultures (Schulz-Bohm et al., 2015 ). Likewise, some antifungal compounds are only produced in diverse bacterial communities, suggesting that less abundant taxa play an important role in antifungal volatile production (Hol et al., 2015 ). However, pathogen-suppressive characteristics can be lost if stress exposure leads to an alteration of the community structure (van Agtmaal et al., 2015 ), effects that can be observed through the analysis of soil VOCs as a function of varying environmental factors over large spatiotemporal scales (McNeal and Herbert, 2009 ; van Agtmaal et al., 2015 ). In this context, but not in the case of VOCs, Sánchez-Gorostiaga et al. ( 2019 ) observed that the contribution of a given species or pair of species to a community function may depend on the presence or absence of other taxa and that the effects are not always additive. Specifically, using mathematical modeling and in vitro experiments, they examined how the amylolytic rate of combinatorial assemblages of six starch-degrading soil bacteria depended on the separate functional contributions from each species and their interactions. They found that the ability of the model to predict community function declined as more species were added to the consortia (Sánchez-Gorostiaga et al., 2019 ), highlighting how complex and challenging it is to predict the functional interactions in microbial communities in bottom-up approaches. Microbe–Microbe Interactions as a Driver of a SynCom Outcome Durán et al. ( 2018 ), using large SynComs (190 members) and a network inference tool to analyze the relationships of operational taxonomic units, revealed that negative interactions dominated between kingdoms (fungi, oomycetes, and bacteria), whereas within kingdoms, positive interactions were more frequent. Moreover, the absence or presence of certain groups (i.e., the bacterial community) had a critical impact on plant growth and health (Durán et al., 2018 ). Selection and competition in host-associated bacterial populations influence rhizobial fitness, which can be explained in part by the host genotype. The symbiotic mutualistic relationship between nitrogen-fixing bacteria, such as Ensifer meliloti , and its legume plant hosts has been well-studied in single-strain experiments (Kraiser et al., 2011 ). However, little is known about the consequences of competition among strains over fitness in the rhizosphere. Burghardt et al. ( 2018 ) set out mixed-strain experiments to evaluate the strength of selection in multistrain competitive environments and estimate the resulting relative strain fitness. Host selection was dependent on the competition among strains, and the more efficient rhizobacteria (i.e., fixing nitrogen and/or contributing to host biomass) were more competitive at forming nodules or being rewarded by hosts (Burghardt et al., 2018 ). SynComs Value to Get Insights on Plant–Microbe Interactions SynCom applications to directly study plant–microbe interactions have proved fruitful. Moccia et al. ( 2020 ) reported not only the existence of plant-driven influence over colonization of some bacterial species, but also several colonization patterns depending on the plant nutrient conditions, or the presence of some microorganisms, discovered through “drop-out communities” assays. The latter is based on the use of different SynComs that comprised the same members, with the difference that in each of the SynCom tested, one member or taxa has been excluded or “dropped out” (Moccia et al., 2020 ). Comparing drop-out communities with the corresponding full SynCom provides an opportunity to determine the individual effects each taxon may have on either the community or the plant host function (Niu et al., 2017 ; Finkel et al., 2019 ; Moccia et al., 2020 ). Plant hormones, such as salicylic acid, are relevant agents in microbial colonization of different plant tissues, as shown by Lebeis et al. ( 2015 ). Plant immune system participation is vital to selectively allow colonization of beneficial nonpathogenic microbes while excluding potential pathogens. Their results showed that this phytohormone is required to assemble a normal root microbiome. The application of SynComs revealed that plant immune signaling drives the selection from available microbial communities to shape the root microbiome (Lebeis et al., 2015 ). Microbiota recruitment can also differ among plant varieties. Zhang et al. ( 2019 ) observed a difference in nitrogen-use efficiency between two rice varieties ( indica and japonica ). After profiling their microbiomes and using genetic approaches, researchers found a nitrogen transporter and sensor ( NRT1.1B ) associated with the recruitment of a large proportion of indica -enriched bacteria (variety with the highest N-use efficiency) and designed bacterial SynComs based on indica - or japonica -enriched operational taxonomic units that were associated with NRT1.1B . Finally, they found that an indica -enriched SynCom had a larger effect on rice growth (Zhang et al., 2019 ). Concerning plant–microbe interactions, several molecules, mainly plant secondary metabolites, are involved in shaping the microbiome structure. Voges et al. ( 2019 ) looked for community shifts that could occur in the absence of plant-secreted specialized small molecules, such as phytoalexins, flavonoids, and coumarins, employing an A. thaliana root-derived bacterial SynCom and plant mutant lines. They observed that a lack of coumarin caused a shift in the root microbial community, specifically under iron deficiency (Voges et al., 2019 ). It has been shown that these compounds alter the root microbiota composition and are required for microbiota-mediated plant iron uptake and immune regulation, contributing to a beneficial plant–microbiota interaction (Harbort et al., 2020 ). Regarding the effects of nutrient conditions on plant–microbe interactions, such as phosphate stress or starvation, Finkel et al. ( 2019 ) used large SynComs (185 strain members) to investigate how inorganic orthophosphate deficiency influences microbiome structure and, in turn, how this can affect plant health. They found that phosphate-stressed plants are susceptible to colonization by latent opportunistic competitors, thus exacerbating the plant's inorganic phosphate starvation condition (Finkel et al., 2019 ). The same SynCom was used in Finkel et al. ( 2020 ), where the authors worked with four modules of coexisting microbes, to assess how the interaction between microorganisms influences the Arabidopsis root growth. Interestingly, they found that a single genus, Variovorax , was responsible for maintaining the plant root growth program, overriding the drastic inhibition of root growth that was induced by a wide diversity of bacterial strains and by the entire community. Furthermore, they found that a single operon for auxin degradation was necessary and sufficient to observe the effect of these Variovorax strains in roots and that this genus is among a limited group of core bacterial genera found in 30 different plant species, suggesting its vital role in the bacteria–bacteria–plant communication networks (Finkel et al., 2020 ). SynCom Design and Assessment: Context and Guidelines for Comparable Experiments Not much vocabulary standardization was observed in the revised literature; terms like consortia, inocula, enrichment, community, and others were frequently used rather vaguely (definition of these and other terms is provided or suggested in Supplementary Table 3 ). We propose using the term “SynCom” when three or more isolates are used, each member of the consortium is well-known, each member is traceable (by any means), and the community is designed to address a specific research question. Additionally, we suggest choosing the term in vivo or in planta when SynComs are inoculated in plants growing in any substrate or cultivation system. SynCom sizes were quite variable: a small consortium (up to 10 strains) may not be as taxonomically representative as a large one (>100 strains). However, it could be as functionally diverse to allow these communities' normal ecological development (Toju et al., 2018 ). Species-rich communities are often more efficient and more productive than species-poor communities as they use limited resources more efficiently (Menéndez and Paço, 2020 ). For some studies, SynComs are designed to represent the core microbiome (Bulgarelli et al., 2012 , 2013 ; Lundberg et al., 2012 ) of a given plant (Kong et al., 2018 ), but it is important to note that satellite microbes, that is, changing microorganisms not generally found in the core microbiome, possibly play critical modulatory roles under particular environmental conditions (Shi et al., 2016 ; Compant et al., 2019 ; O'Banion et al., 2020 ). Therefore, as low relative abundance microbes may play significant roles in plant ecosystem functioning (Schulz-Bohm et al., 2015 ), less represented species should not go unnoticed when designing a SynCom. Several aspects concerning SynCom structure and dynamics should deserve some attention. Compatibility and antagonistic interactions between species were not tested in all analyzed SynCom reports (Moccia et al., 2020 ; Toju et al., 2020 ). Additionally, the starting inoculum concentration may influence the competitivity of the strains. Consequently, it would be more appropriate to test in vitro compatibility between strains (using different culture media), assessing the proper cell dosage before performing subsequent assays. Moreover, existing commercial isolates, which many laboratories use to mimic the WildCom of a given plant, might be genotypically or functionally distinct from microorganisms that researchers find in their new soil samples, thus not genuinely reflecting interactions between plants and the root microbiome in a given natural soil (Liu et al., 2019 ). Only some reports describe the assessment of antagonism among strains in compatibility tests consisting of small experiments such as the cross-streak method, spot inoculation in the vicinity, and evaluation of culture supernatants, among others (Vijayakumar and Muriana, 2015 ; Moran et al., 2016 ; Ijaz et al., 2019 ; Ishizawa et al., 2019 ; Olanrewaju and Babalola, 2019 ). Some authors claim they do not perform compatibility tests when members come from the same plant's natural environment. Nevertheless, having the same origin does not mean that the community members coexist without competition or antagonism. The metabolomic profile of a given community depends on the interactions among its members, for example, a different VOC profile is found for single bacterial strains (Ledger et al., 2016 ) compared to bacterial mixtures, a profile that may be susceptible to the presence or absence of some nonabundant strains (Schulz-Bohm et al., 2015 ). In this context, auxotrophic interactions may allow auxotrophs—unable to synthesize particular compounds required for their growth—to exist and colonize a niche, thanks to the coexistence with other species (Johnson et al., 2020 ). On the same line, genes related to biosynthesis can be absent in auxotrophic microbes, but genes encoding for proteins involved in the efficient transport of specific metabolites may be present instead (Jiang et al., 2018 ). Likewise, specific plant growth–promoting traits present in some species may be an indirect consequence of functions in other microbes that allow these species to develop beneficial metabolites for the plant. In other words, the contribution of a given species or pair of species to a community function may also depend on the presence or absence of other taxa (Jiang et al., 2018 ; Sánchez-Gorostiaga et al., 2019 ). A pivotal issue to assess the effects or performance of SynCom is to track the presence of these members in the plant study system. Unfortunately, not all applications of SynComs employed some method to follow up on the presence and persistence of the strains until the end of the experiment (Sánchez-Gorostiaga et al., 2019 ). Therefore, it is not always correct to link the observed effects with SynCom member functions. Several techniques improve the chances to perform a specific follow-up of different strains in an experiment, such as biomarkers, labels (i.e., tagged genes for further detection through spectrophotometry or microscopy), and advances in three-dimensional images with confocal microscopy (O'Banion et al., 2020 ). On the other hand, comparative studies using plant or microorganism mutants, or transgenic lines where gain and loss of function are tested, have also been especially useful to characterize different traits present in distinct species, giving a better understanding of microbial functioning or evolution (Zúñiga et al., 2013 ; Allard-Massicotte et al., 2016 ; Poupin et al., 2016 ). Another factor to have in consideration is both the short- and long-term scales of the experiments. This is because proliferation, abundance, and population densities within the microbiome are in constant control and modulation across the plant life cycle (Chaparro et al., 2014 ), even in small time scales (Staley et al., 2017 ; Hubbard et al., 2018 ). Besides, time scales should be considered, as the order of species' arrival to the niche can result in significant differences in the mature community structure. Early colonizers often have an advantage because they can use resources before other microorganisms' appearance or because they can produce physical barriers or antibiotics that slow the colonization of subsequent microorganisms (Hu et al., 2020 ). Thus, in general, microbial species introduced early into communities are more likely to persist, controlling the assembly of latecomer species (Werner and Kiers, 2015 ; Sprockett et al., 2018 ; Toju et al., 2020 ). Finally, a proper plant model should be selected, depending on the purpose of the study. If the aim is to explore the relationship of the plant ontogeny with the microbiota, an already well-described plant should be appropriate (Chaparro et al., 2013 , 2014 ). For agronomical purposes, the yield of the crop might be considered. If the environmental effects are being evaluated, such as abiotic stress caused by saline, poor, acidic, or alkaline soils, one might want to explore the holobiont of native plants from these environments. When the aim is to study the underlying molecular mechanisms, it is better to connect this knowledge with plant genetics, biochemistry, metabolism, and molecular physiology (Poupin et al., 2013 , 2016 ; Lebeis et al., 2015 ; Ledger et al., 2016 ; Finkel et al., 2019 ). Finally, if the aim is to study the microbiota solely as a biological community, comparative studies may be suitable (Zhang et al., 2019 ), and sometimes employing a plant model may not be essential. Here, high-throughput technologies, bioinformatics, and predictive models of microbial samples associated with plants can replace or complement plant-growth systems."
} | 6,990 |
28289729 | PMC5340861 | pmc | 5,052 | {
"abstract": "In this study, we show how invasive plant species drive rapid shifts in the soil environment from surrounding native communities. Each of the three plant invaders had different but consistent effects on soils. Thus, there does not appear to be a one-size-fits-all strategy for how plant invaders alter grassland soil environments. This work represents a crucial step toward understanding how invaders might be able to prevent or impair native reestablishment by changing soil biotic and abiotic properties.",
"conclusion": "Conclusions. Our hypothesis that plant invasions would cause reproducible and possibly invader-specific shifts in soil biotic and abiotic properties was supported. Soil nutrient availabilities differed among invaders, whereas microbial life histories shifted according to plant functional groups—possibly mediated by altered resource allocations. Overall, these changes in the soil environment are likely to contribute to the hysteresis we see in these systems, where it is very difficult to reestablish native vegetation (i.e., invasion legacies). Successional timescales are key for restoration of invaded grasslands ( 49 ), and our results indicate that early intervention (<3 years after establishment of invader) is crucial to prevent invasion-mediated alterations in soil chemistry and soil microbial communities. Future work should focus on restoration strategies that prevent or reverse these belowground shifts to disrupt the invader-dominated state.",
"introduction": "INTRODUCTION A major issue affecting grassland ecosystems worldwide is the introduction of exotic plant species ( 1 , 2 ), which is often associated with decreased plant community diversity and increased net primary productivity ( 1 , 3 ). The increased productivity of invaders may be due to lower predation or disease rates ( 4 ) or to an ability to access and use resources more efficiently than the native plant community ( 5 ). Millions of acres of grasslands in the Rocky Mountain West are dominated by noxious Eurasian weeds, such as spotted knapweed ( Centaurea stoebe ; perennial forb), leafy spurge ( Euphorbia esula ; perennial forb), and cheatgrass ( Bromus tectorum ; annual grass). Part of the success of these invaders is due to their expanded temporal niche breadth relative to native plants in the region ( 6 – 8 ), but it may also result from persistent invasion-mediated shifts in the biotic and abiotic soil environment. These shifts can complicate ecological restoration ( 9 , 10 ), and management strategies that suppress one invader often result in the establishment of a second invader ( 11 ). Even in the absence of direct competition from invasive plants, diverse native communities are difficult to restore in soils that once supported invasive plants ( 12 ). Soil microbial community composition has been shown to influence plant community diversity, productivity, and stability ( 13 – 15 ). Interactions between soil microbes and invaders have received more attention recently ( 16 ), but much remains unknown. For example, many previous studies are limited to single invaders ( 10 , 14 ), are based only on field surveys ( 17 , 18 ), and/or look at coarse-grained (e.g., pathogen versus mutualist) microbial communities ( 19 ), which complicate generalizations of invader effects. As a result, we have a limited understanding of the potential differences among invaders as well as the successional timescales of interactions between aboveground and belowground factors that may lead to invasive soil legacies ( 20 ). To better understand how invaders reshape the belowground environment, we conducted three independent studies. First, we surveyed spatially replicated field plots to determine whether forb and grass invaders are associated with consistent changes in abiotic and biotic soil properties across the landscape. We sampled communities invaded by leafy spurge, spotted knapweed, and cheatgrass, along with adjacent native plant communities. We collected a data set that encompassed the entire ecosystem: vegetation, edaphic properties, soil bacterial and fungal community composition (i.e., 16S rRNA gene and internal transcribed spacer [ITS] region amplicon sequencing) and microbial functional potential (i.e., shotgun metagenome sequencing). Second, for each plant invader, we sampled naturally occurring spatial gradients from invader-dominated to native-dominated communities. The goal of this study was to assess whether the effect of the invader would be more pronounced near the center of an established invasion where the exotics have likely had more time to influence the soil. Finally, to assess causation and to better understand the timescales over which soil legacies might develop, we sampled from a common garden where replicate plots of monodominant invaders and plots with mixtures of native plants were grown under controlled conditions for 3 years. We propose the following two hypotheses. (i) Independent invasions are associated with consistent species-specific soil characteristics that differ from surrounding native plant communities. (ii) Invaders are responsible for causing belowground changes, rather than simply being recruited to sites with preexisting characteristics. Indeed, we demonstrate that invaders rapidly cause species-specific shifts in edaphic properties and that these alterations subsequently drive changes in soil microbial community structure and function, which in turn may reinforce invasive soil legacies.",
"discussion": "DISCUSSION Invaders are associated with consistent changes in soil biotic and abiotic properties. We found that invasive plants can push native grassland soils into invader-specific ecological states that are consistent across sites. Cheatgrass, leafy spurge, and spotted knapweed invasions reduced native plant diversity, likely due to competitive interactions ( 24 ), although unlike some previous findings ( 1 , 3 ), this was not associated with substantial and consistent increases in productivity. Like previous findings, we found consistent, and often invader-specific, differences in soil chemistry ( Fig. 2 ). For example, leafy spurge plots showed elevated pH and nitrate levels relative to native plots, supporting prior work at this same study location ( 25 ) and elsewhere ( 3 , 26 , 27 ), whereas cheatgrass plots were enriched in phosphate but depleted in most other nutrients relative to native plots. Prior studies suggest variable correlations between above and belowground alpha-diversity ( 25 , 28 – 30 ). Invasive plots showed significantly reduced plant community richness (see Table S1 in the supplemental material). However, we found no relationship between invasive plant prevalence and prokaryotic or fungal community alpha-diversity, with the exception of phylogenetic diversity (PD) ( Table S1 ). The higher PD in cheatgrass plots may suggest phylogenetic overdispersion, which might be indicative of increased resource competition ( 31 ). Prior work has suggested that cheatgrass is a poor arbuscular mycorrhizal fungus (AMF) host, and thus does not likely allocate much carbon belowground ( 25 ). Leafy spurge and spotted knapweed, on the other hand, are both highly mycotrophic forbs ( 14 , 32 , 33 ), and higher respiration rates and aboveground biomass in spotted knapweed experimental plots and leafy spurge invasion gradients, respectively, suggest a potential for greater belowground carbon allocation and/or turnover ( 34 ) relative to native communities and cheatgrass invasions. There were no large-scale shifts in microbial community structure across plant community types, which could be due to a combination of low biological signal and potentially high technical noise associated with sequencing data. However, at a higher-resolution level, different plant functional groups (grasses versus forbs) did show different effects on soil microbial composition and diversity ( 25 , 35 – 38 ). As expected, spotted knapweed and leafy spurge tended to enrich for copiotrophic bacterial taxa (e.g., Bacteroidetes , Firmicutes , and Proteobacteria ), while oligotrophs were often depleted (e.g., Verrucomicrobia and Acidobacteria ) ( 22 , 39 – 41 ). Concordantly, we found that organisms with higher rRNA copy number, indicative of fast-growing copiotrophs ( 42 – 44 ), were enriched in forb-invaded soils. Stimulation of copiotrophs may have a soil-priming effect that would allow the microbial community to unlock nutrients from more recalcitrant soil organic matter ( 45 ). The higher pH found in leafy spurge and spotted knapweed plots may also contribute indirectly to enhanced SOM degradation ( 46 ). Higher N -acetylglutamine rates and the enrichment of ammonia oxidizers and nitrogen metabolism genes in leafy spurge plots corresponded with greater nitrate concentrations ( Fig. 2 ). Spotted knapweed invasions exhibited an increased prevalence of genes involved in organic matter catabolism, which is consistent with potentially higher respiration rates in these plots. Soil priming could explain the higher nutrient levels in forb-invaded plots. If this were the case, invasive plants could fundamentally alter the soil environment by reshaping the distribution of life history strategies among soil microbes. This soil priming hypothesis fits well with the greater nutrient availabilities across many different types of invasions and may be a general mechanism for invasive soil legacy establishment ( 3 , 26 ). This increase in nutrient availability may also help explain the invasion melt-down phenomenon ( 11 ). For example, the increase in nitrate within leafy spurge plots probably contributes to the greater prevalence of cheatgrass in these plots, because cheatgrass is a superior competitor to native plants under high-nutrient conditions ( 47 , 48 ). Invaders cause belowground changes over multiyear timescales. It is difficult to distinguish between whether the variation in soil chemistry between field sites or across gradients existed prior to invasion or whether the invader caused these differences. However, invasion gradients correlated with soil chemical and microbial shifts, which were most pronounced at the center of a mature invasion and less pronounced at the fringes ( Fig. 5 and 6 ). Soils near the invaded end had putatively been exposed to the exotic plants for longer than samples at the fringe of the invasion. This potentially causal influence was also supported by the common garden experiment, where experimental plots showed consistent shifts in soil chemistry ( Fig. 7 ). Invader-associated changes in nitrate concentrations were evident in the experimental plots after 1 year of growth. Shifts in soil chemistry became more prominent after 3 years ( Fig. 7 ). Many of the changes in belowground properties were concordant with our survey results, despite large differences in initial soil chemistry (much higher nutrients in the experimental garden site due to past fertilization compared to field soils) and the intense herbivore pressure on leafy spurge from flea beetles in the experimental plots. It is noteworthy that after 3 years, we did not observe any significant shifts in overall bacterial and fungal communities, suggesting that perhaps rapid invader-mediated shifts in soil abiotic properties drive subsequent shifts in biotic properties over the longer term. Conclusions. Our hypothesis that plant invasions would cause reproducible and possibly invader-specific shifts in soil biotic and abiotic properties was supported. Soil nutrient availabilities differed among invaders, whereas microbial life histories shifted according to plant functional groups—possibly mediated by altered resource allocations. Overall, these changes in the soil environment are likely to contribute to the hysteresis we see in these systems, where it is very difficult to reestablish native vegetation (i.e., invasion legacies). Successional timescales are key for restoration of invaded grasslands ( 49 ), and our results indicate that early intervention (<3 years after establishment of invader) is crucial to prevent invasion-mediated alterations in soil chemistry and soil microbial communities. Future work should focus on restoration strategies that prevent or reverse these belowground shifts to disrupt the invader-dominated state."
} | 3,073 |
24809026 | PMC4010749 | pmc | 5,053 | {
"abstract": "Bacteria frequently acquire novel genes by horizontal gene transfer (HGT). HGT through the process of bacterial conjugation is highly efficient and depends on the presence of conjugative plasmids (CPs) or integrated conjugative elements (ICEs) that provide the necessary genes for DNA transmission. This review focuses on recent advancements in our understanding of ssDNA transfer systems and regulatory networks ensuring timely and spatially controlled DNA transfer ( tra ) gene expression. As will become obvious by comparing different systems, by default, tra genes are shut off in cells in which conjugative elements are present. Only when conditions are optimal, donor cells—through epigenetic alleviation of negatively acting roadblocks and direct stimulation of DNA transfer genes—become transfer competent. These transfer competent cells have developmentally transformed into specialized cells capable of secreting ssDNA via a T4S (type IV secretion) complex directly into recipient cells. Intriguingly, even under optimal conditions, only a fraction of the population undergoes this transition, a finding that indicates specialization and cooperative, social behavior. Thereby, at the population level, the metabolic burden and other negative consequences of tra gene expression are greatly reduced without compromising the ability to horizontally transfer genes to novel bacterial hosts. This undoubtedly intelligent strategy may explain why conjugative elements—CPs and ICEs—have been successfully kept in and evolved with bacteria to constitute a major driving force of bacterial evolution.",
"conclusion": "Conclusion and outlook Although the molecular details of how regulatory networks control tra gene expression are different in the conjugation systems presented in this review, there is a common theme: As a default, tra genes are OFF and whenever positive stimuli are present, not the whole population transits to the ON stage but only a fraction of the cells carrying a conjugative element. In this way the metabolic burden (fitness cost) imposed by expression of tra genes and assembly of a cell envelope localized DNA secretion machine (a T4SS) is carried not by the whole population but distributed to only a few cells within a population. Further studies at the single cell level are needed to reveal whether the transformation of only a fraction of a donor cell population into transfer competent cells is due to a stochastic process or depends on different physiological states such as metabolic conditions, cellular fitness and cell age. Moreover, positioning of individual cells in structured communities (microcolonies or biofilms) may influence transition to transfer competence. Undoubtedly intelligent strategies exist to minimize or even eliminate fitness costs associated with the carriage of conjugative elements. Populations harboring CPs (and presumably ICEs) can grow and divide largely unaffected by the presence of these elements. At the same time, some cells within a population do become transfer competent and thereby secure the spread and persistence of conjugation modules in many different bacterial species, among them pathogens causing disease in humans, animals, and plants. Thus, genes carried on the conjugative element, which are beneficial for the host cell in particular habitats (e.g., antibiotic resistance genes), are likely to persist in bacterial populations even without continuous selective pressure. Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.",
"introduction": "Introduction Bacterial conjugation is important not only for bacterial evolution, but also for human health since it represents the most sophisticated form of HGT in bacteria and provides, for instance, a platform for the spread and persistence of antibiotic resistance genes (Norman et al., 2009 ). To efficiently counteract the problems associated with antibiotic resistance it is therefore necessary to understand the mobile genetic elements—conjugative plasmids (CPs) and integrative conjugative elements (ICEs)—that are the vehicles for transfer of antibiotic resistance genes from the large communal gene pool to human pathogenic bacteria. In the following sections we will give an overview on the current knowledge of bacterial conjugation. As will be evident, it is a widely distributed, if not ubiquitous phenomenon in the bacterial world. Special emphasis will be given to regulatory mechanisms ensuring timely and spatially controlled expression of tra genes. Furthermore, we consider recent advancements in understanding population dynamics and coevolution of CPs and host cells. In the context of this manuscript intelligence is understood as cell-cell communication and complex regulatory systems producing cellular responses that maximize successful DNA transmission and at the same time do not impose a burden (or fitness cost) to the whole population of CP carrying cells. Bacterial conjugation modules Bacterial conjugation is a cell-cell contact dependent DNA transfer event. Either dsDNA or ssDNA molecules are transported from donor to recipient bacterial cells. The transfer of dsDNA depending on one single dedicated protein (an FtsK like ATPase) is found only in Actinobacteria (Vogelmann et al., 2011 ; Thoma and Muth, 2012 ) and will not be considered further in this review. ssDNA transfer on the other hand is ubiquitous in the bacterial and archebacterial world and relies on a dedicated cell envelope spanning DNA transfer machinery ancestral to T4SS (type IV secretion systems) which translocate virulence determining effector proteins into target eukaryotic cells (Bhatty et al., 2013 ; Guglielmini et al., 2013 ). Approximately 10–20 proteins (fewer in Gram positive bacteria, see below) constitute the building blocks of the T4SS dedicated to ssDNA and protein transfer. The T4S machinery and additional proteins required for DNA transfer and replication (De la Cruz et al., 2010 ) are encoded by CPs or ICEs (Smillie et al., 2010 ; Guglielmini et al., 2011 ). Other genetic elements such as mobilizable plasmids or genomic islands can be mobilized by either of these self-transmissible elements (Smillie et al., 2010 ; Puymège et al., 2013 ). Unlike in true bacterial conjugation where DNA is transferred directly from a donor to a recipient cell, Neisseria gonorrhoeae secretes ssDNA contact-independently via a T4SS encoded by a genomic island (Ramsey et al., 2011 ). Conjugative plasmids (CPs) and integrated conjugative elements (ICEs) Historically, research on bacterial conjugation focused on the F-plasmid and related CPs from Gram negative bacteria (Willetts and Skurray, 1980 ; Frost et al., 1994 ). The interest shifted to broad host range CPs such as RP4 and R388 which encode a DNA transfer system more similar to the T-DNA transfer machinery encoded by virB genes of Ti (tumor inducing) plasmids of Agrobacterium tumefaciens (Eisenbrandt et al., 1999 ; Gomis-Rüth et al., 2001 ; Schröder and Lanka, 2003 ; Cascales and Christie, 2004 ). For structural studies on the T4S machinery plasmid pKM101 has had a pivotal role since it was possible to determine the 3D structure of a ring like core T4S complex composed of 14-mers of three proteins which span the periplasmic space from the inner to the outer membrane (Rivera-Calzada et al., 2013 ). From genomic sequencing projects and bioinformatics analyses it became evident that the most abundant self transmissible elements are ICEs that are maintained chromosomally similarly to temperate bacteriophages and can be transferred via a plasmid intermediate (Wozniak et al., 2009 ; Guglielmini et al., 2011 ). A schematic comparison of how CPs and ICEs are maintained and transferred is depicted in Figure 1 . Figure 1 Conjugative plasmids (CPs) and integrative conjugative elements (ICEs) . Events leading to horizontal transfer of CPs (A) or ICEs (B) are shown schematically. Before transfer can occur, tra genes must be expressed and a T4SS assembled. After cell-cell contact formation, transfer competent donor cells initiate a rolling circle type replication from circular dsDNA and translocate ssDNA via the T4S machinery into recipient cells. dsDNA is then reconstituted in the recipient (dotted inner circle). (A) CPs can autonomously replicate due to the presence of rep genes. (B) ICEs replicate as integrated elements with the host chromosome (green lines), integration and excision is mediated by int/xis genes required for integration and excision by site-specific recombination via attachment sites (vertical bars). After excision and before integration, ICEs are present in a plasmid-like dsDNA form. ssDNA transfer in gram positive and gram negative bacteria ssDNA transfer through T4S machineries has been explored in detail for Gram positive and Gram negative bacteria and excellent reviews describing and comparing these systems have been published recently (Bhatty et al., 2013 ; Goessweiner-Mohr et al., 2013 ). Only a subset of proteins typically found in Gram negative bacteria is also present in the Gram positives which led to the concept of minimized T4SS that are present in Gram positive bacteria (Zhang et al., 2012 ; Bhatty et al., 2013 ; Goessweiner-Mohr et al., 2013 ). Major differences arising from the specific architecture of the cell envelope of diderms vs. monoderms are: The presence of a more complex T4SS spanning two membranes (including the periplasm and a thin peptidoglycan layer) with a cell-surface attached filamentous pilus composed of multiple subunits of a single protein in Gram negative bacteria; a minimized T4SS for translocating ssDNA across the cytoplasmic membrane with a dedicated peptidoglycan hydrolase for local digestion of the thick cell wall and adhesins that mediate cell-to-cell contact in Gram positive bacteria (Bhatty et al., 2013 ; Goessweiner-Mohr et al., 2013 ). Components and functions of ssDNA transfer machines ssDNA is generated in the donor cell by proteins that can initiate a rolling circle type replication by nicking (cleaving) one strand of the dsDNA at a site termed oriT (origin of transfer). The nucleoprotein complex consists of the nicked plasmid DNA and the proteins required for DNA transfer and replication (also termed Dtr, usually a relaxase/ helicase and auxiliary proteins). Presumably, at this stage, the Dtr complex is docked to the T4S complex which has been pre-assembled in the cell envelope (Zechner et al., 2012 ). The T4S complex consists of (i) ATPases fueling assembly of the T4S apparatus and DNA transfer, (ii) translocon proteins of the inner membrane, (iii) core proteins spanning the cell envelope, and (iv) pilus proteins or adhesins. The Dtr complex physically interacts with the T4S apparatus mainly via protein-protein interactions especially via one of the ATPases of the T4S complex being a substrate receptor (Bhatty et al., 2013 ). In order to start translocating ssDNA, a productive and stable mating pair between a donor and a suitable recipient cell has to be formed. This includes initial contact via the pilus or adhesins, pilus retraction in F conjugation (Clarke et al., 2008 ), and the formation of larger contact zones that have been observed in different conjugation systems (Dürrenberger et al., 1991 ; Samuels et al., 2000 ; Lawley et al., 2002 ). It is not known whether the pilus additionally functions as a device delivering ssDNA by penetration of the recipient cell envelope. Upon an elusive signal, ssDNA with the relaxase covalently bound to the 5 prime end of the ssDNA is transported through the conjugation channel (the T4S apparatus) and reaches the cytoplasm of the recipient where the DNA is recircularized (presumably via the co-transported relaxase) to regenerate a circular ssDNA which can be replicated to dsDNA in the recipient (Zechner et al., 2012 ). Establishment of the ds plasmid DNA in the recipient is aided by ssDNA binding, anti-restriction and SOS inhibition proteins, usually encoded by “leading region” genes which are among the first to enter the recipient cell (Althorpe et al., 1999 ; Wilkins, 2002 ). Overall, conjugative DNA replication is similar to the replication of ssDNA phages in which ssDNA (in that case termed the plus strand) is generated by rolling circle replication from a dsDNA intermediate, packaged into viral proteins and then released into the environment, ready to infect novel recipient cells. Transfer of plasmid DNA into cells already containing the same CP is prevented by blocking cell-to-cell contact formation and entry of the ssDNA into the recipient cell, mechanisms that are termed “surface exclusion” and “entry exclusion,” respectively (reviewed in Garcillán-Barcia and de la Cruz, 2008 )."
} | 3,223 |
40271349 | PMC12014621 | pmc | 5,054 | {
"abstract": "Chloroplasts are critical organelles in plants and algae responsible for accumulating biomass through photosynthetic carbon fixation and cellular maintenance through metabolism in the cell. Chloroplasts are increasingly appreciated for their role in biomanufacturing, as they can produce many useful molecules, and a deeper understanding of chloroplast regulation and function would provide more insight for the biotechnological applications of these organelles. However, traditional genetic approaches to manipulate chloroplasts are slow, and generation of transgenic organisms to study their function can take weeks to months, significantly delaying the pace of research. To develop chloroplasts themselves as a quicker and more defined platform, we isolated chloroplasts from the green algae, Chlamydomonas reinhardtii , and examined their photosynthetic function after extraction. Combined with a metabolic modeling approach using flux-balance analysis, we identified key metabolic reactions essential to chloroplast function and leveraged this information into reagents that can be used in a “chloroplast media” capable of maintaining chloroplast photosynthetic function over time ex vivo compared to buffer alone. We envision this could serve as a model platform to enable more rapid design-build-test-learn cycles to study and improve chloroplast function in combination with genetic modifications and potentially as a starting point for the bottom-up design of a synthetic organelle-containing cell.",
"introduction": "1 Introduction Chloroplasts are critical organelles in plant and algal cells, playing key roles in metabolism and cellular maintenance. Beyond their well-known function in carbon fixation through photosynthesis, chloroplasts are increasingly recognized for their involvement in diverse metabolic pathways, such as lipid biosynthesis ( Kunst et al., 1988 ) and amino acid production ( Reyes-Prieto and Moustafa, 2012 ; Chen et al., 2018 ). These pathways make chloroplasts valuable not only for agriculture, where they can enhance crop yield ( Galmes et al., 2014 ), but also for biomanufacturing, as they have been used to produce biofuels ( Georgianna and Mayfield, 2012 ), pharmaceuticals ( Khan et al., 2018 ), and high-value compounds such as vitamins and pigments ( Gedi et al., 2017 ; Bock, 2014 ; Maliga and Bock, 2011 ). Despite their importance, studying chloroplast function through genetic approaches is laborious and time-consuming, often requiring many weeks to generate mutants even within relatively simple and well-defined model organisms such as the single celled green algae Chlamydomonas reinhardtii ( Wang et al., 2023 ). Additionally, dissecting specific biochemical pathways within chloroplasts can be challenging due to the complexity and “noise” of the cellular environment. These limitations hinder progress in both basic research and the development of biotechnological applications related to chloroplasts. \n In vitro systems potentially offer a more controlled approach to study chloroplast function, removing the interference from other cellular processes. Maintaining chloroplast functionality outside of their native cellular context is difficult though, as they rapidly lose photosynthetic activity after isolation over the course of a few days ( Green et al., 2005 ). However, there is evidence that they can be maintained in a prolonged viable state in non-native contexts. Notably, various sacoglossan sea slugs have demonstrated the ability to “steal” chloroplasts from their algal food sources, termed kleptoplasty, and maintain them in a photosynthetically active state for weeks to months inside their own cells ( Cruz and Cartaxana, 2022 ). Intrigued by this, we posited that there must exist a set of conditions under which chloroplasts can artificially be kept photosynthetically functional outside of their host cell. Here, we leveraged the model organism C. reinhardtii and computational modeling to simulate key metabolic networks within chloroplasts and develop an initial defined chemical medium for maintaining chloroplast function ex vivo . We performed flux balance analysis (FBA) using a system-scale model of the C. reinhardtii chloroplast ( Bjerkelund Rokke et al., 2020 ) to analyze single-reaction knockout phenotypes. Through these simulations, we identified essential reactions and metabolites spanning several classes of molecules, notably amino acids, nucleotides, and magnesium ions. Noting the importance of protein synthesis in chloroplast function ( Sundby et al., 1993a ), we expanded the medium to involve components commonly used in in vitro transcription/translation buffers and showed improved maintenance of photosynthetic function of chloroplasts ex vivo compared to storage in sorbitol cushioned HEPES buffer alone.",
"discussion": "3 Discussion The world population is expected to grow beyond nine billion by 2050 ( United Nations Department of Economic and Social Affairs and Population Division, 2024 ), increasing demands on food production and energy infrastructure. Improvements to existing crop yields and alternative energy sources utilizing bioenergy are prospective solutions to these challenges. Central to both processes are chloroplasts, which are mature plastids found in plants and algae that perform carbon fixation through photosynthesis, among other metabolic processes. Chloroplasts have become increasingly appreciated for their potential as biotechnology workhorses, proving instrumental in producing biomolecules for therapeutics ( Specht et al., 2010 ) and lipids used in biofuel production ( Nawkarkar et al., 2020 ; Cecchin et al., 2020 ). However, traditional methods of studying and manipulating chloroplasts within cells rely on labor-intensive techniques, often requiring many generations to achieve homoplasy, significantly slowing down research progress and hindering rapid experimental cycles. Our study sought to approach this issue from a different angle and take the initial steps towards developing an ex vivo chloroplast platform. Isolated chloroplasts provide a defined system that minimizes cellular complexity, yet their rapid functional decline post-extraction remains a major challenge. We took inspiration from Sacoglassan sea slug kleptoplasty, which can maintain photosynthetically active chloroplasts from their algal sources for months after consumption ( Cruz and Cartaxana, 2022 ). In fact, many species of unicellular organisms have displayed the ability to maintain active chloroplasts in completely non-native contexts (reviewed in ( Miyagishima, 2023 )), and we hypothesized that artificially defined environments could also prolong ex vivo chloroplast function. Here, we successfully isolated chloroplasts from C. reinhardtii and identified a set of metabolites that could slow the decline of photosynthetic function after extraction. We used a C. reinhardtii chloroplast metabolic model and performed flux balance analysis (FBA) across the known metabolic reactions present in the model. Modeling single-reaction knockouts and optimizing for ATP production and biomass accumulation in the chloroplast, we identified a list of metabolic reactions and essential metabolites imported into chloroplasts and developed our initial media recipe. While our initial media (EM) did not significantly maintain chloroplast function at 30°C, we were encouraged by the findings. We noticed an abundance of transcription and translation related components in EM media, which led us to infer several key points: 1) These components are crucial since photosynthesis itself damages the photosynthetic machinery within chloroplasts ( Aro et al., 1993 ; Tyystjarvi and Aro, 1996 ; Russell et al., 1995 ); 2) supporting protein synthesis and turnover is critical for maintaining function; and 3) CFPS reaction buffers could satisfy this requirement. CFPS enables protein production using cell extracts, and typical E. coli PANOx-SP ( Jewett and Swartz, 2004 ) CFPS buffers contain most of the components in EM media, along with additional amino acids, nucleotides, salts, and cofactors. When we incubated extracted chloroplasts in EEM media at 30°C over time, we observed significant delays in chloroplast functional decline through PAM measurements. Together with our morphological analyses, our data showed that chloroplasts in EEM media were less degraded and maintained better membrane integrity, suggesting that EEM media better supported their structure and function over time. However, like any computational approach, our results depend on the assumptions and accuracy of the underlying metabolic model. The model we used assumes an intact cell, with defined metabolic reactions in both cytosolic and chloroplast compartments. While the chloroplast genome encodes the most critical components of the photosynthetic machinery, most metabolic enzymes and proteins inside the chloroplast are nucleus encoded ( Gallaher et al., 2018 ). Isolation of chloroplasts from the host would preclude protein resupply and ensure a limited functional lifetime tied to their rates of degradation. Isolated chloroplasts are translation competent ( Leu et al., 1984 ; Mendiola-Morgenthaler et al., 1985 ) and could likely replace essential photosystem proteins such as D1 for a short duration, but indefinite sustained function ex vivo would ultimately require supplementation of nuclear-encoded factors. Nevertheless, we observed a measurable difference in photosynthetic function with EEM media over time at 30°C compared to HEPES buffer alone, suggesting that a purely biochemical formulation could still temporarily maintain function. Our analysis identified several additional components absent in both the EM and EEM media that could possibly improve future media formulations, including adenosyl-containing amino acids, UDP-linked sugars, and fatty acid cofactors. S-adenosyl-homocysteine (AdoCys) can be converted into S-adenosyl-L-methionine (AdoMet), which serves as a universal methyl donor for methylation reactions and is a precursor for polyamine synthesis, which are critical for cell growth, division, and lipid accumulation ( Tehlivets et al., 2013 ; Chiang et al., 1996 ). Although AdoCys and AdoMet are absent in the media, downstream polyamines like putrescine and spermidine are present in the EEM media, potentially compensating for this absence. External polyamines, such as putrescsine and spermidine, are known to influence C. Reinhardtii growth and may be taken up in limited amounts ( Theiss et al., 2002 ; Freudenberg et al., 2022 ), though whether extracted chloroplasts retain these uptake mechanisms remains unexplored. UDP-glucose, UDP-galactose, and the long- and very long-chain fatty acyl CoA’s we identified likely play important roles in lipid production, membrane maintenance, and fatty acid metabolism ( Warakanont et al., 2019 ; Yang et al., 2015 ). UDP-glucose and UDP-galactose are interconvertible, with UDP-galactose serving as a key component of two photosynthetic membrane galactolipids: Monogalactosyldiacylglycerol (MGDG) and digalactosyldiacylglycerol (DGDG) ( Boudiere et al., 2014 ; Li-Beisson et al., 2013 ). These lipids are highly conserved across photosynthetic organisms and can be produced entirely within chloroplasts via the “prokaryotic pathway” or in cooperation with the endoplasmic reticulum (ER) through the “eukaryotic pathway” which involves lipid trafficking between the chloroplast and ER. Both MGDG and DGDG serve critical structural and functional roles in thylakoid membranes and their associated photosystem complexes ( Wang et al., 2014 ). Although extracted chloroplasts in EEM media maintained higher photosynthetic yield than the other conditions, we still observed a noticeable increase in the amount of degraded or membrane-damaged chloroplasts by 8 h and a sharp drop-off in photosynthetic yield by 24 h at 30°C. Although we did not explore the role of these lipid components, these findings suggest that membrane integrity may be involved in ex vivo chloroplast longevity, and we speculate that supplementation of lipid precursors could further prolong chloroplast function. In this study, incubation of extracted chloroplasts in EEM media significantly delayed the decline of photosynthetic function compared to HEPES buffer and EM media. While the exact mechanism by which EEM media prolongs chloroplast function is unknown, we speculate the included components provide the necessary support for transcription, translation, and protein turnover within the chloroplast. Identifying media formulations that can sustain chloroplast functionality ex vivo for even brief periods could facilitate their direct manipulation and create a more efficient research and biomanufacturing ( Miller et al., 2020 ) platform. Refining the media formulation by incorporating additional metabolites, such as fatty acid cofactors and UDP-linked sugars, and/or combining it with genetic modifications or direct protein supplementation may further extend chloroplast functional lifetimes and potentially enable faster design-build-test-learn cycles for chloroplast research and biotechnological applications. Additionally, advancements in extension of chloroplast functional lifetimes outside of their host cells may support efforts in constructing synthetic organelle-containing cells from the bottom up."
} | 3,347 |
39106075 | PMC11340025 | pmc | 5,055 | {
"abstract": "The self-sustained\nmotion of fluids on gradient substrates\nis a\nspectacular phenomenon, which can be employed and controlled in applications\nby carefully engineering the substrate properties. Here, we report\non a design of a gel substrate with stiffness gradient, which can\ncause the spontaneous motion of a droplet along (durotaxis) or to\nthe opposite (antidurotaxis) direction of the gradient, depending\non the droplet affinity to the substrate. By using extensive molecular\ndynamics simulations of a coarse-grained model, we find that the mechanisms\nof the durotaxis and antidurotaxis droplet motion are distinct, require\nthe minimization of the interfacial energy between the droplet and\nthe substrate, and share similarities with those mechanisms previously\nobserved for brush substrates with stiffness gradient. Moreover, durotaxis\nmotion takes place over a wider range of affinities and is generally\nmore efficient (faster motion) than antidurotaxis. Thus, our study\npoints to further possibilities and guidelines for realizing both\nantidurotaxis and durotaxis motion on the same gradient substrate\nfor applications in microfluidics, energy conservation, and biology.",
"conclusion": "Conclusions In this study, we have\nproposed and investigated\na novel substrate\ndesign based on a gel material. Importantly, we have been able to\ndemonstrate that durotaxis and antidurotaxis motion of a droplet is\npossible on the same substrate and the direction of motion only depends\non the fluid. To our knowledge, this is the first time that this possibility\nis realized for gel substrates. As in the case of durotaxis onto brush\nsubstrates, 24 , 25 we have found that the minimization\nof the interfacial energy between the droplet and the substrate is\nthe dominant driving force responsible for the motion of the droplet.\nThis takes place by the substantial penetration of the substrate by\nthe droplet in the case of antidurotaxis or the droplet motion toward\nareas with smaller surface fluctuations on the top of the gel in the\ncase of durotaxis. As a result, the trajectories of the droplet motion\nappear to be more diffusive in the durotaxis cases than in the antidurotaxis\ncases, where in the latter the droplet motion is hindered by the gel\nunits. Moreover, recent experiments 68 have\nreported on the spontaneous droplet motion on soft, gel substrates\nwith stiffness gradient created by varying the degree of cross-linking\nin the gel. In this case, results have pointed out to the minimization\nof the interfacial energy between the substrate and the droplet as\nthe driving force for the durotaxis motion of the droplet, as in the\ncase of simulation experiments here and in previous studies. 17 , 24 , 25 We have also found that durotaxis\ntakes place for a wide range of droplet–substrate affinities\nwith lower affinities leading to more efficient durotaxis motion,\nwhile fully successful antidurotaxis motion has only been observed\nfor a high value of droplet–substrate affinity. Our study\nprovides further evidence that both durotaxis and antidurotaxis\nmotion can be realized on the same gel substrate. Thus, we anticipate\nthat our work highlights the new venues of possibilities in the autonomous\nmotion of fluids based on gradient gel-substrates and provides insights\ninto the motion of droplets driven by stiffness gradients, thus enhancing\nour understanding of similar phenomena, encountered in nature.",
"introduction": "Introduction The autonomous motion of fluids on gradient\nsubstrates has been\nobserved in various contexts, for example, in the case of microfluidics,\nmicrofabrication, coatings, energy conversion, and biology. 1 − 13 Moreover, both the efficiency and the direction of motion can be\ncontrolled by carefully engineering the gradient of a substrate property.\nIn the case of moving cells on tissues, 11 , 12 , 14 − 16 their motion has been attributed\nto gradients in the stiffness of the underlying tissue, a phenomenon\nknown as durotaxis. Inspired by biological systems, efforts to foster\nnew possibilities of sustained motion on substrates with gradually\nchanging properties along a certain direction have taken place, in\nview of the spectrum of possible applications in diverse areas. This\nalso includes nano-objects of different type (e.g., fluids, nanosheets)\non a wide range of different substrates, which have been studied in\nthe context of theoretical and simulation work, 17 − 26 as well as experiments. 27 , 28 The exciting\naspect of durotaxis is the autonomously sustained\nmotion, that is no energy supply from an external source is required\nfor setting in and sustaining the motion of the nano-object. While\nin connection with durotaxis, a gradient in the stiffness is responsible\nfor the motion, such motion can actually be observed in other scenarios\nas well, for example, when the gradient reflects changes in the pattern\nof the substrate. Here, a characteristic example is rugotaxis, where\na fluid motion is caused by a gradient in the wavelength characterizing\na wavy substrate. 28 , 29 Other examples include curvotaxis,\nthat is motion attributed to curvature changes, such as that observed\nin the context of curved protein complexes at the cell. 30 Further possibilities, include small condensate\ndroplets that can move due to the presence of asymmetric pillars, 31 three-dimensional (3D) capillary ratchets, 32 or pinning and depinning effects at the three-phase\ncontact line. 33 Interestingly, in the case\nof capillary ratchets, the surface tension can play a role in determining\nthe direction of motion, whether this is along or against the gradient. 32 In addition, substrates with wettability gradients\nhave been reported as a possibility for the autonomous motion of liquids, 34 − 36 for example, due to corrosion, 13 while\nlong-range transport has been realized by using electrostatic 37 , 38 or triboelectric charges. 39 In the presence\nof an external energy source, motion is also possible, with characteristic\nexamples being electrotaxis 40 and thermotaxis. 41 For example, in the latter case, the motion\nis caused by a temperature gradient that requires to be maintained\nalong the substrate by means of an external energy source. Further\nexamples of motion due to external sources include motion caused by\nelectrical current, 42 − 45 charge, 46 − 48 or even simple stretching. 49 Situations where droplets are chemically driven have also been reported\nin the literature, 50 , 51 as well as droplets on vibrated\nsubstrates 52 − 55 or wettability ratchets. 56 − 59 Motivated by relevant experiments with liquid\ndroplets, 27 , 28 we have previously proposed and investigated\nby computer simulation\nvarious substrate designs that can cause a sustained droplet motion. 17 , 24 , 25 , 29 More specifically, we have proposed two designs of brush substrates\nwith stiffness gradient that can cause such motion either along or\nagainst the gradient direction. 24 , 25 In the first design,\nthe brush substrate had a constant density of grafted polymer chains. 24 In this case, the stiffness gradient was a result\nof changes in the stiffness of the individual polymer chains along\nthe gradient direction. We have found that the droplet can move toward\nareas of higher stiffness (durotaxis), where a larger number of contacts\nbetween the droplet and the substrate can be established, due to a\nlower substrate roughness in these areas. In the second design of\na brush substrate, the grafted polymer chains were fully flexible\nand the stiffness gradient was imposed by changing the grafting density\nalong a particular direction. 25 In this\ncase, the droplet could move toward softer parts of the substrate\n(antidurotaxis), establishing more pair contacts as it penetrated\ninto the substrate. Interestingly, the latter antidurotaxis motion\nmight share similarities with experiments of droplets on soft substrates\nwith stiffness gradient, where droplet motion was also observed from\nstiffer toward softer areas of the substrate. 27 Moreover, in this case, larger droplets seem to perform antidurotaxis\nmotion more efficiently (faster), an effect that might not be attributed\nto gravity effects due to the weight of the droplet, as experiments\nwere carried out for micrometer-sized water droplets, i.e., smaller\nthan the capillary length (∼2.5 mm). Thus, far, experimental\nsubstrates 11 , 12 , 14 − 16 , 27 and simulation models 17 , 18 , 21 , 24 , 25 have mostly demonstrated either durotaxis\nor antidurotaxis motion\nfor a given substrate. Here, building upon our previous experience\nwith durotaxis and antidurotaxis droplet motion onto brush substrates, 24 , 25 we show that a novel gel substrate can demonstrate both antidurotaxis\nand durotaxis droplet motion depending on the type of liquid. To achieve\nthis result, a gradient in the bonding stiffness between the gel chemical\nunits is used in our model to create the stiffness gradient along\na specific direction of the gel substrate. Furthermore, by means of\nextensive molecular dynamics (MD) simulations of a coarse-grained\nmodel, we elucidate the mechanisms for both the durotaxis and antidurotaxis\nmotions and their efficiency for a range of parameters relevant for\nthis substrate design. Interestingly, we observe similarities for\nthese mechanisms with what we have previously seen for brush substrates. 24 , 25 Thus, this may point to more universal features of such substrates\nthat can cause durotaxis and antidurotaxis motion of fluids, and holds\nhope for the experimental realization of such substrates. In the following,\nwe provide details of the system, simulation model and methodology.\nThen, we will present and discuss the obtained results, while we will\ndraw the conclusions resulting from our investigations in the final\nsection.",
"discussion": "Results and Discussion Given the\nconstant gradient\nmaintained in each of our simulation\nexperiments, which is optimally chosen to facilitate our properties\nexploration, the first aspect of our research concerns the possibility\nof causing durotaxis or antidurotaxis motion and the probability of\nsuch motion for a range of droplet–substrate affinities. To\naddress this issue, a droplet is placed either on the softest or the\nstiffest part of the substrate and the outcome of the simulation is\nmonitored. Figure 3 visually summarizes our conclusions. For values ε dg < 0.2 ε, the interaction between the droplet and the gel\nsubstrate is weak. Hence, in this case the droplet detaches from the\nsubstrate due to the thermal fluctuations and this case deserves no\nfurther consideration here. Durotaxis motion takes place when 0.2\nε < ε dg < 0.8 ε. For this range\nof affinity strength between the droplet and the substrate, we observe\nthat the droplet moves from softer to stiffer parts of the gel substrate\ncovering its full length in the x direction, a manifestation\nof successful durotaxis motion for the droplet. While for 0.3 ε\n≤ ε dg ≤ 0.6 ε the probability\nthat the droplet successfully moves from the softest to the stiffest\nside of the substrate is 1.0 as calculated from an ensemble of five\ndifferent trajectories for each affinity case, this probability becomes\nless than unity when ε dg = 0.7 ε. Moreover,\nwe were able to only detect partial motion along the substrate, when\nε dg = 0.8 ε, reporting threrefore this case\nas unsuccessful. This provides a first indication that the droplet\nmotion may become less effective for larger values of ε dg . Indeed, this is corroborated by monitoring the average\nvelocity of the droplet for different values ε dg ( Figure 3 ), which clearly\nindicates that an increased affinity between the droplet and the substrate\nwill lead to a smaller average durotaxis velocity. Further increase\nof the affinity, namely ε dg = 0.9 ε, leads\nto successful antidurotaxis motion. In this case, the droplet reached\nthe softest part of the gel substrate and the recorded average velocity\nwas of the same magnitude as in the durotaxis case with ε dg = 0.7 ε. Finally, antidurotaxis motion for ε dg = ε was observed, but in this case the droplet was\nnot able to cover the full distance from the one to the other side\nof the gel substrate for any of our five trajectories and therefore\nthis case was considered unsuccessful, as was the case of partial\ndurotaxis droplet motion for ε dg = 0.8 ε. The\nabove observations may allow us to conclude that both durotaxis and\nantidurotaxis motions are possible on the same substrate. Since this\ntakes place by varying the droplet–substrate affinity in our\nsimulation, we may argue that the direction of motion eventually depends\non the choice of liquid for the droplet. Also, durotaxis motion on\ngel substrates is overall more efficient than the antidurotaxis motion,\nespecially when the droplet–substrate affinity is lower. Figure 3 Average speed\nof the droplet as calculated from successful durotaxis/antidurotaxis\nexperiments ( N s is the number of the successful\ncases) from a total ensemble of five ( N total = 5) trajectories for each case, as indicated. Inset shows the probability,\nζ = N s / N total , of the droplet moving from one side of the substrate to the other.\nThis probability for antidurotaxis cases is illustrated by purple\nbars, while that for the durotaxis cases by brown color. For ε dg < 0.8 ε durotaxis is observed, while antidurotaxis\nwas recorded for ε dg = 0.9 ε. For ε dg = 0.8 ε, only partial droplet motion was observed\nfrom each trajectory and therefore no successful cases are reported\nin the plot. As in our previous studies, 17 , 24 , 25 , 29 we attempted\nto identify the\ndriving force for both antidurotaxis and durotaxis cases. X in Figure 4 indicates the coordinate of the center-of-mass of the droplet in\nthe x direction with the zero value corresponding\nto the center of the gel substrate. Z is the coordinate\nof the center-of-mass of the droplet in the z direction\nwith the zero indicating the position of the substrate boundary, which\nwas determined through the inflection point in the density profile\nof each substrate as done in our previous work. 25 Moreover, the peculiarities of the gel–droplet interface\nhave been explored recently in detail. 67 On the basis of our analysis for the durotaxis cases, we observe\nthat the interfacial energy between the droplet and the substrate\ndecreases as a function of the center-of-mass position of the droplet\nin both the x ( Figure 4 a) and the z directions ( Figure 4 b), which suggests\nthat the droplet establish a larger number of contacts with the gel\nas it moves along the substrate (see also Movie\nS1 in the Supporting Information ).\nAs a result, the droplet is more strongly attracted by the gel as\nit moves toward the stiffer parts, which results in a decrease in\nthe position Z of the center-of-mass of the droplet,\nbut with the droplet however remaining on top of the substrate. Moreover,\nwe observe that the slope in the energy reduction of the interfacial\nenergy as a function of the position X of the center-of-mass\nof the droplet is larger for smaller values of the attraction strength\nε dg ( Figure 4 a), which reflects the conclusions relating to the average\nvelocity of the droplet presented in Figure 3 , that is a lower adhesion of the droplet\nto the gel substrate offers a more efficient (in terms of droplet\nspeed) durotaxis motion. This motion mechanism of the droplet shares\nsimilarities with the durotaxis motion previously observed on brush\nsubstrates, 24 where the droplet moves to\nthe areas of smaller surface fluctuations of the substrate, that is\nsubstrate parts of lower roughness. Figure 4 (a) Interfacial energy of the droplet\nnormalized by the number\nof substrate–droplet bead pairs as a function of the X coordinate of the center-of-mass of the droplet in the x direction for a range of different durotaxis cases with\ndifferent ε dg , as indicated. The vertical dashed\nline indicates the X position for the center-of-mass\nconsidered for determining the successful translocation of the droplet\ntoward the stiffest end of the substrate. (b) The normalized interfacial\nenergy is plotted against the coordinate of the center-of-mass of\nthe droplet in the z direction. The vertical dashed\nline denotes the position of gel’s surface, calculated by the\ninflection point of the density profile of the gel, as is done in\nour previous study. 25 The results of Figure 4 for the durotaxis cases can be compared\nwith those for the\nantidurotaxis cases presented in Figure 5 . Notably, we observe that the interfacial\nenergy is much more reduced for the antidurotaxis cases in comparison\nwith the durotaxis ones. More importantly, we also see that the droplet\npenetrates deeper into the substrate in the case of antidurotaxis\ndroplet motion and the center-of-mass of the droplet eventually lies\nbelow the top of the substrate as the antidurotaxis motion completes\n(see also Movie S2 of the Supporting Information ). This mechanism is therefore more\nsimilar to the one observed in the case of antidurotaxis motion for\nbrush substrates with gradient in the grafting density of the polymer\nchains. 25 In this case, the minimization\nof the interfacial energy was due to the penetration of the droplet\ninto the brush substrate. For this reason, the droplet motion is much\nless efficient than that in the case of durotaxis simulations, since\nthe droplet faces a larger resistance in carrying out the motion along\nthe substrate by bypassing the gel beads. Figure 5 (a) Interfacial energy\nof the droplet normalized by the number\nof substrate–droplet bead pairs as a function of the X coordinate of the center-of-mass of the droplet in the x direction for a range of different antidurotaxis cases\nwith different ε dg , as indicated. (b) The interfacial\nenergy is plotted against the coordinate of the center-of-mass of\nthe droplet in the z direction. Finally, we monitored the trajectories of the center-of-mass\nof\nthe droplet onto the x – y plane ( Figure 6 ).\nA different behavior of the droplet motion is observed between durotaxis\nand antidurotaxis cases. In particular, we see that the droplet motion\nis more influenced by thermal fluctuations as indicated by the lateral\nmotion in the y direction in the case of durotaxis\n( Figure 6 a). The droplet\nclearly initially moves at a higher instantaneous speed toward the\nstiffer areas and then slightly slows down. This pattern of motion\nis observed for both the lowest and the highest affinity between the\ndroplet and the substrate, which may indicate that the affinity might\nplay a lesser role in determining the exact trajectory of the particle.\nThe weakening of the gradient effect on the droplet velocity as the\ndroplet reaches the ever stiffer parts of the substrate has been thus\nfar observed in all previous durotaxis/antidurotaxis studies. 17 , 24 , 25 In the case of antidurotaxis\nexperiments ( Figure 6 b), the droplet appears to only move in the x direction\nwith minimal lateral (diffusive) motion in the y direction,\nwhich may suggest that the motion in this case is dominated by the\ndroplet–substrate interactions. This takes place to a larger\ndegree as the droplet moves to the softer parts of the substrate.\nHence, we can see that the droplet motion fundamentally differs in\nthe case of antidurotaxis and durotaxis cases, with the antidurotaxis\nmotion providing a more certain path for the trajectory of the droplet\nmoving along the substrate during the simulation experiments. Figure 6 Typical trajectories\nfor (a) durotaxis and (b) antidurotaxis cases\nfor values of ε dg , as indicated. The points of the\ntrajectory are collected by tracking the center-of-mass of the droplet\non the x – y plane."
} | 4,911 |
38542638 | PMC10972092 | pmc | 5,060 | {
"abstract": "For some so-called computationally difficult problems, using the method of Boolean logic is fundamentally inefficient. For example, the vertex coloring problem looks very simple, but the number of possible solutions increases sharply with the increase of graph vertices. This is the difficulty of the problem. This complexity has been widely studied because of its wide applications in the fields of data science, life science, social science, and engineering technology. Consequently, it has inspired the use of alternative and more effective non-Boolean methods for obtaining solutions to similar problems. In this paper, we explore the research on a new generation of computers that use local active memristors coupling. First, we study the dynamics of the memristor coupling network. Then, the simplified system phase model is obtained. This research not only clarifies a physics-based calculation method but also provides a foundation for the construction of customized analog computers to effectively solve NP-hard problems.",
"introduction": "1. Introduction Chua was the first to propose the concept of a memristor [ 1 , 2 ], and it was not until May and June 2008 that Nature published three consecutive articles, reporting the discovery of memristors [ 3 , 4 , 5 ]. HP Laboratories discovered that a two-layer titanium dioxide film, sandwiched between two platinum sheets, exhibited the characteristics of a memristor. This finding was the first to theoretically and experimentally confirm the physical existence of nano memristors, causing great shock in both the industry and academia. Thus, memristors have become a new hot research field [ 6 ]. Memristors are a type of memory-based nonlinear resistor with nanoscale dimensions, precisely adjustable resistance, nonvolatility, and low power consumption characteristics. Their voltage–current pinched hysteresis characteristics and frequency dependence on periodic excitation signals are the main distinguishing features of memristors [ 7 ]. Current research indicates that memristors have extremely important application potential in fields such as non-volatile memory, artificial neural networks, logic circuits, and nonlinear circuits. The direct application of a binary nonvolatile memristor is to build a new generation of nonvolatile resistive memory (ReRAM), which has the advantages of scalability, low power consumption, high density, and compatibility with CMOS. It is the preferred alternative for future Flash, SRAM, and DRAM. Research indicates that, in the future, ReRAM can be scaled down to below 10 nm, and the development of a 20 nm 1 Gb 2-layer 3D ReRAM has already been achieved. The density of 3D ReRAM can exceed that of 2D and 3D flash memory [ 8 ]. The cross-array of 8 nm × 8 nm memristors has been reported [ 9 ], and the development of 1 nm-level memristors is anticipated [ 10 ]. The exponential growth in data volume and computing demand has led to the Von Neumann bottleneck [ 11 ] in traditional storage computing architectures, and computing technologies driven by transistor density are gradually reaching their physical limits. The emergence of neural morphology computing, which uses neural morphology devices to simulate the behavior of neurons and synapses in the human brain to process information, has become the best candidate for the new generation of computing architecture due to its high parallelism, extremely low power consumption, and integrated storage and computing advantages. The nonvolatile and adjustable resistance characteristics of memristors can emulate the memory and weight regulation behavior of synapses, while local active memristors can emulate various firing characteristics of neurons [ 7 , 8 , 9 , 10 ], and the nanostructure of memristors can make neural networks highly integrated. Therefore, memristors can directly emulate the behavior of neurons and synapses from a physical level [ 12 , 13 ], naturally achieving a computing architecture that integrates memory and computing, which precisely meets the needs of neural morphology devices and has become one of the best candidates for current neural morphology devices. Binary memristors have “0” and “1” logical states, which can be used to implement logical operations. Memristive logic circuits are the most promising alternative computing solution for traditional integrated circuits. Using memristors for both memory and processing functions, can break the computational framework of von Neumann’s separation of memory and computation, thus achieving a new architecture of integrated memory and computation. This is a more optimized approach for the new generation of computing machines. Logic circuits based on memristors mainly include memristor/CMOS hybrid logic circuits, memristor logic circuits, logic operation circuits based on programmable nanowire technology, embedded logic circuits, etc. [ 10 , 11 ]. In addition to the above applications, memristors can also be applied to amplifiers, filters, and nonlinear oscillation and chaos circuits [ 14 ]. Memristors provide another new development space for circuit design, especially in the field of memristive chaotic oscillators. Special phenomena, such as infinite equilibrium points (or line, plane, or even three-dimensional space equilibrium points), coexisting chaotic attractors (or parameter-free bifurcations), hidden chaotic attractors, and excessive stability, have been discovered [ 15 , 16 , 17 ]. Memristor chaotic circuits are not only easy to integrate, but also play an extremely important role in applications within chaotic neural networks. There is evidence to suggest that neural networks working under chaotic edge mechanisms can optimize neural computation and global search [ 18 ]. Recently, Nature published a third-order nano integrated circuit component based on a local active memristor [ 19 ]. This component can simulate 15 functions of a single neuron, and can exhibit complex dynamic characteristics of neurons, such as chaotic oscillation, unimodal periodic oscillation, and action potential under different DC biases. After the discovery of memristors, Chua et al. expanded their research to include the concepts of memcapacitors and inductors [ 20 ] and conducted some fundamental theoretical studies. Although some phenomena with characteristics of memory containers and sensors have been preliminarily discovered, such as bistable elastic thin films [ 21 ] exhibiting chaotic characteristics under certain excitations, practical and useful artificial physical devices have not yet been realized. Although Boolean logic has always been the backbone of digital information processing, there are some so-called NP-hard problems. The logical method is fundamentally inefficient [ 22 , 23 ], which inspires the use of alternative, more effective non-Boolean methods to obtain a solution to this problem [ 24 , 25 , 26 , 27 , 28 ]. Based on phase dynamics, it is a good idea to build an Ising computer to solve NP-hard problems. At present, there are many ways to implement the oscillator element [ 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 ]. However, the current oscillator implementations also have some shortcomings and problems. For example, the operational costs of a quantum annealer based on qubits are high and become complex due to the requirement of a low-temperature environment. In addition, although the optical coherent Ising computer has a competitive advantage over the quantum annealing machine, it needs a long fiber ring cavity for implementing Ising spin through time multiplexing. Additionally, it relies on an ultra-high-speed and huge-power-consumption field-programmable gate array (FPGA) to facilitate coupling in the measurement feedback scheme. The functioning of a digital CMOS annealing machine depends on random numbers, which are generated to introduce randomness, but maintaining real randomness is still technically challenging and requires a lot of post-processing. In our research, we will study and implement a local active memristor oscillator-based computer. The computer is using the negative resistance of the local active memristor and a capacitor, together with an external resistor, to constitute the oscillator, so the oscillator network forms a continuous-time dynamic system. The bistable of the oscillator phase simulates the Ising spin. The optimization problem is mapped by carefully selecting the coupling matrix for the memristor oscillatory network. The dynamic evolution of this physical system has reached the energy minimization point, which represents the solution of the optimization problem. The dynamics of this physical system have been explored in a wide range of applications, including understanding neural activities, realizing robot motion control, and using discrete-time Hopfield networks to solve optimization problems. Our work will be published in a series of papers. In order to study the coupling in a local active memristor-based computer, the key is to study the interaction between oscillators. It is not practical to use the circuit simulation software SPICE’s (LTspice 24.0.9) transient simulation method because, with the expansion of the network scale, simulation time also increases significantly. This is because the single oscillator satisfies M individual differential equations; thus, the oscillator coupling network will have M × N differential equations to solve. In the past 40 years, the research has shifted to using the oscillator phase model to study the dynamic characteristics of the network, which will reduce the number of differential equations to N , saving a lot of simulation time. Among them, the most representative is the Kuramoto network model [ 41 , 42 , 43 , 44 , 45 , 46 ]. The advantage of this model is that the equation is simple, and it is easy to predict the phase evolution of the system. However, this model assumes that the oscillators are sinusoidally coupled, which limits its scope of application and accuracy. It is obvious that the network with local active memristor oscillators lacks sinusoidal coupling, so it is necessary to find another way to model them. In this paper, we use the research methods of neuro-dynamics for reference and introduce the concept of weak coupling [ 47 , 48 ]. We then solve the adjoint of the differential equations to further calculate the interaction function between the oscillators, expanding it into a Fourier series, so as to obtain the phase model of the system. Although our research object is the oscillator pair, the same research method can be easily extended to the oscillator coupling network, so as to provide a reference for the research on the memristor oscillatory network computer. The structure of this paper is organized as follows: In Section 2 , we briefly introduce the Kuramoto model because of its strong relevance to our research. We begin by detailing Winfree’s method, then Kuramoto’s, and then we proceed by comparing the two. In Section 3 , we describe the analysis of the memristor phase dynamics comprehensively. Firstly, we give Chua’s unfolding model of the memristor, which is different from the traditional physical model. However, it has characteristics such as a simple structure, clear physical image, and ease of analysis. Secondly, as an additional segment, we will introduce the weak coupling theory because it is fundamentally important to our analysis. Thirdly, based on the theory and model, we solve the adjoint of the memristor differential equations so as to determine the interaction function between the oscillators. Subsequently, we analyze the case of one memristor oscillator. Fourthly, the interaction function of oscillator pairs is expanded into Fourier series, and then the phase model is obtained. Section 4 is the summary of our research."
} | 2,958 |
22518097 | PMC3324817 | pmc | 5,061 | {
"abstract": "Most scene segmentation and categorization architectures for the extraction of features in images and patches make exhaustive use of 2D convolution operations for template matching, template search, and denoising. Convolutional Neural Networks (ConvNets) are one example of such architectures that can implement general-purpose bio-inspired vision systems. In standard digital computers 2D convolutions are usually expensive in terms of resource consumption and impose severe limitations for efficient real-time applications. Nevertheless, neuro-cortex inspired solutions, like dedicated Frame-Based or Frame-Free Spiking ConvNet Convolution Processors, are advancing real-time visual processing. These two approaches share the neural inspiration, but each of them solves the problem in different ways. Frame-Based ConvNets process frame by frame video information in a very robust and fast way that requires to use and share the available hardware resources (such as: multipliers, adders). Hardware resources are fixed- and time-multiplexed by fetching data in and out. Thus memory bandwidth and size is important for good performance. On the other hand, spike-based convolution processors are a frame-free alternative that is able to perform convolution of a spike-based source of visual information with very low latency, which makes ideal for very high-speed applications. However, hardware resources need to be available all the time and cannot be time-multiplexed. Thus, hardware should be modular, reconfigurable, and expansible. Hardware implementations in both VLSI custom integrated circuits (digital and analog) and FPGA have been already used to demonstrate the performance of these systems. In this paper we present a comparison study of these two neuro-inspired solutions. A brief description of both systems is presented and also discussions about their differences, pros and cons.",
"conclusion": "4 Conclusion We have presented a comparison analysis between Frame-Constrained and Frame-Free Implementations of ConvNet Systems for application in object recognition for vision. We have presented example implementations of Frame-Constrained FPGA realization of a full ConvNet system, and partial convolution processing stages (or combination of stages) using spiking AER convolution hardware using either VLSI convolution chips or FPGA realizations. The differences between the two approaches in terms of signal representations, computation speed, scalability, and hardware multiplexing have been established.",
"introduction": "1 Introduction Conventional vision systems process sequences of frames captured by video sources, like webcams, camcorders (CCD sensors), etc. For performing complex object recognition algorithms, sequences of computational operations are performed for each frame. The computational power and speed required makes it difficult to develop a real-time autonomous system. But brains perform powerful and fast vision processing using small and slow cells (neurons) working in parallel in a totally different way. Vision sensing and object recognition in the mammalian brain is not performed frame by frame. Sensing and processing are performed in a continuous way, spike by spike, without any notion of frames. The visual cortex is composed by a set of layers (Shepherd, 1990 ; Serre, 2006 ), starting from the retina. The processing starts beginning at the time the information is captured by the retina. Although cortex has feedback connections, it is known that a very fast and purely feed-forward recognition path exists in the visual cortex (Thorpe et al., 1996 ; Serre, 2006 ). In recent years significant progress has been made toward the understanding of the computational principles exploited by the visual cortex. Many artificial systems that implement bio-inspired software models use biological-like (convolution based) processing that outperform more conventionally engineered machines (Neubauer, 1998 ). These systems run at low speeds when implemented as software programs on conventional computers. For real-time solutions direct hardware implementations of these models are required. However, hardware engineers face a large hurdle when trying to mimic the bio-inspired layered structure and the massive connectivity within and between layers. A growing number of research groups world-wide are mapping some of these computational principles onto both real-time spiking hardware through the development and exploitation of the so-called AER (Address-Event-Representation) technology, and real-time streaming Frame-Based ConvNets on FPGAs. ConvNets have been successfully used in many recognition and classification tasks including document recognition (LeCun et al., 1998a ), object recognition (Huang and LeCun, 2006 ; Ranzato et al., 2007 ; Jarrett et al., 2009 ), face detection (Osadchy et al., 2005 ), and robot navigation (Hadsell et al., 2007 , 2009 ). A ConvNet consists of multiple layers of filter banks followed by non-linearities and spatial pooling. Each layer takes as input the output of previous layer and by combining multiple features and pooling over space, extracts composite features over a larger input area. Once the parameters of a ConvNet are trained, the recognition operation is performed by a simple feed-forward pass. The simplicity of the feed-forward pass has pushed several groups to implement it as custom hardware architectures. Most of ConvNet hardware implementations reported over the years are for the frame-constrained fix-pixel-value version, as they map directly from the software versions. The first one was the ANNA chip, a mixed high-end, analog-digital processor that could compute 64 simultaneous 8 × 8 convolutions at a peak rate of 4.109 MACs (multiply-accumulate operations per second; Boser et al., 1991 ; Säckinger et al., 1992 ). Subsequently, Cloutier et al. proposed an FPGA implementation of ConvNets (Cloutier et al., 1996 ), but fitting it into the limited-capacity FPGAs available at those times required the use of extremely low-accuracy arithmetic. Modern DSP-oriented FPGAs include large numbers of hard-wired multiply-accumulate units that can greatly speed up compute-intensive operations, such as convolutions. The frame-constrained system presented in this paper takes full advantage of the highly parallel nature of ConvNet operations, and the high-degree of parallelism provided by modern DSP-oriented FPGAs. Achieved peak rates are in the order of 10 11 MACs. On the other hand, Frame-free Spiking-Dynamic-Pixel ConvNets compute in the spike domain. No frames are used for sensing and processing the visual information. In this case, special sensors are required with a spike-based output. Spike-based sensors and processors typically use AER (Address-Event-Representation) in order to transmit the internal state and/or results of the neurons inside a chip or FPGA. AER was originally proposed almost twenty years back in Mead’s Caltech research lab (Sivilotti, 1991 ). Since then AER has been used fundamentally in vision (retina) sensors, such as simple light intensity to frequency transformations (Culurciello et al., 2003 ; Posch et al., 2010 ), time-to-first-spike coding (Ruedi et al., 2003 ; Chen and Bermak, 2007 ), foveated sensors (Azadmehr et al., 2005 ), spatial contrast (Costas-Santos et al., 2007 ; Massari et al., 2008 ; Ruedi et al., 2009 ; Leñero-Bardallo et al., 2010 ), temporal contrast (Lichtsteiner et al., 1998 ; Posch et al., 2010 ; Leñero-Bardallo et al., 2011 ), motion sensing and computation, (Boahen, 1999 ), and combined spatial and temporal contrast sensing (Zaghloul and Boahen, 2004 ). But AER has also been used for auditory systems (Chan et al., 2007 ), competition and winner-takes-all networks (Chicca et al., 2007 ; Oster et al., 2008 ), and even for systems distributed over wireless networks (Teixeira et al., 2006 ). After sensing, we need Spiking Signal Event Representation techniques capable of efficiently processing the signal flow coming out from the sensors. For simple per-event heuristic processing and filtering, direct software based solutions can be used (Delbrück, 2005 , 2008 ). Other schemes rely on look-up table re-routing and event repetitions followed by single-event integration (Vogelstein et al., 2007 ). Alternatively, we can find some pioneering work in the literature aiming at performing convolutional filteringon the AER flow produced by spiking retinas, (Vernier et al., 1997 ; Choi et al., 2005 ), where the shape of the filter kernel was hard-wired (either elliptic or Gabor). Since 2006, working AER Convolution chips have been reported with arbitrary shape programmable kernel of size up to 32 × 32 pixels pre-loaded onto an internal kernel-RAM (Serrano-Gotarredona et al., 2006 , 2008 ; Camuñas-Mesa et al., 2011 , 2012 ). This opens the possibility of implementing in AER spiking hardware generic ConvNets, where large number of convolutional modules with arbitrary size and shape kernels are required. In this paper we present, discuss and compare two different neuro-cortex inspired approaches for real-time visual processing based on convolutions: Frame-based fix-pixel-value and Frame-free dynamic-pixel-spiking ConvNet Processing hardware. Section 2 describes generic ConvNets and their structure. Section 3 briefly describes frame-free ConvNet types of implementations, and Section 4 describes a frame-constrained FPGA implementation. Implemention details will be given in a very concise manner, so the reader can grasp the main ideas behind each implementation. For more detailed descriptions the reader is refer to the corresponding references. Finally, Section 5 provides a comparison of both cases indicating pros and cons of each."
} | 2,438 |
30401427 | null | s2 | 5,062 | {
"abstract": "No abstract available"
} | 5 |
31277471 | PMC6681080 | pmc | 5,063 | {
"abstract": "Flooding limits biomass production in agriculture. Leguminous plants, important agricultural crops, use atmospheric dinitrogen gas as nitrogen nutrition by symbiotic nitrogen fixation with rhizobia, but this root-nodule symbiosis is sometimes broken down by flooding of the root system. In this study, we analyzed the effect of flooding on the symbiotic system of transgenic Lotus japonicus lines which overexpressed class 1 phytoglobin ( Glb1 ) of L. japonicus ( LjGlb1-1 ) or ectopically expressed that of Alnus firma ( AfGlb1 ). In the roots of wild-type plants, flooding increased nitric oxide (NO) level and expression of senescence-related genes and decreased nitrogenase activity; in the roots of transgenic lines, these effects were absent or less pronounced. The decrease of chlorophyll content in leaves and the increase of reactive oxygen species (ROS) in roots and leaves caused by flooding were also suppressed in these lines. These results suggest that increased levels of Glb1 help maintain nodule symbiosis under flooding by scavenging NO and controlling ROS.",
"conclusion": "5. Conclusions Glb1 contributes to maintaining nodule symbiosis under flooding conditions and controls ROS by scavenging NO.",
"introduction": "1. Introduction Flooding often reduces crop growth and yield, causing serious problems for farmers. Plant growth is hampered by flooding because it exposes plants to hypoxia, which inhibits aerobic respiration and photosynthesis, reducing ATP production. Hypoxia inhibits photosynthesis by inducing production of reactive oxygen species (ROS), which damage the chloroplast membrane and decrease the photosynthetic potential [ 1 , 2 ]. Excess ROS also lead to lipid peroxidation and alterations in lipid composition, electrolyte leakage, and ultimately cell death [ 3 , 4 , 5 ]. Another reactive molecule, nitric oxide (NO), is produced in plants in response to biotic and abiotic stresses, including hypoxia [ 6 , 7 , 8 , 9 ]. NO serves as a signal molecule in various physiological and pathogenic responses of plants such as stomatal opening and closing [ 10 ], protein S -nitrosylation, and cGMP nitration [ 11 ]. Excess NO is toxic and inhibits plant growth; plants regulate NO levels by producing plant hemoglobin (phytoglobin, Glb) [ 12 , 13 ]. Glbs are divided into three classes: Glb1, Glb2, and Glb3 [ 14 , 15 , 16 , 17 ]. Leghemoglobin (Lb) of leguminous plants, which was the first identified Glb [ 18 ], belongs to Glb2 and is essential for legume–rhizobia symbiosis because it regulates oxygen partial pressure in root nodules [ 19 ]. Although the function of Glb3 is unknown, it may interact with NO [ 20 ]. Glb1 has extremely high affinity for oxygen [ 21 ] and scavenges NO by oxidizing it to nitrate [ 12 , 13 , 17 , 22 ]. Under hypoxia, overexpression of Glb1 ameliorates the energy status and growth of both maize cells and alfalfa roots [ 6 , 8 , 12 ], and enhances the survival of Arabidopsis thaliana [ 16 ]; in all cases, consistent low NO level strongly suggests the role of NO-scavenging activity of Glb1 in tolerance to hypoxia. Overexpression or ectopic expression of tobacco gene NtHb1 enhances Cd tolerance by reducing Cd and NO levels in Nicotiana tabacum and A. thaliana [ 23 , 24 ]. At least eight Glb genes have been identified in the genome of Lotus japonicus : two Glb1 s ( LjGlb1-1 , LjGlb1-2 ), four Glb2 s ( LjGlb2 and three Lb genes) and two Glb3 s ( LjGlb3-1 , LjGlb3-2 ) [ 25 , 26 ]. LjGlb1-1 is the only NO-inducible Glb gene of L. japonicus [ 26 ]. The NO-scavenging activity of Glb1 is required for establishing proper root nodule symbiosis [ 27 ]. In the L. japonicus–Mesorhizobium loti symbiosis, inoculation with M. loti induces NO production in roots with the simultaneous expression of the Glb1 gene ( LjGlb1-1 ) [ 28 ]. NO inhibits nitrogenase [ 29 , 30 ] and promotes nodule senescence [ 31 ]. A null mutant line of LjGlb1-1 shows low infection and low nitrogenase activity of the nodules [ 27 ], whereas overexpression of LjGlb1-1 increases nitrogenase activity [ 32 , 33 ]. The beneficial effects of LjGlb1-1 overexpression may be attributed to a reduced level of NO [ 33 ]. No drastic differences have been observed in the shape and growth of plants among these overexpression lines, the null mutant, and the wild type with supply of nitrogen source [ 27 , 33 ], although the timing of bolting and flowering has not been statistically compared. In the Alnus firma (actinorhizal plant)– Frankia (actinobacterium) symbiosis, Glb1 of A. firma (AfGlb1, accession number AB221344 in DDBJ database) may support the nitrogenase activity of the nodules as a NO scavenger [ 34 ]. Flooding adversely affects nodule symbiosis; NO produced in the nodules in response to flooding decreases nitrogenase expression and activity [ 35 , 36 , 37 ]. NO might attack nodule cells during flooding and delay the recovery of the symbiotic activity of the nodules after flooding. Because NO-scavenging activity contributes to hypoxia tolerance, we hypothesized that Glb1 overexpression might improve the tolerance of nodule symbiosis to hypoxia. In this study, we examined the tolerance of the nodule symbiosis to flooding in two transgenic L. japonicus lines that express LjGlb1-1 or AfGlb1 driven by the CaMV 35S promoter. Our results suggest that Glb1 overexpression improves nodule symbiosis by controlling not only NO but also ROS.",
"discussion": "4. Discussion Flooding inhibits plant growth by causing hypoxic stress in roots [ 52 ], which induces NO production and Glb1 expression [ 6 , 12 , 16 ]. Overexpression of Glb1 enhances NO scavenging activity, increases tolerance to hypoxic stress [ 6 , 16 , 53 ], increases nitrogenase activity, and slows down the aging of nodule symbiosis [ 33 ]. In this study, the lines of L. japonicus stably transformed with LjGlb1-1 (Ox1) or AfGlb1 (Afx1) allowed us to study the effect of Glb1 overexpression on the hypoxic tolerance of nodule symbiosis. Flooding significantly decreased nitrogenase activity and increased the NO level of WT and 96642 nodules but did not alter the high nitrogenase activity and low NO level in the Ox1 and Afx1 nodules ( Figure 1 and Figure 2 ). These results suggest that increased Glb1 expression enhanced NO-scavenging activity and improved flooding tolerance of the nodules. Flooding increased the expression of three senescence-related genes more in WT and 96642 than in Ox1 and Afx1 ( Figure 3 ); the number and size of the vacuoles of the infected nodule cells were increased in WT and 96642, which is typical of aged nodules [ 51 ]. The accumulation of starch granules in the nodule also indicates senescence. The decreased nitrogenase activity of the aged nodules does not consume the product of photosynthesis transported from the leaves, and the residual carbon source accumulates as starch granules. However, starch granules were not observed in any lines despite the decrease in nitrogenase activity in WT and 96642 ( Figure 4 ). Flooding significantly reduces the photosynthetic capacity and transpiration rate [ 2 , 54 ]; in our experiments, flooding might have reduced the photosynthetic capacity or the transport of photosynthetic products to roots. Soil flooding can perturb the photosynthetic machinery, reducing photosynthetic potential, possibly because of ROS production [ 1 ]. ROS damage the structure of chloroplast membranes and inhibit the photosystem function [ 2 ]. In the present study, flooding decreased the total amount of chlorophylls in WT and 96642, but not in Ox1 or Afx1 ( Figure 5 A). Leaf DAB staining was strong in 96642 and flooding increased it in WT, with no obvious difference in Ox1 and Afx1. These results suggest that flooding increased H 2 O 2 levels in leaves, and that Glb1s were involved. Flooding reduced the leaf O 2 − levels in all lines, with no obvious difference among them ( Supplementary Figure S1 ). ROS lead to electrolyte leakage [ 3 , 4 , 5 ], but we detected no significant difference among the lines in electrolyte leakage rate ( Supplementary Figure S2 ), despite the higher level of H 2 O 2 in the leaves of WT and 96642 ( Figure 5 B). The metabolism might have adapted to flooding to reduce the damage by ROS. During flooding, roots are exposed to hypoxic stress more than the other plant parts. In our study, ROS levels increased in the roots and leaves of WT and 96642 ( Figure 5 B and Figure 6 ), but ROS and NO levels remained low in the lines with increased Glb1 levels ( Figure 1 B,C and Figure 6 ). Although the results of NBT and DAB were similar to those of CellROX fluorescence in roots, the difference among the lines in DAB staining was less clear than that in CellROX staining ( Figure 6 ), possibly because of different sensitivity and specificity of these staining methods: CellROX detects various ROS including H 2 O 2 and O 2 − , whereas DAB detects only H 2 O 2 , and NBT detects only O 2 − . Under control conditions, there was no difference in the amount of ROS between the roots of 96642 and WT, but it was slightly higher in the leaves of 96642. We do not have a good explanation for these inconsistent results. The null mutant 96642 showed higher NO levels in the roots ( Figure 1 B,C) and higher expression levels of some senescence-related genes in the nodules than WT ( Figure 3 ). These physiological differences might affect the amount of ROS in leaves. In cultured alfalfa roots, Glb1 overexpression improves the antioxidant status by increasing ascorbate levels and the activity of enzymes involved in H 2 O 2 metabolism [ 55 ]. In corn plants, Glb1 overexpression alleviates flooding stress by limiting ROS-induced damage and ensuring a sustained photosynthetic rate through improvement of the ascorbate antioxidant status and an increase in activities of several ROS-scavenging enzymes [ 56 ]. In L. japonicus , Glb1 overexpression might alleviate flooding stress by improving ROS metabolism, although we did not investigate the expression of genes related to the regulation of ROS."
} | 2,509 |
34390211 | PMC8529448 | pmc | 5,064 | {
"abstract": "Abstract The triboelectric effect is a ubiquitous phenomenon in which the surfaces of two materials are easily charged during the contact‐separation process. Despite the widespread consequences and applications, the charging mechanisms are not sufficiently understood. Here, the authors report that, in the presence of a strain gradient, the charge transfer is a result of competition between flexoelectricity and triboelectricity, which could enhance charge transfer during triboelectric measurements when the charge transfers of both effects are in the same direction. When they are in the opposite directions, the direction and amount of charge transfer could be modulated by the competition between flexoelectric and triboelectric effects, which leads to a distinctive phenomenon, that is, the charge transfer is reversed with varying forces. The subsequent results on the electrical power output signals from the triboelectrification support the proposed mechanism. Therefore, the present study emphasizes the key role of the flexoelectric effect through experimental approaches, and suggests that both the amount and direction of charge transfer can be modulated by manipulating the mixed triboelectric and flexoelectric effects. This finding may provide important information on the triboelectric effect and can be further extended to serve as a guideline for material selection during a nanopatterned device design.",
"conclusion": "3 Conclusion In summary, we demonstrated the coupling between triboelectric and flexoelectric effects at the nanoscale, which could enhance charge transfer during triboelectric measurements when the charge transfers of both effects are in the same directions, or modulate the direction and amount of resulted charge transfer when they are in the opposite directions. In the AFM measurements, flexoelectricity is inevitably induced in the TiO 2 film because of the strain gradient induced by the inhomogeneous force applied through the AFM tip. We showed that triboelectric and flexoelectric effects are fundamentally inseparable in nanoscale systems involving inhomogeneous stress/strain. The relative relationship between them could be modulated by normal force, which makes it possible to manipulate the amount as well as direction of the charge transfer in the case of opposite triboelectric and flexoelectric charging. In contrast, triboelectric charging can be improved through flexoelectricity when the charge transfer directions are the same. The mechanism can be applied to general scenarios involving an inhomogeneous strain at the nanoscale. The different performances of flat and pyramid‐featured triboelectric devices fabricated with different materials further support the proposed mechanism. This study can contribute to a fundamental understanding of the triboelectric effect where a dielectric material subjected to an inhomogeneous force is involved and could be further extended to increase the efficiency of the device performance in energy harvesting.",
"introduction": "1 Introduction Triboelectricity is a process by which material surfaces become electrically charged as a result of touching or rubbing another surface. [ \n \n 1 \n , \n 2 \n , \n 3 \n \n ] This effect is well known and has widespread and significant influences. Based on a charge transfer, triboelectricity has garnered wide attention because of its great potential in applications in various areas such as electrostatic self‐assembly, [ \n \n 4 \n \n ] ionic electret, [ \n \n 5 \n \n ] and triboelectric nanogenerators, [ \n \n 6 \n , \n 7 \n \n ] which have been widely studied for applications in various fields. [ \n \n 8 \n , \n 9 \n , \n 10 \n \n ] Despite its elementary nature and high application value, the fundamental mechanism of triboelectricity has not been fully understood, [ \n \n 1 \n \n ] particularly the involvement of dielectrics and the reduction of size to the micro/nanoscale, which are common in state‐of‐the‐art nanotechnologies, make the question even more complicated. In contrast, flexoelectricity, the coupling between an electric polarization and strain gradient, [ \n \n 11 \n , \n 12 \n , \n 13 \n , \n 14 \n , \n 15 \n \n ] has been demonstrated in recent studies on various material systems to modulate physical properties, for example, a photovoltaic effect, Schottky barrier, resistance, or polarization, has been reported in BiFeO 3 , [ \n \n 16 \n \n ] halide perovskite, [ \n \n 17 \n \n ] Si, [ \n \n 18 \n \n ] TiO 2 , [ \n \n 18 \n , \n 19 \n \n ] and (Nb‐)SrTiO 3 , [ \n \n 14 \n , \n 15 \n , \n 18 \n , \n 19 \n \n ] to name a few. Basically, flexoelectricity is a universal property of all dielectric materials including centrosymmetric materials, which occurs under an inhomogeneous strain field. [ \n \n 14 \n , \n 17 \n \n ] Although it is negligible at the bulk scale level owing to a small strain gradient, [ \n \n 20 \n \n ] it cannot be ignored at the nanoscale level because the strain gradient at this level is several times larger than at the bulk scale level. [ \n \n 12 \n \n ] Therefore, when a dielectric material is subjected to an inhomogeneous force at the nanoscale, the strain gradient breaks the original symmetry and forms a polarization with a preferred direction. [ \n \n 19 \n \n ] It has been reported the polarization in ferroelectrics can modulate the triboelectric charge transfer, [ \n \n 21 \n \n ] likewise, the flexoelectric effect in centrosymmetric dielectrics may also contribute to the charge transfer in the presence of a strain gradient. Recently, theoretical analyses suggested that bipolar current in the triboelectric devices were originated from the surface potential difference induced by flexoelectricity, that is, the triboelectricity is a result of flexoelectricity. [ \n \n 22 \n \n ] Based on conventional Herzian and Johnson–Kendall–Roberts contact models, the authors studied surface potential difference by flexoelectricity in indentation and pull‐off cases and suggested flexoelectricity was a thermodynamic driver in triboelectric phenomena. Indeed, following studies by similar method and density functional theory suggested it could be used to theoretically explain the charge transfer between identical materials, where the work function difference is absence. [ \n \n 23 \n , \n 24 \n \n ] Nevertheless, it cannot provide a satisfactory interpretation on many reports on charge transfer between different materials, which are readily explained by work function difference. [ \n \n 3 \n \n ] On the other hand, direct experimental evidence on the role of flexoelectricity on the charge transfer between different materials, that is, in the presence of work function difference, is scarce to date. Therefore, more efforts are needed to show the direct evidence and have a clear understanding of the contribution of flexoelectricity to triboelectric charge transfer, especially between different materials. The condition described above is easily achievable in an atomic force microscopy (AFM)‐based measurement, [ \n \n 14 \n , \n 25 \n , \n 26 \n , \n 27 \n \n ] which is also a general approach to settle the fundamental questions of the triboelectric effect at the nanoscale. [ \n \n 2 \n , \n 21 \n , \n 28 \n , \n 29 \n \n ] The triboelectric charge transfer is usually realized by rubbing the sample surface with a sharp AFM tip under various conditions. [ \n \n 30 \n , \n 31 \n \n ] During this process, an inhomogeneous force is exerted on the surface, inducing a strain gradient in the sample beneath the tip; thus, the flexoelectric effect can play a role in the charge transfer when rubbing. Because of the high sensitivity to the strain gradient, and thus force, the flexoelectric effect might be distinguishable by varying the force exerted on the surface while keeping the other conditions unchanged. Nevertheless, to the best of our knowledge, detailed force‐dependent triboelectric charge transfer at nanoscale has yet to be experimentally reported. Furthermore, making this question clear is not only of significance to the fundamental scientific understanding of triboelectric charging, but also important for the device design and applications. For instance, it was observed that the electrical power output from the triboelectrification can be significantly modified by fabricating nano‐ and micro‐scale structures on one of the surfaces. [ \n \n 32 \n , \n 33 \n , \n 34 \n , \n 35 \n , \n 36 \n \n ] In these cases, although the variation in the contact area has been employed to explain the observations, the flexoelectric effect might also play a role in modulating the power output because the flexoelectric effect becomes non‐negligible at the nanoscale owing to the gigantic strain gradient. [ \n \n 25 \n \n ] Consequently, experimentally confirming how the flexoelectric effect influences the charge transfer during triboelectric measurement is an important aspect of the triboelectric mechanism. Herein, we propose that the charge transfer between a sharp metal and flat oxide thin film can be attributed to mixed triboelectric and flexoelectric effects instead of pure triboelectric or flexoelectric effects, and the enhancement of charge transfer could be achieved when the charge transfers of both effects are in the same direction. To demonstrate this, force‐dependent triboelectricity was explored using AFM. Contact mode AFM was used to rub the sample surface using an AFM tip with a controllable normal force. Subsequently, the triboelectric charge distribution and subsequent diffusion on the surface were characterized using Kelvin probe force microscopy (KPFM). [ \n \n 37 \n \n ] Because the surface potential is determined by the work function difference between the AFM tip and sample surface, [ \n \n 38 \n \n ] the variation of the work function or charge transfer can be analyzed through surface potential images. It turns out that, in the case of opposite triboelectric and flexoelectric charges, the dominant contributor alters from the triboelectricity in the low‐force regime to flexoelectricity in the high‐force regime. Furthermore, triboelectric charging can be improved by flexoelectricity when they are in the same direction. The subsequent results obtained by applying a small positive/negative tip bias further validate this concept. In addition, the electrical power outputs from triboelectrification between two flat surfaces and flat‐nano pattern pairs further confirm the validity of the proposed mechanism. The present study suggests that triboelectricity is inevitably coupled with flexoelectricity in nanoscale systems where a strain gradient is generated. We demonstrated an experimental pathway to visualize the contribution of flexoelectricity to triboelectricity, which sheds light on the understanding of triboelectric charges at the nanoscale and provides a guideline for the design of high‐performance nanoscale triboelectric devices by selecting suitable materials.",
"discussion": "2 Results and Discussion To explore how the flexoelectricity contributes to the triboelectric measurements, we chose a TiO 2 thin film, of which the physical parameters relevant to flexoelectricity have been reported in various studies, [ \n \n 18 \n , \n 19 \n \n ] and a conductive diamond‐coated tip (CDT‐FMR), the high hardness of which can minimize the tip wear during measurements, as a model system. It has been suggested that a downward flexoelectric polarization associated with the strain gradient could be generated beneath the tip when an AFM tip is pressed against the film surface, which resembles the case of a positive voltage applied to the tip in the case of TiO 2 . [ \n \n 26 \n , \n 39 \n , \n 40 \n , \n 41 \n , \n 42 \n , \n 43 \n \n ] Therefore, the flexoelectricity is supposed to favor electron transfer from the sample to the tip. In this case, an opposite triboelectric charge transfer, that is, from the tip to the sample, would be ideal to study the coupling between the two, which requires a higher work function for the sample than the tip. The calibrated work functions of the tip and sample are φ \n tip,eff = 4.52 eV and φ \n sample,eff = 4.8 eV, respectively (see the details in Figure S1 , Supporting Information). These characteristics make them an optical platform for our purposes. The occurrence of triboelectricity only requires physical contact, whereas flexoelectricity is linearly dependent on the strain gradient, which is sensitive to the normal force. Therefore, we examined the charge transfer dependence on the normal force applied by the AFM tip during rubbing. The amount and direction of transferred charges between the tip and sample can be evaluated by the change in the surface potential between the rubbed and pristine areas. Figure \n \n 1 a displays the surface potential distribution in the areas rubbed with normal forces ranging from 10 to 800 n n . After being rubbed by a low normal force (<50 n n ), the area is relatively negatively charged because it has a lower surface potential than the surrounding pristine area. In contrast, a high normal force (>200 n n ) induces a higher surface potential than the surrounding pristine area, indicating that it is relatively positively charged. This normal force‐dependent surface potential variation is more clearly visible in both surface potential line profiles (Figure 1b ) and surface potential difference (Δ V \n SP ) between the rubbed ( V \n SP,r ) and surrounding pristine ( V \n SP,p ) areas (Δ V \n SP = V \n SP,r − V \n SP,p ) as a function of the normal force (Figure 1c ). The sign of Δ V \n SP changes nonlinearly from negative to positive across zero as the normal force increases. Figure 1 a) Surface potential images after rubbing with different normal forces on TiO 2 thin film. Scale bar is 2 µm. b) Corresponding surface potential profiles, c) Δ V \n SP between rubbed and pristine regions, d) histograms of friction force at different normal forces, and e) friction force as function of normal force corresponding to (d). Solid line in (e) is fitted using linear function. Numbers in (a) and black arrows in (b,d,e) indicate measurement sequence. For all data, scan rate is 0.5 Hz with CDT‐FMR. During surface rubbing, the friction signal, or lateral signal (in volts) distribution at each normal force, was simultaneously recorded. Through a friction force calibration using the modified Wedge–Flat method, [ \n \n 44 \n , \n 45 \n \n ] the friction force (in nanonewtons) was obtained from the friction signal (see the details in Section S1 and Figure S2 , Supporting Information). The histograms of the friction force depending on the normal force and the relationship between the normal and friction forces are shown in Figures 1d and 1e , respectively. It can be seen that the friction force increases linearly with the normal force, as expected. Therefore, we can infer that, because the friction force increases linearly with an increasing normal force, a higher normal force can induce a greater charge transfer associated with the triboelectric effect. First, we attempted to use triboelectricity to explain the intriguing charge transfer results without considering the flexoelectricity. The schematics in Figure \n \n 2 a–c shows the charge flow driven by the work function: when the tip and sample are physically in a contact state, electrons flow from one to another driven by the difference in work function (see Figure 2c ). [ \n \n 46 \n , \n 47 \n \n ] According to the calibrated work functions, the Fermi level ( E \n f ) of the tip is higher than the highest occupied surface state level of the sample. Therefore, the electrons in the tip will migrate to the sample surface to fill up the surface energy states at the same height as E \n f of the tip when they are in contact, [ \n \n 48 \n \n ] as shown in Figure 2c . Consequently, a lower surface potential is observed in the rubbed area compared with the pristine state as an indicator of a lower work function, as depicted in Figure 2b , confirming the dominant role of the triboelectricity at a low normal force. When the normal force increases, the triboelectric charge transfer is considered to increase because of the increased friction force. [ \n \n 49 \n , \n 50 \n \n ] Meanwhile, the contact area will increase until reaching an extreme state, [ \n \n 51 \n \n ] which has been reported to promote the triboelectricity. [ \n \n 52 \n \n ] If triboelectricity is the only contribution, as generally considered, the triboelectric charge should have a positive correlation with the normal force, as well as the friction force. In other words, more triboelectric charges would transfer from the tip to the sample at a higher normal/friction force, and the charge flow direction would not be altered by the force. Thus, we can rule out the influence of increasing contact area on the observed sign reversal result. Consequently, the triboelectric effect alone provides an unsatisfactory interpretation of the sign reversal phenomenon shown in Figure 1 . Figure 2 a,d) Schematics of tip‐sample contact under a) low and d) high normal force and b,e) corresponding (top) representative work function images and (bottom) illustrative surface potential profile across rubbed region under b) low and e) high normal forces. Scale bar is 2 µm. c) Triboelectric charge transfer driven by work function difference where E \n f and E \n 0 are Fermi level of tip and highest occupied surface state level of sample, respectively. f) Simulated flexoelectric surface potential on TiO 2 thin film at normal force of 500 n n , that is, high normal force, from tip. Alternatively, when considering the coupling of flexoelectricity with triboelectricity, we can find a plausible explanation for these fascinating observations. In the case of an extremely low normal force, the flexoelectric effect is assumed to be negligible because of the small strain gradient. With increasing normal force, the strain gradient in the TiO 2 thin film gradually increases to a degree that cannot be ignored, and the flexoelectricity becomes involved as a competitor against the triboelectricity. The schematics in Figure 2d,e illustrates the strain gradient and charge transfer driven by the flexoelectricity at a high normal force, which is in the opposite direction compared to that of triboelectricity, that is, from sample to tip, in the current system. Accordingly, when the flexoelectricity charge is dominant over the triboelectric charge, the rubbed area exhibits a higher surface potential. To confirm the flexoelectric contribution, we use contact mechanics analysis to simulate the strain gradient, flexoelectric field, which results from flexoelectric polarization, and the corresponding electric potential in the TiO 2 thin film under an AFM tip with relatively low and high normal forces. [ \n \n 53 \n \n ] The details can be found in Section 3 , Supporting Information. As expected, the strength and coverage of the strain gradient and flexoelectric field were pronounced at higher normal forces, as shown in Figure S3 , Supporting Information. Furthermore, we were able to obtain the electric potential on the surface associated with flexoelectricity, that is, the flexoelectric surface potential ( V \n flexo ), as demonstrated in Figure 2f and Figure S3 , Supporting Information. Indeed, the simulated flexoelectric surface potential is relatively higher than that of the surrounding area, as shown in Figure 2e . Based on the calculation, the dependence of V \n flexo on the normal force can be determined theoretically (Figure S4b , Supporting Information). It is worth noting that the V \n flexo used here is the maximum value of the flexoelectric surface potential under a certain condition, [ \n \n 22 \n \n ] which could differ from the actual value in the experiment. Furthermore, the Schottky barrier can be modulated through flexoelectric polarization, which could be another contributor to triboelectric charging. [ \n \n 18 \n \n ] \n The observed charge transfer in our case is in nature a result of mixed triboelectric and flexoelectric effects, particularly at high normal forces. The triboelectric and flexoelectric effects generate negative and positive charges, respectively, on the TiO 2 thin‐film surface. Therefore, the resultant charge transfer depends on the competition between the triboelectric and flexoelectric effects. At a low normal force, because the triboelectric effect is dominant owing to the relatively weak flexoelectric effect, the rubbed area is negatively charged and shows a lower surface potential. When the normal force increases, the flexoelectricity gradually surpasses the triboelectricity and thus results in a charge sign that changes from negative to positive, as shown in Figure 1 . According to our calculations and reported studies, the flexoelectric effect exhibits an exponential correlation with the force. [ \n \n 54 \n , \n 55 \n \n ] Note that, because the amount of change in Figure 1 is not significantly high, the saturated‐like behavior at a high normal force might not be relevant to a Coulombic repulsion. [ \n \n 56 \n , \n 57 \n , \n 58 \n \n ] Furthermore, the ionic motion can be excluded as a main contribution owing to the reversibility. It is known that the triboelectric effect could generate heat and raised the temperature at the contact interface, which would also affect the charge transfer behavior by accelerating charge dissipation in the contextual situation. [ \n \n 59 \n \n ] However, it cannot explain the change of charge transfer direction and increase positive surface potential at high normal force. Besides, the temperature change is not supposed to be high considering the small contact radius and pressure by comparing with references. [ \n \n 60 \n , \n 61 \n \n ] Overall, the influence of raised temperature by triboelectricity is not supposed to have a significant influence on the charge transfer observed in this work. To examine the universality of the mechanism, we conducted experiments in different systems with an alternative cantilever and sample, respectively. As shown in Figure \n \n 3 a,b and Figures S5 and S6 , Supporting Information, a similar tendency of decreasing charge transfer with increasing normal force was observed when we changed the cantilever (Multi75E‐G, Pt‐coated) or sample (SiO 2 thin film). Thus, we concluded that this mechanism may be universally valid in other systems, rather than merely valid for specific tips and samples. Meanwhile, we should note that because the triboelectricity is strongly dependent on the surface state of the sample or environmental conditions, that is, humidity, surface defects, and adsorbates, these additional factors could have a tremendous influence on the triboelectric charging behavior. [ \n \n 62 \n , \n 63 \n \n ] In addition, the variation of the local I – V curve with normal force, as shown in Figure S7 , Supporting Information, also suggests a contribution from the flexoelectricity, the details of which can be found in the Supporting Information. Figure 3 a,b) Surface potential profiles after rubbing with different normal forces in systems of a) Pt‐coated tip and TiO 2 thin film, and b) conductive diamond‐coated tip and SiO 2 thin film. Black arrows indicate the measurement sequence. Insets are representative surface potential images. Details are revealed in Figures S5 and S6 , Supporting Information. Scale bar in inset is 3 µm. c) Surface potential images after rubbing with tip bias of i) −300 and ii) +300 mV at different normal forces on TiO 2 thin film with conductive diamond‐coated tip and d) corresponding surface potential profiles. e) Normal force‐dependent Δ V \n SP obtained from Figure 3d and evaluated flexoelectric contribution. f) Δ V \n SP as a function of normal force without tip bias as in Figure 1c , Δ V \n SP,flexo as in Figure 3e , and evaluated triboelectric contribution (Δ V \n SP,tribo = Δ V \n SP – Δ V \n SP,flexo ). According to the concept proposed above, the charge transfer in question is the result of mixed triboelectric and flexoelectric effects instead of the triboelectric or flexoelectric effect alone. Thus, the flexoelectric effect is considered to enhance the triboelectric charge transfer when in the same direction, that is, a positive surface potential difference will be observed at a low normal force and increases with normal force. To avoid the possible uncertainties relevant to the physical properties, for example, flexoelectric coefficient, surface state, and defects, to name a few, examining the idea in the same system is ideal. The application of positive or negative bias to the tip can nullify the work function difference between the tip and sample, which resembles the basic operational principle of contact KPFM. [ \n \n 64 \n \n ] Thus, if one can nullify the work function difference by applying a bias to the tip while rubbing, the triboelectric charge transfer driven by the work function difference can be minimized or even cause a reverse charge transfer owing to the flexoelectric effect. In contrast, a triboelectric charge transfer can be enhanced if an opposite sign of the bias is applied to the tip. Figure 3c(i) shows the surface potential images after rubbing the TiO 2 film surface with different normal forces while applying a tip bias of −300 mV, which is opposite the nullifying bias. It shows a similar dependence on the normal force as in the case of zero tip bias (Figure 1 ), whereas the triboelectric charge transfer is enhanced and the critical normal force, at which no charge transfer is observed, increases. In contrast, the application of +300 mV, which is similar to (but slightly higher than) the work function difference between the tip and sample in Figure 1 , nearly nullifies the original work function difference and leads to an opposite charge transfer direction. In addition, the charge transfer was enhanced with the normal force, as expected. The surface potential line profiles in Figure 3d and the calculated Δ V \n SP in Figure 3e present a distinct contrast in the surface potential after rubbing with a tip bias at various normal forces. With a slight discrepancy, the tip bias of +300 mV nullifies the work function difference between the tip and sample; therefore, Δ V \n SP (+300 mV) is considered to be mainly contributed to by the flexoelectricity. We fitted the normal force‐dependent Δ V \n SP (+300 mV) by using the exponential relationship as in the theoretical analysis and introduced an offset considering the discrepancy. Thus, the flexoelectric charge transfer (Δ V \n SP,flexo ) can be evaluated by subtracting the offset from the experimental Δ V \n SP (+300 mV), as demonstrated in Figure 3e . By assuming that Δ V \n SP obtained in the experiment only contains triboelectric and flexoelectric contributions, it can be expressed as Δ V \n SP = Δ V \n SP,tribo + Δ V \n SP,flexo , where Δ V \n SP,tribo indicates the triboelectric contribution to the experimental results. Therefore, Δ V \n SP,tribo can be roughly estimated by subtracting Δ V \n SP,flexo (Figure 3e ) from Δ V \n SP (Figure 1c ), as shown in Figure 3f . The different tendency at low and high normal force regions could be a result of band bending induced through flexoelectric polarization and an increased contact area, [ \n \n 65 \n \n ] suggesting a more complicated coupling between the triboelectric and flexoelectric effects rather than simply mixing. In Figure 3f , there is a crossover between the triboelectric and flexoelectric effects, where a near‐zero charge transfer is observed in Figure 1c , as marked by a gray shadow. It is noteworthy that the crossover of the flexoelectricity and triboelectricity, and thus the resultant normal force‐dependent charge transfer tendency, depend on various factors, such as the environmental condition, the work function difference between the tip and sample, the radius and Poisson's ratio of the tip, and the flexoelectric coefficient of the sample. Therefore, a crossover can be not observable in some circumstances. In our work, a sharp metal and flat oxide thin films were utilized considering it is a common system in AFM‐based triboelectric studies. Nevertheless, the sharp metal can be replaced by any kind of material as long as it can exert a non‐uniform force on the material it contacts to generate a large strain gradient. We note that, when the sharp metal is replaced by a dielectric material, the flexoelectricity in that material should be also considered. For samples, the analysis approach of flexoelectricity in this work could be applicable to most crystalline dielectric materials such as TiO 2 and SiO 2 shown here, noting physical properties relevant to flexoelectricity are disparate in different materials, for example, materials with large dielectric permittivity usually exhibit large flexoelectricity. [ \n \n 13 \n \n ] Meanwhile, the redistribution of defects should be also taken into consideration when the dielectric materials are doped with high concentration of defects, for example, oxygen vacancies. If the dielectric thin film itself is not flat (e.g., corrugation surface) and exerted by a non‐uniform force, there will be intrinsic flexoelectric polarization by corrugation surface, [ \n \n 40 \n \n ] which would also contribute to the resultant charge transfer. In the case of dielectric bulks, flexoelectricity is generally supposed to be negligible because of relatively small strain gradient, while in certain cases, for example, bending a flake, flexoelectricity contribute to charge transfer in a similar way as described here. In the event of two metals, there might be no need to take flexoelectricity into consideration. Regarding non‐crystalline or semi‐crystalline materials, that is, polymers, the flexoelectricity is also supposed to affect the charge transfer in triboelectric measurement, nevertheless, it should be carefully analyzed because there could be plastic deformation under high forces and the mechanism of flexoelectricity in polymers is different from crystalline materials and could be dependent on various factors, such as motion and rotation of chains and cation sizes. [ \n \n 66 \n , \n 67 \n \n ] \n To further confirm the collaborative flexoelectric and triboelectric effects observed in the AFM, we fabricated triboelectric devices using a flat dielectric thin film and pyramid‐featured Pt thin film as the top and bottom layers, respectively, to simulate the conditions in the AFM measurements ( Figure \n \n 4 a(i) ). For comparison, a flat triboelectric device with flat layers of dielectric and Pt thin films (Figure 4a(ii) ) was also fabricated and tested as a prototype, where only the triboelectric effect contributed to the output voltage. In both devices, the output voltage performance of the triboelectric devices was measured using a pushing tester (the upper image of Figure 4a ). For comparison, two dielectric materials, SiO 2 and Si 3 N 4 , were chosen for the triboelectric device measurements owing to the higher and lower work functions than Pt, which was examined based on KPFM measurements (Figures S8a and S8b , Supporting Information). In the case of pyramid‐featured Pt and SiO 2 , the triboelectric effect is suppressed by the flexoelectric effect owing to opposite charge transfer directions, which leads to a smaller peak‐to‐peak voltage in the pyramid‐featured device than in the flat one, as shown in Figure 4b . Furthermore, increasing the force causes a larger difference because of the increased flexoelectric effect. The detailed output voltages at different forces are shown in Figure S8c,d , Supporting Information. In contrast, the surface potential of Si 3 N 4 is opposite that of SiO 2 , suggesting an opposite triboelectric charge transfer direction, which means that triboelectricity can be improved through the flexoelectricity. The results in Figure 4c distinctly indicate an enhanced triboelectric charging in the pyramid‐featured device, that is, the peak‐to‐peak voltage of the pyramid device is larger than that of the flat device (detailed output voltages are shown in Figure S8e,f , Supporting Information). Similarly, the difference between the pyramid‐featured and flat devices increases with increasing force. In both cases, the increasing contact area with force can be excluded as the main reason. The output voltage performance of the triboelectric device using TiO 2 as dielectric material is also shown in Figure S8g , Supporting Information. The output voltage of nanopatterned device is smaller than that of flat one, similar with the SiO 2 devices with slight difference, details can be found in Supporting Information. Consequently, the results of the flat and pyramid‐featured devices further verify the considerable contribution of concurrent flexoelectricity in triboelectric devices and suggest the selection of suitable material is important in a design of high efficiency nanopatterned devices. Figure 4 a) Schematic diagrams of triboelectric devices: i) pyramid‐featured Pt and ii) flat Pt. b) Peak‐to‐peak voltage as function of applied force (1 kgf = 9.8 n ) in triboelectric devices of flat Pt and pyramid‐featured Pt contact b) SiO 2 and c) Si 3 N 4 thin films."
} | 8,287 |
40253432 | PMC12009515 | pmc | 5,065 | {
"abstract": "Microbiome research is revolutionizing human and environmental health, but the value and reuse of microbiome data are significantly hampered by the limited development and adoption of data standards. While several ongoing efforts are aimed at improving microbiome data management, significant gaps still remain in terms of defining and promoting adoption of consensus standards for these datasets. The Strengthening the Organization and Reporting of Microbiome Studies (STORMS) guidelines for human microbiome research have been endorsed and successfully utilized by many research organizations, publishers, and funding agencies, and have been recognized as a consensus community standard. No equivalent effort has occurred for environmental, synthetic, and non-human host-associated microbiomes. To address this growing need within the microbiome research community, we convened the Microbiome Data Management in Action Workshop (June 12–13, 2024, in Atlanta, GA, USA), to bring together key decision makers in microbiome science including researchers, publishers, funders, and data repositories. The 50 attendees, representing the diverse and interdisciplinary nature of microbiome research, discussed recent progress and challenges, and brainstormed actionable recommendations and paths forward for coordinated environmental microbiome data management and the modifications necessary for the STORMS guidelines to be applied to environmental, non-human host, and synthetic microbiomes. The outcomes of this workshop will form the basis of a formalized data management roadmap to be implemented across the field. These best practices will drive scientific innovation now and in years to come as these data continue to be used not only in targeted reanalyses but in large-scale models and machine learning efforts.",
"conclusion": "Conclusions The Microbiome Data Management in Action Workshop assembled fifty participants across the microbiome research field to begin creating a consensus set of guidelines for environmental, synthetic community, and non-human host-associated microbiome data management and reporting. The participants emphasized that these goals require immediate action from every facet of microbiome research, and everyone has a responsibility to do their part in improving microbiome data management. The outputs from this workshop will assist in moving the field towards more FAIR microbiome data across agency borders, working to make the microbiome research field more equitable and inclusive. Data management best practices will benefit scientific innovation now and in years to come, as these data continue to be used in large-scale models and machine learning efforts. The outcomes of this workshop are intended to directly lead to an increased ability to more consistently and comprehensively manage microbiome data to enable reuse, thus unlocking research opportunities for all researchers, and improving the microbiome data ecosystem. This will positively impact environmental, synthetic community, and non-human host-associated microbiome research, which have demonstrated their importance to ecosystem health, food security, and environmental change.",
"introduction": "Introduction The field of microbiome research is rapidly growing with innovations spanning human, animal, plant, marine, and soil health, and implications for ecosystem resilience, nutrient cycling, food security, and environmental responses to extreme conditions [ 1 – 3 ]. Institutions, companies, and government agencies around the world have collectively invested billions of dollars into understanding human microbiomes and their relationships to health and disease [ 4 – 6 ]. The importance of environmental, non-human host, built environment, and synthetic microbiomes and their contributions to ecosystems is also broadly acknowledged within the research community, but the investment, infrastructure, and public recognition have historically lagged behind human-associated microbiomes [ 7 ]. However, this field is also experiencing a rapid rate of growth as the sense of urgency for preserving environmental health in the face of changing conditions becomes increasingly pertinent. Researchers study microbiomes using techniques that interrogate their composition and function, for example, by examining their associated DNA, RNA, proteins, and chemical signatures. Combinations of these investigations, known as multi-omics, require the generation of large data files that are difficult to produce, store, process, and share. These data are also produced by a range of methods and instruments, such that data produced by different researchers or institutions may not be directly cross-comparable or interoperable. The generation of microbiome data has vastly outpaced the development of data management infrastructure and consensus reporting standards. This disconnect hinders data reuse, including in meta-analyses and large-scale modeling efforts. Despite efforts to improve microbiome data management [ 8 – 12 ], these data are often not generated, stored, and shared according to the FAIR (Findable, Accessible, Interoperable, and Reusable) data principles that emphasize data reuse [ 13 ], which limits the growth of this field and compounds existing research siloes. In 2021, a group of human microbiome researchers collectively released the Strengthening The Organization and Reporting of Microbiome Studies (STORMS, https://www.stormsmicrobiome.org ) checklist which outlines reporting guidelines for human microbiome research publications [ 14 ]. This consensus checklist has been rapidly and widely adopted, and is already recognized as integral to promoting standardization across human microbiome and epidemiological data and studies. Data management across non-human host-associated microbiomes, the environmental sciences, and synthetic communities is in need of a STORMS-like reporting framework. However, across these diverse domains, it is currently challenging to consistently understand even basic study design or environmental parameters. While the existing Minimum Information about any (x) Sequence (MIxS) standard [ 15 ] supports sequence-specific metadata specifications including environmental ontologies, these do not encompass the minimal requisite information about microbiome datasets (e.g., study and sampling design, sample preparation or synthetic material construction, and other essential parameters to enable cross-study comparisons and data reuse). Significant gaps also remain in the adoption of existing standards and their application across diverse environment types and also multi-omics data [ 16 ]. The Microbiome Data Management in Action Workshop was convened from June 12–13, 2024, in Atlanta, GA, USA. The workshop was funded by the National Science Foundation (NSF) and was closely coordinated with the organizers of the subsequent American Society for Microbiology Microbe conference. The workshop aimed to address the increasing need for a robust data management ecosystem for microbiome research following the recommendation of the 2018 Interagency Strategic Plan for Microbiome Research [ 17 ] “to support the development of platform technologies to enhance data access and sharing.” The workshop brought together researchers, publishers, funders, and data repository representatives to identify local and national priorities. The workshop sought to tackle the theme of data management broadly, with a more specific focus on standardized reporting guidelines. The workshop attendees began generating a roadmap for microbiome data standards across non-human host-associated microbiomes, the environmental sciences, and synthetic communities leveraging the framework and lessons learned from the STORMS reporting guidelines."
} | 1,939 |
32796828 | PMC7429506 | pmc | 5,067 | {
"abstract": "Ecological communities often show changes in populations and their interactions over time. To date, however, it has been challenging to effectively untangle the mechanisms shaping such dynamics. One approach that has yet to be fully explored is to treat the varying structure of empirical communities—i.e. their network of interactions—as time series. Here, we follow this approach by applying a network-comparison technique to study the seasonal dynamics of plant-pollinator networks. We find that the structure of these networks is extremely variable, where species constantly change how they interact with each other within seasons. Most importantly, we find the holistic dynamic of plants and pollinators to be remarkably coherent across years, allowing us to reveal general rules by which species first enter, then change their roles, and finally leave the networks. Overall, our results disentangle key aspects of species’ interaction turnover, phenology, and seasonal assembly/disassembly processes in empirical plant-pollinator communities.",
"introduction": "Introduction Ecological communities are inherently dynamic. Their species composition is in constant change due to species’ intrinsic phenologies 1 , environmental variability 2 , 3 , and disturbances such as habitat fragmentation and invasive species 4 , 5 . In turn, the presence, absence, and intensity of ecological interactions also vary over time 6 . This happens either by the direct turnover of interacting species or by higher-order effects of changes in the community composition 7 . That is, the arrival of a new species in a community will come hand-in-hand with a new set of interactions, and these changes in the community will also indirectly interfere with existing interactions (e.g. potentially generating new cases of apparent competition between species 8 ). There is a longstanding tradition in ecology of exploring community dynamics using mathematical models 9 , 10 . The empirical basis of such models is generally static interaction networks, where nodes and links represent the different species and their observed interactions aggregated over sampling seasons 7 , 11 , 12 . While species’ interaction strengths are linked to their abundances in these models, the aforementioned dynamic nature of ecological interactions makes the underlying static representation of a community unrealistic—for example, plant-pollinator systems have been shown to present high levels of within-season species and interaction turnover 13 , 14 . Fortunately, several pioneering empirical studies have laid the groundwork for analysing natural systems over different time scales, providing crucial examples of the way ecological communities change within seasons 1 , 15 – 17 , across seasons 18 , 19 , and over much longer time scales 20 , 21 . These examples can often be represented as network time series, providing glimpses of ecological dynamics at whole-community scales. Such a network representation offers a way to assess how species enter and leave the community, and how they change their interactions over time. It therefore provides a valuable way to empirically interrogate three stages of community dynamics: assembly, intermediate dynamics, and disassembly 22 . There are several hypotheses on how each of these three stages might progress in natural communities. Within-season assembly of plant-pollinator networks seems to be described by ‘preferential attachment’, the mechanism by which newcomer species are more likely to attach to generalist ones 1 , 23 . However, this preferential attachment hypothesis is not without contention, as contrasting mechanisms seem to also explain the assembly process in other mutualistic systems 24 and time scales (e.g. ‘opportunistic attachment’ 19 ). The way communities disassemble, on the other hand, has been less frequently studied. Although some studies have shown the disassembly process in plant-pollinator communities to showcase the preferential loss of less-connected species (i.e. ‘preferential detachment’ 5 , 20 ), how disassembly plays a role within seasons is often unexplored. Finally, the bridge between community assembly and disassembly—its intermediate dynamics—is largely missing (but see Tylianakis et al. 25 ). The non-random structure of mutualistic communities suggest, nevertheless, a coherent dynamic in the way species enter the community, change their interactions, and leave the community. Perhaps one of the main obstacles for untangling this coherent dynamic is finding the ‘appropriate’ scale. For example, some studies have focused on the change of species composition over time, adopting a ‘full-network’ perspective to community dynamics. Unfortunately, while it appears useful to explain observed species distributions 26 , 27 , these studies often need to assume that interactions are independent from local changes in species abundances, thereby washing away a key component of community dynamics. Alternative approaches have found success quantifying temporal interaction turnover and linking such turnover to species’ phenologies 14 , 28 . Nevertheless, these ‘species-level’ approaches are often centred around quantifying variation of species interactions and lack the resolution to understand how such variation transforms the overall structure of ecological networks. That is, changes in species composition or interactions might not always translate into meaningful changes in the community structure (or vice versa). We employ an approach here to study the complete seasonal dynamics of plant-pollinator communities using the technique of network alignment 29 . Conceptually, aligning any two ecological networks proceeds by pairing up the species that play similar structural roles in each community 21 , 30 , 31 . This pairing essentially identifies species with analogous ‘positions’ across communities (Fig. 1 )—i.e. species that are similarly embedded in the corresponding network of interactions. It also offers a suitable scale to study community dynamics, one in which the state of any given species is always defined relative to all the other species in the community. This scale allows us to synthesize the information encoded within network time series, providing a comprehensive conceptual mapping of the changes in the communities and their many components. Fig. 1 Dynamics of a bipartite network. a An example of a time series for a plant-pollinator network. The circles and squares represent pollinators and plants, respectively. The links characterize interactions between these species. The colored species and links identify the changes made to the network over time. b The change in species' positions in the network time series represented in a . The different numbers describe different pollinator positions, and the different Greek letters describe different plant positions. The colored dotted lines indicate the position of two specific species a 3 and b 2 . On the one side, species a 3 change its position over time, starting in position ‘2' and ending up in position ‘1'. On the other, species b 2 preserves the same position 'β ' over time. Note how several species can have the same position in the network. In our study, we first use network alignment to assess the extent to which the positions of individual plant and pollinator species are variable within seasons. That is, given the alignment between the network observed in a community at two points in time, we use the information about who gets paired with whom to reveal whether and how species change their positions over time. We then evaluate the similarity of these positions across all of the data. In particular, we identify the set of distinct groups of species’ positions found across networks, representing the characteristic ways in which species tend to be embedded in their community. This allows us to synthesize the complex dynamics of individual species over time into something much simpler: the movement of species across groups of positions within the network. Characterizing this movement, we display a road map on how plants and pollinators first enter, then comprise, and ultimately leave the networks. This enables us to prune down the seasonal assembly and disassembly processes in empirical plant-pollinator communities, respectively. Likewise, this road map allows us to reveal the mechanisms by which species vary their positions in the community over the course of a season, which species will stay in the network the longest, and what positions species occupy before leaving the community. Overall, our study uncovers the underlying structural dynamics of plant-pollinator communities, leading us towards general rules regarding species’ interaction turnover, phenology, and assembly/disassembly processes in empirical plant-pollinator communities.",
"discussion": "Discussion In this work, we use a network-alignment technique as a way to disentangle the seasonal dynamics of plant-pollinator networks. First, we studied the uniqueness and variability of species’ positions within and across networks, respectively. We found that species have unique network positions at every time point, but they also tend to change such positions over time. Assessing the similarity of positions over time, we then found that there are major groups of positions characterizing plant-pollinator communities. These groups of positions provide a suitable scale to synthesizing complex ecological dynamics. For pollinator species, for example, they can be broadly described as: (i) specialist pollinators that interact with at least one generalist plant, (ii) generalist pollinators that interact with at least one other generalist plant, and (iii) specialist pollinators that interact with other specialist plants. Using these groups of positions, we estimated the underlying dynamics of species within seasons and found general rules regarding species’ seasonal dynamics within plant-pollinator communities. Putting this all together, our results suggest that the structure of plant-pollinator networks is extremely dynamic, where species rapidly switch positions within the network over a season. This structural dynamic, however, is also coherent across years, and one can predict the changes in species’ positions within networks over time. The study of network time series is challenging due to the many levels of information that these systems encompass. One could, for example, adopt a full-network perspective and study community dynamics using general network metrics 33 . Unfortunately, network metrics lack the resolution to distill the mechanisms by which species change positions over time 19 . Indeed, the study of metrics such as nestedness and connectance has shown certain mutualistic networks to exhibit generally constant structures over time 15 , 34 , 35 (but see CaraDonna and Waser 32 ). While this may be useful for understanding their dynamical stability and functioning 10 , 36 , 37 , these metrics are particularly ill-suited to understand the full scope of plant-pollinator seasonal dynamics. Alternatively, one could use single-species approaches. Ecological data, however, are often clouded by environmental variability 38 , 39 or sampling errors in the data collection 40 , both of which can add considerable noise to single-species dynamics. Perhaps most importantly, these approaches could also easily be overwhelmed by species’ natural idiosyncrasies 41 , which could mask potential general rules governing community dynamics. Indeed, we observed the effects of such idiosyncrasies when studying the uniqueness and variability of species’ positions. The high uniqueness of species’ positions indicates how singularly different species are embedded within a network; and, the high variability shows how sensitive these positions are to changes in the network structure. Noticeably, our observations on the variability of species positions also agree with recent work showcasing constant temporal switching of species’ interactions in empirical plant-pollinator communities 14 , 42 . Therefore, the commonly used static network representation, albeit useful in many cases, might strongly constrain our understanding of the dynamic nature of some ecological communities. Much like the artificial nature of the geographic boundaries between networks 43 , one could argue that temporal boundaries are just as artificial 44 . Here, we illustrate how it is possible to find a useful middle ground between full-network and single-species approaches. In particular, we focused on identifying distinct groups of positions within networks by clustering species with similar positions. This group scale allowed us to strategically prune down plant-pollinator dynamics. Assessing the movement of species across these groups of positions, one could, for example, focus on how pollinator species enter the community. As expected, we found the degree of newcomers to be generally low; we observe species entering the community mostly as specialists (group A from Fig. 3 ). These new-coming species tend to interact with at least one generalist plant, showing consistency with the idea of preferential attachment 1 . Likewise, the detachment of pollinators from networks often comes from groups of less-connected species (groups A and C from Fig. 3 ), also in agreement with the idea of preferential detachment 5 , 20 . The symmetry between these two processes—preferential attachment and detachment—has been showcased at longer time scales 25 as well as hypothesised to generate and maintain network patterns promoting stability 22 . For example, Tylianakis et al. 25 used simulations to show how commonly observed nested patterns can arise as a result of such symmetric processes. Our results provide a road map for how species change positions within the community, which positions are those that species take before exiting the network, and which species will likely remain in the network the longest. For example, our approach reveals that once in group A (specialists interacting with generalists), a species will often remain in this group, or simply leave the network. This pattern is observed with Anthophora terminalis (orange-tipped wood digger bee), which briefly appears in the network across years as a specialist pollinator interacting with generalist plants. Other species, however, may change their interactions in such a way that moves them from one group to the next (e.g. from group A, specialists interacting with generalists, to group B, generalists interacting with generalists). Species in group B will either stay in this group, or move to group A (rarely do we observe species in group B leaving the network). Bombus bifarius (two-form bumble bee) is such an example, and is often found in the networks as a generalist pollinator interacting with generalist plants. Finally, our results also reveal that when species are in group C (specialists interacting with other specialists), they will most likely move to group A or leave the network; Arctophila flagrans (a flower fly) showcases these transitions across years. Overall, we found species’ network-position dynamics to be consistent across the three different sampling seasons from our dataset. In other words, these general patterns we have reported—i.e. the way in which species first enter, then comprise, and ultimately leave the networks—are consistent from one year to the next. This is important because it showcases how network-position groups could be used as fundamental building blocks for understanding plant-pollinator community dynamics (similar to the concept of trophic components in predator-prey food webs 45 , 46 ). For the sake of simplicity, we chose the ‘short random walks’ algorithm to identify these position groups, as it provided us with a conveniently sized grouping of species positions. However, different community detection methods can partition the alignment matrix differently, providing different degrees of resolution to the dynamics of species across groups. Although, we showed how more complex partitions display more resolved dynamics (Supplementary Fig. 8 ), we also found that finer resolutions might lead to groups of positions that can be difficult to discern from each other in purely ecological terms (Supplementary Fig. 14 ). There are multiple factors that could play a role in explaining plant-pollinator dynamics beyond species’ degree. Information on species’ abundances is likely one such factor, as species’ phenologies have already been linked to interaction turnover in this system 14 , and abundances are good predictors of network structural properties in other ecological systems 47 , 48 . However, the evolutionary fingerprint underlying empirical communities 49 and evidence regarding pollinators’ plant preferences 50 , 51 point towards the idea that species’ abundances cannot represent the full picture regarding species’ interaction dynamics 52 . In some pollinator systems, for example, the pollinator costs of searching for trait-matching resources has been shown to be lower than the cost of switching to more abundant ones 53 , 54 . Indeed, species’ abundances are often considered emergent population-level properties that are ultimately constrained by species’ traits, and these same traits have been shown to effectively predict empirical plant-pollinator interactions 55 . That being said, the link between species’ abundances and networks’ structural dynamics define two ends of an interesting conceptual spectrum: one in which interaction variability can be explained solely based on species’ abundances, and the other where species interaction changes are completely independent from their abundances. Previous research indicates that empirical communities likely fall somewhere in between these two cases 6 , 25 , 56 , and our results also seem to support this idea. While pollinator observations and flower counts here showed positive relationships with species’ relative degree, these weak correlations left a lot of variation unexplained (Supplementary Fig. 9 ). Finally, we identify three areas we feel represent key steps from which to move forward. First, the approach used in the present work is not limited to plant-pollinator networks. Indeed, it could be used to shed light on the mechanisms governing many other systems, including food webs 57 , host-parasite communities 58 or other types of temporal networks 59 . Second, while we focused on temporal variation, another interesting perspective would be to put the same tools to work across other type of gradients 60 . For example, one could focus on the structural variability of plant-frugivore networks along forest-farmland gradients 3 , which could reveal how bird species change positions within networks in order to adapt to different environmental conditions. Third, we defined species’ positions purely based on the structure of plant-pollinator communities. Nevertheless, these positions could easily also account for other species’ properties such as species’ ecological traits and evolutionary histories 29 . This would allow us, for example, to study network dynamics from a functional diversity or evolutionary perspective, potentially untangling the eco-evolutionary mechanisms governing complex community dynamics."
} | 4,834 |
32782254 | PMC7419493 | pmc | 5,068 | {
"abstract": "Cooperative breeding may be selected for in animals when, on average, it confers greater benefits than solitary breeding. In a number of eusocial insects (i.e., ants, bees, wasps, and termites), queens join together to co-create new nests, a phenomenon known as colony co-founding. It has been hypothesised that co-founding evolved because queens obtain several fitness benefits. However, in ants, previous work has suggested that co-founding is a random process that results from high queen density and low nest-site availability. We experimentally examined nest-founding behaviour in the black garden ant, Lasius niger . We gave newly mated queens the choice between two empty nesting chambers, and compared their distribution across the two chambers with that expected under random allocation. We found that queens formed associations of various sizes; in most instances, queens group together in a single chamber. Across all experiments, the frequency of larger groups of queens was significantly higher than expected given random assortment. These results indicate colony co-founding in ants may actually be an active process resulting from mutual attraction among queens. That said, under natural conditions, ecological constraints may limit encounters among newly mated queens.",
"introduction": "Introduction Cooperative breeding is a social system in which organisms create communal nests, and it has evolved repeatedly in a range of taxa, including insects, fish, birds, and mammals 1 – 7 . In cooperative breeding, several adults engage in social behaviours that benefit both themselves individually and the group as a whole. This system may be selected for when ecological constraints (e.g., nest-site limitation, predation, parasitism, unpredictable resource availability) and competition greatly diminish the expected fitness payoff of solitary breeding. Cooperative breeding can result in greater nesting success because it enhances survival and reproduction, alloparental care, and/or collective nest defence 8 – 13 . Related individuals may nest together because they obtain fitness benefits, either directly or indirectly (i.e., via kin selection) 14 . Unrelated individuals may also nest together because they derive benefits arising from mutualism, reciprocity, and/or group selection 15 – 19 . Ecological constraints on solitary breeding appear to be major drivers of collaborative colony founding in the four main groups of eusocial insects—ants, bees, wasps and termites 20 – 25 . In the majority of ant species, foundation of a colony is the deadliest phase of the life cycle because newly mated queens are exposed to predation, starvation, disease, competition, and adverse environmental conditions (e.g. desiccation). Colony founding events have a very high failure rate, as high as 99% in some species [ 26 – 29 and references therein]. Although new colonies are created by single queens (haplometrosis) in most ants, the process can involve multiple queens (pleometrosis) in several species 20 , 26 , 30 . Founding associations have been documented across a dozen genera from three different ant subfamilies 20 . Collaborative colony founding, hereafter referred to as colony co-founding, is usually carried out by unrelated queens; therefore, it is unlikely to have evolved as a result of indirect fitness benefits 20 , 26 , 28 , 30 . In ants, colony co-founding enhances the productivity and success of incipient colonies because it increases queen survival during the early founding phase 31 – 33 ; promotes faster brood development 31 – 39 ; and boosts the size of the initial workforce, providing greater protection against intraspecific brood raiding, predation, and/or adverse abiotic conditions 32 , 36 , 37 , 40 – 43 . However, there is a cost associated with colony co-founding. In most species, the collaboration among queens is transient, and, after the first workers emerge, all but one of the queens are usually eliminated via queen fighting and/or culling by workers 20 , 42 . Co-founding a colony is therefore a risky endeavour: while the surviving queen will reap the full reproductive benefits of the colony, the defeated queens will have zero fitness. Thus, co-founding should be selected for when, on average, queens achieve higher fitness than they could have as solitary foundresses; conversely, it should be selected against when fitness benefits are significantly lower. Then, co-founding would result from random encounters when co-founding and solitary founding provide roughly equal benefits. Although considerable attention has been paid to the benefits of colony co-founding in eusocial insects, the proximate factors underlying the phenomenon have remained largely unexplored. In particular, it is unclear whether co-founding results from a random process in which queens are simply tolerant of one another (i.e., there is neither attraction nor repulsion) or whether it results from attraction among queens. Studies have shown that group size increases with increasing queen density in some ant species 43 – 45 . However, whether or not such associations were random was unclear. A laboratory study of co-founding in the ant Lasius pallitarsis suggested that queen association resulted from random allocation, but mutual attraction and active co-founding could occur with large queen density 45 . In the tree-nesting ant Crematogaster scutellaris , the number of groups formed by queens under natural conditions did not differ from that expected based on random allocation 46 , suggesting that newly mated queens were not actively co-founding colonies. However, this study did not take into account spatial variation in nest-site availability or the density of newly mated queens. Here, we examined whether colony co-founding could result from queens actively grouping together. We used the black garden ant, Lasius niger , as a model system (Fig. 1 a,b). In this species, mating occurs during large-scale nuptial flights, where thousands of sexuals from many colonies gather for a few hours. Once mated, queens land in an unknown environment, lose their wings, and quickly find a nesting site (small burrows in the open soil or under stones). In about 18–25% of cases, groups of 2–5 unrelated queens co-found colonies 42 , 47 . However, after the first workers emerge, queens start fighting with each other. Ultimately, only one queen survives, and she alone benefits from the colony’s future reproductive success. An experimental study of colony founding in L. niger offered newly mated queens an asymmetrical binary choice of nesting chambers: queens could settle either in an empty chamber or in a chamber containing another newly mated queen 42 . The study found that queens did not display a preference for either scenario, supporting the conclusion that colony co-founding was likely a random process promoted by high queen densities. To better understand the forces driving colony co-founding, we explored whether newly mated queens actively nested in groups. To this end, we presented newly mated queens with a symmetrical binary choice between two nesting chambers that were both initially unoccupied. We investigated how queen number affected the grouping patterns of queens across the two chambers by carrying out experimental trials involving two, four, and eight queens. Queens were allowed to move freely between the two chambers. We compared the observed grouping patterns of the queens across the two chambers after 24 h with the expected grouping patterns given random allocation based on stochastic simulations. Under conditions of random allocation, there would be no attraction among queens, and the queens would have an equal probability of ending up in either chamber ( p = 0.5). If queens were actively grouping together, frequencies of larger groups of queens would be higher than expected based on random allocation. Conversely, if queens were actively avoiding each other, frequencies of larger groups of queens would be lower than expected based on random allocation. Figure 1 Grouping patterns of L. niger founding queens. ( a ) Virgin winged L. niger queens embarking on their mating flight from their nest of origin. In the centre of the image is a male standing on the wings of a queen. Picture: Q. Willot. ( b ) After mating, queens land and then lose or tear off their wings. They subsequently search for small burrows in the ground in which they found new colonies, either alone or with other queens. Picture: H. Darras. ( c ) Queen grouping during one of the experimental trials: 8 newly mated queens have clustered in a single nesting chamber. ( d ) Proportion of observations as a function of the number of queens in the largest group sheltering within a chamber (dark grey), and theoretical distribution (light grey) based on the assumption of random assortment (see “ Methods ”). Experimental trials were performed with N = 2 queens ( n = 34), 4 queens ( n = 23), and 8 queens ( n = 25). For example, three situations were possible in trials with 4 queens: 2 queens in each chamber (2); 3 queens in one chamber and 1 queen in the other chamber (3); and all 4 queens in a single chamber (4). The graphs only show results for experimental trials in which all the queens were sheltered (i.e., none remained in the arena).",
"discussion": "Discussion We show that newly mated queens actively formed groups when given the choice between two empty nesting chambers. This suggests that colony co-founding in L. niger is an active process that results from mutual attraction among queens. Our results contrast with those from the few previous studies that have examined colony co-founding in ants, which assumed that the phenomenon resulted from queens simply being drawn to the safety of an enclosed nesting place rather than being drawn by the presence of other queens [see “ Introduction ” 42 , 46 , 48 , 49 ]. One study specifically stated that there was no attraction or repulsion between L. niger foundresses 42 . The discrepancy between their findings and our findings may stem from differences in methodology. In our study, queens were given a symmetrical choice between two empty nesting chambers. In contrast, in Sommer and Hölldobler’s study 42 , queens were given an asymmetrical choice: they could shelter in an empty chamber or in a chamber that already contained a queen. However, the study did not make clear how the latter queen was kept in the chamber or whether the potential retention method affected the queen’s behaviour. Moreover, the sample size was small, so the probability of making a type II error (wrongly failing to reject the null hypothesis of random allocation) was high. Another possibility is that queens from different populations differ in their colony founding strategies, as has been observed in the seed-harvester ant Pogonomyrmex californicus : in some populations, queens found colonies solitarily, whereas, in other populations, unrelated queens co-found colonies 30 , 50 . In the latter case, the associations persist as the colony matures, which means that colonies are headed by several reproductive queens ( i.e ., primary polygyny) 33 . Colony founding strategy is correlated with aggressiveness in P. californicus queens, and aggressiveness and tolerance phenotypes are strongly influenced by genetics 35 , 50 – 53 . Although it cannot entirely be excluded, this scenario seems unlikely in L. niger since (i) queens sampled in different parts of Europe have been observed to group together in the laboratory [e.g. 42 , 54 – 60 ] and (ii) collaboration among queens is unstable and always transforms into intense fighting when the first workers emerge, a phase that only a single queen survives. Our study was time limited and restricted to the grouping patterns of queens after 24 h. Clearly, additional studies should help decipher the mechanisms involved in the nesting choice of founding queens. Among these, is the probability for a queen to enter a chamber a function of the number of foundresses already present? Does the time spent searching for a shelter influence the probability for a queen to join other queens? What are the exact behavioural interactions among co-founding queens? Do queens move between shelters under laboratory or natural conditions and, if so, does the probability of leaving a shelter vary with the number of congeners in the same chamber? Also, the density of newly mated queens was probably much greater than in the field, a situation that increased the likelihood of queens clustering in the same chamber. However, our results clearly show that the queens’ grouping patterns were not random; they indicate that there was mutual attraction among queens. In L. niger , colony co-founding has been shown to confer clear demographic advantages, since multiple-foundress colonies have a higher rate of worker production than do single-foundress colonies 47 , 55 . The creation of a larger workforce within a shorter time period presumably enhances colony survival under natural conditions. Altogether, these findings suggest that the low frequency of colony co-founding in L. niger in nature (18–25% of incipient colonies) 42 , 47 is due to a lower likelihood of queens encountering each other. This encounter frequency could be diminished by low local densities of newly mated queens, high abundances of nest sites, and/or the need for queens to move into the first nest site they find to avoid desiccation or predation. In addition, the propensity of queens to co-found could depend on intrinsic factors, such as body weight or size, metabolic resources and, ultimately, the probability of surviving the conflict during reversion to single-queen colonies 20 , 45 , 54 , 61 , 62 . Joining behaviour indeed appears to be influenced by queen condition in the ant Lasius pallitarsis , where heavier queens are significantly more likely to join others than lighter queens, consistent with predictions of improved competitive ability 45 . So far, queen condition was however not shown to affect co-founding in the black garden ant 42 . In short, and in contrast to previous studies, this study shows that colony co-founding in the black garden ant, L. niger , is an active process likely rooted in strong mutual attraction among newly mated queens. It is possible that the same is true in other Formicidae. Identifying the specific mechanisms mediating this mutual attraction may be challenging from a technical perspective, but they should be explored in future research."
} | 3,633 |
35042817 | PMC8794879 | pmc | 5,069 | {
"abstract": "Significance Biofilms are multicellular, soft microbial communities that are able to colonize synthetic surfaces as well as living organisms. To survive sudden environmental changes and efficiently share their common resources, cells in a biofilm divide into subgroups with distinct functions, leading to phenotypic heterogeneity. Here, by studying intact biofilms by synchrotron X-ray diffraction and fluorescence, we revealed correlations between biofilm macroscopic, architectural heterogeneity and the spatiotemporal distribution of extracellular matrix, spores, water, and metal ions. Our findings demonstrate that biofilm heterogeneity is not only affected by local genetic expression and cellular differentiation but also by passive effects resulting from the physicochemical properties of the molecules secreted by the cells, leading to differential distribution of nutrients that propagate through macroscopic length scales.",
"discussion": "Discussion Spatiotemporal analysis of molecular structures within B. subtilis biofilms reveals that cellular spatial differentiation by function also translates into structural and elemental heterogeneity across whole biofilms at the molecular level. Probing intact biofilms with XRD, we have found that B. subtilis biofilms convey three characteristic structural signatures originating from spores, ECM components, and water in bound and free state. Concomitant spatial elemental analysis by XRF showed that the metal ions, mainly, Ca, Zn, Mn, and Fe, known to play a crucial role in bacterial metabolism and sporulation ( 69 ), differentially accumulate in B. subtilis biofilms. While Ca and K are evenly distributed across the biofilms, other metal ions, mainly Zn, Mn, and Fe, accumulate along biofilm wrinkles. Based on our observations, we suggest an inclusive view of biofilm development linking between macroscopic features, namely biofilm wrinkles, to elemental and structural heterogeneities in biofilms, as illustrated schematically in Fig. 6 . We suggest a dual role for the ECM in biofilms. It provides structural support and leads to wrinkle formation, and it also serves as filter, selectively binding Ca over other metal ions. These two properties result with heterogeneous distribution of metal ions and spores, as detailed in the next paragraph. Fig. 6. Schematic representation of the relationships between biofilm structures, metal ion distribution, and their implications on biofilm physiology. ( A ) A side (cross-section) view of a WT B. subtilis biofilm with a central water-filled (blue) wrinkle. Bacterial cells are represented by brown ovals; spores are represented by circles with a dark brown circumference. The straight arrows represent water and nutrient uptake into the biofilm, and the curled arrows represent water evaporation from the biofilm surface. ( B ) The biofilm is composed of cells and ECM (represented by brown background and darker lines). ( C ) TasA fibrils in the ECM contain short cross β-sheet domains. ( D ) Zoom into a column across the biofilm wrinkle (black rectangle); water (blue background) and metal ions (we only refer to the metal ions that were observed in this study, Ca: purple and Fe/Zn/K/Mn: red and green) are free inside water channels ( Bottom ). The metals are adsorbed by the ECM ( Middle ), with preference to Ca that remains mostly bound to the ECM. Zn, Mn, and Fe are free to diffuse in the matrix, and they concentrate on the biofilm wrinkles ( Top ), in which water evaporation is the largest. Metal ion accumulation on biofilm wrinkles possibly leads to sporulation. This image is for illustration purposes, and it is not drawn to scale. Metal ions, initially residing in the medium, accumulate in the ECM biopolymers with preferred Ca binding. This process is driven by water evaporation, which occurs throughout the film (curled blue arrows in Fig. 6 ). However, Wilking et al. ( 33 ) have shown that biofilm wrinkles act as water channels, in which water flow is driven by enhanced evaporation along the wrinkles. The water flow carries nutrients, but, as Ca is selectively bound by the matrix and therefore filtered out, the solution is relatively enriched with Zn, Mn, and Fe, which are only weakly bound by the matrix or taken up by cells. These ions eventually accumulate along the wrinkle because of water evaporation. The cooccurrence of these metal ions and spores along wrinkles is consistent with their essential role in sporulation ( 65 , 70 , 71 ) and points at the possibility that spores act as a drainage for metal ions, as a means to circumvent toxicity. Our study also suggests that, in the absence of wrinkles in ECM mutant biofilms, a Ca/metal ion ratio required for sporulation cannot be achieved, providing reason to the delayed sporulation in ECM mutant biofilms. Our model therefore offers a functional link between ECM properties, macroscopic wrinkles, and sporulation via heterogenic metal ions distribution, showing that biofilm heterogeneity is not only affected by genetic expression and cellular differentiation but also by passive processes that stem from physicochemical properties of molecules secreted by the cells in the biofilm. These lead to a differential distribution of nutrients that propagates through macroscopic length scales. Multiscale approaches to biofilm internal structures and metal ion relationships may hold the key to understanding biofilm physiology and multicellularity and their relation to subpopulation survival in B. subtilis , as well as in other biofilms of single or mixed bacterial species."
} | 1,394 |
30514205 | PMC6280343 | pmc | 5,070 | {
"abstract": "Background Constraint-based modeling is a widely used and powerful methodology to assess the metabolic phenotypes and capabilities of an organism. The starting point and cornerstone of all such modeling is a genome-scale metabolic network reconstruction. The creation, further development, and application of such networks is a growing field of research thanks to a plethora of readily accessible computational tools. While the majority of studies are focused on single-species analyses, typically of a microbe, the computational study of communities of organisms is gaining attention. Similarly, reconstructions that are unified for a multi-cellular organism have gained in popularity. Consequently, the rapid generation of genome-scale metabolic reconstructed networks is crucial. While multiple web-based or stand-alone tools are available for automated network reconstruction, there is, however, currently no publicly available tool that allows the swift assembly of draft reconstructions of community metabolic networks and consolidated metabolic networks for a specified list of organisms. Results Here, we present AutoKEGGRec, an automated tool that creates first draft metabolic network reconstructions of single organisms, community reconstructions based on a list of organisms, and finally a consolidated reconstruction for a list of organisms or strains. AutoKEGGRec is developed in Matlab and works seamlessly with the COBRA Toolbox v3, and it is based on only using the KEGG database as external input. The generated first draft reconstructions are stored in SBML files and consist of all reactions for a KEGG organism ID and corresponding linked genes. This provides a comprehensive starting point for further refinement and curation using the host of COBRA toolbox functions or other preferred tools. Through the data structures created, the tool also facilitates a comparative analysis of metabolic content in any given number of organisms present in the KEGG database. Conclusion AutoKEGGRec provides a first step in a metabolic network reconstruction process, filling a gap for tools creating community and consolidated metabolic networks. Based only on KEGG data as external input, the generated reconstructions consist of data with a directly traceable foundation and pedigree. With AutoKEGGRec, this kind of modeling is made accessible to a wider part of the genome-scale metabolic analysis community. Electronic supplementary material The online version of this article (10.1186/s12859-018-2472-z) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusion Here we present AutoKEGGRec, a fast tool to be used within the COBRA toolbox in Matlab that is able to create single organism, community, and consolidated first draft reconstructions based on the KEGG database. Unlike most other available tools, AutoKEGGRec is designed from the ground up to allow for community and consolidated metabolic network reconstructions. Also, being based only on data from KEGG, it has a clear and transparent data ancestry. AutoKEGGRec does not provide a fully functioning model since elements such as a biomass function, ATP maintenance function, and most transport reactions for the given organism(s) are not included. The addition of these elements would require estimations and guesswork beyond the data available in KEGG at present. However, AutoKEGGRec does provide a very well annotated first draft reconstruction based on all data available in KEGG. As with reconstructed networks generated with any software package, the user should perform additional experiments, and add biomass function(s), ATP maintenance, and relevant import reactions to further enhance the reconstruction; all tasks which may be performed using existing functions in the COBRA toolbox. Consequently, AutoKEGGRec fills an important niche for the modeling community with its reliable and rapid generation of well annotated single, community, and consolidated FDRs within the COBRA toolbox environment based only on the clear, high-quality, and freely accessible data in KEGG.",
"discussion": "Discussion AutoKEGGRec is an easy-to-use tool for the rapid assembly of first-draft reconstructions for genome-scale metabolic modeling including as much data as possible. While there exists a selection of tools and toolkits that aid in the assembly and refinement of first-draft GEMs, these tend to be stand-alone and/or online, both of which, especially in combination with small development teams in temporary academic positions, may cause any number of issues due to lack of support. These tools mostly create single-organism FDRs, as most are not designed for or allow the creation of community or generic/consolidated models. Any automatic reconstruction necessitates steps such as gap-filling, mass and charge balance for reactions, reaction reversibility corrections, and adjustment of exchange, ATP maintenance and biomass reactions. These steps being mostly automated might appear as a convenience, but it is important to note that these tools often make implicit assumptions in the course of the reconstruction work without marking them as such. This is done in order to achieve functional models, i.e. models that are capable of producing the constituents of a biomass reaction, but, importantly, there are usually many ways of achieving this, and which way is chosen matters for the correctness and quality of the model, making control and traceability of the process an important issue. Many such educated guesses are based on template models or common knowledge, such as the inclusion of seemingly universally essential cofactors [ 33 ] in the biomass function. However, the way in which this information is retrieved and assembled is often not completely transparent, leaving important details of the model’s workings outside of a user’s control or awareness. Any resulting FDR may therefore contain misleading information, which is further propagated in reconstruction projects. Biomass reactions in particular are known to be fickle with regards to cofactor coefficients [ 34 ]. This may lead to false growth predictions with regards to media and knockouts when naively propagated from skeleton models. Therefore, the manual curation of all reactions, and especially the biomass reaction, is still the recommended approach for most applications, and the added value of including template features such as biomass functions is debatable. To varying degrees, these published tools automate the generation of first draft reconstructions, and greatly facilitate the reconstruction of GEMs. AutoKEGGRec, however, is not intended to automatically generate a first draft model with all requisite transport reactions and a biomass function. Since all information is based only on KEGG, every detail is eminently traceable, which may be of particular interest to modelers, as many “new” models are based on previous models, and may therefore inherit previous poor annotations and assumptions. The intent is to generate an unbiased and clean FDR of the metabolic network based only on the genome annotation in KEGG. It is implemented as a Matlab function fully integrated with the COBRA toolbox, with its open, community-maintained and state-of-the-art suite of tools for model curation and manipulation. Additionally, AutoKEGGRec contains the complete annotation for all reactions and compounds offered by KEGG, putting more information right at the modeler’s fingertips than what is offered by most other comparable reconstruction tools. In its rapid assembly of a draft reconstruction, AutoKEGGRec mostly covers the currently automatable parts of the first stage in the protocol to generate high-quality genome-scale metabolic reconstructions [ 35 ]. Additionally, it allows not only the creation of an FDR for a single organism, but also consolidated and community FDRs. These features set AutoKEGGRec apart from published methods. Consequently, it is a valuable tool for the current-day practitioner of genome-scale metabolic modeling and reconstruction. While the reconstruction of microbial consortia remains an immature method, AutoKEGGRec can help speed up the process. Further work in assembling the reconstruction is still expected to rise steadily with the number of organisms, as not only does each separate organism’s metabolic network needs to be curated, but also the interactions between the different networks and their global effect on the environment. However, due to the chosen naming convention for compartments within the community models, transport reactions can easily be added simultaneously for multiple organisms if necessary."
} | 2,161 |
24348089 | PMC3856425 | pmc | 5,071 | {
"abstract": "Several laboratories are pursuing the synthesis of cellular systems from different directions, including those that begin with simple chemicals to those that exploit existing cells. The methods that begin with nonliving components tend to focus on mimicking specific features of life, such as genomic replication, protein synthesis, sensory systems, and compartment formation, growth, and division. Conversely, the more prevalent synthetic biology approaches begin with something that is already alive and seek to impart new behavior on existing cells. Here we discuss advances in building cell-like systems that mimic key features of life with defined components."
} | 166 |
31844276 | PMC6986341 | pmc | 5,072 | {
"abstract": "Early insights into the unique structure and properties of native silk suggested that β-sheet nanocrystallites in silk would degrade prior to melting when subjected to thermal processing. Since then, canonical approaches for fabricating silk-based materials typically involve solution-derived processing methods, which have inherent limitations with respect to silk protein solubility, stability in solution, and time and cost efficiency. Here we report a thermal processing method for the direct solid-state molding of regenerated silk into bulk ‘parts’ or devices with tunable mechanical properties. At elevated temperature and pressure, regenerated amorphous silk nanomaterials with ultralow β-sheet content undergo thermal fusion via molecular rearrangement and self-assembly assisted by bound water to form a robust bulk material that retains biocompatibility, degradability and machinability. This technique reverses presumptions about the limitations of direct thermal processing of silk into a wide range of new material formats and composite materials with tailored properties and functionalities."
} | 276 |
35814004 | PMC9260013 | pmc | 5,075 | {
"abstract": "Global warming is approaching an alarming level due to the anthropogenic emission of carbon dioxide (CO 2 ). To overcome the challenge, the reliance on fossil fuels needs to be alleviated, and a significant amount of CO 2 needs to be sequestrated from the atmosphere. In this endeavor, carbon-neutral and carbon-negative biotechnologies are promising ways. Especially, carbon-negative bioprocesses, based on the microbial CO 2 -metabolizing chassis, possess unique advantages in fixing CO 2 directly for the production of fuels and value-added chemicals. In order to fully uncover the potential of CO 2 -metabolizing chassis, synthetic biology tools, such as CRISPR-Cas systems, have been developed and applied to engineer these microorganisms, revolutionizing carbon-negative biotechnology. Herein, we review the recent advances in the adaption of CRISPR-Cas systems, including CRISPR-Cas based genome editing and CRISPR interference/activation, in cyanobacteria, acetogens, and methanogens. We also envision future innovations via the implementation of rising CRISPR-Cas systems, such as base editing, prime editing, and transposon-mediated genome editing.",
"conclusion": "Concluding Remarks In this review, we summarized recent advances in developing and applying CRISPR-Cas systems for CO 2 -metabolizing chassis. CRISPR-Cas-based genome editing and CRISPRi, have been reported in these microbes, and the methods have been advancing biorefinery and bioproduction with CO 2 as the carbon source, exhibiting great potential in alleviating CO 2 emissions and in reducing atmospheric CO 2 levels. However, more efforts are imperative to awake the full power of CRISPR-Cas systems in these CO 2 -metabolizing chassis. CRISPRa, base editing, prime editing, and transposon-mediated integration may offer encouraging future directions in developing novel CRISPR-Cas systems for CO 2 -metabolizing microorganisms. Moreover, discoveries of new CRISPR-Cas systems with special properties (e.g., a thermostable Cas9) are needed to engineer CO 2 -metabolizing microorganisms, such as thermophilic strains Thermoanaerobacter kivui and Methanothermobacter thermautotrophicus ( Moon et al., 2019 ; Fink et al., 2021 ). Besides CO 2 , CO, methane, methanol and formate are also important greenhouse gases and C1 compounds that can be obtained from waste gases and products or byproducts of clean energy industries. As such, natural and engineered C1-metabolizing microbes, including but not constrained in the autotrophs discussed here, will also be favorable microbial chassis for sustainable bioproduction. The development of novel synthetic biological tools, such as CRISPR-Cas systems, for C1 metabolizing organisms, will significantly foster innovations in carbon-negative biotechnologies.",
"introduction": "Introduction Anthropogenic emission of carbon dioxide (CO 2 ) has driven an unprecedented high level of CO 2 in the atmosphere, leading to an approximately 1.1°C increase in the average global temperature ( Tollefson, 2021 ). This increase has reached an alarming level and has caused global and local climate issues. It also leaves a very small window to achieve the 1.5°C target settled in the Paris Agreement and reinforced in the UN Climate Conference in Glasgow (COP26). The atmospheric CO 2 level must be lowered by a significant amount by controlling the emission of CO 2 and sequestering CO 2 from the atmosphere at the same time. In this endeavor, biotechnology provides promising routes. In one way, carbon-neutral biotechnology utilizes sustainable carbon sources (e.g., agriculture and forest wastes) to produce chemicals (e.g., ethanol, butanol, and 2,3-butanediol), alleviating the reliance on fossil fuels and reducing CO 2 emissions ( Liu et al., 2020 ). In another way, carbon-negative biotechnology directly consumes industrial or atmospheric CO 2 for the bioproduction of fuels and value-added chemicals ( Liu et al., 2020 ; Liew et al., 2022 ). With metabolic engineering and synthetic biology, the inventory of products from biological routes has been greatly expanded, and the production and yield have been improved. For instance, the baker’s yeast Saccharomyces cerevisiae has been genetically engineered to convert lignocellulosic feedstock to bioethanol and chemicals, exhibiting the potential of carbon-neutral biotechnology ( Wei et al., 2013 ; Sun et al., 2021 ). Recently, S cerevisiae has been engineered to utilize wasteful CO 2 accumulated during lignocellulosic sugar fermentation by the installation of a CO 2 fixation pathway, transforming the correlated biotechnology from a carbon-neutral process to a carbon-negative technology ( Li et al., 2017 ; Xia et al., 2017 ). Inspiringly, Gassler et al. (2020) generated an engineered yeast Pichia pastoris capably of growing with CO 2 and methanol, opening a new window for heterotrophic yeast to use one-carbon (C1) compounds as sole carbon sources. Similar enterprises have been made in Escherichia coli , and artificial autotrophic E. coli has been generated via the implementation of CO 2 fixation pathways and adaptive laboratory evolution ( Antonovsky et al., 2016 ; Gleizer et al., 2019 ; Flamholz et al., 2020 ). Another biological path is to employ microorganisms that metabolize CO 2 innately, such as photoautotrophic cyanobacteria and chemoautotrophs, including acetogens and methanogens. These organisms can use CO 2 as a carbon source from either industrial waste gases or the atmosphere ( Fackler et al., 2021 ). CO 2 -metabolizing microorganisms have shown great potential as microbial chassis, and industrial attempts have been made ( Liu et al., 2020 ; Liew et al., 2022 ). Given the advances in synthetic biology, these microbes play more important roles on the path towards a sustainable future with enhanced CO 2 utilization efficiency and an expanded spectrum of products. For instance, cyanobacterium Synechocystis sp. PCC 6803 has been modularly engineered to produce a high titer of 1-butanol, short/medium-chain carbohydrate, and lactate from CO 2 ( Liu X. et al., 2019 ; Shabestary et al., 2021 ; Yunus et al., 2022 ). Lately, a pioneer study conducted by LanzaTech, Inc. (Skokie, IL, United States) shows that Clostridium autoethanogenum can convert syngas (consisting of CO 2 , CO, and H 2 ) to acetone and isopropanol, and a pilot-scale fermentation in a 125-L scalable reactor was demonstrated ( Liew et al., 2022 ). These advances have validated the capability of CO 2 -metabolizing chassis in the fixation of CO 2 and production of value-added chemicals, and these succuss illustrated the ever-increasing power of synthetic biology in biotechnology. CRISPR-Cas systems, the bacterial and archaeal immune systems, have been repurposed as synthetic biology tools for gene editing and regulation ( Knott and Doudna, 2018 ). They have been revolutionizing biotechnology in fundamental ways. Though still in its infant stage, multiple CRISPR-Cas-based synthetic biology tools have been developed for cyanobacteria, acetogens, and methanogens, driving the rising of novel biotechnologies based on CO 2 -metabolizing microbes. Herein, we summarize the current progress of CRISPR-Cas systems in genetically engineering microbial CO 2 -metabolizing chassis, especially cyanobacteria, acetogens, and methanogens, for the conversion of CO 2 to biofuels and value-added products, and we discuss the challenges and future endeavors in developing more efficient synthetic biology tools."
} | 1,875 |
35480677 | PMC9037800 | pmc | 5,077 | {
"abstract": "Motion is a basic behavioral attribute of organisms, and it is a behavioral response of organisms to the external environment and internal state changes. Materials with switchable mechanical properties are widespread in living organisms and play crucial roles in the motion of organisms. Therefore, significant efforts have been made toward mimicking such architectures and motion behaviors by making full use of the properties of stimulus-responsive materials to design smart materials/machines with specific functions. In recent years, the biomimetic motions based on micro/nanomotors, actuators and soft robots constructed from smart response materials have been developed gradually. However, a comprehensive discussion on various categories of biomimetic motions in this field is still missing. This review aims to provide such a panoramic overview. From nano-to macroscales, we summarize various biomimetic motions based on micro/nanomotors, actuators and soft robotics. For each biomimetic motion, we discuss the driving modes and the key functions. The challenges and opportunities of biomimetic motions are also discussed. With rapidly increasing innovation, advanced, intelligent and multifunctional biomimetic motions based on micro/nanomotors, actuators and soft robotics will certainly bring profound impacts and changes for human life in the near future.",
"conclusion": "5. Conclusions and outlook In conclusion, an important characteristic of living organisms is the intelligent response to external stimuli. Building stimuli-responsive materials and machines could help or even replace the human body to perform a variety of functions, which will have a significant influence on society. The appearance of bionic technology provides a new idea for our research and provides a reference for designing excellent micro/nanomotors, actuators and soft robotics. Up to date there are a large number of methods developed to use smart response materials to prepare various micro/nanomotors, actuators and soft robotics driven by chemical/biochemical reactions and external fields, which then are used to mimic the motion behaviors of organisms ( Table 1 ). Remarkable progresses have been made in both understanding and controlling the motion behaviors based on micro/nanomotors, actuators and soft robotics. The extensive research into biomimetic motions from nano-to macroscales has achieved many key functions of organisms, such as capturing, jumping, swimming, collective behavior and so on. However, there are still many challenges for biomimetic motions, including the limitation of the practical application environment, the same active deformation ability as the creatures in nature, and the difficulty of high maneuvering and flexible motion. Therefore, extensive efforts will still be made to biomimetic motions based on micro/nanomotors, actuators and soft robotics in innovative ways. A brief summary of biomimetic motions based on micro/nanomotors, actuators and soft robotics Categories Typical examples Power sources Motion behaviors Limitations Micro/nanomotors Micro/nanorods, micro/nanoparticles, micro/nanotubes Chemical reaction, light, magnetic, ultrasonic, electric Movement, collective behaviors Precise regulation of motions and performance of specific tasks in special environments cannot be achieved. Actuators Bilayer structures, gripper, flower, etc. Light, electronic, magnetic, temperature Flicking, jumping, walking, etc. Soft robotics Soft grasping robot, soft crawling robot, soft camouflage robot, soft growth robot and soft mechanical fish Biological muscle cells, osmotic actuation, friction, light, electronic, magnetic, etc. Swimming, grasping, crawling, releasing, curling In the future, with the further development of science and technology, with the more research of perception, special functional structure, motion mechanics and motion control, we believe it is possible to realize truly perception and movement in the real environment, and which will be applied to industrial and agricultural production and benefit mankind.",
"introduction": "1. Introduction Nature has evolved high-performance objects using commonly found materials from the nanoscale to the macroscale. Biologically inspired design or adaptation or derivation from nature is referred to as ‘biomimetics’. Biomimetic motions are derived from the many different functional materials and/or intricate and highly organized structure of the biological material from the molecular to the nanoscale, microscale and macroscale. For example, the early developments of the wing design of airplanes was inspired by birds' several consecutive rows of covering flexible feathers on the wings, which develop the lift by movable flaps. The sandfish found in the Sahara desert moves over the sand. The scales on its body and biomaterials used in the body provide wear resistance. In addition, there are a large number of objects, including bacteria, plants, land and aquatic animals from nature and their selected functions will sever as the inspiration for various biomimetic motions. An important characteristic of living matter is the intelligent behavioral response to external stimuli. In nature, the communication between living organisms is well-established and is essential for the ecosystems, which gives us a lot of inspirations. 1–4 For example, geckos are able to crawl by the adhesion of tens of thousands of extremely fine bristle fibrils on the soles of their feet with the surface structure. 5–8 Scyphomedusae ephyra , the juvenile of the most widely distributed jellyfish, can smartly control the fluidic flow around their body to realize diverse functionalities. 9–13 Plants exhibit hydration-trigged changes in their morphology due to differences in local swelling behaviour that arise from the directional orientation of stiff cellulose fibrils within plant cell walls. 14–17 Inspired by plants and animals in nature, a variety of stimuli-responsive materials and machines are fabricated to realize various functions and motion behaviors by mimicking the mechanisms and motion behaviors of various biological systems. 18–21 The biomimetic behaviors could help or even replace the human body to perform a variety of functions, which will have a significant influence on society. 22,23 For example, Whitesides et al. 22 constructed the multi-legged robots that mimic some of the important musculoskeletal features of arthropods, which is well suited for safe robot–human interaction. In addition, R. V. Martinez et al. 23 reported soft actuators and robots that can display remarkable resistance to mechanical damage. These soft robotic structures have the potential to be used in congested and hazardous spaces, which will protect humans from harm. Materials with switchable mechanical properties are widespread in living organisms and endow many species with traits that are essential for their survival. Therefore, materials with stimuli-responsive properties are the foundation of achieving biomimetic motion. When stimuli-responsive materials are exposed to a highly specific or predefined stimulus, the materials are able to change their mechanical properties on command or stimulation. In addition to the stimuli-responsive materials, the design of the structure and the mechanism of movement are the key factors that determine the motion. Bioinspired design is a general approach of applying biological principles to develop new solutions. 24–27 By fully understanding the motion mechanism of organisms, researchers design and realize biomimetic motion by taking advantage of the switchable mechanical properties of materials. It is instructive that most of the mechanically adaptive materials found in nature rely on simple mechanisms and building blocks. The structures of building elements and modulation of interactions are the key factors to achieve different and complex functions. 28,29 Many examples of artificial stimuli-responsive materials being designed to mimic the motion behaviors of living things have been reported, with applications ranging from micro/nanomotors to soft robotics. 30–42 This review is aimed to provide such a panoramic discussion of existing various biomimetic motions. Based on the motion mechanism of organisms, the intelligent materials/machines with specific functions are designed by making full use of the properties of stimulus-responsive materials. In this review, the motion behaviors from macroscopic to microscopic are divided into the following three categories ( Fig. 1 ): (1) biomimetic motions based on micro/nanomotors. In this section, we mainly introduce various structures and driven modes of micro/nanomotors and the applications in biomedicine, environmental remediation and so on. What's more, the swarm behavior of micro/nano motors imitating organisms is also discussed. (2) Biomimetic motions based on actuators. In this section, the compositions and biomimetic behaviors (jumping, locomotion, walking, capturing, etc. ) of actuators based on smart materials such as hydrogels, shape memory materials and liquid crystal elastomers are reviewed. (3) Biomimetic motions based on soft robotics. A variety of biomimetic motions, such as snake-like, tendril-like and underwater jellyfish-like have been designed, and the motion mechanisms and motion behaviors have been summarized in this section. We finally briefly review the opportunities and challenges of biomimetic motions. Fig. 1 Overview of various biomimetic motions based on micro/nanomotors, actuators and soft robotics."
} | 2,385 |
19249281 | null | s2 | 5,079 | {
"abstract": "Memory storage on short timescales is thought to be maintained by neuronal activity that persists after the remembered stimulus is removed. Although previous work suggested that positive feedback is necessary to maintain persistent activity, here it is demonstrated how neuronal responses can instead be maintained by a purely feedforward mechanism in which activity is passed sequentially through a chain of network states. This feedforward form of memory storage is shown to occur both in architecturally feedforward networks and in recurrent networks that nevertheless function in a feedforward manner. The networks can be tuned to be perfect integrators of their inputs or to reproduce the time-varying firing patterns observed during some working memory tasks but not easily reproduced by feedback-based attractor models. This work illustrates a mechanism for maintaining short-term memory in which both feedforward and feedback processes interact to govern network behavior."
} | 245 |
39934908 | PMC11817178 | pmc | 5,080 | {
"abstract": "Background Beta-diversity is a fundamental ecological metric for exploring dissimilarities between microbial communities. On the functional dimension, metaproteomics data can be used to quantify beta-diversity to understand how microbial community functional profiles vary under different environmental conditions. Conventional approaches to metaproteomic functional beta-diversity often treat protein functions as independent features, ignoring the evolutionary relationships among microbial taxa from which different proteins originate. A more informative functional distance metric that incorporates evolutionary relatedness is needed to better understand microbiome functional dissimilarities. Results Here, we introduce PhyloFunc, a novel functional beta-diversity metric that incorporates microbiome phylogeny to inform on metaproteomic functional distance. Leveraging the phylogenetic framework of weighted UniFrac distance, PhyloFunc innovatively utilizes branch lengths to weigh between-sample functional distances for each taxon, rather than differences in taxonomic abundance as in weighted UniFrac. Proof of concept using a simulated toy dataset and a real dataset from mouse inoculated with a synthetic gut microbiome and fed different diets show that PhyloFunc successfully captured functional compensatory effects between phylogenetically related taxa. We further tested a third dataset of complex human gut microbiomes treated with five different drugs to compare PhyloFunc’s performance with other traditional distance methods. PCoA and machine learning-based classification algorithms revealed higher sensitivity of PhyloFunc in microbiome responses to paracetamol. We provide PhyloFunc as an open-source Python package (available at https://pypi.org/project/phylofunc/ ), enabling efficient calculation of functional beta-diversity distances between a pair of samples or the generation of a distance matrix for all samples within a dataset. Conclusions Unlike traditional approaches that consider metaproteomics features as independent and unrelated, PhyloFunc acknowledges the role of phylogenetic context in shaping the functional landscape in metaproteomes. In particular, we report that PhyloFunc accounts for the functional compensatory effect of taxonomically related species. Its effectiveness, ecological relevance, and enhanced sensitivity in distinguishing group variations are demonstrated through the specific applications presented in this study. \n Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40168-024-02015-4.",
"conclusion": "Conclusions In this work, we introduce a novel metric PhyloFunc and provide its method of computation. The PhyloFunc metric integrates phylogenetic information with taxonomic and functional data to better capture beta-diversity in gut metaproteomes, offering sensitive insights into microbial ecology responses in health and disease applications. To streamline the calculation of PhyloFunc distances, we developed the Python package PhyloFunc , which automates the process of calculating functional distances between sample pairs and generates comprehensive distance matrices for multiple samples. This enables efficient assessment of metaproteomic functional beta-diversity across datasets.",
"discussion": "Discussion Metaproteomics is an informative approach to studying the functionality of the human gut microbiome and its implications in human health and disease. Evaluation of beta-diversity is often one of the initial steps in metaproteomics data exploration. However, there has been a lack of a measurement tool that effectively captures the ecology-centric variations in metaproteomics data. The beta-diversity of gut metaproteome samples is influenced not only by the abundance of taxa and taxon-specific functional compositions but also by the phylogenetic relatedness between taxa. Therefore, including phylogenetic information with protein group taxonomic and functional annotations can better empower researchers to explore both the functional and ecological dynamics of microbial communities, offering insights much overlooked by solely considering taxonomic and functional abundances. Here, we proposed a novel beta-diversity metric, PhyloFunc, which provides a comprehensive perspective to better detect functional responses to drugs by incorporating phylogenetic information to inform functional distances. Through a simulated dataset, we illustrated the calculation process and indication of the PhyloFunc distance method. This simple toy dataset makes it possible for readers to follow the calculations and understand the hierarchy of PhyloFunc algorithm more effectively. It hierarchically incorporates functional abundance of proteins, taxonomic abundance, and phylogenetic relationship between taxa. As demonstrated by the proof-of-concept toy dataset, as well as a real-world dataset, we report that PhyloFunc distance can account for the functional compensatory effect among taxonomically related species and offered a more ecologically relevant measurement of functional diversity compared to the three established distance methods tested. Functional compensation can mitigate the impact of species loss or functional changes on the overall ecosystem function, thereby helping maintain ecosystem functions. Research has shown that functional compensation among closely related species with harboring functional redundancy is a key mechanism in sustaining ecosystem functions in response to environmental stimulants [ 31 , 32 ]. Our PhyloFunc metric is built on such a mechanism, leveraging the functional roles of related taxa to provide a more ecologically relevant measure of ecological beta-diversity. Furthermore, we tested PhyloFunc using a dataset of in vitro drug responses of a human gut microbiome. We first showed that for drugs exhibiting strong effects, PhyloFunc distance showed agreements with other distance metrics. Interestingly, we further observed that for drugs exerting milder effects, the PhyloFunc method can detect new responses and achieve better classification evaluation results than the other tested distance measures, providing deeper insights into drug-microbiome interactions. This result suggests PhyloFunc’s potential for clinical applications. By offering deeper insights into how various drugs affect the functional ecology of the human gut microbiome, PhyloFunc could be useful in developing personalized medicine approaches [ 33 ], optimizing drug therapies, and understanding the microbial basis of drug efficacy and side effects. Apart from drug-microbiome interactions, the PhyloFunc metric has significant potential across an even-broader range of applications. These applications extend to any area where evaluation of microbial ecology responses is required, including but not limited to personalized nutrition, prebiotics/probiotics development, disease diagnostics, etc."
} | 1,735 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.