pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
34552138 | PMC8458349 | pmc | 2,692 | {
"abstract": "Dinoflagellates in the family Symbiodiniaceae are obligate endosymbionts of diverse marine invertebrates, including corals, and impact the capacity of their hosts to respond to climate change-driven ocean warming. Understanding the conditions under which increased genetic variation in Symbiodiniaceae arises via sexual recombination can support efforts to evolve thermal tolerance in these symbionts and ultimately mitigate coral bleaching, the breakdown of the coral-Symbiodiniaceae partnership under stress. However, direct observations of meiosis in Symbiodiniaceae have not been reported, despite various lines of indirect evidence that it occurs. We present the first cytological evidence of sex in Symbiodiniaceae based on nuclear DNA content and morphology using Image Flow Cytometry, Cell Sorting and Confocal Microscopy. We show the Symbiodiniaceae species, Cladocopium latusorum , undergoes gamete conjugation, zygote formation, and meiosis within a dominant reef-building coral in situ. On average, sex was detected in 1.5% of the cells analyzed (N = 10,000–40,000 cells observed per sample in a total of 20 samples obtained from 3 Pocillopora colonies). We hypothesize that meiosis follows a two-step process described in other dinoflagellates, in which diploid zygotes form dyads during meiosis I, and triads and tetrads as final products of meiosis II. This study sets the stage for investigating environmental triggers of Symbiodiniaceae sexuality and can accelerate the assisted evolution of a key coral symbiont in order to combat reef degradation.",
"conclusion": "Conclusion This study is the first to categorically demonstrate sexual reproduction in Symbiodiniaceae, establishing a foundation from which to explore the potential role of symbiont evolution in coral resilience to global change. Based on DNA content and morphological evidence, we propose that Symbiodiniaceae species may follow the same two-step meiotic process described for other dinoflagellates, in which the first meiotic division produces a dyad of cells, whereas the second division produces an intermediate triad state, with meiosis II ultimately resulting in a tetrad stage of haplontic cells. This process may be under circadian control, as most putatively meiotic cells were detected at night. Beyond basic biology, understanding sexuality in Symbiodiniaceae can advance experimental evolution work on this group, with the goal of enhancing the capacity of coral holobionts to cope with warming ocean temperatures and other stressors under rapid global change.",
"introduction": "Introduction Reef-building corals and other marine invertebrates establish obligate symbioses with a diverse group of dinoflagellates in the family Symbiodiniaceae (reviewed in 1 , 2 ). This symbiosis can be disrupted by environmental stressors including elevated sea surface temperatures (SSTs) and increased UV radiation, resulting in bleaching—the mass loss of Symbiodiniaceae cells and/or chlorophyll from the host—and frequently, host mortality 3 , 4 . Thermal stress due to anthropogenic climate change is recognized as the leading cause of coral reef degradation 5 – 7 . Despite this, the onset of coral bleaching from 2007 to 2017 occurred at significantly higher SSTs (+ ~ 0.5 °C) than the preceding decade 8 . This suggests that thermally susceptible genotypes may have adapted and/or declined such that the thermal threshold for bleaching has increased. As reefs continue to experience thermal stress under committed (and likely additional) warming due to climate change, supporting the assisted evolution of thermal tolerance in Symbiodiniaceae 9 is critical to increasing reef resilience 10 and contributing to the restoration of ecologically and economically valuable ecosystems 11 . The most direct mechanism for adaptation to environmental challenges is sex 12 , 13 . Sexual recombination of parental genotypes during meiosis promotes new (and potentially beneficial) genetic combinations in offspring, the basic prerequisite for evolution via natural selection. Indeed, various field observations and experimental evolution studies across diverse taxa have documented that stressful or novel environments can select for higher levels of sexuality 14 , and microorganisms, including Symbiodiniaceae, are predicted to have a high adaptive capacity in selective environments 15 , 16 . Meiosis is the hallmark of sex, consisting of two nuclear divisions (karyokinesis) and one simultaneous or two successive cytoplasmic divisions (cytokinesis). Recombination in meiosis ‘mixes’ genetic material from both parents to increase genetic variation in the progeny, in contrast to mitosis—the division typical of ordinary cell growth—where daughter cells have the same number and kind of chromosomes as the parent cell 17 . More than 10% of the approximately 2000 known marine dinoflagellate species produce cysts and are thought to exhibit facultative sexuality during part of their life cycle 18 . In these dinoflagellates, reproduction is primarily asexual (through mitosis, Fig. 1 A), but sex can be induced within a subset of cells in a population under certain environmental conditions. Foundational studies, dating back to the 1970s, linked dinoflagellate sexuality to the formation of highly resistant, benthic stages (‘resting cysts’), considered a mechanism for surviving harsh environmental conditions 19 . When resting cysts germinate, meiosis results in the release of novel genotypes that are potentially better adapted to local conditions. Although dinoflagellate sex was first proposed to be rare in nature 20 , research over the last decade revealed that sex in these microeukaryotes is a relatively frequently and flexibly utilized reproductive mechanism. The capacity for sex in dinoflagellates is also now recognized as independent of a species’ ability to form resting cysts (see reviews by 21 , 22 ). Initial studies of Crypthecodinium cohnii suggested that dinoflagellates could undergo only a one-step meiosis 23 (Fig. 1 B.1), but later works on different species consistently reported the existence of a two-step meiotic process ( 24 and references therein). In two-step meiosis, there is a delay in meiosis II: a single division occurs in the zygote, whereas the second division takes place at postzygotic stages (Fig. 1 B.2). Despite these advances, sexuality remains difficult to identify in most dinoflagellate species due to (i) morphological similarities between sexual and vegetative stages; and (ii) the potential for co-occurrence of 2C DNA content stages derived from both mitosis (haploid) and gamete fusion (diploid) within the same population of cells. Given this, a general consensus has emerged that the detection of a fourfold DNA content stage, which is formed during meiosis (but not mitosis), is key to identifying sex in dinoflagellates 25 – 29 . Figure 1 Differences in DNA content and ploidy state between the mitotic and the meiotic cycle, including the two meiotic processes proposed for dinoflagellates (one-step and two-step meiosis). A growing body of molecular evidence shows that Symbiodiniaceae possess functional sexual machinery, and thus suggests that these key coral reef symbionts can reproduce sexually. Indirect evidence for sexual reproduction in this group includes: (i) the existence of a sufficient inventory of essential Symbiodiniaceae meiotic genes 30 , 31 , as well as genes related to gamete formation 32 ; and (ii) population-level genetic signatures 33 – 36 and codon usage trends 32 most parsimoniously interpreted as arising from meiotic recombination. Upregulation of meiosis-related genes has also been documented to occur under thermal stress 37 , 38 . This temperature-associated regulation suggests that sexual reproduction may be key for the adaptation of Symbiodiniaceae under current warming trajectories, driven by climate change. In contrast to this growing body of evidence, genomic evidence for the absence of canonical synaptonemal complex (SC), as well as for a reduced set of cohesin complex genes 32 , have been reported from this dinoflagellate family. The synaptonemal complex (SC) mediates the pairing of homologous chromosomes during the early stages of meiotic prophase I and cohesin proteins play a role in sister chromatid cohesion. However, the absence of the SC and reduction of cohesin complex genes does not preclude meiotic capability in Symbiodiniaceae; similar patterns have been reported in other dinoflagellates known to be sexual 32 . Despite strong molecular evidence of sexual reproduction in Symbiodiniaceae, no direct cytological proof for fertilization and meiosis in this group have been available. The first cytological descriptions of Symbiodiniaceae (which was previously recognized as a single genus, Symbiodinium 2 ) life cycle stages 39 indicated the existence of motile gymnodinoid zoospores and vegetative cells (the dominant, non-motile stage). Freudenthal 39 observed that vegetative cells (haploid) either divide by binary fission or form cysts, which are characterized by a thicker wall (Fig. 2 A). Cysts could divide or turn into a zoosporangium, which could either release a swimming gymnodinoid zoospore or remain as a non-motile spore (aplanospores, Fig. 2 A). In cultures described as “old”, which could indicate nutritional deficiencies, cysts were observed to contain dividing autospores (according to their external morphology, typically two and rarely four). Under certain conditions (not clarified, although the cultures used were clonal), cysts could even give rise to multiple cells resembling a process of gametogenesis. However, morphologies related to gamete conjugation were not detected 39 . Fitt and Trench 40 subsequently argued that the term “coccoid stage” should be used to describe the non-motile form of Symbiodiniaceae (as opposed to “cyst”). This is because “cyst” in dinoflagellates is usually related to a dormant (non-active), resistant (thick wall) stage, whereas “coccoid stage” can be used independently of a cell’s metabolic activity or cell wall thickness (a highly variable character). Instead, the haploidy of the coccoid (vegetative) stage was considered key to sexuality by Fitt and Trench 40 , who argued that if the coccoid stage was haploid, doublets and emerging motile cells result from a mitotic division, whereas tetrads could represent sexual stages resulting from meiotic division. A summary of this proposed life cycle 41 is shown in Fig. 2 A. Later works based on nuclear reconstructions 42 and microsatellites 43 supported this hypothesis of a sexual cycle in Symbiodiniaceae, as they provided molecular evidence of haploidy in vegetative stages of diverse species in Breviolum (a Symbiodiniaceae genus formerly known as ‘ Symbiodinium clade B’ 2 , 43 ). Previous work has shown that algal endosymbionts of other dinoflagellate taxa (i.e., Peridinium balticum ) can sexually reproduce 44 , providing general support for hypothesized sexuality in intracellular symbionts. Although seminal, previous cytological descriptions of the Symbiodiniaceae life cycle and potential sexual stages remain incomplete (e.g., without evidence of gamete fusion) and lack supporting nuclear images and DNA content analyses 39 , 40 ; additional cytological analyses of Symbiodiniaceae life stages are necessary to directly demonstrate sexuality in this key dinoflagellate family of reef symbionts. Figure 2 A. Summary of previous direct observations and hypotheses regarding the Symbiodiniaceae life cycle, as well as new information generated by this study. ( A ) Schematic figure (modified from LaJeunesse ( http://tolweb.org/ ), which was based on (Fitt and Trench 40 ) summarizing previous direct observations and hypotheses regarding the Symbiodiniaceae life cycle. Previously published direct observations include the production of two mobile haploid cells (mastigotes, referred to as ‘zoospores’ by Freudenthal 39 ) from mitosis within the coccoid stage (termed ‘cysts’ or ‘aplanospores’ by Freudenthal 39 ), which could behave as isogametes or transform into coccoid stages. The formation of zygotes through gamete fusion, as well as the formation of tetrads (called ‘autospores’ by Freudenthal 39 ) via meiosis were hypothesized but not documented. ( B ) Schematic view of the results of the present study in relation to the previously proposed Symbiodiniaceae life cycle (in A ). Discriminating morphological features (nuclei, pyrenoids and accumulation bodies) are shown in the sexual stages unless in dyads, triads and tetrads, as these stages are transitory and were found in different evolving grades. ( C ) Confocal images corresponding to the proposed sexual stages depicted in ( B ). Here, we provide new cytological evidence for sexuality in Symbiodiniaceae, focusing on nuclear processes (regardless of motility stage). A combination of flow cytometry techniques (image flow cytometry and sorting) and high-resolution confocal microscope imaging were conducted on populations of the Symbiodiniaceae species, Cladocopium latusorum , fixed from the tissues of a dominant coral genus (Pocillopora spp.) sampled on a South Pacific reef . Our work provides the first direct cytological evidence of meiosis and gamete function in Symbiodiniaceae and suggests that sexual reproduction can occur in hospite under natural conditions. These findings open the door to exploring the conditions that promote sex, as well as potential variation in sexual recombination rates, among Symbiodiniaceae species.",
"discussion": "Discussion Foundational studies previously generated evidence delineating much of the Symbiodiniaceae life cycle, and strong molecular evidence indirectly supported the existence of a sexual cycle in this group of dinoflagellates. However, cytological proof of sexual reproduction in Symbiodiniaceae was still needed to advance our understanding of the basic biology of the ecologically and economically valuable Symbiodiniaceae-coral mutualism, and to catalyze subsequent research into when, where, and how sex occurs in this dinoflagellate group. This study is the first to apply cutting edge approaches (IFC, sorting and confocal analyses) to identify Symbiodiniaceae cells with DNA content and nuclear processes that can definitively be interpreted as sexual activity, including the identification of fusing gametes, zygotes and cells in profase I of meiosis (“4C”, uninuclear cells). Although not a conclusive proof of meiosis, the formation of dyads, triads and tetrads aligns with a meiotic two-step process already described in other dinoflagellates (e.g. 29 , 45 , 46 ). However, a two-round, asynchronous mitosis cannot be discarded with the available data. Below, we highlight key DNA content and cell morphology observations that allow us to establish differences between mitosis and meiosis in Symbiodiniaceae cells, compare our hypothesis for sex in Symbiodiniaceae to the sexual stages reported in other dinoflagellate species, and highlight outstanding questions regarding the conditions that promote Symbiodiniaceae sex. Identifying sexual stages: key differences between mitotic and meiotic cells Image flow cytometry (IFC) indicated that most of the Symbiodiniaceae cells processed in this study fell into a single group representing the vegetative stage, characterized by low DNA content (“1C”, haploid) and a single, oval to round-ish nucleus. As recently shown in other dinoflagellate species 47 , 48 , close examination of the haploid cells with confocal microscopy showed that chromosomes were not all identical as previously thought 49 , but in fact, highly variable in size (e.g., Fig. 4 C’). Other single nucleus cells fell between “1C” and “2C” DNA content and were interpreted to be replicating their DNA as part of the mitotic cycle (“S” phase). However, some cells with a single nucleus had a DNA content higher than “2C”; these cells were consistent with a replicating zygote in meiosis I during a two-step meiosis (Fig. 1 B.2). Following a similar dichotomy to the cells with one nucleus, cells with two nuclei could have either “2C” DNA content and be in a mitotic cycle, or have a DNA content > 2C and be in a non-mitotic cycle. The two nuclei of cells in this latter group varied in size, shape and chromatin condensation state; such cells were interpreted as precursors to a triad stage (i.e., cell with three nuclei). To distinguish cells that were part of this non-mitotic sequence, cells with two nuclei and DNA content between “2C” and “4C” were considered dyads, as opposed to what we will simply call “mitotic coccoid stages”. Cells with three and four nuclei (triads and tetrads, respectively) were also detected. Although the morphologies of these stages were difficult to analyze in the IFC images (Fig. 3 D,E), such cells could be examined at higher magnification using sorting and confocal microscopy. The variability observed in tetrad morphology (Fig. 9 ) indicates these cells were dividing stages derived from triads (Fig. 8 ), in what we interpret to be a delayed meiosis II. Taken together, we infer that the observed Symbiodiniaceae tetrad cells were undergoing two-step meiosis (Fig. 1 B.2); this reproductive strategy has been observed in most studied dinoflagellates. For example, Prorocentrum micans and Prorocentrum minimum form tetrads as a final meiotic product 29 . Additionally, asynchronous divisions of the zygote in these Prorocentrum species also lead to the formation of triads 29 . It should be noted that some free-living dinoflagellates (e.g., members of the genus Alexandrium ) produce chains of cells during two consecutive mitotic divisions. In these Alexandrium species, cell chains are formed frequently and are readily detectable in any growing culture. Since Symbiodiniaceae do not form such chains in culture, it is unlikely that the low percentage of tetrad cells observed in our study represent the product of a two-round mitosis. However, this possibility cannot be totally ruled out with the present data. Instead, this study documents Symbiodiniaceae cells with (> 2C–4C) DNA content and a single nucleus that have a morphology consistent with a replicating zygote—such a cell stage is not possible within a mitotic cycle and thus constitutes the first direct proof of meiosis in this dinoflagellate family. If Symbiodiniaceae engage in sexual reproduction, albeit at low levels within the observed cell populations, then other sexual stages, such as zygotes, should also be observable. Here, we briefly review how mitosis proceeds in Symbiodiniaceae since cells diverging from this process can be recognized as zygotes. According to Freudenthal 39 , during the mitotic division process, all cellular inclusions are equally distributed among the daughter cells with the exception of the accumulation body, which persists as a single unit in the parent cell. Karyokinesis occurs followed by cytoplasmic division, which is initiated by the formation of an equatorial zone of constriction. Following mitotic division, each new daughter cell produces a new cell wall within the old cell wall of the parent cell 50 , 51 . The old cell wall material is then degraded via an unknown process 52 , releasing the daughter cells. Based on this description of mitosis, only one accumulation body should be present in mitotically dividing cell stages (as shown in Fig. 5 ). Some of the 2C-single nucleus cells in this study were therefore identifiable as zygotes because they contained two accumulation bodies (Fig. 6 A–C); such cells could not have been undergoing mitotic division. The nuclear morphology and pyrenoid count of some cells also allowed zygotes to be distinguished from mitotically dividing cells. For example, 2C-two nuclei cells with duplicated pyrenoids would only be observed at advanced stages of mitotic division; such cells were frequently detected in our study (e.g., Fig. 5 C,D). However, we also observed cells that contained only one “2C” DNA content nucleus but had two well-developed pyrenoids (Figs. 6 D, 7 A); such cells could not have been undergoing mitosis and were interpreted as fusing gamete products or early stage zygotes (these differences in accumulation bodies and pyrenoids are summarized in Fig. 2 B.2 mitotic stage versus 2B.3 zygote). Sexual stages of Symbiodiniaceae are similar to those reported in free-living dinoflagellates In free-living dinoflagellates, the identification of zygotes in non-resting stage cells (planozygotes) has been impeded due to the high morphological similarity between planozygotes and mitotic cells. In past studies, the number of flagella was considered the hallmark of a zygote (although this characteristic is unreliable due to the weak nature of flagella under fixation). Number of flagella is inapplicable to Symbiodiniaceae, however, since this family alternates between mobile and coccoid stages (Fig. 2 A), the latter of which lacks flagella. Therefore, the key to morphological discrimination of zygotes in Symbiodiniaceae (and other dinoflagellates in non-resting stages) lies in differentiating features of sex from mitosis, either during nuclear fusion of gametes or the meiotic process. The process of zygote formation we posit for Symbiodiniaceae here is very similar to that observed in other dinoflagellates. For example, during gamete fusion in the naked dinoflagellates Gymnodinium catenatum and Gymnodinium nolleri , karyogamy occurs first, and, during the process, one gamete nucleus migrates to the position of the other, and they fuse. This occurs while the cell wall is in early stages of fusion, allowing two fusing cytoplasms to be distinguished 45 , 46 (Fig. 10 , first row). In other species, such as Prorocentrum micans , the process looks similar at the nuclear level, but the cytoplasms never fuse, and instead, one of them degenerates 26 . Early stages of meiosis in Gymnodiniaceae are characterized by a big and round-ish zygotic nucleus that changes into a bi-lobed form with a central ‘cytoplasmic channel’, and a DNA-decondensed state in which chromosomes appear thinner (Fig. 10 , second row). In dinoflagellates, chromosome segregation occurs via binding to the nuclear envelope surrounding the cytoplasmic channels and microtubule bundles 53 . Although the formation of cytoplasmic channels was first described during mitosis 54 , 55 , it was later also confirmed to occur during meiosis. Specifically, a main channel centrally positioned during meiosis I is often visible via conventional fluorescence microscopy 45 , 46 . Given this, some of the one-nucleus Symbiodiniaceae cells with (2C–4C) DNA content recorded here (Fig. 7 A,B) could represent zygotes in early meiosis I. Figure 10 Comparative images show similarities between sexual stages of Gymnodiniaceae and Symbiodiniaceae. ( A – C ) Fusing gametes . ( A ) Putative gamete fusion in Symbiodiniaceae. ( B , C ) Nuclear fusion during syngamy in Gymnodinium catenatum . ( D – F ) Zygotes. ( D ) Putative zygote nucleus in Symbiodiniaceae. ( E , F ) Zygote nuclei in G. catenatum and Gymnodinium nolleri in early meiosis I . A central cytoplasmic channel (arrow, cc) is observed in some zygotes, becoming the bilobed nucleus during meiosis I. Gymnodiniaceae pictures are original, corresponding to the same time lapse series published in 45 , 46 . Some characteristics and processes described here for Symbiodiniaceae have also been reported for members of the plant kingdom. For example, the final product of male meiosis in flowering plants is a tetrad of haploid microspores enclosed in a polysaccharide cell wall, and meiosis I often leads to the formation of dyads. Each dyad can divide again to form tetrads through an asynchronous meiosis II division 56 . In Arabidopsis , two nuclear divisions occur before simultaneous cytokinesis yields a tetrad of haploid cells. Additionally, in some Arabidopsis mutants, cell divisions are delayed, resulting in the formation of abnormal intermediates, most frequently dyad meiotic products 57 . The hypothesized process for gamete conjugation proposed here aligns with previous observations of Symbiodiniaceae by Freudenthal 39 and Fitt and Trench 40 . These researchers indicated that during Symbiodiniaceae cell division, karyokinesis (nuclear division) occurs first, and later an equatorial zone of constriction in the cytoplasm separates the two daughter cells, which split pyrenoids but not accumulation bodies. As occurs in other dinoflagellate species (e.g. 45 , 46 ), we propose that nuclear fusion is faster than cytoplasmic fusion during the process of gamete conjugation, given the existence of cells with elongated external shapes and duplicated pyrenoids and accumulation bodies but a single “2C” DNA content nucleus; such cells further corroborate ongoing gametogenesis and zygote formation. Thus, the hypothesized Symbiodiniaceae life cycle (summarized in Fig. 2 A) put forward by initial, foundational works 39 , 40 constitutes the foundation of our updated life cycle (Figs. 2 B,C, 11 ). This study documents the entire meiotic process in Symbiodiniaceae and includes: (i) the novel observation of (2C–4C) DNA content cells with a single nucleus and duplicated pyrenoids and accumulation bodies (i.e., direct proof of meiosis); (ii) the identification of previously unpredicted intermediate stages, as dyads and triads; and (iii) the first images of tetrads, which had a relatively linear morphology, compared to the previously described coccoid morphology (Figs. 2 B4, 9 vs Fig. 2 A4, respectively). Integrating our observations with foundational works, we provide a revised proposed life cycle for Symbiodiniaceae (Fig. 11 ). Figure 11 Updated proposed life cycle hypothesis for Symbiodiniaceae, based on the findings of this study. Vegetative cells are characterized by the presence of three unique elements: nucleus (N), pyrenoid (PY) and accumulation body (AB). Mitotic cells replicate their DNA, forming a larger (2C) nucleus. Nuclear division (karyokinesis) occurs first, followed by cytoplasmic division (cytokinesis). Pyrenoids are observed to duplicate in advanced cytokinesis stages, when the outer cellular morphology is already indicative of two cells. Zygotes (2 N, diploid) have a larger nucleus than vegetative cells due to the nuclear fusion of gametes. Two pyrenoids and accumulation bodies are present after fusion as a result of the cytoplasmic contribution of each gamete. DNA replicates once, giving rise to a 4C DNA content cell, which then divides into two cells (dyad) during meiosis I. Asynchronous division during meiosis II leads to the formation of triads and eventually to a 4-cell stage (tetrad) of haploid, 1C DNA content cells. Outstanding questions regarding Symbiodiniaceae sex Both techniques (IFC and conventional flow cytometry) applied in this study indicated that sex occurs at relatively low levels in Symbiodiniaceae in hospite (typically occurring in less than 1% of cells observed in a sample and at a maximum of ~ 5% of cells); sex was only detectable using high resolution imaging following cell sorting. Conventional flow cytometry has a lower capacity to discriminate cells from other fluorescent particles or aggregates; this is likely why the technique reported a slightly higher percentage of cells with (2C–4C) DNA content than IFC analyses (Table 1 , Fig. 3 B). Regardless, findings from both IFC and conventional flow cytometry agree with previous molecular analyses that indicated Symbiodiniaceae display a mixed reproductive strategy, which is mainly asexual with occasional to frequent sex 35 . In free-living dinoflagellates, sex also originally appeared to occur rarely, such as in the case of severe nutrient deficiency 20 . However, other studies concluded that sexual reproduction in dinoflagellates is probably more common and flexible in nature than previously thought, but induced under species-specific environmental conditions 21 , 22 . Now that sexual stages are confirmed for Symbiodiniaceae, subsequent studies can investigate the biotic and abiotic factors promoting sex, as well as physiological details of Symbiodiniaceae sexual stages. It is worth noting that each colony analyzed in this study contained only Cladocopium latusorum symbionts. Rates of sexuality for Symbiodiniaceae in hospite may correlate with population- and/or species-level genetic variation; this should be tested in future works. In our study, sexuality occurred more frequently from 18 to 0 h (compared to 6 h and 12 h); this correlation agrees with a previous report for Alexandrium minutum 28 , in which sexual 4C peaks were mainly detected during the dark period, after gradually increasing from the final hours of the light period. However, in this study, we did not identify a relationship between sexuality and temperature stress (or colony ID). This is similar to work by Bellantuono et al. 38 , which did not identify enriched GO terms for meiosis I in a Durusdinium trenchii strain exposed to elevated versus ambient temperatures, whether the strain was in hospite or in culture. Levin et al. 37 , however, reported ≥ fourfold up-regulation of meiosis genes, as well as enrichment of meiosis functional gene groups, in two heterogeneous cultured populations of Cladocopium exposed to elevated temperatures. Additional studies that more comprehensively test for the abiotic and biotic triggers of sexuality in a diversity of Symbiodiniaceae strains, populations, species and genera are needed (Fig. 12 ). Advancing our understanding of the role of Symbiodiniaceae sex in nature, as well as constraints on sex in this group, can potentially be leveraged to enhance the resilience of reef coral colonies to climate change. For example, the induction of new genetic diversity within a Symbiodiniaceae species via enhanced sexuality, may contribute to rapid symbiont adaptation to thermal stress 58 . Indeed, sexual reproduction could underlie local adaptation of Symbiodiniaceae populations to thermal stress, and help explain some within host species variation in bleaching susceptibility 59 – 61 . Assisted evolution efforts could potentially leverage this by generating more thermally robust populations of homologous symbionts, which can then be provided to aposymbiotic recruits of horizontally transmitting coral species 62 , 63 . Figure 12 Outstanding question box. Summary of important questions for future research."
} | 7,625 |
31068721 | null | s2 | 2,693 | {
"abstract": "Software implementations of brain-inspired computing underlie many important computational tasks, from image processing to speech recognition, artificial intelligence and deep learning applications. Yet, unlike real neural tissue, traditional computing architectures physically separate the core computing functions of memory and processing, making fast, efficient and low-energy computing difficult to achieve. To overcome such limitations, an attractive alternative is to design hardware that mimics neurons and synapses. Such hardware, when connected in networks or neuromorphic systems, processes information in a way more analogous to brains. Here we present an all-optical version of such a neurosynaptic system, capable of supervised and unsupervised learning. We exploit wavelength division multiplexing techniques to implement a scalable circuit architecture for photonic neural networks, successfully demonstrating pattern recognition directly in the optical domain. Such photonic neurosynaptic networks promise access to the high speed and high bandwidth inherent to optical systems, thus enabling the direct processing of optical telecommunication and visual data."
} | 294 |
39234547 | PMC11371792 | pmc | 2,694 | {
"abstract": "Thermodynamics has predicted many different kinds of microbial metabolism by determining which pairs of electron acceptors and donors will react to produce an exergonic reaction (a negative net change in Gibbs free energy). In energy-limited environments, such as the deep subsurface, such an approach can reveal the potential for unexpected or counter-intuitive energy sources for microbial metabolism. Up until recently, these thermodynamic calculations have been carried out with the assumption that chemical species appearing on the reactant and product side of a reaction formula have a constant concentration, and thus do not count towards net concentration changes and the overall direction of the reaction. This assumption is reasonable considering microorganisms are too small (~1 μm) for any significant differences in concentration to overcome diffusion. However, recent discoveries have demonstrated that the reductive and oxidative halves of reactions can be separated by much larger distances, from millimetres to centimetres via conductive filamentous bacteria, mineral conductivity, and biofilm conductivity. This means that the concentrations of reactants and products can indeed be different, and that concentration differences can contribute to the net negative change in Gibbs free energy. It even means that the same redox reaction, simultaneously running in forward and reverse, can drive energy conservation, in an ElectroMicrobiological Concentration Cell (EMCC). This paper presents a model to investigate this phenomenon and predict under which circumstances such concentration-driven metabolism might take place. The specific cases of oxygen concentration cells, sulfide concentration cells, and hydrogen concentration cells are examined in more detail.",
"conclusion": "5 Conclusion Electromicrobiological concentration cells are fascinating theoretical constructs which, if found to exist in the real world, could greatly impact our understanding of microbial element cycling, physiology, and the origin of life on Earth and elsewhere. I hope that other scientists will be inspired by this paper to explore this concept further, both theoretically and experimentally.",
"introduction": "1 Introduction Microbial communities inhabit spatially heterogeneous and complex environments that resist simple understanding. While most of our understanding of microbial physiology is the result of more than a century of cultivating and isolating pure cultures of microorganisms, this approach has proven inadequate for understanding all microbial diversity, with 22–87% of archaeal and bacterial genera remaining uncultivated depending on the environment ( Lloyd et al., 2018 ). Spatially heterogeneous environments such as sediment, the terrestrial subsurface, and soil contain the highest fractions of uncultivated taxa ( Lloyd et al., 2018 ). Such spatially heterogeneous environments contain concentration gradients of microbial substrates, such as O 2 , sulfide, and H 2 . It has recently been shown that certain microorganisms can link metabolic reactions taking place across these gradients via long-distance electron transport between microbial cells, thus allowing them to better exploit optimal substrate concentrations than a single cell. One example of this is the cable bacteria Electronema and Electrothrix ( Trojan et al., 2016 ; Plum-Jensen et al., 2024 ), which were discovered through careful measurement of chemical microprofiles in aquatic sediment ( Pfeffer et al., 2012 ). Cable bacteria are filamentous bacteria that couple the oxidation of sulfide in deep, anoxic sediment to the reduction of oxygen in shallow sediment. They conduct electricity over several centimetres through as-yet uncharacterized conductive structures in their periplasm. This discovery changed our conception of what living organisms are capable of – we previously thought that oxidative and reductive halves of an organism’s metabolism had to exist in the same cell, cable bacteria demonstrated that it is possible to oxidize an electron donor in one cell and use that electron to reduce an electron acceptor in another cell, possibly centimetres away, by transporting that electron through a conductive biological wire much faster than diffusion could transport an electron carrier molecule. Similarly, it has been shown that microorganisms can transfer electrons centimetre distances via networks of dissolved organic matter and organo-mineral associations ( Bai et al., 2023 ), or across millimetre distances in biofilms ( Li et al., 2016 ). There is also increasing evidence that some anaerobic microorganisms can reverse their metabolism depending on the surrounding chemical environment. For example, acetogens have been shown to oxidize acetate in a H 2 -consuming co-culture ( Hattori et al., 2005 ). Obviously, a single-celled microorganism cannot conserve energy through the simultaneous oxidation and reduction of the same compound – this reaction would not be exergonic. However, if we combine this concept with the long-distance transport of electrons one can imagine two distant, yet electrically connected, bacterial cells in different chemical environments – one environment with a higher concentration of electron donor favouring the forward reaction and the other with a higher concentration of the oxidized form of the substrate favouring the reverse reaction. The metabolism of these microorganisms would be powered by an electrochemical concentration cell. An electrochemical concentration cell is an electrochemical system where a single redox reaction, run both in forward and reverse directions, can result in an electrical current between two different concentration regimes ( Foulkes, 2012 ). Such concentration cells are well known in some fields, such as the contribution of oxygen concentration cells to metal corrosion ( Iverson, 1987 ). Traditionally, concentration cells have not been seen as relevant for individual microorganisms, as a ~1 μm cell is much too small to span multiple concentration regimes. However, the discovery of long-distance electron transport in heterogeneous subsurface environments by cable bacteria, conductive minerals, and biofilms makes this topic newly relevant. Concentration differences could drive microbial metabolism in the subsurface in a way that has not been previously recognized, in what could be called an ElectroMicrobiological Concentration Cell (EMCC). The energy yields would be small, and turnover rates may be small compared to other processes driving the concentration differences, but the energy-starved subsurface has previously shown to be homes for such low-energy-yield metabolisms, like anaerobic methane oxidation ( Knab et al., 2008 ) or acetogenesis ( Lever, 2012 ), and low metabolic rates are also common in subsurface environments ( Jørgensen and Marshall, 2016 ). As long as steep concentration gradients exist, for example at geochemical transition zones or boundaries between sediments or rocks of different chemical compositions, then EMCCs are thermodynamically possible. The minimum energy requirement to support oxidative phosphorylation is thought to be as low as −10 kJ/mol ( Hoehler et al., 2001 ). At least one microbial cell involved in the reaction would have to conserve energy at any given time, and if multiple cells conserved energy simultaneously this would multiply energy needs. There may also be situations where such concentration cells do not support microbial energy conservation, but their activity still may impact geochemical element cycling. The goal of this paper is to develop a theoretical framework for understanding microbial metabolism through concentration cells. I will present three concentration-cell scenarios based on H 2 O/O 2 , SO 4 2− /S 2− , and H + /H 2 .",
"discussion": "3 Results and discussion 3.1 A general description of electromicrobiological concentration cells (EMCC) The EMCC concept, outlined in Figure 1 , can seem counter intuitive at first: a reaction where substrates and products are the same is usually understood to result in no net changes in any chemical concentrations and therefore no possibility for an organism to conserve energy. However, the key here is that the substrates and products are in fact not the “same” as they exist in areas with different substrate and product concentrations. The reaction is driven forward by coupling the two reactions together through electrical conductivity, allowing the concentration difference to drive the reaction and permit energy conservation. This concentration difference would have to be maintained by some other biotic or abiotic process to keep the EMCC running, as the EMCC will diminish this concentration difference over time. This is just like any other chemotrophic microorganism—substrates are used up over time, leading to the cessation of metabolism if substrates are not replenished. Figure 1 Generic overview of the electromicrobiological concentration cell model. The oxidized and reduced compartments may be a consequence of a concentration gradient (as the examples in this paper show) or a physical conductive barrier preventing diffusion from one compartment to the other. Dotted blue lines show that some other abiotic or biotic process will need to maintain the concentration difference for an EMCC to remain energetically favourable over time. Figure 2 presents a generalised model for this process, showing how different variables can affect the ∆G. The assumption is that the electron acceptor varies in concentration over a distance of 1 millimetre, with the lowest concentration being 1 nM and the highest concentration being a number of orders of magnitude above 1 nM. This electron acceptor is reduced by a certain number of electrons to make an electron donor. The calculated Gibbs free energy is for the difference between the concentration at 0 mm and the concentration at the indicated depth. The energy conservation itself does not necessarily occur at the indicated depth, but the indicated ∆G shows the total free energy available for an EMCC extending over that depth. The exact energy yield at each depth would depend on whether the energy was conserved in cells carrying out the oxidation reaction or the reduction reaction. Figure 2 (A) General model for an electromicrobiological concentration cell for a 4-electron reaction with 6 orders of magnitude concentration difference, with a minimum. (B) General models for a range of different electron numbers and concentration differences, with the minimum Gibbs free energy and distance at <−10 kJ/mol (the assumed minimum necessary for ATP production) shown for reactions with different numbers of electrons transferred and different concentration ranges. pH is assumed to be constant. Figure 2 provides several key insights into the EMCC concept. Firstly, order-of-magnitude changes in concentration differences lead to linear changes in Gibbs free energy. Secondly, the number of electrons transferred change the nFED term in the ∆G calculation and thus alter the slope in the linear part of the calculation. Fewer transferred electrons mean less energy loss along the conductor, resulting in a longer distance that a favourable ∆G can be maintained. This difference in energy loss rate reflects the fact that higher currents result in higher potential loss according to Ohm’s law (V = IR)—assuming a constant substrate turnover rate and resistance, redox reactions transferring more electrons will have a proportionally higher current that results in a higher loss in potential. Put together this means that the most energetically favourable EMCCs involve high concentration differences and low numbers of electrons transferred. The model for EMCCs in Figures 1 , 2 is best understood when applied to certain specific scenarios, some where the model may help to explain certain difficult to understand observations. Figures 3 – 5 illustrate models for EMCCs based on O 2 , H 2 S and H 2 gradients, with the concentration of all other chemical species assumed as constant. These gradients transition from maximum concentration to minimum concentration linearly across a given penetration depth. Penetration depths and maximum substrate concentrations were chosen based on observed literature values ( Jørgensen et al., 1979 ; Nielsen et al., 2015 ). Supplementary Figures show ∆G calculated from real-world microprofiles for O 2 , S 2− , and pH – this is a small selection of possible models based on publicly available data, and should not be interpreted as representative of all real-world possibilities. No suitable publicly available H 2 /pH microprofile datasets were found. Choosing a minimum substrate concentration presented an interesting challenge – setting this to 0 μM would make the ∆G impossible to calculate, as the logarithm of zero is infinite. We therefore need to choose an arbitrarily low number for our minimum substrate concentration, but how low this arbitrarily low number is will have a big impact on the resulting ∆G – each order of magnitude change in the minimum concentration could lower the ∆G by up to −5.6 kJ/mol. For each model I have chosen to set the minimum concentration to 1 nM, which is below the 2–20 nM detection limit for the current most sensitive microscale detection methods for O 2 ( Revsbech et al., 2009 ), H 2 S (e.g., https://unisense.com/products/h2s-microsensor/ ), and H 2 ( Nielsen et al., 2015 ) and therefore our best understanding of “zero” at present. Figure 3 Modelled EMCC based on an O 2 concentration gradient, assuming an oxygen penetration depth of 0.5 mm, a maximum O 2 concentration of 300 μM and a minimum concentration of 1 nM. pH is assumed to be constant. Figure 4 Modelled EMCC based on a H 2 S concentration gradient, assuming most sulfide is removed in the upper 7 millimetres, a maximum H 2 S concentration of 2 mM and a minimum concentration of 1 nM. pH is assumed to be constant. Figure 5 Modelled EMCC based on a H 2 gradient, assuming H 2 is produced and concentration diminishes within 1 mm of the production surface. Maximum H 2 concentration is 2.5 μM, minimum H 2 concentration is 1 nM. One interesting aspect of EMCCs is the large impact of pH on the resulting ∆G. Protons are produced where oxidation takes place, and consumed where reduction takes place. Multiple protons lead to exponents on the proton concentration in the ∆G calculation, leading to exponential impacts on the overall calculation: an O 2 /H 2 O EMCC involves the transfer of 4 electrons and thus 4 protons, with [H + ] 4 then appearing in the numerator and denominator of the reaction quotient. The H 2 S/SO 4 2− EMCC involves 8 protons and the H 2 /H + EMCC involves 2 protons. This means that the pH concentration profile of an environment has a large impact on whether an EMCC is possible. A lower pH at the reducing environment than the oxidizing environment will inhibit an EMCC, while a higher pH in the reducing environment will make an EMCC more likely to be exergonic. Figure 2 and the other theoretical models in this study assume a constant pH, but the models based on real-world data ( Supplementary Figures ) take pH differences into account and show that in many cases the reaction could still be thermodynamically favourable even when the reducing end of the EMCC has a lower pH. As an EMCC runs, more protons are produced at the reducing end and consumed at the oxidizing end, making the pH profile less favourable for an EMCC as time goes on, with the rate of this degradation in ∆G a function of how well buffered the system is. In a poorly buffered system, the pH difference will render the reaction unfavourable before the other system components due to the higher exponents. The calculated changes in Gibbs free energy closest to the zero position are a function of whether the chemical species at the zero position is consumed (O 2 and H 2 ) or produced (H 2 S), with consumption resulting in a concave curve and production in a convex curve. The linear part of the curve starts where the concentration becomes constant – here the energy change is a function of the distance and the number of electrons transferred, with more electrons resulting in a steeper slope ( Figures 2 – 5 ). 3.2 Oxygen reduction coupled to water splitting \n 2 H 2 O r e d + 4 H o x + + O 2 , o x → 2 H 2 O o x + 4 H r e d + + O 2 , r e d \n \n Δ G = R T ln ( [ H r e d + ] 4 [ O 2 , r e d ] [ H o x + ] 4 [ O 2 , o x ] ) + n F E D \n The oxic/anoxic interface in aquatic sediments is a possible environment where EMCCs may occur, as shown by the theoretical model used here ( Figure 3 ) and the models based on O 2 and pH microprofiles from freshwater and marine environments ( Supplementary Figures SF1–F6 ). Oxygen from the water column is consumed by aerobic microorganisms, often in the uppermost millimetre of sediments with ample electron donor. This creates a steep O 2 gradient in the sediment, meaning that a conductive structure could connect O 2 -reducing and H 2 O-oxidizing processes on either end of this gradient without losing too much potential. The steepness of the O 2 gradient appears to be critical for whether or not an EMCC will be possible – O 2 penetration depths of about 1–2 mm as found in Lake Baldegg ( Supplementary Figure SF1 ), Lake Greifen ( Supplementary Figure SF2 ), the shallow part of Lake Lucerne ( Supplementary Figure SF3 ), Lake Zug ( Supplementary Figure SF4 ), and Amon Mud Volcano ( Supplementary Figure SF5 ) result in ∆G values below −10 kJ/mol, while O 2 penetration depths of about 1 cm or more found in the two deep cores from Lake Lucerne (SF3) and the Nordic Margin (SF6) result in positive ΔG values. This is because the potential difference lost over the longer O 2 penetration depth is greater than the potential difference created by the concentration difference. The fact that steeper gradients are more likely to produce EMCCs introduces an interesting paradox – O 2 is depleted rapidly by aerobic microorganisms oxidizing organic matter, and more organic matter (such as in the eutrophic lakes Baldegg, Greifen, and Zug) results in a steeper decline in O 2 concentration than environments with less organic matter, like Lake Lucerne. However, higher organic matter concentrations will also mean that an EMCC microorganism is more likely to be outcompeted by an aerobic chemoorganotroph in the oxic sediment layer. The EMCC organism could only become established if there was some factor preventing direct competition for oxygen, such as physically extending further into the oxic zone above the sediment surface. Similar processes have been observed in cable bacteria emerging from the sediment under conditions of oxygen limitation ( Burdorf et al., 2018 ). Cable bacteria form conductive structures from the oxic layer to the anoxic layer, allowing an exergonic production of O 2 from water in the uppermost part of the anoxic layer. Such O 2 production may explain the observation of so-called “flocking” bacteria, microorganisms capable of aerobic respiration that flock around cable bacteria in the anoxic zone of a slide designed to reproduce O 2 gradients as in sediment ( Bjerg et al., 2023 ; Lustermans et al., 2023 ). These flocking bacteria would consume O 2 produced by the cable bacteria, keeping the O 2 concentration in the anoxic sediment low and allowing the O 2 -producing reaction to continue to proceed. In a way this would extend the influence of O 2 several millimetres below the oxic/anoxic interface, even allowing O 2 -dependent reactions such as aerobic methane and ammonium oxidation to proceed. There is currently no direct evidence for such cryptic O 2 production in sediment, only some evidence of O 2 production by cable bacteria filament sheaths in an in vitro concentration cell ( Digel et al., 2023 ), but the model presented here shows that such a process is thermodynamically possible. These thoughts about O 2 production in cable bacteria come amid an increasing wave of interest in trace amounts of O 2 produced in ostensibly anoxic environments and making aerobic respiration possible ( Berg et al., 2022 ; Kraft et al., 2022 ; Ruff et al., 2023 ). While EMCCs cannot explain many of the observations made until now, especially in the water column, the potential for such concentration-driven production of O 2 ought to be considered a possible explanation for some otherwise difficult-to-explain aerobic metabolism. 3.3 Sulfate reduction coupled to sulfide oxidation \n 4 H 2 O r e d + H 2 S r e d + 8 H o x + + SO 4 2 − o x → 4 H 2 O o x + H 2 S o x + 8 H r e d + + SO 4 2 − r e d \n \n Δ G = R T ln ( [ H 2 S o x ] [ H r e d + ] 8 [ SO 4 2 − r e d ] [ H 2 S r e d ] [ H o x + ] 8 [ SO 4 2 − o x ] ) + n F E D \n Sulfate reduction is a strong candidate for EMCCs, as the dissimilatory sulfite reduction (DSR) pathway is already known to be able to run in reverse, oxidizing sulfide or reducing sulfate. It used to be thought that microorganisms using the oxidative DSR pathway were phylogenetically distinct from sulfate-reducing organisms ( Müller et al., 2015 ), but sulfide-oxidizing organisms with a dsrAB genes that belong to sulfate-reducing clades have since been found. These include cable bacteria ( Kjeldsen et al., 2019 ) and Desulfurivibrio alkaliphilus ( Thorup et al., 2017 ). Similarly, many sulfate-reducing bacteria have been shown to oxidize sulfide while reducing O 2 , albeit without observed growth ( Dannenberg et al., 1992 ). While a single organism capable of energy conservation using either sulfate reduction and sulfide oxidation has not yet been observed, it has not been ruled out yet either, especially not using an electrode as electron donor or acceptor rather than a chemical substrate. One could imagine a situation where an organism capable of both sulfide oxidation and sulfate reduction could couple these two processes together electrically from an environment with high sulfide concentration to low sulfide concentration ( Figure 4 ). An examination of sulfide and pH gradients in nature shows that in at least one case such a sulfide-driven EMCC would be exergonic ( Supplementary Figure SF7 ). One interesting consequence of this model would be that abiotic consumption of sulfide could enable an EMCC to use alternative electron donors indirectly. For example, cable bacteria are not known to reduce metal oxides or organic matter, and nor do they have any clear indications in their genome that such electron acceptors could be used ( Kjeldsen et al., 2019 ). However, metal oxide minerals ( Peiffer et al., 1992 ) and organic matter ( Yu et al., 2015 ) can abiotically oxidize sulfide, creating a low sulfide zone around their surfaces. A cable bacteria EMCC could drive its metabolism using this area around a mineral with a lower sulfide concentration against a relatively constant sulfate concentration background, thus indirectly accessing alternative electron acceptors with abiotic sulfide/sulfate cycling as an intermediate. Cable bacteria activity results in iron oxide near the sediment surface ( Seitaj et al., 2015 ), could these iron oxides support cable bacteria survival during annual bottom water hypoxia? A laboratory analogue to such sulfur-mediated use of alternative electron acceptors may already have been observed. There have been several observations of cable bacteria being attracted to and potentially growing on graphite anodes in the place of other electron acceptors ( Reimers et al., 2017 ; Li et al., 2021 ; Bonné et al., 2024 ). Could these graphite electrodes be abiotically oxidizing sulfide released from the sediment, creating a low-sulfide zone around the electrode that can enable the coupling of sulfate reduction close to the electrode to the oxidation of sulfide distant from the electrode? Abiotic sulfide oxidation at graphite electrodes has been observed before ( Ateya and Al-Kharafi, 2002 ). The observed increase in current compared to cable-free controls could be explained by the cable bacteria increasing the effective surface area of the electrode. This explanation would avoid the need for any direct-interspecies electron transfer to the electrodes by cable bacteria. 3.4 Proton reduction coupled to hydrogen consumption \n H 2 , r e d + 2 H o x + → H 2 , o x + 2 H r e d + \n \n Δ G = R T ln ( [ H 2 , o x ] [ H r e d + ] 2 [ H 2 , r e d ] [ H o x + ] 2 ) + n F E D \n Hydrogen metabolism is widespread amongst prokaryotes, with both H 2 oxidation and H 2 production by H + reduction coupled to many other redox reductive processes. Several types of hydrogenases exist to mediate these reactions, and [NiFe]-hydrogenases are known to catalyse both the oxidation of H 2 and the reduction of H + ( Vignais and Billoud, 2007 ). H 2 as an electron donor for microorganisms can be produced geologically ( Nealson et al., 2005 ). One could imagine a subsurface source of H 2 where the H 2 diffuses away, creating a high-H 2 to low-H 2 gradient, ideal for an electromicrobiological concentration cell driven by the H 2 concentration difference, assuming that pH remains relatively constant ( Figure 5 ). The idea of an H 2 EMCC is particularly interesting in the context of the earliest forms of metabolism in the early Earth or other planets. One challenge for the first life forms was the absence of electron acceptors—the early Earth atmosphere was very reduced ( Shaw, 2008 ), and the first life has often been thought to use nitrogen oxides ( Mancinelli and McKay, 1988 ) or carbon dioxide ( Walker, 1985 ). But a H 2 -based concentration cell gives us another possible electron acceptor for early life: H + , coupled to the oxidation of H 2 . In this way early life would not have had to find a way to couple energy conservation to CO 2 fixation all at once (as in the complex methanogenesis pathway), but could have had an alternative energy source from the beginning. One key advantage for early life would be that a single reversible proto-hydrogenase could function both to oxidize the electron donor and reduce the electron acceptor, limiting the complexity for a minimal metabolism to work. Extracellular H 2 oxidation produces a higher H + concentration outside the cell, generating a proton motive force that could have driven the first metabolism with a single catalyst. This is not the first suggestion that concentration gradients could have helped drive metabolism in the earliest life (e.g., Ooka et al., 2019 ), but it is striking how simple the earliest life form could be if concentration cells are taken into account."
} | 6,653 |
37835994 | PMC10575179 | pmc | 2,695 | {
"abstract": "Organosilicon polymers (silicones) are an important part of material chemistry and a well-established commercial product segment with a wide range of applications. Silicones are of enduring interest due to their unique properties and utility. Recently, new application areas for silicone-based materials have emerged, such as stretchable electronics, wearable stress sensors, smart coatings, and soft robotics. For this reason, research interest over the past decade has been directed towards new methods of crosslinking and increasing the mechanical strength of polyorganosiloxanes. The introduction of self-healing mechanisms may be a promising alternative for such high-value materials. This approach has gained both growing research interest and a rapidly expanding range of applications. Inherent extrinsic and intrinsic self-healing methods have been used in the self-healing of silicones and have resulted in significant advances in polymer composites and coatings, including multicomponent systems. In this review, we present a summary of research work dedicated to the synthesis and applications of self-healing hybrid materials containing polysiloxane segments, with a focus on antimicrobial and antifouling coatings.",
"conclusion": "6. Conclusions Recent years have seen significant advances in the field of self-healing silicone elastomers with antimicrobial and antifouling properties. This is due to the emergence of new application areas for silicone-based materials, such as stretchable electronics, wearable stress sensors, smart coatings, and soft robotics. However, such high-value silicones have much broader interest in many other areas, including marine antifouling coatings, anti-icing coatings, microfluidics, medical devices, electrochemical biosensors, and beyond. In this manuscript, we have summarized research work on the synthesis, physical and mechanical properties, and applications of self-healing hybrid materials containing polysiloxane segments. Extrinsic and intrinsic self-healing methods are reviewed, which have recently resulted in significant advances in siloxane-based polymer composites and coatings, including multicomponent systems. This particular branch of silicone elastomer chemistry and materials engineering can be expected to flourish in the near future due to the unique potential that such materials have. Similarly to the trends observed for S-H materials in general, the new generations of self-healing silicones are not based on single mechanisms that lead to the self-repair of the polymer network. The design of their overall characteristics relies on a multi-level combination of several types of dynamic bonds. Antimicrobial properties are highly desirable, and this feature should ideally be incorporated into the self-regeneration mechanism. The first idea for the use of S-H silicones was antibacterial coatings. Since then, this branch of polysiloxane-based elastomers has developed greatly. Antifouling S-H silicone coatings, especially those for marine applications, are bound to self-heal under special conditions. Nowadays, it is not only about antifouling and self-repair coatings; however, S-H mechanisms are also involved in promoting adhesion to various substrates, including metals, which is not trivial. In silicone S-H elastomers, bioactive inorganic or metallic nanoparticles are often used, as well as biocomponents with antimicrobial properties such as tannic acid, coumarin, or eugenol. In the case of silicones for biomedical applications, in addition to antifouling coatings, the focus is on S-H siloxane-based smart hydrogel materials with drug molecules and antimicrobials encapsulated in a pH-sensitive network structure for long-term delivery and monitoring. Most recent, but very important, are advanced silicones designed for modern materials science and prospective directions. The future of S-H antimicrobial silicones will be largely related to applications in soft robotics and flexible wearable electronics (including sensors). Such materials need to meet a range of criteria in which self-healing and antimicrobial properties are components of a multifactorial mechanism that works through a number of properties such as electrical conductivity, luminescence, magnetic properties, and high mechanical strength. The application potential of antibacterial S-H silicone elastomers has not yet been fully explored. The development of this field of material chemistry is challenging and difficult. In the future, such polymers and composites will need to incorporate additional qualities and properties that will be required in widening application areas, including, for example, smart protective coatings, materials with antiviral properties, actuators and artificial muscles, sensors, or smart materials for additive manufacturing.",
"introduction": "1. Introduction Polymers made of siloxane bonds (silicones: polysiloxanes and polysilsesquioxanes) are recognized high-value materials of excellent low-temperature flexibility, high and low-temperature stability, low surface energy, high permeability, electrical resistance, high oxidation stability, and resistance to environmental conditions, as well as biocompatibility, sterilization tolerance, biological durability, and hemocompatibility [ 1 , 2 , 3 , 4 ]. Owing to their intrinsic properties, silicones can be used not only in the electronics or aerospace industry but also in the biomedical field and the food industry, where durability, sterility, and chemical purity of polymeric materials are required. Poly(dimethylsiloxane) (PDMS) is the most well-known example of silicone-based materials, a precursor of covalently cross-linked elastomers that are soft, stretchable, and almost creep-free. The mechanical performance of siloxane polymers can be tuned by adjusting their molecular weight or degree of cross-linking. However, conventional chemical cross-linking does not guarantee durability, and thus the main disadvantage of traditional silicone elastomers is their susceptibility to physical damage. While silicone materials of this type are still in high demand, modern technologies, much more oriented towards sustainability, recyclability, and reusability, tend to focus on the dynamic crosslinking of siloxanes with tailored mechanical properties by means of self-healing (S-H). The introduction of a S-H mechanism into silicone materials may be a promising alternative to covalent network formation, and self-regenerating polysiloxane elastomers that are capable of healing minor physical damage are the focus of recent research [ 5 , 6 , 7 , 8 , 9 ]. The self-repair capability can not only extend the life of silicone coatings but also help meet the specific requirements of emerging new applications such as stretchable electronics, wearable stress sensors, smart coatings, and soft robotics. An important concern for materials intended for such applications, aside from their self-healing effectiveness and excellent mechanical properties, is their resistance to biological factors. Frequent and prolonged contact with skin, moisture, and the external environment makes each type of silicone material susceptible to unwanted microbial deposition and biofilm growth. Any major damage but also any small changes in the continuity of polymer coatings (e.g., minor scratches) can create favorable conditions for biofilm formation and have a huge impact on the rate of surface colonization by microorganisms. Scratches can become specific habitats—protective environmental niches for bacterial growth. In this context, the capability for effective re-mending and, at the same time, discouraging microorganisms from colonizing the surface is a very desirable and advantageous feature. High-value materials of this type are gaining increasing attention. In this review, we have focused on research related to self-healing siloxane materials with antibacterial and antibiofilm properties that were directly designed for or have the potential to be applied in the area of flexible electronics but also for biomedical devices and as antifouling coatings."
} | 2,017 |
38559741 | PMC10976600 | pmc | 2,696 | {
"abstract": "The ordered arrangement of nanoparticles can generate\nunique physicochemical\nproperties, rendering it a pivotal direction in the field of nanotechnology.\nDNA-based chemical encoding has emerged as an unparalleled strategy\nfor orchestrating precise and controlled nanoparticle assemblies.\nNonetheless, it is often time-consuming and has limited assembly\nefficiency. In this study, we developed a strategy for the rapid and\nordered assembly of DNA origami-framed nanoparticles assisted by dynamic\ninterfaces. By assembling Au nanoparticles (AuNPs) onto DNA origami\nwith different sticky ends in various directions, we endowed them\nwith anisotropic specific affinities. After assembling DNA origami-framed\nAuNPs onto supported lipid bilayers with freely diffusing single-stranded\nDNA via DNA hybridization, we found that DNA origami-framed AuNPs\ncould form larger ordered assemblies than those in 3D solution within\nequivalent time frames. Furthermore, we also achieved rapid and ordered\nassembly of liposome nanoparticles by employing the aforementioned\nstrategy. Our work provides a novel avenue for efficient and rapid\nassembly of nanoparticles across two-dimensional interfaces, which\nis expected to promote the application of ordered nanoparticle assemblies\nin sensor and biomimetic system construction.",
"conclusion": "Conclusion In conclusion, this study developed a strategy\nfor the rapid and\nordered assembly of DNA origami-framed nanoparticles assisted by fluidic\nsupported lipid bilayers (SLBs) as dynamic interfaces. By employing\ntriangular DNA origami to encode both gold nanoparticles (AuNPs) and\nliposomes with anisotropic affinities, we successfully realized the\nrapid self-assembly of these nanoparticles on fluidic SLBs. Notably,\nour findings demonstrated that the DNA origami-framed anisotropic\nnanoparticles assembled on SLBs could form 1D arrays exceeding one\nmicron in length within a remarkably short time frame of 2 h, surpassing\nthe assembly outcomes observed in solution environments. By providing\na streamlined method for the rapid preparation of both inorganic and\norganic nanoparticle arrays on two-dimensional interfaces, our work\nopens avenues for advancing the development of nanotechnology-enabled\ndevices and systems.",
"introduction": "Introduction The self-assembly of nanoparticles into\nordered structures is of\nsignificant importance for harnessing the excellent physicochemical\nproperties of nanomaterials. 1 − 3 Although van der Waals forces,\nelectrostatic interactions, hydrogen bonds, molecular dipole interactions,\nand DNA base pairing can drive the ordered arrangement of nanoparticles, 4 , 5 it is challenging to assemble nanoparticles into customized morphologies\nalong controlled pathways. The advancement of structural DNA nanotechnology\nhas facilitated the controllable assembly of nanoparticles. 6 − 11 On one hand, by leveraging the addressability of DNA nanostructures,\nDNA-modified nanoparticles can be assembled onto individual DNA nanostructures\nwith controlled copy numbers and spacing; 12 − 14 on the other\nhand, DNA nanostructures can encode individual nanoparticles, enabling\nthem to acquire anisotropic specific affinity, thereby allowing nanoparticles\nto assemble into arrays along predefined pathways. 15 − 17 However, in\na solution environment, the assembly rate of the DNA nanostructures\nthemselves, as well as DNA nanostructure-encoded nanoparticles, is\nrelatively slow. Stringent annealing conditions and several tens of\nhours are often required to obtain arrays with the desired morphology. 18 In recent years, research on the interaction\nbetween DNA nanostructures\nand solid supported lipid bilayers (SLBs) have inspired innovative\napproaches to fabricating DNA nanoarrays. 19 − 22 DNA nanostructures can bind onto\nSLBs through various pathways. 20 , 21 , 23 One approach involves the electrostatic adsorption of DNA nanostructures\nmediated by divalent metal ions to zwitterionic bilayers. 20 Another approach entails the hybridization of\nanchoring DNA chains on DNA nanostructures with the single-stranded\nDNA (ssDNA) on the surface of SLBs. 21 Additionally,\nDNA nanostructures can also be embedded into phospholipid bilayers\nthrough lipid molecules carried by themselves, such as cholesterol. 23 Biological membranes can confine membrane-bound\nproteins in 2D space, thereby increasing the probability of their\ninteractions. Therefore, DNA nanostructures bound to SLBs should also\nbe able to self-assemble into higher-order structures more efficiently\ncompared to in a 3D solution environment, as evidenced by multiple\nstudies. 20 , 24 For example, cross-shaped DNA origami units\nwith dimensions of 64 nm in length and width can form micrometer-scale\n2D arrays on SLBs composed of zwitterionic 1,2-dioleoyl- sn -glycero-3-phosphocholine (DOPC) within minutes in the presence of\nMg 2+ . 20 Therefore, in\nthis study, we proposed the use of SLBs as a dynamic\ninterface to assist the rapid self-assembly of DNA origami-framed\nanisotropic nanoparticles. We first employed triangular DNA origami\nto encode gold nanoparticles (AuNPs) and liposomes with anisotropic\naffinities and then assembled the triangular origami-framed anisotropic\nAuNPs onto fluidic SLBs through DNA hybridization. We found that the\ntriangular DNA origami-framed anisotropic AuNPs assembled on SLBs\ncould undergo free two-dimensional diffusion and form 1D arrays with\nlengths exceeding one micron within 2 h. In comparison, in a solution\nenvironment, assemblies formed by only 2–5 triangular DNA origami-framed\nAuNPs were observed within the same time frame. Furthermore, we also\nrealized the rapid self-assembly of DNA origami-framed liposomes on\nSLBs with the strategy described above. Therefore, our work provided\na novel strategy for the rapid preparation of inorganic and organic\nnanoparticle arrays on two-dimensional interfaces, holding significant\npotential for applications in sensor and biomimetic system construction.",
"discussion": "Results and Discussion Before attempting to promote\nthe self-assembly of DNA origami-framed\nanisotropic nanoparticles using SLBs, we planned to first compare\nthe differences in the self-assembly efficiency of DNA origamis themselves\non SLBs and in solution environments. As shown in Figures 1 and S1 , we designed two types of DNA origami units (units 1 and 2) based\non previously reported triangular DNA origami. Each unit had five\nprotruding arm chains on both the A and B sides (U1A1–U1A5,\nU1B1–U1B5, U2A1–U2A5, U2B1–U2B5) for the assembly\nof Unit 1 and Unit 2. Of note, arm chains U1A1, U1A2, U1A3, U1A4,\nand U1A5 on Unit 1 could specifically hybridize with arm chains U2A5,\nU2A4, U2A3, U2A2, and U2A1 on Unit 2, respectively; while arm chains\nU1B1, U1B2, U1B3, U1B4, and U1B5 could specifically hybridize with\narm chains U2B5, U2B4, U2A3, U2A2, and U2A1 on Unit 2, respectively;\nthereby mediating the self-assembly of Unit 1 and Unit 2 into one-dimensional\narrays. Additionally, to assemble Units 1 and 2 onto SLBs and enable\ntheir free diffusion, we designed three anchoring chains on each unit\nwhich could hybridize with the cholesterol-modified ssDNA embedded\nin the SLBs. The reason we designed three anchoring chains was that\naccording to our previous studies, more anchoring chains may reduce\nthe diffusion rate of DNA origami on SLBs, potentially affecting the\nefficiency of self-assembly. 25 Figure 1 Schematic showing\nthe design of triangular DNA origami Units 1\nand 2 (a) and 2D dynamic interface (SLBs)-assisted rapid self-assembly\nof Units 1 and 2. Chol-ssDNA: cholesterol-modified ssDNA. Then, we planned to use atomic force microscopy\n(AFM) imaging to\ndemonstrate whether Unit 1 and Unit 2 exhibit higher self-assembly\nefficiency on the SLBs. For this purpose, we first prepared liposomes\ncontaining DOPC and 1,2-dipalmitoyl- sn -glycero-3-phosphocholine\n(DPPC) (molar ratio of DOPC:DPPC = 70:30), and we further prepared\nSLBs by incubating liposomes with clean mica surfaces. The choice\nof liposomes containing DOPC and DPPC for SLBs preparation was because\nthe difference in phase transition temperatures between the two phospholipids\ncould lead to phase separation. Additionally, due to the difference\nin molecular structures, the height of the regions where DPPC aggregated\nwould be higher, which aided our in determining the formation of SLBs\nvia AFM imaging ( Figure S2 ). Upon the formation\nof SLBs, incubation of cholesterol-modified ssDNA with the SLBs resulted\nin the formation of freely diffusing DNA monolayers on the SLBs, as\ndemonstrated in our previous study. 25 Subsequently,\nthrough AFM imaging, we demonstrated that Unit 1 and Unit 2 could\nbe assembled onto the SLBs through DNA hybridization and form linear\nassembles comprising more than ten DNA origami units within 2 h ( Figures 2 a,b,d and S3 ). In contrast, in a solution environment, incubation\nof both Unit 1 and Unit 2 in a 1:1 molar ratio for 2 h only resulted\nin the formation of assemblies mostly consisting of up to five monomers\n( Figure 2 c,d). Figure 2 (a,b) Atomic\nforce microscopy imaging results of the self-assembly\nof Units 1 and 2 on the SLBs (a) and representative height maps of\n1D DNA origami arrays (b). (c) Atomic force microscopy imaging results\nof the self-assembly of Units 1 and 2 in a solution environment. (d)\nHistogram of the frequency distribution of the number of DNA origami\nunits in self-assembled structures under different conditions. (e,f)\nTIRFM imaging results of the self-assembly of Units 1 and 2 on the\nSLBs (e) and in solution (f). Scale bar is 200 nm in parts a and c;\nscale bar is 3 μm in parts e and f. Furthermore, we labeled Unit 1 and Unit 2 with\nfluorescent molecules\n(Cy3) and observed their self-assembly on SLBs by using total internal\nreflection fluorescence microscopy (TIRFM) ( Figure\nS1 ). In this part, SLBs were prepared on the glass bottom of\nclean cell culture dishes. Similarly, fluorescently labeled DNA origami\ncould be specifically assembled onto SLBs via DNA hybridization ( Figure S4 ). Moreover, when only one type of Unit\nwas assembled on the SLBs or when both Units were assembled but lacked\narm chains, we observed only discrete spots of fluorescence ( Figure S5 ). Furthermore, when both Unit 1 and\nUnit 2 were coincubated with SLBs, numerous linear spots of fluorescence\nwith lengths exceeding one micron were observed after approximately\none h ( Figure 2 e).\nHowever, when Unit 1 and Unit 2 were first incubated in solution for\n1 h and then transferred to SLBs for immediate observation, no significant\nlinear spots of fluorescence were observed ( Figure 2 f). These results indicated that indeed DNA\norigami could self-assemble more rapidly on SLBs. Next, we attempted\nto utilize SLBs to promote the assembly of DNA\norigami-framed AuNPs. As shown in Figure 3 a, for real-time observation of the assembly\nprocess using TIRFM, we prepared two types of ssDNA-modified AuNPs,\none labeled with Cy3 and the other labeled with carboxyfluorescein\n(FAM). Cy3-labeled AuNPs and FAM-labeled AuNPs could be respectively\nassembled onto Units 1 and 2 through DNA hybridization ( Figure S6 ), thus enabling us to observe whether\nthe two types of AuNPs assembled together to form 1D arrays on SLBs\nthrough fluorescence colocalization. Prior to this, we demonstrated\nthat both 15 and 30 nm AuNPs could be efficiently assembled onto\ntriangular DNA origami through AFM imaging ( Figure\nS7 ). Additionally, AFM imaging results also confirmed that\nUnit 1- and Unit 2-framed AuNPs could form linear assemblies in a\nsolution environment, but with low efficiency. Only assemblies consisting\nof 2–5 units were formed within 24 h ( Figures\nS8 and S9 ). Figure 3 (a) Schematic showing the design of DNA origami-framed\nAuNPs, and\nrepresentative AFM images showing the self-assembly of Unit 1- and\nUnit 2-framed AuNPs. (b,c) TIRFM imaging demonstrating the self-assembly\nof Unit 1- and Unit 2-framed 15 nm AuNPs (b) and 30 nm AuNPs (c) on\nSLBs at different time points. Scale bars in b and c are 2 and 1 μm,\nrespectively. Subsequently, through TIRFM imaging, we found that\nUnit 1- and\nUnit 2-framed AuNPs could also be specifically assembled onto the\nSLBs through DNA hybridization ( Figure S10 ). Moreover, the DNA origami-framed AuNPs could freely move on the\nSLBs ( SI Video ). After simultaneously assembling\nUnit 1- and Unit 2-framed 15 nm AuNPs onto the SLBs, we conducted\na real-time observation of the assembly process. It was found that\nboth types of AuNPs were dispersed on the SLBs at 0 min. However,\nafter 120 min, the spots of both types of AuNPs aggregated into linear\nshapes, and the fluorescence from FAM and Cy3 could be colocalized\n( Figure 3 b). This was\nbecause after self-assembly of Unit 1- and Unit 2-framed AuNPs into\nlinear arrays, their diffusion rate on the SLBs became very slow.\nAs a control, if only Unit 1- or Unit 2-framed AuNPs were assembled\non the SLBs, the fluorescent spots of AuNPs remained dispersed ( Figure S11 ). Additionally, we found that Unit\n1- or Unit 2-framed 30 nm AuNPs could also efficiently assemble into\nlinear arrays on the SLBs ( Figure 3 c). Next, we planned to adopt the strategy described\nabove in the\nself-assembly of DNA origami-framed lipid nanoparticles. To achieve\nthis, we first prepared liposomes with a diameter of 50 nm labeled\nwith 1,2-dimyristoyl- sn -glycero-3-phosphoethanolamine- N -(7-nitro-2-1,3-benzoxadiazol-4-yl) (NBD) (excited by 488\nnm laser), and then modified liposomes with cholesterol-labeled ssDNA\n( Figure S12 ). To assemble the liposomes\nonto Units 1 and 2, we designed 12 arm DNA chains on the inner side\nof Units 1 and 2 ( Figure S13 ), which were\ncomplementary to the ssDNA on the liposomes. Additionally, we decorated\neach Unit 1 and Unit 2 with one Cy3 for observing the assembly of\nUnit 1- and Unit 2-framed liposomes using TIRFM ( Figure 4 a). After simultaneously assembling\nUnit 1- and Unit 2-framed liposomes onto the SLBs, we found that after\n60 min, Unit 1- and Unit 2-framed liposomes could assemble into small-sized\nlinear arrays, as indicated by colocalized linear fluorescence spots\nof Cy3 and NBD. By 180 min, Unit 1- and Unit 2-framed liposomes assembled\ninto larger-sized linear arrays ( Figure 4 b). These results indicated that SLBs could\nalso promote the rapid self-assembly of DNA origami-framed lipid nanoparticles. Figure 4 (a) Schematic\nshowing the design of DNA origami-framed liposomes\nand their self-assembly. (b) TIRFM imaging demonstrating the self-assembly\nof Unit 1- and Unit 2-framed liposome on SLB at different time point.\nScale bar: 2 μm."
} | 3,635 |
40229524 | PMC11996954 | pmc | 2,697 | {
"abstract": "Rhizosphere bacteria work in synergy with mycorrhizal fungi to promote plant growth. The community structure of rhizosphere bacteria may be influenced by continuous changes in fungal associations with host plants. Asiatic herbaceous plant Pyrola japonica (Ericaceae) forms arbutoid mycorrhizas without fungal mantles, with its mycorrhizal development being visually distinguishable at the cellular level. This study aimed to investigate roles of rhizosphere bacteria and their community shifts along with mycorrhizal developments. We examined bacterial communities at three different developmental stages of mycorrhizal roots—limited, full, and digested—via a partial 16S rRNA amplicon sequencing. Both α- and β-diversities in the full condition were significantly lower than those in the limited and digested conditions. Significant clusters of bacterial compositions were found among all treatments. In terms of ecological processes of community assembly, communities in limited conditions and bulk soil were influenced by both deterministic and stochastic processes, whereas those in full and digested conditions were regulated only by stochastic ways. Furthermore, the order Rhizobiales and Actinomycetales known as mycorrhizal helper bacteria were characterized in the full and digested conditions through phylogenetic analysis and detection of indicator taxa. These results suggest that mycorrhizal fungi may play ecologically important roles not only as temporal drivers initiating the formation rhizosphere bacterial communities but also as key founders exerting continuous influences to establish priority effects. Moreover, the rhizosphere bacterial community remains after mycorrhizal degeneration and their historical continuity may contribute to maintaining plant-mycorrhizal fungi-bacterial associations. Supplementary Information The online version contains supplementary material available at 10.1007/s00248-025-02526-z.",
"conclusion": "Conclusion The present study is the first to clarify the assemblage pattern of the rhizosphere bacterial communities of arbutoid mycorrhizal P. japonica , along with different mycorrhizal developments from the limited, full, and digested conditions of the associated fungi. As results, mycorrhizal developmental levels affected irreversibly of rhizosphere bacterial communities. Under the full condition, where mycorrhizal associations are likely to be active, both α- and β-diversity of the bacterial community were the lowest compared to other developments, and the community was historically maintained after the mycorrhizal fungal degeneration in the digested condition. Furthermore, some bacterial taxa known as mycorrhizal helper bacteria were characterized from the communities in the full and digested conditions. This study highlights the dynamic change of rhizosphere bacterial communities synchronized with the continuum of mycorrhizal developmental conditions within a single root system.",
"introduction": "Introduction The rhizosphere is a unique biological and chemical soil environment where some soil microbial communities attracted by the carbon resources as root exudates and rhizodeposition [ 1 , 2 ]. In the rhizosphere, microbial consortia comprising fungi, bacteria, and archaea can form complex biological interactions that affect plant growth [ 2 ]. Among rhizosphere microbes, mycorrhizal fungi are directly associated with the roots of ca . 80% of terrestrial plants [ 3 ]. Fungi receive plant-derived photosynthetic products and provide their hosts with soil nutrients and water [ 4 , 5 ]. Other rhizosphere microbes, known as plant growth-promoting rhizobacteria, positively affect plant growth by enhancing nutrient uptake and improving disease resistance [ 6 ]. Some of these bacteria are called mycorrhizal helper bacteria (MHB) because they contribute to the maintenance and function of mycorrhizal associations [ 7 , 8 ]. To date, many MHB have been reported to function in promoting mycorrhizal fungal activities, improving soil nutrient acquisitions, and inhibiting pathogenic infections [ 7 – 9 ]. These reports focused mainly on two major arbuscular mycorrhizal and ectomycorrhizal (ECM) plants, which are found in diverse plant species that are geographically widespread on Earth [ 3 , 8 ]. Positive effects on plants due to the cooperation between ECM fungi and MHB have been reported in inoculation experiments [ 10 ]. Moreover, rhizosphere bacterial communities, even as small as single ECM root tips, vary depending on the presence or absence of mycorrhizal fungal colonization [ 11 , 12 ]. These findings suggest that cooperative bacteria in plants may be selectively screened by mycorrhizal fungal metabolic products [ 1 ]. However, the subsequent shifts of the bacterial community shaped by mycorrhizal fungi have not been investigated, despite their importance for the maintenance of mycorrhization, nutrient acquisition, and against pathogenic microbes. The rhizosphere bacterial community may shift as a result of mycorrhizal fungal degeneration caused by root turnover. An ecological importance of historical impacts in community assembly was conceptualized [ 13 ], but a recent review noted the lack of standardized study frameworks in ECM-bacterial interactions [ 14 ]. One potential reason is that monitoring the colonization progress of ECM roots is hardly applicable visually owing to the formation of fungal mantles. Monitoring of mycorrhizal fungi-bacterial relationships along with root growth is essential to reveal the tripartite mechanistic connections to maintain plant growth, mycorrhizal formation, and bacterial habitat niches. Arbutoid mycorrhiza is a type of mycorrhizal structures where fungi form hyphal coils inside epidermal root cells [ 15 ]. Pyrola japonica (Ericaceae) is categorized as an arbutoid mycorrhiza, lacking fungal mantles. Transparent epidermal cells allows visual assessment of the extent of mycorrhizal development [ 15 , 16 ]. Their mycorrhizas formed on the plant are presumably attributable to Russulaceae fungi that are known as well-known ECM fungi [ 16 – 22 ]. Moreover, the mycorrhizal developmental progress varies with the position of the root system from peripheral to proximal positions. For the nutrient acquisition strategy, P. japonica is known to be a mixotrophic plant that obtains carbon sources partially from their associated mycorrhizal fungi other than autotrophic photosynthesis [ 16 , 17 , 23 ]. Because P. japonica has various sets of cells with different levels of fungal colonization, the plant is an ideal model for studying the assemblage patterns of rhizosphere bacterial communities along with mycorrhizal development. This study aimed to (1) characterize bacterial taxa associated with different conditions of mycorrhizal development and (2) determine whether the rhizosphere bacterial community structure changes with the development of P. japonica. To achieve these, rhizosphere bacterial communities were analyzed at different stages of mycorrhizal development of P. japonica roots that were visually distinguished using amplicon sequencing of the MiSeq system.",
"discussion": "Discussion This study demonstrated progressive changes in the rhizosphere bacterial community structures of different mycorrhizal developments in P. japonica root systems. To the best of our knowledge, this is the first study to clarify the bacterial community structure associated with arbutoid mycorrhizas. Moreover, progressive changes in the community along with mycorrhizal development in situ have not been reported for any mycorrhizal associations. Here, we discuss the tripartite interactions among plants, mycorrhizal fungi, and rhizosphere bacteria. We also highlighted the potential of a novel nutrient acquisition pathway via bacteria, which has not been well considered in previous mixotrophic studies. Diversity of Bacterial Community Changes Along with Mycorrhizal Development We demonstrated that rhizosphere bacterial diversity varied among different levels of mycorrhizal development. The number of ASVs and UniFrac distances in the full condition were significantly lower than those in the other conditions and bulk soil. In full conditions, mycorrhizal fungal hyphae are filled within epidermal cells, indicating active nutrient movement considering the mixotrophic nature of P. japonica [ 16 , 42 ]. In contrast, other mycorrhizal developments, i.e. limited and digested, are unlikely to function because mycorrhizal fungi have not yet colonized under limited conditions, or hyphal coils have disintegrated in the digested condition [ 16 ]. Because the predominance of Russulaceae ECM fungi was found in the mycorrhizal fungal communities of P. japonica [ 16 – 19 , 22 ], they may be abundantly colonized under full conditions. If this is the case, extramatrical mycelia can extend from epidermal cells into the soil, forming hyphal networks for carbon acquisition [ 17 ]. Russulaceae is a common ECM fungal taxon that is ubiquitously distributed in mature forests [ 20 , 21 ], and mycorrhizal associations in this family are closely related to MHB belonging to Streptomycetaceae, Bacillaceae, Rhizobiaceae, and Burkholderiaceae [ 8 , 43 ]. Previous MHB studies on woody ECM plants have indicated that the presence or absence of fungal mantles can have a significant impact on rhizosphere bacterial communities [ 11 , 12 ]. For P. japonica mycorrhizas, no fungal mantles but fungal coils within epidermal root cells were formed, and growth of extramatrical mycelium was observed [ 16 , 17 ]. Thus, rhizosphere bacteria in P. japonica mycorrhizal associations are likely to be involved in nutrient exchange between plants and mycorrhizal fungi. Lower α- and β-diversities in the full condition may be due to convergence into certain bacterial taxa in the associations. In contrast, higher α- and β-diversities in the limited and digested condition may be an admix effect of both opportunistic associated and MHB-like selected bacteria. In the case of the limited condition, the root tip of P. japonica would have encountered soil bacteria by chance due to root growth. The intermediate level of β-diversity in the digested condition can be ascribed to the recruitment of new bacterial taxa as aging roots of full condition. Detached or decayed root tissues can provide the rhizosphere with root rhizodeposition, attracting neighboring saprophytic bacteria [ 44 ]. Therefore, the full condition may have bacterial community assemblies due to mycorrhizal colonization compared with the limited and digested conditions. The Assembly Pattern of Bacterial Communities Along with Progressive Mycorrhizal Developments Contrasting assembly patterns of bacterial community structures were found under limited and full conditions, whereas community compositions were consistent regardless of mycorrhizal development stages. The number of unique ASVs was significantly different, being highest in the limited condition and lowest in the full condition (Fig. 3 ). Physically, continuum habitat transitions from bulk soil to the rhizosphere allow for the sharing of common bacteria [ 2 ]. This was evidenced by 11.3% of all ASVs (88/776 ASVs) being shared among all treatments (Fig. 3 ). Thus, the unique ASVs in each condition were considered to be the remaining shared ASVs and minor members of the community with a low detection probability. The high number of unique ASVs under limited conditions suggests non-selective bacterial gathering owing to opportunistic associations. In contrast, under full conditions, a low number of unique ASVs may indicate the need for elaborate screening systems for bacterial requests during the development of mycorrhizal associations. The clustering pattern of the bacterial communities, as in the NMDS plot, suggested significant discrimination among the treatments (Fig. 4 a). However, community compositions among mycorrhizal developments were highly similar, based on pairwise treatment comparisons (bulk soil vs. limited condition, p < 0.01; bulk soil vs. full condition, p < 0.01; bulk soil vs. digested condition, p < 0.01; Table S3 ). Thus, clustering was significantly different between the bulk soil and pooled data for all mycorrhizal developments. In this respect, community composition was stable regardless of mycorrhizal development. A nested analysis, incidentally, slightly suggested the community structure under the full condition is likely a subset of other conditions based on. However, owing to possible stochastic PCR and sequencing biases in the NGS short-reading approach, no statistical significance was found (NODF = 8; data not shown). The bacterial community composition remained highly similar irrespective of mycorrhizal development, but bacterial diversity decreased only under full conditions. This phenomenon can be explained as follows: First, the growing root tip encounters soil bacteria in a non-selective manner, leading to diversification under limited conditions. Second, under full conditions, a bacterial community that was not affected by mycorrhizal fungi was maintained. Finally, in the digested condition, certain members of the bacterial community from the previous full condition were almost completely retained, and additional bacteria that fed on rhizodepositions were recruited from the outside rhizosphere. The explanation was also supported by the quantification of ecological processes in community assembly that showed different patterns between the limited condition and post-full conditions (Fig. 5 ). Notably, the bacteria communities in the limited condition and bulk soil were partially regulated by deterministic processes, although the communities in the full and digested conditions were shaped only by stochastic ways (Fig. 5 ). A similarity of the processes between the limited condition and bulk soil is additional evidence that the bacterial community in the limited condition is likely to be established contingently via surrounding soil bacteria (Figs. 2 , 3 , and 5 ). Only the community in the limited condition was partially regulated by homogeneous selection, where environment is uniform and community variation is less obvious (Fig. 5 ). This result indicates that certain bacteria groups may be selected in the rhizosphere during the formation of root tips (Fig. 5 ). Additionally, the homogeneous selection process was not predicted in subsequent mycorrhizal developmental stages, indicating that the phylogenetic selection of rhizosphere bacteria may only occur at an early stage of root growth (Fig. 5 ). Although both deterministic and stochastic ecological processes involve the regulation of bacterial communities with varying degree of contributions, the communities are more likely to be governed by the former process under extreme environmental situations (e.g., high pH) [ 45 ]. The rhizosphere environment in the limited condition and bulk soil may be less stable than those in the full and digested conditions owing to lacks of mitigation by the existence of plants or mycorrhizal fungi [ 1 , 2 ]. Furthermore, dispersal limitation accounted for ca. 60% in the community assembly even after mycorrhizal fungi have degenerated (Fig. 5 ), inferring that bacterial communities were maintained largely unchanged in the rhizosphere. High similarities in bacterial communities among mycorrhizal developments within the same individuals—plants 1, 5, and 8—also supported the importance of historical contingency in rhizosphere community assembly (Fig. 4 b). The historical impacts were conceptualized as “priority effect” in community ecology that occurs when the order and timing of species arrivals influence community assembly [ 13 , 46 ]. In the case of P. japonica , roots remain functioning as conduits even after the mycorrhizal degeneration and are connected to newly formed mycorrhizal parts. Therefore, the historical continuity of rhizosphere bacterial communities may be a reasonable strategy for persistence in both nutrient acquisition and pathogenetic resistance. Surprisingly, the presence of mycorrhizal fungi did not determine bacterial communities as a biotic factor (Fig. 5 ); if mycorrhizal fungi attracted certain bacterial groups, homogeneous selection would have contributed to the community in the full and digested conditions. This result suggests that mycorrhizal fungi may function as an antagonist to pathogenic bacteria, rather than selecting bacteria to form a preferred bacterial community. Thus, some opportunistic bacteria can remain in the rhizosphere irrespective of mycorrhizal developments. This may result in diffused phylogenies and highly similar communities of bacteria in the full condition. These results indicate ecological importance that mycorrhizal fungi work not only as temporal drivers initiating the formation rhizosphere bacterial community but also as key founders exerting continuous influences to establish priority effects. Potential Functions Characterized in Rhizosphere Bacterial Communities After Mycorrhizal Formation In the phylogenetic trees of the top three taxa, four clades were composed of ASVs derived only from the full condition with reference sequences, such as Streptomyces spp. (ASV651 and ASV654, Fig. S3 ), uncultured Alphaproteobacteria bacterium (ASV378, Fig. S4 ), Rhizobium spp. (ASV1553, Fig. S5 ), and Mesorhizobium spp. (ASV295, Fig. S5 ). From our observations, epidermal cells in the full condition were harbored by healthy hyphal coils (Fig. 1 b), indicating that the ASVs in the clades are likely to function in some roles with P. japonica and its associated mycorrhizal fungi. The reference sequences in three of the four clades were retained as MHB isolates, such as actinomycetes and rhizobia [ 8 , 47 ]. Some mycorrhizal fungi in the genus Lactarius (Russulaceae) are symbionts of P. japonica and show co-occurrence relationships with Streptomycetaceae belonging to Actinomycetales [ 17 , 22 , 43 ]; therefore, it is not unexpected to list them as symbiotic candidates. For example, the candidate MHB Streptomyces sp. AcH 505, isolated from the ECM roots of Amanita muscaria , inhibits the growth of other fungi [ 48 , 49 ]. Other ASVs were clustered with either Rhizobium spp. or Mesorhizobium spp., which belong to the Rhizobiaceae, known as nitrogen-fixing bacteria [ 50 , 51 ]. However, they are not necessarily obligate symbionts and sometimes occur not only as free-living in the soil but also as MHB [ 52 ]. Considering the tripartite interactions among host plants, ECM fungi, and nitrogen-fixing bacteria [ 53 ], the two ASVs (ASV295 and ASV1553) detected in this study may have contributed to nitrogen acquisition by P. japonica . Furthermore, the detection of indicator taxa revealed the potential contribution of bacterial community to host plant and mycorrhizal fungi (Table S3 ). Notably, four indicator ASVs selected from mycorrhizal developments belonged to Rhizobiales (ASV032 from the full condition; ASV012, ASV034, and ASV003 from the digested condition; Table S3 ). Among them, two ASVs were clustered with reference sequences derived from ECM mycorrhizosphere or nodules with bootstrap values ≥ 99% (Fig. S5 ). Moreover, these ASVs were detected in all mycorrhizal conditions (Fig. S5 ), suggesting their continuous association with P. japonica and mycorrhizal fungi. Combined with our understanding of the community analysis, rhizosphere bacterial communities associated with the plant may have been historically involved in nitrogen cycling. Considering the mixotrophic nature of P. japonica , the nutrient acquisition process may function complementarily with rhizosphere microbes, with mycorrhizal fungi obtaining carbon and MHB to obtain nitrogen. Furthermore, MHB may indirectly contribute to carbon acquisition by assisting in mycorrhizal formation. Rhizosphere bacteria are potential new pathways for nutrient acquisition in mycoheterotrophic plants [ 54 , 55 ]. The high nitrogen content found in mycoheterotrophic and mixotrophic plants has long been of interest [ 56 ]. Further understanding of mycoheterotrophic lifestyles would require clarification of the contribution of MHB. However, we acknowledge that the DNA approach with short-read sequencing provides a lower resolution for bacterial phylogenetic diversity. Functional evaluations using isolated bacterial strains in vitro are essential to examine the interactions between P. japonica and MHB. As a caveat in the study, we focused on the mycorrhizal development as a factor regulating bacterial communities, while we did not fully consider plant-derived influences. The results of this study supported that bacterial shifts were caused by the mycorrhizal development but may be due to metabolic changes in root organs themselves along with plant growth. Since the only visual assessment of mycorrhizal developments was conducted in this study (Fig. S1 ), metabolic-functional aspects of how rhizosphere bacterial communities are regulated by plants and mycorrhizal fungi along with mycorrhizal development were remained to be unanswered."
} | 5,288 |
22125358 | null | s2 | 2,699 | {
"abstract": "This article reviews the physical and chemical constraints of environments on biofilm formation. We provide a perspective on how materials science and engineering can address fundamental questions and unmet technological challenges in this area of microbiology, such as biofilm prevention. Specifically, we discuss three factors that impact the development and organization of bacterial communities. (1) Physical properties of surfaces regulate cell attachment and physiology and affect early stages of biofilm formation. (2) Chemical properties influence the adhesion of cells to surfaces and their development into biofilms and communities. (3) Chemical communication between cells attenuates growth and influences the organization of communities. Mechanisms of spatial and temporal confinement control the dimensions of communities and the diffusion path length for chemical communication between biofilms, which, in turn, influences biofilm phenotypes. Armed with a detailed understanding of biofilm formation, researchers are applying the tools and techniques of materials science and engineering to revolutionize the study and control of bacterial communities growing at interfaces."
} | 297 |
31732577 | PMC6974649 | pmc | 2,700 | {
"abstract": "Lignin is the most abundant aromatic polymer on Earth and a resource that could eventually substitute for fossil fuels as a source of aromatic compounds for industrial and biotechnological applications. Engineering microorganisms for the production of aromatic-based biochemicals requires detailed knowledge of the metabolic pathways for the degradation of aromatics that are present in lignin. Our isolation and analysis of a Rhodopseudomonas palustris strain capable of syringic acid degradation reveal a previously unknown metabolic route for aromatic degradation in R. palustris . This study highlights several key features of this pathway and sets the stage for a more complete understanding of the microbial metabolic repertoire required to metabolize aromatic compounds from lignin and other renewable sources.",
"conclusion": "Concluding remarks. meta -Methoxylated aromatics are present at significant levels in the lignin of different plants and are potential sources of compounds for industrial applications. In this work, we isolated a strain of R. palustris that acquired the ability to use syringic acid as a growth substrate under photoheterotrophic conditions. Our strategy of incrementally exposing cultures to higher concentrations of syringic acid, while at the same time reducing the availability of the known growth substrates benzoic acid and 4-HBA, has been shown to be conducive to adaptation and the acquisition of new metabolic activities in R. palustris ( 38 , 39 ) and other bacteria ( 40 ). Our analysis of this adapted strain, SA008.1.07, has provided important new knowledge on the bacterial metabolism of syringic acid. First, we found that syringic acid degradation does not occur through or induce the expression of the genes in the well-characterized BAD pathway. This finding makes syringic acid an aromatic compound whose photoheterotrophic metabolism does not utilize the BAD pathway in R. palustris . In addition, the increased abundance of vanARB transcripts in SA008.1.07 cultures grown in the presence of syringic acid and the requirement of vanAB for the growth of this adapted strain on this methylated aromatic provide evidence for a heretofore unknown role of the VanAB enzyme in the anaerobic metabolism of this compound. Since the previously reported function of vanAB is in the aerobic demethylation of vanillic acid ( 31 , 32 ), our observations suggest that the VanAB enzyme may have an additional unrealized function under anaerobic conditions. Known homologues of VanAB are reported to contain an oxygen-sensitive iron sulfur cluster ( 32 ), so our findings reinforce the suggestion that additional experiments are needed to test the role of this enzyme in the anaerobic metabolism of syringic acid. Our analysis of syringic acid metabolism by R. palustris SA008.1.07 sets the stage for further studies of the metabolism of this and other aromatics by this and other bacteria and for the evaluation of previously unexplored functions of the VanAB enzyme. Elucidating such novel pathways and metabolic functions could expand our ability to use microbial transformations of lignin and other renewable resources as biomass-based sources of compounds with potential uses in the energy, chemical, pharmaceutical, and other industries.",
"introduction": "INTRODUCTION As one of the major biopolymers present in plant tissues, lignin has the potential to serve as a renewable source of carbon for the biomass-based production of compounds that are currently derived from petroleum. Unfortunately, the ability to derive chemicals of commercial, chemical, or medicinal value from lignin is limited by a lack of the information needed to improve the biological conversion of the aromatics in lignin into valuable products. We are interested in improving our understanding of how bacteria metabolize the aromatic building blocks in lignin and using this information to develop strategies that allow the conversion of this major component of plant cell walls into valuable products. Syringic acid and other meta -methoxy-substituted phenolic compounds are plant-derived aromatics that present both a hindrance and a potential source of value to the chemical, fuel, and biotechnology industries ( 1 – 3 ). Originating from the guaiacyl (coniferyl alcohol) and syringyl (sinapyl alcohol) phenylpropanoids that are polymerized into lignin during secondary cell wall formation ( 1 ), meta -methoxylated aromatics are frequently present in products generated from deconstructed biomass ( 4 ). While present at low concentrations in sugar-rich lignocellulosic hydrolysates, these methoxylated aromatics can nonetheless induce stress responses ( 2 , 5 ) and cause toxicity ( 6 , 7 ) in non-aromatic-degrading microbes, leading to decreases in both microbial growth and biofuel yield during fermentation ( 8 , 9 ). Further, these phenolics are present at much higher concentrations in solubilized lignin streams produced with emerging technologies ( 10 – 13 ). Incorporation of meta -methoxylated aromatics into the metabolism of an appropriate, genetically tractable microorganism could provide a promising and efficient route for monolignol valorization through the identification and optimization of the biochemical pathways involved. To expand the ability of microbes to metabolize syringic acid and related plant-derived aromatic compounds, we are studying Rhodopseudomonas palustris , a metabolically versatile, well-characterized, and genetically tractable purple nonsulfur alphaproteobacterium ( 14 – 16 ) that has a proven and well-understood ability to utilize aromatic monomers ( 17 , 18 ). Under anaerobic conditions, R. palustris uses the benzoyl coenzyme A (CoA) degradation (BAD) pathway to cleave the aromatic ring of monoaromatic compounds after activation of the molecule via coenzyme A ligation ( 19 ). The diversity of aromatic compounds that R. palustris can degrade depends on the existence of accessory pathways that transform aromatic monomers to the common BAD pathway intermediates benzoyl-CoA or 4-hydroxybenzoyl-CoA ( 20 , 21 ). In addition, previous studies have shown that the growth of R. palustris in lignocellulosic hydrolysates that contain a mixture of plant-derived organic compounds allows for the degradation of aromatic monomers that do not support growth when supplied as the sole carbon source in defined medium ( 21 ). Here we describe studies aimed at understanding the metabolism of syringic acid by an adapted R. palustris strain. By supplying syringic acid to a series of successive cultures, we isolated a strain of R. palustris capable of utilizing this meta -methoxylated aromatic as the sole source of organic carbon. We analyzed the degradation of syringic acid by this adapted isolate, R. palustris SA008.1.07, in defined laboratory medium to provide insight into the mechanisms involved in the degradation of this aromatic monomer.",
"discussion": "RESULTS AND DISCUSSION Isolation of a syringic acid-degrading R. palustris strain. R. palustris CGA009 is reported to be unable to grow photoheterotrophically with syringic acid as the sole organic carbon source ( 14 ). To explore the potential for R. palustris to evolve the capacity to degrade syringic acid, we established a series of anaerobic cultures in which CGA009 was provided with a combination of syringic acid, benzoic acid, and 4-hydroxybenzoic acid (4-HBA), with the last two being established growth substrates for this strain ( 22 , 23 ). Cultures were kept under illumination and anaerobic conditions for at least 1 week after growth had reached stationary phase. At the conclusion of each growth phase, extracellular samples from each culture were assayed for the presence of aromatic acids. Cultures showing some decrease in the extracellular syringic acid concentration were used as an inoculum for new cultures containing an equal or higher proportion of syringic acid in the medium ( Fig. 1 ). This process was iterated five times with increases in the proportion of syringic acid in the medium until cells were growing on medium in which syringic acid represented 80% of the organic carbon added (measured as chemical oxygen demand [COD]). The highest-performing culture at this stage, as determined by total syringic acid consumption from the medium (culture 5.14 in Fig. 1 ), was plated photoheterotrophically onto solid medium containing this compound as the sole source of organic carbon, and 14 colonies were picked after 2 weeks of incubation. The isolated colonies were then used to inoculate separate liquid photoheterotrophic cultures containing syringic acid as the sole source of organic carbon, and the highest performing of these cultures were incubated in a second round of liquid photoheterotrophic growth on medium containing syringic acid as the sole organic carbon source. From a second anaerobic plating (from culture 7.07 in Fig. 1 ), 12 colonies were obtained. To further test that these cells acquired the ability to grow solely on syringic acid, cells in isolated colonies were first grown photoheterotrophically on succinate and then subcultured to a medium containing syringic acid as the sole photoheterotrophic carbon source. The isolate that degraded the most syringic acid under photoheterotrophic conditions ( Fig. 2 ), hereafter referred to as strain SA008.1.07, was selected for further testing. FIG 1 Final cell density (bars; Klett units) and percentage of syringic acid in the culture medium (blue triangles) during sequential anaerobic incubations. Culture 1.02 was started from a colony of R. palustris CGA009 that did not exhibit significant metabolism or growth on syringic acid as a sole carbon source. Each culture was seeded from a subculture of the prior one, except in the two instances indicated as 1st plating and 2nd plating in the figure. Cells were plated and single colonies were selected for isolation prior to the inoculation of cultures 6.08 and 8.1.07. The initial COD of the medium, used as a measurement of bioavailable organic carbon, was maintained at 1 g COD/liter in all cultures by decreasing the proportion of benzoic acid and 4-HBA upon increases in the syringic acid concentration. All cultures were grown anaerobically at 30°C in sealed glass tubes under constant illumination. FIG 2 Syringic acid consumption by 12 strains isolated from culture 7.07 ( Fig. 1 ). Strain SA008.1.07 had the highest syringic acid transformation and was selected for further study. The initial concentration of syringic acid in these cultures was 3.47 mM. Identification of DMBQ as a compound that accumulates extracellularly during growth on syringic acid by SA008.1.07. We found that when SA008.1.07 used syringic acid as a sole source of organic carbon under anaerobic, photoheterotrophic conditions ( Fig. 3 ), an orange-yellow tint appeared during early stages of culture growth. However, as growth progressed, the color of the culture became dark and distinguishable from the deep-red color of the accumulating biomass. FIG 3 Anaerobic growth of R. palustris SA008.1.07 (red) on 5 mM syringic acid compared to that of parent strain CGA009 (blue) and a light-exposed abiotic control (gray). Solid lines show growth (in Klett units) (●), and dashed lines track the concentrations of syringic acid (□). SA008.1.07 consumed approximately half of the syringic acid initially present in the medium, while CGA009 did not grow on syringic acid. Error bars represent standard deviations from experiments performed in triplicate. High-performance liquid chromatography (HPLC) analysis of the medium before and after the growth of SA008.1.07 revealed the accumulation of a light-absorbing unknown product that eluted at 8.4 min ( Fig. 4A ). By analyzing standards of aromatics that are known or potential syringic acid degradation by-products (3- O -methyl gallic acid, gallic acid, vanillic acid, protocatechuic acid) by HPLC, we determined that none of these compounds were found at detectable levels in supernatants from SA008.1.07 cultures. A liquid chromatography-tandem mass spectrometry (LC-MS/MS) examination of the extracellular unknown indicated an m/z ratio of 169.05 ( Fig. 4B ). For further analysis of this unknown, an extractive procedure was performed on the medium, partitioning the compounds into ethyl acetate (EtOAc) or dichloromethane (DCM) (see Materials and Methods), and both fractions were analyzed by nuclear magnetic resonance (NMR). Syringic acid was identified as the major product in the 1 H NMR of the DCM extract, based both on its spectrum and on a comparison to that of a commercially purchased standard ( Fig. 4E ). The 1 H NMR of the EtOAc extract ( Fig. 4E ) contained two major peaks, indicative of methoxy groups and hydrogen atoms on an aromatic ring. Neither of these signals were split, indicating a lack of coupling to adjacent hydrogen atoms in the compound. The predicted molecular weight of the unknown (∼168.05 g/mol, based on the positive ionization mass spectrometry [MS] spectrum) and the 1 H NMR pattern suggested that 3,5-dimethoxy-1,4-benzoquinone (DMBQ) was the compound that accumulated during growth on syringic acid. Indeed, NMR ( Fig. 4E ) and MS analysis ( Fig. 4D ) of a commercial DMBQ standard, which also has an orange-yellow tint (CAS number 530-55-2), showed that it was indistinguishable from the extracellular product that accumulates when SA008.1.07 is grown on syringic acid. FIG 4 Identification of DMBQ as a soluble extracellular product of SA008.1.07 grown on syringic acid cultures. (A) HPLC contour view of PM-syringic acid medium after SA008.1.07 growth, showing peaks at 8.4 and 9.7 min, with the latter peak corresponding to syringic acid. (B) An LC-MS/MS trace of the compound isolated from the peak collected at 8.4 min suggests an m/z ratio of 169.04 g/mol (molecular weight, ∼168 g/mol). (C) HPLC contour view of the DMBQ standard, showing that the retention time matches that of the unknown peak in panel A. The peak at 4 min is DMSO. (D) LC-MS/MS trace of commercially purchased DMBQ, showing a match to the MS spectrum of the unknown peak in panel B. (E) NMR trace of EtOAc-extracted culture medium, an authentic DMBQ standard, DCM-extracted culture medium, and an authentic syringic acid standard. DMBQ inhibits the growth of R. palustris SA008.1.07. Since syringic acid was not totally degraded in the SA008.1.07 cultures ( Fig. 3 ), we investigated whether the presence of DMBQ affected syringic acid metabolism by this strain. In one test of this hypothesis, we analyzed the photoheterotrophic growth of SA008.1.07 in cultures containing 3 mM syringic acid and various concentrations of DMBQ ( Fig. 5 ). When the initial DMBQ concentration was 0.15 mM or above, we observed complete inhibition of growth (as scored by cell density) and of syringic acid degradation ( Fig. 5 ). In experiments with initial DMBQ concentrations of less than 0.15 mM, growth and syringic acid degradation occurred, and extracellular DMBQ concentrations increased to about 0.19 mM. Thus, the results of this experiment suggested that, over the range of concentrations tested, DMBQ had an inhibitory effect on syringic acid degradation and cell growth. The inhibitory effect increased as the DMBQ concentration increased, suggesting that the buildup of DMBQ in medium containing syringic acid can prevent its total degradation by SA008.1.07. To test this hypothesis, we added 0.3 mM DMBQ (a concentration that approximates the amount found in stationary-phase syringic acid-grown cultures) to an SA008.1.07 culture when growth on syringic acid was detected (see Fig. S1 in the supplemental material). We found that the addition of 0.3 mM DMBQ arrested growth and blocked further syringic acid degradation in this culture compared to the findings for a control not receiving any added DMBQ. FIG 5 Effect of DMBQ on syringic acid degradation by SA008.1.07. The cultures received 3 mM syringic acid and various starting concentrations of DMBQ (black, 0 mM; violet, 0.03 mM; blue, 0.06 mM; green, 0.15 mM; yellow, 0.3 mM; red, 0.6 mM). (A) Solid lines show the cell density (in Klett units) (●); dashed lines show the syringic acid concentration (□). (B) DMBQ concentrations. As the initial concentration of DMBQ increased, cell growth and syringic acid degradation decreased. Cultures with DMBQ concentrations of 0.15 mM or greater showed no growth. To test whether the negative impact of DMBQ on growth was seen in cells grown in the presence of other aromatic substrates, we tested its effects on photoheterotrophic cultures grown on equimolar amounts of benzoic acid and 4-HBA. In this case, we found that addition of 0.3 mM DMBQ to growing SA008.1.07 cultures reduced the rates of growth and of aromatic degradation compared to those for a control not receiving DMBQ (Fig. S2). However, the extracellular DMBQ concentrations decreased in these cultures, suggesting a low rate of DMBQ degradation that was not evident in experiments with syringic acid. To test the effect of DMBQ on cells growing on a nonaromatic substrate, SA008.1.07 was grown on succinate with various concentrations of DMBQ (Fig. S3). In this case, a lag phase was observed when DMBQ concentrations were 0.06 and 0.3 mM, and complete growth inhibition was observed at 0.6 mM. There was also the apparent degradation of DMBQ in these cultures (Fig. S3). These results indicate that the inhibitory effect of DMBQ on growth or substrate utilization is not specific to cells that are using syringic acid as a sole organic carbon source. However, the inhibitory effect of exogenous DMBQ was more pronounced in cultures growing on aromatic substrates than in those growing on succinate as an organic carbon source. Furthermore, the evidence obtained with these experiments is not sufficient to determine whether DMBQ is in the syringic acid degradation pathway. For instance, a benzoquinone has been described to be a toxic intermediate in the degradation pathway of pentachlorophenol by Sphingobium chlorophenolicum ( 24 ). The decrease in the DMBQ concentration observed in experiments with 4-HBA and succinate could be a result of DMBQ either being slowly degraded or reacting with cellular components, as described for tetrachlorobenzoquinone in S. chlorophenolicum ( 24 ). Syringic acid degradation by R. palustris SA008.1.07 does not require the BAD pathway. To date, the only known route for photoheterotrophic degradation of aromatic compounds in R. palustris is through the BAD pathway ( 19 ) (Fig. S4). To examine the role of the BAD pathway in syringic acid degradation by SA008.1.07, we created SAΔbadE, a mutant of this adapted strain lacking the benzoyl-CoA reductase gene. This deletion is sufficient to block the anaerobic degradation of all tested aromatic substrates in wild-type strain R. palustris CGA009 ( 19 ). We found that the SAΔbadE mutant strain lacks the ability to consume benzoic acid or 4-HBA, as expected ( Table 1 ). However, we also found that SAΔbadE grows on syringic acid, exhibiting a behavior similar to that of the parent strain, SA008.1.07 ( Fig. 6 ). We also examined the role of the peripheral HBA pathway, responsible for the conversion of 4-HBA into benzoyl-CoA (Fig. S4), in the growth of strain SA008.1.07 on syringic acid. To do this, we created SAΔhbaB, a mutant of SA008.1.07 which lacks the 4-hydroxybenzoyl-CoA reductase gene, which is known to be required for 4-HBA metabolism in R. palustris CGA009 ( 25 ). As expected, we found that the SAΔhbaB mutant lacks the ability to degrade 4-HBA, yet it can degrade benzoic acid ( Table 1 ). As with the SAΔbadE mutant, we found that SAΔhbaB maintained the ability to grow on and degrade syringic acid ( Fig. 6 ). TABLE 1 Endpoint analysis of R. palustris SA008.1.07 bad and hba mutants grown in PM medium containing benzoic acid and 4-HBA a Culture Benzoic acid concn (mM) 4-HBA concn (mM) Final cell density (Klett units) SA008.1.07 ND ND 172 SAΔbadE 1.36 1.45 18 SAΔhbaB ND 1.43 81 a The PM medium contained benzoic acid at 1.41 mM and 4-HBA at 1.63 mM. ND, not detected. FIG 6 Photoheterotrophic degradation of syringic acid by R. palustris strains SA008.1.07 (blue), SAΔbadE (red), and SAΔhbaB (gray). Solid lines show the cell density (in Klett units) (●), dashed lines show the concentrations of syringic acid (□), and dotted lines show the DMBQ concentrations (Δ). From these experiments, we conclude that neither the peripheral HBA pathway nor the BAD pathway is required for the degradation of syringic acid by R. palustris SA008.1.07. This was a surprising result because the BAD pathway is the only known route for anaerobic aromatic metabolism in R. palustris ( 16 , 19 ). Growth on syringic acid does not induce expression of BAD pathways in R. palustris SA008.1.07. We used RNA sequencing (RNA-seq) to compare the global changes in transcript levels in cultures of SA008.1.07 anaerobically grown on syringic acid, 4-HBA, and succinate ( Tables 2 and 3 ). Comparing growth on 4-HBA to growth on succinate revealed the expected increase in the transcript abundance of genes involved in the BAD pathway and the peripheral HBA pathway ( Table 2 ). This is consistent with the above-mentioned finding that SA008.1.07 uses the BAD pathway for 4-HBA metabolism ( 18 ). However, the abundance of transcripts from these genes was much lower and mostly not significantly differentially expressed ( P > 0.05) when the growth of SA008.1.07 on syringic acid and succinate was compared ( Table 2 ). Therefore, in addition to SA008.1.07 not needing the BAD and HBA pathways for growth on syringic acid ( Fig. 6 ), the transcriptomics data show that growth in the presence of syringic acid does not induce the expression of known genes within the BAD and HBA pathways. TABLE 2 Fold change in transcript abundance for genes predicted to be associated with the BAD and peripheral pathways when strain SA008.1.07 is anaerobically grown on 4-HBA or syringic acid compared to that when it is grown on succinate Gene a \n Name Predicted product Log 2 fold change b \n 4-HBA to succinate SA to succinate rpa0669 hbaA 4-Hydroxybenzoate-CoA ligase 9.98* 3.63 rpa0670 hbaB 4-Hydroxybenzoyl-CoA reductase subunit 8.40* 3.01 rpa0671 hbaC 4-Hydroxybenzoyl-CoA reductase subunit 8.37* 2.58 rpa0653 badI 2-Ketocyclohexanecarboxyl-CoA hydrolase 7.84* 2.94* rpa0658 badE Benzoyl-CoA reductase subunit 7.68* 0.36* rpa0659 badF Benzoyl-CoA reductase subunit 7.39* 1.13 rpa0660 badG Benzoyl-CoA reductase subunit 7.10* 0.95 rpa0656 badC Alcohol dehydrogenase 6.82* 0.83 rpa0654 badH 2-Hydroxycyclohexanecarboxyl-CoA dehydrogenase 6.70* 2.00 rpa0651 aliA Cyclohexanecarboxylate-CoA ligase 6.38* 1.00 rpa0672 hbaD 4-Hydroxybenzoyl-CoA reductase subunit 6.35* 0.83 rpa0657 badD Benzoyl-CoA reductase subunit 6.09* −0.40 rpa0655 badR Benzoate anaerobic degradation transcription regulator 5.82* 1.37 rpa0652 aliB Cyclohexanecarboxyl-CoA dehydrogenase 5.70* 0.87 rpa0650 badK Cyclohex-1-ene-1-carboxyl-CoA hydratase 5.62* 0.69 rpa0667 hbaF Inner membrane translocator 5.50* 0.98 rpa0662 badB Ferredoxin 5.01* 0.55 rpa0661 badA Benzoate-CoA ligase 4.91* 0.77 rpa0668 hbaE ABC transporter subunit substrate-binding component 4.83* 0.95 rpa0665 hbaH ABC transporter ATP-binding protein 4.67* 0.50 rpa0673 hbaR Hydroxybenzoate anaerobic degradation regulatory protein 4.19* 0.46* rpa0666 hbaG ABC transporter ATP-binding protein 4.10* −0.01 rpa0664 badL Acetyltransferase 3.64* 0.04 rpa3714 pimC Pimeloyl-CoA dehydrogenase large subunit 3.60* 0.67 rpa3713 pimD Pimeloyl-CoA dehydrogenase small subunit 3.43* 0.23 rpa0663 badM Transcriptional regulator BadM 3.02* −0.03 rpa3717 pimF Enoyl-CoA hydratase 2.63* 0.45 rpa3715 pimB Acetyl-CoA acetyltransferase 2.58* −0.22 rpa3716 pimA AMP-dependent synthetase/ligase 2.53* 0.53 a Genes are sorted, in descending order, with respect to the log 2 fold change in abundance when growth is on 4-HBA compared to that when growth is on succinate. b SA, syringic acid. *, statistically significant difference ( P < 0.05). TABLE 3 Transcripts with the highest increase in abundance when strain SA008.1.07 is anaerobically grown on syringic acid compared to that when it is grown on succinate Gene a \n Name Predicted product Log 2 fold change b \n SA to succinate 4-HBA to succinate rpa0910 Pirin family protein 9.70 1.95 rpa2160 3-Oxoacyl-ACP reductase 7.88* −1.36 rpa0909 wrbA NAD(P)H dehydrogenase (quinone) 6.80* 0.28 rpa3619 vanA Aromatic ring-hydroxylating dioxygenase subunit alpha 6.66* 0.73 rpa2717 Hypothetical protein 6.27* 4.22 rpa3621 vanB Oxidoreductase 6.11* −0.29 rpa0005 hppD 4-Hydroxyphenylpyruvate dioxygenase 6.07 1.72* rpa3620 vanR GntR family transcriptional regulator 6.01* −1.34 rpa4284 Polyisoprenoid-binding protein 5.95* −0.71 rpa4222 Hypothetical protein 5.91* 1.32* rpa0319 Hypothetical protein 5.90 7.58* rpa3329 Hypothetical protein 5.87 3.42 rpa1475 Hypothetical protein 5.56 0.59 rpa3631 3-Oxoacyl-ACP reductase 5.52* −1.20 rpa4285 Malonic semialdehyde reductase 5.51* −1.73 rpa3565 l,d -Transpeptidase 5.39 −2.17* rpa3943 Ferritin-like domain-containing protein 5.09* 0.55* rpa4394 Isocitrate lyase 5.05* 6.21* rpa3308 Ferritin-like domain-containing protein 5.03 1.12 rpa0320 4-Coumaroyl-homoserine lactone synthase 5.00 6.71* rpa1089 Hypothetical protein 4.99 1.32 rpa0214 Hypothetical protein 4.92* 0.66* rpa4220 l,d -Transpeptidase 4.92 0.68 rpa4286 Dioxygenase 4.90* −2.03 rpa2895 Hsp20/alpha crystallin family protein 4.87 0.92* a Genes are sorted, in descending order, with respect to the log 2 fold change in abundance when growth is on syringic acid compared to that when growth is on succinate. Shaded rows indicate that the gene was explored in this study. b SA, syringic acid. *, statistically significant difference ( P < 0.05). Identification of a gene cluster required for syringic acid degradation by R. palustris SA008.1.07. The global gene expression analysis was also used to identify genes with increased transcript abundance when SA008.1.07 was grown on syringic acid compared to that when it was grown on either 4-HBA or succinate ( Table 3 ). Among the transcripts showing the largest increase in abundance are those derived from genes within a putative vanARB ( rpa3619 , rpa3620 , rpa3621 ) operon. The vanARB genes are annotated as coding for the VanAB proteins and a GntR-family transcriptional regulator (VanR), homologues of which are known or proposed to act as repressors of the vanAB genes ( 27 – 30 ). The VanAB proteins are known or predicted subunits of an enzyme (VanAB) with aromatic ring-hydroxylating activity ( 16 , 27 ). Homologues of VanAB are known or predicted to catalyze the oxidation of vanillic acid to protocatechuic acid and formaldehyde in Bradyrhizobium diazoefficiens ( B. japonicum ) ( 31 ) and Pseudomonas sp. strain HR199 ( 32 ). In addition, a VanAB homologue in a Streptomyces strain has the reported ability to demethylate syringic acid as well as other aromatic compounds ( 33 ). A global gene expression analysis of R. palustris SA008.1.07 grown aerobically on vanillic acid (Fig. S5; Table S3) confirmed the predicted role of the vanAB genes in aerobic vanillic acid degradation, since there was an increased abundance of transcripts encoding these genes along with others in a predicted pathway, with protocatechuic acid and formaldehyde being intermediate metabolites of aerobic vanillic acid degradation (Fig. S5). The increased transcript abundance of the vanARB genes when SA008.1.07 was grown anaerobically on syringic acid was unexpected, given that the RNA was isolated from cells grown under anaerobic photoheterotrophic conditions. As described in Materials and Methods, for the RNA-seq experiments, the cultures were continuously bubbled with N 2 and CO 2 to avoid air entering the cultures. For all other experiments, the culture tubes were completely filled with medium, leaving no headspace, and when samples were withdrawn from the cultures for chemical analyses, the resulting headspace was flushed with argon gas to prevent the introduction of air into the cultures. These are standard techniques that have been successfully employed to grow anaerobic bacterial cultures and isolate oxygen-sensitive proteins in their active form ( 34 ). We also monitored the abundance of diagnostic transcripts as a reporter for the presence of oxygen in our photoheterotrophic cultures. Analysis of the transcript abundance of photoheterotrophically grown cultures shows that there was a relatively low abundance of those encoding HemF, an oxygen-dependent coproporphyrinogen oxidase (RPA1514), or subunits of the low-affinity enzymes in the aerobic respiratory chain, such as cytochrome bd (RPA1319, RPA4452, and RPA4793-RPA4794) or cytochrome aa 3 oxidases (RPA1453, RPA4183, and RPA0831 to RPA0836) (Table S4). In contrast, transcripts from the following genes were, on average, ∼32-fold more abundant in the photoheterotrophic cultures than in those mentioned above which are associated with growth in the presence of oxygen: genes encoding subunits of the high-affinity cytochrome cbb 3 oxidase (RPA0015 to RPA0019); genes encoding the oxygen-independent coproporphyrinogen oxidase HemN (RPA1666); those needed for anaerobic growth in the light ( 15 , 35 ), including ones that encode pigment biosynthetic enzymes or pigment-binding proteins of the photosynthetic apparatus (RPA1505 to RPA1507, RPA1521 to RPA1548, RPA1667-RPA1668, RPA3568); plus other genes whose induction requires the global anaerobic regulator FixK (RPA1006-RPA1007, RPA1554) ( 36 , 37 ) ( P = 0.01, unpaired t test) (Table S4). This analysis provides independent experimental evidence that the photoheterotrophic cultures used as a source of RNA or for other experiments in this study were anaerobic. Nevertheless, to further test whether oxygen influences the ability of SA008.1.07 to degrade syringic acid, we performed additional experiments. First, when we tested SA008.1.07 for aerobic growth on the methoxylated aromatics syringic acid and vanillic acid (Fig. S6), we found that this adapted strain could not grow on syringic acid aerobically but could grow aerobically on vanillic acid. We also performed growth experiments in which additional steps were taken to eliminate oxygen from the medium. In one experiment, we used 100-ml serum bottles with PM medium ( 22 ) containing syringic acid and sealed them with rubber septa and aluminum crimp caps. We then flushed the PM medium with argon gas for 20 min and then applied vacuum to remove gases from the bottles and reflushed them with argon. This process was repeated three times to remove as much oxygen as possible. As a control that simulated the conditions used in the experiments described earlier, another group of 100-ml serum bottles was used, but in this case, the bottles were sealed without using the degassing procedure. SA008.1.07 was inoculated into both sets of bottles through sterilized syringes and needles. In these experiments, we observed no significant difference on the growth of SA008.1.07, the consumption of syringic acid, or the production of DMBQ between the degassed bottles and the nondegassed controls (Fig. S7), demonstrating that any traces of oxygen potentially present at the initiation of the incubations did not influence the ability of R. palustris SA008.1.07 to grow on syringic acid under anaerobic photoheterotrophic conditions. In a separate experiment, we used l -cysteine as a reducing agent and resazurin as an oxygen indicator (Fig. S8). Inoculation of R. palustris SA008.1.07 was performed after the resazurin was colorless, indicating the absence of oxygen. Syringic acid degradation and DMBQ production were observed, similar to the findings of the experiments performed with other techniques, further confirming that oxygen is not involved in the transformation of syringic acid. Based on these results, we proceeded to investigate whether the vanARB operon participated in anaerobic syringic acid degradation by SA008.1.07. To do this, we deleted the entire vanARB operon in SA008.1.07 (producing strain SAΔvan; Table 4 ) and found that this strain lost its ability to grow anaerobically on syringic acid ( Fig. 7A ). In addition, we found that transforming SAΔvan with a plasmid carrying either the wild-type vanARB operon or only wild-type vanAB (producing strain SAΔvan/pBRvanARB or SAΔvan/pBRvanAB, respectively; Table 4 ) under the control of a constitutive promoter rescued the ability of SAΔvan to grow on and degrade syringic acid under anaerobic conditions, although cell densities were lower than those for SA008.1.07 ( Fig. 7A ). Thus, we conclude that the vanAB genes in the R. palustris \n van cluster are required for the anaerobic degradation of syringic acid by SA008.1.07. In control experiments, we found that, as expected, the activities of the BAD and HBA aromatic pathways were not affected by the loss of vanARB , as SAΔvan was able to grow photoheterotrophically on 4-HBA or benzoic acid ( Fig. 7B ). Placing the same vanARB plasmid in the wild-type CGA009 strain (A9/pBRvanARB; Table 4 ) did not confer on this strain the ability to grow on syringic acid ( Fig. 7A ), indicating that yet to be identified genes outside this operon are required for syringic acid metabolism by SA008.1.07. TABLE 4 Strains and plasmids used in this study Strain or plasmid Description Source or reference Strains E. coli DH5α supE44 lacU169 (ϕ80dΔ lacZ M15) hsdR178 recA1 endA1 gyrA96 thi-1 relA1 Invitrogen-THF S17-1 C600::RP-4 2-(Tc::Mu) (Kn::Tn 7 ) thi pro hsdR HsdM + \n recA 55 NEB 5α fhuA2 Δ( argF-lacZ ) U169 phoA glnV44 ϕ80Δ( lacZ )M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17 NEB R. palustris CGA009 Wild-type strain 22 SA008.1.07 Derivative of CGA009 able to grow on syringic acid This work SAΔbadE Deletion of 3′ end of badD , whole badE gene, and 5' end of badF in SA008.1.07; ΩKn r cassette insertion in place of deleted nucleotides This work SAΔhbaB Deletion of hbaB in SA008.1.07 This work SAΔvan Deletion of vanARB operon in SA008.1.07 This work A9/pBRvanARB Gm r ; CGA009 carrying pBRvanARB vector This work SAΔvan/pBRvanARB Gm r ; SAΔvan carrying pBRvanARB vector This work SAΔvan/pBRvanAB Gm r ; SAΔvan carrying pBRvanAB vector This work SAΔ2160 Deletion of rpa2160 in SA008.1.07 This work SAΔ4286 Deletion of rpa4286 in SA008.1.07 This work SAΔ1972 Deletion of rpa1972 in SA008.1.07 This work A9Δ1972 Deletion of rpa1972 in CGA009 This work Plasmids pSUP202 Mobilizable suicide plasmid 55 pK18mobsacB oriV oriT mob sacB Kn r 53 pS202badE 3.7-kb fragment containing badE and most of surrounding genes badD and badF cloned into HindIII/BamHI sites of pSUP202 This work pS202ΔbadE Deletion of 2.1-kb fragment containing badE , 3' end of badD , and 5′ end of badF and insertion of 2.3-kb ΩKn r cassette in pS202badE This work pK18hbaB Kn r ; 2.1-kb fragment containing hbaB and 800-bp flanking regions cloned into XbaI/HindIII sites of pK18mobsacB This work pK18ΔhbaB Kn r ; deletion of hbaB in pK18hbaB This work pKΔvanARB Kn r ; ∼1.5 kb upstream and ∼1.5 kb downstream flanking regions of vanARB operon cloned into the XbaI/HindIII sites of pK18mobsacB This work pBBR1MCS-5 IncA/C Gm r ; broad-host-range cloning vector 54 pBRvanARB Gm r ; vanARB operon cloned into pBBR1MCS-5 vector This work pBRvanAB Gm r ; vanA and vanB genes cloned into pBBR1MCS-5 vector This work FIG 7 (A) R. palustris SAΔvan (red), a mutant culture of SA008.1.07 (black) with the vanARB operon deleted, does not grow on syringic acid. The complementation of vanARB on expression plasmids pBRVanARB and pBRVanAB restores syringic acid-degrading activity in SAΔvan/pBRvanARB (green) and SAΔvan/pBRvanAB (blue). The expression plasmid pBRVanARB does not impart syringic acid-degrading activity when inserted into wild-type strain CGA009 (A9/pBRvanARB; gray). Solid lines show growth (in Klett units) (●), dashed lines track the concentrations of syringic acid (□), and dotted lines track the DMBQ concentration (Δ). (B) Both R. palustris SA008.1.07 (black) and SAΔvan (red) grow on benzoic acid (circles) and 4-HBA (triangles). Solid lines show growth (in Klett units), and dashed lines indicate aromatic concentrations. In addition to the genes in the vanARB operon, we also tested the effect of deleting two other genes that showed an increased transcript abundance during growth on syringic acid. One gene encodes an oxidoreductase and had one of the highest increases in transcript abundance ( rpa2160 ), and the other gene was annotated as encoding a dioxygenase and had a lower increase in transcript abundance ( rpa4286 ) ( Table 3 ). Experiments with deletion mutants of SA008.1.07 lacking these genes, SAΔ2160 and SAΔ4286, respectively ( Table 4 ), showed that neither deletion affected photoheterotrophic growth on syringic acid (Fig. S9), indicating that these genes are not required for the breakdown of syringic acid by SA008.1.07. Identification of mutations in strains adapted to grow on syringic acid. In an attempt to identify additional mutations that could confer on R. palustris SA008.1.07 the ability to grow photoheterotrophically on syringic acid, we resequenced strain SA008.1.07 along with 16 other R. palustris isolates that had acquired the same metabolic ability by performing the same enrichment and isolation experiments described above (Table S5). When the genome sequences of this panel of isolates were compared to the genome sequence of R. palustris CGA009 (Table S6), only 4 mutations were found in the majority of the strains ( Table 5 ). One mutation was an indel upstream of rpa0746 , a gene annotated as encoding a cytochrome c -type cytochrome of unknown function. A second mutation was a frameshift in rpa1972 , a gene annotated as encoding a two-component sensor histidine kinase, for which no function is known. The other two mutations were nonsynonymous, causing amino acid changes in rpa2457 , which encodes a hypothetical protein, and rpa3268 , which encodes the β subunit of RNA polymerase. No mutations were detected in the vanARB operon in any of the syringic acid-metabolizing strains that were sequenced. TABLE 5 Mutations identified in more than half of the 17 adapted R. palustris strains conferring the ability of syringic acid degradation compared with the genome of CGA009 Position Reference Alteration Mutation type Amino acid change Gene Name Function Occurrence of mutation 826195 C A Substitution Upstream rpa0746 Cytochrome c -type cytochrome 17 of 17 strains 2221051 TC T Frameshift at GLU242 Premature stop codon at 2220962 rpa1972 Two-component sensor histidine kinase 17 of 17 strains 2795227 G A Nonsynonymous Glycine → aspartic acid rpa2457 Hypothetical protein 16 of 17 strains 3685350 T G Nonsynonymous Threonine → proline rpa3268 rpoB RNA polymerase β subunit 11 of 17 strains We were unsuccessful in our attempts to delete rpa2457 and rpa3268 from SA008.1.07 using the methods used in this study, which is not surprising, since both of these genes have been shown to be essential for the growth of R. palustris ( 15 ). We successfully deleted rpa1972 in both CGA009 and SA008.1.07, creating strains SAΔ1972 and A9Δ1972, respectively. To test the hypothesis that the observed frameshift in rpa1972 altered the function of this predicted histidine kinase and somehow influenced syringic acid degradation by SA008.1.07, we evaluated the photoheterotrophic growth of both SAΔ1972 and A9Δ1972 on syringic acid. This experiment showed that deletion of rpa1972 in CGA009 did not enable A9Δ1972 to grow on syringic acid, nor did deletion of this gene in SA008.1.07 prevent SAΔ1972 from growing on syringic acid (Fig. S10). Therefore, additional efforts are needed to identify single or synergistic combinations of mutations in SA008.1.07 or other adapted strains that contribute to anaerobic growth on syringic acid. Concluding remarks. meta -Methoxylated aromatics are present at significant levels in the lignin of different plants and are potential sources of compounds for industrial applications. In this work, we isolated a strain of R. palustris that acquired the ability to use syringic acid as a growth substrate under photoheterotrophic conditions. Our strategy of incrementally exposing cultures to higher concentrations of syringic acid, while at the same time reducing the availability of the known growth substrates benzoic acid and 4-HBA, has been shown to be conducive to adaptation and the acquisition of new metabolic activities in R. palustris ( 38 , 39 ) and other bacteria ( 40 ). Our analysis of this adapted strain, SA008.1.07, has provided important new knowledge on the bacterial metabolism of syringic acid. First, we found that syringic acid degradation does not occur through or induce the expression of the genes in the well-characterized BAD pathway. This finding makes syringic acid an aromatic compound whose photoheterotrophic metabolism does not utilize the BAD pathway in R. palustris . In addition, the increased abundance of vanARB transcripts in SA008.1.07 cultures grown in the presence of syringic acid and the requirement of vanAB for the growth of this adapted strain on this methylated aromatic provide evidence for a heretofore unknown role of the VanAB enzyme in the anaerobic metabolism of this compound. Since the previously reported function of vanAB is in the aerobic demethylation of vanillic acid ( 31 , 32 ), our observations suggest that the VanAB enzyme may have an additional unrealized function under anaerobic conditions. Known homologues of VanAB are reported to contain an oxygen-sensitive iron sulfur cluster ( 32 ), so our findings reinforce the suggestion that additional experiments are needed to test the role of this enzyme in the anaerobic metabolism of syringic acid. Our analysis of syringic acid metabolism by R. palustris SA008.1.07 sets the stage for further studies of the metabolism of this and other aromatics by this and other bacteria and for the evaluation of previously unexplored functions of the VanAB enzyme. Elucidating such novel pathways and metabolic functions could expand our ability to use microbial transformations of lignin and other renewable resources as biomass-based sources of compounds with potential uses in the energy, chemical, pharmaceutical, and other industries."
} | 10,673 |
32914160 | null | s2 | 2,703 | {
"abstract": "Quorum sensing (QS) is a mechanism by which bacteria regulate cell density-dependent group behaviors. Gram-positive bacteria generally rely on auto-inducing peptide (AIP)-based QS signaling to regulate their group behaviors. To develop synthetic modulators of these behaviors, the natural peptide needs to be identified and its structure-activity relationships (SARs) with its cognate receptor (either membrane-bound or cytosolic) need to be understood. SAR information allows for the rational design of peptides or peptide mimics with enhanced characteristics, which in turn can be utilized in studies to understand species-specific QS mechanisms and as lead scaffolds for the development of therapeutic candidates that target QS. In this review, we discuss recent work associated with the approaches used towards forwarding each of these steps in Gram-positive bacteria, with a focus on species that have received less attention."
} | 232 |
22470577 | PMC3314632 | pmc | 2,706 | {
"abstract": "Background Stability is a crucial ecosystem feature gaining particular importance in face of increasing anthropogenic stressors. Biodiversity is considered to be a driving biotic force maintaining stability, and in this study we investigate how different indices of biodiversity affect the stability of communities in varied abiotic (composition of available resources) and biotic (invasion) contexts. Methodology/Principal Findings We set up microbial microcosms to study the effects of genotypic diversity on the reliability of community productivity, defined as the inverse of the coefficient of variation of across-treatment productivity, in different environmental contexts. We established a bacterial diversity gradient ranging from 1 to 8 Pseudomonas fluorescens genotypes and grew the communities in different resource environments or in the presence of model invasive species. Biodiversity significantly stabilized community productivity across treatments in both experiments. Path analyses revealed that different aspects of diversity determined stability: genotypic richness stabilized community productivity across resource environments, whereas functional diversity determined stability when subjected to invasion. Conclusions/Significance Biodiversity increases the stability of microbial communities against both biotic and abiotic environmental perturbations. Depending on stressor type, varying aspects of biodiversity contribute to the stability of ecosystem functions. The results suggest that both genetic and functional diversity need to be preserved to ensure buffering of communities against abiotic and biotic stresses.",
"introduction": "Introduction Human activities are affecting the functioning of virtually all Earth's ecosystems via multiple and exacerbating environmental changes [1] . This evoked a scientific quest for stabilizing mechanisms within ecosystems and unifying across-ecosystem theories [2] – [6] . Biodiversity is contentiously discussed as one biotic ecosystem property determining stability and may thus be essential for human well-being [5] , [7] . However, varying definitions of stability as well as the underrepresentation of certain study systems and stability measures, and differences in experimental designs complicate this discussion [3] , [5] , [8] . Though inconsistent reports still fan the ongoing debate [2] , evidence accumulates that biodiversity significantly determines major facets of ecosystem stability, such as temporal [9] and spatial variability [10] , resistance against perturbations [11] and invasions [12] , resilience [13] and reliability [14] . The underlying notion is that diverse communities host a variety of life strategies that are able to respond differentially to environmental perturbations and maintain ecosystem functioning through a plethora of traits [3] , [9] , [15] . This means that asynchrony in species' responses to environmental fluctuations due to niche differences stabilize ecosystem functions at high diversity [6] , [16] – [18] . This explanation may not only be crucial for the stability of ecosystem functions in response to temporal fluctuations but also to spatial fluctuations or varied environmental contexts, such as differences in resource availability/composition or when affected by biotic invasion. In fact, spatial and temporal stability are closely interrelated [19] . Microorganisms represent the functional backbone of virtually any ecosystem [20] , [21] , and it is essential to understand their response under changing abiotic and biotic conditions. Therefore, diversity–stability relationships in microbial communities need closer consideration [4] , [14] , [22] – [24] . Here we address this issue by considering various stability measures of microbial productivity as a function of genotypic richness and functional diversity, two of the most prominent indices of biodiversity. Recent research stressed that different aspects of diversity (e.g., species richness, functional diversity and phylogenetic diversity) are responsible for ecosystem functioning [25] – [27] , and we propose that this also applies to ecosystem stability. Phylogenetic and/or functional diversity may better predict ecosystem functioning than species richness per se \n [26] , [28] . In particular functional diversity of communities may be a major driver of their performance [26] , [27] and stability, while species richness has been manipulated in the vast majority of previous studies [29] . We thus investigated the impacts of genotypic richness and functional diversity of bacterial communities in the present study. An important feature of communities is the reliability [14] , [30] or predictability [22] , [31] of functioning, the “probability that a system (specific community) provides a consistent level of functioning” [14] after a certain amount of time, i.e., a consistent level of functioning in varied abiotic or biotic contexts. In contrast to temporal stability, reliability is commonly measured as the across-treatment variation in ecosystem functioning [30] – [32] . This measure can thus be used to investigate the stability of ecosystem functioning of a given community in varied abiotic and biotic contexts. To investigate the linkage between biodiversity and reliability in multiple contexts, we manipulated the diversity of Pseudomonas fluorescens communities. Diversity was expressed as genotypic richness and functional diversity (as defined by Petchey and Gaston [33] ). We subjected the communities to varied resource environments and to invasion by functionally similar invaders, thereby simulating two of the most important human induced stressors of ecosystems [1] . We measured the reliability of the communities as the stability of productivity of a given bacterial community across the treatments. Since diverse communities are more likely to contain genotypes with different responses to varied abiotic and biotic contexts [6] , [16] , [17] , we expected the productivity of genetically and functionally more diverse communities to be more stable.",
"discussion": "Discussion Biodiversity is a major predictor of the reliability of various communities including plants and microbes [3] , [14] , [31] . Our results indicate that biodiversity stabilizes the productivity of microbial communities in varied abiotic and biotic contexts for the first time using an across-replicate comparison of microbial productivity. Importantly, different biodiversity indices had to be considered to predict the stability of bacterial productivity. Genotypic richness, i.e., the number of genotypes present in a community, was the main driver of stability in varied resource environments, while functional diversity was closely related to the stability of communities subjected to invasion. This finding is surprising since genotypic richness and functional diversity were highly correlated in the present study, but the two diversity indices differed substantially in their explanatory power regarding stability in different environmental contexts. Although the mechanisms underlying the differential significance of genotypic richness and functional diversity cannot be uncovered with the design of the present study, the results provide guidelines for future experiments targeting these mechanisms. The increase in the stability of community productivity with genotypic richness in varied resource environments is likely to be related to an insurance effect [18] : diverse communities are more likely to contain genotypes being able to use new and/or varied resources. If the increased growth of some genotypes is sufficient to compensate the lower growth of the ones that are unable to use the new resources, then aggregate community performance will remain stable across treatments [6] , [16] , [17] . This, however, implies that dominance between the strains will vary, and further experiments investigating the relative performance of the different genotypes are needed to understand the stability of community composition across treatments. Moreover, genotypic richness may have encompassed functional traits not captured by our functional diversity index [44] . Stability in the varied invader experiment was best explained by functional diversity. Niche preemption by functionally diverse resident communities reduces the success of invasive species [12] , [45] . The higher stability of diverse communities in our experiments therefore suggests that niche preemption reduced the effect size of invaders. Hence, functional diversity, i.e., the diversity of functional traits involved in resource capture, likely was responsible for the increased invader resistance of (N. Eisenhauer, W. Schulz, A. Jousset, S. Scheu, unpublished data) and the decreased invader effect size within more diverse bacterial communities. Our results indicate that communities of low diversity are likely to be more sensitive to environmental changes, while diverse communities more stably maintain their functioning. The importance of biodiversity for ecosystem functioning [4] , [20] , [22] , [23] and stability [2] , [3] , [5] is well established. Adding to these findings, we showed that in varied abiotic and biotic contexts different aspects of microbial diversity account for stability. Genotypic richness increased the reliability of community productivity across different resource treatments. This indicates that species-rich communities may be buffered against changes in resource composition, and maintain their function in case of environmental changes or habitat degradation. In contrast, functional diversity was the best predictor for the reliability of community productivity when subjected to invasion. This suggests that functionally diverse communities cope better with new species, and maintain their functionality in presence of invaders. Different aspects of biodiversity of a given community therefore complement each other in warranting the stability of communities facing multiple stressors. Overall, the results suggest that for maximizing the stability of functions of natural communities facing multiple perturbations, such as those increased by anthropogenic activity, as many aspects of biodiversity as possible should be conserved."
} | 2,574 |
33070100 | null | s2 | 2,707 | {
"abstract": "Extracellular electron transfer via filamentous protein appendages called 'microbial nanowires' has long been studied in Geobacter and other bacteria because of their crucial role in globally-important environmental processes and their applications for bioenergy, biofuels, and bioelectronics. Thousands of papers thought these nanowires as pili without direct evidence. Here, we summarize recent discoveries that could help resolve two decades of confounding observations. Using cryo-electron microscopy with multimodal functional imaging and a suite of electrical, biochemical, and physiological studies, we find that rather than pili, nanowires are composed of cytochromes OmcS and OmcZ that transport electrons via seamless stacking of hemes over micrometers. We discuss the physiological need for two different nanowires and their potential applications for sensing, synthesis, and energy production."
} | 226 |
30202440 | PMC6128992 | pmc | 2,708 | {
"abstract": "Background Biogenic and biogenic-thermogenic coalbed methane (CBM) are important energy reserves for unconventional natural gas. Thus, to investigate biogenic gas formation mechanisms, a series of fresh coal samples from several representative areas of China were analyzed to detect hydrogen-producing bacteria and methanogens in an in situ coal seam. Complete microbial DNA sequences were extracted from enrichment cultures grown on coal using the Miseq high-throughput sequencing technique to study the diversity of microbial communities. The species present and differences between the dominant hydrogen-producing bacteria and methanogens in the coal seam are then considered based on environmental factors. Results Sequences in the Archaea domain were classified into four phyla and included members from Euryarchaeota , Thaumarchaeota , Woesearchaeota , and Pacearchaeota . The Bacteria domain included members of the phyla: Firmicutes , Proteobacteria , Bacteroidetes , Actinobacteria , Acidobacteria , Verrucomicrobia , Planctomycetes , Chloroflexi , and Nitrospirae. The hydrogen-producing bacteria was dominated by the genera: Clostridium , Enterobacter , Klebsiella , Citrobacter , and Bacillus ; the methanogens included the genera: Methanorix , Methanosarcina , Methanoculleus , Methanobrevibacter , Methanobacterium , Methanofollis , and Methanomassiliicoccus. Conclusion Traces of hydrogen-producing bacteria and methanogens were detected in both biogenic and non-biogenic CBM areas. The diversity and abundance of bacteria in the biogenic CBM areas are relatively higher than in the areas without biogenic CBM. The community structure and distribution characteristics depend on coal rank, trace metal elements, temperature, depth and groundwater dynamic conditions. Biogenic gas was mainly composed of hydrogen and methane, the difference and diversity were caused by microbe-specific fermentation of substrates; as well as by the environmental conditions. This discovery is a significant contribution to extreme microbiology, and thus lays the foundation for research on biogenic CBM. Electronic supplementary material The online version of this article (10.1186/s13068-018-1237-2) contains supplementary material, which is available to authorized users.",
"conclusion": "Conclusion The data presented in this study reveals a level of in situ hydrogen-producing bacteria and methanogen diversity within the coal seam. Indeed, in some areas of biogenic CBM, microbial consortia consist of coal microorganisms to such an extent that two, or more, microbial groups with complementary metabolic activity comprise these specific systems. Thus, through the exchange of metabolic substrates and microenvironmental regulation symbiosis, groups compete with one another for resources to generate biological methane. In contrast, in areas where biogenic CBM is absent, traces of hydrogen-producing bacteria and methanogens were detected, and it is suspected that the geological and environmental conditions in these regions were unable to provide optimal growth environments for microbial communities involved in metabolic processes. The diversity and abundance of in situ hydrogen-producing bacteria and methanogens within the coal seam studied in this paper are influenced by both biological and non-biological factors. The influence of coal rank on species abundance and diversity is such that microorganisms can grow even in high rank samples to some extent, where the bacterial community is more diverse than the archaeal community. Good groundwater conditions are also more likely to affect the growth of methanogens than is coal rank, as seen in the Jiaozuo (C8) sample. Coal bed temperatures between 25 and 27 °C are the most conducive for the growth of hydrogen and methane-producing bacteria and lead to higher species abundances. The presence of the elements Fe, Co, and Ni also promotes the growth and metabolism of both hydrogen and methane-producing bacteria, whereas mutual competition and promotion amongst microbial communities are also key factors influencing community distribution characteristics. The interactions between these variables also determine the fermentation and metabolism pathways of methanogens in the coal seam.",
"discussion": "Discussion Analysis of factors affecting the microbial community structure To further investigate possible relationships between the environment factors and community variance, RDA analysis was created (Fig. 7 ). In total, eight environment factors, including trace elements Fe, Co, Ni, temperature, salinity, depth, moisture, and R O . The depth and reservoir temperature were measured in the sampling location and other information were obtained from the geological data of local mines (see Table 2 ). Data shown in Fig. 7 a revealed that the bacterial community compositions found in this study were significantly affected by Fe, Ni, moisture, salinity, and R O . All communities, other than C4, C7, C10, are positively correlated with R O ; C4, C7, and C10 positively correlated with Fe, Ni and moisture. Co is required for co-enzyme M methyl-transferase, which is an important enzyme in the biochemical metabolism of methanogens [ 53 ]; therefore, the effect of Co on archaeal community is greater than that of bacterial community. The elements Fe, Co, and Ni; as well as moisture, appeared to be the most significant environmental factors followed by R O and salinity in the archaeal community. There is a significant positive correlation between Co and the C1, C8, C9, C10 communities. All communities except C2, C4, C6, and C5 were negatively correlated with salinity (Fig. 7 b). Fig. 7 The RDA (redundancy analysis) based on the level of bacteria ( a ) and archaea ( b ) with the coal bed environmental factors and coal characteristics. The length of the impact factor is longer, the contribution of the impact is higher, and conversely, when the impact factor is shorter, the contribution of the impact is lighter. When the environmental factor is acutely angled with the sample, there is a positive correlation, and when the environmental factor and the sample angle are obtuse, there is a negative correlation \n Table 2 The environment information and coal samples Coal no. Location Depth (m) Temperature (°C) Salinity (g/L) C1 Yima 500 24.90 0.75 C2 Liyazhuang 578 26.30 0.80 C3 Shaqu 557 26.80 0.50 C4 Hebi 580 25.40 1.00 C5 Jingcheng 550 25.90 0.90 C6 Suzhou 590 26.50 0.70 C7 Neimeng 880 26.00 1.25 C8 Jiaozuo 465 27.20 0.85 C9 Shoushan 820 34.60 0.33 C10 Pingdingshan 670 40.10 0.50 \n Coal rank The coalification “jump” refers to a series of physical and chemical changes under the temperature and pressure of coal during geological history. Coal has thus undergone a process of gradual to sudden change. The four jumps correspond to R O of 0.6, 1.3, 2.5, and 3.0%. Regardless of the archaeal or the bacterial community under consideration, the coal rank has a certain influence on the diversity and abundance of the bacteria. With an increase of coal rank, in both the archaeal and the bacterial communities, the diversity of community shows a certain downward trend overall (Fig. 8 ). Moreover, microorganisms may impact the composition of the coal controlled by the coal ranks. The middle and low rank coals contain large amounts of plant evolved substances in Group 1, which contains a lot of plant evolved substances. Here, there is higher content of hydrogen, oxygen, and nitrogen; and the nutrients required by the bacteria are abundant. In the process of coalification, organic substances generate a lot of moisture and liquid hydrocarbons. At the same time, the side chains of hydrogen and oxygen contained in coal are also abundant. These liquid and solid substances provide the foundation of life for bacteria. As a result, the abundance and diversity of hydrogen-producing bacteria and methanogens in coal in this region are relatively high. With the increase of R o, the side chain content of hydrogen and oxygen in coal is drastically reduced and the components available to the microorganisms are also reduced. Therefore, the species abundance and diversity of the bacterial and archaeal communities in Group 2 and Group 3 are reduced overall. So far, the coal ranks of biogenic coalbed methane have been found in Nature to have a reflectivity of 2.0% (C4 Hebi). After R O > 2.5%, the organic compounds that can be converted to small molecules have been very rare, but there has been a higher diversity and abundance in Group 4. We speculate that the nutrients introduced by groundwater at this time are available for bacterial reproduction. The nutrient components brought about by groundwater in different regions and in different seasons may have contributed to diversity of species. One reason for the higher flora diversity in Group 4 may be that C8 Jiaozuo Jiulishan area has better groundwater runoff conditions and stronger recharge. It can transport nutrients for the flora, so the diversity and abundance are higher than Group 2. It is worth noting that the diversity and abundance of the archaeal communities are negatively correlated with coal ranks to a certain extent. However, the species abundance in the bacterial communities is positively correlated with coal ranks, and the diversity shows a downward trend. With the rise of coal ranks, some bacterial groups gradually adapted to the environment of various coal ranks and can grow and multiply in large numbers, and methanogens are difficult to adapt to coal ranks. Fig. 8 Chao1′s (dark grey) and Shannon’s (light grey) index for the four groups (Coal samples were divided into four groups according to the value of RO for the bacterial community ( a ) and archaea community ( b ), Group 1 represents a value less than 0.6%, Group 2 represents the value between 0.8 and 1.1%, Group 3 represents the value between 1.4 and 1.8%, Group 4 represents the value between 2.67 and 3.15%) derived from regions. 25th and 75th percentiles are indicated by the outer edges of boxes while the maximum and minimum values are showed by the ends of the whiskers and the median by a horizontal line within each box \n Trace metal elements Trace metal elements can promote the growth of microorganisms within a certain range, where the cell maintains homeostasis of the elements through metabolic regulation. Trace metal elements can also exist in various enzymes, which can be absorbed and used by microorganisms in the process of anaerobic metabolism, which has an influence on the community structure of hydrogen-producing bacteria and methanogens (Table 3 ). Table 3 The role of Fe, Co and Ni in reaction and transformation in anaerobic metabolism [ 61 , 62 ] Element functions Element functions Element functions Fe Hydrogenase CO-dehydrogenase Methane monooxygenase NO-reductase Superoxide dismutase Nitrite and Nitrate reductase Nitrogenase Ni CO-dehydrogenase Acetyl-CoA synthase Methyl-CoM reductase (F430) Urease Stabilize DNA, RNA Hydrogenase Co B12-enzymes CO-dehydrogenase Methyltransferase \n Fe and Ni have a greater effect on hydrogen-producing bacteria than Co [ 54 – 56 ]. Fe and Ni can participate in the synthesis and metabolism of hydrogenases and other metalloenzymes in microorganisms. As the content of Fe and Ni increases within an certain range, so does the abundance and diversity of hydrogen-producing bacterial populations. The content of Fe and Ni in C7 is much higher than that in other regions, and this work has found that Clostridium are hydrogen-producing bacteria. This finding indicates that excessive levels of the trace elements may have a toxic effect on the growth of microorganisms and inhibit the activity of metalloenzymes. Levels of Fe in C4, C6, and C9 were not significantly different and were stable at 3500 mg Kg −1 (Fig. 9 ). The relative abundance of Ni in the three areas is C6 > C9 > C4, which corresponds to abundance order (also C6 > C9 > C4), but the order of diversity is C6 > C4 > C9. Members of genera: Clostridium , Klebsiella , Enterobacter and Citrobacter were detected in the C4, C6 and C9 communities; including those at higher abundances and levels of diversity than from other regions. Fig. 9 The trace metal elements content of Fe, Co, Ni in coal samples \n In the archaeal community, the influence of Fe, Co, and Ni on the methanogens is even more important. Co is a key element in the synthesis of methanogenic coenzyme F 430 [ 57 ], and the Co content in the top three is C8 > C7 > C10, methanogen species and abundance being C8 > C7 > C10. Content of Co is positively correlated to the abundance and diversity of methanogens to a certain extent. Although the content of Fe in C7 is much higher than that of other regions, it does not affect the distribution of methanogens in the region. There are only few types of methanogens that may contain Fe—in previous studies, only one species, named Methanothermobacter , was discovered. The presence of monoferric hydrogenase in methanogens of M. marburgensis catalyzes the reversible reaction of methenyl-H 4 MPT + and H 2 to generate methylene-H 4 MPT and H + ; producing methane from CO 2 and H 2 [ 58 ]. Methanogens using hydrogenotrophic metabolism may also contain similar enzymes. In addition, a large proportion of methylotrophic methanogens are also speculated of harboring such enzymes except the Methanoculleus and Methanobacteria. It is speculated that there may be a metalloenzyme associated with Fe in the methylotrophic methanogens. Groundwater conditions Groundwater directly or indirectly provides an ecological basis for the growth and metabolism of extremophiles in the coal seam. On the one hand, groundwater recharge supplies large amounts of nutrients for bacterial and archaeal communities; on the other hand, groundwater environmental conditions (Eh, pH, salinity, ion composition, and trace elements) directly affect microbial growth and metabolic enzyme activity. Groundwater environmental conditions are directly related to the use and degradation of coal, and the microorganisms located in the coal seam show different community structures and functional characteristics. Microbial nutrient substrates are generally dissolved. The runoff zone in the mining area can allows the survival of coal seams. High permeability reservoirs have a positive impact on the growth and reproduction of hydrogen-producing bacteria and methanogens, whereas metamorphism has a significant negative impact on coal permeability in coal reservoirs [ 59 , 60 ]. In areas with biogenic CBM, the C2, C4, C6 and C7 communities have all been well documented. These communities belong to the low and medium coal rank, the porosity of coal is relatively higher than high rank coal, groundwater can provide nutrients to the microbes in the coal seam in time. The current CBM development zone within the Powder River Basin in the U.S. is mainly concentrated in the groundwater runoff zone. The gas stable isotope data from a shallow CBM well in the C6 mining area also confirmed the presence of biogenetic CBM in the area. However, gas stable isotope data from another deep CBM well indicated that the CBM is mainly thermogenic. These results show that as the depth of burial increases, the runoff conditions will weaken and it will be difficult to transport nutrients for the microorganism, which will result in a decrease in the abundance and diversity of the community. The roof and floor of the No. 2 coal seam in C2 area have relatively stable layers of mudstone and clay rock, which makes it difficult for the hydrogen-producing bacteria and methanogens in the coal seam to obtain liquid nutrients, and limits their growth and metabolism and, therefore, their community diversity and abundance. Note that in this area the Chao1 index is 240 and the Shannon index is 1.38 in the bacteria community. The Chao1 index of methanogens is 82, the Shannon index is 0.56. The sandstone fissured aquifer roof in C4 area of the No. 2 1 coal seam has better recharge conditions and fills the coal seam with water. It is possible that the microbial community experiences cumulative effects from the sufficient availability of different nutrients, which affects transportation, compared with the abundance and diversity of the microorganism community in the C2 area, which has greatly improved. In this area, the Chao1 index of hydrogen-producing bacteria is 148, the Shannon index is 1.52; the Chao1 index of methanogens is 368, and the Shannon index is 2.35. The sandstone fissured aquifer of C6 area is a direct water-filled aquifer of No. 3 coal seam. Fracture development within the layer and moderate aqueosity also plays an active role in the abundance and diversity of community. Here, the Chao1 index of hydrogen-producing bacteria is 472, the Shannon index is 1.56; the Chao1 index of methanogens is 384 and the Shannon index is 1.08. This is also the case in the C7 area, the No.5 coal seam has a directly fractured aquifer with good recharge conditions, the abundance index of the hydrogen-producing bacteria is 458, the Shannon index is 1.98; and the Chao1 index of the methanogens is 256 and the Shannon index is 2.47. Therefore, the species diversity of hydrogen-producing bacteria and methanogens in C4, C6, and C7 is higher than that in C2. Groundwater environmental conditions will directly affect the growth and metabolism of microorganisms. The pH value of coal-bed groundwater is generally neutral, but the pH value varies between 6.5 and 8.4 in sandstone fractured aquifer in C4 area No. 2 1 coal seam and the salinity is 1.0 g L −1 . In the direct aquifer layer of the No. 3 coal seam in the C6 area, the pH ranges between pH 6.8 to 8.0, and the salinity is 0.7 g L −1 . The groundwater pH value of the C7 area is 6.1–7.3, and the salinity is 1.25 g L −1 . The pH value in C4, C6, and C7 is close to neutral and the degree of mineralization is low, where the microorganism community has better growth, higher abundance and higher diversity. In addition, the groundwater salinity and ion composition are closely related to the anaerobic reduction environment of the coal seam. For example, SO 4 2− is used to evaluate the closed conditions of groundwater, and HCO 3 − is the product of the anaerobic desulfurization reaction of SO 4 2− , so high HCO 3 − can be used as a sign of good sealing and strong reduction of coal-bed groundwater [ 63 ]. The water chemistry in the C4 area is HCO 3 ·SO 4 –Ca·Mg, the water chemistry in the C7 area is similar to the C4 area, HCO 3 ·SO 4 –Ca·Na, and provides a relatively closed anaerobic environment. In this case, the Chao1 index of hydrogen-producing bacteria in C4 is 148, the Shannon index is 1.52; the Chao1 index of methanogens is 368, the Shannon index is 2.35. The Chao1 index of hydrogen-producing bacteria in C7 is 458, the Shannon index is 1.98, the Chao1 index is 256, and the Shannon index is 2.47. In C6, the water chemistry is SO 4 ·HCO 3 –K·Na, and SO 4 2− is dominant, whereas the Chao1 index of hydrogen-producing bacteria in C6 is 472, the Shannon index is 1.56, the Chao1 index of methanogens is 384, and the Shannon index is 1.08. Data show that diversity in C6 is slightly lower than C4 and C7. Some hydrogen-producing bacteria and methanogens were detected in C8 and C9 areas in the area where no biomethane was found. It was also noteworthy that the groundwater conditions of these two areas are similar to those in the above-mentioned biogenic methane areas, which are located in the groundwater runoff zone and the groundwater recharge is more able to transport some organic matter into the coal seam, so that a large number of bacteria grow and multiply, which is one of the reasons for the higher species abundance and diversity of C8 and C9. Temperature Temperature and trace metal elements influence the abundance and diversity of microbial communities by directly changing both the growth and metabolism of microorganisms and their metabolic environment. Thus, from a microbiological point-of-view, optimum temperature is one of the most important factors that influences microorganism growth and metabolism. Figure 5 b show that temperature exerts a relatively weak influence on the abundance and diversity of methanogens, even though hydrogen-producing bacteria exist within a narrow ecological amplitude and are sensitive to temperature change. This variable is correlated with species abundance and diversity, and the results of this study show that the temperature of the coal seam (i.e., between 25 and 27 °C) is positively correlated with bacterial population abundance. At C8, the temperature was 27.2 °C, the highest temperature recorded in this study. The Chao1 index of hydrogen-producing bacteria was 510 and the Shannon index was 2.61, also the highest among samples (C1–C8). The lowest temperature, 24.9 °C, was found at C1, where the Chao1 index of bacterial community was the lowest. The abundance and diversity of microbial species increases with temperature in C3 > C6 > C2 > C7 > C5 > C4. Geothermal gradient anomalies at C9 and C10 caused much higher temperatures; the ambient temperatures at C9 and C10 were 34.60 °C, and 40.10 °C, respectively. The Chao1 index of hydrogen-producing bacteria in C9 was 176, and the Shannon index was 1.28; the Chao1 index is 237, Shannon’s index is 1.58. Compared to the first eight areas, abundance and diversity has slightly decreased. Here, both hydrogen-producing bacteria and methanogens can grow and reproduce at the ambient temperature. Microbial syntrophic interactions In the extreme environment of the coal seam, consortia of bacteria are formed among the microorganisms in the coal seam. Through the exchange of metabolites and micro-environmentally controlled symbiosis, competition and resource allocations maintain the specific functions of the microbial community, which determines the biomethane production pathway in the coal seam. Methanothrix, which convert acetic acid into methane, is the dominant genus in the methanogen community of the C1 area. The bacteria associated with Alkalibaculum and Desulfosporosinus are homoacetogenic bacteria that use H 2 as the electron donor to produce acetic acid. They are the main competitors for hydrogenotrophic methanogens and also provide metabolic substrate for methanogens. Hydrogen-producing bacteria such as Clostridium and Tissierella also provide acetic acid, and thus the high abundance of hydrogen-producing bacteria provides a rich metabolic substrate for Methanothrix . Together, the methanogens and the hydrogen-producing bacteria are in syntrophic interaction and the methane generation pathway in this area is determined by the decomposition of acetic acid. The methanogens in C2, C4, and C6 are mainly hydrogenotrophic methanogens. Hydrolytic fermentation bacteria and acetogens both contribute to the production of acetic acid and H 2 . They also produce enzymes, cofactors and metabolic signals to regulate the hydrogen production. Furthermore, homoacetogenic bacteria and acetogens do not compete in these areas. Hydrogenotrophic methanogens can produce methane from CO 2 and H 2 produced in the previous stage. Therefore, metabolic pathways in these areas are mainly used for H 2 , formate, and other substances. More than 99% of the C3 area harbors methylotrophic methanogens, such as: Methanolobus . Brevibacter , Paenibacillu s, Brochothrix , and Lactococcus . Previous studies showed that methoxyaromatic compounds (an important part of lignocellulose), are degraded to produce methanol and other substances [ 64 ]. Microorganisms in this region may degrade the lignocellulose-like matter of coal to provide resources to methylotrophic methanogens. This simple microbial community cannot provide sufficient substrates for methanogens that consume H 2 . The biomethane production pathway in this area is based on the consumption of methyl compounds. Staphylococcu s was also detected in the C3 area. Recently, Staphylococcus AntiMn - 1 was isolated from deep-sea sediments in the Clarion-Clipperton area with high manganese content. It contained genes with high resistance to manganese, which is thought to be an adaptation to the marine sedimentary environment [ 65 ]. The heavy metal content in the C3 area is relatively high. It may be that the coal seam environment can effectively induce the expression of resistance genes, which may have antagonistic and detoxifying effects on the transport and toxicity of heavy metals within microorganisms. Staphylococcus in this area may contain resistance genes to adapt to the coal seam environment so that it may also participate in the fermentation metabolism of coal. There are many different species of methanogens in C7, and the hydrogen-producing bacteria is dominated by Clostridium , Bacillus , Citrobacter , and other anaerobes, which provide substrates for acetoclastic methanogens and also H 2 , CO 2 , and formate for hydrogenotrophic methanogens. Furthermore, accumulating acetic acid reduces sulfate-reducing bacteria, including Desulfosporosinus and Desulfitobacterium . The SRB have a stronger affinity for acetic acid than acetoclastic methanogens, but they do not compete with methylotrophic methanogens for certain substrates, such as methanol. Thus, the metabolisms of both sulfate-reducing bacteria and methanogens can proceed simultaneously in this area [ 66 ]. Metabolism in C7 was dominated by methylotrophic methanogens, followed by acetic acid fermentation and then carbon dioxide reduction."
} | 6,403 |
21617741 | null | s2 | 2,709 | {
"abstract": "We present an approach to design spiking silicon neurons based on dynamical systems theory. Dynamical systems theory aids in choosing the appropriate level of abstraction, prescribing a neuron model with the desired dynamics while maintaining simplicity. Further, we provide a procedure to transform the prescribed equations into subthreshold current-mode circuits. We present a circuit design example, a positive-feedback integrate-and-fire neuron, fabricated in 0.25 μm CMOS. We analyze and characterize the circuit, and demonstrate that it can be configured to exhibit desired behaviors, including spike-frequency adaptation and two forms of bursting."
} | 163 |
37938315 | PMC9723713 | pmc | 2,710 | {
"abstract": "Stony coral tissue loss disease (SCTLD) is a widespread and deadly disease that affects nearly half of Caribbean coral species. To understand the microbial community response to this disease, we performed a disease transmission experiment on US Virgin Island (USVI) corals, exposing six species of coral with varying susceptibility to SCTLD. The microbial community of the surface mucus and tissue layers were examined separately using a small subunit ribosomal RNA gene-based sequencing approach, and data were analyzed to identify microbial community shifts following disease acquisition, potential causative pathogens, as well as compare microbiota composition to field-based corals from the USVI and Florida outbreaks. While all species displayed similar microbiome composition with disease acquisition, microbiome similarity patterns differed by both species and mucus or tissue microhabitat. Further, disease exposed but not lesioned corals harbored a mucus microbial community similar to those showing disease signs, suggesting that mucus may serve as an early warning detection for the onset of SCTLD. Like other SCTLD studies in Florida, Rhodobacteraceae, Arcobacteraceae, Desulfovibrionaceae, Peptostreptococcaceae, Fusibacter, Marinifilaceae, and Vibrionaceae dominated diseased corals. This study demonstrates the differential response of the mucus and tissue microorganisms to SCTLD and suggests that mucus microorganisms may be diagnostic for early disease exposure.",
"conclusion": "Conclusions The present study is the first to examine diseased coral and tissue microbiome samples separately, and the first disease transmission experiment conducted on stony coral tissue loss disease infected corals from the US Virgin Islands. This approach allowed us to control for variation in the microbial community in response to environmental changes, and also to make comparisons between the same coral individuals in a control and disease state for six coral species. A similar response to SCTLD infection– a shift to a more dissimilar microbial community – was seen in coral mucus when corals became infected suggesting that the disease is causing dysbiosis in the mucus layer, potentially impacting the immune function. In contrast, we found a species-specific response in the tissue community that is reflective of SCTLD susceptibility patterns, with the most susceptible corals having the greatest shift, converging to a highly similar microbial community within diseased coral tissue. A strength of the current study was the separate investigation of the mucus and tissue microhabitats, and we suggest future studies continue to focus on these communities separately to gain a clearer picture of the changes occurring in the coral microbiome.",
"introduction": "Introduction Marine diseases are growing in magnitude and severity causing economic and biodiversity impacts on marine ecosystems [ 1 – 5 ]. Historically the role of individual microorganisms as causative agents has been the focus of marine disease studies, but it is increasingly recognized that microbiomes, assemblages of associated microorganisms, play a critical role in organismal health and immune function [ 6 ]. The tendency for microbial communities to remain stable when exposed to stressors, or undergo dysbiosis, the breakdown of microstructure and diversity, are emerging as important elements of stress and disease responses. In the marine environment, specific diseases generally affect one species, and it is rare that similar disease signs are exhibited across diverse species [ 7 , 8 ]. However, host fidelity appears less restrictive within diseases of scleractinian corals. Stony coral tissue loss disease (SCTLD) affects phylogenetically diverse Caribbean corals (>22 species) with distinct morphologies and growth rates [ 9 ]. Examining microbiome stability and dysbiosis associated with the onset of disease signs within a multi-species experimental framework provides an opportunity to uncover the characteristics of microorganisms or communities that may be involved in maintaining disease resilience. Despite the toll SCTLD is having on Caribbean coral reefs [ 10 – 14 ], little is known about the microbial contributions to this disease. The diversity, species specificity, and variability of coral-microbial associations all contribute to the challenges of identifying a disease pathogen. Antibiotic treatment has been shown to halt progression of SCTLD, suggesting a bacterial origin [ 15 , 16 ]. Spatial epidemiological models of disease spread have suggested that the pathogen(s) may be waterborne [ 17 , 18 ]. Recently, evidence suggests that a virus may be associated with the disease [ 19 , 20 ]. While there have been some bacterial taxa correlated with the disease in field-based surveys [ 21 – 23 ], no etiological agent(s) have been identified and no transmission studies have specifically examined coral microbiomes during SCTLD onset. Further, why some coral species are susceptible to SCTLD while others are not remains unclear but may be due to species-specific traits such as differential host gene expression, or symbiont (e.g., dinoflagellate endosymbionts of corals in the family Symbiodiniaceae) and microbiome characteristics, which ultimately affect the innate host immune function. Understanding SCTLD-associated microbiome dynamics in a species-specific, as well as coral microhabitat (mucus or tissue) framework is necessary to address this question. Previous studies have found that coral “microhabitats” (e.g., mucus, tissue, skeleton) have distinct microbial communities [ 24 , 25 ]. Further, differentiation between these microhabitats allows for a more targeted understanding of microbial dynamics upon disease acquisition. Mucus layers are found in a vast variety of organisms, from humans to cnidarians, yet the function of this layer is similar among species, in providing a protective barrier between the organism and the external environment. Mucus microbial communities tend to be influenced by both environmental and host factors [ 25 ], playing an important role in coral immunity as a physical barrier that traps and immobilizes pathogens [ 26 ]. Furthermore, this layer is inhabited by beneficial microbes that exclude pathogens from penetrating the mucus layer and thus infecting the coral tissue through both competition and the secretion of antibiotic substances [ 27 , 28 ]. The microbes intimately associated with coral tissue (exclusive of mucus or skeleton) are less studied. Coral tissue is composed of three layers: epidermis, gastrodermis, and mesoglea, with Symbiodiniaceae symbionts located in the gastrodermis and microbial aggregates identified within epidermal and gastrodermal tissues of healthy corals [ 29 , 30 ]. Histological evidence from Landsberg et al. [ 31 ] suggests that SCTLD first affects the Symbiodiniaceae, with lesions originating in the gastrodermis. Thus, focusing on identifying tissue-associated microbes may provide more specific information about potential causative agents or perhaps secondary or opportunistic pathogens responding to tissue sloughing. Advances in human microbiome studies have shown that many stressors and disease result in microbial dysbiosis -- a shift to a microbial community that is detrimental to the organism’s health [ 32 , 33 ]. This shift in the microbiome from mutualistic to dysbiotic can be seen as a reduction in beneficial microbes, an increase in pathogenic microbes, and both reductions and increases in microbial diversity [ 34 – 37 ]. For coral diseases, few have an identified etiological agent, and it is unclear if disease is caused by a single pathogen or rather a shift that occurs in the microbial community. This shift to a dysbiotic community may reduce the efficacy of the microbes in the mucus layer to protect the host and also may reduce the host’s ability to fight infection [ 33 , 36 ] and may be seen as an increase in pathogens present in both the mucus and tissue. In a disease transmission study by Macknight et al. [ 33 ] the microbial community of white plague disease infected coral holobiont samples converged to a community with reduced microbial diversity that was dominated by pathogens. Further, the changes to the microbiome followed a species susceptibility pattern. To understand coral mucus and tissue microbiome dysbiosis upon exposure to SCTLD, we conducted a laboratory-based transmission study with six coral species that vary in terms of their disease susceptibility, reported lesion progression rates in the field [ 38 ], and their representation of phylogenetic and ecological diversity. We hypothesized that both the mucus and tissue microbiomes of diseased corals would significantly differ from that of healthy colonies, and that diseased mucus and tissue would show different microbiome disease signatures. Further, we hypothesized that, upon disease acquisition, the most susceptible species would show greater microbial community similarities among affected colonies compared to the least susceptible coral species.",
"discussion": "Discussion We applied a transmission experiment approach complimented with a subset of field samples, to examine responses of six species of coral to SCTLD. This is the first disease study to separate the mucus and tissue microbiomes of SCTLD infected corals. Mucus microbiome alterations were detected in both diseased (with lesion) and disease exposed (but without a visible lesion) colonies, suggestive of a mucus microbial response to SCTLD that occurs prior to visible lesions. In contrast, diseased tissue microbiomes showed differential species responses that followed species disease susceptibility. Microbiome similarity patterns among colonies also differed between the mucus and tissue compartments upon contracting the disease, which could relate to differences in the role of the mucus and tissue microorganisms in SCTLD. Lastly, we identified common disease-associated bacteria that may serve as indicators for SCTLD, 16 of which were identified in other SCTLD studies. SCTLD mucus microbiomes Here we observed a significant shift in the diseased mucus microbial community of all species compared to apparently healthy controls, along with a community composition that changed to a less similar make up (except for S. siderea ). No other study has examined mucus microbiomes separately from tissue in SCTLD affected corals and doing so allowed us to uncover opposing patterns in mucus and tissue microbiome similarity. The most susceptible coral species showed the smallest divergence in the mucus microbiome, which suggests that dysbiosis is occurring in this mucus layer and may be affecting the immune function provided by coral mucus, and that the most susceptible species are the most sensitive to loss of this protection offered by the mucus. Our experiment showed novel trends in the mucus microbiomes of some SCTLD exposed (without lesion) colonies. Three species, P. strigosa, M. cavernosa and P. astreoides , showed significant differences in the mucus microbiomes of SCTLD exposed colonies, compared to controls, while S. siderea did not. While our observations bring up an interesting idea that the mucus microbiome could serve as an early diagnostic indicator of SCTLD exposure, more time-series type research is needed on susceptible colonies to conclude whether these genets were resistant to SCTLD, or in the early transitional period before lesion development. For future work, it would be informative to pair mucus sampling with histological investigations, to determine if the mucus microbiomes coincide with any of the observed histological changes in the coral tissue [ 39 ]. SCTLD tissue microbiome alterations followed disease susceptibility patterns In this study, the tissue microbiotas of the SCTLD lesioned colonies converged on a general disease signature (Fig. 2 , NMDS). Half of the species, P. strigosa, O. annularis and C. natans , showed significant differences in their tissue microbiotas between lesioned and control colonies. These three species also showed increased similarity in the tissue microbiomes in the lesioned compared to control colonies. These results suggest that the associated microbes may be causative or highly reflective of the disease state in P. strigosa, O. annularis and C. natans . In contrast, P. astreoides, S. siderea and M. cavernosa lesioned tissue microbiomes were not significantly different than controls. These microbiome similarity patterns are reflective of SCTLD susceptibility patterns. Based on results from Meiling et al. [ 39 ] and in corroboration with the SCTLD case definition from the Florida Department of Environmental Protection [ 9 , 53 ], the two most susceptible species in this study were P. strigosa and C. natans . Interestingly, these two species had the greatest percent change (300 and 212% respectively, Fig. 4 ) in the tissue microbial community with disease. The intermediately susceptible species included M. cavernosa (29%) and O. annularis (21%), which had a more moderate convergence in the tissue microbiomes compared to the two highly susceptible species. S. siderea is also an intermediately susceptible species, however, we found a slight divergence in the tissue microbiome. One caveat to consider is that there is uncertainty about whether or not S. siderea in fact contracts SCTLD based on differences in lesion morphology and progression [ 53 ]. However, based on the rapid lesion progression rates as well as no control colonies becoming diseased, we believe that S. siderea did in fact contract SCTLD in this study, and this species-specific microbial response may partly explain why the lesions are so different (Supplementary Fig. 4 ). P. astreoides is considered a rarely susceptible species, but in this study, over half of the treatment colonies developed lesions by the end of the experiment. Notably the tissue microbial community behaved differently compared with the other species, diverging to a less similar community with disease. The high rates of infection in this study may be due to high pathogen loads within an enclosed system. Healthy P. astreoides tissue and mucus microbiomes in this study were highly similar (>65% among colonies, Supplementary Figure 5 ) and enriched with Endozoicomonas (Supplementary Table 2 ), which were lost upon lesion acquisition. Additional research is needed to understand why P. astreoides is more susceptible to SCTLD transmission in a lab setting, and if Endozoicomonas , prominent symbionts of P. astreoides [ 24 , 54 ] provide protection against SCTLD in the field. Conceptually, our results demonstrating microbiome change alongside coral health alterations align with the idea that the mucus microbiome serves as a primary immune response system for corals [ 26 , 55 ]. Further, by showing microbiome divergence between colonies of each species (except S. siderea ), our results suggest that there are likely several factors influencing SCTLD mucus microbiomes. For example, the alterations in the mucus microbiomes upon lesion development likely reflect primary and/or secondary infections, as well as potential reductions in beneficial microbes that contribute to host immune protection. The mucus and tissue microbial responses considered together suggest the microbiome responses among species are tied to species susceptibility, with the most susceptible species having the least dramatic divergence in the mucus microbiome and the most dramatic convergence to a very similar disease community, and the least susceptible species having a less similar tissue microbiome. This suggests that the shift to a dysbiotic community in the mucus microbiome results in a loss of the protective functions of the mucus and allows the tissue microbial community to become dominated by pathogens. SCTLD indicator bacteria This laboratory-based experiment allowed us to sample the lesions early, likely reducing the number of secondary and saprophytic colonizers, and the observation of microbiome consistency among diseased colonies was also observed in the field-based samples, and thus is not likely an artifact of the laboratory setting (Supplementary Figs. 1 and 5 ). There were multiple bacteria that were exact sequence matches or similar (>97%) to those found in SCTLD field studies in the USVI and Florida, including multiple Rhodobacteraceae, Arcobacteraceae, Desulfovibrionaceae, Peptostreptococcaceae, Fusibacter, Marinifilaceae, and Vibrionaceae. Uncovering exact sequence matches across time and geographically disparate locations that were present, even if not identified as statistically more abundant, in all species and sample types suggests an important role that these ASVs are likely playing in this disease. However, it is unclear if these bacteria are causative agents, secondary pathogens, or associated with some part of the tissue breakdown process. Rhodobacteraceae were the most common Family of bacteria enriched in disease mucus and tissue samples (12 total, Table 2 ) and are common associates of both healthy and diseased corals. Some Rhodobacteraceae metabolize dimethylsulfoniopropionate (DMSP), and it is possible that they are attracted to this or other osmolytes released during the tissue sloughing. The genus Nautella includes Nautella italica R11, a pathogen that causes bleaching of red macroalgae, by secreting compounds that inhibit photosynthesis and that aid in algal cell wall penetration [ 56 , 57 ]. Additionally, genomics studies have found that Vibrio have several virulence associated genes, allowing them to deploy many different tools to attack coral and their algal symbionts, such as toxins that cause photoinactivation in Symbiodiniaceae and tissue damage to the coral [ 58 ]. These bacteria warrant further investigation when considered alongside results from Landsberg et al. [ 31 ] which suggest that SCTLD first affects the coral’s algal endosymbiont, Symbiodiniaceae, and is a result of toxicosis. Arcobacter was only found to be enriched in disease mucus samples but not tissue. Arcobacter are found in various coral diseases globally which may point to its potential as an opportunistic bacterium (Supplementary Table 3 ). Desulfovibrionaceae thrive in anoxic, sulfide-rich environments and are a key secondary pathogen in the polymicrobial Black Band Disease, where its role as a sulfate-reducing bacteria results in sulfide production causing coral tissue death [ 59 , 60 ]. Similarly, this bacterium may be acting as a secondary pathogen in SCTLD infected corals. Peptostreptococcaceae and Fusibacter are both anaerobic bacteria, and Marinifilaceae are facultatively anaerobic, further supporting that there is a reduction in the oxygen availability at the disease lesion. Histopathological and transmission electron microscopy has provided evidence that SCTLD is a result of toxicosis [ 31 ] or viral infection [ 19 ]. In the context of our microbial data, we are unable to determine if the bacteria identified are causing the disease or are opportunistic to the altered lesion conditions."
} | 4,785 |
35205878 | PMC8878519 | pmc | 2,711 | {
"abstract": "As an important resource for screening microbial strains capable of conferring stress tolerance in plants, the fungal community associated with the plants grown in stressful environments has received great attention. In this study, high-throughput sequencing was employed to study the rhizosphere fungal community in the reclaimed area (i.e., sites F, H, and T) of the eastern coast of China. Moreover, endophytic fungi from the root of six plant species colonizing the investigated sites were isolated and identified. The differences in soil physicochemical parameters, fungal diversity, and community structure were detected among the sampling sites and between the seasons. Ectomycorrhizal (ECM) fungi (e.g., genera Tuber and Geopora ) were dominant at site F, which was characterized by high soil total carbon (SC) and total nitrogen (SN) contents and low soil electrical conductivity (EC) value. Arbuscular mycorrhizal (AM) fungi, including genera Glomus , Rhizophagus , and Entrophospora were dominant at sites H (winter), H (summer), and T (summer), respectively. The positive relationship between the EC value and the abundance of genus Glomus indicated the ability of this AM fungus to protect plants against the salt stress. Endophytic fungi at sites F ( Aspergillus and Tetracladium ), H ( Nigrospora ), and T ( Nigrospora , Coniochaeta and Zopfiella ) were recognized as the biomarkers or keystone taxa, among which only genus Aspergillus was isolated from the plant roots. The aforementioned AM fungi and endophytic fungi could contribute to the promotion of plant growth in the newly reclaimed land.",
"conclusion": "4. Conclusions The differences in soil physicochemical parameters (i.e., the EC values and contents of SN and SC), fungal diversity, and community structure were detected among the sampling sites (i.e., sites F, H and T) and between the seasons (summer and winter 2020). Based on FunGuild ecological guild assignments, the highest relative abundances of Ectomycorrhizal (ECM) fungi (83.5%), arbuscular mycorrhizal (AM) fungi (21.4%), and endophytic fungi (7.46%) were observed in the samples FS, TS, and TW, respectively. Genera Tuber and Geopora that belong to ECM fungi were dominant at site F, which was characterized by high SC and SN contents and low soil EC value. AM fungi, including genera Glomus , Rhizophagus , and Entrophospora , were dominant in the samples HW, HS, and TS, respectively. The positive relationship between the EC value and the abundance of genus Glomus indicated the ability of AM fungi to protect plants against the salt stress. According to LEfSe analysis and co-occurrence analysis, some biomarkers or keystone taxa belonging to endophytic fungi were recognized at sites F ( Aspergillus and Tetracladium ), H ( Nigrospora ), and T ( Nigrospora , Coniochaeta and Zopfiella ), which could contribute to the promotion of plant growth. However, most of them were not isolated from the plant roots in the investigated sites. A total of 15 endophytic fungi were detected by both root isolation and ITS sequencing of soil samples. It deserves further investigation to isolate more endophytic fungal strains from the plant roots and to evaluate their potential function in conferring stress tolerance to plants.",
"introduction": "1. Introduction The coastal area (Yellow Sea) in Jiangsu province of China has a great length of coastline (1039.7 km) and a large mudflat area (6520.6 km 2 ) [ 1 ]. Large-scale coastal reclamation activities have been performed in China since 1950 to mitigate the conflict between the growing population and shrinking usable land [ 2 ]. Reclamation of coastal areas for agriculture, aquaculture, and forestry was considered as the preferred strategy to increase the food supply and improve the eco-environment [ 3 ]. So far, lots of reclamation activities have been done. However, the process of plants development in the reclaimed area is lagging behind due to some abiotic factors, among which saline stress is the most detrimental to plant growth [ 4 ]. Therefore, it is difficult to promote the exploitation and utilization of the reclaimed area, and more attention should be paid to improve saline stress tolerance of plants. In saline habitats, only the halophytes that constitute about 1% of the world’s flora can thrive, contributing important eco-functions in the desert and coastal areas [ 5 ]. Halophytes are capable of coping with multiple environmental stresses, such as high salinity, tidal flooding, and nutrient deprivation [ 6 , 7 ]. Along with the reclamation processes, the vegetation will gradually shift from halophytes to non-halophytes (salt tolerant grasses, shrubs, and trees) in the old reclaimed regions. In a preliminary investigation in the reclaimed area, Salix was observed as the only tree species growing in all the investigated sites. As a pioneer tree species, willow could naturally grow in harsh soil conditions, which was partly ascribed to the development of dual mycorrhizal symbiosis (i.e., forming both arbuscular mycorrhizas and ectomycorrhizas within the same root system) [ 8 ]. The ability of some tree species (e.g., genera Alnus , Eucalyptus , Populus and Salix ) to form dual mycorrhizal symbiosis is recognized as a key factor to improve their adaptions to unfavorable habitat conditions (e.g., the fluctuations of soil temperature, water content, salinity, and nutrients) [ 9 , 10 ]. It is necessary to explore the fungal community associated with plants (e.g., Salix ) that survived or thrived in the reclaimed area. Great attention has been put on the soil fungal communities, which maintains the ecosystem balance by contributing to the soil organic matter decomposition and mineralization [ 11 ]. It is noteworthy that some types of fungi in soil could penetrate into the plant tissue through the roots and wounds or by horizontal transmission through spores, forming extensive symbiotic relations with the host plant [ 12 ]. Being the two major groups of root fungal symbionts, mycorrhizal and endophytic fungi are generally believed to be critically important to improve host fitness, including productivity and abiotic tolerance, especially under stressful conditions [ 13 , 14 ]. There is relatively limited information concerning the role of the ectomycorrhizal (ECM) symbiosis in enhancing salt tolerance of plants. In contrast, considerable evidence indicates that arbuscular mycorrhizal (AM) fungi colonization in plant roots augments water and nutrient uptake capacities and enhances plant resistance to salinity stresses [ 15 ]. Unfortunately, AM fungi can hardly be grown in pure culture to obtain a large amount of inoculum, which restrained its large-scale application into integrated management of plants development in harsh environments [ 16 , 17 ]. Instead, endophytic fungi that can easily meet the requirements of mass production are gradually recognized as a highly promising mutualistic partners of plants. Endophytic fungi benefit their hosts by promoting resistance against high salinity stress via acting as elicitors in the process of resistance induction [ 18 ]. Nevertheless, endophytic fungi are known to shift their functional role between pathogenicity and mutualism depending on fungal genotype, host, and abiotic conditions [ 19 ]. The available information concerning the diversity of endophytic fungi and their potential function is limited, which deserves extensive investigation. The research on plant-associated fungi in the coastal area has attracted global attention [ 20 , 21 , 22 , 23 ]. Most of the studies on fungi in the coastal area of China focused on mangroves ecosystems along the south China coast [ 24 ] and the salt marsh ecosystem in the Yellow River Delta [ 25 ]. Along the east China coast, limited information was obtained concerning either the root endophytic fungi or the soil fungi community in the reclaimed area. Moreover, the relationship between the soil fungal communities and the endophytic fungi is so complicated [ 26 ] that it deserves further investigation. The aims of this study include: (i) investigate the fungal community composition in the reclaimed regions and discover the critical factor mediating fungal genera succession; (ii) explore the potential relationship between the fungi in rhizosphere soil and the root endophytic fungi; (iii) examine the functional microbial taxa of the soil fungal community in the reclaimed regions. Identifying fungal community composition in the rhizosphere of the plants survived in the reclaimed area will be helpful in understanding the development of the host plants and their co-evolution with symbionts. Furthermore, collecting the isolated root endophytic fungi and constructing a fungal resource bank can pave the way for the ecosystem reconstruction of the reclaimed area along the east China coast.",
"discussion": "3. Results and Discussion 3.1. Soil Physicochemical Properties A two-way ANOVA indicated that the physicochemical parameters (i.e., the EC values and contents of SN and SC) of the soil samples differed significantly ( p < 0.05) among the three sampling sites (i.e., sites F, H, and T) and between the two seasons (summer and winter 2020) ( Table 1 ). There was a significant difference in pH values of the soil samples between two seasons ( p < 0.05). A significant interaction term (i.e., sampling sites × seasons) for EC value and SC content ( p < 0.05) was also detected. In both seasons, the average percentages of TN and SC content in the rhizosphere soil samples were gradually decreased in the order of sites F, H, and T. The salinity level (EC value) in the rhizosphere soils was the highest at site H, while it was not so high as expected at site T. The investigated sites were located in the coastal area of Nantong, Jiangsu Province of China. There is a long history of the coastal mudflats’ reclamation in this area. Although the detailed reclamation ages of the sampling sites were unknown, the vertical distance from the location of the sites (in the order of sites F, H, and T) to the coastline could indicate a chronosequence reclamation of coastal mudflat. If the coastal areas were reclaimed at different stages, the impact of reclamation on soil properties could be varied with each other [ 2 , 27 ]. In general, the average percentages of soil SN and SC contents increased with increasing reclamation ages. At site T, the low EC value could be due to the desalination measures employed in the newly reclaimed land. 3.2. Soil Fungal Diversity and Community Structure In the soil samples, the species diversity was evaluated using the Shannon and Simpson index, while ACE and Chao1 indexes were used to reflect the species richness of fungi communities ( Table 2 ). In terms of the fungal diversity and richness in the investigated area, the diversity of fungi was highest at site F, followed by the sites H and T, and the index values were higher in winter compared with those in summer. In this study, six major phyla were identified in all collected soil samples, including the phyla Ascomycota , Basidiomycota , Mortierellomycota , Chytridiomycota , Rozellomycota , and Glomeromycota ( Figure 1 a). Except for the sample HS, the majority of OTUs in the investigated sites were belonging to phylum Ascomycota . In the sample HS, the highest abundance of phylum Basidiomycota was found, and its abundance was even higher than phylum Ascomycota . The findings that Ascomycota and Basidiomycota were predominant phyla in the sampling sites were consistent with an investigation on the reclaimed land in the Sanjiang plain, China [ 11 ]. FUNGuild was used to examine the fungal community from an ecological perspective, which could explore fungal trophic type and metabolic function characteristics rather than taxonomic identity [ 28 ]. Based on FunGuild ecological guild assignments ( Figure 1 b), the highest relative abundances of AM fungi and endophytic fungi were observed in samples TS (21.4%) and TW (7.46%), respectively. A large proportion of the OTUs at site F belonged to ECM fungi, and the relative abundances of ECM fungi were 47.0% and 83.5% for samples FW and FS, respectively. According to the location of the sampling sites, AM fungi and endophytic fungi were more frequent at the sites close to the sea embankment (i.e., TS and TW, respectively), while ECM fungi were dominant at the site located in the protection forest along the highway (i.e., FS). The difference in relative abundances of AM and ECM fungi among the investigated sites was probably due to the different vegetation types. In general, AM fungi were prevalent in herbaceous plants, while ECM prefer to form symbiosis with temperate woody plant species [ 29 ]. Additionally, AM fungi generally establish symbiotic associations with plants in the early stages of growth, while ECM fungi dominate in mature plants [ 9 , 21 , 30 ]. The establishment of a dense woody plant (i.e., 5-year-old willow) cover at site F could explain the dominance of ECM. At the site close to a farm along the coastal area (i.e., samples HS and HW), FUNGuild database analyses revealed a higher proportion of saprophytic fungi. The relative abundance of dung saprotroph, plant saprotroph, and wood saprotroph were 27.4%, 26.9%, and 27.1%, respectively, in the HS sample, indicating a disturbance from agricultural industry. It is noteworthy that plant pathogens accounted for 38.8% abundance of the fungal OTUs in sample TW, which cannot be ignored due to potential harmful effects on plant health. 3.3. Isolation of Endophytic Fungi from Plant Roots A total of 29 endophytic fungal taxa were isolated from the roots of the plant distributed in the investigated sites. Among them, genera Alternaria , Fusarium , Monosporascus , and Podospora were the four most common genera that were detected in the roots of many plant species ( Table S1 ). Alternaria was observed mostly in willow, while Fusarium , Monosporascus , and Podospora were mostly isolated in plant roots from site H. Besides the four genera, Aspergillus , Aureobasidium , Cladosporium , Clonostachys , Dactylonectria , Diaporthe , Mortierella , Penicillium , Sarocladium , Stemphylium , and Talaromyces were detected by both root isolation and soil ITS sequencing. According to Hardoim et al. [ 31 ], the endophytic community can be classified into facultative endophytes, obligate endophytes, and passenger endophytes. The aforementioned genera belonged to the facultative endophytes that can live inside the host plant and the soil habitats. Strains from genera Aspergillus , Penicillium , and Talaromyces were distributed in the plant roots with low frequency of appearance. However, they were widely distributed with relatively low abundance in soil samples according to the ITS sequencing results. These three genera might belong to passenger endophytes, which randomly enter the plant. The fungi that were isolated from the plant roots but were not detected in the soil samples via ITS sequencing ( Arthrobotrys , Botrytis , Macrophomina , Paecilomyces , Parengyodontium , Peniophora , Phaeomyces , Phialophora , Phoma , Pithomyces , and Rhizoctonia ) could be classified into obligate endophytes. Among the isolated fungi, genera Alternaria [ 32 ], Aspergillus [ 33 ], Cladosporium [ 34 ], Phoma [ 35 ], and Phialophora [ 36 ] contain strains classified as dark septate endophytes (DSE). Positive effects of DSE on the growth of host plants were reported, including the production of bioactive metabolites against pathogens [ 37 ], the synthesis of phytohormones, and the mineralization of organic N-containing compounds [ 38 ]. Additionally, strains from the genera Aspergillus , Aureobasidium , Cladosporium , Monosporascus , and Sarocladium have been reported as halophilic fungi [ 39 ]. Among them, genera Aspergillus , Aureobasidium , and Monosporascus were only isolated in plant roots from site H. Strains from the genera Mortierella , Penicillium , and Talaromyces have been recognized as plant growth promoting fungi [ 40 , 41 ]. The presence of the aforementioned fungal endophytes in the plant roots from the investigated sites indicates their important role in plant–fungus interactions. Nevertheless, previous literature showed that the majority of the genera listed ( Table S1 ) include pathogenic fungal strains [ 42 , 43 ]. Among them, genus Fusarium is widely accepted as one of the major groups of soil-borne root pathogenic fungi [ 44 ]. In vitro tests need to be employed to evaluate the pathogenicity and confirm the function of the isolated fungi. 3.4. Spatial and Temporal Variation of Soil Fungal Communities As the EC values and the average percentages of SN and SC contents in the soil samples differed significantly among the sampling sites and between the seasons, it was expected that these factors would be the key drivers for the distribution of fungi community. Indeed, the NMDS plot showed that the samples collected at different sites in two seasons were separated from each other, and the ANOSIM revealed a significant difference among them ( R = 0.983, p = 0.001) ( Figure 2 ). Seasonality is commonly regarded as an important parameter influencing plant-associated microbial communities [ 45 ]. In this study, the samples collected at different seasons were separated from each other. Besides providing essential nutrition (e.g., carbon and nitrogen source) for microbial growth, soil also affects the growth and distribution of microbe by its physicochemical properties (e.g., EC and pH) [ 10 ]. The samples collected at site F (distributed at the bottom right of the coordinate plane) were separated from those sampled at sites H and T. At site F, a dense tree canopy and relatively lower EC value and SN and SC contents were observed. The preferable environmental conditions could promote the dominance of ECM fungi, which explain the clustering pattern of the samples in the NMDS plot. The random forest analysis was performed to identify the signature microbiota in response to the variation of the investigated sites and seasons, respectively. It is expected that random-forest analysis could be used to classify the fungi community as mirroring sampling sites and seasons, suggesting essential factors shaping soil fungi community. Figure 3 shows the result of the random forest classification preformed using genera-level composition data of the rhizosphere soil samples. The abundance of fungi genera with the 30 highest Gini scores in the random forest analysis is presented in the heatmap. Among these selected genera, four dominant genera ( Tuber , Peziza , Tomentella , and Geopora ) belonging to ECM fungi were the potential contributors to identify the sites at which rhizosphere soil samples were collected ( Figure 3 a). ECM fungi were abundant in the FS ( Tuber ) and HS ( Tomentella ) samples, while it was seldom observed at site T. The site T located in a young reclaimed region, in which the C% and N% contents of the soil samples were the lowest. Five genera ( Pichia , Barnettozyma , Candida , Nigrospora , and Coprinopsis ) with high abundance in winter contributed to distinguishing the winter samples from the summer samples ( Figure 3 b), among which three genera ( Pichia , Barnettozyma , and Candida ) were identified as yeast, and their dominance was observed in winter. LEfSe analysis (LDA threshold 4.0) was conducted to identify and compare unique fungal taxa in the soil samples with statistical difference among different groups ( Figure 4 a). For samples from site F, fungi that were differentially abundant include genera Tuber and Peziza in the FS sample and genera Geopora and Tetracladium in the FW sample. At site H, biomarkers mainly comprised of genera Podospora , Coprinellus and Tomentella in the sample HS and genera Preussia , Glomus and Candida in the sample HW. At site T, genus Zopfiella was the biomarker in the TS sample, while the biomarkers in the TW sample include genera Nigrospora , Coniochaeta , Alternaria , Stemphylium , and Pichia . A network interface was constructed to show the topological and taxonomic characteristics of the fungal co-occurrence patterns in each region ( Figure 4 b–d). The densities of the co-occurrence network at sites F, H, and T were 0.189, 0.241, and 0.409, respectively. These results suggested that the fungal network at site T was more connected than that at sites F and H. Nodes with high degree values and centrality metrics (betweenness or closeness centrality) were recognized as keystone species in the co-occurrence network. The keystone species at sites F, H, and T were identified, which could play critical roles in the co-occurrence network. Genera Clitopilus and Aspergillus were identified as the keystone genera at site F. At site H, keystone genera including Rhizophagus , Nigrospora , Pichia , and Kazachstania were identified. Genera Papiliotrema , Phanerochaete , Entrophospora , and Zopfiella were recognized as the keystone taxa at site T. In the FS sample, the relative abundances of genera Tuber and Peziza were the highest. Genus Geopora was detected in most of the investigated sites, with the highest abundance in the sample FW. Genera Tuber , Peziza , and Geopora belong to order Pezizales . Many species in order Pezizales were recognized as ECM fungi that grow in symbiosis with the plant roots [ 46 ]. Both RF and LEfSe analyses indicated that genus Tuber had a preference for its distribution among the three sites. At site F, the higher soil SC and SN contents and lower EC values compared with the other two sites could be responsible for the dominance of genus Tuber . As the keystone genera at site F, genus Clitopilus was only detected in sample FS, while genus Aspergillus was identified in all samples. Genus Clitopilus belongs to family Entolomataceae , in which many strains can form ectomycorrhizae with a wide range of tree species; being considered as a common saprotrophic genus, it was reported to increase plant growth via facilitated potassium uptake [ 47 ]. Genus Aspergillus was widely distributed in all the investigated sites. Liang et al. [ 48 ] reported a ribosomal protein from halophilic strain ( Aspergillus glaucus ) that could confer salt tolerance in heterologous organisms. It is noteworthy that genus Aspergillus was recognized as a culturable fungal endophyte, and it was only isolated from the plant roots at site H. Genus Tetracladium was mainly identified at site F, with a higher abundance in the sample FW than that in the sample FS. Sati and Pant [ 49 ] reported a Tetracladium strain isolated from healthy roots of Berberris vulgaris growing in a riparian area had the potential to solubilize phosphate. As an endophytic fungus, genus Tetracladium could be capable of promoting plant growth at site F. In the sample HS, the highest abundance of genus Podospora was detected. Genus Coprinellus were dominant in the sample HS, while they were seldom observed in other samples. The abundance of genus Preussia in winter was higher than that in summer, and its highest abundance were observed in the sample HW. Genera Podospora , Coprinellus , and Preussia include coprophilous species inhabiting the dung of various herbivores [ 50 , 51 , 52 ]. The site H is close to a poultry farm, which might probably release poultry excrement to the surrounding environment. It could explain why the coprophilous species were dominant at site H. In the sample HS, genus Tomentella were dominant. It was commonly found as an ectomycorrhizal partner of many trees, such as willow, birch, or alder, with the tendency to dominate in unfavorable environmental conditions (e.g., salinity stress) [ 46 ]. The dominance of genera Glomus and Rhizophagus (belonging to AM fungi) was observed in HW and HS, respectively. Glomus species was reported to increase the number of fruits and yield in papaya plants [ 53 ], while genus Rhizophagus could stabilize soil aggregates, enhance plant nutrition, and improve plant growth [ 16 ]. AM fungi form a mutualistic association with the roots of 90% of the terrestrial plants, and the symbiosis between AM fungi and plant roots is a well-recognized beneficial interaction occurring in soil [ 54 ]. They were reported to be capable of improving rooting, enhancing plant nutrition, favoring nutrient renovation, and promoting tolerance to biotic and abiotic stress [ 53 ]. Unfortunately, the use of AM fungi as an inoculant on a large scale is not yet widely used because of several difficulties in obtaining a large amount of inoculum (e.g., low growth and high competition with native AM fungi) [ 16 ]. Genera Candida , Pichia , and Kazachstania were described as ascomycetous yeast species [ 55 , 56 ]. Yeasts are particularly suitable as biocontrol agents against several kinds of rot fungi because their activity does not usually rely on the production of toxic metabolites, but rather on their ability to compete with pathogens for space and nutrients [ 57 ]. Some species belonging to genus Candida were reported to stimulate rice seedling growth [ 58 ]. Their dominance at site H (especially in winter) could be beneficial for the plants to defend against pathogens and survive under the salinity stress. However, a Pichia strain ( Pichia fermentans ) with dual activity (an effective biocontrol agent and an aggressive pathogen) was reported [ 55 ]. To avoid unpredictable effects, it needs a thorough risk analysis before the application of yeast species as biocontrol agents. Many strains in the genus Nigrospora are characterized as plant endophytic micro-organisms. This genus produces a broad range of bioactive secondary metabolites, which could be used to inhibit the growth of plant pathogenic fungi [ 59 ]. In the sample TS, the dominance of genus Entrophospora was observed. It was classified as an AM fungus [ 53 ], which could exchange nutrients gathered from soil for carbon compounds provided by plants. They are important microbial symbionts for plants, especially when soil total phosphorus (TP) and TN are limiting [ 60 ]. Although soil TP content was not measured in this study, the lower SC and SN contents at site T indicated a possible deficit in TP contents. The appearance of genus Entrophospora might favor plant growth at site T. Genera Nigrospora and Coniochaeta were observed in winter, especially in the sample TW. Genus Zopfiella was observed with the highest abundance in the sample TS. Some strains of the genera Nigrospora , Coniochaeta , and Zopfiella are characterized as plant endophytic micro-organisms, and they are capable of promoting the stress resistance of the host plant and producing antifungal compounds [ 61 , 62 , 63 ]. Genera Alternaria and Stemphylium include both plant-pathogenic and saprophytic species [ 64 , 65 ]. Their dominance in winter, especially in the sample TW, indicated possible damage or decay of plants after establishing pathogenic, saprotrophic, or endophytic relationships with the plant hosts. Additionally, genus Phanerochaete that was recognized as a saprotrophic fungus appeared in the sample TS. It is capable of degrading lignin and mineralizing a wide variety of priority aromatic pollutants [ 66 ]. 3.5. The Relationship between Fungi Genera and Soil Physical-Chemical Parameters The relationships between the fungi (top 10 dominant genera, biomarkers, and keystone taxa) and selected environmental factors were explored by CCA ( Figure 5 ). The two CCA axes explained a substantial proportion of the variation (88.8%) in the fungi–environment relationship. The abundance of Tuber and Geopora (belonging to ECM fungi) and Peziza (containing both ECM and saprotrophic fungi) was positively related with SC content and negatively related with EC level. High soil salinity was generally recognized to have a negative impact on ECM associations [ 46 ]. Genera Tuber and Geopora were dominant at site F, which was characterized by high SC and SN contents and low soil EC value. ECM fungi are capable of excreting oxidative and hydrolytic enzymes to break down soil organic matter and liberate unavailable nutrients [ 8 , 67 ]. Their predominance (e.g., Tuber and Geopora ) at site F could be helpful to transfer nutrients from organic matter to their host tree. There are also some ECM fungi that are highly adapted to saline conditions [ 10 ], and their appearance (e.g., Tomentella ) at site H (with relatively high values of soil EC) was observed. Additionally, the positive relationship between the abundance of Tetracladium and SC content was also observed. Some strain ( Tetracladium setigerum ) belonging to Tetracladium was reported as a phosphate solubilization fungi, which has the potential to solubilize various source of phosphates through the production of phytases and phosphatases enzymes and the organic acids [ 49 ]. The dominance of genus Tetracladium at site F could provide sufficient P for the good development of shelter forest. Soil pH value has a positive influence on the distribution of the genera Zopfiella , Tomentella , and Coprinellus that were dominant in summer during the investigation. Zhao et al. [ 68 ] reported that genera Zopfiella was dominated in the soil with the highest pH values after disinfestation treatment using bean dregs. During the investigation on fungi in protected coastal Salix repens communities in the Netherlands, Geml et al. [ 60 ] reported that Tomentella showed a strong preference for alkaline soils. Liu et al. [ 69 ] indicated that Coprinellus is preferably dwelling alkaline environments. Additionally, several yeasts ( Candida , Pichia , and Barnettozyma ) were negatively related with pH values. Birkhofer et al. [ 70 ] investigated the soil yeast community composition and abundance in different land use types in Germany, and reported that yeasts were highly abundant in the region with a low soil pH and high soil moisture. There was a positive relationship between the EC value and the abundance of genera Stemphylium and Glomus . Gonçalves et al. [ 13 ] isolated Stemphylium from the field-collected Salicornia , and found that the growth of Stemphylium sp. was unaffected by the presence of NaCl (200–800 mM) in the growth medium. Furthermore, the inoculation of Salicornia with the isolated Stemphylium strain positively influenced total biomass production and N concentration in roots in salinity conditions (150 mM NaCl). It is already known that AM fungi (e.g., Glomus ) widely exist in salt-affected soils, and they were reported to play a vital role in mitigating the adverse effects of salinity on plants by the processes including the up-regulation of the antioxidant system, the modulation of the biosynthesis of osmoprotectants, and the compartmentalization of excessive toxic ions into the vacuole [ 71 , 72 , 73 ]."
} | 7,766 |
34765169 | PMC8571575 | pmc | 2,713 | {
"abstract": "Abstract Across the globe, ecological communities are confronted with multiple global environmental change drivers, and they are responding in complex ways ranging from behavioral, physiological, and morphological changes within populations to changes in community composition and food web structure with consequences for ecosystem functioning. A better understanding of global change‐induced alterations of multitrophic biodiversity and the ecosystem‐level responses in terrestrial ecosystems requires holistic and integrative experimental approaches to manipulate and study complex communities and processes above and below the ground. We argue that mesocosm experiments fill a critical gap in this context, especially when based on ecological theory and coupled with microcosm experiments, field experiments, and observational studies of macroecological patterns. We describe the design and specifications of a novel terrestrial mesocosm facility, the iDiv Ecotron. It was developed to allow the setup and maintenance of complex communities and the manipulation of several abiotic factors in a near‐natural way, while simultaneously measuring multiple ecosystem functions. To demonstrate the capabilities of the facility, we provide a case study. This study shows that changes in aboveground multitrophic interactions caused by decreased predator densities can have cascading effects on the composition of belowground communities. The iDiv Ecotrons technical features, which allow for the assembly of an endless spectrum of ecosystem components, create the opportunity for collaboration among researchers with an equally broad spectrum of expertise. In the last part, we outline some of such components that will be implemented in future ecological experiments to be realized in the iDiv Ecotron.",
"introduction": "1 INTRODUCTION Ecosystems are threatened by a multitude of environmental change drivers (Díaz et al., 2019 ; Maxwell et al., 2016 ; Murphy & Romanuk, 2014 ; Newbold et al., 2015 ; Pereira et al., 2012 ). Over the last few decades, there has been an explosion of studies examining changes in ecological communities and environmental conditions (Hines et al., 2019 ; Liu et al., 2011 ; Stork & Astrin, 2014 ). The desire to draw generalizable conclusions from these studies led to a period of synthesis, during which information from individual studies was compiled allowing for quantitative evaluation of the variation in ecological changes across systems (Gurevitch et al., 1992 ; Halpern et al., 2020 ; Hillebrand et al., 2020 ). Such comprehensive and quantitative synthesis studies enabled researchers to identify generalizable patterns in biodiversity (Calatayud et al., 2020 ), trends in biodiversity change (Blowes et al., 2019 ; Dornelas et al., 2014 ), and relationships between biodiversity and ecosystem functioning (e.g., Cardinale et al., 2012 ; Gessner et al., 2010 ; Lefcheck et al., 2015 ; Soliveres et al., 2016 ). These high‐impact synthesis studies can also serve as a roadmap for designing future experiments, as they help to identify important knowledge gaps which need to be filled in order to better understand the functioning of ecosystems and predict the consequences of climate change. We have limited empirical evidence for at least three key aspects of environmental changes in ecosystems and communities that draw a roadmap for future research. First, there are limited numbers of ecosystem response variables that have been consistently studied across systems. For example, the most commonly reported response variables are primary production and decomposition (Cardinale et al., 2006 ; Schmidt, Auge, et al., 2015 ; Schmidt, John, et al., 2015 ). However, the few existing multitrophic biodiversity studies indicate that the interactions of higher trophic levels may be particularly important for multiple ecosystem functions (Hines, van der Putten, et al., 2015 ; Lefcheck et al., 2015 ; Naeem et al., 1994 ; Soliveres et al., 2016 ) and that especially these species might be very vulnerable to environmental changes (Hines, Eisenhauer, et al., 2015 ; Voigt et al., 2003 ). Second, studies tend to investigate limited types of mechanisms and processes underlying changes in biodiversity, ecosystem functioning, and the relationship between the two (Hillebrand et al., 2020 ). That is, while there is strong emphasis on the effects of global change drivers on changes in species richness (Tilman & Downing, 1994 ; Harpole et al., 2016 , 1994 ; Seabloom et al., 2021 , but see Dornelas et al., 2014 ; Vellend et al., 2013 ), there is less known about the ecosystem consequences of changes in behavior (Cordero‐Rivera, 2017 ; Wilson et al., 2020 ) and community composition (Hillebrand et al., 2018 ; Spaak et al., 2017 ) of species that persist in communities. Third, although ecosystems are confronted with complex cocktails of global change drivers (Bowler et al., 2020 ), so far only a limited number of their types and combinations have been studied in realistic experiments (Rineau et al., 2019 ; Rillig et al., 2019 , but see Schädler et al., 2019 ; Korell et al., 2020 ). Especially with regard to climate change, understanding interactions between different environmental variables such as temperature and precipitation, land use or biodiversity on ecosystem functioning is essential to make predictions for future ecosystem developments and the potential consequences for society (Roy et al., 2017 ). To address our current knowledge gaps, we need experiments which can simultaneously manipulate and measure different global change drivers (Vanderkelen et al., 2020 ) and investigate their impacts on a wide range of functional groups and trophic levels of organisms (De Boeck et al., 2020 ; Komatsu et al., 2019 ; Korell et al., 2020 ). Combining such “meta‐scale” studies with laboratory and field studies, especially large‐scale climate change experiments (like Schädler et al., 2019 ), provides the opportunity to understand the complex patterns of biodiversity–ecosystem function relationships and their responses to environmental changes as well as the underlying processes that operate across organizational levels of life (cell‐individual‐population‐community‐ecosystem; Ferlian et al., 2018 ). Here, we introduce the iDiv Ecotron platform (iDiv stands for the German Centre for Integrative Biodiversity Research Halle‐Jena‐Leipzig in Germany). This platform is a highly flexible experimental infrastructure that was specifically designed to perform multitrophic biodiversity experiments in terrestrial ecosystems (Eisenhauer & Türke, 2018 ). In the following sections, we describe the iDiv Ecotron specifications and functioning, we highlight a case study experiment as an application possibility, and we provide an outlook on the potential contributions of future ecotron experiments. The concept of the iDiv Ecotron was to create a facility which allows the setup and maintenance of complex communities and manipulation of several abiotic factors in a near‐natural way, while simultaneously measuring multiple ecosystem functions. Environmental conditions, such as humidity, nutrient supply, light, and precipitation, can be fully controlled and monitored (for details see Appendix 1 ), which allows the iDiv Ecotron to be used for the simulation of multiple abiotic scenarios together with scenarios of above‐belowground community change. The iDiv Ecotron offers the possibility to study a wide range of ecosystem responses, including above‐belowground interactions of plants, microbes, and invertebrates. The platform can accommodate stand‐alone experiments and also provides complementary information to small‐ and large‐scale experiments (lab‐ecotron‐field). Therefore, the iDiv Ecotron links investigations at multiple experimental and spatial scales and serves as a key component for collaborations between researchers from different disciplines to conduct interdisciplinary studies on the drivers of, and relationship between, biodiversity and ecosystem functioning. Consequently, this platform is likely to provide novel insights into ecosystem responses to global change.",
"discussion": "3.4 Discussion In contrast to our expectations, beans did not generally benefit from growing in herb communities, while being suppressed by more dominant nitrophilous grasses (Eisenhauer & Scheu, 2008 ). We observed opposing effects for the two grass species and for the two herb species on bean biomass. Among the four neighboring plant species, H . lanatus produced by far the highest amount of aboveground plant biomass (139.5 g) at the end of the experiment compared to the other three species ( F . pratensis : 92.1 g, C . jacea : 51.1 g, B . perennis : 5.3 g), and, as graminoid species typically produce a dense and large root system, we speculate that also root biomass was highest (not assessed in this study). Thus, both enhanced aboveground light competition and belowground competition for resources may have contributed to an overall advantage in resource acquisition over the bean, causing low bean biomass. Indeed, it has been often confirmed that grasses are stronger competitors compared to herbaceous species (Del‐Val & Crawley, 2005 ; Tilman, 1982 ). Moreover, another potential explanation for the patterns found in our study may be that in patches of low biomass, for example, in B . perennis patches, the habitat structure for predators was comparably low leading to a migration to more favorable habitat structures. This effect may have cascaded to lower trophic levels increasing abundances of herbivores and decreasing plant performance (Romero & Koricheva, 2011 ). The importance of such non‐trophic interactions based on habitat structure has been often highlighted (Kalinkat et al., 2013 ; Majdi et al., 2014 ). Our results confirm the often found tritrophic relationships between predators, herbivores, and primary producers, where predators, in our case ladybirds, exert a top‐down control on aphid abundances which, in turn, have a top‐down effect on the bean (Romero & Koricheva, 2011 ). Surprisingly, the effects of plant neighbor species on aphid abundances were opposing for communities without and with belowground invertebrates. These findings highlight the significance of aboveground–belowground interactions and show that decomposers can influence aboveground multitrophic interactions by altering the competition between plants (Wardle et al., 2004 ). Moreover, we found that trickle‐down effects of aboveground invertebrates on soil food webs (here represented by soil nematode species richness) depend on plant community composition. This finding suggests that the competitive environment of a focal plant can alter its effects on soil community composition, potentially through changes in the amount and quality of plant‐derived resources entering the soil (Hooper et al., 2000 ). Taken together, our study shows distinct interaction effects between aboveground and belowground invertebrate communities on multitrophic interactions and community composition in the sub‐compartments. These changes are likely to alter how communities function, which may have subsequent feedback effects on nutrient cycling and community composition. The results of our study highlight the need for infrastructures that allow to manipulate food webs of high complexity, which can hardly be realized experimentally under field or simplified laboratory conditions (Beyers & Odum, 1993 ), and at the same time, taking advantage of measuring and controlling a large fraction of other non‐targeted parameters including environmental conditions."
} | 2,915 |
34301053 | PMC8309462 | pmc | 2,714 | {
"abstract": "Self-healing materials have been developed since the 1990s and are currently used in various applications. Their performance in extreme environments and their mechanical properties have become a topic of research interest. Herein, we discuss cutting-edge self-healing technologies for hard materials and their expected healing processes. The progress that has been made, including advances in and applications of novel self-healing fiber-reinforced plastic composites, concrete, and metal materials is summarized. This perspective focuses on research at the frontier of self-healing structural materials.",
"conclusion": "6. Concluding Remarks Over the past 30 years, new materials and successful approaches that expand the field of self-healing applications have been explored. In the next few decades, the performance of self-healing materials under extreme conditions must be considered. A viable self-healing technology must be developed to resolve safety issues that inevitably arise for large structures. FRP composites, concrete, and metals, which are representative building materials, are more important in this respect. We should devote more attention to endowing these hard materials with the ability to undergo self-healing, which was considered impossible until now. Thus, self-healing technologies make it possible to contribute to the protection of lives and properties by using new materials and groundbreaking approaches.",
"introduction": "1. Introduction Nature-inspired self-repairing strategies have been explored in biomimetic engineering with the aim of restoring damaged materials. A variety of types of infrastructure must be protected as human activity expands from being terrestrial-based into such environments as submarine and space. Various self-healing methods have recently been developed and tested. These bioinspired engineered materials, i.e., materials that “self-heal” after external damage, have been studied since the early 1990s [ 1 , 2 ]. The damage to engineering materials is mostly repaired via the process of systematic transport and the polymerization of healing materials in the damaged area. For this reason, the majority of early studies on self-healing materials were conducted using soft materials [ 2 ]. Polymeric materials were found to be more manageable when encapsulated in microcapsules, prepared as nanofibers, or used in a reversible form, such as a thermoplastic or supramolecular materials; in addition, these materials were viable and did not require external stimulation to initiate the healing process [ 3 , 4 ]. However, from the perspective of mechanical strength, healing materials are inevitably soft and weak. To date, numerous self-healing materials have been reported; however, most of these are mechanically soft and weak and are thus not appropriate for practical applications involving structural materials. Therefore, the next goal in the development of self-healing materials is to devise materials with sufficient mechanical strength for use as structural materials. For example, the walls of a nuclear power plant or an airplane frame must possess sufficient strength under extreme conditions, including long-term fatigue conditions. Thus, healing materials for use in these applications are also required to be strong, meaning that traditional polymer-based materials are inherently unsuitable. Moreover, large structures are difficult and expensive to maintain, and detecting and repairing defects is challenging from a technical perspective. Solutions that are realistic from both technical and material points of view are required to identify and repair structural problems that may arise in thick reactor walls or fuselages of aircrafts during flight. To this end, new approaches have been investigated to repair hard materials, such as concrete composites [ 5 , 6 , 7 , 8 , 9 ], carbon fiber-reinforced plastic (CFRP) composites [ 10 , 11 , 12 ], steel, and aluminum [ 13 , 14 , 15 , 16 ]. Owing to the wide range of applications for self-healing materials, novel strategies and new materials are being actively developed. The challenges that have been overcome to integrate these self-healing materials into full-fledged structures on real construction sites are highlighted in this review. Furthermore, the effects of the various self-healing techniques on panel motion under load and the techniques used to heal the cracks that develop are discussed. This study surveys cutting-edge self-healing approaches, with a particular focus on structural materials. There is an adage that states “a small crack breaks a big dam.” High-rise buildings and long bridges are susceptible to accidents and disasters every hour of every day; therefore, the focus of self-healing technology must necessarily shift toward such large structures. In the past, the main purpose behind the use of self-healing materials was to repair scratches and prevent rust; however, the research community concerned with self-healing materials should now focus on preventing accidents caused by the collapse of bridges or the rupture of aircraft structures. Once motivated by the desire to repair physical damage, self-healing technology is predicted to be used in future to safeguard the world’s infrastructure and protect lives."
} | 1,311 |
35218086 | PMC9311646 | pmc | 2,715 | {
"abstract": "Abstract Coral reefs are in global decline due to climate change and anthropogenic influences (Hughes et al., Conservation Biology , 27: 261–269, 2013). Near coastal cities or other densely populated areas, coral reefs face a range of additional challenges. While considerable progress has been made in understanding coral responses to acute individual stressors (Dominoni et al., Nature Ecology & Evolution , 4: 502–511, 2020), the impacts of chronic exposure to varying combinations of sensory pollutants are largely unknown. To investigate the impacts of urban proximity on corals, we conducted a year‐long in‐natura study—incorporating sampling at diel, monthly, and seasonal time points—in which we compared corals from an urban area to corals from a proximal non‐urban area. Here we reveal that despite appearing relatively healthy, natural biorhythms and environmental sensory systems were extensively disturbed in corals from the urban environment. Transcriptomic data indicated poor symbiont performance, disturbance to gametogenic cycles, and loss or shifted seasonality of vital biological processes. Altered seasonality patterns were also observed in the microbiomes of the urban coral population, signifying the impact of urbanization on the holobiont, rather than the coral host alone. These results should raise alarm regarding the largely unknown long‐term impacts of sensory pollution on the resilience and survival of coral reefs close to coastal communities.",
"introduction": "1 INTRODUCTION Human activities have impaired normal ecosystem functioning across most of the Earth's surface (Dominoni et al., 2020 ), highlighted by the loss of biodiversity, even in remote areas (Motesharrei et al., 2016 ; Vitousek et al., 1997 ). Coastal ecosystems proximal to urban centers in the tropics are likely to be among the most vulnerable, as they face not only climate change but unpredictable and fluctuating nutrient (nitrogen and phosphorus) and xenobiotic (hormones and other organic contaminants) local pollution, besides chronic exposure to noise and light pollution (Duarte et al., 2021 ; Heery et al., 2018 ). Thus, urbanization of coastal areas near coral reefs is a global issue and is gaining momentum; not only has there been continuous growth of cities such as Jakarta, Singapore, and Hong Kong, but also major new developments have occurred or are planned that are likely to directly impact coral reefs in the near future. The population of the Chinese coastal city of Shenzhen has grown from <1 million in 1990 to >12.5 million in 2021, further expansion being predicted ( https://www.macrotrends.net/cities/20667/shenzhen/population ), and the rapid development of the Jeddah Corniche on the Red Sea coast of Saudi Arabia is of particular concern given the unique nature of the adjacent reef system (Kleinhaus et al., 2020 ). Human activity has impacted animal and marine habitats in almost every conceivable way. This includes urbanization, buildings, lights at night, chemicals from industry or farming, tourism etc. which are known as anthropogenic or sensory pollutants. Sensory pollutants can mask environmental cues, interfere with the cellular processing of information, or alter cue perception leading to distracted responses by the organism (Dominoni et al., 2020 ). Therefore, formerly dependable cues may no longer be reliable in environments altered by humans. These anthropogenic stimuli can decrease animal survival and reproductive success and may ultimately alter populations and ecological communities. To understand and mitigate the effect of these stimuli, it is crucial to study the impact underlying the sensory reception of these pollutants, on marine habitats like coral reefs (Halfwerk & Slabbekoorn, 2015 ). Although light pollution has been shown to disrupt the timing of coral spawning (Ayalon et al., 2020 ; Loya, 2004 ), with potentially devastating consequences, the broader impacts of urbanization on corals are unknown, as are the underlying molecular mechanisms. Coral reefs, which support the highest concentrations of marine biodiversity and provide essential ecosystem services to millions of people, are among the most impacted coastal ecosystems in response to human act (Hughes et al., 2013 ). One consequence of increased human activity near coastlines is that the community structure of many reefs has changed—in extreme cases to dominance by algae or other taxa rather than corals (Guest et al., 2016 ; McManus & Polsenberg, 2004 ). Coral reefs exposed to anthropogenic stress typically exhibit lower structural complexity, are dominated by “stress‐resistant,” generalist coral species, and exhibit decreased coral cover, all of which compromise ecosystem function (Brandl et al., 2019 ; Heery et al., 2018 ). Even though some corals have survived the selective pressures of “city life,” growing for years under chronic anthropogenic stress, they can face sudden fluctuations in pollutant levels to which corals are not adapted, thus having little chance to acclimatize. Urban corals can be distinguished from corals in otherwise degraded reef ecosystems by differences in physiology, cellular processes, and growth characteristics (Heery et al., 2018 ; Nyström et al., 2000 ). Here we determined the effects of proximity to an urban environment on coral biorhythms over diel, monthly, and annual cycles, resolved in‐natura, using nitrogen and carbon stable isotopes, physiological monitoring, profiling gene expression in the coral host, and analyses of microbiome assemblage patterns. Our comparison between colonies from an urban area to a non‐urban area (Figure S1a ) reveals the extent to which coral biorhythms are disrupted and modified by anthropogenic influences. Despite being sessile organisms lacking specialized sensory organs, corals can sense chemical or physical environmental cues (Armoza‐Zvuloni et al., 2016 ; Paul & Puglisi, 2004 ) via complex repertoires of receptors that respond to the external cues by triggering signal‐transduction pathways to initiate specific biological processes (Armoza‐Zvuloni et al., 2016 ; Levy et al., 2007 ). The ability of these sensory systems to detect natural environmental cycles has been refined over millions of years of evolution. In contrast, anthropogenic influences are recent and can interfere with these mechanisms through which corals synchronize with the environment (Ayalon et al., 2019 , 2020 ; Loya, 2004 ; Rosenberg, Doniger, & Levy, 2019 ; Shlesinger & Loya, 2019 ). For example, recent works (Ayalon et al., 2020 ; Shlesinger & Loya, 2019 ) clearly implicated light pollution in the impairment of coral gametogenesis and spawning synchrony. This study focuses on the common coral species Acropora eurystoma from the Gulf of Aqaba in the northern Red Sea to determine the impacts of urbanization on biological rhythms in corals. The fringing reefs in the Gulf of Aqaba are located at an unusually close distance, a few meters, from the shore (Loya, 2004 ). They are, therefore, particularly exposed to the impact of the surrounding urban environment (Ayalon et al., 2019 ; Loya, 1972 , 2004 ; Loya et al., 2004 ; Rosenberg, Doniger, & Levy, 2019 ; Shlesinger & Loya, 2019 ). The General Circulation Model in the Gulf of Aqaba points on north to south current while a later model claim that coral larvae connectivity between “source” and “sink” in the Gulf can range between 9 ± 13 km (Berenshtein, 2018 ). Extrapolating to a global scale, the impacts of chronic exposure to sensory pollutants will likely further decrease the resilience of coral reefs but are not considered in current projections of the future of coastal coral reefs.",
"discussion": "4 DISCUSSION Our year‐long in‐natura experiment comparing corals from urban and non‐urban areas indicate that normal diel, monthly, and annual biorhythms of corals are considerably disrupted by the urban conditions in the Gulf of Aqaba. The Gulf of Aqaba is an ideal study site because corals exhibit remarkable thermal tolerance, facilitating the determination of the effects of urbanization, without having to take into consideration the confounding impact of ongoing and increasing thermal stress by ocean warming (Savary et al., 2021 ; Voolstra et al., 2021 ). At the same time, the Gulf of Aqaba is a very small and constrained water body, which made the incorporation of more study sites (ideally, replicated urbanized and non‐urbanized areas) impossible. Rather, the urban and non‐urban study areas were only 6 km apart. It is noteworthy, that this spatial proximity renders a potential influence of genetic isolation or divergence that could arguably contribute to the observed differences unlikely. In this regard, the apparent chronic disruption of coral physiology is particularly troubling, because it directly aligns with coastal urbanization, that is, the presence of humans in the study area. Despite the presence of large colonies of A. eurystoma of healthy appearance at similar densities in both areas, the comprehensive analyses conducted here suggest that major differences are apparent at physiological and metabolic level (Figure 6 ). The enrichment of genes involved in N fixation and photosynthesis in the non‐urban samples suggest that diazothrophs and phototrophs were more abundant at this area than in the urban one. The signature of “heavier” carbon and nitrogen isotope ratio in both coral tissue and symbiont samples from the urban area, together with the abundance of ABC transporter genes, supports our notion that there is an anthropogenic local disturbance of eutrophication (Figure S1b ). This probably effects the metabolism and photosynthesis performances which impact the isotopic fractionation and photopigment synthesis (Ferrier‐Pagès & Leal, 2019 ; Muscatine, 1994 ; Muscatine & Cernichiari, 1969 ; Muscatine et al., 1989 ; Rädecker et al., 2015 ; Wall et al., 2019 ; Figure 1b,c ; Figures S1b and S4d–f ). The variations in δ 13 C values between the urban and non‐urban samples could indicate changes in the biomass composition (protein: lipid: carbohydrate ratios) of corals from each sampling area. In addition, the measured δ 15 N values could reflect on the temporal variability in nitrogen sources in both areas affecting the symbiont nitrogen demands (Levy et al., 2010 ; Wall et al., 2019 ). FIGURE 6 Conceptual illustration of urbanization effects on coral reefs. Environmental conditions are represented by arrows pointed from the outside towards the coral. Biological outputs are represented by arrows pointed from the coral towards to outside Artificial light at night (ALAN) is likely to play a significant role in disrupting the normal diel, lunar, and annual cycles of physiology and gene expression. Direct effects of ALAN on the circadian clock have been documented in a range of organisms, including corals (Ayalon et al., 2019 ; Davis et al., 2001 ; Rich & Longcore, 2013 ; Rosenberg, Doniger, Harii, et al., 2019 ; Rosenberg, Doniger, & Levy, 2019 ). The disrupted lunar rhythms of gene expression in the urban corals documented here are consistent with ALAN interfering with moonlight‐sensing systems as observed in other marine invertebrates (Kronfeld‐Schor et al., 2013 ). As seen by the monthly cycle of enriched processes in the non‐urban corals, most processes respond to the full moon where illumination is the strongest but, even at the full moon, the intensity of artificial light penetrates the water column at the urban area is greater (Tamir et al., 2017 ). The monthly enriched processes found in the non‐urban corals represent features of cellular organization, tightly connected to the actin and myosin filaments, enriched during the full moon (correlated to previously published work (Rosenberg et al., 2017 ). The reproductive processes, which are all aligned with the moon phase, are also abolished in the urban corals. Our results support the notion of moon light regulating expression patterns of clock genes in corals (Brady et al., 2016 ; Hoadley et al., 2011 ; Levy et al., 2007 ; Reitzel et al., 2013 ; Rosenberg, Doniger, Harii, et al., 2019 ) and the destructive effect of artificial light, over riding the moon light, causing delayed gametogenesis and loss of synchrony in gamete release as observed in many coral species across the globe (Ayalon et al., 2020 ; Jokiel et al., 1985 ; Kaniewska et al., 2015 ; Van Woesik et al., 2006 ). Our molecular data for these two cycles (diel and moon phase) emphasize the ability of light pollution to override the natural light/dark cycle and moon light, as well as masking biological processes associated with these cycles effecting corals in urban areas. Additionally, recent work (Lin et al., 2021 ) showed that dim light during the night suppressed spawning in the coral Dipsastraea speciosa . Importantly, this later study showed that the period of darkness between sunset and moonrise is essential to trigger synchronized mass spawning. Normal seasonal rhythms of gene expression were also severely disrupted in the urban area, as the GO term enrichment analyses clearly indicate (Figure 4a ). Seasonal GO term enrichment of sexual reproductive and cell cycle processes was only observed at the non‐urban area or in some cases, seasonally shifted between the two areas (e.g., RNA splicing). In the GO term analysis, the few processes that were only enriched in the urban area most likely reflect stress responses. Those included heme biosynthesis which is a biomarker for evaluating contamination in marine environments (Bogovski et al., 1998 ; Hongo et al., 2017 ), cell redox homeostasis which refers to the capacity of cells to continuously deal with challenges brought by different stressors, metabolic or environmental (Ursini et al., 2016 ), and ATP metabolic process, which is known indicator for the presence of contaminants (Kroll, 2009 ). Diel and seasonal light regimes drive respiration and primary production, and in turn, are fundamental determinants of nutrient transformation and heterotrophic microbial diversity. We hypothesize that ALAN will affect these ecological processes and, therefore, diminish the microbial biorhythmicity as reefs are impacted by anthropogenic perturbation. Correspondingly, we found that the biorhythmicity of coral‐associated bacteria was strongly dependent not only on the season but also on the sampling area (i.e., reefs with different degrees of anthropogenic perturbation; Roder et al., 2015 ; Ziegler et al., 2019 ). Moreover, microbial community composition differed between moon phases in urban and non‐urban areas. The bacterial taxa that significantly responded to moon phases were common to both areas but fluctuated asynchronously between new and full moon at different seasons. In addition, lower bacterial beta diversity across seasons in the urban area indicated a less pronounced seasonality compared to the non‐urban area. Our results suggest that anthropogenic‐derived ALAN affects the biorhythmicity of coral microbiomes due to the potential demotion of light as an essential seasonal cue, as shown for microbial communities in sediment over time (Hölker et al., 2015 ). Coral‐associated bacteria are primarily heterotrophic but have been previously shown to respond to diel fluctuations driven by primary producer‐derived dissolved organic carbon (Kelly et al., 2019 ). Similarly, our results show that higher light availability states (i.e., full moon and daytime) were associated with a stronger microbial response, evidenced by a higher number of enriched bacterial taxa (ASVs). However, consistent with previous work (Baquiran et al., 2020 ; Silveira et al., 2017 ), we found no overall changes in the bacterial community composition (activity) between diel cycles. Instead, we identified taxa from bacterial families such as Rhodobacteraceae and Sphingomonadaceae oscillating between day and night, which putatively have food webs tightly coupled with Symbiodiniaceae and/or phytoplankton‐derived organic matter. Notably, our approach to assess community changes based on the active microbiome (cDNA‐based) may have helped resolve such differences. According to the predictions of bacterial functional signatures, photoautotrophs and diazotrophs were more abundant in urban samples than in non‐urban samples dominated by heterotrophs. This opens the possibility that bacterial communities in urban sites prefer autotrophic metabolism contrary to the heterotrophic metabolism preferred in urban sites. However, the accuracy of functional predictions for non‐human samples are limited, and the derived hypotheses require further investigation. Overall, our data suggest seasonal microbiome variation is affected by urbanization, indicating the impact of urbanization on the holobiont in general and not only the coral host. Contemporary coral reef ecosystems are thought to have evolved in the last 45–50 million years (Close et al., 2020 ). During this period, reefs have experienced a wide range of natural disturbance regimes differing in magnitude, duration, and frequency, to which these complex ecosystems and their reef‐building corals have adapted and evolved (Buddemeier & Smith, 1999 ; Hatcher, 1997 ; Nyström et al., 2000 ). These natural disturbance regimes have led to the high species diversity, complex community structure and dynamics characteristic of pre‐industrial coral reefs (Pandolfi, 1999 ). Conversely, human‐induced disturbances often happen in a more persistent manner and occur at frequencies that prohibit adaptation, acclimatization, or recovery (Connell, 1997 ). In the longer run, even low levels of chronic stress can have severe impacts on coral reef ecosystems, causing decreased reproduction and growth rates, and compromising coral immunity (Richmond, 1993 ). In this presented work we have focused on interpreting the differences between urban and non‐urban sites mainly in regard to light pollution, nutrients, and eutrophication. This is since we have active monitoring data from the Israeli National Monitoring Program, regrading those stressors and knowledge of how they can impact the lifestyle of coral reefs from physiological and molecular aspects. However, we are not ignoring the fact that other sensory pollutants, such as chemical and hormonal pollution, which are not monitored, might add another layer of stress to the system. Given the critical importance of the coral holobiont to the fabric of coral reef ecosystems, the impact of increasing urbanization on coral biorhythms adds a further level of threat to an already compromised reef ecosystem unaccounted for in current projections of reef loss (Anthony et al., 2008 ; Fitt & Warner, 1995 ; Jokiel & Coles, 1990 ; Negri et al., 2005 ; Rj, 1997 ). Finally, the coral reefs in the northern part of the Red Sea are considered as coral refuge from climate change and ocean acidification (Krueger et al., 2017 ). The increased economic interest, future development planned along the Red Sea coastlines, which are still not heavily populated, will eventually expose the Red Sea fringing reefs to human‐based disturbances in addition to the environmental threats (Fine et al., 2019 ; Loya, 2004 ); therefore, we hope our study can serve as warning to the potential sensory pollutants chronic disturbances can impair coral reefs."
} | 4,847 |
27279223 | PMC4900452 | pmc | 2,716 | {
"abstract": "Summary Breviatea form a lineage of free living, unicellular protists, distantly\nrelated to animals and fungi 1 – 3 . This lineage emerged almost one billion\nyears ago, when the oceanic oxygen content was low, and extant Breviatea have\nevolved or retained an anaerobic lifestyle 4 . Here we report the cultivation of Lenisia\nlimosa , gen. et sp. nov., a newly discovered breviate colonized by\nrelatives of animal-associated Arcobacter . Physiological\nexperiments showed that the association of L. limosa with\n Arcobacter was driven by the transfer of hydrogen and was\nmutualistic, providing benefits to both partners. With whole genome sequencing\nand differential proteomics we show that an experimentally observed fitness gain\nof L. limosa could be explained by the activity of a so far\nunknown type of NAD(P)H accepting hydrogenase, which was expressed in the\npresence, but not in the absence of Arcobacter . Differential\nproteomics further revealed that the presence of Lenisia \nstimulated expression of known “virulence” factors by\n Arcobacter . These proteins typically enable colonization of\nanimal cells during infection 5 , but may in\nthe present case act for mutual benefit. Finally, re-investigation of two\ncurrently available transcriptomic datasets of other Breviatea 4 revealed the presence and activity of\nrelated hydrogen-consuming Arcobacter , indicating that\nmutualistic interaction between these two groups of microbes might be pervasive.\nOur results support the notion that molecular mechanisms involved in virulence\ncan also support mutualism 6 as shown here\nfor Arcobacter and Breviatea."
} | 405 |
35421312 | PMC9048697 | pmc | 2,717 | {
"abstract": "The development of\na superhydrophobic and, even, water-repellent\nmetal alloy surface is reported utilizing a simple, fast, and economical\nway that requires minimum demands on the necessary equipment and/or\nmethods used. The procedure involves an initial irradiation of the\nmetallic specimen using a femtosecond laser, which results in a randomly\nroughened surface, that is subsequently followed by placing the item\nin an environment under moderate vacuum (pressure 10 –2 mbar) and/or under low-temperature heating (at temperatures below\n120 °C). The effects of both temperature and low pressure on\nthe surface properties (water contact angle and contact angle hysteresis)\nare investigated and surfaces with similar superhydrophobicity are\nobtained in both cases; however, a significant difference concerning\ntheir water-repellent ability is obtained. The surfaces that remained\nunder vacuum were water-repellent, exhibiting very high values of\ncontact angle with a very low contact angle hysteresis, whereas the\nsurfaces,\nwhich underwent thermal processing, exhibited superhydrophobicity\nwith high water adhesion, where water droplets did not roll off even\nafter a significant inclination of the surface. The kinetics of the\ndevelopment of superhydrophobic behavior was investigated as well.\nThe findings were understood when the surface roughness characteristics\nwere considered together with the chemical composition of the surface.",
"conclusion": "Conclusions The surface properties of a Ti6Al4V metal alloy were investigated\nfollowing laser irradiation, and the effects of vacuum pressure, temperature,\nand environment on the wetting properties of the surface were evaluated.\nThe initial smooth surface of the Ti6Al4V alloy can be characterized\nas hydrophilic since its contact angle is 60 ± 2°. By irradiating\nthe surface with a femtosecond laser, a random roughness is fabricated,\nwhich, together with the surface polar groups introduced during irradiation,\nrenders the surface superhydrophilic belonging to the Wenzel wetting\nregime. The wetting properties of the surface can be significantly\naltered by residing in ambient air, heating at various (low) temperatures,\nand under vacuum in a vacuum chamber; however, the kinetics of the\nmodification is very different in the three cases. An irradiated surface,\nwhich remained under ambient conditions, becomes hydrophobic with\na maximum contact angle of 129 ± 2°; however, the change\ntakes a long time and it lasts for a finite time interval. When the\nsurface is heated following irradiation, the higher the temperature\nat which the surface is heated, the greater the contact angle that\ncan be achieved and the faster this value is attained; for temperatures\nequal to or higher than 120 °C, the surface can become hydrophobic\nin less than 24 h. On the other hand, there is a significant and not\ntrivial dependence of the contact angle on the pressure of the vacuum\nchamber in which the surface resides. It is both the pressure and\ntime that the surface remains at the specific vacuum that affect the\nbehavior. The lower the pressure, the higher the hydrophobicity of\nthe surface. When the surface resides for at least 3 h under dynamic\nvacuum conditions, its contact angle reaches a value of about 149\n± 1°. It is shown that these mild post-irradiation\ntreatments do not\nmodify the characteristics of the micro/nano-structured morphology\nof the metallic surfaces that is developed by the laser processing;\nit is only its chemical composition that is modified by the exposure\nto the different environments, which results in a modification of\nthe wetting behavior. For the irradiated surfaces staying in air or\nundergoing temperature or vacuum treatments, the surface composition\nshows higher carbon contents due to the adsorption of hydrocarbons\nfrom the environment rendering the surfaces hydrophobic. This, together\nwith the existing micro/nano-structuring of the surfaces, makes them\nsuperhydrophobic. It is not clear, however, whether those surfaces\ncan be considered belonging to the Wenzel or to the Cassie–Baxter\nregime. Although both temperature and pressure lead to surfaces\nwith similar\nhydrophobicity, i.e., in both cases, the maximum measured contact\nangle was the same, there is a significant difference concerning their\nwater-repellent properties; surfaces that remained under vacuum were\nsuperhydrophobic and water-repellent (very low contact angle hysteresis),\nwhereas the ones that had gone through thermal processing were superhydrophobic\nwith high water adhesion (a water drop did not roll off even when\nthe surface was tilted by 90°). Therefore, temperature\nand vacuum can provide facile, economic,\nand eco-friendly ways to fabricate superhydrophobic surfaces with\nlow or high water adhesion to be utilized in different applications\nthat preclude the use of organic coatings for surface modification.",
"introduction": "Introduction Superhydrophobic surfaces\nhave attracted significant scientific\ninterest due to their importance in both fundamental research and\npractical applications; 1 − 5 such applications include self-cleaning surfaces, 6 − 9 antifogging materials, 10 , 11 anti-icing, 12 , 13 antifouling, 14 sensing, 15 , 16 microfluidics, 17 − 19 biomedical\napplications, 20 , 21 fabrics, 22 − 25 etc. The existence of hierarchical\nsurface roughness and the appropriate chemical composition are critical\nparameters that control the behavior. 1 , 6 , 26 − 31 Surfaces can be described as superhydrophobic and water-repellent\n(with low water adhesion) when the contact angle of a water droplet\nis as high as 150°, whereas the contact angle hysteresis is less\nthan 10° 1 , 27 , 32 or as superhydrophobic with high water adhesion, when the water\ncontact angle is similarly as high as 150° and the contact angle\nhysteresis is high. 33 − 35 The inspiration for developing superhydrophobic surfaces\ncomes from nature; a plant that is extensively studied for its properties\nis the lotus leaf ( Nelumbo mucifera ), which is characterized by high contact angle, ultralow water adhesion,\nand self-cleaning properties. 27 , 36 In addition to the\nlotus leaf, there are a wide variety of plants and insects that exhibit\namazing properties such as Rosa montana , Strelitzia reginae , Oryza sativa leaves, water striders, Nambibian beetles,\ncicadae, and butterfly wings. 37 − 40 A special category among those are the parahydrophobic\nplants, 41 which exhibit high contact angles,\nbut much less than 150° (e.g., advancing angle of ∼110°),\nwith high water adhesion (contact angle hysteresis of 27°) such\nas various thermogenic plants. Fabricating superhydrophobic surfaces\nis relevant for very different materials like polymers, ceramics,\nand metals; depending on the material, various techniques were utilized\nto achieve it, such as photolithography, nanoimprinting, laser beam\nmachining, wet chemical etching, chemical coating, and molding. 6 , 42 − 46 Among the various materials, metals and metal alloys have\nbeen\nwidely utilized in many applications because of their mechanical and\nthermal properties. However, the surface of metals and metal oxides\nis usually hydrophilic due to its high surface energy; in this case,\nthe introduction of hierarchical roughness or geometrical patterns\nrather enhances the hydrophilicity and the application of a low-surface-energy\nmaterial as coating is necessary to alter the behavior. 47 Titanium and its alloys are especially important\nstructural metals, which, because of their excellent physicochemical\nproperties like low density, high specific strength, good resistance\nto corrosion, and nonmagnetic character, are widely used in many applications\nincluding aerospace, marine applications, and in biomedicine as implant\nmaterials and/or as dental and orthopedic prostheses. 48 , 49 Ti6Al4V is the most frequently used titanium alloy in aircraft structural\nparts, aero-engines, low-pressure steam turbine materials as well\nas in the biomedical field due to its excellent mechanical properties\nsuch as high modulus of elasticity, fatigue strength, fracture toughness,\nas well as corrosion resistance, biocompatibility, and bioadhesion. 50 − 52 Altering the surface properties of titanium alloys and developing\ntitanium-based superhydrophobic materials can raise even further its\nvalue in the aircraft and shipping industries. It is noted that, in\nsuch applications, utilization of an organic coating to alter the\nsurface behavior is, in most cases, not a desirable option because\nof the mechanical softness of the coatings and the deterioration of\nits wetting properties with time. To fabricate superhydrophobic\nmetallic surfaces, different methods\nhave been utilized, such as solution immersion, sol–gel, chemical\netching, and electrodeposition. 53 − 59 However, these techniques have the disadvantages of being complicated,\nexpensive, and resulting in poor mechanical properties of the fabricated\nsurface structures. Recently, laser surface modification is considered\nas one of the simpler and most effective approaches to directly produce\nsuperhydrophobic surfaces on a wide variety of metals since it provides\nprecise control of the three-dimensional hierarchical micro/nanostructures. 60 It is a noncontact, nonpolluting, cost-effective,\nand flexible method that offers efficient and high precision texturing.\nAt the same time, it does not require a specific environment as high\nvacuum or clean room facility during the process, and it can create\na multiscale hierarchical roughness in a one-step process. 61 Various works have investigated laser structuring\nof metal substrates, such as Ti6Al4V, 62 stainless steel, 63 , 64 copper, 65 magnesium alloy, 66 aluminum, 67 and others, and demonstrated the formation of\nhierarchical structures or laser-induced periodic surface structures\n(LIPSS). 68 It is well accepted that\nfreshly processed laser metal surfaces\nare hydrophilic 69 − 74 due to the formation of metal oxides with high surface energy. 75 − 78 One way to alter this is the application of an appropriate coating\nutilizing a low-surface-energy modifier. 79 , 80 Alternatively, leaving the irradiated material in ambient air can\nmodify the properties of the surface from superhydrophilic to hydrophobic\nor even superhydrophobic; however, this procedure, although being\nvery eco-friendly and not requiring any further processing, takes\na very long time to be completed, which ranges between a few days\nto a few weeks. 69 , 70 , 72 , 73 In a few works, researchers attempted to\nshorten the time for the alteration of the behavior using either very\nhigh vacuum 71 , 75 − 78 , 81 or high temperatures. 74 , 82 In most of those works,\nthe authors investigated the effect of the irradiation characteristics\non the shape, size, and periodicity of the obtained surface structures\nand, thus, on the water contact angle; the effect of an environment\nrich in CO 2 , O 2 , N 2 , H 2 O, or organic moieties on the hydrophilic to superhydrophobic conversion\nwas investigated as well. 69 , 71 , 72 The modification of the time scale for obtaining hydrophobic properties\nwas attributed to the effective adsorption of hydrocarbons due to\nthe low partial vapor pressure of water. However, there has not been\nany detailed investigation of the contact angle change as a function\nof the value of the pressure (vacuum) or of the residence time under\na low-pressure environment. In this work, superhydrophobic metallic\nsurfaces are fabricated\nusing a very simple procedure. More specifically, Ti6Al4V alloy surfaces\nwere initially irradiated with femtosecond laser pulses. This was\nfollowed by placing the surfaces under a low vacuum, which was easily\nachieved utilizing a rotary pump, and/or by heating at low temperatures.\nThe surface immediately after the irradiation is superhydrophilic\nwith water droplets spontaneously and completely wetting it. However,\nheating of the irradiated surface at different temperatures results\nin a modification of its surface properties and in the manifestation\nof a superhydrophobic behavior with a contact angle of 149 ±\n2°, especially for temperatures higher than 120°. A similar\neffect was observed when an irradiated surface is placed in a vacuum\nchamber under relatively weak vacuum conditions (pressure 10 –2 mbar); after a minimum of 3 h, the surface was converted to a superhydrophobic\none with a contact angle of 149 ± 1°. Despite the similar\nvalues of water contact angle after processing under vacuum or heating\nat a low temperature, a different behavior is observed concerning\nthe ability of the surfaces to repel water. The superhydrophobic surfaces\nobtained following heating show a significantly high contact angle\nhysteresis, thus showing high water adhesion, whereas the ones exposed\nto low vacuum exhibited very low contact angle hysteresis, thus showing\nwater-repellent behavior. Moreover, the effect of the time that the\nsurface remains in air before further processing (in vacuum or at\na temperature) was investigated. The observed behavior can be understood\ntaking into account the change in the surface chemical composition\nbecause of hydrocarbon absorption due to the post-irradiation processing\nthat is amplified by the effect of the surface micro/nano-roughening\nby the laser irradiation.",
"discussion": "Results and Discussion Figure 1 shows photographs\nof representative water droplets during the contact angle (CA) measurements\non a Ti6Al4V surface together with the respective images of the surface\nmorphology prior ( Figure 1 a,b) and following ( Figure 1 c,d) laser irradiation. The untreated smooth surface\ncan be characterized as hydrophilic exhibiting a contact angle of\nabout 60 ± 2°; it is noted that this value is the average\nof at least five measurements at different positions on the surface.\nIts surface energy was evaluated as σ = 40.7 ± 0.3 mN/m\nutilizing the OWRK (Owens, Wendt, Rabel και Kaelble)\nmethod using four different fluids (water, glycerol, ethylene glycol,\nand dimethylsulfoxide). The topology of the Ti6Al4V surface was imaged\nby scanning electron microscopy (SEM) as shown in Figure 1 b, where the smooth surface\nof the untreated alloy is demonstrated. Figure 1 Photographs of representative\nwater droplets on a flat Ti6Al4V\nsurface (a) and immediately after laser irradiation of the Ti6Al4V\nsurface (c) and the corresponding SEM images of a flat (b) and a micro/nanostructured\n(d) Ti6Al4V surface. During direct fs-laser\nwriting, the laser–matter interaction\nleads to the removal of material (ablation) and, ultimately, to the\nformation of random micro/nanostructures on the surface. The SEM image\nof the irradiated surface, shown in Figure 1 d, shows that significant roughness has been\ndeveloped on the surface and emphasizes the difference in its topography\nfrom the corresponding one of the nonirradiated surface. When the\ncontact angle measurement is performed immediately after the irradiation,\nthe water drop completely wets the surface, making the determination\nof the contact angle value very difficult ( Figure 1 c); this can only be estimated to be smaller\nthan ∼5°. Therefore, the irradiation results in superhydrophilic\nsurfaces. The change in the wetting behavior of the surface can be\nattributed both to the increased roughness as a result of the femtosecond\nlaser irradiation and to the formation of polar functional groups\nthat are formed during the ablation process. It is noted that according\nto the Wenzel model that assumes a homogeneous wetting, roughness\nmakes a hydrophilic surface even more hydrophilic; therefore, the\nmicro/nanostructured surface right after irradiation belongs to the\nWenzel regime of wettability. The observed increase in hydrophilicity\nfollowing laser irradiation is in agreement with previous investigations\non various metallic materials. 69 − 74 Following the irradiation, the surface was left under ambient\nconditions\nfor variable time intervals and the wetting properties were evaluated. Figure 2 a,b shows a photograph\nof a representative water droplet during the contact angle measurement\nand a SEM image for the surface, which was left in air for 32 days\nafter laser irradiation. A significant increase in its hydrophobicity\nis observed from the state immediately after irradiation; the contact\nangle has reached a value of 129 ± 2° ( Figure 2 a). Figure 2 (a) Photographs of representative\nwater droplets on a Ti6Al4V surface\n32 days after irradiation and (b) the corresponding SEM image. (c)\nWater contact angle values as a function of time that a surface remains\nunder ambient conditions following the femtosecond laser irradiation.\nError bars are included although they are smaller than or similar\nto the size of the points. (d) Photographs of representative water\ndroplets on a Ti6Al4V surface 82 days after its irradiation and (e)\nthe corresponding SEM image. Figure 2 c illustrates\nthe dependence of the measured contact angles as a function of time\nthe specimen is left in air following irradiation; the contact angle\nvalues increase systematically from its initial superhydrophilic behavior\nup to a certain value during the first 20 days after the irradiation,\nwhereas for more or less the next 40 days, the surface properties\nremain unchanged as the value of the water contact angle remains constant\nand eventually starts to decrease again. Therefore, when an irradiated\nsurface remains at room temperature and pressure in open air, its\nwetting properties can be altered due to its contamination by organic\nmolecules that exist in the environment. However, this property is\nnot permanent. The instability of the surface over time can\nbe attributed to the\nfact that the surface begins to wear out ( Figure 2 e) leading to a decrease in contact angle\n( Figure 2 d) that reaches\nthe value of 90 ± 1° after 82 days. The morphology of the\nsurface as soon as it is irradiated is somewhat “crowded”\n( Figure 1 d), resulting\nin a superhydrophilic behavior ( Figure 1 c). Over time, the morphology changes and appears to\n“solidify”, acquiring hydrophobic properties ( Figure 2 b). However, after\n∼65 days, the contact angle begins to decrease because the\nsurface loses its homogeneity. The increase in the water contact angle\non a freshly irradiated surface that has remained in ambient air has\nbeen reported in the past, as well as the long time that it takes\nuntil the surface becomes hydrophobic; 69 , 70 , 72 , 73 however, to our knowledge,\nthis is the first time that it is reported that this change is not\npermanent but it deteriorates with time; this is a finding that would\nprohibit the use of such surfaces in current industrial applications. Effect\nof Heating Treatment The irradiated metal alloy\nsurfaces were placed in an oven at various temperatures between 25\nand 200 °C for different time intervals up to 96 h, and their\nwetting characteristics were investigated. Figure 3 shows an image of a representative water\ndrop on an alloy surface that has remained in an oven at 120 °C\nfor 24 h after irradiation and its corresponding SEM image. Figure 3 (a) Photograph\nof a representative water droplet on a Ti6Al4V surface\nthat was heated at 120 °C for 24 h after irradiation and (b)\nthe respective SEM image. The measured water contact angle is 149 ± 2°, i.e., heating\nthe surface at this temperature turns the surface superhydrophobic.\nAt the same time, the morphology of the surface shows certain heterogeneities\nthat resemble the topography of a surface that was left at RT for\nmore than 30 days after irradiation ( Figure 2 b, with higher-magnification images shown\nin Figure S1 ). Note that, when a flat surface\nis similarly heated, no significant change is observed either in its\nsurface properties (the water contact angle is measured as 65 ±\n1°) or in its surface morphology (shown in Figure S2 ). The modification of the wetting behavior\nof the irradiated surface\nby heating depends on both temperature and residence time at that\ntemperature. The temperature dependence of the contact angle is shown\nin Figure 4 for various\ntimes the surface remains at each temperature. As temperature increases,\nthe water contact angles increase and, above 100 °C, they reach\nvalues of ∼150°. When an irradiated surface is left at\neach temperature for 24 h, the water contact angle values change from\n∼10° (complete wetting) when annealed at 25 °C to\n∼70° when annealed at 80 °C (hydrophilic surface).\nAnnealing, however, at even higher temperatures results in a sharp\nincrease in the value of the contact angle, which reaches the value\nof 135 ± 2° when annealed at 90 °C and ∼150°\nwhen annealed at 120 °C or higher. When the irradiated surfaces\nremain at various temperatures for longer times, the behavior is qualitatively\nsimilar to contact angles increasing faster with annealing temperature\nfor temperatures up to 80 °C followed by a smaller jump to values\nclose to 149 ± 2° for higher annealing temperatures (from\n112 ± 1 to 142 ± 2° for 72 h and from 125 ± 1\nto 146 ± 2° for 96 h). Figure 4 Water equilibrium contact angles on a\nTi6Al4V surface that was\nheated at different temperatures for various times. Error bars are\nincluded in all points although they are smaller than the size of\nthe points. If one prefers to analyze the\ncontact angle data as a function\nof time for each temperature, one would notice that for temperatures\nhigher than 120 °C, the maximum of the contact angle is obtained\neven when the sample remains at this temperature for 24 h, whereas\nany additional increase of the residence time for these temperatures\ndoes not result in any further effect on the contact angle. At 90\n°C temperature, the contact angle is already high enough (135\n± 2°) during the first 24 h, and it shows a weak dependence\non residence time since it reaches the value of 146 ± 1°\nafter 96 h. At lower temperatures, however, the contact angle after\n24 h treatment is even lower and the time to reach the maximum value\nis assumed to require much longer than 96 h (since it was not reached\nin any of the cases). Moreover, in the extreme case of the surface\nthat has remained at room temperature, the maximum value of the contact\nangle was not achieved even after ∼60 days when the deterioration\nof the surface had started, as discussed above ( Figure 1 c). It is noted that, when the surface that\nhas undergone a heating treatment is immersed in water, no effect\nis observed on the measured contact angles, whereas, when it is immersed\nin ethanol, the water contact angle values are reduced to 120 ±\n1°. Complete characterization of the wetting properties\nof a surface\nrequires the measurement of the contact angle hysteresis besides the\nmeasurements of the equilibrium contact angle. To measure the contact\nangle hysteresis, a water droplet is placed on the surface and, then,\nthe OCA system begins to tilt until the drop begins to roll. Figure 5 shows photographs\nof representative water droplets on a surface that had remained at\n120 °C for 24 h (i.e., for a surface that shows a superhydrophobic\nbehavior) as deposited on the surface, where it shows a 149°\ncontact angle ( Figure 5 a), at a tilt angle of 45°, where the difference between the\nadvancing and the receding angles are shown ( Figure 5 b) and at a tilt angle of 90°, where,\nevidently, the droplet still does not roll off the surface ( Figure 5 c). A video of the\nbehavior of a water droplet during the tilting of the surface that\nhas remained at 120 °C for 24 h is shown in the Supporting Information\nas Video S1 . Therefore, the particular\nTi6Al4V surface that was heated at a relatively high temperature has\nbecome superhydrophobic with, however, high water adhesion, and it\ncan be characterized as exhibiting the rose petal effect. In contrast,\nwhen Al 74 or Cu 82 surfaces were irradiated with a nanosecond laser and were, subsequently,\nheated at 100 °C for 24 and 13 h, respectively, superhydrophobic\nsurfaces with low water adhesion (low sliding angles) were obtained. Figure 5 Photographs\nof a representative water droplet on a Ti6Al4V surface\nthat was heated at 120 °C for 24 h after irradiation (a) as deposited\nonto the surface, (b) when the surface is at a tilt angle of 45°,\nand (c) when the surface is at a tilt angle of 90°. Effect of Vacuum Treatment Figure 6 shows an image of a representative water\ndroplet on an irradiated Ti6Al4V surface that was placed in a vacuum\noven for 24 h under dynamic conditions following its irradiation together\nwith a SEM image that illustrates its topography. It is clear that\nresiding in vacuum rendered the initially superhydrophilic surface\nafter irradiation to a superhydrophobic one with a contact angle of\n149 ± 1° ( Figure 6 a). Figure 6 (a) Photograph of a representative water droplet on a Ti6Al4V surface\nthat remained under dynamic vacuum for 24 h after irradiation and\n(b) its corresponding SEM image. At the same time, the morphology of the surface ( Figure 6 b, with higher-magnification\nimages shown in Figure S3 ) is significantly\ndifferent from the respective one of the irradiated surface right\nafter irradiation, whereas it resembles the one that has been heated\nat high temperatures (e.g., Figure 3 b). It is noted that, when a smooth nonirradiated surface\nremains under the same vacuum conditions, no effect on its wetting\nproperties was observed and its contact angle was measured at 67 ±\n1°, i.e., it showed a behavior very similar to a surface that\ndid not reside under vacuum; this is illustrated in Figure S4 . Therefore, the roughness that was obtained via\nirradiation is a critical parameter to alter the wetting properties\nof the metal alloy. Considering the possibility that some kind\nof contamination of\nthe treated surface may have occurred due to the lubricating oil of\nthe vacuum pump, the procedure was repeated placing the surface in\na chamber where the vacuum is attained by an oil-free pump. The latter\nsurface exhibited a similar morphology to the one shown in Figure 6 b; thus, a similar\neffect on hydrophobicity could be anticipated. However, the surface\nthat remains in the oil-free pump vacuum chamber did not show any\nincrease in the contact angle, i.e., the water droplets continue to\ncompletely wet the surface. Therefore, it is the chemical composition\nof the two surfaces that should be different since the morphology\nis similar but the contact angle and, therefore, the wetting properties\nare significantly different; the latter depends both on the composition\nand the roughness of the surface. The stability over time of\nthe wetting properties of the superhydrophobic\nsurface after the vacuum treatment is of utmost importance. It is\nnoted that, as discussed above, a deterioration of the behavior was\nobserved following the initial increase in the contact angle for surfaces\nthat had simply remained in air after irradiation. Figure 7 a shows the equilibrium water\ncontact angles on three different surfaces that resided under vacuum\nfor 24 h after irradiation as a function of time in air after vacuum.\nThe data illustrate the reproducibility of the measurements and the\nquality of the developed surfaces, which are very stable over time\nwith contact angles showing constant values of 149 ± 1°.\nThus, the obtained surfaces are significantly stable over time, in\ncontrast to surfaces, which remained in air without having undergone\nany vacuum treatment at all. This is corroborated, as well, by the\nSEM image of a surface obtained after 82 days in air following its\nplacement under vacuum for 24 h ( Figure 7 b). Moreover, it is noted that the water\ncontact angles on the surfaces that remained under ambient conditions\nin air, even before their deterioration, are significantly lower than\nthose on surfaces that remained in vacuum, showing smaller contact\nangles by ∼20°. In conclusion, when the surface is treated\nunder vacuum, it acquires a large contact angle, which is stable over\nlong times. Moreover, it is noted that when the vacuum-treated surface\nis immersed in water, no effect is observed on the measured contact\nangles, whereas, when it is immersed in ethanol, the water contact\nangle values are reduced to 127 ± 1°. Figure 7 (a) Equilibrium contact\nangle values as a function of the time\nthat the surfaces have remained in air following 24 h under vacuum\n(red, blue, and green points) or directly (black points) after irradiation.\nError bars are included in all cases although they are smaller than\nor similar to the size of the points. (b) SEM images of an irradiated\nTi6Al4V after 82 days in air following its placement under vacuum\nfor 24 h. The parameters of the processing\nsteps that a surface undergoes\nsignificantly affect the final wettability of the surfaces. Such parameters\nare related, on one hand, to the time between the surface irradiation\nand its placement under vacuum and, on the other hand, to the details\nof the processing of the surface under vacuum. Figure 8 shows the corresponding contact angle data\nof an irradiated surface as a function of the time that the surface\nremained in air before it is placed under vacuum for 24 h. It is clear\nthat the longer the surface stays in air before it is placed under\nvacuum in the vacuum chamber, the smaller the contact angle that the\nsurface reaches after 24 h under dynamic conditions. More specifically,\nwhen the surface is placed under vacuum immediately after irradiation,\na superhydrophobic surface is developed with a contact angle of 149 ±\n1°; however, if the surface remains in air for 4 h and, then,\nis placed under dynamic vacuum for 24 h, it barely becomes hydrophobic\nsince the contact angle is measured as 90 ± 2°. Thus, to\nachieve superhydrophobicity, the surface should be placed under vacuum\nimmediately following its irradiation. Figure 8 Equilibrium contact angle\nvalues as a function of the time that\nthe surface remains in air before it is placed in vacuum under dynamic\nconditions for 24 h. Error bars are included even when they are smaller\nthan the size of the points. On the other hand, Figure 9 a shows the water contact angle values on surfaces, which\nwere introduced in the vacuum chamber immediately after irradiation\nand have remained continuously under dynamic vacuum for different\ntime periods. The contact angles increase weakly for small time intervals,\nwhereas a jump in the contact angle values is observed after ∼50\nmin and, finally, a plateau is reached with the maximum obtained contact\nangles being 149 ± 1°. More specifically, when the surface\nremains under vacuum for 40 min, the contact angle becomes 45 ±\n1°, whereas when it remains for 60 min, i.e., 20 min longer,\nits value jumps to 133 ± 1°. Finally, it seems that staying\nin vacuum for 180 min is enough for the surfaces to reach the most\nsuperhydrophobic state expressed via contact angles of 149 ±\n1°. Residing under vacuum for even much longer times does not\nresult in any further increase in the contact angle. Figure 9 (a) Equilibrium contact\nangle values as a function of the total\ntime that the sample remained in the vacuum chamber continuously (the\nsurfaces were placed in vacuum immediately after irradiation). The\ninset shows the measured pressure inside the vacuum chamber as a function\nof pumping times. (b) Equilibrium contact angle values as a function\nof the total time that the sample remained in the chamber in a noncontinuous\nway (the surfaces were placed in vacuum immediately after irradiation).\nThe error bars are smaller than the size of the points. Figure 9 b\nshows\nthe static water contact angle values on surfaces that have been placed\nunder vacuum immediately after irradiation not continuously but with\nintermediate breaks as a function of the accumulated times. A behavior\nsimilar to that of Figure 9 a with increasing contact angles is observed as well. More\nspecifically, for a total time from 10 to 120 min (measured in steps\nof 10–30 min), the contact angles increase weakly with time.\nHowever, following this initial increase, a jump appears in the contact\nangle value since, after 120 min total time, the contact angle is\nmeasured at 50 ± 2°, whereas after an additional 30 min\n(i.e., total time of 150 min), the contact angle has increased up\nto 120 ± 1°. After that jump, the contact angle keeps increasing,\nhowever, with a slower rate, reaching a maximum contact angle of about\n149 ± 2° at a total time of ∼270 min. This is the\nmaximum value of contact angle that can be obtained since it does\nnot increase further even if the surface remains under vacuum for\nmuch longer times. The fact that the contact angle increases\nand the superhydrophobic\nstate is obtained faster when the surfaces reside under vacuum continuously\ncompared to when they are placed in vacuum in discrete time intervals\nindicates that the critical parameter is not simply the time that\nthe surface remains under vacuum but rather the air pressure, which\nis reached in the chamber; the longer the vacuum pump operates, the\nlower the pressure inside the vacuum chamber (inset of Figure 9 a). Therefore, the difference\nin the contact angle values between the two experiments described\nabove (continuous time intervals under vacuum, Figure 9 a, and total time with discrete measuring\nslots, Figure 9 b) can\nbe explained by considering the values of the pressure in the chamber\nas a function of pumping time. When the vacuum pump starts, the pressure\ninside the vacuum chamber decreases continuously as a function of\ntime as shown in the inset of Figure 9 a. Therefore, at the beginning, the pressure has a\nhigh value, which decreases with time to the minimum value of ∼0.020\nmbar when the pump is running continuously for 24 h. The top x -axis of Figure 9 a illustrates the values of the pressure in the chamber for\nthe various times the pump operates. It is clear that high values\nof the water contact angles correspond to the lower values of the\npressure in the chamber. Moreover, the maximum value of contact angle\nof 149 ± 1° is obtained at a pressure of 0.039 mbar (corresponding\nto a pumping time of 3 h). In conclusion, by leaving the surface in\nthe chamber for shorter times, the pressure in the vacuum chamber\nis not sufficiently low to provide a satisfactory effect on the hydrophobicity\nof the surface. However, pressure is not the only parameter\nthat influences the\nwetting properties of the surfaces; successive placing of the irradiated\nmetal alloy even at the same pressure results in an increase in its\nhydrophobicity. As one can see in Figure 9 b, when the surface is left under dynamic\nvacuum for 10 min, i.e., when the final pressure reached in the vacuum\nchamber is 0.3 mbar (from the inset of Figure 9 a), a contact angle of 8 ± 2° is\nobtained; following the contact angle measurement and the subsequent\nplacement of the surface back in the chamber for 10 min more (i.e.,\ndown to the same pressure as before), the contact angle increases\nto 10 ± 2°, whereas, after the third iteration, the contact\nangle increases to 17 ± 1°. After that, the surface was\nplaced under dynamic vacuum for an additional 30 min, i.e., until\na pressure of 0.085 mbar was reached (from the inset of Figure 9 a). The contact angle was measured\nat 35 ± 1°, and then, it increased to 40 ± 2 and to\n50 ± 1° after the second and third times the same surface\nremained for an additional 30 min each time, respectively. During\nthe fourth iteration of 30 min, the contact angle displayed a jump\nfrom 50 ± 1 to 120 ± 2° and continued to increase up\nto 130 ± 2 and 137 ± 2° for more 30 min periods in\nvacuum. Finally, when the same surface was placed in vacuum for an\nadditional 60 min, i.e., when a pressure of 0.059 mbar was reached,\nthe maximum contact angle of 149 ± 2° is achieved. If one\nplaces the surface in the vacuum chamber for even more time, this\ndoes not have any further effect on its hydrophobicity and the contact\nangle does not increase anymore. Therefore, the sequence of processing\nsteps, i.e., whether the surface is placed in vacuum continuously\nor with intermediate breaks, determines the changes in the surface\nproperties of an irradiated surface and the enhancement of its hydrophobicity. These findings can be easily explained if one considers the relation\nbetween pressure and time in the vacuum chamber. As shown in the inset\nof Figure 9 a, when\nthe surface is left under pumping for 10 min in the vacuum chamber,\nthe pressure reaches the value of 0.3 mbar, which means that when\nthe surface is placed under vacuum three times 10 min each (for a\ntotal of 30 min), the pressure is 0.3 mbar at the end of each time\ninterval, but when it is left for 30 min continuously, the pressure\nreaches a value of 0.049 mbar. Therefore, for the difference between\nleaving the surface continuously for 30 min or leaving it cumulatively\nfor 30 min but for lower time intervals, the vacuum has a different\nvalue. The same applies to other times like 30, 60 min, etc. However,\na memory effect exists, and, thus, an increase in the contact angle\nvalues is observed even when a surface is placed repeatedly at a certain\npressure. Residence of an irradiated Ti6Al4V surface in vacuum\nrenders it\nsuperhydrophobic; it is important, however, as it has already been\ndiscussed, to investigate the behavior of the surface with respect\nto water adhesion, which determines its water repellency. Figure 10 shows representative\nwater droplets on a Ti6Al4V surface that remained under vacuum for\n24 h right after irradiation; Figure 10 a shows the droplet as soon as it was deposited on\nthe surface exhibiting an equilibrium contact angle of 149 ±\n1° and after rotating the stage to monitor the angle at which\nthe drop rolls off ( Figure 10 b). The surface exhibits a very weak water adhesion since\nthe drop completely rolls off the surface at a sliding angle of ∼1.5°;\ntherefore, this particular surface is superhydrophobic and water-repellent\nand it can be characterized as exhibiting the Lotus leaf effect. A\nvideo that illustrates the rolling off of a water droplet from an\nirradiated surface that has undergone a vacuum treatment for 24 h\nis shown in the Supporting Information as Video S2 . It is reminded that, in the case of the surfaces that underwent\nheating after irradiation, where the contact angle was 149 ±\n2°, as well, the surface was not water-repellent but exhibited\nhigh water adhesion with the droplet being adhered onto the surface\neven after rotation of the system by 90°. Therefore, both treatments\nof the surface following laser irradiation result in superhydrophobic\nbehavior, however with different water adhesion characteristics. The\neffect of vacuum treatment on the superhydrophilic–superhydrophobic\nconversion was investigated in a few works in the past; however, in\nthe majority of such studies, an ultrahigh vacuum O (1 × 10 –4 Pa) was applied to influence the surface properties. 75 − 78 , 81 In only one previous work, a\nmoderate and easily accessible vacuum was applied to render copper\nsurfaces hydrophobic; in that case, however, it took ∼8 days\nto achieve a water contact angle of 120°. 70 Therefore, it is the first time, at least to our knowledge,\nthat superhydrophobicity and water repellency are attained utilizing\nan easy, fast, and economic way achievable in every laboratory. Figure 10 Photographs\nof representative water droplets on a Ti6Al4V surface\nthat remained in a vacuum oven under dynamic conditions for 24 h (a)\nas deposited onto the surface and (b) after a 1.5° rotation of\nthe system. Chemical Analysis In general, it is the chemical composition\nof a flat surface that determines its hydrophobicity or hydrophilicity.\nThe effects of the chemistry of the surface are further modified (usually\nenhanced) by the surface morphology, especially by the presence of\nhierarchical roughness. 1 , 69 , 85 , 86 As far as metals and metal alloys are concerned,\ntheir surfaces are usually covered by films of the respective native\noxides, which give them high surface energy and, thus, hydrophilic\nproperties. However, the situation changes, as discussed above, when\nthe surface is exposed to laser irradiation or other aggressive conditions\nthat initiate adsorption, chemisorption, and chemical interactions\nof gases, vapors, or moisture present in ambient air. It is well known\nthat hydrophilic surfaces with high surface energy are rich in polar\nfunctional groups in contrast to hydrophobic ones, which are rich\nin nonpolar groups. Therefore, assuming the same surface morphology,\nas the oxygen content on the surface increases, the surface hydrophilicity\nis expected to increase, while an increase in the carbon content,\nlike, for example, by adsorption or chemisorption of nonpolar hydrocarbons\npresent in air, leads to hydrophobic behavior. In the case of the\npresent systems, it is the femtosecond laser irradiation of the surface\nthat causes its morphological changes and no further modification\nin its structural characteristics or roughness can be imposed when\nit resides under ambient air, temperature, or vacuum. It is, thus,\nanticipated that the alteration of the surface wetting properties\nis completely due to changes in its chemical composition. To\nquantify the changes in the chemical composition of the Ti6Al4V surfaces\nfollowing the laser irradiation and the subsequent treatment in air,\nat a particular temperature and/or under vacuum, energy-dispersive\nX-ray spectroscopic (EDS) analysis was employed. Table 1 shows the results of the EDS\nanalysis for all of the surfaces investigated. EDS results show that\nthe C:O:Ti atomic ratio of a smooth Ti6Al4V surface is 4.8:4.1:81.9,\nand it remains almost unaffected when this surface stays under vacuum\nor heating. However, when the surface is irradiated, a significant\nincrease in the amount of the surface oxygen is observed with the\nratio C:O:Ti becoming 6.4:58.1:30.7, which justifies the superhydrophilicity\nmeasured by the contact angle measurements. This can be understood\nbased on a passivation layer formed by water molecules (moisture)\non the surface after the irradiation and on the presence of a large\nnumber of coordinatively unsaturated metal and oxygen atoms that promote\nthe heterolytic dissociative adsorption of water molecules; this gives\nrise to a hydroxylated layer on the oxide surface, which in turn can\nadsorb a second water layer via hydrogen bonds. 75 , 81 Fabrication of a superhydrophobic surface requires the reduction\nof its surface energy, which can be achieved via adsorption of organic\nmolecules and/or the formation of a thin carbonaceous layer. Following\nthe irradiation of the surface, its presence in air or under vacuum\nor at high temperature causes a significant increase in the amount\nof carbon in agreement with previous studies; 72 , 75 − 77 , 81 this amount increases\nmore and much faster in the two latter cases of vacuum treatment or\nheating. The increase in the amount of carbon even in ambient air\nis due to the adsorption or chemisorption of hydrocarbon molecules\npresent in air, which may form carboxylates via, e.g., esterification\nonto the hydroxylated metal oxide surfaces that becomes significantly\nmore effective under vacuum, due to the very low water vapor content\ninside the vacuum chamber since the passivation of reactive OH sites\nby hydrogen-bonded water molecules is avoided. The effect is also\nenhanced by heating because of the effect of temperature on the kinetics\nof adsorption but to a less significant degree. Table 1 EDS Analysis of the Studied Surfaces\n(Expressed as Atom %) a C O Ti Al V hydrophobicity smooth surface 4.8 4.1 81.9 8.5 0.7 hydrophilic smooth surface in vacuum 5.0 4.2 81.6 8.6 0.7 hydrophilic smooth surface at 120 °C for 24 h 4.7 4.1 82.2 8.3 0.7 hydrophilic just irradiated surface 6.4 58.1 30.7 4.1 0.7 superhydrophilic irradiated surface and 24 h in vacuum (oil-free pump) 6.8 54.2 34.6 4.4 0.8 superhydrophilic irradiated\nsurface and 5\ndays in air 8.9 55.7 31.6 4.1 0.6 hydrophilic irradiated surface and 24 h heating at 120 °C 14.9 47.6 32.2 4.7 0.6 superhydrophobic irradiated\nsurface and 24 h in vacuum (oil pump) 15.4 50.7 28.9 4.2 0.7 superhydrophobic a C: carbon; O: oxygen; Ti: titanium;\nAl: aluminum; V: vanadium."
} | 11,019 |
28461887 | null | s2 | 2,718 | {
"abstract": "Altough double network (DN) hydrogels are extremly tough, they are irreversibly softened during large strain deformation. We incorporated mussel-inspired adhesive moiety, catechol, and a synthetic nano-silicate, Laponite, into DN to examine the effect of strong, reversible crosslinks on the DN's ability to recover its mechanical properties during successive loading cycles. The introduction of catechol and Laponite drastically increased the compressive strength and toughness of DN without compromising the compliance of the hydrogel. After 2 hours of recovery at room temperature, the nanocomposite DN hydrogel recovered over 95 and 82 % of its strain energy and hysteresis, respectively, during successive compressive loading to a strain of 0.5. Both equilibrium swelling and oscillatory rheometry data confirmed that there were minimal changes to the network crosslinking density and stiffness after large strain compressive deformation, indicating that mechanical loading did not result in irreversible structural damage. Strong catechol-Laponite interactions can be repeatedly broken and reform to dissipate fracture energy and enable the recovery of DN hydrogel."
} | 292 |
33728823 | PMC8252474 | pmc | 2,719 | {
"abstract": "Abstract Quantifying changes in functional community structure driven by disturbance is critical to anticipate potential shifts in ecosystem functioning. However, how marine heatwaves (MHWs) affect the functional structure of temperate coral‐dominated communities is poorly understood. Here, we used five long‐term (> 10 years) records of Mediterranean coralligenous assemblages in a multi‐taxa, trait‐based analysis to investigate MHW‐driven changes in functional structure. We show that, despite stability in functional richness (i.e. the range of species functional traits), MHW‐impacted assemblages experienced long‐term directional changes in functional identity (i.e. their dominant trait values). Declining traits included large sizes, long lifespans, arborescent morphologies, filter‐feeding strategies or calcified skeletons. These traits, which were mostly supported by few sensitive and irreplaceable species from a single functional group (habitat‐forming octocorals), disproportionally influence certain ecosystem functions (e.g. 3D‐habitat provision). Hence, MHWs are leading to assemblages that are deficient in key functional traits, with likely consequences for the ecosystem functioning.",
"introduction": "INTRODUCTION Marine life is increasingly threatened by anthropogenic climate change (Smale et al . 2019 ). Global impacts such as ocean warming are altering the biology and ecology of many organisms, populations and species (Scheffers et al . 2016 ). As a consequence, community‐level biodiversity changes are emerging in the oceans, with potentially far‐reaching consequences for ecosystems’ functioning (e.g. Poloczanska et al . 2016 ; Antão et al . 2020 ). The natural processes (physical, chemical or biological) determining the movement or storage of energy and materials within an ecosystem or its self‐maintenance over time are called ecosystem functions (Paterson et al . 2012 ). The joint effects of all individual functions determine the overall ecosystem functioning (Reiss et al . 2009 ). Abiotic (e.g. light, temperature, pH, nutrient) and biotic (e.g. biodiversity or species interactions) factors influence ecosystem functioning in multiple interconnected ways (Reiss et al . 2009 ). Yet, the role of biodiversity has traditionally been considered as highly influential and thus has been the focus of much scientific research (e.g. Hooper et al . 2005 ; Balvanera et al . 2006 ; Reiss et al . 2009 ). In recent decades, the emergence of trait‐based approaches is providing new opportunities to understand how changes in community structure translate to changes in its functioning. Specifically, these approaches are shifting from the taxonomic perspective of traditional biodiversity–ecosystem function research to a functional one, and in doing so, proposing that changes in ecosystem function can be better estimated when considering the functional roles among species, as measured by their traits (e.g. McGill et al . 2006 ; Mokany et al . 2008 ; Mouillot et al . 2013 ; Madin et al . 2016 ; Hughes et al . 2018 ). Two major complementary hypotheses link changes in trait composition to alteration of ecosystem function: the diversity and the mass ratio hypothesis. According to the diversity hypothesis (Tilman et al . 1997 ), both the species and their associated functional traits influence ecosystem processes through mechanisms such as complementary resource use. Therefore, variation in the range of functional traits in a given community (i.e. its functional richness; Frich ) affect its functioning (Díaz & Cabido, 2001 ; Tilman, 2001 ). Alternatively, the mass‐ratio hypothesis (Grime, 1998 ) states that the functional traits of the dominant species are the primary drivers of ecosystem function. Therefore, changes in community composition or species relative abundance may shift the community dominant traits (i.e. its functional identity; FI ) and subsequently, its functioning (Mouillot et al . 2013 ; Weigel et al . 2016 ). Determining how the richness and mass‐ratio of functional traits respond to ecological disturbances is therefore necessary to forecast their functional trajectories (Mouillot et al . 2011 ; Gagic et al . 2015 ). Furthermore, as some functions in the ecosystems are more likely to be influenced by key taxa rather than overall diversity patterns (e.g. bioerosion in oceanic reefs; Bellwood et al . 2003 ), the vulnerability of species or species’ groups presenting key trait values needs to be also carefully considered (Bellwood et al . 2004 ; Bellwood et al., \n 2019a , b ). Increasingly frequent and intense marine heatwaves (MHWs) have recently triggered devastating warming‐induced mass mortality events worldwide, affecting a wide range of different species‐rich benthic communities such as coral reefs, seagrass meadows or kelp forests (Wernberg et al . 2013 ; Hughes et al . 2017 ; Carlson et al . 2018 ; Smale et al . 2019 ). In the Mediterranean, these extreme warming events have recurrently impacted the coralligenous assemblages, which are endemic reefs home to approximately 10% of Mediterranean species (Cerrano et al . 2000 ; Ballesteros, 2006 ; Garrabou et al . 2009 , 2019 ). Therefore, MHWs are likely causing changes in the structure and functioning of one of the most biodiverse systems in the Mediterranean. However, field surveys (e.g. Garrabou et al . 2009 ; Verdura et al . 2019 ) and aquaria thermotolerance experiments (e.g. Pagès‐Escolà et al . 2018 ; Gómez‐Gras et al . 2019 ) have suggested contrasting vulnerabilities to warming among co‐occurring coralligenous species, in terms of tolerances to or regeneration after MHWs. This phenomenon, called 'response diversity' (Elmqvist et al . 2003 ), can act to stabilise functioning if the more vulnerable species are being replaced by functionally similar (i.e. 'redundant'), but more resistant species (Yachi & Loreau, 1999 ). On coral reefs, for example a mortality outbreak of the staghorn coral Acropora cervicornis , which occurred in Belize during the 1980s due to disease and high temperature, was partially compensated for by the previously uncommon, functionally similar and more thermally resistant lettuce coral, Agaricia tenuifolia , that became the main reef builder (Nyström, 2006 ). Further examples of this stabilising effect can be found in other marine (Steneck et al., \n 2002 ; McLean et al . 2019 ) and terrestrial ecosystems (e.g. Walker et al . 1999 ; Stavert et al . 2017 ). However, if vulnerable species are not replaced, or are replaced by species that do not contribute similarly to a given ecosystem process, important functions are likely to be compromised (e.g. provision of 3D habitats, surface stability or benthic‐pelagic coupling in the case of coral reefs), with potential detrimental consequences for the associated ecosystem services (Gili & Coma, 1998 ; Bellwood et al . 2003 ; Nyström, 2006 ; Cardinale et al . 2012 ). In this study, we combined long‐term (10–15 years) ecological data and in situ temperature data to examine MHW‐induced functional changes in Mediterranean coralligenous assemblages. By quantifying multidimensional trait spaces, we investigated: (1) whether MHWs have driven fine‐scale changes in their functional structure (i.e. Frich and FI ), and (2) whether some functional groups (i.e. clusters of coarsely functionally redundant species sharing similar combinations of traits) are more vulnerable than others to MHWs, which may imply consequences for the maintenance of critical functions in the ecosystem. Our results provide empirical insights into MHW‐driven functional changes in one of the most species‐rich communities in the Mediterranean. Accordingly, this study takes us a step towards understanding the role of climate change as a driver of functional change in coral‐dominated benthic assemblages in temperate regions.",
"discussion": "DISCUSSION MHWs linked to climate change have recurrently impacted Mediterranean temperate reefs in recent decades, leading to mass mortality events and changes in patterns of biodiversity (Cerrano et al . 2000 ; Garrabou et al . 2009 , 2019 ; Verdura et al . 2019 ). Here, we show that MHWs have also induced marked changes in functional trait composition that are likely to impact ecosystem functioning. We examined abundance distributions of functional entities (species sharing identical combinations of traits) across the trait space and found that, whereas non‐impacted coralligenous assemblages maintained their functional richness (range of traits values) and functional identity (dominant trait values) through time, MHW‐impacted assemblages exhibited shifts in their functional identity. In particular, MHWs decreased the abundances of taxa with large sizes, arborescent and massive morphologies, coloniality, high physical defences, slow‐growing and long‐lived life histories or heterotrophic filter‐feeding strategies. For benthic systems such as tropical and temperate reefs, these are traits that confer important ecosystem functions, including the provision of habitat structure, nutrient cycling, carbon storage or benthic pelagic coupling (Gili & Coma, 1998 ; Loya et al . 2001 ; Graham & Nash, 2013 ; Darling et al . 2017 ; Paoli et al . 2017 ; Coppari et al . 2019 ). Thus, their decline in MHW‐impacted assemblages in benefit of other traits (e.g. small sizes, fast‐growing and reproduction, epibiotic strategies, less complex encrusting and foliose morphologies, less calcification or autotrophic photosynthetic strategies) may likely lead to changes in the ecosystem functioning (Figure 5 ). For instance a decrease of large, colonial, calcifying, morphologically complex, or slow‐growing, long‐lived organisms that have a high investment in long‐term maintenance (e.g. arborescent corals, massive sponges or erect bryozoans) is likely to lead to reductions in the creation of three‐dimensional, long‐term, biogenic habitats that could also store carbon for decades (Darling et al . 2017 ; Coppari et al . 2019 ). In benthic ecosystems, the three‐dimensionality of organisms plays a fundamental role in the organisation, function and resilience. It has been associated to positive effects on biodiversity (e.g. by ameliorating physical and biological stresses for the associated species), productivity, invasion resistance or stability over time (Angelini et al . 2011 ; Ponti et al . 2014 , 2018 ; Darling et al . 2017 ; Verdura et al . 2019 ; De la Torriente et al . 2020 ). Moreover, since structural complexity determines water flow disruption, its reduction in MHW‐impacted assemblages may minimise the time that suspended particles remain close to the benthos, depleting prey capture chances, larval settlement probabilities and sedimentation, and thus also the benthic‐pelagic coupling and nutrient cycling (Gili & Coma, 1998 ). Similarly, since heterotrophic filter feeders significantly interact with the water column by depleting food particles and sediments and by transferring energy and nutrients from the water column to the benthos, their decrease in some impacted assemblages may also reduce the benthic‐pelagic coupling, the nutrient cycling, the carbon storage, the energy transfer through the food webs or the lithification processes that contribute to the building up of the coralligenous structure (Cloern 1982 ; Officer et al . 1982 ; Marshall, 1983 ; Kimmerer et al . 1994 ; Gili & Coma, 1998 ). The increased abundance of epibiotic organisms in some impacted assemblages may also hinder the assemblage resilience, since it may lead to the reduction of the resistance and recovery capacity of structural species. In particular, an increase in epibiosis may reduce the fitness of the overgrown organisms by disrupting the energy and material fluxes between their surfaces and the environment (Wahl, 2008 ). Likewise, the observed increase in ruderal organisms with fast life‐history strategies (e.g. algal turfs or weedy macroalgae) may contribute to a rapid colonisation of free spaces after MHWs, thus potentially hindering the recruitment and recovery of the previously dominant macro‐invertebrates and the overall assemblage resilience (Kuffner et al . 2006 ; Linares et al . 2012 ). Figure 5 Schematic representation of functional traits shifts (light yellow boxes) observed in MHW‐impacted C. rubrum and P. clavata dominated assemblages and their potential consequences for different ecological processes (in grey bubbles) and their associated ecosystem functions (in yellow bubbles). Red arrows indicate a reduction of a given process or function while green arrows represent an increase. Overall, the observed MHW‐driven changes in functional identity suggest that environmental filtering is taking place with particular functions being gradually reduced (losers), in detriment of others (winners). This indicates a limited capacity for response diversity and functional redundancy for maintaining the functioning in MHW‐impacted coralligenous assemblages (McWilliam et al . 2020 ). Interestingly, our clustering analysis suggests that the observed changes in functional identity could be mostly the consequence of the lack of both functional redundancy and response diversity in a single functional group (FG); the Cluster 8 grouping the habitat‐forming octocorals. In fact, whereas most FGs were highly redundant and were represented by many species (up to 24 in some cases), Cluster 8 was no redundant and only consisted of one species in most sites. Such lack of redundancy in an abundant cluster indicates that the constitutive species may present unique traits that may contribute disproportionally to the ecosystem functioning and stability (Ellison et al . 2005 ). Therefore, if the species are also highly vulnerable and get their abundance reduced, no functional compensation will occur and serious detrimental consequences for the overall ecosystem functioning may unfold (Bellwood et al . 2003 ; Nyström, 2006 ). Unfortunately, Cluster 8 was not only low in redundancy, but low resistant as well. This FG exhibited collapsing trajectories in coverage (from 65 to 93% reductions) in all MHW‐impacted assemblages (Fig. 4l‐p & Fig. S7). Consequently, although all the other FGs and their underlying ecosystem functions count on certain degree of insurance against MHWs, the quality of those functions that are disproportionately influenced by Cluster 8 (e.g., provision of long‐term 3D habitats, the benthic‐pelagic coupling, the nutrient cycling, the resilience, or the long‐term carbon storage) might be highly compromised. For instance, although the increasingly dominant erect macroalgae could provide three‐dimensionality in some MHW‐impacted assemblages, the function will presumably be of less quality and less durability than when the larger, calcified and longer‐lived species from Cluster 8 used to dominate. Similarly, the emergent encrusting sponges or bryozoans may partially replace the filtering role of the collapsing habitat‐forming octocorals. Yet, not only the type of captured prey would be different, but the overall associated lesser structural complexity would likely reduce their capturing rates as well, affecting the amount of energy and matter transferred from the water column. Overall, our analyses at different levels of trait‐based dissimilarity show MHW‐induced changes in community structure to assemblages that are now deficient in key functional traits. Similarly to what has been observed in tropical coral reefs (e.g. Hughes et al . 2018 ), our results indicate that MHWs are likely inducing severe changes in the ecosystem functioning of Mediterranean temperate reefs. Given the predicted increase in the frequency and intensity of MHWs (Oliver et al . 2019 ), identifying and preserving the mechanisms of reef stability that maintain essential functions and services is critical (Bellwood et al . ,b 2004 , 2019a ; Hughes et al . 2018 ). Here, we have shown that the functional stability (considered here as the maintenance of functional structure over time in terms of both functional richness and identity, and thus of what functions are in the system and how they are performed) of Mediterranean coralligenous assemblages can be highly compromised by the decline of just few pivotal species with unique trait values. Thus, if we aim to preserve these temperate reefs in a way in which their essential ecological functions are maintained, further efforts will be needed to; i) globally reduce CO2 emissions and ii) further investigate the effectiveness of ocean‐based solutions that could promote the resilience of their key habitat‐forming species to MHWs (e.g. the operationalisation of a climate‐responsive design and management of a fully protected network of MPAs in the Mediterranean; Gattuso et al . 2018 ; Bates et al . 2019 )."
} | 4,233 |
31700000 | PMC6838125 | pmc | 2,720 | {
"abstract": "Vegetation impacts on ecosystem functioning are mediated by mycorrhizas, plant–fungal associations formed by most plant species. Ecosystems dominated by distinct mycorrhizal types differ strongly in their biogeochemistry. Quantitative analyses of mycorrhizal impacts on ecosystem functioning are hindered by the scarcity of information on mycorrhizal distributions. Here we present global, high-resolution maps of vegetation biomass distribution by dominant mycorrhizal associations. Arbuscular, ectomycorrhizal, and ericoid mycorrhizal vegetation store, respectively, 241 ± 15, 100 ± 17, and 7 ± 1.8 GT carbon in aboveground biomass, whereas non-mycorrhizal vegetation stores 29 ± 5.5 GT carbon. Soil carbon stocks in both topsoil and subsoil are positively related to the community-level biomass fraction of ectomycorrhizal plants, though the strength of this relationship varies across biomes. We show that human-induced transformations of Earth’s ecosystems have reduced ectomycorrhizal vegetation, with potential ramifications to terrestrial carbon stocks. Our work provides a benchmark for spatially explicit and globally quantitative assessments of mycorrhizal impacts on ecosystem functioning and biogeochemical cycling.",
"introduction": "Introduction Mycorrhizas are mutualistic relationships between plants and fungi, in which fungi supply plants with nutrients and plants provide carbon to fungi 1 . Among mycorrhizal types, arbuscular mycorrhiza (AM), ectomycorrhiza (EcM) and ericoid mycorrhiza (ErM) are geographically the most widespread, colonizing over 85% of vascular plants across vegetated terrestrial biomes 1 – 4 . Due to the facilitation of plant nutrient acquisition 1 and the large biomass of fungal networks in soil 5 , the presence and type of mycorrhiza are among the key determinants of ecosystem functioning 6 – 9 and biogeochemical cycling 10 – 13 . Thus, the types of mycorrhizal associations present likely also affect the global distribution of soil carbon stocks. There is growing evidence that ecosystems dominated by EcM and ErM vegetation exhibit higher topsoil carbon to nitrogen ratios (C/N) compared with ecosystems dominated by AM plants 11 , 12 , 14 , 15 , although in temperate forests the pattern may be reversed in deeper soil layers 16 . The mechanisms driving these differences are heavily debated in current literature, with distinct physiological traits of mycorrhizal fungi most likely playing a critical role 16 – 20 . Although it can be argued that high abundance of EcM plants is a consequence rather than a driver of high soil C stocks, a large body of recent findings provides evidence that EcM symbionts may be the key drivers of topsoil carbon accumulation through two interacting mechanisms. First, EcM fungi produce greater biomass of more recalcitrant mycelium compared to AM fungi 5 . Second, while EcM fungi are more efficient in taking up N in N-poor soils than AM fungi or roots 10 , 21 , EcM fungi immobilize most of the N in their own biomass. This suppresses saprotrophic decomposition process 22 and reinforces the competitive advantage of EcM and ErM plants via enhanced N limitation 22 , 23 . A full understanding of global carbon and nitrogen stocks requires quantitative models on the distribution of mycorrhizal types in ecosystems 18 . Despite the existence of regional maps of current 24 , 25 and past 26 mycorrhizal vegetation, and on the distribution of mycorrhizal fungal species 27 , 28 we still lack global information on the distribution of biomass of mycorrhizal plants, which is a much better proxy for mycorrhizal impacts on ecosystem functioning than the biodiversity of mycorrhizal symbionts. While current terrestrial biosphere models simulate feedbacks between the carbon cycle and vegetation distribution 29 , most models ignore mycorrhizal types and their effects on nutrient cycling. Integration of such information is expected to provide a more realistic simulation of carbon and nutrient fluxes associated with plant nutrition 17 , 18 , 21 and soil carbon cycles 15 , 21 . Quantitative models of the distribution of mycorrhizal vegetation constitute an important missing link between the known effects of mycorrhizas in biogeochemical cycles and their global impacts 30 . Human activities such as forest logging, urbanization and agricultural practices have altered 50–75% of the Earth’s terrestrial ecosystems 31 , transforming areas with previously natural EcM and ErM vegetation into AM and non-mycorrhizal (NM) vegetation. However, the impact of anthropogenic land use shifts on biogeochemical cycles associated with mycorrhiza have remained poorly known due to the lack of appropriate spatial information. Based on a comprehensive quantitative evaluation of plant-mycorrhizal associations and the distribution of vascular plant species across biomes and continents, we assembled high-resolution digital maps of the global distribution of biomass fractions of AM, EcM, ErM and NM plants. Building on these maps, we assessed: (i) the amount of aboveground biomass carbon currently stored in each type of mycorrhizal vegetation; (ii) the impact of conversion of natural ecosystems to croplands on the distribution of mycorrhizal types globally; and (iii) the relationships between relative abundance of AM and EcM plants in an ecosystem and soil carbon content in topsoil (0–20 cm), medium (20–60) and deep (60–100 cm) subsoil layers.",
"discussion": "Results and Discussion Assembly of mycorrhizal vegetation maps To generate global maps of mycorrhizal vegetation, we estimated biomass fractions of AM, EcM, ErM and NM plants within each combination of continent × ecoregion × land cover type. Supplementary Fig. 1 illustrates the data assembly processes for the maps. Ecoregions follow Bailey 32 (Supplementary Data 1 ), and land cover types were retrieved from the ESA CCI land cover map 33 , which specifies cover and biomass fractions of trees, shrubs and herbaceous plants (Supplementary Data 2 ). For each combination, we determined the dominant species or group of species from 1568 vegetation surveys (Supplementary Data 3 ). For these species, we determined mycorrhizal type using the FungalRoot database v1.0 34 (see Supplementary Data 4 for data sources). Integrating these data, we obtained mycorrhizal plant biomass fractions of AM, EcM, ErM and NM plants for each combination of Bailey ecoregion, continent, and land cover type (Supplementary Data 5 and 6 ). These fractions were overlain on a global grid. Our maps (Fig. 1 ) provide quantitative estimates of the distribution of aboveground biomass fractions among AM, EcM and ErM plants within areal units of 10 arcmin. The use of a detailed map of ecoregions 32 provides much greater resolution compared with the biome-based patterns of mycorrhizal distributions reported by Read 3 > 25 years ago, whereas the land cover map 33 enabled us to provide accurate spatial positioning of ecosystem boundaries based on satellite-derived data, explicitly taking into account human-driven transformations of vegetation. Fig. 1 Percentage of aboveground plant biomass of mycorrhizal vegetation. a Arbuscular mycorrhizal plants, b ectomycorrhizal plants, c ericoid mycorrhizal plants, and d non-mycorrhizal plants. The map resolution is 10 arcmin. See Supplementary Fig. 4 for associated uncertainty values. Source data are provided as a Source Data file We validated the map data using four independent datasets: (i) forest biomass structure for Eurasia 35 , (ii) a global dataset of forest biomass structure used for an analysis of mycorrhizal impacts on carbon vs nitrogen dynamics 19 , (iii) estimates of mycorrhizal associations in the USA based on satellite remote sensing 36 , and (iv) West Australian map of mycorrhizal root abundance 24 (Supplementary Fig. 2 ). This validation revealed that the vast majority of the data (87% of the AM data points and 89% of the EcM data points) deviate by < 25% from the measurements 19 , 35 , 36 , when excluding ESA land use classes comprising poorly resolved combinations (i.e. mixed classes of land cover types 33 , such as “Tree cover, broadleaved, evergreen, closed to open (>15%)”, see Methods for details) that were difficult to couple to our classification scheme. The relationship between the validation data and our estimates is shown in Supplementary Fig. 3 . Our maps of mycorrhizal vegetation were assembled based on multiple published datasets, using a number of conversion factors to obtain per pixel values of mycorrhizal plants biomass fractions. These conversions as well as the fact that the plant species distribution data (Supplementary Data 4 ) originates from multiple sources constitute important uncertainty sources in our dataset. We examined the uncertainty of our maps based on uncertainties of tree, shrubs and herbaceous plant fractions within the land cover types 37 , and the number of data sources used to assess mycorrhizal fractions of plant biomass within each combination of Bailey ecoregion × continent; see Methods for details. Supplementary Fig. 4 shows spatial distribution of uncertainties. The mean uncertainties of AM, EcM, ER and NM maps are 19.6, 17.6, 14.6 and 15.0% at the 90% confidence interval. Overall, tropical areas have the highest uncertainties of the mycorrhizal fraction data, reaching 50% (AM) in the Amazon region. Therefore, our maps should be used with caution for these areas. Future sampling efforts of mycorrhizal vegetation distribution should be more focussed on tropical areas of Asia, Africa and South America. Mycorrhizal vegetation and aboveground carbon stocks By linking our maps of mycorrhizal vegetation to satellite observations of global aboveground biomass carbon 38 , we estimated the amount of aboveground biomass carbon stored in arbuscular, ecto-, ericoid and non-mycorrhizal vegetation as 241 ± 15, 100 ± 17, 7 ± 1.8 and 29 ± 5.5 GT (mean values ± uncertainty at 90% confidence interval; Fig. 2 ). In this analysis, the data were scaled to a resolution of 15 arcmin to match biomass estimates 38 . Most of the aboveground carbon stock stored in arbuscular mycorrhizal vegetation is situated in tropical forests (Fig. 2a ). Supplementary Table 1 shows per-biome distribution of the carbon stocks among mycorrhizal types. Fig. 2 Amount of carbon stored in plant biomass in vegetation of different mycorrhizal types (Mt C per-grid cell of 15 arcmin). a Arbuscular mycorrhizal plants, b ectomycorrhizal plants, c ericoid mycorrhizal plants, d non-mycorrhizal plants. The amount of aboveground biomass carbon stored in arbuscular, ecto-, ericoid and non-mycorrhizal vegetation is 241 ± 15, 100 ± 17, 7 ± 1.8 and 29 ± 5.5 GT (mean values ± uncertainty at 90% confidence interval), respectively Impacts of land transformations on mycorrhizal vegetation Agricultural practices drive the replacement of natural vegetation by facultatively AM crops 1 , 39 , which could also be de facto non-mycorrhizal due to destruction of hyphal networks by ploughing and excess fertilisation 40 , 41 . Using past vegetation estimates, Swaty et al. 26 showed that across conterminous USA, agriculture has reduced the relative abundance of ectomycorrhizal plants compared with other mycorrhizal types. However, global quantifications of agricultural impacts on distribution of mycorrhizas have not been possible until now. Based on the current land use data underlying our maps (Supplementary Data 4 ), we assessed mycorrhizal distributions on Earth in the absence of croplands. For each ecoregion-continent-land cover combination that contained croplands, we replaced current biomass fractions by estimates of per-grid cell biomass fractions in AM, EcM, ErM and NM plants that would be expected at these locations based on natural vegetation types (see Methods for details, and Supplementary Data 7 – 8 for data). Based on these data, we generated maps presenting potential natural distributions of biomass fraction of AM, EcM, ErM and NM plants in a cropland-free world (Supplementary Figs. 6 and 7 ). The current biomass fractions of AM plants have increased in Europe, parts of Asia and North America, but declined in Africa, Asia (mostly India) and South America, coinciding with increase in non-mycorrhizal vegetation (Fig. 3 , Supplementary Fig. 8 ). Our analysis suggests that EcM biomass has declined in all continents, primarily due to a replacement of natural forests by agricultural lands, whereas ErM biomass has remained unchanged. Fig. 3 Changes in biomass fractions of mycorrhizal vegetation induced by crop cultivation and pastures. a Arbuscular mycorrhizal plants, b ectomycorrhizal plants, c ericoid mycorrhizal plants, d non-mycorrhizal plants. Purple colours indicate losses, green colours indicate gains. Uncertainties are shown in Supplementary Fig. 7 . Source data are provided as a Source Data file Biomass fractions of mycorrhizal types and soil C stocks Recent field research in the US temperate forests suggests that soil carbon content increases with increasing EcM abundance in topsoil layers; but, depending on forest type, this relationship may be reversed in deeper soil 16 , 42 . This is in agreement with the Microbial Efficiency-Matrix Stabilization hypothesis, which predicts that ecosystems with rapid decomposition, such as most AM-dominated forests 19 , enhance soil organic matter (SOM) stabilization by accelerating the production and deposition of microbial residues 43 – 45 . To separate the effects of biome and mycorrhizal type on soil C on a global scale, we modelled the relationships between soil carbon content, biome type 46 , and biomass fractions of AM and EcM plants. We did not analyse the relationship between ErM plant cover and soil C content, due to a small proportion of the ErM plant biomass in the majority of ecosystems. We conducted separate analyses for the topsoil (uppermost 20 cm soil layer) and subsoil (20–60 and 60–100 cm soil layers), as obtained from the ISRIC-WISE Soil Property Databases, at a resolution of 30 arcsec 47 . The data sources used for these analyses are independent: the ecoregion classification, and hence mycorrhizal type distribution, does not account for edaphic parameters, whereas soil C data are unrelated to that of vegetation. Model comparisons were based on the Akaike information criterion (AIC). The relative importance of each predictor was examined using the Lindemann-Merenda-Gold (LMG) metric, providing the fraction of variance explained by each predictor, within the total variance explained by the model. Our global analysis revealed a moderately strong positive relationship between the biomass fractions of EcM plants and both topsoil and subsoil carbon (Fig. 4 , Supplementary Figs. 8 and 9 ). Consistent with the current paradigm 45 , cf. ref. 11 , our analysis revealed that biome is the main predictor of soil carbon stocks. Even so, the grid cell biomass fraction of EcM plants still accounted for one third of the explained variation in both topsoil carbon and in subsoil (Table 1 ). The interaction between biome and EcM biomass fractions was significant ( P < 0.001) but only marginally important (LMG = 1%), suggesting that the increase in topsoil carbon along with an increase in EcM plant biomass is mostly independent from the environment (Table 1 , Fig. 4 , Supplementary Figs. 8 , 9 , Supplementary Table 2 ). The total aboveground carbon stock stored in the EcM plants was a relatively worse predictor of soil carbon stocks compared with the EcM relative biomass fraction, showing in all cases higher AIC and lower R 2 . In contrast to EcM, AM biomass fractions per-grid cell showed negative but inconsistent relationships to soil carbon, with a contrasting positive trend in tundra (Fig. 4 , Supplementary Fig. 10 ). The latter relationship explained only a small amount of variance in tundra soil carbon stocks (12% in top 0–20 cm, 1.7% in 20–60 cm layer and 0.2% in 60–100 cm layer), and could arise from less accurate data of soil C in tundra 47 due to local landscape heterogeneity and prevalence of facultative AM plants with no or low mycorrhizal colonization 48 . Fig. 4 Quantitative relationships between topsoil (0–20 cm) C and biomass fraction of mycorrhizal vegetation in natural ecosystems. a EcM plants and b AM plants. The outcomes of individual models are presented in the Supplementary Table 2 . Croplands were excluded from the analysis. Per-biome predictions are shown in different colours. Source data are provided as a Source Data file Table 1 Summary of generalized linear models (glm) predicting soil carbon stocks Predicted variable Model \n R \n 2 \n Predictor P -value LMG (%) Topsoil C 0–20 cm EcM + Biome + EcM × Biome 0.53 EcM <0.001 42 Biome <0.001 57 EcM × Biome <0.001 1 Subsoil C 20–60 cm EcM + Biome + EcM × Biome 0.38 EcM <0.001 39 Biome <0.001 60 EcM × Biome <0.001 1 Subsoil C 60–100 cm EcM + Biome + EcM × Biome 0.33 EcM <0.001 35 Biome <0.001 64 EcM × Biome <0.001 1 Topsoil C 0–20 cm AM + Biome + AM × Biome 0.54 AM <0.001 38 Biome <0.001 56 AM × Biome <0.001 6 Subsoil C 20–60 cm AM + Biome + AM × Biome 0.33 AM <0.001 29 Biome <0.001 67 AM × Biome <0.001 2 Subsoil C 60–100 cm AM + Biome + AM × Biome 0.32 AM <0.001 31 Biome <0.001 67 AM × Biome <0.001 2 Predictions are made for C at 0–20, 20–60, and 60–100 cm depth and are based on biome and fraction of EcM or AM plants in vegetation biomass. R 2 —Cragg and Uhler’s pseudo R 2 . LMG—relative importance of individual predictors in a model examined through the Lindemann-Merenda-Gold metric. The LMG shows the percentage of variance explained by each of model predictors within the entire variance explained. The P values show the outcome of ANOVA type I models ( n = 78883 in all models). Source data are provided as a Source Data file Our analysis of the relationships between mycorrhizal type and soil carbon is based on ancillary maps, which feature large uncertainties. Analyses of relationships between ISRIC-WISE predicted soil carbon and the original data that were used to generate the ISRIC soil map yield R 2 -values in the range of 0.4–0.6 49 , 50 . This uncertainty adds ambiguity to our analysis, and reduces the reliability of quantitative estimates of the relationships between EcM plant biomass fractions and soil C. However, these uncertainties equally apply to the analysis of AM vs soil C as well as to that of the EcM vs soil C, due to the fact that the analysis is based on the same geographical data points. Therefore, we consider that the high uncertainty of the ISRIC-WISE soil data is unlikely to affect the qualitative nature of our conclusion that AM and EcM vegetation differently relate to soil C. Implications Our work provides the quantitative estimates of the global biomass distribution of arbuscular, ecto-, ericoid and non-mycorrhizal plants, accounting for human-induced transformation of habitats. Previous research has shed light onto the distribution patterns of mycorrhizal plant and fungal species richness 27 , 28 , and onto low-resolution (1 arcdegree) distribution patterns of mycorrhizal trees 51 . In contrast, our maps directly reflect the global distribution of biomass fractions of mycorrhizal plants across all biomes and all main vegetation types. Availability of such data at high resolution of 10 arcmin provides an opportunity for multiple potential analyses aimed at unravelling mycorrhizal impacts on ecosystem functioning at large-geographical scales. Our maps were derived at spatial resolutions allowing identification of the global patterns of mycorrhizal distributions and are most appropriate for global and large-geographical scale analyses of mycorrhizal impacts on ecosystem functioning and global drivers thereof. Recent estimates suggest that the total soil carbon loss due to agricultural practices accounts for 133 GT, with great acceleration of losses during the past 200 years 52 . Our analyses accounting for pre-agricultural patterns in EcM plant distribution point to large-scale losses of ectomycorrhizal vegetation, with potentially strong effect on the amount of C stored in soils. Analyses of agricultural impacts presented in this paper are based on the assumption that these impacts are limited to shifts in plant species composition, and do not encompass shifts in soil water and nutrient availability, which could affect activity of mycorrhizal fungi, depth distribution of mycorrhizas in the soil, and shifts among mycorrhizal fungal species composition due to the introduction of exotic species. While such simplifications are necessary for the analyses reported in this paper, they should be considered when interpreting our results. Furthermore, our analyses do not address other human impacts that can lead to shifts among AM and EcM vegetation, such as climate change, introduction of invasive species and nitrogen deposition. The latter is known to be an especially important driver of mycorrhizal vegetation shifts 42 , 53 , as it negatively affects abundance of ectomycorrhizal fungi in soil 54 . Given that nitrogen deposition leads to replacement of ectomycorrhizal plants by arbuscular and non-mycorrhizal vegetation, it further enhances agricultural impacts on soil carbon losses. The question whether increased domination of EcM plants in an ecosystem is associated with higher soil carbon content across both top- and subsoil is heavily debated 11 , 12 , 16 , 18 – 20 , 30 . However previous studies have been based on a limited number of observations 11 or regional-scale analyses 11 , 12 , 16 , 53 , 55 . Our analysis shows that across large geographical scales, higher cover of EcM vegetation is broadly associated with greater soil C stocks in both topsoil and subsoil, while AM vegetation has more variable, weaker and mostly negative relationships. This analysis does not provide evidence of causality of this relationship as multiple environmental variables such as climate, soil nutrients, especially nitrogen availability, and soil texture may affect both soil carbon and mycorrhizal plant distributions. Nonetheless, our study establishes a quantitative framework to test the relation between the dominance of mycorrhizal types and soil C stocks. Complete and directional understanding of the complex nature of the hierarchy of environmental drivers controlling soil carbon patterns requires further detailed (experimental) investigations of the hierarchy of different predictors and importance of local edaphic variables. Our estimates of the carbon stocks in AM and EcM aboveground biomass together with the quantitative relationships between soil carbon stocks and AM or EcM plant dominance in ecosystems provide qualitative insights into the global carbon cycle, highlighting the substantial role of mycorrhizas therein. Tropical forests, usually dominated by AM symbiosis (Fig. 2 ), contain 162 GT (44%) of global aboveground biomass 38 , whereas the predominately EcM temperate and boreal forests altogether store only 21% of global aboveground biomass carbon 38 , indicating that contribution of EcM vegetation to the aboveground biomass carbon is relatively small. In contrast, belowground carbon stocks are positively correlated to the proportion of EcM plant biomass, suggesting that mycorrhizal contribution to large carbon stocks in these regions occurs primarily through the carbon supply to belowground organs and mycorrhizal fungi, which is further emphasized with slowed decomposition processes 22 . Furthermore, our analyses revealed a relatively stronger relationship between soil C and EcM plant biomass fraction (%) than between soil C and the amount of carbon stored in EcM plants. These findings suggest that belowground carbon allocation by plants through mycorrhizal pathways is not directly proportional to the aboveground plant biomass, supporting the view of the importance of mutualism-parasitism trade-off in ectomycorrhizal associations 56 across biomes. Taken together, this study provides a benchmark for relating ecosystem processes to the functioning of distinct types of mycorrhizas on a global scale. So far, quantitative global information about mycorrhizal distribution was virtually absent despite the high demand for such data 12 , 13 , 30 . In spite of some uncertainty, our mycorrhizal distribution maps provide an essential source for systematic analyses of mycorrhizal biogeography and environmental drivers. Because our maps are based on field data, and not on a machine-learning model trained with environmental variables, they provide independent data for examining the relationships between mycorrhizal status and ecosystem functioning, without introducing a circular reasoning caused by the use of common environmental variables. Inclusion of mycorrhizal distribution into vegetation models would provide a benchmark for testing hypotheses about mycorrhizal impacts on ecosystem functioning and related ecosystem services. Our maps enable quantifying relationships between mycorrhizal abundances in ecosystems as well as soil and vegetation carbon content in global-scale analyses of biogeochemical cycles. In particular, the results of our study suggest that restoration of native vegetation especially in abandoned agricultural and barren land may help alleviate anthropogenic soil carbon losses and ameliorate increases in atmospheric greenhouse gases."
} | 6,361 |
35494353 | PMC9043574 | pmc | 2,722 | {
"abstract": "Despite being widely investigated for their memristive behavior, ferroelectrics are barely studied as channel materials in field-effect transistor (FET) configurations. In this work, we use multilayer α-In 2 Se 3 to realize a ferroelectric channel semiconductor FET, i.e. , FeS-FET, whose gate-triggered and polarization-induced resistive switching is then exploited to mimic an artificial synapse. The FeS-FET exhibits key signatures of a synapse such as excitatory and inhibitory postsynaptic current, potentiation/depression, and paired pulsed facilitation. Multiple stable conductance states obtained by tuning the device are then used as synaptic weights to demonstrate pattern recognition by invoking a hidden layer perceptron model. Detailed artificial neural network (ANN) simulations are performed on binary scale MNIST data digits, invoking 784 input (28 × 28 pixels) and 10 output neurons which are used in the training of 42 000 MNIST data digits. By updating the synaptic weights with conductance weight values on 18 000 digits, we achieved a successful recognition rate of 93% on the testing data. Introduction of 0.10 variance of noise pixels results in an accuracy of more than 70% showing the strong fault-tolerant nature of the conductance states. These synaptic functionalities, learning rules, and device to system-level simulation results based on α-In 2 Se 3 could facilitate the development of more complex neuromorphic hardware systems based on FeS-FETs.",
"conclusion": "Conclusion In summary, we fabricated and demonstrated synaptic devices that use multilayer ferroelectric α-In 2 Se 3 as the channel in a back-gated FET configuration with Ta 2 O 5 as the dielectric. Devices exhibited excellent nonvolatile resistance switching modulated by source-drain voltage ( V ds ) and back-gate voltage ( V G ), allowing it to be used as both memristors as well as a synaptic device. The gradual change in conductivity of the channel through partial polarization switching between the In 2 Se 3 layers is an important step for studying the adaptive learning in ANNs. The potentiation and depression properties in FeS-FET were measured by applying incremental excitatory and inhibitory bias pulses that exhibit excellent weight-update properties, with nonlinearity as low as 0.12, appropriate variation margin (Δ G = 7–12) a low energy consumption of 10 pJ per spike and a large number of stable conductance states(>64). Leveraging these synaptic parameters, an artificial neural network was simulated corresponding to different N states (10, 20 and 64) that showed 93% recognition accuracy in recognizing handwritten digits from the MNIST database. Introduction of a noise (variance value of 0.10 in) in background pixels showed accuracy of more than 70% indicating strong fault-tolerant nature of the conductance states. Using 2D α-In 2 Se 3 as a ferroelectric channel in synaptic transistor has shown high performance in stable conduction states, low power consumption of 10 pJ per spike, and high endurance. These synaptic functionalities, learning rules and device to system-level simulation results based on α-In 2 Se 3 are expected to facilitate the development of more complex neuromorphic hardware systems based on FeS-FETs. Characterization techniques All the electrical characteristics of the FeS-FET were measured under ambient conditions and room temperature using a probe station (Agilent B1500 Semiconductor Parameter Analyzer). Presynaptic spikes were applied on the back-gate electrodes, and postsynaptic current output was measured by applying a voltage ( V ds = 0.10 V) between the source and drain electrodes. The thickness of Ta 2 O 5 was measured by spectroscopic ellipsometer (M-2000, J.A. Woollam). The structural properties of the α-In 2 Se 3 flake were measured using FEI TEM Themis (60–300). The thickness of flake was measured in tapping-mode using Bruker Scanasyst atomic Force Microscopy. All ANN simulations were performed using MATLAB and Python.",
"introduction": "Introduction Advancements in neuromorphic computing techniques in conjunction with brain-inspired hardware platforms are leading to an emergence of next-generation energy-efficient computing systems which can potentially outperform conventional von Neumann computers. 1,2 Solid-state devices that mimic the biological synapse and neurons have been attracting interest 3–5 in building Artificial Neural Networks (ANN) towards efficient and adaptive implementation of cognitive tasks like classification, speech, and pattern recognition. The basic principle in such tasks is based on weight optimization associated with the individual neurons to achieve excellent efficiencies in supervised learnings. 6 For an artificial synaptic device to be effectively used in an ANN, it is desirable that it includes linearly spaced conductance weight states, high endurance, longer retention and low energy consumption. To date, multiple device concepts have been proposed to realize artificial synapse, such as resistive random access memristors (RRAMs), 7,8 phase-change memory (PCM), 9 charge trapping transistors, 10 ion movement in electrolytes, 11 and ferroelectric based devices with excellent performance and high integration capability. However, solid state devices are far from even remotely reaching the performance of actual biological synapse. For example, the random switching mechanism in RRAMs and phase-change memory PCM suffer from a significant variation in conductivities due to high cycle to cycle variation. Traditional charge-based memories use a charge trapping layer within a transistor configuration that requires a relatively high operating voltage, high thermal budget and suffers from a non-linear weight update. Electrolyte-gate transistors (EGTs), 12 exhibit better weight updating performance thanks to their gate modulation, yet they are not compatible with CMOS technology due to the liquid electrolyte-gate or organic channel material involved. In this context, ferroelectric materials have emerged as a promising candidate for enabling synaptic devices as they lead to fast operation, 13 nondestructive readout, 14 low-power, low variations, and high on/off ratios. 15–17 The partial polarization switching behaviors of the ferroelectric material maybe exploited to emulate the biological synaptic functions by gradually modulating the channel conductance through an external electrical field. In recent years there have been various studies using ferroelectric material such as PbZr 0.2 Ti 0.8 O 3 (PZT), 18 Hf 0.5 Zr 0.5 O 2 , 19 CuInP 2 S 6 (ref. 20 ) for different device designs such as capacitors, ferroelectric tunnel junctions (FTJ), and ferroelectric field-effect transistors (FeFETs). 21 However, among all the reported studies, ferroelectric materials have been studied primarily in metal–semiconductor–metal geometry or as a dielectric layer in a transistor to modulate channel conductance. For example, Zheng et al. chose PbZr 0.2 Ti 0.8 O 3 (PZT) as a thick (100 nm) ferroelectric gate dielectric, demonstrating the polarization switching albeit at a high operating voltage (∼10 V). Wang et al. , have shown a fully organic electrochemical synapse using a ferroelectric dielectric for sensory memory system but is limited by stability issues being organic material. 22 Considering the synaptic properties by HfO 2 based films are demonstrated at much lower thicknesses and operating voltages, these are limited by difficulty in fabricating and decrease in residual polarization with cumulative switching cycles, resulting in memory with limited endurance. However, ferroelectrics as a channel material in a FET configuration for mimicking the synapse are barely reported. 23,24 FeS-FET, or ferroelectric semiconductor FET 25 uses a ferroelectric material as the transistor channel, and is expected to offer more operational freedom to tune conductivity through partial polarization switching at much lower thicknesses. If the ferroelectric channel material happens to be a layered 2D semiconductor, the promises for superior and/or novel functionalities are expected to multiply. This is because layered 2D materials enable aggressive atomic-layer scaling, ultralow power consumption and provide unique advantage: these can be transferred on almost any platform via weak van der Waals bonding which circumvents lattice mismatch issues in traditional heterostructures. This opens up the possibility of realizing heterogenous integration toward system scaling. Devices based on layered 2D semiconductors have been extensively studied toward achieving synaptic functionalities as well as demonstrating cognitive tasks. In this context, a 2D ferroelectric semiconductor such as indium selenide (In 2 Se 3 ) would provide an opportunity to exploit properties of both 2D and ferroelectric materials and could offer exciting avenues into neuromorphic hardware platforms. 26–29 Unlike traditional insulating ferroelectrics, In 2 Se 3 is a semiconductor with a direct bandgap of about 1.36 eV, making it attractive for conventional electronic and optoelectronic applications also. In this work, we use multilayer α-In 2 Se 3 as a channel material for FeS-FET with Ta 2 O 5 as a back-gated high- k dielectric. First, we demonstrate the memristive switching of In 2 Se 3 -based FeS-FET by exploiting the polarization switching in the material via the gate terminal. The use of high- k dielectric ( ε = 15) allows us to work at much lower operating voltages. 30 The devices can emulate various characteristics of an artificial synapse such as excitatory and inhibitory postsynaptic current (PSC), paired pulse facilitation/depression (PPF/PPD), and long-term potentiation/depression (LTP/LTD). We have also explored the continuous wight modulation through partial polarization of the channel displaying an excellent linear weight update trajectory with multiple stable conductance states. Finally, leveraging this device-level emulation of synaptic dynamics, the recorded data from a single device is converted to conductance weights for pattern recognition tasks. We achieved an accuracy of 93% using grayscale MNIST (Modified National Institute of Standards and Technology) datasets. This demonstration of ferroelectric channel transistors with synaptic functionalities and subsequent deployment in pattern recognition is expected to open new vistas in hardware neural networks using layered 2D ferroelectrics. Material characterization Indium selenide (In 2 Se 3 ) is known to exist in 5 different phases (α, β, γ, δ, and κ), which are formed by different arrangements of indium and selenium atoms within a single layer. 31 For this study, we have used α-In 2 Se 3, which is known to be ferroelectric at room temperature. A high-resolution transmission electron microscopy (HRTEM) imaging of α-In 2 Se 3 was carried out to study its crystal structure (see ESI Note 1 † for sample preparation). Fig. 1(a) shows the top view HRTEM image of the few-layered exfoliated α-In 2 Se 3 , the staggered atomic arrangement of the structural layer can be attributed to the multi-layered structure, which can be confirmed from the thickness contrast at multiple areas. The inset to Fig. 1(a) shows the FFT pattern of the highlighted area. The lattice parameter calculated is 0.36 nm, which confirms a hexagonal structure for α-In 2 Se 3 . Fig. 1(b) shows the enlarged image of the lattice fringes from Fig. 1(a) ; the high contrast of indium atoms depicts the hexagonal structure. Fig. 1(c) shows the selected area diffraction pattern of the sample. Fig. 1(d) shows the typical Raman spectra of a multi-layered α-In 2 Se 3 flake used in this study, which shows four clear peaks at 89, 105, 180, and 194 cm −1 . These peaks correspond to E , A (LO + TO), A (TO), and A (TO) phonon modes, respectively, which are attributed to reported room-temperature ferroelectricity in both out-of-plane (OOP) and in-plane (IP) directions displayed by α-In 2 Se 3 . 32 , 33 Fig. 1 (a) HRTEM image of multi-layered exfoliated α-In 2 Se 3 . The inset shows the FFT image of the selected region enlarged in (b). (c) Selected Area Diffraction Pattern (SADP) of the sample. (d) Raman Spectra of α-In 2 Se 3 measured at room temperature. (e) Schematic structure of α-In 2 Se 3 synaptic transistor with 50 nm Ta 2 O 5 and 10 nm Al 2 O 3 as a capping layer with source and drain terminal acting as pre and post neuron respectively as shown in the cartoon (f) The thickness of the flake measured using AFM (g) Atomic Force Microscopy (AFM) scan of the α-In 2 Se 3 based synaptic transistor. Fabrication and experimental details \n Fig. 1(e) shows the schematic of a FeS-FET used in this study where the gate and drain electrodes act as pre and postsynaptic terminals, respectively. 40 nm of Ta 2 O 5 dielectric film is deposited by magnetron sputtering on p + doped-Si substrate. Multilayer In 2 Se 3 flakes were exfoliated by using a conventional scotch tape approach from a bulk In 2 Se 3 crystal (bought from HQ graphene). Fig. 1(f) shows the AFM height scan of one such flake used for device fabrication. E-beam lithography was used to pattern source-drain contact fingers of Ni/Au (50/70 nm) while the channel length and width in the α-In 2 Se 3 flake were defined 2 μm and 4 μm, respectively. 20 nm of e-beam evaporated Al 2 O 3 was used as a capping layer to encapsulate the device. Fig. 1(g) shows an AFM scan of the final processed device.",
"discussion": "Result and discussion When measured for the lateral 2-terminal and 3-terminal electrical characteristics, the α-In 2 Se 3 devices exhibit clear hysteresis, indicative of polarization due to ferroelectricity. Fig. 2(a) shows the output electrical characteristics of a representative FET with the gate floating, exhibiting significant hysteresis in both forward and reverse biases. The direction of hysteresis is anticlockwise from high resistive state (HRS) to low resistive state (LRS). As indicated by the direction of the arrow for a positive voltage sweep from 0 to V max and turns to a low resistance state from a sweep voltage of V max to 0 V (sweep ii), thus acting as an HRS-LRS (LRS-HRS) memristive device for sweep i (iii) and ii (iv) respectively. The devices produce reproducible hysteresis for different ranges of sweep voltage, a typical characteristic of memristors. The notable asymmetric transport behavior observed in forward and reverse bias can be attributed to the ferroelectric polarization charges that enhance the Schottky barrier of one interface while reducing the barrier of the opposite interface as has been previously reported. 33 , 34 Under the 2-terminal configuration with the gate floating as discussed above, the hysteresis window and the switching ratio is found to gradually increase with an increase in maximum sweep voltage. (ESI Fig. S1 † ) shows that the switching ratio between HRS to LRS at V read of −0.5 V approaches 10 2 as the maximum sweep voltage exceeds 5–6 V. Fig. 2(b) shows the endurance characteristics of the FET that had been switched 100 times between the HRS and the LRS using full-sweep cycles from 0 to ±4 V (please see ESI Fig. S1 † for extended data). Fig. 2(c) shows the retention characteristics of FeS-FET at V G = 0 V and V D = 2 V with high (HRS) and low (LRS) stable states. Formation of built-in electric field due to the presence of mobile charges strengthens the polarization of the ferroelectric dipole and improves the endurance. The switching ratio between the LRS and HRS remains one order of magnitude for more then 1100 s showing the substantial endurance of the devices. Fig. 2(d) shows the typical n-type transfer characteristics of the back-gated FeS-FET with drain current of 2 μA μm −1 at a drain bias of V ds = 0.5 V and an on/off ratio of 10 5 . Forward and reverse transfer-curve sweeps evoked typical clockwise hysteresis, which occurs due to the partial polarization switching of the In 2 Se 3 layers (see extended electrical data in ESI Fig. S2 † ). As previously reported for a back-gated In 2 Se 3 transistor with high dielectric thickness, the vertical electric field is not enough to completely switch the polarization in the In 2 Se 3 channel. 25 The mechanism can be understood as follows: when the applied gate voltage is −6 V, positive bound charges get accumulated at the bottom of the α-In 2 Se 3 layer (oxide/semiconductor interface), and negatively bound charges get accumulated at the top layer of the channel. This is known as the polarization downstate as illustrated in Fig. 3(a) with a schematic energy band diagram. Fig. 2 (a) Output characteristics of FeS-FET exhibiting ferroelectric hysteresis with maximum sweep range of V D varying from (−3 V, 3 V) to (−6 V, 6 V). The sweeping directions are indicated by the arrows. (b) Resistance switching between the HRS and LRS over 100 cycles. The amplitude of the write voltage were set at 2 V. (c) Retention test, in which the ratio of the HRS and LRS states remained over one order of magnitude for up to 1100 s. (d) Transfer characteristics and gate leakage of In 2 Se 3 FeS-FET showing on/off current ratio of 10 5 . Fig. 3 (a and b) Schematics showing the device operation and corresponding band diagrams, of the synaptic transistor explaining clockwise hysteresis loop achieved due to partial polarization switching between multiple layers. As the gate voltage is swept from −6 V towards positive voltage, the free carriers in the channel cause an increase in electron density in the bottom layer giving rise to an increase in channel conductance. With a further increase in gate voltage, the ferroelectric polarization in the bottom layer starts changing sequentially from downward to upward, but the vertical field is not high enough to flip the polarization of the top layer completely. The top layer remains in a downward polarization state, giving rise to domain walls in the channel. A situation known as partially polarized switching emerges due to this, and it is observed at +6 V in this case. With a further change in the gate voltage, the upward polarization in the bottom layer starts flipping, leading to a decrease in the channel conductance as shown in the band diagram [ Fig. 3(b) ]. The switching of polarization in our devices is gradual and not abrupt, as observed in other ferroelectric oxide films where the gate oxide (and not the channel) is ferroelectric. Low-temperature measurements were carried out on a representative α-In 2 Se 3 FeS-FET, exhibiting a similar hysteresis window as room temperature. This confirmed that the ferroelectric nature of α-In 2 Se 3 flakes is responsible for the hysteresis instead of charge trapping between the oxide and semiconductor interface. (The details of the measurements are shown in ESI Fig. S3 † ). To mimic various synaptic functionalities of the In 2 Se 3 based FeS-FET structure, single and sequentially triggered (or presynaptic) pulses are applied at the gate. The channel conductance increases or decreases depending upon the amplitude and the duration of the applied presynaptic pulse. The conductance of In 2 Se 3 is equivalent to the synaptic weight of the biological synapse, modulated by the gate voltages. When a single negative pulse with an amplitude of −2 V and duration of 40 ms at V D = 0.5 is applied at the gate of the transistor, a typical excitatory postsynaptic current (EPSC) of 20 nA is observed as the channel conductance increases, as shown in Fig. 4(a) . Since In 2 Se 3 is an n-type semiconductor, ideally the current should decrease as we apply a negative pulse but the reverse phenomenon is observed. This is due to the coupled IP and OOP ferroelectricity present in In 2 Se 3 layers. 33 On the application of a negative pulse, an additional in-plane electric field gets strengthened opposite to the built-in electric field leading to overall change in drain current, thus we observe a positive current for a negative pulse. Also, an inhibitory current (IPSC) of −3 nA is observed corresponding to a positive pulse with an amplitude of 2 V and duration of 40 ms, as shown in Fig. 4(b) . The corresponding insets show the EPSC and IPSC responses to the presynaptic pulses of different pulse amplitudes with the same pulse width of 40 ms, showing the increase or decrease in the channel conductivity as a function of the applied pulse amplitude. As the amplitude of the gate voltage pulse increases, the additional electric field inside the In 2 Se 3 starts increasing, providing a higher number of ferroelectric domains switching in the In 2 Se 3 channel layer. Hence, the conductivity of the channel increases or decreases with an increase in the gate bias, as testified by the current values. Fig. 4 (a) EPSC response to a programming gate pulse of −2 V. (b) IPSC response to an erasing gate pulse of 2 V. The corresponding insets show the EPSC and IPSC response of synaptic transistor at different back-gate voltages ranging from (−4 to +4 V). (c) Potentiation and depression responses of the channel to 100 sequential pulses at V LTP and V LTD of ±2 V and 30 ms for different devices. The upper and lower dashed lines represent the maximum and minimum values of the conductivity, and the solid line represents the average value of the conductance states observed for different devices with minimal variation. (d) Top part shows the pulse scheme applied at the back-gate terminal of the In 2 Se 3 synaptic transistor. PPF index determined by the time interval of two input presynaptic pulses (Δ t PPF = t pre2 − t pre1 ). Dashed lines indicate the fitting, (e) Sequential 30 excitatory postsynaptic current corresponding to gate pulses of different pulse widths (10 ms 30 ms and 50 ms) at fixed pulse amplitude of 2 V. Long-term synaptic plasticity, which consists of long-term potentiation (LTP) and long-term depression (LTD) of synaptic weights, is essential to learning and memory functions in human brain. 35 Implementation of long-term plasticity is an essential phenomenon for better accuracy and reliability in analyzing various cognitive tasks like pattern and speech recognition. In previous studies, ferroelectric-based memristive devices have demonstrated long-term plasticity by utilizing the dynamic ferroelectric domain evolution and robust domain stability. 36,37 In our FeS-FETs, long term potentiation/long-term depression are observed when trains of 100 gate voltage pulses V LTP (−2 V) and V LTD (+2 V), are sequentially applied as shown in Fig. 4(c) . Negative pulses triggered at the back gate lead to an increasing polarization switching in the In 2 Se 3 layer, enhancing the channel conductivity with increasing pulse number, which manifests as potentiation. Depression is the opposite of potentiation, and corresponds to a decrease in channel conductivity with increasing pulse number. The upper and lower dashed lines represent the maximum and minimum values of the conductivity, and the solid line represents the average value of the conductance states observed for different devices with minimal variation. We further confirm the emulation of paired pulse facilitation/depression (PPF/PPD), a characteristic that can be exploited for time-dependent learning algorithms. 38 In a biological synapse, the time interval Δ t between two successive presynaptic inputs affects the magnitude of the postsynaptic current. This dependence of the change of postsynaptic current on Δ t can be captured by measuring the channel current for two gate pulses of the same amplitude separated by different values of Δ t ( Fig. 4(d) ). In our devices, we applied the gate pulses of amplitude −2 V (or +2 V) while the drain bias is 0.1 V. PPF and PPD are quantified in an index called the PPF (or PPD) index, which is defined by the ratio of ( E 2 − E 1 )/ E 1 where E 2 and E 1 are the channel current values corresponding to the 2 nd and the 1 st gate pulses respectively. Time constants of 110 ms and 77 ms for depression and facilitation respectively, are estimated based on a double exponential fit to the data points. These characteristic time constants are comparable in scale to those of a biological synapse. Further, we studied spike rate dependent plasticity (SRDP) i.e. ; the impact of the pulse width for 30 sequential excitatory pulses with a fixed pulse amplitude of 2 V, as shown in Fig. 4(e) . The postsynaptic current decreases from 250 nA to 100 nA when the gate pulse width is reduced from 50 ms to 10 ms. As the pulse width increases, the In 2 Se 3 layers in the channel get more orderly polarized, resulting in higher current. As a synaptic device, the stability of the device is of critical importance. Thus, the endurance of the devices is examined for multiple, successive potentiation and depression cycles. ESI Fig. S4 † shows the endurance characteristics of the FeS-FET for 30 excitatory and inhibitory pulses at ±2 V, 30 ms. No degradation in performance is observed in the devices, as confirmed by the consistent stability exhibited up to 30 cycles. ESI Fig. S4 † also shows the endurance for 100 consecutive excitatory and inhibitory pulses for a total of 600 pulses with multiple stable conductance states, demonstrating the robustness of FeS-FET devices for ANN in neuromorphic computing. Next we studied the performance of our synaptic devices in the implementation of neural networks, i.e. , recognition accuracy (pattern recognition) of MNIST data digits. The recognition performance of memristive devices is affected by various parameters such as (a) weight conductance non-linearity (NL), (b) total variation margin (Δ G ) in conductance, and (c) conductance weight states ( N states ), 39,40 as discussed below. (a) Weight conductance non-linearity (NL) The conductance of a synaptic device usually increases steeply for the first few potentiation and depression pulses and then becomes saturated as the number of pulses increases. 5,41 Every pulse leads to a different response in terms of weight modulation trajectory, and its cumulative effect does not follow a simple linear relation. This is known as non-linearity (NL). The NL factor is used to analyze its effect on learning efficiency for gradual conductance changes. 42 Increasing the NL degrades the recognition accuracy due to the difficulty in the convergence of conductance weight into a stable value; a smaller NL is thus more desirable for better recognition accuracy. ESI Fig. S5 † shows the non-linearity observed in LTP and LTD curves for different pulse widths. Our device exhibits good linearity with NL_LTP (0.12, 0.31 and 0.81 for 10 ms 30 ms and 50 ms) respectively, with a near symmetric LTP and LTD curves. The observed values are amongst the lowest reported values using a ferroelectric channel for artificial synapse. (b) Total variation margin (Δ G ) in conductance Another critical factor which affects the accuracy of ANN-based implementation of cognitive tasks such as pattern recognition is Δ G defined by the minimum ( G min ) and maximum ( G max ) conductance weight variation. Previous studies on Δ G have concluded that a more extensive variation margin provides more analog states to store information but at the cost of recognition accuracy, especially for Δ G >20. Thus, it is necessary to optimize the synaptic device specifications for improved system-level performances. 40,43 For the FeS-FET synaptic device under study, we observed a total variation margin of 7, 9 and 12 for 10 ms 30 ms and 50 ms respectively, for a hundred sequential excitatory pulses. A comprehensive study of various parameters like timing between two consecutive pulses and pulse amplitude is required to improve the values of non-linearity and variation margin values to study their effect on learning accuracy. (c) Conductance weight states ( N states ) In neuromorphic computing, a higher number of conductance weight states guarantees reliability in data retention or endurance. 44,45 In this study, we have shown that even 20 conductance weight states are sufficient to provide stable recognition accuracy of 85%. The multiple conductance states in our devices could be attributed to the controlled polarization of individual layers in the semiconducting channel. For the α-In 2 Se 3 FeS-FET under study, ESI Fig. S6 † shows more than 64 stable conductance states obtained after applying a hundred excitatory pulses of magnitude 2 V and 10 ms duration. The value of conductance state at each pulse is increasing in linear fashion, with a few conductance states either overlapping or less than the previous pulse value. In order to avoid this overlap, we extracted the stable conductance states by setting threshold value for Δ G . The effective stable conductance states were extracted using ( G N − G N−1 ) conductance weight values when the difference between G 1 and G 2 exceeds a certain threshold. Here G 1 and G 2 are conductance weight values from pulse-1 and pulse-2, respectively. A threshold of 0.2%, 0.3% and 0.5% was used to extract the stable weight values from 100 pulses. ESI Fig. S6 † shows the extracted values of three different cases, corresponding to N states = 10 (case-1), N states = 20 (case-2) and N states = 64 (case-3), to identify its impact on pattern recognition. It can be seen that our device shows high number of conductance states (>64) with much less non-linearity in LTP confirming potential of our FeS-FET in ANN for future high performance neuromorphic computing. Benchmarking of various synaptic parameters reported in the literature for different synaptic devices vis-à-vis the FeS-FET in this study is made in Table 1 . Benchmarking with state of art Synapses Devices HfAlyOx ReRAM 47 HZO-Si FINFET 45 Charge trapping transistor 48 GrFeFET 49 HfZrOx-FeFET 50 CNT-transistor 51 FeS-FET (this work) \n G \n max / G min 25 4.98 — — 45 57 12 NL >5 1.58–7.57 0.06–0.89 2.8 1.75–1.46 0.82 \n 0.1–1.8 \n Conductance states ∼25 >32 35 — 32 — \n >64 \n Operating voltage 2–8 V 3.2–3.7 V 10 V 16 V 3.4–3.7 V 1–5 V \n 1–2 V \n Neural network 320–3 528–250–125–10 400–100–10 1000–100–10 400–100–10 784–10 784–256–10 Number of layers 2 3 2 2 >2 2 2 Total training data Yale face database 100 000 MNIST 60 000 MNIST 60 000 MNIST 100 000 MNIST 60 000 MNIST 60 000 MNIST Energy consumption 30 nJ 12.1 pJ — 50 nJ — 10 nJ \n 10 pJ \n Recognition accuracy ∼0.90 0.8 0.87 0.94 0.9 0.7 0.93 In this study, we have investigated the impact of N states of the α-In 2 Se 3 synaptic device to assess its applicability to pattern recognition of the system. We present detailed artificial neural network (ANN) simulations performed on the Modified National Institute of Standard and Technology (MNIST) database to demonstrate neuromorphic computing. Here, we categorize 28 × 28 pixels of the MNIST dataset using a single-layer perceptron model, performing supervised learning. As explained in Fig. 5(a) , the 784 input neurons ( X 1 to X 784 ) correspond to 28 × 28 pixels of an original image, and the 10 output neurons relate to 10 classes of digits ( Z 0 to Z 9 ). During the simulation, 42 000 patterns were used as a training dataset, and 18 000 images were used to test the recognition accuracy of each epoch/training cycle. As illustrated in flowchart Fig. 5(b) , every training cycle is divided into two halves, i.e. , forward propagation and backward propagation. During forward propagation, each neuron in the input layer receives a value corresponding to a pixel in the image and is assigned to an input vector ( X i ), which is then transformed into 256 hidden layer neurons through a weight matrix ( W i , j ) and further down to 10 output neurons ( Z i , k ) through output layer weight matrix ( V j , k ). The summation of weights and input vector are converted to output vector by a sigmoid activation function. Next, during backward propagation, we fine-tune the weights ( W i , j , V j , k ) based on the difference between the output value and the label value. The trained input ( W i , j ) and output ( V j , k ) weight values obtained after reaching maximum efficiency are then replaced by synaptic conductance weight values ( N states ) of the FeS-FET under study. ESI Fig. S5 † shows the learning accuracy using the different number of middle layer neurons. The higher the number of neurons, the higher the accuracy achieved. Using 256 middle neurons, a linear transformation of the device conductance values is performed so that the conductance range is consistent with the weight range using the following relation 46 1 C j = AI j + B where C j represents the weight value after the linear transformation and A and B are linear transformation coefficients. In the case of 64 weight states, the linear transformation coefficients were A 1 = 2.68 × 10 6 , A 2 = 7.30 × 10 6 and B 1 = −5.473 and B 2 = −1.5083. Details are provided in the ESI † for extracted values corresponding to N states = 10 and 20. The minima for each C j value in the weight matrix is then calculated using | W i , j − C j | and | V j , k − C j |; the new matrix thus obtained has all the values corresponding to the conductance values of the device. Fig. 6(a) shows the recognition accuracy using the updated weight values corresponding to different conductance states ( N states ) for 120 training cycles. The higher the number of stable conductance weight states, the higher is the accuracy obtained. Case-3 starts showing a stable recognition accuracy after 20 training cycles. Case-2 shows higher recognition accuracy of 85% than case-1 (75%), but both case-1 and case-2 follow a similar trend of switching between high and low numbers at lower training cycles in classifying the accuracy. This shows that the effective number of conductance weights ( N states = 10 and 20) is sufficient to achieve stable recognition accuracy at higher training cycles. As the number of training cycles increases, the synaptic weight values get further improved for numerical pattern recognition. Fig. 6(b) shows the reconstructed and visualized pattern corresponding to synaptic weights with case-3 after 10 th and 240 th training cycles. Fig. 5 (a) Artificial neural network comprises of three layers, containing 783 input neurons, 256 middle (hidden) layer neurons, and 10 output neurons. (b) Flowchart of the training recognition cycle, where N represents the total number of training images, i is in the range 1–784 input neuron, j is in the range 30–256 hidden layer neuron, k is in the range 1–10 output neuron, and these indices imply the sequence number of the input neurons, middle (hidden) layer neurons and output results, respectively. Fig. 6 (a) Recognition accuracy as a function of the training cycle for different N states (10, 20, and 64). (b) Weight mapping images after 10 th and 240 th training cycles. (c) Average confusion matrix of the testing data set for 18 000 MNIST data. The digits highlighted in dark color represents the number of identified digits for all three cases, whereas those with lighter color code shows the confusion between two digits, for example digit 2, 3 for case 1. (d) Recognition accuracy as a function of the training cycle for different noise pixels introduced in the test data. Numerical digit 4 represents different % of variance introduced as background noise pixels for calculating the classification accuracy. Introducing background noise leads to difficulty in convergence state of conductance weight value. Next, we studied the average confusion/error created by the proposed algorithm for different conductance weight values. Fig. 6(c) shows the average confusion matrix over the 10 digits of the MNIST test data for all three cases. Given the recognition accuracy was 75% for case-1 and 85% for case-2, most of the input digits were identified correctly in both the cases, with the most common confusions for digits 2, 3, and 5 being higher for case-1 compared to case-2. It was found that case-3 corresponding to N states = 64 exhibited less confusion in precisely identifying the input digits than the other two cases. Besides this, we studied the learning accuracy with 64 conductance weight states by introducing noise to the background pixels. Gaussian noise with mean μ = 0 and variance σ = 0.01, 0.05, 0.10, 0.20 is introduced to the background pixels to check the stability of the weight states ( Fig. 6(d) ). More than 70% accuracy is achieved for a variance value of 0.10, which indicates that the conductance weight states hold great potential in the classification of the digits. This establishes the fault tolerance nature of the α-In 2 Se 3 FeS-FET. The accuracy can be further improved by increasing the training cycles but at the cost of higher energy consumption involved in simulations. Further, the energy efficiency of the proposed FeS-FET based synaptic device is evaluated. The energy consumption per pulse is governed using the following relation: E = I × V × t , where ( V ) indicates the programming voltage, ( I ) the channel current, and ( t ) the pulse width. The energy consumption per switching of the device can be lowered significantly by reducing the pulse width. The smallest value is estimated to be 10 pJ ∼ (1 V, 10 ms), which is lower than that of graphene/Fe-FET (∼50 nJ) 49 and can be further decreased by decreasing the pulse widths to an even lower value. The learning accuracy of FeS-FET synapse is benchmarked against other ReRAM, FINFeT, Fe-FET and charge trapping transistors in Table 1 . FeS-FETs exhibit a clear accuracy advantage compared to other ReRAM and charge trapping transistor synapses. The high recognition accuracy (93%) results from a low non-linearity factor, reasonable Δ G variation and high number of conductance states. The low non-linearity (0.12) values can be attributed to ordered partial polarization switching mechanism in different layers of In 2 Se 3 thus providing higher number of conductance weight states compared to ReRAM and FiNFET. The fabricated FeFET synapse exhibits (>64) analog states that can be modulated symmetrically (potentiation and depression) using applied electric field, whereas in charge trapping transistors, there is a linear increase in conductance weights for initial spikes which gets saturated as the number of pulse increases. The Δ G values extracted for different pulse schemes are compared with other reported values in Table 1 . The required variation margin should be higher than 10 to achieve accuracies of >80%. 8 Compared to (ref. 48 ), we achieved 85% accuracy with as low as 20 conductance states providing a reasonable balance between accuracy and variation margin (Δ G ∼9). It should be noted that it is unclear what NL and Δ G values are appropriate for the reliable operation of a neuromorphic system. In other words, an investigation on the impact of high/low NL and Δ G is necessary, as is particularly evident when investigating the effect on the pattern recognition accuracy. Though the recognition accuracy observed for graphene based FeFET is higher then our FeS-FET, but they are operated at much higher voltages (16 V) providing a disadvantage when using for neuromorphic computing systems. Nevertheless, the results presented here highlight the potential of using In 2 Se 3 based FeS-FET based non-volatile synaptic transitor for training of neural network that outperforms other devices in terms of non-linearity (0.12–1.8), high number of conductance states (>64), desirable Δ G , low energy consumption (10 pJ) and electric-field controlled switching."
} | 9,924 |
33444876 | PMC8012881 | pmc | 2,723 | {
"abstract": "Graphical abstract",
"conclusion": "Conclusions Fourteen distinct microbial reactions that transform nitrogen between its redox states form the basis of processes such as nitrification, denitrification and anammox that are applied for biological nitrogen removal. These reactions are carried out by a wide variety of physiologically diverse microorganisms found throughout the tree of life. The in-depth physiological understanding of the involved microorganisms is key to address persistent problems associated with nitrogen removal, in particular emission of climate-active gases and energy consumption. Moving forward, biological and physicochemical ammonia recovery should be considered as a sustainable nitrogen management approach compared to the energy-intensive conversion of industrially produced ammonium back to N 2 during wastewater treatment.",
"introduction": "Introduction Domestic and industrial wastewaters are a major source of anthropogenic nitrogen deposition into the environment. The main reactive nitrogen species in wastewater is ammonium, and the overall aim of biological nitrogen removal is the conversion of ammonium back into N 2 . To reach this aim and meet the strict discharge standards for nitrogen-containing wastewaters, diverse biological nitrogen removal systems have been developed. Furthermore, the continual discovery of new nitrogen-transforming pathways and microorganisms results in relatively less energy- and resource-intensive nitrogen removal strategies [ 1 , 2 ]. Whereas both conventional and novel processes achieve efficient nitrogen removal from wastewater, they still consume considerable amounts of energy and lead to the production of climate-active gases nitric oxide (NO) and nitrous oxide (N 2 O), which are turned over by many nitrogen-transforming microorganisms [ 1 , 3 •• , 4 ]. Currently, there are no discharge standards for climate-active gases produced during wastewater treatment; however, the inclusion of wastewater treatment plants (wwtp) as a source of greenhouse gases to the recently refined guidelines for the Intergovernmental Panel on Climate Change [ 5 ] and the great public interest in global warming and climate change highlight the growing demand for more efficient biological nitrogen removal that needs to include minimizing greenhouse gas emissions. Furthermore, biological ammonium conversion to N 2 requires up to 91% of the energy (MJ/kg N) that is consumed to convert N 2 to ammonium and, consequently, is an intrinsically unsustainable nitrogen management method, which could theoretically be replaced by ammonium recovery technologies [ 3 •• ]. In this review, we focus on the recently gained insights into the physiology of nitrogen-transforming microorganisms and the current and prospective application of these for sustainable nitrogen management."
} | 701 |
37257137 | PMC10278177 | pmc | 2,724 | {
"abstract": "DNA nanotechnology enables straightforward fabrication\nof user-defined\nand nanometer-precise templates for a cornucopia of different uses.\nTo date, most of these DNA assemblies have been static, but dynamic\nstructures are increasingly coming into view. The programmability\nof DNA not only allows for encoding of the DNA object shape but also\nit may be equally used in defining the mechanism of action and the\ntype of stimuli-responsiveness of the dynamic structures. However,\nthese “robotic” features of DNA nanostructures are usually\ndemonstrated for only small, discrete, and device-like objects rather\nthan for collectively behaving higher-order systems. Here, we show\nhow a large-scale, two-dimensional (2D) and pH-responsive DNA origami-based\nlattice can be assembled into two different configurations (“open”\nand “closed” states) on a mica substrate and further\nswitched from one to the other distinct state upon a pH change of\nthe surrounding solution. The control over these two configurations\nis achieved by equipping the arms of the lattice-forming DNA origami\nunits with “pH-latches” that form Hoogsteen-type triplexes\nat low pH. In short, we demonstrate how the electrostatic control\nover the adhesion and mobility of the DNA origami units on the surface\ncan be used both in the large lattice formation (with the help of\ndirected polymerization) and in the conformational switching of the\nwhole lattice. To further emphasize the feasibility of the method,\nwe also demonstrate the formation of pH-responsive 2D gold nanoparticle\nlattices. We believe this work can bridge the nanometer-precise DNA\norigami templates and higher-order large-scale systems with the stimuli-induced\ndynamicity.",
"conclusion": "Conclusions In this work, we have presented a strategy\nfor constructing pH-responsive\nand dynamically reconfigurable lattices using DNA origami as the building\nblock. The pH-responsiveness of the lattice is achieved by equipping\nthe arms of the pliers-like, lattice-forming DNA origami unit with\npH latches that form Hoogsteen-type of triplexes in low pH. Therefore,\nthe unit could rapidly switch between an open “+”-shaped\nand a closed “X”-shaped configuration upon a pH change.\nNevertheless, the high level of programmability of the DNA origami\nwould equally enable other stimuli-responsive elements, such as photoresponsive\nmolecules 49 and thermoresponsive polymers, 50 to be implemented into the basic building block\nof the lattice, thus allowing reconfigurable lattices that undergo\nconformational changes in response to different external stimuli.\nFurthermore, the high addressability of DNA origami allows not only\nAuNPs (as demonstrated here) but also a wide variety of other compounds\nto be precisely positioned onto DNA origami frameworks. Therefore,\nwe believe that our demonstrated system as well as other recently\nreported reconfigurable DNA-based lattices 28 − 31 , 51 , 52 will contribute to the development of more\nsophisticated stimuli-responsive and functional materials in future.",
"discussion": "Results and Discussion Design and Characterization of the Reconfigurable pH-Responsive\nDNA Origami Unit To assemble the pH-responsive and dynamic\nlattice, we first constructed and characterized the pH-sensitive DNA\norigami unit, the basic building block of the lattice. The pliers-like\nDNA origami unit consists of two bar-shaped arms (86 nm × 12\nnm × 6 nm) that are connected to each other through the pivot\nwhich is two single-stranded DNA (ssDNA) scaffold crossovers (analogous\nto the Holliday junction) ( Figure 1 a). The unit is designed with two rationally engineered\npH-sensitive “latches”. Therefore, depending on the\npH of the surrounding solution, the unit may adopt either an open\n(arms rotate freely with respect to each other, and thus the observed\nvertex angle between the arms varies from α ≈ 20°\nto α ≈ 90°) or a closed configuration (vertex angle\nα ≈ 30°). In more detail, the latches are staple-strand\nextensions and consist of two counterparts positioned on different\narms of the unit: a hairpin with a 20-base pair (bp) double-stranded\nDNA (dsDNA) region and a complementary 20-nucleotide (nt) ssDNA sequence.\nAt high pH, the hairpin and the ssDNA do not interact with each other,\nthus allowing for free rotation of the arms. At low pH, for one, these\ntwo counterparts can form a parallel DNA triplex through Hoogsteen\ninteractions, which locks the two arms at a fixed position. Both pH\nlatches have different base sequences but an identical T-A·T\nbase content of 60%, which ensures that both latches have a transition\npH value of p K a ∼ 7.2 and thus\nthey will open/close at the same pH. 38 However,\nthe pH range at which the opening/closing takes place can be rationally\ntuned by adjusting the T-A·T base content of the latch sequences. 35 , 39 In addition to the pH-sensitive unit, we also designed and prepared\ntwo control units: a permanently open unit with no latch sequences\n(Op) and a permanently closed unit (Cl), in which the pH-sensitive\nlatch sequences have been replaced with complementary ssDNA overhangs. To confirm both the correct folding of the units and the functionality\nof the pH-sensitive latches, poly-T passivated DNA origami units (8-nt\npolythymine extensions at each helix to avoid end-to-end stacking)\nwere first analyzed by agarose gel electrophoresis (AGE) ( Figure 1 b and Figure S2 ). The closed unit is more compact than\nthe open unit and therefore the closed unit exhibits a higher electrophoretic\nmobility in the gel. This allows for separation of these two configurations\nby AGE. The first gel was run at pH 8.2 ( Figure 1 b, top panel), which is above the p K a value, and thus, it was also expected that\nsamples prepared at pH 8.2 (initially open) and at pH 6.0 (initially\nclosed) will both adapt the open configuration. This is indeed the\ncase, as the both samples exhibit equal mobility which further matches\nthe mobility of the permanently open (Op) control sample. The second\ngel ( Figure 1 b, bottom\npanel), was run at pH 6.0, which is well below the p K a value. Here, the (initially closed) sample at pH 6.0\nremains predominantly in its closed configuration, while the (initially\nopen) sample prepared at pH 8.2 shows a slightly broader band. This\nindicates that the sample is a blend of both open and closed configurations\ndue to the slow closing kinetics of the initially open unit. 38 , 39 The opening kinetics is faster, and therefore, the initially closed\nunit will swiftly open in the pH 8.2 gel, resulting in a clear and\nnarrow band. In addition to AGE, we also used TEM to characterize\nthe DNA origami\nunits at both pH 8.2 and pH 6.0. In both cases, TEM reveals distinct,\ncorrectly folded units ( Figure 1 c,d and Figures S3 and S4 ). At\npH 8.2, the unit equipped with pH-sensitive latches adapted the open\nconfiguration with a wide (α ≈ 20–90°) and\nflat vertex angle distribution ( Figure 1 c). At pH 6.0, on the other hand, most of the units\nadapted the closed configuration with an vertex angle of α ≈\n30° (∼75% of the units have vertex angles within 20–40°)\n( Figure 1 d). Despite\nthis pronounced and narrow vertex angle distribution, both TEM and\nAGE analysis additionally reveal that a small fraction of the pH-sensitive\nunits still remains at the open configuration at pH 6.0. The same\ntrend was also observed for a pH-sensitive unit variant with different\nlatch configurations, thus allowing for closing of the arms in the\nopposite direction ( Figure S5 ) and the\npermanently closed control unit (at both pH 8.2 and pH 6.0, see Figures S6 and S7 ). The effect was even more\npronounced at low cation concentrations (see Figure S1 ), indicating that the electrostatic repulsion between the\ntwo arms is strong enough to prevent some units from closing. Nevertheless,\nthe observed closing yield is in good agreement with previously reported\nclosing efficiencies for similar pH-responsive DNA origami structures. 40 Selective Assembly of DNA Origami Dimers For the lattice\nformation, it is crucial that the units are connected together in\na programmable fashion without undesired interconnection of the top\narm and the bottom arm that are located in different planes. In order\nto selectively connect only specific ends of the arms, we designed\n“connector oligonucleotides” for seven helices in each\nof the two arms ( Figure 2 a and Figure S45 ). To further minimize\nthe undesired interactions between the two arms, the connector oligonucleotides\nwere arranged in different patterns for the top and the bottom arms\nof the unit, while the rest of the helix-ends remained untouched (blue\nhelices in Figure 2 a cross-section). In total, 14 strands of the connector oligonucleotides\n(7 per each arm, 4 at one end and 3 at the other) contain a 3-nt long\nprotruding 3′-end-overhang. Each overhang is complementary\nto a 3-nt long scaffold sequence, which is located in the same helix\nbut at the opposite site of the arm (3 and 4 recession sites at the\nopposite edges of the arm). Therefore, these interlocking complementary\nsequences can efficiently bridge the side scaffold loops of these\ntwo adjacent DNA origami units. 41 The combination\nof short hybridizing sequences and shape complementarity provides\nthe needed specificity for correctly joining the units together; however,\nthe interactions are still weak enough to allow for rearrangements\nbetween the units and thus to help avoiding misaligned lattice formation. 42 Figure 2 Formation of dynamic DNA origami dimers. (a) Dimers are\nformed\nby mixing equimolar amounts of both units (folded separately). The\nDNA origami units are selectively linked together by bridging the\nside scaffold loops with connector oligonucleotides. To connect the\nscaffold loops, seven of the connector oligonucleotides have a 3-nt\noverhang (in the 3′ end) complementary to the scaffold sequence\non the opposite end of the arm. (b) Characterization of the dimer\nformation by AGE at pH 8.2 (top panel) and pH 6.0 (bottom panel).\nIf not otherwise specified, the pH of the samples are 8.2 in the top\ngel and 6.0 in the bottom gel. TEM images of (c) dimers formed at\npH 8.2 by combining A and A′ units ( c = 5.7\nnM), (d) the same dimer solution as in (c) (A and A′ units, c = 5.4 nM) after the pH has been decreased to 6.0 with\nacetic acid, and (e) a mixture of B and A′ units ( c = 5.4 nM). These units do not have matching connector oligonucleotides,\nand therefore, no dimers are formed. The samples in TEM images are\nnegatively stained with 2% (w/v) uranyl formate. To demonstrate the selectivity of the connector\noligonucleotides,\nwe, as a proof of concept, prepared different versions of pH-responsive\nDNA origami dimers ( Figure 2 a and Figures S9–S17 ). The\ntwo units (marked with A and A′ if the connector oligonucleotides\nare in the bottom arm) were folded and purified from excess staple\nstrands in separate batches, after which the dimers were formed by\nmixing equimolar amounts of both units. In order to prevent multimerization\nof the units, the interfaces of the arm ends not involved in dimer\nformation were poly-T-passivated (8-nt long polythymine overhangs).\nAGE revealed that the band corresponding to the single units almost\ncompletely vanished in the dimer mixture, whereas another band with\nlower electrophoretic mobility appeared in the gel, indicating a successful\ndimerization ( Figure 2 b and Figures S9 and S13 ). Importantly,\na control sample with mismatching units (unit B with connectors in\ntop arm combined with unit A′) did not form any dimers, demonstrating\nthat our strategy to connect the units is indeed highly selective.\nTo further confirm that the two units interact with each other correctly,\nwe used TEM to visualize the formed dimers. The TEM images of the\ndimers that were assembled at pH 8.2 show, as expected, perfectly\naligned and well-defined DNA origami dimers with the arms open ( Figure 2 c and Figures S10 and S11 ). Furthermore, by adding\nacetic acid to this dimer solution, the arms of the dimer could be\nlocked into the closed configuration ( Figure 2 d and Figures S10 and S11 ). Equally, the dimers could be formed from the units initially\nat the closed state at pH 6.0, after which the arms could be released\nagain by increasing the pH with sodium hydroxide ( Figures S13–S15 ). As indicated above, the B and A′\nunits neither have the required shape complementarity nor the matching\nsequences, and therefore, only discrete, unconnected DNA origami units\nwere observed in TEM ( Figure 2 e and Figures S12 and S16 ). Formation of 1D Arrays Using the DNA Origami Unit To\nfurther explore the possibility of using the DNA origami unit for\nthe construction of large-scale lattices, we formed 1D arrays using\nthe DNA origami unit. To this end, we prepared a unit with the polymerizing\nconnector oligonucleotides on the bottom arm (A and A′ interactions)\nand fully poly-T-passivated interfaces on the top arm. To avoid undesired\nmultimerization and formation of kinetically trapped configurations\nduring the folding, the unit was prepared without the connector oligonucleotides.\nThe polymerization of the units into linear arrays was initiated in\na subsequent step by adding connector oligonucleotides in 10-fold\nexcess to units that were earlier purified from the excess staple\nstrands used in folding ( Figure 3 a, step 1). Initially, the assembly was carried out\nin solution by incubating the sample mixture at room temperature for\nat least 24 h. Although we recognized correctly formed linear chains\nwhen imaging the sample by TEM ( Figure 3 b and Figures S18 and S19 ), the tendency of the sample to form highly entangled structures\nset limitations to the analysis of the chain formation. Figure 3 Formation of\none-dimensional (1D) arrays using the DNA origami\nunit. (a) Polymerization of the units into chains is initiated by\nthe addition of connector oligonucleotides. For the surface-mediated\nassembly, the mixture is immediately deposited onto a mica substrate.\n(b) TEM image of a negatively stained linear 1D array formed in solution\nat pH 8.2 (25 h incubation at room temperature, c unit = 10.0 nM, but sample diluted 1:2 in 1× FOB\n(1× TAE, 20 mM MgCl 2 , 5 mM NaCl) before deposited\nonto the TEM grid). (c) Atomic force microscopy (AFM) image of DNA\norigami chains formed on a mica substrate at pH 6.0 (3 h incubation).\n(d) Observed chain length distribution for the 1D arrays assembled\non a mica substrate at pH 6.0 (determined from AFM images). As an alternative to the solution-phase formation,\nwe also assembled\nthe DNA origami chains on a mica substrate at the solid–liquid\ninterface. The interface restricts the movement of the units to the\n2D plane and may thus provide additional control of the lattice formation\nand growth. 34 , 43 For the surface-assisted assembly,\nthe units and the connector oligonucleotides were mixed together in\na buffer supplemented with MgCl 2 and NaCl and immediately\nafter that deposited onto a mica substrate ( Figure 3 a). Linear arrays were grown at both pH 6.0\n( Figure 3 c and Figure S20 ) and pH 8.2 ( Figures S21 and S22 ), and in both cases, discrete chains of various\nlengths were formed. Nineteen percent of the units assembled into\n>1 μm long chains (>11 units), while the majority of them\nformed\nchains of 3–10 units (pH 6.0, n = 275) ( Figure 3 d). This is also\nin line with the previously reported chain lengths for similar linear\nDNA origami arrays. 44 Assembly of pH-Responsive and Reconfigurable 2D DNA Origami\nLattices By introducing connector oligonucleotides on both\nthe bottom and the top arms of the unit (A and A′ interactions\nas well as B and B′ interactions), we constructed dynamic 2D\nlattices ( Figure 4 a,b).\nThe two pH-sensitive conformations of the unit allow the lattice to\nadopt either an open or a closed configuration depending on the pH\nof the assembly solution. The 2D lattices were assembled directly\nonto the mica substrate by employing a previously established protocol 24 that we further developed and optimized for\nour system. For successful formation of large hierarchical DNA origami\nassemblies on mica, the electrostatic interactions between the DNA\norigami and the surface have to be carefully controlled, which is\nusually accomplished by tuning the relative amounts of Na + and Mg 2+ in the assembly buffer. 43 , 45 The divalent Mg 2+ ions mediate the DNA origami adsorption\nonto mica by forming salt bridges, whereas the competitive Na + ions weaken these interactions and enhance the DNA origami\nmobility on the surface. Depending on the assembly pH, we observed\na clear difference in the DNA origami adsorption, which also affected\nthe lattice growth. Therefore, we investigated the influence of the\nMg 2+ concentration on lattice formation on the mica substrate\nduring 3 h by keeping the Na + concentration constant at\n75 mM ( Figure 4 c and Figures S23 and S24 ). At pH 8.2, the optimum\nMg 2+ concentration was found to be 10 mM, which is well\nin agreement with previously optimized conditions. 46 At pH 6.0, for one, the electrostatic interactions were\nnoticeably weaker and a Mg 2+ concentration of 12.5 mM was\nneeded to obtain sufficient DNA origami adsorption for the subsequent\nlattice growth. The observed pH-dependent difference in the required\nMg 2+ concentration may be explained by silicate protonation\nand thus a reduced surface charge of mica at low pH. In addition,\nincreasing the Mg 2+ concentrations of the assembly solution\nbeyond these optimized values results in high DNA origami adsorption\nand low DNA origami mobility on the surface, which considerably decrease\nthe lattice order. Figure 4 Assembly of pH-responsive and reconfigurable two-dimensional\n(2D)\nDNA origami lattices. (a and b) Connector oligonucleotides for both\narms of the unit initiate the assembly of a 2D lattice on a mica substrate.\nThe formed lattice could adapt either an open or a closed configuration\ndepending on the pH of the surrounding solution. The reconfigurable\nlattice could be expanded or squeezed also after the initial assembly\nby increasing or decreasing the pH. AFM images (1 μm ×\n1 μm) of the different lattice configurations are shown below\nthe schematics. (c) AFM images (1 μm × 1 μm) of the\n2D lattice formation at different Mg 2+ concentrations.\nThe Na + concentration is kept constant at 75 mM, and the\nassembly time is 3 h. The DNA origami lattice could guide gold nanoparticles\n(AuNPs) into either (d) a square lattice at pH 8.2 or (e) a oblique\nlattice at pH 6.0. The top panel shows an AFM image (500 nm ×\n500 nm) of the AuNP lattice, and the area marked with dotted lines\nis schematically presented next to the image. The bottom panel show\nthe observed lattice constant distributions for the formed AuNP lattice\n(determined from the AFM images). The AuNP lattices are assembled\nduring 3 h. In (a), (b), (d), and (e), the Mg 2+ concentration\nis 10 mM for lattices at pH 8.2 and 12.5 mM for lattices at pH 6.0. Depending on the assembly pH, the DNA origami lattice\nhas two clearly\ndistinguishable configurations ( Figures 4 a,b, bottom left). At pH 8.2, the unit will\nadapt the open configuration and the formed lattice will be in an\nexpanded state. At pH 6.0, on the other hand, the units are predominantly\nin the closed configuration, and therefore, a more compact lattice\nis formed. Nonetheless, in both cases, the obtained lattice is polycrystalline\nand composed of smaller crystalline domains of various sizes in close\nproximity to each other. The order and the size of the crystal domains\ncorrelate with the assembly duration, and therefore, the crystal growth\ncould be considerably improved by increasing the assembly time ( Figures S25–S28 ). The crystal domains\nare generally also larger at pH 6.0, which could be explained by the\nenhanced rigidity of the unit when the arms are tied together and\nthus not able to rotate freely. Moreover, replacing Mg 2+ with Ca 2+ has been shown to enhance the lattice order\nfor closed-packed lattices of symmetric DNA origami units that do\nnot bind to each other via basepairing, 47 but for our system, this replacement had no significant effect ( Figure S29 ). Thus far, most of the reported\nDNA origami-based frameworks have\nbeen static, meaning that their lattice parameters have been fixed\nonce they have been assembled. However, approaches allowing a stimuli-induced\ndynamic symmetry conversion after the assembly would be highly desirable.\nTherefore, we next studied whether our assembled pH-responsive and\nreconfigurable lattices could be readily expanded and squeezed by\nsimply increasing or decreasing the pH. For these experiments, we\nfirst assembled the lattices on the mica surface for 5 h at pH 8.2\nor 6.0, washed away weakly interacting and unbound assemblies, deposited\na different buffer solution with lower/higher pH, and incubated for\nadditional 2 h (pH increase from 6.0 to 8.2) or 20 h (pH decrease\nfrom 8.2 to 6.0). When the pH was increased from 6.0 to 8.2, a clear\nchange from the closed state toward the open lattice configuration\nwas observed ( Figure 4 a, bottom right and Figures S30–S32 ), indicating that the formed lattices are rather mobile on the surface.\nClosing of the lattice after assembly, (pH decrease from 8.2 to 6.0),\nfor one, required much longer time, and the overall change in the\nlattice configuration was not as pronounced as in the case of opening\nthe lattice ( Figure 4 b, bottom right and Figures S33–S35 ). Interestingly, we also observed that, as long as the lattices\nwere not attached to the mica substrate with NiCl 2 , the\nonce dried lattices (for AFM imaging) could be rehydrated and their\nconfiguration altered by increasing or decreasing the pH ( Figures S36–S40 ). This further demonstrates\nthat the lattices are mobile enough on the surface to rearrange also\nafter the initial assembly. Assembly of DNA-Templated, pH-Responsive, and Reconfigurable\n2D AuNP Lattices It is known that spatially well-defined\narrangements of metal nanoparticles possess intriguing optical, plasmonic,\nelectronic, and magnetic properties, 2 but\nfabrication of highly ordered dynamic nanoparticle lattices is rather\nchallenging. As already mentioned, programmable and modular DNA-based\nstructures are suitable templates for guiding nanoparticles into complex,\nmostly static lattices using either DNA hybridization 21 , 25 , 26 or electrostatic interactions. 48 In order to demonstrate that our pH-sensitive\nlattice could be used as a template to create reconfigurable nanoparticle\nlattices, we modified the DNA origami unit by adding an anchoring\nsite for an oligonucleotide-coated gold nanoparticle (AuNP, 10 nm\nin diameter) in the middle of the unit ( Figure S8 ). AFM images of the prepared lattices show, as expected,\ntwo distinct lattice configurations depending on the assembly pH or\nunit used; a 2D square lattice at pH 8.2 ( Figure 4 d and Figure S41 ) and a 2D oblique lattice at pH 6.0 ( Figure 4 e and Figure S42 ) or if a permanently closed unit is used ( Figure S43 ). Furthermore, the average lattice constants determined\nby AFM are a = 85 ± 13 nm for the square lattice\nand a = 87 ± 10 nm, b = 55\n± 14 nm for the oblique lattice. The highest frequency was observed\nfor a = 90–92 nm for the square lattice and a = 86–88 nm and b = 44–46\nnm for the oblique lattice. The DNA origami unit is rather flexible\nat pH 8.2, and taking that into account, the observed lattice constants\nare well in agreement with the theoretical ones ( a = 86 nm (both for square and oblique lattices) and b = 45 nm, assuming a vertex angle of 30°)."
} | 5,891 |
21325039 | PMC3039438 | pmc | 2,725 | {
"abstract": "Many members of the LuxR family of acyl-homoserine lactone (acyl-HSL)-dependent quorum-sensing transcriptional activators are thought to have the unusual characteristics of requiring the signal ligand during polypeptide synthesis to fold into an active conformation and of binding signal extraordinarily tightly. This is the case for the N -3-oxo-dodecanoyl-HSL-dependent Pseudomonas aeruginosa virulence regulator LasR. We present evidence that LasR can fold into an active conformation in vivo in the absence of the acyl-HSL ligand. We also present evidence indicating that in the cellular environment, LasR and N -3-oxo-dodecanoyl-HSL readily dissociate. After dissociation, LasR can remain in a properly folded conformation capable of reassociating with signal. We present a new model for the folding and signal binding of LasR and other members of the family of transcription factors to which LasR belongs. Our findings have important implications concerning the cellular responses to decreased environmental concentrations of signals and have implications about potential quorum-sensing inhibition strategies.",
"introduction": "INTRODUCTION Acyl-homoserine lactone (acyl-HSL) quorum sensing controls a battery of genes involved in virulence of Pseudomonas aeruginosa ( 1 , 2 ). This opportunistic pathogen has two acyl-HSL synthases, LasI and RhlI, which produce N -3-oxo-dodecanoyl homoserine lactone (3OC12-HSL) and N -butanoyl homoserine lactone (C4-HSL), respectively ( 3 – 8 ). The lasI gene is linked to lasR , which codes for a 3OC12-HSL-dependent transcription factor. The rhlI gene is linked to rhlR , which codes for a C4-HSL-dependent transcription factor. There is an additional 3OC12-HSL-dependent transcription factor called QscR ( 9 , 10 ). LasR, RhlR, and QscR control overlapping sets of P. aeruginosa genes. LasI and RhlI are members of a family of acyl-HSL synthases distributed widely amongst Proteobacteria , the LuxI protein family. Likewise, LasR, RhlR, and QscR are members of the LuxR family of transcription factors. Acyl-HSLs can move out of and into cells by diffusion, although 3OC12-HSL likely partitions somewhat to membranes where efflux pumps aid in moving it out of cells ( 11 ). It is generally believed that LasR and many other LuxR family members will not fold into an active polypeptide in the absence of an appropriate acyl-HSL ( 10 , 12 – 16 ). The idea, first developed by Zhu and Winans who showed that the Agrobacterium tumefaciens TraR is rapidly degraded in the absence of the cognate acyl-HSL signal, is that during elongation the nascent polypeptide must fold around the acyl-HSL to achieve an active conformation ( 16 ). Furthermore, LasR and TraR bind their signal very tightly, virtually irreversibly ( 12 , 15 ). This may be true of other but not all LuxR homologs ( 17 ). On the basis of available evidence, it is reasonable to conclude that these unusual characteristics of a ligand requirement for proper protein folding and virtually irreversible ligand binding are features of at least certain LuxR homologs. If true, there are important implications for quorum-sensing control of gene expression in P. aeruginosa . The evidence that has led to these conclusions includes the following: LasR, TraR, and other LuxR homologs are almost exclusively in an insoluble aggregated form when expressed in the absence of signal, and active protein has not been obtained from cells grown without signal ( 10 , 12 – 15 ). Considerable amounts of soluble active protein can be obtained from cells grown in the presence of an appropriate acyl-HSL. When taken through a purification procedure, LasR (and TraR) retains stoichiometric amounts of signal, and additional signal is not required for binding of these transcriptional activators to target DNA ( 12 , 15 ). We point out that although many LuxR homologs are thought to require signal as a scaffold for folding during polypeptide synthesis, it is clearly not true of all LuxR homologs. The Mesorhizobium tianshanense transcriptional activator MrtR requires its cognate signal for dimerization but not for folding ( 18 ). Likewise, the Pantoea stewartii EsaR protein does not require signal for folding. EsaR is perhaps an unusual example, as it binds target DNA in the absence of signal and serves as a repressor. Signal binding results in derepression ( 19 ). If this belief that LasR and certain other LasR homologs require their signal as a folding scaffold during their synthesis is correct, it has important implications about cellular responses to decreasing environmental acyl-HSL levels and about designing inhibitors of quorum sensing that might have utility as antivirulence therapeutics. For example, if signal binding is essentially irreversible, then one would predict that gene activation by 3OC12-HSL–LasR would persist when a cell moves from a high-population-density (high-3OC12-HSL concentration) environment to a low-population-density (low-3OC12-HSL) environment, at least until active LasR is decreased by proteolysis or cell division. One might also predict that competitive inhibitors of 3OC12-HSL might work via binding to the nascent LasR polypeptide chain prior to 3OC12-HSL binding but would not affect preexisting LasR–3OC12-HSL complexes. There has been an enormous amount of effort aimed at understanding the influence of increased acyl-HSL levels on the activity of LuxR homologs, but almost no attention has been paid to the immediate cellular response to decreases in acyl-HSL concentrations. We are aware of only three reports on the influence of a drastic decrease in signal levels on LuxR homolog-dependent gene transcription ( 20 – 22 ). When Vibrio fischeri , the bacterium in which acyl-HSL quorum sensing was first discovered, is rapidly transferred from a signal-replete medium to a signal-depleted medium but kept at the same population density, transcription of quorum-sensing-controlled genes returns to the basal level within minutes. This suggests that LuxR binds to its cognate signal, 3OC6-HSL, reversibly. This is consistent with the fact that, unlike LasR and TraR, purified LuxR does not appear to bind signal irreversibly ( 13 ). LuxR is considered an exception rather than the rule. Somewhat similar experiments on the A. tumefaciens LuxR homolog TraR showed a precipitous drop in the environmental signal concentration that resulted in a rather rapid decrease in TraR-dependent reporter transcription, but a low level of transcription appeared to persist for several cell divisions. The investigators came to the conclusion that the low level of transcription indicated that some TraR retained signal and continued to function ( 22 ), but they did not discriminate this possibility from the possibility that the low level of transcription represented basal-level TraR-independent transcription of the reporter. None of these studies ( 20 – 22 ) addressed the issue of whether functional acyl-HSL receptors capable of signal binding and DNA binding remained after signal removal from the medium. In light of the above-described considerations, we have examined the activity of LasR in recombinant Escherichia coli and in P. aeruginosa after a rapid decrease in 3OC12-HSL from above a threshold level for LasR activity to below that level. The experiments we present are consistent with the view that LasR folds into an active conformation in the absence of any acyl-HSL signal and can remain in a properly folded signal-free state. Our experiments are also consistent with the view that, in the context of the cellular environment, signal binding is reversible.",
"discussion": "DISCUSSION Acyl-HSL-dependent transcriptional activators such as TraR and LasR form insoluble aggregates when expressed in recombinant bacteria grown without the cognate signal but show increased solubility when cells are grown in the presence of signal ( 10 , 12 , 13 , 15 ). What little soluble protein there might be in extracts of cells grown without signal does not appear to be active as assessed by EMSA in the absence or presence of the cognate signal ( 12 ). Active TraR or LasR can be obtained from cells grown with the cognate signal. When these proteins are purified, they retain signal even after purification steps involving dialysis and column chromatography ( 12 , 15 ). Moreover, LasR retained signal even after extensive dialysis against signal-free buffer. Furthermore, the crystal structure of TraR in its DNA-bound state shows that the acyl-HSL signal is completely buried in the protein ( 14 ). These sorts of data have led to the view that during polypeptide synthesis many LuxR homologs require their cognate signal to fold properly and that once the active protein has formed, it retains signal ( 17 ). If this were the case, these would represent very unusual properties. This sort of behavior would have important implications related to the ability of cells to respond rapidly to decreases in environmental signal levels and the ability of small molecules to competitively inhibit activity of functional LasR or TraR proteins. Because there is little known about cellular responses to precipitous decreases in external acyl-HSL signals, we sought to investigate such a response. We chose to study LasR because production of significant levels of soluble LasR in recombinant bacteria requires its cognate acyl-HSL, it retains the acyl-HSL even after extensive dialysis against signal-free buffer, and it has been a target for development of quorum-sensing inhibitors that might have value as anti- Pseudomonas virulence therapeutics ( 12 , 17 ). Our in vivo experiments with either recombinant E. coli or P. aeruginosa show that there is a rapid cessation in transcription of a LasR-dependent gene after 3OC12-HSL is decreased from a saturating level to a level below the threshold for gene activation ( Fig. 1 ). These experiments indicate that in contrast to in vitro results, 3OC12-HSL can rapidly dissociate with LasR in vivo . These experiments also show that LasR-dependent quorum control of gene expression can respond rapidly to decreased signal concentration just as it can respond rapidly to increased signal concentration. If it is true that LasR and 3OC12-HSL can dissociate rapidly in vivo , it becomes an open question as to whether signal-free LasR can bind added 3OC12-HSL and regain its ability to activate quorum-controlled genes. We addressed this question in two ways. First, we used recombinant E. coli containing an arabinose-inducible lasR and a LasR-responsive reporter in the presence of arabinose and 3OC12-HSL. During the induction phase of the reporter, we removed arabinose and 3OC12-HSL, and we asked whether the existing pool of LasR maintained an ability to activate the reporter in response to 3OC12-HSL added over increasing periods of time ( Fig. 2 ). Our evidence indicates that the bacteria retained a pool of functional, 3OC12-HSL-responsive LasR for at least 20 min ( Fig. 2 and 3 ). Second, we were able to demonstrate the existence of active LasR in extracts of bacteria that were grown without any added acyl-HSL ( Fig. 4 ). We believe that this has been problematic for LasR and other LasR homologs in the past because unless they are bound to their signal, these proteins are very unstable in cell-free extracts. We overcame this obstacle by breaking the cells in ice-cold buffer with 3OC12-HSL. On the basis of all of the available results on LasR and related transcription factors, we propose a new model for the interactions of LasR with its cognate signal ligand 3OC12-HSL. We believe this model might hold for TraR, but we have no experimental data to bring to bear on this possibility. Our model is perhaps more congruent with our general understanding of protein-ligand interactions than the previous model. We believe that LasR can fold into a functional conformation in the absence of an acyl-HSL ligand. Consistent with previous reports on TraR and QscR ( 15 , 24 ), we believe that signal-free LasR is relatively unstable in comparison to signal-bound LasR. However, it is sufficiently stable for a pool of functional cellular signal-free LasR to exist. We believe that like other transcription factors, LasR binds to its coinducing ligand (3OC12-HSL) reversibly. It is generally assumed that members of the LuxR family fall into one of three general categories based on interactions with their signals ( 17 ). One category is represented by LasR and TraR, both of which were thought to require signal as a folding scaffold during synthesis and to bind signal in a virtually irreversible manner. Those like LuxR from V. fischeri were thought to require signal as a scaffold but once folded exhibit reversible signal binding. Finally, members of a class represented by M. tianshanense MrtR and P. stewartii EsaR do not require an acyl-HSL as a folding scaffold. Our new evidence leads us to speculate that all LuxR homologs might interact with their signals in a fashion analogous to MrtR and EsaR in that they can fold into active forms in the absence of an acyl-HSL. Our recent biochemical analysis of QscR ( 24 ) is consistent with this new view, and it suggests that the differences in behavior of different LuxR homologs can be accounted for by signal and DNA binding affinities as well as stability of signal-free forms in vivo . Considerable attention has been given to cellular responses to increasing acyl-HSL levels, either natural accumulations or artificial additions. Unfortunately, little attention has been given to the responses resulting from rapid decreases in signal levels. This report shows there is information to be gained by studying the effects of decreasing environmental signal concentrations on quorum-sensing regulation of gene expression."
} | 3,453 |
38459235 | PMC11164683 | pmc | 2,726 | {
"abstract": "Stimuli-responsive hydrogels with programmable shape changes are promising materials for soft robots, four-dimensional printing, biomedical devices and artificial intelligence systems. However, these applications require the fabrication of hydrogels with complex, heterogeneous and reconfigurable structures and customizable functions. Here we report the fabrication of hydrogel assemblies with these features by reversibly gluing hydrogel units using a photocontrolled metallopolymer adhesive. The metallopolymer adhesive firmly attached individual hydrogel units via metal–ligand coordination and polymer chain entanglement. Hydrogel assemblies containing temperature- and pH-responsive hydrogel units showed controllable shape changes and motions in response to these external stimuli. To reconfigure their structures, the hydrogel assemblies were disassembled by irradiating the metallopolymer adhesive with light; the disassembled hydrogel units were then reassembled using the metallopolymer adhesive with heating. The shape change and structure reconfiguration abilities allow us to reprogramme the functions of hydrogel assemblies. The development of reconfigurable hydrogel assemblies using reversible adhesives provides a strategy for designing intelligent materials and soft robots with user-defined functions.",
"conclusion": "Conclusion In conclusion, we have demonstrated the fabrication of complex, heterogeneous, multiresponsive hydrogel assemblies with reconfigurable structures and reprogrammable functions by reversibly gluing hydrogel units using a metallopolymer adhesive. The newly designed metallopolymer adhesive adheres firmly and reversibly to the wet surfaces of hydrogels. It adapts to the actuation and shape changes of hydrogels. It is also tolerant to external stimuli (for example, pH or temperature) for hydrogel actuation. These features are unique. The combination of photocontrolled Ru–S coordination and polymer chain entanglement is distinct from the mussel-inspired chemistry and supramolecular chemistry for adhesives reported in the literature and represents a strategy for designing strong yet reversible adhesives. We anticipate that, similar to the actuators and soft robots reported here, other intelligent materials with multiple components, reconfigurability, programmability and customizable functions can be fabricated by reversibly gluing intelligent building blocks. Our study opens up an avenue for the design of responsive materials, four-dimensional printing materials, biomaterials and soft robots.",
"discussion": "Results and discussion Reversible Ru–S coordination To demonstrate reversible Ru–S coordination, we synthesized Ru complexes (Ru–H 2 O and Ru–SL) and a thioether ligand (SL) as model compounds (Fig. 1b and Supplementary Figs. 1 – 19 ). We studied reversible Ru–S coordination using 1 H nuclear magnetic resonance ( 1 H NMR) spectroscopy (Fig. 1c ). When the mixture of Ru–H 2 O (4.36 mM) and SL (43.6 mM) in D 2 O was heated to 70 °C for 30 min, the signal from Ru–H 2 O at 9.50 ppm disappeared, and a new signal from Ru–SL at 9.76 ppm appeared. This result suggests that SL coordinated with the Ru centre. Then, the Ru–SL solution was irradiated with blue light (470 nm, 60 mW cm − 2 ) for 15 min. The signal at 9.76 ppm notably increased, and the signal at 9.50 ppm greatly decreased, which indicated that most Ru–SL was hydrolysed to form Ru–H 2 O upon light irradiation. The 1 H NMR data showed that Ru–SL reformed when the sample was heated to 70 °C for 30 min. Thus, Ru–S coordination was reversible upon heating and light irradiation. We also studied Ru–S coordination using ultraviolet–visible (UV–vis) absorption spectroscopy (Fig. 1d ). Initially, the absorption maximum for the mixture of Ru–H 2 O and SL was at 476 nm. This was attributed to the metal-to-ligand charge transfer band of Ru–H 2 O (Supplementary Fig. 10 ). When the mixture was heated to 70 °C for 2 h, the absorption maximum shifted to 452 nm, which was the same as that of pure Ru–SL (Supplementary Fig. 19 ). These results showed that SL coordinated with the Ru centre upon heating. Then, the sample was irradiated with blue light (470 nm, 60 mW/cm 2 ) for 6 min. The absorption band reverted to the initial state, which suggested that Ru–SL was hydrolysed to form Ru–H 2 O upon light irradiation. Subsequently, the sample was heated again at 70 °C for 2 h, and its absorption maximum reverted to 452 nm. The formation and dissociation of the Ru–S coordination bond was cycled four times (Fig. 1e ), which revealed that the Ru–S coordination was reversible. Different from conventional supramolecular interactions or host–guest interactions, the Ru–S coordination did not dissociate upon dilution (Supplementary Fig. 20 ), which revealed that the Ru–S coordination is a stable yet reversible bond. Reversible adhesives for hydrogels To prepare reversible adhesives based on Ru–S coordination, we synthesized a Ru-containing polymer (P-Ru) and a thioether-containing polymer (P-S) (Fig. 2a and Supplementary Figs. 21 – 41 ). P-Ru and P-S were water soluble, and gelation occurred when the mixture of P-Ru (1 wt%) and P-S (1 wt%) in water was heated to 70 °C (Fig. 2b and Supplementary Fig. 45 ). 1 H NMR spectroscopy showed that the gelation was attributable to crosslinking via Ru–S coordination (Supplementary Fig. 46 ). Irradiating the P-Ru/P-S gel with blue light induced a gel-to-sol transition due to light-induced dissociation of Ru–S bonds (Fig. 2b and Supplementary Fig. 46 ). The sol–gel transitions upon heating and light irradiation were reversible. Fig. 2 Reversible adhesion of P1 gels based on reversible sol–gel transitions of the P-Ru/P-S adhesive. a , Chemical structures of P-Ru ( x / y = 95.3/4.7, number average molecular weight M n = 16.4 kg mol −1 and polydispersity index PDI of 2.91) and P-S ( m / n = 79/21, M n = 14.9 kg mol −1 and PDI of 1.35). The counterion of P-Ru is Cl − . b , Schematic and photos of reversible sol–gel transitions of P-Ru/P-S on heating and light irradiation. c , Schematic illustration of reversibly adhering two P1 gels using the P-Ru/P-S adhesive. d , Photos showing reversible adhesion of two P1 gels using the P-Ru/P-S adhesive. Scale bars, 10 mm. e , Adhesion strength of adhered P1 gels before irradiation, after light irradiation and after subsequent readhesion. The adhesion measurements were conducted after the samples were immersed in water for 24 h. f – h , SEM images at the junctions of two adhered P1 gels before irradiation ( f ), after light irradiation ( g ) and after readhesion ( h ). The P1-rich and P-Ru/P-S-rich networks are indicated by the white and yellow frames in f , respectively. The interpenetrating networks, which show mixed morphologies of P1 and P-Ru/P-S, are indicated using the red arrows in f . Source data To demonstrate that the P-Ru/P-S mixture can be used as a reversible adhesive, we glued two P1 gels (cross-linked N -hydroxyethyl acrylamide) together (Fig. 2c ). To glue the P1 gels, P-Ru/P-S sol (6 wt%) was placed between the gels and heated to 70 °C. Heating induced the sol-to-gel transition (Fig. 2d ). To separate the glued P1 gels, light irradiation was applied to induce a gel-to-sol transition. The separated P1 gels were readhered by adding P-Ru/P-S sol and heating. Control experiments showed that P1 gels could not be glued using P-Ru or P-S alone (Supplementary Fig. 47 ), which demonstrated that the sol–gel transition of P-Ru/P-S is essential for adhesion. We quantified the adhesion of P-Ru/P-S adhesives to P1 gels via lap shear tests (Fig. 2e and Supplementary Figs. 48 – 51 ). The adhesion strength of P1 gels glued by P-Ru/P-S (6 wt%) was 1.18 kPa. Light irradiation induced a gel-to-sol transition, and the adhesion strength decreased to almost zero. Thus, P1 gels could be separated after light irradiation. Subsequently, the separated P1 gels were readhered using P-Ru/P-S upon heating. The readhered sample had almost the same adhesion strength as the initial sample. To understand why the P-Ru/P-S adhesive exhibited strong yet reversible adhesion to P1 gels, the morphologies at the junctions of two P1 gels glued by a P-Ru/P-S adhesive were studied using scanning electron microscopy (SEM) (Fig. 2f–h ). Both P1 and P-Ru/P-S formed porous gel networks (Supplementary Fig. 52 ). The average pore diameter of the P-Ru/P-S gel was approximately three times larger than that of the P1 gel. For the adhered and readhered samples, some network structures with both small and large pores overlapped at the junctions of the P1 and P-Ru/P-S gels (Fig. 2f,h ), which indicated that the networks of P1 and P-Ru/P-S interpenetrated with each other. For the sample after light irradiation, the P-Ru/P-S network disappeared (Fig. 2g ), which resulted in de-adhesion. To characterize the structures at the junctions of P1 and P-Ru/P-S, energy-dispersive spectroscopy (EDS) of SEM was used. The signals of S in P-S and Ru in P-Ru were measured to show the distributions of P-S and P-Ru. The EDS data showed that the contents of S in a P-Ru/P-S-rich region (region 1), junctional region (region 2) and P1-rich region (region 3) were 0.40%, 0.19% and 0.13%, respectively (Fig. 3a,b ). These data showed that P-S penetrated into P1. EDS also detected Ru in these three regions (Supplementary Fig. 53 ). However, the contents of Ru measured by EDS were lower than 0.1%, which is the lower limit for quantitative analysis. The contents of Ru were lower than those of S because the ratio of Ru/S in P-Ru/P-S was 4.7/21. To further investigate the distribution of P-Ru, we used the Raman mapping technique, which is described below. Fig. 3 Characterization of the interpenetrating network and the mechanism for its formation. a , SEM image at the junctions of two P1 gels glued by a P-Ru/P-S adhesive. The P-Ru/P-S-rich region, junctional region and P1-rich region are labelled with 1, 2 and 3, respectively. b , EDS data of regions 1, 2 and 3. The contents of S atoms in these regions are 0.40%, 0.19% and 0.13%, respectively. c – e , Raman mapping at the junctions of two P1 gels glued by a P-Ru/P-S adhesive based on the Raman signals at 1,480 cm − 1 ( c ) and 661 cm − 1 ( d ) and Raman spectra at different areas in the Raman maps ( e ). The areas highlighted by coloured boxes in c and d correspond to the signals in e . f , Scheme for the FCS experiment. The distribution of the fluorescently labelled P-S along the Z direction was scanned by adjusting the distance between the sample and the objective. g , Fluorescence intensity along the Z direction. h , Normalized fluorescence intensity autocorrelation functions G ( τ ) recorded for the fluorescently labelled P-S in a P1 gel (red circles) and in the water phase (black squares). The solid lines represent the fittings with Supplementary Equation 1 . Source data P-Ru/P-S showed strong Raman signals. The C–C and C=N stretching vibrations of the polypyridine skeleton of P-Ru were at 1,480, 1,544 and 1,600 cm − 1 , and the C–S–C stretching vibration of P-S was at 661 cm − 1 (Supplementary Fig. 54 ). The P1 gel did not have any characteristic peak in the Raman spectrum (Supplementary Fig. 55 ). Raman maps at the junctions of two P1 gels glued by a P-Ru/P-S adhesive were obtained by scanning a 100 µm 2 × 100 µm 2 region using a focused laser beam with a diameter of ~2 µm (Fig. 3c–e and Supplementary Fig. 56 ). Although the Raman signals of P-Ru/P-S decreased as the laser focus moved from the P-Ru/P-S-rich region to the P1-rich region, P-Ru/P-S was detected in the P1-rich region. These results showed that P-Ru/P-S interpenetrated with P1. To understand the mechanism for the formation of the interpenetration network, we synthesized fluorescently labelled P-S (Supplementary Figs. 42 – 44 ) and studied its diffusion using fluorescence correlation spectroscopy (FCS). A P1 gel was placed in a sample cell that contained an aqueous solution of fluorescently labelled P-S (10 nM). A layer of aqueous solution remained between the P1 gel and the coverslip at the bottom of the sample cell. The sample cell was placed on top of the objective of the inverted confocal microscope (Fig. 3f ). The fluorescence intensity along the perpendicular direction was scanned (Fig. 3g ). The appearance of fluorescence in the P1 gel suggested that some fluorescently labelled P-S diffused from the water phase into the P1 gel. Next, FCS autocorrelation curves were recorded in both the water phase and the P1 gel (Fig. 3h ) and fitted using Supplementary Equation 1 . The diffusion coefficients of the fluorescently labelled P-S in water and the P1 gel were 1.3 × 10 − 10 and 0.9 × 10 − 10 m 2 s −1 , respectively. The P1 gel caused a 31% slowdown of the diffusion, which indicated that the network of the P1 gel hindered the diffusion of the fluorescently labelled P-S. Because the decrease of the diffusion coefficient was not large, the fluorescently labelled P-S can still penetrate into the P1 gel. To provide molecular-level insight into the interpenetration, we performed computer simulations (Supplementary Figs. 57 – 61 ). Initially, P-Ru and P-S out of the P1 gel diffused freely. After diffusion for a few nanoseconds, P-S and P-Ru came into contact with P1, and hydrogen bonds between P1 and P-Ru/P-S were formed. Although steric hindrance and hydrogen bonding slowed their diffusion, P-S and P-Ru still penetrated into the P1 gel because P1 is porous and hydrogen bonds are dynamic, which could break and reconfigure. The computer simulation revealed that P-Ru and P-S penetrated into the P1 gel via free diffusion and diffusion under steric hindrance and hydrogen bonding reconfiguration. To study the effects of interpenetration on adhesion, we performed control experiments by replacing P1 gels with polyethylene and Teflon substrates (Supplementary Fig. 62 ). Because polyethylene and Teflon substrates are hydrophobic and water insoluble, the aqueous solutions of the adhesive cannot penetrate into them. The adhesion energy of P-Ru/P-S-glued P1 gels is more than 375% of that of P-Ru/P-S-glued polyethylene or Teflon substrates, which suggested that the interpenetration of P-Ru/P-S with P1 enhanced the adhesion. Hydrogel actuators assembled by the reversible adhesive Importantly, the P-Ru/P-S adhesive is tolerant to temperature and pH, both of which are frequently used stimuli for hydrogel actuation. After the P1 gels, glued with a P-Ru/P-S adhesive, were treated at different temperatures (25 °C and 70 °C) and pH values (4, 7 and 10) for 24 h, the adhesion strength did not change (Supplementary Fig. 63 ). This environmental tolerance enabled the fabrication of responsive hydrogel actuators using the P-Ru/P-S adhesive, as described below. To prepare responsive hydrogel actuators, we used the P1 gel, thermoresponsive P2 gel, and pH-responsive P3 gel as units (Fig. 4a ). P1/P2 gel assemblies and P1/P3 gel assemblies were fabricated by gluing the corresponding gels using the P-Ru/P-S adhesive. The P1 gel was inert to temperature or pH change. The P2 gel swelled and shrank upon heating to 70 °C and cooling to 25 °C because of the hydration and dehydration of the zwitterionic moiety (Supplementary Fig. 64 ). Therefore, the P1/P2 gel assembly reversibly bent and unbent upon cooling and heating (Fig. 4b ). We prepared a butterfly-shaped P1/P2 assembly, which beat its wings via temperature actuation (Fig. 4d ). Fig. 4 Preparation and actuation of responsive hydrogel assemblies. a , Schematic showing preparation of responsive hydrogel assemblies using P-Ru/P-S as an adhesive and shape changes of the hydrogel assemblies actuated by temperature or pH. b , Reversible bending of a P1/P2 gel assembly at different temperatures. The bending angles were measured after immersing the gel in water at 25 °C or 70 °C for 1 h. c , Reversible bending of a P1/P3 gel assembly at different pH values. The bending angles were measured after immersing the gel in an aqueous solution with pH 4 or 10 for 30 min. d , Photos of a butterfly-shaped P1/P2 gel assembly at different temperatures. e , Photos of a hand-shaped P1/P3 gel assembly at different pH values. f , Reversible detachment and adhesion of some arms of a five-arm P1/P3 gel assembly and shape changes of the gel assembly at different pH values. Scale bars, 5 mm. Source data We also prepared pH-actuated P1/P3 gel assemblies (Fig. 4c,e ). The P3 gel swelled at pH 10 and shrank at pH 4 because of the deprotonation and protonation of the acrylic acid moiety (Supplementary Fig. 65 ). Thus, pH changes induced bending/unbending of a rectangular P1/P3 gel assembly and triggered opening/closing of a hand-shaped P1/P3 gel assembly (Fig. 4c,e ). Bending and unbending were fully reversible for at least ten cycles (Fig. 4c ). This observation demonstrated that the P-Ru/P-S adhesive glued pH-responsive hydrogel was firmly assembled, even under large shape changes. Lap shear measurements showed that the P1/P2 gel assembly and P1/P3 gel assembly were glued firmly even after ten bending/unbending cycles (Supplementary Figs. 66 and 67 ). We interpret that the strong adhesion is attributed to the following reasons: (1) Ru–S coordination and polymer chain entanglement cooperate (Fig. 3 ); (2) the adhesion of P-Ru/P-S is independent of pH and temperature (Supplementary Fig. 63 ); and (3) the P-Ru/P-S adhesive is a hydrogel that can change shape and maintain its integrated network structure during swelling and shrinkage (Supplementary Fig. 68 ). Thus, the P-Ru/P-S adhesive can adapt to the shape changes of the gel assemblies. To verify that the P-Ru/P-S adhesive is strong yet reversible, we prepared a five-arm P1/P3 gel assembly using the adhesive (Fig. 4f ). The gel assembly showed reversible shape changes when varying the pH (Fig. 4f , f1 and f4). To illustrate the advantage of light and its high spatial resolution, some of the adhered arms were selectively detached via light irradiation (Figs. 4f , f2 and f3), and the gel assembly was changed to different shapes via pH activation. The detached arms could be readhered to form the initial five-arm structure (Fig. 4f , f4). Reconfigurable hydrogel assemblies with customized shapes The reversible P-Ru/P-S adhesive enables the fabrication of reconfigurable hydrogel assemblies with reprogrammable shape changes. We prepared assembly 1 with complex and heterogeneous structures by gluing a P1 gel, two P2 gels and two P3 gels together (Fig. 5a ). Assembly 1 changed into four shapes at different temperatures and pH values (Fig. 5b , first column). Moreover, the structure of assembly 1 could be reconfigured by light-induced detachment and readhesion using the P-Ru/P-S adhesive. Therefore, we prepared assemblies 2, 3 and 4 via reconfiguration using the same hydrogel units. Each gel assembly changed into four shapes upon pH and temperature stimulation (Fig. 5b ). These hydrogel assemblies are intelligent multishape transformers. The results showed that the use of the reversible metallopolymer adhesive is a strategy for the preparation of complex, heterogeneous and reconfigurable hydrogel assemblies with reprogrammable actuation functions. Fig. 5 Reconfiguring the structures of responsive hydrogel assemblies for multiple customized actuation. a , Schematic showing fabrication and reconfiguration of hydrogel assemblies with complex and heterogeneous structures and shape changes at different temperatures and pH values. Five hydrogel units can be glued using the P-Ru/P-S adhesive to form assembly 1. Assembly 1 can be reconfigured to assemblies 2–4. Each assembly can be changed to four shapes. b , Photos of hydrogel assemblies at different temperatures and pH values. Each of the four assemblies underwent a different shape change under each of the four conditions tested. Scale bars, 10 mm. Soft robot based on hydrogel assemblies for maze navigation We fabricated a soft robot with complex and heterogeneous structures by gluing a non-responsive P1 gel, a pH-responsive P3 gel and a magnetic particle-containing P4 gel using the P-Ru/P-S adhesive (Fig. 6a ). The P4 gel contained magnetic Ni particles so that it could move in a magnetic field. The P1/P3/P4 robot was firmly glued using the P-Ru/P-S adhesive. Thus, the P1 and P3 units followed the movement of the P4 unit in a magnetic field (Supplementary Movie 1 ). Fig. 6 Soft robot based on a responsive gel assembly for maze navigation. a , Schematic of a P1/P3/P4 robot prepared by gluing the gel units using the P-Ru/P-S adhesive. P1 is a non-responsive gel, P3 is a pH-responsive gel and P4 is a magnetic particle-containing gel. b , c , Schematic illustration ( b ) and photos ( c ) of the P1/P3/P4 robot passing through a maze under the control of pH and a magnetic field. The inset of c1 shows that the robot was too tall to pass through the gate. The inset of c2 shows that the flattened robot passed the gate. Scale bar: 20 mm. d , P1/P3/P4 robot under the action of a magnetic field (d1 and d2). After light irradiation, the P1/P3 gels and P4 gel were separated (d3 and d4). Scale bar, 5 mm. The P1/P3/P4 robot could go through a maze under the stimulation of pH and a magnetic field (Supplementary Movie 2 ). At pH 10, the robot was too tall to pass through Gate 1 (Fig. 6b,c , b2 and c1). When the pH was adjusted to 4, the robot became flat and passed through gate 1 under the guidance of a magnetic field (Figs. 6b,c , b2 and c2). However, the flattened robot could not pass through Gate 2 (Fig. 6b , b1). Therefore, the robot was reshaped by changing the pH to 10, and it passed through gate 2 under the guidance of a magnetic field (Figs. 6b,c , b1 and c3). In similar ways, the robot went through the additional gates and tortuous path of the maze (Fig. 6c , c4 and c5). To demonstrate that the robot is reconfigurable, the P1/P3 gels and P4 gel were separated via light irradiation (Fig. 6d and Supplementary Movie 3 ). The P1/P3 gels were reused as an actuator, which is the same as the actuator demonstrated in Fig. 4f . The P4 gel was reused by gluing it to a five-arm P1/P2 assembly to prepare a P1/P2/P4 robot (Supplementary Fig. 69 and Supplementary Movie 4 ). Because P2 was thermoresponsive and P4 was a magnetic particle-containing gel, the P1/P2/P4 robot passed through a maze under the control of temperature and a magnetic field (Supplementary Fig. 69 and Supplementary Movie 5 )."
} | 5,633 |
35663639 | null | s2 | 2,727 | {
"abstract": "Emerging bottom-up fabrication methods have enabled the assembly of synthetic colloids, microrobots, living cells, and organoids to create intricate structures with unique properties that transcend their individual components. This review provides an access point to the latest developments in externally driven assembly of synthetic and biological components. In particular, we emphasize reversibility, which enables the fabrication of multiscale systems that would not be possible under traditional techniques. Magnetic, acoustic, optical, and electric fields are the most promising methods for controlling the reversible assembly of biological and synthetic subunits since they can reprogram their assembly by switching on/off the external field or shaping these fields. We feature capabilities to dynamically actuate the assembly configuration by modulating the properties of the external stimuli, including frequency and amplitude. We describe the design principles which enable the assembly of reconfigurable structures. Finally, we foresee that the high degree of control capabilities offered by externally driven assembly will enable broad access to increasingly robust design principles towards building advanced dynamic intelligent systems."
} | 312 |
22632113 | null | s2 | 2,728 | {
"abstract": "Although natural silk fibers have excellent strength and flexibility, the regenerated silk materials generally become brittle in the dry state. How to reconstruct the flexibility for silk fibroin has bewildered scientists for many years. In the present study, the flexible regenerated silk fibroin films were achieved by simulating the natural forming and spinning process. Silk fibroin films composed of silk I structure were first prepared by slow drying process. Then, the silk fibroin films were stretched in the wet state, following the structural transition from silk I to silk II. The difference between the flexible film and different brittle regenerated films was investigated to reveal the critical factors in regulating the flexibility of regenerated silk materials. Compared with the methanol-treated silk films, although having similar silk II structure and water content, the flexible silk films contained more bound water rather than free water, implying the great influence of bound water on the flexibility. Then, further studies revealed that the distribution of bound water was also a critical factor in improving silk flexibility in the dry state, which could be regulated by the nanoassembly of silk fibroin. Importantly, the results further elucidate the relation between mechanical properties and silk fibroin structures, pointing to a new mode of generating new types of silk materials with enhanced mechanical properties in the dry state, which would facilitate the fabrication and application of regenerated silk fibroin materials in different fields."
} | 394 |
24705510 | PMC4012984 | pmc | 2,732 | {
"abstract": "Biological systems are collections of discrete molecular objects that move around and collide with each other. Cells carry out elaborate processes by precisely controlling these collisions, but developing artificial machines that can interface with and control such interactions remains a significant challenge. DNA is a natural substrate for computing and has been used to implement a diverse set of mathematical problems 1 - 3 , logic circuits 4 - 6 and robotics 7 - 9 . The molecule also naturally interfaces with living systems, and different forms of DNA-based biocomputing have previously been demonstrated 10 - 13 . Here we show that DNA origami 14 - 16 can be used to fabricate nanoscale robots that are capable of dynamically interacting with each other 17 - 18 in a living animal. The interactions generate logical outputs, which are relayed to switch molecular payloads on or off. As a proof-of-principle, we use the system to create architectures that emulate various logic gates (AND, OR, XOR, NAND, NOT, CNOT, and a half adder). Following an ex vivo prototyping phase, we successfully employed the DNA origami robots in living cockroaches (Blaberus discoidalis) to control a molecule that targets the cells of the animal."
} | 309 |
25641069 | PMC4648445 | pmc | 2,733 | {
"abstract": "The aims of this study were to evaluate the microbial diversity of different lignocellulosic biomasses during degradation under natural conditions and to isolate, select, characterise new well-adapted bacterial strains to detect potentially improved enzyme-producing bacteria. The microbiota of biomass piles of Arundo donax , Eucalyptus camaldulensis and Populus nigra were evaluated by high-throughput sequencing. A highly complex bacterial community was found, composed of ubiquitous bacteria, with the highest representation by the Actinobacteria , Proteobacteria , Bacteroidetes and Firmicutes phyla. The abundances of the major and minor taxa retrieved during the process were determined by the selective pressure produced by the lignocellulosic plant species and degradation conditions. Moreover, cellulolytic bacteria were isolated using differential substrates and screened for cellulase, cellobiase, xylanase, pectinase and ligninase activities. Forty strains that showed multienzymatic activity were selected and identified. The highest endo-cellulase activity was seen in Promicromonospora sukumoe CE86 and Isoptericola variabilis CA84, which were able to degrade cellulose, cellobiose and xylan. Sixty-two percent of bacterial strains tested exhibited high extracellular endo -1,4-ß-glucanase activity in liquid media. These approaches show that the microbiota of lignocellulosic biomasses can be considered an important source of bacterial strains to upgrade the feasibility of lignocellulose conversion for the ‘greener' technology of second-generation biofuels.",
"discussion": "Discussion In recent years, the competitive production of alternative renewable biofuels has stimulated research into new bacteria as a source of highly active and specific cellulases. They exhibit several advantages such as a fast growth rate, production of enzymes that are often more effective catalysts due to less feedback inhibition, and secretion of a complete multi-enzyme system for an efficient conversion of lignocelluloses into fermentable sugars 11 14 . In this context, particular attention must be given to exploring the biodiversity of natural niches so that cellulase-producing bacteria can be isolated and characterised. For these reasons, in this work, the microbial diversity of natural ecosystems, represented by lignocellulosic biomasses of A. donax , E. camaldulensis and P. nigra , was evaluated by culture-independent and culture-dependent approaches. A highly complex bacterial community was found, in which the most frequently occurring bacteria were those belonging to the Actinobacteria , Proteobacteria , Bacteroidetes and Firmicutes phyla. Proteobacteria was the most abundant taxa recovered in the E. camaldulensis and P. nigra piles, followed by Actinobacteria , Bacteroidetes , Firmicutes and Acidobacteria. These taxa are related to microorganisms previously characterised as biomass degraders. The biodiversity of the microbial community in our study corresponded well with a previous study in which Proteobacteria , Firmicutes and Bacteroidetes , along with members of the class Proteobacteria , comprised 83% of the microbial richness and heavily dominated switchgrass-adapted communities 15 . Moreover, bacterial species belonging to Proteobacteria and Acidobacteria , together with Firmicutes ( Clostridium and Bacillus genera) and followed by Bacteroidetes , Chlamydiae/Verrucomicrobia and Actinobacteria (mainly Streptomyces), are known as the major plant biomass-degrading microbes in peat swamp forests 16 . During the natural biodegradation process of plant substrates, the indigenous bacterial community would initially have grown by utilising the more accessible cellulose and hemicellulose and only later would use the more resilient lignin component 17 . With regard to lignin decomposition, Actinobacteria, Firmicutes and Acidobacteria are the major taxa involved in this process 18 . Acidobacteria was recovered in the late stage of our experiment and its abundance increased during the biodegradation process, especially in the E. camaldulensis piles. Firmicutes showed a similar trend in the A. donax and P. nigra biomasses, with the relative abundance of this taxon gradually increasing in the piles processed under the open field condition. Wu and He 19 reported that Firmicutes could be the main microbes for lignin depolymerisation since a dominance of this phylum was recovered in enriched microbial consortia using a medium with lignin. Moreover, this phylum is common in natural processes such as rice straw compost 20 and decaying wood 21 , suggesting its importance in the degradation of lignocellulolytic materials. Analysing the microbial diversity more deeply, Actinobacteria, γ-Proteobacteria, β-Proteobacteria, α-Proteobacteria, Acidobacteria, Sphingobacteria, Flavobacteria, Bacilli and Acidobacteria were recovered in all samples. In particular, Actinobacteria, α-Proteobacteria, Sphingobacteria and β-Proteobacteria , all potent plant polysaccharide-degrading microbes that play an important role in plant biomass degradation in the tropical peat swamp forest ecosystem 16 , were the most abundant taxa during the biodegradation process in all lignocellulosic piles. Different bacteria belonging to the Actinobacteria class are involved in complex glycoside degradation such as chitin and cellulose and are fundamental in lignin and polyphenol degradation 22 . Martins and co-workers 23 reported that biomass degradation in the composting process, including the deconstruction of recalcitrant lignocellulose, is fully performed by bacterial enzymes, most likely by members of the Clostridiales and Actinomycetales orders. β-Proteobacteria and α-Proteobacteria were also recovered in our study. According to Castillo et al. 22 , the dominance of members of the phylum Proteobacteria such as γ-Proteobacteria is observed only at the beginning of the biodegradation process, and its strong reduction during the experiment could be due to its involvement in lignocellulosic waste declining during the early stages of the process. In the open field experiment, Bacilli increased in both the A. donax and P. nigra biomasses. Members belonging to this taxon are known to have specific genes encoding enzymes involved in cellulose and hemicellulose degradation 24 25 . The constant increase in the relative abundance of Sphingobacteria and Acidobacteria recorded in the E. camaldulensis pile during the degradation process in the open field experiment suggests that these species play a role in the decomposition of lignocellulosic material. Kanokratana et al. 26 , using complementary shotgun pyrosequencing, identified different genes encoding glycosyl hydrolases targeting cellulose and hemicellulose degradation in a bagasse pile, most of which were found in orders Clostridiales , Bacteroidales , Sphingobacteriales and Cytophagales . Moreover, the Acidobactria taxon is able to a use a diversity of carbon sources, from simple sugars to complex plant biomass substrates 26 . Another adapted-lignocellulosic taxon was β-Proteobacteria , which increased in the E. camaldulensis and P. nigra piles. In recent work, Stursova et al. 27 identified β-Proteobacteria, \n Bacteroidetes and Acidobacteria as the primary cellulose decomposers in forest litter. An interesting finding was the high incidence of the uncultured bacterium CH21 in the P. nigra pile during the first phase of the degradation process in the open field. CH21 is a member of the phylum Armatimonadetes , formerly called candidate division OP10 28 . The phylum Armatimonadetes is very poorly studied and its phylogeny is still poorly defined 29 . Members of this phylum are detected in different ecosystems and they are phylogenetically different. This phylum includes species with a wide variety of metabolic potentials 30 . Wang and co-workers 31 reported that their prevalence in plant-fed anaerobic bioreactors indicates a role in degradation of plant material. In our research, CH21 is the only recovered member of this phylum. Since it's relative abundance increased in the Populus nigra pile after 45 days and decreased during the other phases of the degradation process in the open field, it is possible its involving in the biodegradation of this specific lignocellulosic biomass during the first phase of the process. The differing trends observed in this study in terms of taxa abundance during the biodegradation process and between the vegetable species used demonstrated a local selective pressure in the lignocellulosic ecosystems. The culture-dependent methodology used here, which was based on a functional approach of detection and isolation to find new lignocellulose-degrading bacterial strains, provides us with key insight. Special attention to the methodology was required to determine the optimal culture and assay conditions. A comparison with another study 11 revealed that differential substrates containing CMC and Avicel are effective for the enumeration and isolation of putative colonies of cellulolytic microorganisms 32 . According to Soares et al. 6 , exo- and endo-cellulolytic bacterial isolates are found at different frequencies, and the number of microorganisms that were able to grow on Avicel as the sole carbon source was high. The cellulolytic strains isolated from the biomasses showed multienzymatic activities useful to perform the hydrolysis of a complex substrate such as lignocellulose, an important initial step in many technological applications 33 that require the action of different specific enzymes. All forty bacterial strains submitted to the multienzymatic screening showed both endo- and exo-glucanase activities, confirming that these enzymes act synergistically during the saccharification of celluloses. These observations were reinforced by the fact that many of these bacterial strains also possessed β-cellobiase as well as cellulolytic activity. Xylanase activity was also commonly observed, which is unsurprising because a close correlation between cellulase and xylanase activities has been demonstrated and is due to their coexpression in the same operon. The Neighbour-Joining phylogenetic method generated a consensus tree that grouped all 16S rRNA gene sequences of the isolated strains into five different clusters at the class level. Culture-dependent data showed similar predominant bacterial classes detected by high-throughput sequencing in the lignocellulosic biomasses, with an abundance of Actinobacteria , Bacilli , α-Proteobacteria, \n γ-Proteobacteria and Sphingobacteria . Moreover, the bacteria isolated are dominant players since these microbial strains were isolated from a high serial decimal dilutions (10 −6 –10 −7 ). According to phylogenetic research on cellulose-decomposing bacteria isolated from soil carried out by Ulrich et al. 34 , Actinobacteria are the most prevalent bacterial group based on 16S rRNA gene sequences. This cluster included the species Curtobacterium citreum, which is able to use up to six different lignocellulose components and that is phylogenetically related to Microbacterium testaceum and Microbacterium lacus. In particular, previous studies reported the production of enzymes involved in cellulose and xylan degradation by Microbacterium species 35 . The Actinobacteria cluster included other genera involved in lignocellulosic biomass degradation, such as Cellulomonas . Akasaka et al. 36 reported that more than 60% of isolates from rice plant residues was closely related to Cellulomonas and involved in their degradation. Our study revealed the production of cellobiase and pectinase activities in Actinobacteria members, such as Cellulosimicrobium cellulans , Isoptericola variabilis, Promicromonospora sukumoe and Promicromonospora citrea , which belong to the suborder Micrococcineae . Many of the representatives of the Promicrosporaceae and Corinebacteriaceae families can degrade polysaccharides such as cellulose and xylan 37 38 . The Bacilli cluster was primarily represented by the Pediococcus and Bacillus genera on the basis of 16S rRNA gene sequence analysis. Interesting, Zhao and co-workers 39 reported the use of a strain of Pediococcus acidilactici with high tolerance to temperature and a lignocellulose-derived inhibitor in simultaneous saccharification and fermentation (SSF) for high lignocellulosic lactic acid production. The strains CA812, CA81, CA816, CA82 and CA81b, isolated from the A. donax biomass and phylogenetically correlated to the Bacillus genus, showed high endo-cellulolytic and multi-enzymatic activities. The capacity of Bacillus strains to produce large quantities of extracellular enzymes has placed them among the most important industrial enzyme producers isolated from compost, soil and several other natural habitats 40 . In particular, Bacillus amyloliquefaciens, Bacillus subtilis and Bacillus licheniformis , isolated from soil and compost, are able to hydrolyse cellulosic waste-material 8 by both cellulolytic activities 41 and multi-enzyme complexes 42 . The selected bacterial strains could have cross-specificity facilitated by specific or non-specific active sites and also distinct catalytic domains binding to different substrates 33 . Xanthomonas campestris , known as a phytopathogenic bacterium, and Lysobacter gummosus and Lysobacter enzymogenes , potent biocontrol agents that release cellulolytic enzymes such as glucanase 43 , were identified in the phylogenetic group of γ-Proteobacteria . Other strains of this class, such as Enterobacter sp. and Pantoea sp., although are known as insect-associated bacteria that are able to produce bioactive compounds and digestive enzymes that are responsible for lignocellulose degradation 44 , they are less attractive for possible biotechnological application since human disease has been reported to be caused by these bacteria as well as Raoultella terrigena and Sphingobacterium multivorum . The two strains CE77 and CE84, included in the α-Proteobacteria cluster and characterised for their peroxidase activity, were closely related to the species Novosphingobium resinovorum , previously isolated from soil and studied for its capacity to degrade oil resins 45 . Interestingly, the strain Aurantimonas altamirensis SBP73 showed multienzymatic activity. To our knowledge, this study is the first to report the multienzymatic activity of this specie. The bacterial strains screened in liquid medium showed enzymatic activity levels similar to the values reported in previous studies 46 . For example, Ekperigin 47 reported that the maximum activity of the cellulose-degrading enzyme determined for CMC in the culture supernatant of Branhamella spp. was 0.34 U mL −1 . Even though comparing cellulase production across studies is difficult because there are too many differences in the production of extracellular enzymes (e.g., media composition, fermentation conditions and raw materials), the results demonstrated that the bacterial strains isolated and characterised in this study could represent a very interesting biological source for the conversion of lignocellulose carbohydrates into products of commercial significance. Moreover, their biotechnological performance could be enhanced by modifying the biotechnological parameters of the fermentation process such as the temperature 33 . In conclusion, in this work, pyrosequencing-based technology increased our knowledge of lignocellulosic-adapted microbiota and the microbial dynamic during the degradation of three different biomasses under natural conditions. The dominant taxa found during the biodegradation process were members of classes Actinobacteria , γ-Proteobacteria , α-Proteobacteria and Sphingobacteria . However, the abundance of the major and minor taxa retrieved during the process was determined by selective pressure employed by the lignocellulosic plant species and the degradation conditions. In addition, new multifunctional degrading bacteria have been selected and identified that are potential producers of multiple enzymes that have synergistic actions on cellulose and hemicellulose. This step is fundamental for biotechnological applications of interest to industry because they constitute a microbial source of new multifunctional enzymes that can increase the efficiency of the hydrolysis of lignocellulosic biomass into fermentable sugars for biofuel eco-technology."
} | 4,152 |
36823152 | PMC9950430 | pmc | 2,735 | {
"abstract": "Understanding the interactions between plants and microorganisms can inform microbiome management to enhance crop productivity and resilience to stress. Here, we apply a genome-centric approach to identify ecologically important leaf microbiome members on replicated plots of field-grown switchgrass and miscanthus, and to quantify their activities over two growing seasons for switchgrass. We use metagenome and metatranscriptome sequencing and curate 40 medium- and high-quality metagenome-assembled-genomes (MAGs). We find that classes represented by these MAGs (Actinomycetia, Alpha- and Gamma- Proteobacteria, and Bacteroidota) are active in the late season, and upregulate transcripts for short-chain dehydrogenase, molybdopterin oxidoreductase, and polyketide cyclase. Stress-associated pathways are expressed for most MAGs, suggesting engagement with the host environment. We also detect seasonally activated biosynthetic pathways for terpenes and various non-ribosomal peptide pathways that are poorly annotated. Our findings support that leaf-associated bacterial populations are seasonally dynamic and responsive to host cues.",
"conclusion": "Conclusions Many recent review, perspective, and opinion pieces have urged integration of multi-omics approaches to improve understanding of the microbiome and its relationship to the host plant 44 , 78 – 81 . However, most integrative studies have focused almost exclusively on the rhizosphere as the compartment of soil-plant feedbacks and nutrient and water acquisition for the host. Though leaves are readily accessible for sampling, the phyllosphere microbiome is challenging to investigate using throughput, cultivation-independent approaches like metagenomics and metatranscriptomics. There are high levels of host and chloroplast contamination in leaf samples, and relatively low microbial biomass per leaf that that must be first dislodged from tightly-adhered biofilms. A signal from messenger RNA in metatranscriptome analysis is masked by abundant ribosomal RNA signal, leading to further challenges. These two challenges result in a relatively low proportion of usable sequences relative to the total sequencing effort (after quality filtering non-target signal) for leaf microbiome studies, which was true also in this work and is a limitation of it. Because of the combination of all of these challenges, much of our understanding of the phyllosphere, as the largest surface area of microbial habitation on Earth 26 , has been learned from studies that employ model hosts and synthetic or model microbial communities in controlled settings, or from description of the community structure by sequencing of marker genes, amplified and bioinformatically depleted of chloroplast genes to overcome the challenges of low signal and host contamination. Here, we report optimized laboratory protocols (to minimize host and chloroplast signals) combined with a genome-centric bioinformatic approach to perform focused functional gene and transcript analysis of seasonally dynamic yet persistent phyllosphere microbiome members. Our work is an untargeted bacterial metatranscriptomic work performed on the leaf phyllosphere of field-grown crops. Other recent leaf metatranscriptome studies have focused the viral communities of tomato and pepper 82 , soybean 83 , and rice 84 . A key strength of this work is the challenging integration of phyllosphere metagenome and metatranscriptome data, leveraging the higher coverage of the metagenomes with the activity information available from the metatranscriptomes. Despite the relatively limited coverage of the MAGs (due to substantial host and ribosomal DNA contamination), the analysis proved successful by integrating both datasets and focusing on genome-centric interpretation. Thus, there are likely many more prevalent and functionally active populations of the phyllosphere that were not captured in this study, including those players known to be key in the phyllosphere (e.g. 20 , 85 ). Substantial additional sequencing effort or an enrichment strategy would improve signal for a cultivation-independent approach to target those players. While the use of genome-centric approaches has the obvious shortcoming that we have obviously not captured every microbiome member, our approach does allow us to link actively transcribed functions to specific microbial membership. Furthermore, the functional genes and activities documented here are logical given current understanding of microbial adaptation to the host and phyllosphere environment. Overall, this work provides evidence of a thriving, dynamic, functionally diverse, leaf-specialized, and host-responsive microbiome on the phyllosphere of perennial grasses. It provides evidence of specific phyllosphere functions that are seasonally activated in a temperate agroecosystem and suggests several hypotheses of important host-microbe interactions in the phyllosphere, for example via central metabolism, isoprenoid biosynthesis, and stress response engagements. This research contributes to our broad understanding of the dynamics and activities of phyllosphere microbial communities, and points to specific microbial functions to target that could prove useful for plant-microbiome management.",
"introduction": "Introduction Perennial plants are a crucial target for the sustainable development of biofuels 1 – 3 . In addition to yielding high biomass that can be converted to biofuels and bioproducts, perennial crops offer a broad range of ecosystem services that support efforts to mediate climate change, including greenhouse gas mitigation and promotion of soil nutrient cycling 1 , 4 – 6 . Like all plants, perennials harbor diverse microbiota, and many of these microbes are either known or expected to benefit their hosts. For example, plant-associated microbes can increase productivity and protect against environmental stressors. Because of the intimate engagement of many plant-associated microbiome members with the host, management of the plant microbiome is one tool proposed to promote crop vigor and support crop resilience to global climate changes 7 – 10 . Therefore, along with selective breeding and data-informed field management, regulating the plant microbiome is expected to be strategic for the sustainable production of biofuel feedstocks. Plants have anatomical compartments that each are inhabited by distinctive microbial consortia. Generally, the diversity and composition of the plant microbiome narrow from external compartments to internal, and the plant plays an active role in filtering the microbiome composition inward 11 – 13 . External plant compartments include the root zone, rhizosphere and rhizoplane below ground, and the epiphytic phyllosphere above ground 14 . External compartments have a relatively higher representation of transient or commensal microbial taxa, and these compartments engage with and recruit microbes from the immediate environment. Internal compartments include the endosphere of above- and below-ground tissues, and these have relatively low richness and harbor the most selected microbiota 15 , 16 . Of these compartments, the rhizosphere has received the most attention as a critical site of microbial-plant interactions that are important for nutrient and water acquisition (e.g., Kuzyakov and Razavi 17 ). However, members of the microbiota that inhabit the phyllosphere also provide important plant functions, such as pathogen exclusion and immune priming 18 , 19 . Phyllosphere microorganisms have specialized adaptations to their exposed lifestyle 16 , 20 – 22 and contribute to global carbon and other biogeochemical cycling, including transformations relevant to climate change 23 – 25 , and inhabit the largest above-ground surface area 26 . Because perennial biofuel feedstocks often are selected to maximize foliage surface area, understanding the phyllosphere microbiome is expected to provide insights into microbial engagements that benefit the plant to support productivity and stress resilience. There are two general challenges in regulating the microbiome to promote crop vigor and resilience to environmental stress. The first challenge is to distinguish the beneficial members of the plant microbiome from transient or commensal members, with the recognition that some members that provide host benefits likely change situationally, either over plant development or given environmental stress 27 – 29 , while others are stable 30 , 31 . The second challenge is that plants and their agroecosystems are temporally dynamic over the growing season, and their associated microbiota is also dynamic. It is currently unclear what functions may be associated with phyllosphere microbial dynamics and their potential interactions with plant hosts. Previously, we used 16S rRNA gene amplicon analysis to identify a “core” cohort of bacterial and archaeal taxa that were persistently associated with the phyllosphere microbiomes of two perennial biofuel feedstocks, miscanthus, and switchgrass. Persistent membership was established by collecting leaf samples over replicated field plots, over a temperate seasonal cycle, and across two annual growing seasons for switchgrass 32 . Other studies in switchgrass have similarly reported that the leaf can be distinguished from other plant compartments by its microbiome composition (e.g., 33 – 37 ), suggesting selection to the leaf compartment. In the current study, we aimed to understand the functional attributes and activities of persistent phyllosphere taxa, with an interest in specialized adaptations to the leaf and interactions with the host plant that may inform the mechanisms and nature of plant-microbe engagements. Therefore, we performed a seasonal analysis of phyllosphere metagenomes for miscanthus and switchgrass using nucleic acids from the same samples as used for amplicon analysis. We paired our metagenome longitudinal series with metatranscriptome analyses at select time points in switchgrass phenology to determine which functions were active seasonally. We performed genome-centric analyses of the metagenome and focused on understanding the seasonal dynamics and functions of a focal subset of medium- and high-quality metagenome-assembled-genomes (MAGs) that we could bin from these data. Our results reveal functions supportive of a leaf-associated lifestyle and seasonal activities of persistent phyllosphere members. Finally, we provide evidence that these genomes were detected in various sites and years beyond our original study plots, suggesting that they are general, consistent inhabitants of midwestern U.S. bioenergy grasses.",
"discussion": "Discussion Here, we report a multi-year seasonal metagenome and metatranscriptome assessment of plant phyllosphere in agricultural field conditions, focusing on the bacterial functions associated with two promising biofuel feedstocks. We expect these findings to have relevance for other grasses or systems with substantial aerial biomass, including native prairie. Furthermore, our collection of MAGs included phyllosphere members previously reported as abundant, persistent, or significant for microbiome assembly in other plants, including the model Arabidopsis (e.g., 41 , specifically: Sphingomonadales, Pseudomonadales, Actinomycetales, Burkholderiales, and Rhizobiales. Therefore, the patterns and consistently detected functions among this curated collection offer general insights as well as seasonal phyllosphere functions. There are multiple lines of evidence that the focal MAGs discussed here represent ecologically important lineages in the switchgrass and miscanthus phyllosphere. First, there is ample overlap among these focal MAGs and the core taxa of high abundance and occupancy from our previous 16S rRNA gene amplicon analysis of the switchgrass and miscanthus phyllosphere diversity 32 , including a MAG associated with Hymenobacter M9, which is a genus that had high occupancy across fields and over time, as well as several OTUs assigned. This same time series was investigated in our previous amplicon analysis, and the core taxa were prioritized using consistency across replicated fields at the same time point, persistence over time, and relative abundance. This suggests that the populations represented by the MAGs are not rare taxa that are transient to the system. Second, we were able to generate quality assemblies from the complex metagenome data, which is a process that is generally biased towards abundant members. These MAGs are highly represented in read abundance in metagenomes and recruited relatively more metatranscriptome reads as well, suggesting that they are both abundant and active in the phyllosphere. Though it is possible that there are additional, ecologically important lineages missing from our collection, we are confident that those discussed here are among the major host- or environment-selected populations inhabiting the switchgrass and miscanthus phyllosphere. Stress responses: trehalose, betaine, reactive oxygen, IAA Trehalose is a disaccharide that protects cells against salt, water, and osmolyte stress by serving as a stabilizing chemical chaperone, either by displacing water from protein surfaces or by vitrifying around protein structures to shield them 42 . Similarly, betaine is another commonly biosynthesized osmoprotectant used by microorganisms to contend with water, salt, and temperature stress 43 . Both have been hypothesized to be important survival strategies of microorganisms in the phyllosphere, with supporting evidence from isolate genome analyses 44 . Here, we show both trehalose and betaine to be prominent among MAG populations and activated consistently on the leaf surface, suggesting that they are not seasonally activated but rather necessary for the leaf-associated lifestyle. In bacteria, trehalose metabolism prevents cellular overflow metabolism and carbon stress by redirecting glucose-6-phosphate from conversion to pyruvate 42 . Trehalose biosynthesis is common among bacteria and archaea that live in arid, saline, thermal, or seasonally dry environments (e.g., 45 – 47 ) and it has also been reported to be induced in a pseudomonad by ethanol 48 , which can originate inside plant cells 49 or more generally in the roots (e.g., 50 ), especially, during stress, fruit ripening, or senescence 51 . In switchgrass and plants in general, trehalose concentration is increased in response to drought conditions 52 , and its precursor, trehalose-6-phosphate, induces senescence when carbon is readily available 53 . Furthermore, the A. thaliana phyllosphere member Sphingomonas melonis was reported to regulate trehalose biosynthesis during growth conditions that promoted mild stress 54 . Given this, it makes sense that the majority of MAGs had enrichment of transcripts related to trehalose metabolism, which would support a plant-associated lifestyle during drought and host senescence 54 . Betaine biosynthesis in bacteria often begins with the oxidation of choline, which is a part of plant tissues and can be transported into the cell 55 . Indeed, choline degradation was detected in 9/10 Gammaproteobacteria MAGs here (Fig. 6 ). Microbial-derived osmolytes such as betaine and trehalose have been suggested as targets for biotechnology development to support crop stress tolerance. However, plants can biosynthesize betaine and can be divided into groups of those that do and do not accumulate it in concentrations that are supportive of stress tolerance 56 . For example, in switchgrass, the concentration of glycine betaine was not predictive of differences in drought tolerance among different genotypes, while trehalose was, along with abscisic acid, spermine, and fructose 52 . Furthermore, our data suggest no notable microbial limitations in the genetic potential or activation of betaine biosynthesis in the phyllosphere. Reactive oxygen species (ROS) serve as signals for various developmental and cellular processes in plants, and here we detected active pathways for ROS degradation in the majority of focal MAGs. Though the precise mechanisms are unclear, homeostasis of ROS is expected to be involved in senescence 57 , which is relevant to our study given that at least some to a majority of senescing plants were observed per plot in August and September sampling dates, respectively. Additionally, ROS accumulates in plants that are exposed to abiotic stress, to negative effects. ROS degradation is one of many functions phyllosphere microorganisms employ to contend with expected fluctuations in ROS on the leaf surface (though these fluxes are difficult to measure 57 ). Previously, several genes relevant to oxidative stress response were differentially regulated in a wild-type phyllosphere bacteria Sphingomonas melonis strain Fr1 as compared to a knock-out mutant for regulation of general stress response when both were grown in a medium expected to induce low levels of stress 54 . Given that managing plant ROS is a target for reducing crop stress and regulating plant development 57 , 58 , it is possible that manipulating microbial ROS degradation could be applied as one tool to achieve such efforts, but much more research is needed to understand the microbial-host interaction given ROS exposure or accumulation, and any possible ROS signaling between them. IAA is a phytohormone produced by plants to regulate many processes in growth and stress response (e. g., 59 , 60 ). It is also made by many microorganisms, including those shown to support plant growth promotion (e. g., 60 ). Therefore, the activity of IAA degradation pathways by focal MAGs is expected, given the redundancies between plants and microorganisms in synthesizing and responding to IAA, and demonstrates microbiome responsiveness to feedback in the host environment. Biosynthesis of isoprene-related molecules Most of the functions identified in our MAGs suggest general requirements for growth and maintenance given a leaf-associated lifestyle (e.g., carbohydrate and amino acid metabolism, pigment production to protect from radiation, similar to previous reports, e.g., 61 . However, BGC analysis revealed surprising consistency in terpene metabolism pathways, leading us to look more closely at transcript ORFs associated with terpenes. This analysis revealed particular enrichment in pathways and genes associated with isoprenoid biosynthesis. Isoprenoids are a class of volatile terpenes that are generally abundant and reactive, and they engage in indirect and complex feedback with methane and nitrous oxide greenhouse gases 62 . Isoprene is one of the simplest isoprenoids. It is released by many plant species, and much of it is synthesized within the methylerythritol phosphate pathway of the chloroplast (MEP, aka:non-mevalonate) 63 . Isoprene is thought to act as a signaling molecule in stress response 64 . Studies have also found that isoprene emission protects leaf photosynthesis against short episodes of high temperature 65 . Plants emit isoprene from matured, photosynthetically active leaves, and emissions are light-responsive 63 . However, senescing leaves have been reported to decrease in their isoprene emissions relative to their leaves at peak growth 66 . Both switchgrass and miscanthus have been reported to emit relatively low basal levels of isoprene 67 , 68 . We speculate that members of the biofuel feedstock phyllosphere bacterial community may be either compensating for the loss of plant-derived isoprene, engaging in interspecies isoprenoid signaling with the host, protecting plant photosynthesis from thermal damage, quenching reactive oxygen species, or possibly producing isoprenoids as overflow metabolites (as hypothesized for Bacillus subtilus 69 ). Bacterial isoprene degraders and synthesizers are widespread in nature 62 and have been previously investigated in phyllosphere communities of the relatively high isoprene emitter Populus spp 70 , as well as in soils 71 , which can serve as an isoprene sink. Stable isotope assays have been used to determine that a subset of bacteria community members degrade isoprene, including several Actinobacteria ( Rhodococcus spp.) and Variovorax (Proteobacteria) 70 , 71 . Our MAG collection contains several Actinomycetia, a Hymenobacter , several Methylobacterium , and Pseudomonas MAG S28 that show activation of genes involved in isoprenoid biosynthesis and add support for their involvement in related molecular feedbacks in the phyllosphere. In addition, as these activities were detected in three Bacteria phyla, it suggests that biosynthesis of isoprene-related molecules may be a very common phyllosphere microbiome function. As isoprene is a precursor to sidechains needed for several quinones 72 , it could be speculated that leaf bacteria scavenge isoprene emitted by the host plant to supplement bacterial synthesis of these sidechains but then compensate with de novo biosynthesis if the host decreases production. We observe that isoprenoid synthesis increases seasonally in the majority of MAGs containing these pathways and concurrently with when plant isoprene emissions also are expected to decrease, and directs future work to understand these dynamics and potential isoprenoid-mediated bacterial-host engagement. MAGs of interest We highlight three MAG populations that were of interest because of their taxonomy, functions, dynamics, or occupancy. All were shared with taxa in our prior 16S rRNA survey, supporting their inclusion as part of the “core” set that was selected by abundance and occupancy. First, high-quality MAG S28 (>97% complete, <2% contamination) was a prominent pioneer and active colonizer of the leaf (Figs. 4 , S2 Group 1). MAG S28 is related to Pseudomonas cerasi , a species reported to have phytopathogen relatives 73 , but we did not note any disease symptoms on the leaves analyzed. This population had expected traits of a strong surface colonizer, including colonization, adaptation, and motility subsystems. It also had six pathways related to phytohormone responses (out of 7 total phytohormone pathways observed in these data), including activated ethene biosynthesis, ACC deaminase, and degradation of ethylene glycol, putrescine, salicylate, and IAA. These data suggest that S28 has several mechanisms to engage or respond to the host via phytohormones. Next, MAG M9, identified as Hymenobacter , was of interest because it was associated with the most numerous taxonomic group detected in our prior 16S rRNA gene survey 32 and not among the most typically investigated phyllosphere lineages in the literature. While M9 populations were first detected early in the season, their transcripts were enriched in the late season (Figs. 4 and S2 Group 5). MAG M9 had detected and activated galactonate, N-acetylglucosamime, and lactose metabolisms, which were not common among the focal MAGs. It also had activated benzoate, curcumin, and putrescine degradation, as well as cyanide detoxification, type VI secretion, and dihydrogen oxidation. While M9 also had some pathways that were common among these MAGs (e.g., ROS and IAA degradation, terpene biosynthesis), its suite of more sparsely detected pathways and functions suggest a specialized role in the phyllosphere community. Notably, M9 has 65% completion and 0% detected contamination, suggesting more functional potential remains to be discovered for this and similar Hymenobacter lineages inhabiting the phyllosphere. Finally, we selected a representative Actinomycetia MAG M60, a Quadrisphaera lineage that had activated isoprenoid biosynthesis and had increased activity late in the season, along with the majority of focal MAGs (Figs. 4 and S2 Group 3). Studies have found that members of Actinomycetia are an important part of the phyllosphere that contribute to disease prevention and plant growth 74 – 76 . MAG M60 had several oligo/polysaccharide metabolisms that were infrequently detected in these data, including glycogen, melibiose, and trehalose. Despite its high completeness and low contamination (>95% and <5%, respectively), M60 was sparsely annotated by the methods we applied. However, Quadrisphaera taxa have been reported to be highly abundant in the phyllosphere or endosphere of various plants 77 ."
} | 6,076 |
29094002 | null | s2 | 2,736 | {
"abstract": "It is highly desirable to develop a universal nonfouling coating via a simple one-step dip-coating method. Developing such a universal coating method for a hydrophilic polymer onto a variety of surfaces with hydrophobic and hydrophilic properties is very challenging. This work demonstrates a versatile and simple method to attach zwitterionic poly(carboxybetaine methacrylate) (PCB), one of the most hydrophilic polymers, onto both hydrophobic and hydrophilic surfaces to render them nonfouling. This is achieved by the coating of a catechol chain end carboxybetaine methacrylate polymer (DOPA-PCB) assisted by dopamine. The coating process was carried out in water. Water miscible solvents such as methanol and tetrahydrofuran (THF) are added to the coatings if surface wettability is an issue, as for certain hydrophobic surfaces. This versatile coating method was applied to several types of surfaces such as polypropylene (PP), polydimethyl siloxane (PDMS), Teflon, polystyrene (PS), polymethylmethacrylate (PMMA), polyvinyl chloride (PVC) and also on metal oxides such as silicon dioxide."
} | 273 |
28373853 | PMC5357650 | pmc | 2,737 | {
"abstract": "Social interaction occurs across many time scales and varying numbers of agents; from one-on-one to large-scale coordination in organizations, crowds, cities, and colonies. These contexts, are characterized by emergent self-organization that implies higher order coordinated patterns occurring over time that are not due to the actions of any particular agents, but rather due to the collective ordering that occurs from the interactions of the agents. Extant research to understand these social coordination dynamics (SCD) has primarily examined dyadic contexts performing rhythmic tasks. To advance this area of study, we elaborate on attractor dynamics, our ability to depict them visually, and quantitatively model them. Primarily, we combine difference/differential equation modeling with mixture modeling as a way to infer the underlying topological features of the data, which can be described in terms of attractor dynamic patterns. The advantage of this approach is that we are able to quantify the self-organized dynamics that agents exhibit, link these dynamics back to activity from individual agents, and relate it to other variables central to understanding the coordinative functionality of a system's behavior. We present four examples that differ in the number of variables used to depict the attractor dynamics (1, 2, and 6) and range from simulated to non-simulated data sources. We demonstrate that this is a flexible method that advances scientific study of SCD in a variety of multi-agent systems.",
"conclusion": "Conclusion Understanding how large-scale, multi-agent social systems coordinate is challenging and complex. In part, the challenge is due to the fact that there are so many agents, system components, and potential system states that can become coordinated; all of which may change over time (Van Orden et al., 2003 ). These many components interact generating higher order system behavior that is emergent and dynamic. However, knowing the “Dynamics demystifies…emergence” and it can also provide “basic laws for a quantitative description of phenomena that are observed” (Kelso, 2009 ; p. 1540). As such, we have expanded on work in SCD by demonstrating the utility of modeling the attractor dynamics of several systems to characterize their higher-order behavioral patterns and showed how these patterns varied over time and could be linked to meaningful aspects of the systems. Within domains, such as agent based modeling, qualitative depictions of higher-order patterns are often known, but not quantitatively modeled. In SCD, phenomena can be non-rhythmic, and yet dynamically coordinated. They can exhibit stability and multistability. Thus, using attractor dynamic descriptions along with statistical innovations, such as mixture modeling, provide a reasonable solution to understanding the large-scale, multi-agent social coordination. Characterizing the higher order properties of the system in this way forms a foundation for examining the emergent patterns through time in either a confirmatory or exploratory manner. This same technique, as we have shown, can be utilized with simulated as well as observational data. It is our aim that we recognize that we study systems that are inherently open systems (even though simulations are often closed). By examining part of the system (the variables we measure), unobserved aspects of the system function as perturbations to the system. Thus, a system depicting families is open because we are only examining some of the variables involved. To understand how agents exhibit emergent self-organization and coordination, we have advanced a general quantification that can be applied to a range of social systems, such as two individuals that form a couple up to a crowd's behavior. We hope that the widely applicable techniques will be adopted to advance scientific understanding of SCD.",
"introduction": "Introduction For many animals and humans, social interaction is pervasive in daily life. Social interaction occurs across many time scales and varying numbers of agents; from one-on-one to large-scale coordination in organizations, crowds, cities, and colonies. Since social interactions occur at different scales, and in ways that change dynamically over time, they can be quite a complex phenomenon to study without appropriate guiding theoretical and methodological frameworks. In dynamical systems theory, complexity arguably occurs due to the emergent, self-organizational nature of the system. Emergent, self-organization here implies that there are higher order macroscopic patterns occurring over time that are not necessarily due to the actions of any particular controlling agents or components, but rather due to the collective ordering that occurs from the individual interactions of the agents or components of the system (Halley and Winkler, 2008 ). Taken together these diffuse interactions contribute to more macro-scale phenomena that are observed over time. Common examples of this type of emergent, self-organization of social behavior occurs in flocking birds and in schools of fish that appear to move in a highly coordinated fashion (e.g., Couzin and Krause, 2003 ). Because of the multitude of agents (or system components) that give rise to emergent patterns, it is often difficult to determine how one should depict the resultant system. In line with approaches to social coordination dynamics (SCD), we aim to uncover the dynamic processes that underlie the ways in which agents are able to organize their behavior and change together in time (Oullier and Kelso, 2009 ). This emergence is a form of c oordination that specifically implies the occurrence of a functional ordering of components that interact across spatial and temporal dimensions, often with multi-directional relationships (Kelso, 2009 ; Butner et al., 2014a ). We aim to model this emergent, multi-agent coordination through attractor dynamics depictions (which we discuss in detail in the next section). From the SCD perspective, two or more agents are able to coordinate their behavior based on some form of mutual information exchange. This information exchange generates coordinative structures with higher order patterns not easily identifiable from the lower order interactions. The resultant higher order patterns are then depicted through attractor dynamics in which the patterns are attributed with stability properties implied by an underlying topology (Kelso, 1995 ). SCD is thus consistent with notions of weak emergence (Bedau and Humphreys, 2008 ), but the resultant patterns have then been modeled using attractor dynamics descriptions depicting patterning over time that have stable properties under perturbations. One major distinction between SCD and examples of weak emergence (usually through agent-based or cellular automata models) is in the scale of the social systems involved. Agent-based models are usually quite large-scale social systems, while SCD has often focused on a dyadic scale of analysis. SCD has excelled in generating models of intentional and spontaneous dyadic interpersonal rhythmic behavior such as finger or limb oscillations (e.g., Haken et al., 1985 ; Schmidt et al., 1990 ; Oullier et al., 2008 ), swinging pendula (Schmidt and O'Brien, 1997 ), and rocking in chairs (Richardson et al., 2007 ). Some recent work has provided ways to assess social interactions in larger scales such as coordination of groups bigger than dyads (e.g., Richardson et al., 2012 ; Duarte et al., 2013 ). One challenge is to generate models of emergent, multi-agent coordination in social systems where the agents may not behave rhythmically per se , but are following some organizing rules or structures that give rise to coordinated behavior serving a functional purpose. In the current paper we build on SCD approaches, by modeling the results from large-scale agent-based systems as a function of attractor dynamics. Our chosen technique utilizes mixture modeling in conjunction with topological equations to represent attractor dynamics. This approach is particularly attractive in that topological representations of phase space can yield more qualitative information in comparison to other time series approaches (Strogatz, 2014 ), generating a more complete picture of the underlying system dynamics. Specifically, we examine a series of agent-based examples, and model each set of time series as a function of their changes through time. We show how sets of linear equations can depict the higher order emergent patterns in ways consistent with attractor dynamics. The advantage of this approach is that we are able to quantify the self-organized dynamics that agents exhibit, link these dynamics back to activity from individual agents, and relate it to other variables central to understanding the coordinative functionality of a system's behavior. Our goal is to exemplify the strategy. In all, we present four examples that differ in the number of variables used to depict the attractor dynamics (i.e., the dimensionality of the systems) and range from simulated to non-simulated data sources. Attractor dynamics In dynamical systems theory, the concern is often placed on what states a system is drawn toward, or away from, as it changes over time (e.g., Richardson et al., 2014 ). This epitomizes the notion of attractor dynamics. Attractor dynamics are merely a mathematical way of expressing repetitious behavior in the face of constant disruptions to those repetitions. The constant disruptions are inherently part of the system in that open systems are dissipative and function far from equilibrium to maintain patterns (Prigogene and Stangers, 1984 ). By only examining a portion of a system, as is common in empirical research, the unexamined variables are treated as constant disruptions or perturbations to those patterns. These repetitious behaviors describe the most probable system states and their ability to remain in these states (while facing perturbations) conveys the inherent stability of those states. These attractor dynamics can then be modeled using differential/difference equations, allowing for exploration of the inferred dynamics and theorizing the manifolds in which the system functions (Differential equations are based on idealized models for when change in time approaches zero while difference equations estimate models using the observed discrete differences; Butner et al., 2015 ). Assuming a system exhibits stability, the emergence of a limited set of patterns, which can be described in terms of topological features, are plausible. These topological features can be described using map analogies, because there is a strong tie between topology and maps. In fact, differential topology is the math behind maps. Traditionally, topographical maps convey elevation of a landscape. But, the notion of topologies can also be applied as a graphical representation of how data are changing over time. To ease the interpretation of differential topology, we will temporarily link movement on maps to different topological features. This interpretation is directly relevant to several of the examples (although more general definitions are extant; Butner et al., 2015 ). An Attractor is when the agents are drawn toward a particular coordinate over time or a particular directional heading. This is akin to a topographical valley. A Repeller is a coordinate that agents move away from. These would be reflected topographically as a mountain peak. A Saddle occurs when agents are attracted in one dimension and repelled in another. It is analogous to a topographical ridgeline because it can separate different patterns such as two attractors (Abraham and Shaw, 1983 ; Butner et al., 2015 ). A Cycle corresponds to a push/pull of two dimensions on one another. Combined versions of patterns described can also be observed such as spiral attractors where there are circling movements for how agents converge toward an attractor. Saddles and cycles require at least two dimensions and thus will only be possible in the later examples (not merely with heading as it is a one dimensional example). To continue with the link to maps, we will begin with agent-based models that function spatially. As a simplification, we can reduce their behavior to movement along an X and Y axis or merely the directional heading of agents (when we only require a single dimension to depict the system). We can then model the simultaneous change of these variables over time. In this way, we capture the movement of many agents and can characterize them with attractor descriptions. With this information it is possible to examine and identify patterns of change for the overall system using the particular topological features defined above to describe how multiple agents are moving over time (Butner et al., 2015 ). It is in these terms that we gain an understanding of the emergent, coordination of many agents. As a beginning example, consider a Flocking agent based model (Wilensky, 1998 ) in NetLogo v5.2.1 (Wilensky, 1999 ) designed to emulate the self-organized behavior of how flocks of birds might come to match one another's movements creating complex group behavior. Agents start with a random heading and constant velocity in a wrapped environment (makes a torus). The heading for each agent is determined by three rules: (1) alignment states that each agent tends to turn to be moving in the same direction as nearby agents; (2) separation states that each agent will turn to avoid an agent when it gets too close; and (3) cohesion states that agents tend to move toward other agents. As the agents “fly” through the two dimensional environment they update their headings over time. Figure 1 shows time series of the headings for all (300) agents simultaneously over one thousand iterations. It is clear that early in the simulation the full range of headings are observed yet, in later times the range of headings become more restricted and shared by the agents. This is an example of the emergent coordination that occurs within the Flocking model. Figure 1 Time series of the headings for 300 flocking agents . Note that the flock moves toward a very restricted heading. To depict these results topologically requires identifying the underlying map in which the agents are interacting. An attractor, in this case, would be the heading(s) in which the agents move toward and the stability would be the resistance exhibited in the system when an agent begins to diverge from this attractive heading and pulled back thusly. The map is not one of actual hills and valleys, but instead the resultant decrease in heading directions, from the emergent self-organization between agents. Thus, the trajectories for agents imply an underlying pattern that we can infer. One assumption of dynamical systems theory is that there is one—or perhaps multiple—underlying patterns emergent from the interactions of individual agents over time. Interactions result in a consistent pattern, that the system flexibly returns to when interactions or outside forces briefly move the system away from its primary pattern. This notion of consistency in the face of perturbations is stability. While it is easy to observe the convergence of heading amongst the agents in Figure 1 , little information can be drawn in regards to the number of underlying patterns and their inherent stabilities. The flocking example is a useful one in that the implied map is not a map of X and Y coordinates, but one of heading—it is a one dimensional map. One dimensional maps are not very interesting to draw; they are a line showing where the data converges over time. That is, attractor dynamics are time implicit models rather than time explicit ones and thus, are akin to collapsing the X axis in Figure 1 , while adding in notions of where each agent goes next to determine the map. Dynamical systems theory has long provided the theoretical framework and terminology for describing multi-agent self-organized patterning. Returning to Figure 1 , an apt depiction of the Flocking simulation is one that begins with many attractors that cease to exist over time, which produce a limited set of stable attractors. This qualitative description captures the evolving process, without any of the quantitative dynamics. By quantifying them through topological equation representations, we can further differentiate aspects of the system and specify the strength of the attractors. We therefore next cover the step of quantifying the dynamic. A vector based approach There are several ways to estimate differential topological equations. In all cases, we must first express the data in terms of data vectors rather than values. For the heading data illustrated in Figure 1 , the data is structured such that two or more points in time are used to define a data vector, known as a time delay or Toeplitz data structure (e.g., Boker and Laurenceau, 2006 ). The data is structured so that a value at time t and a value at time t +1 are two variables within the model. Further, our models are all estimated in structural equation modeling wherein change was built into the models themselves as latent variables (McArdle, 2009 ). One can also estimate change directly through a discrete difference or various methods for estimating derivatives and thus while we use structural equations to build our models, this is far from a necessity (Boker et al., 2010 ). Different attractor dynamics are then captured through expressions of change predicted by value. For example, Equation (1) expresses the potential dynamics for the headings of the various agents.\n (1) ẋ t = b 0 + b 1 x t + e t \nCurrent heading of a given agent at time t is x, x -dot is an estimate of its derivative with respect to time, b 0 is the intercept, b 1 , the slope with respect to x and e t is error. For clarity, this equation is written in regression form where velocity in heading at each point in time is treated as the criterion and position (current heading) is the predictor. When the slope in Equation 1 is negative, we observe an attractor where the time series are attracted toward a value of − b 0 / b 1 known as the set point (Butner et al., 2015 ). A repeller occurs when the slope is positive instead of negative. The strength of attraction/repulsion is defined by the steepness of the slope relative to zero. Equation (1) is limited in that it can only capture a single topological feature (Butner et al., 2015 ). While the system may converge to a single heading, this convergence is developed over time. Consistent with the qualitative description of the flocking model, we should observe several patterns that cease to exist as time continues. This results in a much more limited set of dynamic patterns that occur at later times. We therefore expand our approach to allow for multiple sets of Equation (1). We did this through an analytic technique known as mixture modeling. Mixture modeling methods Mixture modeling is a taxonomic approach that can be combined with structural equation modeling (Enders, 2006 ) as an alternative way to capture interactions (Jung and Wickrama, 2008 ). Non-linear dynamical systems allow for multiple topological patterns by building non-linear transformations, such as the interaction and therefore mixture modeling can be used as a way to capture the different topological features by slicing up the overarching state space under an assumption that each dynamic is locally linear. One description of mixture modeling is as a multiple group analysis (stacked model), where assignment to group is unknown (Muthen, 2001 ). Multiple group models allow for different parameters across groups. We can extract different equation sets by allowing key parameters to differ across these groups while equating others. Specifically, we allowed the slope coefficients characterizing how position predicted each velocity, the intercepts for the velocity factors, the means for the position factors, the residual variances for the velocity factors, and the variances for the position factors to vary across sets of equations [see Appendix A (Supplementary Material) for an example in Mplus (Muthén and Muthén, 1998–2012 )]. As previously described, the sign of the slope coefficients capture the type (e.g., attractor, repeller, limit cycle) and strength of attraction for the dynamic implied by the equations (see also Butner et al., 2015 ). In addition, the velocity intercepts help determine the set point, or relative position to which the dynamics can be described (e.g., the location of the attractor). Following logic laid out under notions of centering and simple slopes analysis (Cohen et al., 2003 ), the means and variances for the position factors help depict common trajectories implied by the pattern and thus help identify the basin of attraction. By allowing for variation in these parameters across latent classes, we can infer a number of varying topological features, as opposed to a single feature. Mixture modeling can be used as a confirmatory or exploratory method. In either case, there must be established criteria for fit. The current preferred methods are through forms of the Bayesian Information Criterion (BIC) or through forms of model testing such as log likelihood or chi-square comparisons to see if the current number of extracted groups improves description of the data beyond the previous number of groups. Specifically, the BIC and sample size adjusted BIC tend to minimize when the proper number of mixture groups has been extracted (Sclove, 1987 ; Nagin and Tremblay, 1999 ; Nylund et al., 2007 ) and both have been used under different circumstances usually relating to the sample size (sample size adjusted BIC is preferred when n < Bauer and Curran, 2003 ; Lubke and Neale, 2006 ; Enders and Tofighi, 2007 ). Model identification can also be informed by various likelihood ratio tests (LRT), which are used to test relative model fit by testing the null hypothesis that competing models demonstrate comparable fit (Vuong, 1989 ). Within latent variable models such as the present one, the Vuong-Lo-Mendell-Rubin test (Lo et al., 2001 ) is an accepted methodology for testing the equivalence of two associated probability density functions (Henson et al., 2007 ). Simulation studies have indicated that the VLMR test favors selection of more components when used with small samples, resulting in increased Type I error rates; this suggests the need for an adjusted test (aVLMR) with samples less than 300 (Lo et al., 2001 ). For our purposes, we chose to rely on the BIC. Note that our data had an inherent dependency—the nesting of multiple measures through time within each agent. Ignoring a data dependency is known to produce biased standard errors with large alpha inflation as the common result (Cohen et al., 2003 ). However, current mixture modeling practices that incorporate methods for accounting for the dependency preclude any descriptions of predictors. In this case, that would result in the loss of the means and variances for the position factors that detail key information about the basins of attraction. We therefore chose to temporarily ignore the dependency, recognizing that the standard errors for each coefficient may be biased toward Type 1 errors. To better understand the extracted equation groups, we saved out the posterior probabilities for each data vector. This is the probability that each instance in time for a given agent belonged to one of the classes characterized by a particular equation set where the set of posteriors for a given vector sum to one. It is the equivalent of factor scores if mixture groups as likened to a categorical latent variable. The value of the posterior probabilities is that they allow us to specifically link each agent to the various attractor dynamics at each point in time. Through the combination of the description of each attractor dynamic and the posterior probabilities linking the agents to the topologies, we are able to traverse between the observed vectors from the agents to the underlying topology. One dimensional systems What follows is an illustration of the analytic strategy for the flocking example using the headings from all agents. Fit indices of the 300 flocking agents over 1,000 iterations resulted in sixteen unique attractors (as indicated by the BIC at its lowest value). Table 1 contains the estimated parameters for each of the sixteen equations. All sixteen patterns are attractors as indicated by the negative slopes. They vary in their stability, indicated by the range of slopes. The headings to which each pattern indicates a point of attraction is identified by converting the intercepts and slopes into the set point (− b 0 / b 1 ). In essence, the flock example is characterized by a total of sixteen unique attractors. Table 1 Unstandardized coefficients from the sixteen attractor solution for the Flocking model of headings . Pattern Intercept Slope 1 348.528 (1.485) −1.428 (0.007) 2 300.932 (1.632) −1.402 (0.007) 3 235.521 (3.214) −1.306 (0.013) 4 197.095 (3.055) −1.202 (0.093) 5 255.672 (1.519) −1.174 (0.061) 6 147.377 (2.554) −1.103 (0.014) 7 93.272 (2.171) −1.063 (0.013) 8 24.897 (0.650) −1.019 (0.003) 9 * 271.455 (1.389) −0.993 (0.005) 10 332.548 (1.056) −0.972 (0.006) 11 * 259.730 (3.296) −0.952 (0.012) 12 227.678 (6.830) −0.926 (0.021) 13 334.308 (0.936) −0.896 (0.052) 14 21.801 (0.811) −0.881 (0.037) 15 284.039 (5.719) −0.818 (0.021) 16 245.613 (7.467) −0.755 (0.022) Patterns ordered by slope deviation from zero (to match Figure 2 ). Italicized patterns marked with an * match dotted patterns in Figure 2 . We can link the attractors back to the individual agents through the posterior probabilities. For purposes of relating to the initial assessment of the many unique patterns dying off, we chose to illustrate the average posterior probabilities (the average likelihood a given agent is depicted by a given attractor) as a function of time. Figure 2 shows the average posterior probabilities for each attractor dynamic. The legend shows the heading attracted to (set point) and level of attraction (slope) as a function of time. Consistent with Figure 1 (and expectations), initially there were many attractors, but somewhere around iteration 300, two specific attractors started to dominate (dotted lines in Figure 2 ). Figure 2 Time series of the average posterior probabilities for each attractor dynamic pattern . All patterns were attractors as indicated by their negative slopes, but differed in terms of their set points (SP; the heading in which the agents were attracted toward) and the attractor strengths (the steepness of their slopes). At around iteration 300, the system shows a phase transition wherein two patterns with the same heading begin to dominate. Notice that they share the same heading of 273 degrees, but with slightly different degrees of attraction. Recall the three rules that constitute the changes in heading over time: alignment, separation, and cohesion. Alignment and cohesion drive the agents toward a single heading, but separation instead evokes divergence when agents become too close (and specifically overrides the other two rules). What distinguishes the patterns is not the heading they are drawn toward, but in the divergences themselves due to separation that produces a weaker attractor. Note that agents can be switching between the two attractors over time moving to the slightly weaker attractor, as they need to avoid collisions. We gain additional information from the quantitative attractor dynamic description as illustrated in Figure 2 when compared to Figure 1 . Each data vector is now depicted not only in terms of its vector, but also the likely attractor in which it is drawn (through the posterior probabilities). Further, the description is now in terms of the underlying system forces that depict the type of pattern (all attractors since all the slopes were negative), the location to which the patterns are relative (the set points), and their stability under perturbations (the deviation of the slopes from zero). However, thinking topologically becomes even more beneficial as we move toward systems with more dimensions. Two dimensional systems Modeling a two dimensional system can be captured through two simultaneous equations.\n (2) ẋ t = b 0 + b 1 x t + b 2 y t + e x t \n (3) ẏ t = b 3 + b 4 x t + b 5 y t + e y t \nThese equations represent two variables measured simultaneously (x and y) at time t, x-dot and y-dot are their estimated derivatives at time t , b 0 and b 3 are intercepts, b 1 and b 5 are each variable predicting its own derivative, b 2 and b 4 are crossover or coupling relationships, and e xt and e yt are errors in equation. Notice that Equation (2) is identical to Equation (1) with the addition of the other changing variable also predicting velocity in x (or x predicting velocity in y). By having both variables changing simultaneously, we generate a two dimensional depiction. The emergent dynamic (attractor, repeller, etc.) is a function of all the b coefficients in the equations (Gottman et al., 2002 ). Common interpretation is that the own effects (i.e., x predicting change in x and y predicting change in y) depict the stability properties of the dynamic pattern (attractor, repeller, or saddle) such that negative coefficients are indicative of attractive behavior and positive coefficients are indicative of repulsive behavior in the respective dimensions. The crossover relationships (also known as coupling effects ) are commonly interpreted to represent the push-pull of variables that constitute cycles and swirling qualities graphically. The set point is a function of both equations. And as noted earlier, two-dimensional systems can include saddles and cycles, which are topological features that are not possible in one-dimensional systems. While many cases can be interpreted as described in the previous paragraph, some cases do not always conform to the conventional interpretations (and we include some examples of this below). A common violation relates to the notion of collinearity. If all variation in both x and y perfectly map onto one another, then x and y are essentially a single dimension. Under this circumstance the coefficients can be misrepresentative of the dynamic pattern. In our spatial movement circumstance, agents will sometimes capitalize on diagonal movement as a primary, singular dimension. Assessment of the eigenvalues and eigenvectors of the coefficients (treated as a Jacobian matrix of partial derivatives for estimating local Lyapunov exponents; Arabanel et al., 1992 ) is a method for verifying and determining whether to follow the classic interpretation or whether the interpretation should be modified. The ants model Consider the Ants model (Wilensky, 1997 ) in NetLogo v5.2.1 (Wilensky, 1999 ). This agent-based model was designed to simulate ant colony foraging behavior. The simulation consists of 125 ants each with the same instructions, starting at a nest in the center of a two-dimensional space. Ants are released one at a time from the nest, moving at a constant velocity. Three food sources are placed within the two-dimensional space each with a finite quantity of food supply. The ants search the environment for food (following a random direction algorithm) and upon locating and collecting food, return it to the nest. The primary mechanism for the emergent foraging behavior involves the ants releasing digital pheromones while carrying food and that the ants are attracted to this pheromone. This is much like how stigmergy, a form of environmental modification by individual social animals that affords collective coordination, is proposed to work in live ant populations (Theraulaz and Bonabeau, 1999 ). The nest also releases a pheromone signal so that the ants can find the nest. The simulation allows for the manipulation of the evaporation and diffusion rates of the pheromones, which we left at default settings. Figure 3 shows the standard placement of food sources in the environment in relation to the nest at the center. Figure 3 Screenshot of nest and food placement of the Ants model from Netlogo . From visual inspection, several emergent colony behaviors can be observed. Ants will search the environment until a critical threshold of ants find a given food source. At this point the ants will form a trail between the food source and the nest. There are sometimes congestion-like behaviors that occur in the middle of the trail or near the nest as more ants converge toward the strongest pheromone locales. Once the food source is used up, the ants once again spread out into a search pattern until a new food source is found. In this case, we will depict the attractor dynamics of the ant movement in two dimensions as a way to characterize the different ant behavioral patterns. We extracted the horizontal (x) and vertical (y) coordinate position of every ant from the beginning of the simulation until the last food pile was fully exhausted, totaling 1,080 iterations. Figure 4 is a kernel density plot of the ant positions, collapsed across all ants and all iterations. This shows the regions where ants spent most of their time and can be thought of as the probability density function of the data (under the assumption of two dimensions)—a graphical illustration of the integral of the dynamics. The density plot is read in the same fashion as a topographical map, where the lines illustrate more density. Note that the greatest density is at the nest (0,0). This was likely a function of all the ants starting at the nest, including the dispersion algorithm of only a single ant leaving the nest per iteration. It is also a function of all the ants returning to the nest to deliver food. Each branch of the density plot corresponds to one of the food sources, consistent with a trail between the nest and the food source. The densest part for each of the branches was, however, closer to the nest than the food source. Figure 4 Kernel density plot of where the ants spent most of their time during the simulation . Note that the highest densities correspond to the three food source locations and the nest. Figure 5 contains trails of three exemplar ants as vector plots in time to help illustrate the link between individual agents and the model estimated from all agents. Figure 5A shows the trail of an ant that helped collect food from all of the food piles. However, it also shows searching behavior in some of the areas of the world where food did not reside. Figure 5B illustrates an ant that only helped collect food from two piles and also participated in searching behavior in empty quadrants of the world. Figure 5C shows an ant that participated minimally in food collection instead spending more time searching. As a whole, these illustrate that the emergent behavior is not from any one ant. Instead, it is through their interactions with one another (through pheromones) and the environment (food resources relative to the nest) that their behavior becomes emergently coordinated. Figure 5 (A-C) Three example ant trails that illustrate how the ant behavior is shared across all the ants while each ant had unique behavior. Our mixture model identified a total of 7 different patterns in the example ant model (minimized BIC at 7 groups). Table 2 shows all the coefficients for the seven different patterns, labeled by their colors from Figure 6 . The last two columns are the eigenvalues wherein we built matrices of the own and coupling effects in the same order as Equations (2) and (3) (First row: own predicting x, coup predicting x and second row: coup predicting y, own predicting y). The eigenvalue procedure allows us to account for when the coefficients do not directly represent the type of attractor dynamic due to the primary axes for the dynamic depictions being different from the variables used in the equations. When the eigenvalues are both real numbers and negative, the system depicts an attractor. When the eigenvalues are both real and positive, the system depicts a repeller. When one is positive and one is negative, the system depicts a saddle. Imaginary numbers instead depict cyclic behavior with complex numbers being a combination of cyclic and attractive/repulsive at the same time (Abraham and Shaw, 1983 ). Table 2 Unstandardized coefficients (and standard errors) for the seven group solution along with eigenvalues for the Ants model . Pattern Own Coupling Intercept Eigenvalues Blue X −0.009 (0.002) 0.006 (0.002) −0.361 (0.064) −0.011 + 0.007i Y −0.013 (0.001) −0.009 (0.002) 0.096 (0.064) −0.011 − 0.007i Light Blue X −0.020 (0.003) 0.018 (0.004) −0.467 (0.075) −0.040 Y −0.023 (0.004) 0.020 (0.004) 0.522 (0.083) −0.002 Purple X −0.014 (0.003) 0.014 (0.004) −0.146 (0.035) −0.028 Y −0.001 (0.005) 0.027 (0.004) 0.107 (0.049) 0.013 Yellow X 0.005 (0.001) 0.018 (0.003) −0.645 (0.071) −0.011 Y −0.017 (0.003) −0.005 (0.001) 0.605 (0.085) −0.000 Green X −0.003 (0.001) 0.002 (0.001) −0.049 (0.014) −0.001 Y 0.003 (0.001) −0.004 (0.001) 0.002 (0.014) 0.001 Brown X 0.000 (0.001) 0.001 (0.000) 0.022 (0.019) 0+0.002i Y 0.000 (0.000) −0.003 (0.001) 0.046 (0.017) 0−0.002i Red X 0.000 (0.001) 0.001 (0.002) 0.011 (0.004) 0.019 Y 0.019 (0.002) −0.006 (0.001) 0.036 (0.004) 0.000 Figure 6 Topographical illustration of the seven equation solution for the Ant simulation . Figure 6 is a topographical representation of the seven attractor dynamics patterns emergent in the ant behavior. Figure 6 was generated by using the estimated equations from the mixture model in conjunction with the adaptive Runge-Kutta algorithm from the deSolve package (Soetaert et al., 2010 ) in R (R Core Team, 2016 ) to estimate example trajectories iterated over time. In each case, values were chosen using the position means and variances extrapolating in all possible combinations of one standard deviation in X and Y and iterating the trajectories forward in time. Details on each pattern follow. The blue, brown, and green patterns correspond to the food piles while the red pattern corresponds to the ant nest. The yellow pattern corresponds to searching an area where no food existed. The light blue captured the pattern of the ants converging in the middle of the trail as the pheromones were most intense there and the purple captured the dispersal after the food pile in the upper left had been fully collected (it was the first pile found in the simulation). Notice how each pattern is captured through a different attractor dynamic. For example, the red nest pattern shows a repeller in which ants leave the location. If we capture each ant trail of food collection through the other patterns, then what primarily remains is the initial leaving from the nest. The blue and brown patterns, both corresponding to food piles, show cyclic properties (they have imaginary components to their eigenvalues). This is capturing the pattern of getting the food from the pile, bringing it to the nest and returning. The pattern corresponding to the lower left food pile was a saddle, however—attractive in one dimension and repulsive in the other. By having the set point far from the dynamic pattern, the saddle generated curved trails that could then be completed by feeding into other, already established, patterns. Now, we link the agents to these patterns and to key system descriptions—in this case food depletion. Figure 7 shows the decline in the food piles as a function of time. Notably, the ants found the pile in the upper left first, followed by the lower left and then finally the middle right. We ran seven multilevel models treating the posterior probabilities of each pattern as the outcome as a function of the proportion of food remaining in each pile (a three predictor MLM). The fixed and random effects along with intraclass correlations (ICC) are in Table 3 . All random effects were significantly non-zero suggesting that there was variability in their likely pattern as a function of the remaining food piles among the individual ants. The fixed effects can be interpreted as whether or not the likelihood of being in a pattern occurred where a positive sign meant that declines in a food pile corresponded to declines in the pattern and a negative sign meaning that declines in a food pile corresponded to increases in the pattern. Given the order of the food pile depletions, the pattern of effects can also roughly determine when the pattern was more prevalent. Figure 7 Time series plot of the amount of food available in each of the three food piles . The legend describes where in the coordinate space a given food pile was located (see also Figure 3 ). Table 3 Unstandardized coefficients (and standard errors) and intraclass correlations from multilevel models predicting the posterior probabilities of being in each of the seven groups for the Ants models . Upper left food pile Lower left food pile Middle right food pile Intercept ICC FIXED EFFECT Blue 0.014 (0.009) −0.018 (0.010) −0.248 (0.020) * 0.269 (0.016) * 0.046 Light Blue −0.008 (0.006) 0.008 (0.008) −0.239 (0.013) * 0.249 (0.011) * 0.031 Purple −0.019 (0.005) * −0.010 (0.009) −0.221 (0.015) * 0.259 (0.012) * 0.021 Yellow −0.020 (0.011) −0.018 (0.010) 0.034 (0.011) * 0.028 (0.007) * 0.185 Green −0.092 (0.024) * −0.304 (0.023) * 0.479 (0.029) * 0.025 (0.017) * 0.070 Tan −0.219 (0.025) * 0.187 (0.025) * 0.072 (0.027) * 0.094 (0.018) * 0.116 Red 0.345 (0.025) * 0.154 (0.011) * 0.123 (0.012) * 0.076 (0.006) * 0.020 VARIANCE COMPONENT Blue 0.009 (0.001) * 0.012 (0.002) * 0.049 (0.006) * 0.033 (0.004) * Light Blue 0.004 (0.0010 * 0.008 (0.001) * 0.021 (0.003 * 0.014 (0.002) * Purple 0.003 (0.000) * 0.009 (0.001) * 0.029 (0.004) * 0.017 (0.002) * Yellow 0.014 (0.002) * 0.012 (0.001) * 0.016 (0.002) * 0.006 (0.001) * Green 0.072 (0.009) * 0.063 (0.008) * 0.107 (0.014) * 0.037 (0.005) * Tan 0.080 (0.010) * 0.079 (0.010) * 0.087 (0.011) * 0.039 (0.004) * Red 0.079 (0.010) * 0.014 (0.002) * 0.017 (0.002) * 0.004 (0.001) * * Denotes p < 0.05 . The red pattern at the nest was likely when all the food sources were untouched and declined in likelihood as all the food piles declined, consistent with the ants initially leaving the nest to search. The blue, light blue, and purple patterns all associated with the upper left quadrant were all less likely when the last food pile was untouched, but only the purple (the theoretical dispersion after the food pile was depleted) was contingent upon the corresponding upper left food pile. The negative sign was indicative that declines in the first pile increased the likelihood of the purple dispersion pattern consistent with leaving the trail to find another food source once the food in the first pile was depleted. The green (corresponding to the lower left food pile) and brown (corresponding to the middle right food pile) patterns were predicted by all three food piles with negative coefficients suggesting that as any food depleted, these became more likely—consistent with these food piles being found later. Finally, the yellow pattern was only uniquely predicted by the middle right food pile depletion such that as the food pile declined, so did the likelihood of being in the search pattern. Given that as more ants found the last food pile, more converged on it. Once it depletes, however, fewer ants would be in this search pattern. Baboons navigation data So far, we have relied on simulations to illustrate how one can depict higher order emergent coordination for agent interactions using attractor dynamics. Our next two examples are derived from observed data. Figure 8 represents a solution from global positioning system (GPS) data collected from a troop of baboons at the De Hoop Nature Reserve in South Africa. Table 4 contains the coefficients and eigenvalues, again using colors to indicate correspondence. To collect this data, researchers recorded the positions of 14 adult baboons by holding a GPS device over or very close to each animal at different points over a 74 day period (data was made available by Bonnell et al., 2016 ; and further details of the original study can be found at Bonnell et al., 2017 ). Consistent, with the ants data, this example data is in an x/y coordinate space, but now in longitude and latitude. To facilitate estimation due to variability occurring in small decimal places, longitude and latitude were mean-centered and multiplied by 1,000. Figure 8 Topographical solution for the Baboon gps data . Table 4 Unstandardized coefficients (standard errors) and eigenvalues for the 10 pattern solution from the Baboon GPS data . Pattern Own Coupling Eigenvalues Red X −0.821 (0.015) −0.280 (0.013) −0.554 + 0.441i Y −0.314 (0.023) 0.689 (0.023) −0.554 − 0.441i Orange X −0.687 (0.021) 0.084 (0.017) −1.713 Y −0.860 (−0.14) −0.140 (0.039) −0.828 Yellow X −0.025 (0.002) −0.007 (0.001) −0.178 Y −0.005 (0.001) −0.019 (0.002) −0.004 Light Green X −3.720 (0.637) −0.906 (0.182) −0.006 + 0.017i Y 1.595 (0.455) 3.798 (1.660) −0.006 − 0.017i Dark Green X −0.024 (0.003) 0.005 (0.001) −0.424 + 0.311i Y −0.011 (0.001) −0.062 (0.003) −0.424 − 0.311i Light Blue X −0.386 (0.015) 0.164 (0.015) −1.205 + 0.350i Y −1.073 (0.036) −1.103 (0.032) −1.205 − 0.350i Blue X −1.066 (0.019) −0.039 (0.017) −0.857 + 0.239i Y −1.040 (0.040) 0.807 (0.047) −0.857 − 0.239i Navy Blue X −0.011 (0.005) 0.002 (0.002) −0.026 + 0.011i Y −0.035 (0.003) −0.109 (0.009) −0.026 − 0.011i Purple X −0.008 (0.030) 0.014 (0.014) −0.028 + 0.017i Y −0.035 (0.019) −0.064 (0.039) −0.028 − 0.017i Magenta X −0.013 (0.002) −0.020 (0.003) −0.026 Y −0.014 (0.003) −0.007 (0.002) −0.002 Figure 8 illustrates that several of the patterns show cyclic behaviors. In fact, all the eigenvalues were negative with 7 of the 10 showing imaginary eigenvalues consistent with cyclic behaviors. Further, all patterns had at least one negative real eigenvalue suggesting that they all were attractive indicating a pattern of convergence for baboons. Figure 8 clearly shows that the patterns were not equally attractive, however, in that vector length differed dramatically when example trajectories were estimated. This can also be seen by the size of the eigenvalues where some were quite close to zero in their real number portion(s) while others were much smaller numbers approaching and surpassing negative one. Thus, some of these patterns were more stable clusters for the baboons while others were more loose associations around the shared longitude/latitude set point. In their original work, Bonnell et al. ( 2017 ) evaluated whether the movement patterns of a focal individual baboon was influenced by the location of the troop as a collective or by the locations of specific influential members of the troop. Ultimately, their results showed evidence for both of these patterns. In some cases, the focal baboon's movement was highly influenced by the average movement location of the entire troop. In other cases, the focal baboon's movement was quite sensitive to the movements of the alpha female (F1) and the alpha male (M1). To link back to individual baboons, our results suggest a consistent pattern as illustrated in Figure 9 wherein we show the average posterior probabilities for each baboon illustrating which pattern would arguably influence a given baboon the majority of the time (again, colors correspond). Few distinctions existed between the female and dominant male baboons showing preference for the light green (cyclic attractor) and yellow (attractor) patterns. The dominant male (M1) showed slightly more preference for the magenta pattern (also an attractor). Thus, there is evidence of following the primary male baboon, but also one of a female majority. And yet in both cases these most common patterns represent the least attractive patterns (eigenvalues closest to zero) in that there is lots of wandering in comparison to the other patterns inferred from the GPS data. Figure 9 Average posterior probabilities associated with each equation group, by baboon . F is for female and M is for male. Beyond two dimensions As we move beyond two dimensions, it is difficult to make easy to read and meaningful maps of the data. However, our approach is not limited to two dimensions. By relying on the eigenvalues presented earlier, one can derive the higher order patterns to illustrate what is occurring without a means to draw them. Further, it also allows us to point out that any variables can be captured as attractor dynamics—they do not inherently need to be spatial, as illustrated by our next example. Each new dimension corresponds to an additional equation. In the six-dimensional case that follows, we model six simultaneous equations where change in each variable is treated as the outcome from each equation. Further each variable at a given point in time is allowed to freely predict the changes in each equation. The matrix used to generate the eigenvalues is based on the coefficients where, as before, the main diagonal are the own effects and the off diagonals are the coupling relationships. Each matrix row corresponds to a different equation. Affect in families data To show a non-spatial example with more than 2 simultaneous change equations, we modeled positive and negative affect from the PANAS (Watson et al., 1988 ) taken from mothers, fathers, and adolescents from 252 families where the adolescent has type 1 diabetes. The data are taken from the Adolescents with Diabetes and Parents Together study where each family member completed a daily diary for 14 days (further study details can be found at Berg et al., 2009 ). We extracted two stable patterns (three patterns would not properly converge and fit indices supported the two pattern solution). Table 5 provides the estimated coefficients. Notably, the eigenvalues were quite different between the two patterns. The first pattern generates all negative eigenvalues indicating that it forms one large six dimensional attractor (−0.709, −0.522, −0.455, −0.428, −0.285, −0.257). The second pattern, on the other hand had complex numbers for the first two eigenvalues suggesting cyclic behavior as a primary component (−0.613+0.027i, −0.613 −0.027i, −0.420, −0.343, −0.180, −0.157). Table 5 Unstandardized coefficients (and standard errors) from the two pattern solution for the Affect Daily Diary . Mother + Mother − Father + Father − Adolescent + Adolescent − 1 ΔM+ −0.39 (0.03) 0.09 (0.03) 0.07 (0.03) 0.08 (0.04) 0.01 (0.02) 0.02 (0.02) ΔM− 0.06 (0.02) −0.55 (0.03) 0.023 (0.03) −0.11 (0.04) 0.04 (0.02) −0.04 (0.02) ΔF+ 0.09 (0.03) 0.05 (0.03) −0.41 (0.03) 0.09 (0.04) −0.01 (0.02) 0.02 (0.03) ΔF− −0.00 (0.02) −0.05 (0.02) 0.08 (0.02) −0.55 (0.04) 0.02 (0.02) −0.03 (0.02) ΔA+ 0.01 (0.04) 0.09 (0.04) 0.02 (0.04) 0.09 (0.06) −0.41 (0.03) 0.12 (0.03) ΔA− 0.00 (0.03) −0.09 (0.03) 0.01 (0.03) −0.08 (0.02) 0.09 (0.02) −0.34 (0.03) 2 ΔM+ −0.33 (0.03) 0.28 (0.12) 0.07 (0.03) −0.19 (0.21) 0.02 (0.02) 0.03 (0.08) ΔM− −0.00 (0.01) −0.61 (0.04) −0.02 (0.01) −0.04 (0.07) 0.01 (0.01) 0.04 (0.02) ΔF+ 0.04 (0.03) −0.03 (0.09) −0.19 (0.04) −0.32 (.19) 0.02 (0.02) 0.02 (0.07) ΔF− −0.01 (0.01) 0.01 (0.02) 0.00 (0.01) −0.50 (0.07) 0.00 (0.00) −0.01 (0.01) ΔA+ 0.06 (0.03) 0.36 (.12) 0.01 (0.03) 0.35 (.24) −0.20 (0.02) 0.19 (0.09) ΔA− 0.00 (0.01) 0.00 (0.05) −0.01 (0.01) −0.12 (0.08) 0.02 (0.01) −0.41 (0.06) In the form of matrices used to estimate eigenvalues. Table was rounded to second decimal for space. Rows are changes (Δ) in Mother (M), Father (F), and Adolescent (A) . Though we cannot draw a map to represent this higher order pattern, one way to represent the changes in the system is through a network diagram. Figures 10A,B shows only significant (alpha = 0.05, two-tailed) pathways between affect variables. The beginning of an arrow is value and the end of an arrow is change. Blue arrows represent negative relationships and brown ones are positive. Note between Figures 10A,B the connections between individuals breaks down substantially with the cyclic nature relating to the less connected network. The most noteworthy is the changing connections of father's affect to the mother and adolescent. It is noteworthy that these coefficients merely indicate prediction and thus any interpretation of causality would overstate the relationship. That said, fathers were clearly showing less connection in the second pattern. Figure 10 (A,B) Two network diagrams that illustrate the two different equations. Beginning of arrows represent value at time t. Arrow heads represent change in value. To link back to individual families, we built a multilevel model predicting the posterior probabilities for the first pattern as a function of diabetes risk for the adolescent on a given day. We use the variable risk as an easy to interpret indicator as to how well the adolescent was managing their diabetes on a given day. Risk is a rescaled version of daily blood glucose variability and level such that zero indicates perfect maintenance at doctor recommended levels and 100 indicates either going too high or too low repeatedly (both of which can be quite dangerous; see Kovatchev et al., 2006 ). Since posterior probabilities for a given data vector add to one, high probabilities of being in the first pattern inherently implies a low probability of being in the second. Table 6 contains the coefficients. At zero risk on a given day, families were equally likely to be in each pattern (the intercept is the posterior probability when risk was zero). As risk increased, however, families were more likely to fall into the second pattern. That is, on good days we see the more connected attractor pattern and on bad days the father appears less connected and the family affect adopts a cyclic attraction pattern instead. Table 6 Unstandardized coefficients (and standard errors) from multilevel model predicting the posterior probability of the first pattern as a function of Diabetes risk . Fixed coefficients Variance components Intercept 0.502 (0.027) * 0.100 (0.012) * Risk −0.003 (0.001) * 0.000 (0.000) * Denotes p < 0.05 .",
"discussion": "Discussion Kelso ( 2009 ) posited that SCD “unites the spontaneous, self-organizing nature of coordination and the obviously directed, agent-like properties characteristic of animate nature into a single framework” (p. 1540). This logic matches with self-organization from agent-based models, and cases where many agents engage in social coordination, more generally. By connecting attractor dynamics modeling with cases where there are a range of agents and a range of outcomes allows for a generalized approach to quantifying the emergent patterns. Through various examples, we illustrated that the attractor dynamics can be captured using a combination of difference/differential equation modeling and mixture modeling. Further, we showed that these attractor patterns and their occurrence could be linked with different outcomes. For the flocking model, we found sixteen attractor patterns of the agents' heading that converged on fewer attractors over time. For the ants model, we found seven dynamic patterns to depict their motion in a two-dimensional x/y space that roughly corresponded to qualitative depictions of rules the ants follow. For the baboon navigation data, we found ten patterns in two-dimensional longitudinal and latitudinal space in which the probability of exhibiting a particular attractor was contingent upon influential baboons in the troop (e.g., an alpha male). For families where an adolescent has type 1 diabetes, we found two patterns in a six dimensional affect space that corresponded to higher and lower levels of risk from the disease. By using the data from all the agents, the underlying topology is inclusive of all the agents. In the ants model, for example, not all ants illustrated being influenced by every pattern. Instead, ants can exist in a single pattern their entire time or move between them. Thus, the underlying map implied by the set of dynamic patterns generates an inclusive generalization both within and between agents that capitalizes on the most probable systems states over the duration of the observation period. In each circumstance, the technique depicts the topological feature in terms of the implied patterns and the stability of those patterns. Whereas, the flocking model only contained attractors that varied in their set points (attractive headings) and their stabilities, the ants model illustrated all the common possible attractor dynamic patterns including attractors, repellers, saddles, and cycles. The complexity of the underlying pattern is directly related to the number of dimensions. With a single dimension, attractor dynamics may only convey attractors and/or repellers. With two dimensions, cycles and saddles can be inferred. Beyond two dimensions, chaotic (strange) attractors are possible, though all currently known chaotic attractors require non-linear equation forms and the equations herein were restricted to linearity within each equation group. Thus, this is a limitation of the technique provided. In each case, we then linked the quantification back to the individual agents. Through mixture modeling we did this by outputting the posterior probabilities. These probabilities are the probability that a given data vector is under the influence of a given dynamic pattern, the probabilities for a given vector sum to one across all the possible patterns. Therefore, these probabilities maintain the data dependency we inherently ignored in the estimation for the dynamic patterns themselves. We therefore always either examined the probabilities at a collapsed agent level (e.g., averages) or through multilevel modeling wherein the dependencies could be properly taken into account. In each case, it could be linked to possible variables of interest used to depict the system. For the headings, this was illustrated with time in that attractors should collapse as time goes on. For the ants model, this was illustrated through food supply. For the baboons, this was illustrated through the location of the alpha male and the females. For affect in families where the adolescent had type 1 diabetes, it was illustrated with the diabetes risk exhibited that day. In all, this allows one to link the higher order patterns back to meaningful outcomes that characterize when agents behave in certain ways or exhibit theoretically important states. In the spatial examples, we utilized variables that depicted the spatial movement. As an initial foray into understanding attractor dynamics, thinking spatially helps make the concepts more intuitive. But, ultimately, these concepts can be applied in many contexts where relationships are not inherently spatial. Being able to think about the spatial analogs helps ground what is being observed, but does not inherently limit the domains in which attractor dynamics can be examined. Further, individual equation parameters do not always align with the system depiction graphically or through the eigenvalue procedure. In the ants example, this had to do with the reliance on diagonal movement of the ants. By depicting the system through an equation of x and an equation of y, we mask diagonal movement—it is really a straightforward combination of the two dimensions rather than showing some independence. More generally, the coefficients are under an assumption that the dimensions chosen are the primary dimensions for depicting the changes occurring in the system. The eigenvalue procedure bypasses this assumption by instead capitalizing on dimensions that maximize the strength of the attractor dynamics. Once we moved beyond two dimensions, the eigenvalue procedure becomes even more valuable. There is no easy way to graphically “see” the implied dynamic, but the sign and distinctions between real and imaginary portions elucidate the attractor pattern. In practice, anytime we model a system with two or more equations we should adopt the eigenvalue procedure as a means to understand the higher order pattern in addition to any interpretations applied to the individual coefficients themselves. For example, it is common to interpret coupling coefficients as the push/pull of one variable upon another. However, this fails to capture what pattern the push-pull creates as their interpretation is under an assumption that we somehow picked ideal dimensions to represent them. Locally, the coefficients maintain their meaning, but we cannot extrapolate the more global pattern of which they are a part. In regards to equation identification, the technique is not without its limitations. The choice of slicing up the data into a series of locally linear equations is an imperfect method for capturing non-linear dynamic models. Specifically, non-linear dynamic models can have both multistability in which more than one pattern is stable simultaneously and cases where variables differentiate when one pattern is or is not accessible. By slicing up the data into a series of locally linear equations through mixture modeling, these two circumstances are difficult to distinguish. One can begin to distinguish these circumstances by attempting to predict the posterior probabilities. However, ultimately multistability is distinguished by states being probable despite nothing differentiating them (or when the dimensions being examined are all that differentiate them). That is, multistability would occur under a lack of being able to predict differences of when agents would be in one or the other. Thus, this approach provides a limited potential for knowing when multistability exists as opposed to having some variable differentiate them. We may never examine the “right” variable or are instead in the situation of arguing a null finding to support the multistable case. In contrast, it is possible through a cusp catastrophe model in conjunction with multilevel modeling, for example, to allow for differentiating variables (also known as control parameters) without their identification (Butner et al., 2014b ), though knowing which scenario you are observing requires examination of many more qualities than discussed herein (Gilmore, 1981 ). Further, manifolds (the surfaces implied by topological equations) are smooth, while the mixture modeling approach is more patchwork. We do not know the reach of a given attractor dynamic—we chose to represent each dynamic through one standard deviation in each direction from the means when we utilized the Runge-Kutta algorithm to graph plausible trajectories. Notably the means and standard deviations are specific to each dynamic pattern (allowing some to be large and others to be smaller). However, the boundaries of one pattern to another are truly unknown, requiring some inference. Notably, SCD has tended to rely on cyclical descriptions to model the rhythmic coordination of social agents. While the modeling approach illustrated herein allows for cycles, it does not assume their existence. The direct equation link is that SCD generally functions with second order equations where the second derivatives (acceleration or change in velocity) are treated as the outcomes. Within our structural equation model, it would be analogous to building a quadratic growth model on Toeplitz data where the quadratic growth latent variable would be the second derivative predicted by the other two latent variables (Butner and Story, 2010 ). Moving to a second order model automatically implies two dimensions and thus generates cycles. However, it is not without a cost. Specifically, second order modeling in this form assumes that the set point of the cycles must equal zero. Overcoming this assumption is currently something under consideration for modeling dynamic patterns and once resolved will unite these approaches more generally."
} | 16,027 |
37260678 | PMC10227517 | pmc | 2,738 | {
"abstract": "Different crop genotypes showed different adaptability to salt stress, which is partly attributable to the microorganisms in the rhizosphere. Yet, knowledge about how fungal communities of different genotypes in soybean respond to salt stress is limited. Here, qPCR and ITS sequencing were used to assess the response of rhizobial fungal communities of resistant and susceptible soybean to salt stress. Moreover, we isolated two fungal species recruited by resistant soybeans for validation. The assembly of fungal community structure might be strongly linked to alterations in fungal abundance and soil physicochemical properties. Salt stress derived structural differences in fungal communities of resistant and susceptible genotypes. The salt-resistant genotype appeared to recruit some fungal taxa to the rhizosphere to help mitigating salt stress. An increase of fungal taxa with predicted saprotrophic lifestyles might help promoting plant growth by increasing nutrient availability to the plants. Compared with the susceptible genotypes, the resistant genotypes had more stronger network structure of fungi. Lastly, we verified that recruited fungi, such as Penicillium and Aspergillus, can soybean adapt to salt stress. This study provided a promising approach for rhizospheric fungal community to enhance salt tolerance of soybean from the perspective of microbiology and ecology.",
"introduction": "Introduction It is well known that soil salinization is a considerable problem in agricultural system, and that soil salinity can greatly reduce plant productivity and yield value ( Khasanov et al., 2023 ). Due to an increase in global population and the ever-increasing demand for food quality, the issue of how to alleviate the pressure of soil salinity, improve plant resistance to salt stress and eventually increase crop yields is an urgent need to be addressed. Soybean [ Glycine max (Linn.) Merr.], an important source of protein and oil in the world, is very sensitive to salt stress, which can severely restrict nutrient use and growth and development, ultimately reducing yields ( Phang et al., 2008 ). In the last few years, traditional breeding techniques combined with beneficial microorganisms have been widely used to improve the salt resistance of soybeans ( Pathan et al., 2007 ; Hanin et al., 2016 ). Different soybean varieties have different root exudates, which determines the composition of the plant-specific root and microbial communities in rhizosphere area ( Bulgarelli et al., 2013 ; Lian et al., 2019a ). Under salt stress, the amount and type of root exudates secreted by different species are different ( Lian et al., 2020 ). It has been demonstrated that salt-resistant soybeans have a much greater salicin, arbutin 6-phosphate, phosphoglycolate, and 1-methlseleno-N-acetyl-dgalactosamine than salt-susceptible soybeans in soils, which may increase the salt adaptation of soybeans ( Lian et al., 2020 ). Microorganisms have the benefit of promoting health and increasing productivity in plants ( Mendes et al., 2013 ; Li et al., 2014a , b ). Different types and amounts of metabolites from plants or microorganisms could alter the diversity and structure of rhizosphere microbes, which could assist the host to become more resistant to stress ( Wu et al., 2006 ; Qin et al., 2016 ; Hu et al., 2018 ; Lian et al., 2020 ). It is well known that plant growth promoting rhizobacteria (PGPB) have certain functions that can promote plant growth ( Bhatt et al., 2022 ). For example, a variety of metabolites produced by Pseudomonas can lead to salt stress-relieving, including exopolysaccharides, ACC deaminase and hormones (indoleacetic acid and gibberellins; Etesami and Glick, 2020 ; Li et al., 2021 ). However, studies in recent years have largely emphasized on bacteria, neglecting fungal species, with the improvement in nutrient cycling and the resistance to environment, which can also assist plants to mitigate damage caused by abiotic stresses ( Kawai et al., 2000 ; Peltoniemi et al., 2012 ). Penicillium and Aspergillus , which were reported to increase nitrogen and phosphorus to plant roots, stimulate the growth of host plants by increasing the accumulation of nutrient under unfavorable conditions ( Kiers et al., 2011 ), and thus might help plant alleviate the biotic and abiotic stresses. Thus, to understand how salt-resistant soybean better adapt to salt stress, it is necessary to investigate how rhizosphere microbes of salt-resistant soybean genotypes respond to salt stress. In this study, we selected the resistant soybean (Qinong7) or susceptible soybean (Hefeng50), growing at soils under salt and non-salt stress. Then, we analyzed the fungal community structure in rhizosphere through ITS high-throughput sequencing. Moreover, the fungal community structure was investigated in relation to its physicochemical properties. We hypothesized that (1) Salt-R genotype possesses higher fungal diversity compared to Salt-S genotype, and (2) Salt-R genotype will enrich particular Salt-R fungal taxa to the rhizosphere that help mitigating salt stress.",
"discussion": "Discussion This study was conducted to reveal how salt stress affects the structure of the rhizospheric fungal community of salt-tolerant (Salt-R) and susceptible (Salt-S) soybean genotypes. In comparison to the Salt-S genotype, the fungal communities of the Salt-R genotype were higher in abundance and significantly different in structure, but not in diversity ( Figures 1 – 3 ). The amount and type of root exudates secreted by different soybean genotypes was variable, which led to a diverse response to salt stress ( Li et al., 2021 ). Therefore, Salt-R genotypes, except for secreting abundant organic acids directly to dilute NaCl, might also enrich some fungi with ability to secrete organic acids around the rhizosphere, thus increasing their resistance to salinity ( Lian et al., 2020 ). Notably, salt stress usually amplified the segregating trend in fungal community structure between resistant and susceptible genotypes, which was in accordance with the observations of Lian et al. (2020) and Lian et al. (2020) . The composition of the rhizospheric fungal community was greatly altered by salt stress in both genotypes ( Figure 3A ). This was in line with previous studies that fungal community structure was influenced by the complicated effects of saline alkaline soil environments ( Yao et al., 2021 ). There was also, however, genotype-dependance in the fungal community structure with non-salt stress ( Figure 3A ), which was in contrast with previous study ( Wang et al., 2008 ). Wang et al. (2008) reported that fungal communities were not found to be significantly different among the three genotypes at the same growth stage ( Wang et al., 2008 ). This phenomenon might be explained by the lower methodological resolution to test fungal communities ( Gomes et al., 2003 ). It was possible that differences in soil physicochemical properties directly contributed to changes in the structure of fungal communities under salt stress ( Figure 5 ). Moreover, Na + , Olsen-P, NH 4 + , NO 3 − and pH were the most important factors that shaping the rhizosphere microbial community. A number of studies have demonstrated that salt resistant genotypes had the ability of secreting some special root exudates to make plants more adapted to salt stress ( Innes et al., 2004 ; Lian et al., 2020 ). We have previously shown that the resistant genotype can recruit beneficial bacteria and hypothesize that the same is true for fungi ( Lian et al., 2019a ). The relative abundance of several fungal taxa in the rhizosphere of Salt-R was higher compared to Salt-S under salt stress like Talaromyces , Saitozyma and Cladosporium . Moreover, Talaromyces and Cladosporium isolated from the rhizosphere soil were verified that significantly increased the shoot and root biomass of soybean ( Table 4 ). It has been previously revealed that Talaromyces was able to solubilize phosphate at salinity and thus showed high tolerance to salt stress ( López et al., 2020 ). Thus, Talaromyces may be a key species for improving salt tolerance in soybean. However, the other five genera have not yet been found to be associated with soil salt stress and their role needs to be investigated in more detail. The two genotypes also showed differences in the relative abundance of specific trophic groups ( Table 1 ). Saprotrophs were the dominant trophic mode in the present study. Saprotrophs, as the dominant guild, had the highest relative abundance in the Salt-R genotype under salt stress. It has reported that fungi belonging to saprotrophs might have an essential function in promoting nutrient conversion and controlling plant pathogens ( Lian et al., 2019b ). The increased abundance of saprotrophs is again directly linked to the presence of different Talaromyces species. Previous studies have well established that Talaromyces could facilitate plant growth through better utilization of nutrients by plants ( Shi et al., 2022 ). Co-occurrence network analysis revealed that network properties were inherently different among salt-resistant and susceptible genotypes ( Figure 6 ; Table 2 ). Compared to Salt-R genotypes, there was fewer negative correlations and higher modularity in fungal networks of Salt-S genotypes under salt condition, according to network theory, probably because of weaker competitive relationships between microbial species within the rhizosphere ( Saavedra et al., 2011 ; Fan et al., 2018 ). Additionally, Salt-R genotype exhibited a higher number of positive correlations than Salt-S genotype under salt stress, suggesting that most fungal members were connected through a series of cooperative relationships ( Coyte et al., 2015 ; de Vries et al., 2018 ). However, this network structure was considered unstable because fungal members might be strongly influenced by environmental fluctuations, thus increasing unstable coupling ( Coyte et al., 2015 ; de Vries et al., 2018 ). In addition, core species served as a critical pointcut to analyze how to alleviate salt stress ( Table 3 ). For example, ASV23, ASV8, and ASV14 were identified as Aspergillus , which alleviated salt stress by producing organic acids to form organic acid-salt complexes ( Ali et al., 2021 ). However, rhizosphere microbes also include bacteria, which can help soybeans resist salt stress by releasing hormones and promoting plant nutrient uptake, among other things, which cannot be ignored ( Li et al., 2021 ). Bacteria should be explored in future studies and analyses in conjunction with fungi to explore the synergistic role of different microbial communities in helping the host to resist stress. In conclusion, the rhizospheric fungal community of the two genotypes differed under salt stress. The Salt-R genotype recruited salt-resistant fungal species to the root zone to help alleviating salt stress. Different co-occurrence structure of the fungal community associated with the resistant genotype indicate more complex along with environmental changes. Taken together, the study provides new evidence for the important role of the soybean rhizosphere microbiome in conveying resistance to salt stress. In the future, the rhizospheric fungal community could serve as a promising breeding strategy to select for plants that are more resistant towards different stresses."
} | 2,854 |
28295875 | null | s2 | 2,741 | {
"abstract": "This communication describes a simple and effective method for welding electrospun nanofibers at the cross points to enhance the mechanical properties of their nonwoven mats. The welding is achieved by placing a nonwoven mat of the nanofibers in a capped vial with the vapor of a proper solvent. For polycaprolactone (PCL) nanofibers, the solvent is dichloromethane (DCM). The welding can be managed in a controllable fashion by simply varying the partial pressure of DCM and/or the exposure time. Relative to the pristine nanofiber mat, the mechanical strength of the welded PCL nanofiber mat can be increased by as much as 200%. Meanwhile, such a treatment does not cause any major structural changes, including morphology, fiber diameter, and pore size. This study provides a generic method for improving the mechanical properties of nonwoven nanofiber mats, holding great potential in various applications."
} | 227 |
34034081 | null | s2 | 2,742 | {
"abstract": "Microbial interactions are increasingly recognized as an integral part of microbial physiology. Cell-cell communication mediated by quorum sensing and metabolite exchange is a formative element of microbial interactions. However, loss-of-function mutations in quorum-sensing components are common across diverse species. Furthermore, quorum sensing is modulated by small molecules and environmental conditions that may be altered in the presence of other microbial species. Recent evidence highlights how strain heterogeneity impacts microbial interactions. There is great potential for microbial interactions to act as selective pressures that influence the emergence of common mutations in quorum-sensing genes across the bacterial and fungal domains."
} | 188 |
32070253 | PMC7062023 | pmc | 2,743 | {
"abstract": "Rapid and unprecedented ecological change threatens the functioning and stability of ecosystems. On coral reefs, global climate change and local stressors are reducing and reorganizing habitat-forming corals and associated species, with largely unknown implications for critical ecosystem functions such as herbivory. Herbivory mediates coral–algal competition, thereby facilitating ecosystem recovery following disturbance such as coral bleaching events or large storms. However, relationships between coral species composition, the distribution of herbivorous fishes and the delivery of their functional impact are not well understood. Here, we investigate how herbivorous fish assemblages and delivery of two distinct herbivory processes, grazing and browsing, differ among three taxonomically distinct, replicated coral habitats. While grazing on algal turf assemblages was insensitive to different coral configurations, browsing on the macroalga Laurencia cf. obtusa varied considerably among habitats, suggesting that different mechanisms may shape these processes. Variation in browsing among habitats was best predicted by the composition and structural complexity of benthic assemblages (in particular the cover and composition of corals, but not macroalgal cover), and was poorly reflected by visual estimates of browser biomass. Surprisingly, the lowest browsing rates were recorded in the most structurally complex habitat, with the greatest cover of coral (branching Porites habitat). While the mechanism for the variation in browsing is not clear, it may be related to scale-dependent effects of habitat structure on visual occlusion inhibiting foraging activity by browsing fishes, or the relative availability of alternate dietary resources. Our results suggest that maintained functionality may vary among distinct and emerging coral reef configurations due to ecological interactions between reef fishes and their environment determining habitat selection.",
"introduction": "1. Introduction Global climate change and mounting local stressors are degrading ecosystems via species extirpations and introductions, modifying the composition of assemblages and threatening ecological function [ 1 , 2 ]. Non-random species turnover, ordered by the susceptibility of organism traits [ 3 ], is increasing the taxonomic and functional similarity of communities [ 4 – 6 ]. These changes can disrupt ecosystem processes, such as habitat provisioning [ 7 , 8 ], primary productivity [ 9 ], trophic energy flow [ 10 ], nutrient cycling [ 11 , 12 ] and pollination [ 13 ]. While evidence exposes a coherent pattern of ecological change across biomes [ 14 ], variation exists from the individual to community level in how ecological structure, ecosystem processes and ongoing disturbance dynamics interact [ 15 , 16 ]. For effective and adaptive local management, better understanding is needed of the extent to which different, and in some cases emerging, species configurations support processes critical to ecological stability [ 17 ]. We focus on coral reefs, one of the most biodiverse but threatened ecosystems [ 18 ], to elucidate how the composition of habitat-building species (i.e. corals) influences key ecosystem functions. Climatic changes and local human impacts have reduced populations of corals, resulting in unprecedented loss of coral cover and marked shifts in coral species composition due to differential susceptibilities of corals to thermal stress, severe storms, predation by crown-of-thorns starfish and poor water quality [ 19 , 20 ]. Typically stress-sensitive, topographically complex branching corals (e.g. Acroporidae) are replaced by more robust, prostrate corals (e.g. Mussidae, Poritidae) following disturbance [ 20 , 21 ]. The composition and cover of coral species are key determinants of the structural complexity of reef habitats [ 21 , 22 ], and can exert considerable influence over the taxonomic and functional structure of reef fish assemblages [ 6 ]. However, the capacity of altered coral species configurations to support key ecosystem processes despite ongoing disturbance is largely unknown and of growing concern [ 20 , 23 ]. Herbivory, the consumption of algal material, is dominated by fishes on coral reefs with relatively intact fish assemblages. Herbivory processes can promote coral dominance by reducing the cover and/or biomass of algae, though the amount of herbivory necessary will depend on the extent of substrate available to algae, background nutrient levels that can accelerate algal increase [ 24 ] and the effect of anthropogenic ocean warming on corals [ 25 ]. If herbivory is sufficient, it can mediate competitive interactions with corals [ 26 ], mitigate shifts to macroalgal dominance following extensive coral mortality, and facilitate recovery of coral populations [ 27 ]. However, the distribution of herbivorous fishes and their rates of herbivory can be highly spatially variable; among regions [ 28 , 29 ], latitudes [ 30 ], across the continental shelf [ 31 ], with the amount of nutrients entering the system [ 32 ] and among reef zones [ 33 , 34 ]. Importantly, rates of herbivory by fishes often vary among sites within-reef zones [ 35 , 36 ], with studies relating variation to differences in habitat structural complexity [ 28 ], the cover of live coral [ 29 , 37 ], the relative palatability of resident algal communities [ 34 , 38 ], predation pressure or competition for resources [ 39 ]. Where variation in herbivory is driven by the differential composition of benthic reef habitats [ 35 ], this may carry implications for the variable functioning of distinct coral species configurations. However, relationships between coral species composition and herbivory processes by fishes at the within-reef scale remain unclear. Herbivory processes are diverse, carried out by multiple species that perform complementary, and in some cases functionally overlapping, roles in removing algae from the reef substrate [ 40 , 41 ]. For example, grazing fishes (including algal croppers/detritivores, scrapers and excavators) feed on surfaces covered by epilithic algal matrices (EAM: a conglomerate of algal turfs, macroalgal propagules, sediment, detritus and microbes [ 42 ]), but have limited capacity to remove large fleshy macroalgae [ 38 ]. By feeding on EAM covered surfaces, grazers maintain algal communities in a cropped state, reduce the growth of macroalgal propagules within the EAM, reduce coral–algal competition and thereby facilitate settlement, growth and survival of corals and coralline algae [ 41 ]. By contrast, macroalgal browsers typically feed on larger fleshy macroalgae and have the potential to reverse phase shifts by removing macroalgae biomass, facilitating the recovery of coral populations [ 27 , 43 ]. Understanding the extent to which different configurations of structurally distinct corals maintain populations of herbivorous fishes and the critical functions they provide is paramount for the management of ecological integrity yet is largely unknown. The primary objective of this study was to investigate how grazing and browsing herbivory processes by reef fishes varied among coral habitats that differed in coral species composition and structural complexity across within-reef scales [ 22 ]. Using a combination of in situ surveys and transplanted algal assays across three replicated habitats characterized by the predominance of distinct coral taxa ( sensu [ 22 ]), we specifically ask the following questions. (i) Do the structure of herbivorous fish assemblages and rates of grazing and browsing vary among reefs characterized by distinct coral habitats? (ii) What is the relative influence of coral species composition and structural complexity, and herbivore biomass on these herbivory functions within reefs?",
"discussion": "4. Discussion Shifts in the composition of habitat-forming species and consequences for the function of ecosystems pose new challenges for conservation as the composition of assemblages that rely on habitats for food and shelter reorganize [ 7 , 51 ]. Focusing on coral reefs, we show that the taxonomic and functional composition of herbivorous fish assemblages, and rates of browsing, but not grazing, differed among taxonomically distinct coral habitats. Browsing on the red macroalga Laurencia was greatest in soft coral and mixed coral habitats, and lowest in branching Porites habitats. These differences in the consumption of Laurencia were best predicted by variation in both the composition and cover of benthic assemblages, with the highest rates of removal in habitats with the lowest coral cover, lowest structural complexity, and highest cover of dead substrata and macroalgae. Interestingly, rates of browsing on Laurencia were poorly reflected by visual estimates of the biomass of browsing fishes, despite browsing fishes being recorded in all three habitats. By contrast to browsing rates, grazing on algal turfs did not differ among habitats. This contrast highlights that different environmental mechanisms, such as those determined by the influence of differential habitat characteristics on foraging behaviour, may shape the functional impact of key species and functional groups such that shifts in species configurations under mounting disturbances may have varied consequences for maintained ecosystem function [ 7 , 8 ]. The observed variation in rates of browsing among habitats was best predicted by the cover and composition of benthic communities, indicating that particular habitat characteristics may influence foraging behaviour and/or habitat selection by browsing reef fishes. The cover of live coral and structural complexity of reef habitats typically have positive effects on the abundance, biomass, and diversity of herbivorous fish communities [ 33 , 52 ], and rates of herbivory [ 35 , 36 ]. By contrast, however, we found that browsing on Laurencia was greater in habitats with lower coral cover that had lower structural complexity, and higher cover of dead substrata and macroalgae (e.g. mostly mixed coral habitats, largely characterized by massive and branching Porites , Sarcophyton , Lobophyton ). Conversely, while branching Porites habitats were the most structurally complex [ 22 ], had the highest coral cover, and the greatest observed biomass of browsing fishes among habitats, no significant reduction in Laurencia biomass was detected over a 24 h period. The negative relationship between the cover of structurally complex corals (and conversely the positive relationship with the cover of dead substrata and macroalgae) and browsing rates may be related to increased levels of visual occlusion during feeding in high-relief habitats and hence greater risk of foraging [ 53 , 54 ]. Studies show the physical topography of structurally complex habitats can inhibit access to algal resources at fine scales (i.e. between coral branches [ 55 ]), and can alter the foraging behaviour of fishes by reducing their visual fields and thereby enhancing perceived predation risk [ 53 ]. Such findings reflect patterns of habitat use in other terrestrial and aquatic systems where foraging species favour open over structurally complex habitats due to the enhanced ability to detect approaching predators (e.g. African savannahs [ 56 , 57 ]; temperate intertidal rocky shores and mudflats [ 58 ]; alpine forests [ 59 ]; European grasslands [ 60 ]; temperate arable areas [ 61 ]). Indeed, evidence shows that visual obstruction can increase vigilant predator-scanning behaviour at the cost of time spent foraging in various taxa [ 57 , 60 ]. Moreover, perceived predation risk can also be mediated by body size with larger prey less susceptible to predation [ 56 ]. Of the four main species recorded feeding on Laurencia in our study, only P. sexstriatus was observed feeding within the structurally complex branching Porites habitat, despite N. brevirostris and S. doliatus being recorded in visual surveys of that habitat. P. sexstriatus was the largest-bodied species observed (mean biomass ± s.e.: 670 g ± 77; other species mean biomass 195–539 g), potentially reducing predation risk and enabling less discriminant foraging activity. The positive relationship between browsing and the cover of dead substrata and macroalgae (which was highly collinear with the cover of live coral), also suggests that habitat condition may influence the foraging behaviour of herbivore fishes. Indeed, feeding rates by herbivorous reef fishes can be higher in degraded areas, of often lower topographic complexity [ 37 ]. By feeding where food resources are more abundant, animals may maximize net energy gain by reducing energetic costs of movement [ 62 , 63 ], and risk of predation associated with moving larger distances between resource patches [ 64 ]. In our study, differential browsing rates may relate to the differential availability of algal dietary resources [ 35 , 39 ] following the bleaching event that caused coral loss and increased the cover of turf algae ( figure 1 a ) [ 6 ] at our study sites (between 52.4 and 71.4% cover of dead substrata). Browsing on Laurencia was greatest in mixed coral habitats that also had the highest cover of dead substrata and macroalgae as a result of the bleaching (due to loss of mainly Acropora and soft coral taxa [ 6 ]), and highest biomass and diversity of herbivorous fish. Increased cover of algae (predominately turf communities) following large-scale bleaching-induced coral mortality and subsequent increases in the abundance and/or biomass of herbivorous fishes (e.g. [ 65 ]), has led to suggestions that herbivorous fish populations may be food limited in areas of high coral cover [ 66 ]. However, this relationship may not hold at very low levels of macroalgal cover [ 34 ], such as those observed in the present study (mean: 0.3–1.4% cover). While visual census estimates show macroalgal browsing herbivores are present in each of the studied habitats, browser biomass was a poor predictor of browsing rates. This is consistent with previous studies of herbivorous coral reef fishes [ 36 , 50 ] and processes in other systems (e.g. the decomposition of dung by invertebrates [ 67 ]; pollination by bees [ 13 ]) in which abundance shows little relation to their functional impact. The discrepancy between observed browser presence and function in our study may also reflect the high mobility and opportunistic foraging behaviour of roving herbivores [ 68 ], or the diver-negative behaviours of some fishes [ 69 ]. The utility of using the density or biomass of browsing herbivores as a proxy for macroalgal removal may be further hindered by the plasticity and opportunistic diets among herbivorous fishes [ 47 ], and a potential bias in the literature classifying browsers as those species known to feed on large fleshy brown macroalgae versus those that consume other fleshy macroalgae [ 48 ]. By contrast to browsing, there were no detectable differences in grazing on the algal turf assays among habitats. This provides further evidence of a disconnect between the observed density and realized the impact of functional groups of herbivorous fishes. Despite no detectable differences in grazing rates, among-habitat differences in herbivore assemblages were largely driven by differences in the biomass of grazing species. The lack of among-habitat variation in grazing may be related to the high diversity of fishes that feed on algal-turf covered substrata [ 41 ], and their response diversity to changes in benthic composition [ 70 ]. Similarly, the lack of observed differences may be due to grazing herbivores preferentially targeting sparse and short early successional turfs and avoiding later successional dense turf assemblages [ 71 ]. Feeding rates and foraging behaviours of grazing coral reef fish species have been shown to vary with the condition and structure of reef habitats and algal communities, however, responses tend to be species specific [ 37 ]. The among-habitat variation in the changes in the turf height on caged tiles was interesting as, despite feeding by large herbivorous fishes being excluded, there was a decline in height in soft coral habitat and increase in branching Porites habitat which may be related to grazing by small invertebrates and/or differences in algal productivity [ 72 ]. Similarly, negative values of turf height loss for both caged and exposed assays in branching Porites habitats may be due to high algal productivity in that habitat, warranting further investigation. Our results provide new evidence of the variable influence of the composition and cover of habitat-building corals on two key functions on coral reefs—grazing and browsing—based on comparisons among three taxonomically distinct coral habitats. While the use of Laurencia has provided valuable information on the variable browsing behaviour among habitats, previous studies have shown rates of macroalgal browsing can be dependent on the macroalgae used due to feeding preferences of local herbivore assemblages [ 48 , 49 ]. Therefore, further investigation using other commonly occurring macroalgae may offer insight into behavioural variation among habitats of a broader suite of herbivores. Similarly, herbivory processes can vary with depth, exposure and reef zonation [ 33 , 73 , 74 ]. Our study compared relatively small experimental assay units among habitats within in a sheltered lagoon environment. Therefore, further study across a wider range of environmental gradients, reef zones, across additional coral species configurations and across broader spatial scales is now needed. Our study coincided with a large-scale bleaching event [ 44 ], resulting in rapid coral loss and changes in reef fish assemblage structure among our study sites [ 6 ], and likely affected the foraging behaviour of a range of reef fish species including herbivores [ 15 , 65 , 75 ]. Although the present study provides clear evidence of how herbivory processes can vary with coral species composition, it was carried out in the context of this disturbance. Disturbance dynamics are complex [ 15 , 70 ], and it is likely that fish assemblages are in transition with changes in coral cover. Further research into the spatio-temporal variation in foraging behaviour of individuals and functional groups across such disturbances would improve our understanding of how changing reef configurations interact with climate change impacts to influence critical ecological functions [ 15 , 16 ]. Understanding causal links between habitat species composition and ecosystem function is of growing concern in this era of unprecedented and rapid ecological change [ 5 , 7 , 9 ]. In particular, elucidating how the increasing modification of ecological communities affects ecosystem processes is central to our capacity to anticipate whether new species configurations will continue to provide goods and services as required by societies that depend on them [ 14 , 17 , 23 ]. On coral reefs, whether herbivores can compensate for increased algal production as coral cover decreases, and maintain critical rates of algal consumption will be fundamental to the persistence of reconfigured coral-dominated systems [ 66 ]. Our results show that herbivore assemblage structure varied among the studied habitats, however, did not reflect the observed variation in herbivory rates. While grazing was insensitive to variation in coral composition, browsing varied considerably, indicating that different mechanisms determined by specific habitat characteristics may be shaping these key processes. While the precise mechanisms are not known, variation in browsing was best predicted by the composition and cover of benthic communities, and conversely the cover of dead substrata and macroalgae, characteristics that underscore the structural complexity of reef habitats and which may have influenced differential foraging behaviour. With ongoing degradation of coral reefs and the homogenization of both coral and fish assemblages [ 6 , 20 ], these results suggest that, within reefs, key ecosystem functions will likely vary among altered coral configurations, according to the differential vulnerability of corals to disturbances and ecological interactions between reef fishes and their environment [ 15 ]. More generally, our results emphazise the role of differential habitat characteristics and provide explicit support for assigning greater concern to the composition and structure—as well as cover—of habitat-building species in assessments and management of ecosystem function [ 7 , 23 ]."
} | 5,165 |
26579146 | PMC4621405 | pmc | 2,744 | {
"abstract": "Continuous hydrogen photo-production under sulfur deprivation was studied in the Chlamydomonas reinhardtii pgr5 pgrl1 double mutant and respective single mutants. Under medium light conditions, the pgr5 exhibited the highest performance and produced about eight times more hydrogen than the wild type, making pgr5 one of the most efficient hydrogen producer reported so far. The pgr5 pgrl1 double mutant showed an increased hydrogen burst at the beginning of sulfur deprivation under high light conditions, but in this case the overall amount of hydrogen produced by pgr5 pgrl1 as well as pgr5 was diminished due to photo-inhibition and increased degradation of PSI. In contrast, the pgrl1 was effective in hydrogen production in both high and low light. Blocking photosynthetic electron transfer by DCMU stopped hydrogen production almost completely in the mutant strains, indicating that the main pathway of electrons toward enhanced hydrogen production is via linear electron transport. Indeed, PSII remained more active and stable in the pgr mutant strains as compared to the wild type. Since transition to anaerobiosis was faster and could be maintained due to an increased oxygen consumption capacity, this likely preserves PSII from photo-oxidative damage in the pgr mutants. Hence, we conclude that increased hydrogen production under sulfur deprivation in the pgr5 and pgrl1 mutants is caused by an increased stability of PSII permitting sustainable light-driven hydrogen production in Chlamydomonas reinhardtii .",
"conclusion": "Conclusion Enhanced hydrogen production rates in the pgrl1, pgr5 , and the pgr5 pgrl1 double mutant under S deplete conditions lead to the highest continuous photobiological produced hydrogen amounts of eukaryotic cells reported so far. These rates are achieved by a prolonged residual PSII activity providing an increased electron supply toward the hydrogenase. PSII activity can be maintained in the mutants without inhibiting the oxygen sensitive hydrogenase, because the oxygen consumption capacity is increased. Our results suggest that respiration and light-dependent O 2 -uptake rates are higher in the pgr mutants, and that this is responsible for the faster transition to anaerobiosis, especially when the greater residual PSII activity of the mutants is taken into account.",
"introduction": "Introduction Solar fuels are an important motive for the development of future renewable energy systems with zero CO 2 emission. This development is necessary to meet one of the most urgent challenges of our society today, to counter the problems of global warming, fossil fuel depletion, concurrent increasing energy demand, and consequently the maintenance of economic and political stability ( Organisation for Economic Co-operation and Development (OECD)/International Energy Agency (IEA), 2011 ). Among other fuels, hydrogen is considered to be one of the most effective and clean fuels ( Hankamer et al., 2007 ). Solar-driven H 2 production by photosynthetic microorganisms, particularly cyanobacteria and microalgae, is a promising complement to clean and sustainable technologies of hydrogen production beside chemical techniques. Photobiological hydrogen production was first discovered by Gaffron and Rubin, 1942 . In this process, electrons and protons from water splitting are directed via photosynthesis toward specific H 2 -evolving enzymes, the hydrogenases. The algal Fe-Fe hydrogenase is very efficient compared to other hydrogenases (turnover rate in thousands per second, 100-fold higher than other hydrogenases; Volgusheva et al., 2013 ; Lubitz et al., 2014 ). However, direct light-to-hydrogen conversion efficiency is very low, because hydrogenase activity is extremely sensitive to oxygen ( Ghirardi et al., 1997 ; Rupprecht et al., 2006 ; Stripp et al., 2009 ), thus oxygenic photosynthesis cannot easily be directly coupled to hydrogen production in green algae. Therefore, hydrogen production is a transient phenomenon in nature and stops after a few minutes of illumination ( Bishop and Gaffron, 1963 ). Melis et al. (2000) proposed an experimental protocol for prolong H 2 evolution based on sulfur deprivation that circumvents this limitation. This method allows the separation of photosynthetic oxygen evolution and hydrogen production by a two stage process: In the first phase of cell cultivation, oxygenic photosynthesis drives production of biomass and carbohydrate stores in the presence of acetate as an additional carbon source. Shifting cells to sulfur depleted medium in sealed flasks induces the switch to the second anaerobic stage, inducing hydrogenase expression and sustainable H 2 production for several days. During acclimation to this nutrient stress, cells stop dividing and undergo morphological changes ( Zhang et al., 2002 ). Both light and dark reactions of photosynthesis are down-regulated, with the amounts of Rubisco being substantially reduced within the first 24 h of sulfur starvation ( Zhang et al., 2002 ). Photosystem II activity drops gradually, attributed to an impaired PSII repair cycle due to the restricted de novo synthesis of the D1 reaction center protein (which contains methionines and cysteines) by S-limiting conditions, driving down O 2 evolution ( Wykoff et al., 1998 ). When O 2 consumption overtakes O 2 evolution, anaerobic conditions are established. However, active PSII in the first hours of S-depletion was shown to be essential for H 2 generation, as no hydrogen evolution could be observed when the PSII inhibitor DCMU was added directly after transfer to S-free medium ( Fouchard et al., 2005 ; Hemschemeier et al., 2008 ). Indeed, PSII activity was reported to contribute a substantial amount of electrons from water oxidation at PSII to hydrogen production ( Antal et al., 2003 ; Kosourov et al., 2003 ). Beside this direct PSII-dependent pathway, electrons toward the hydrogenase can also derive from an indirect pathway that relies on non-photochemical reduction of PQs from metabolites such as starch ( Fouchard et al., 2005 ; Jans et al., 2008 ; Chochois et al., 2009 ). Starch is massively accumulated during the first hours of –S conditions and is subsequently degraded ( Zhang et al., 2002 ). The direct sunlight to hydrogen pathway has nevertheless the potential for higher energy conversion efficiency. The improvement in hydrogen production by the method of sulfur deprivation implicated the highest efficiency for photobiological systems reported by then ( Rupprecht et al., 2006 ). Yet, this H 2 production efficiency has still to be advanced to reach economic profitability. Optimization of the electron supply to the hydrogenase appears to be a critical issue. Since the majority of electrons toward the hydrogenase have been shown to derive from PSII activity, but O 2 evolution by PSII has to be prevented not to inhibit hydrogenase activity, possible solution scenarios involve either a more O 2 tolerant hydrogenase or a more efficient oxygen scavenging system. Engineering resulting in a decrease in the O 2 sensitivity of the Fe–Fe hydrogenase has not been reported yet. Among some other Chlamydomonas reinhardtii mutant strains, the state transition 6 ( stm6 ) mutant lacking the mitochondrial respiratory chain assembly factor moc1 , and the proton gradient regulation like 1 ( pgrl1 ) mutant displayed the highest improved hydrogen production rates ( Kruse et al., 2005 ; Tolleter et al., 2011 ). The PGRL1 protein was first discovered in Arabidopsis ( DalCorso et al., 2008 ) as an essential component of the PGR5 (proton gradient regulation 5) dependent cyclic electron flow (CEF) pathway ( Munekage et al., 2002 ). Moreover, PGRL1 has been suggested to operate as the elusive ferredoxin-plastoquinone reductase ( Hertle et al., 2013 ). In Chlamydomonas , PGRL1 is also required for effective CEF ( Petroutsos et al., 2009 ; Tolleter et al., 2011 ) and suggested to be a functional component of a CEF-supercomplex formed under CEF promoting conditions ( Iwai et al., 2010 ; Terashima et al., 2012 ). A pgr5 mutant has recently been characterized in Chlamydomonas ( Johnson et al., 2014 ), which revealed that PGR5 deficiency leads to a diminished proton gradient across the thylakoid membrane accompanied with less effective CEF capacity. The proton gradient across the thylakoid membrane was shown to restrict electron flow toward the hydrogenase in pgrl1 , and also proton uncouplers like carbonyl cyanide- p -trifluoromethoxyphenylhydrazone (FCCP) and nigericin increased hydrogen evolution ( Antal et al., 2009 ; Tolleter et al., 2011 ). Interestingly, both stm6 and pgrl1 share the characteristic of an increased respiration rate ( Kruse et al., 2005 ; Petroutsos et al., 2009 ; Dang et al., 2014 ). Enhanced H 2 production in stm6 was recently linked to prolonged PSII activity under sulfur starvation ( Volgusheva et al., 2013 ). In this study, we investigated hydrogen production under similar conditions in the PGR5-deficient mutant and a pgr5 pgrl1 double mutant in comparison to the pgrl1 mutant. These mutants also exhibit enhanced hydrogen production and higher PSII stability under –S conditions. Since blocking PSII activity by DCMU abolished hydrogen production almost completely, we conclude that electrons for enhanced hydrogen production mainly derive from PSII. Anaerobiosis and therefore hydrogenase activity can still be maintained due to a higher oxygen consumption capacity in the mutants.",
"discussion": "Discussion In the present study, we investigated pgr5 and pgr5 pgrl1 double mutants in comparison to a PGRL1 deficient mutant and wild type in their capacity of light-driven hydrogen production. As already reported earlier for PGRL1 deficiency in C. reinhardtii ( Tolleter et al., 2011 ), the absence of the pgr5 gene product also increased the capacity of hydrogen production as compared to wild type significantly. Importantly, the pgr5 mutant is, with hydrogen volumes of up to 10 times higher than the wild type, one of the highest hydrogen producers reported to date. The maximum amounts of 850 ml L -1 culture (at 15 μg ml -1 chlorophyll referring to 2.4 μmol H 2 μg -1 chlorophyll) exceed reported volumes of pgrl1 and other high hydrogen producers like stm6 , stm6glc4 , LO1 , or the PSII D1 mutant L159I-N230Y (see Table 1 , Kruse et al., 2005 ; Doebbe et al., 2007 ; Torzillo et al., 2009 ; Oey et al., 2013 ). Our data indicate that (i) the ability to maintain linear photosynthetic electron flow rates under S deplete conditions in conjunction with (ii) an increased oxygen consumption capacity explains the increased hydrogen production in the pgr mutants. Table 1 Comparison of hydrogen amounts produced by different Chlamydomonas mutants. Mutant H 2 volume (ml/L culture) Chlorophyll (μg/ml) H 2 volume (μmol H2/μg chl) PBR size (ml, ø in cm) Reference pgr5 850 15 2.4 250, 6.5 pgrl1 580 15 1.6 250, 6.5 Also Tolleter et al., 2011 pgr5 pgrl1 610 15 1.7 250, 6.5 stm6 540 26 0.9 500, 8.0 Kruse et al., 2005 stm6 glc4 150% of stm6 26 1.3 500, 8.0 Doebbe et al., 2007 LO1 400 15 1.1 500, 8.0 Oey et al., 2013 L159I-N230Y 504 12 1.75 1000, 5.0 Torzillo et al., 2009 Absolute H 2 volumes are taken from specified publications. For a better comparison, H 2 amounts are equalized to chlorophyll content (column 4). Photobioreactor (PBR) sizes are given as one possible factor introducing some variance between studies due to differences in experimental set-ups. Elevated Hydrogen Production in the pgr Mutants: Interplay between Capacities of Photosynthetic Electron Transfer and Oxygen Consumption It is known that the hydrogenase is extremely sensitive to oxygen ( Ghirardi et al., 1997 ; Stripp et al., 2009 ). On the other hand, linear electron transfer will produce oxygen. Compared to maximal wild type rates, the hydrogen amounts produced per hour were four times higher in the pgr mutants with maximal rates of 7 ml L -1 culture per hour over a period of 2–3 days ( Figure 1A ). The wild type stopped producing hydrogen slightly earlier than the mutants ( Figures 1A and 2B ). After 2 days of sulfur deprivation (time point of DCMU addition, Figure 1B ), the wild type produced already 46% vs. only 12% of hydrogen produced by the pgr mutants compared to the total hydrogen amounts produced over 9 days. This time point correlates with the drop in PSII activity in the wild type ( Figure 1C ). Thus, PSII activity of the pgr mutants remained higher until the late phase of sulfur deprivation ( Figures 1C and 3 ) suggesting that linear electron flow from PSII is responsible for the prolonged and significantly higher hydrogen production rates in the pgr mutants. Why the hydrogenase is not inhibited by elevated PSII activity in the pgr mutants? This is because anaerobiosis can be maintained in the pgr mutants due to an increased oxygen consumption rate, as oxygen is removed much faster from the measuring system compared to wild type ( Figures 1D and 2C,E ). This is further supported by the fact that light-induced oxygen uptake is increased in the pgr5 mutant ( Figure 4 ). Increased respiration rates were already reported for the pgrl1 ko and kd lines ( Petroutsos et al., 2009 ; Tolleter et al., 2011 ). Recently, Dang et al. (2014) reported increased mitochondrial cooperation and increased O 2 photo-reduction in the pgrl1 mutant and also in Arabidopsis pgr5 , this cooperation was enhanced ( Yoshida et al., 2007 ). Metabolic shuttles such as the malate-oxaloacetate shuttle can export reducing power from the stroma to the mitochondria ( Scheibe, 2004 ; Shen et al., 2006 ). This cooperation is particularly enhanced under stress conditions, where ATP demand is increased ( Lemaire et al., 1988 ). Mitochondrial inhibitors induced a drop in the PSII yield in pgrl1 ( Dang et al., 2014 ). Also light-dependent O 2 uptake and the abundance of flavodiiron proteins were higher in the mutant, indicating an increased capacity of Mehler-like reactions and possibly photorespiration. High antioxidant capacity was reported for Arabidopsis pgr5 ( Suorsa et al., 2012 ). A recent study showed an up-regulation of those proteins in the early acclimation phase of sulfur deprivation, suggesting an involvement of flavodiiron proteins in acclimation to anoxia during hydrogen production ( Jokel et al., 2015 ). In addition to the very similar phenotype of the pgrl1 and pgr5 mutants in regard to CEF impairment, NPQ reduction and PSI photo-inhibition ( Tolleter et al., 2011 ; Johnson et al., 2014 ; Kukuczka et al., 2014 ), O 2 consuming mechanisms are also up-regulated under S-deplete mixotrophic conditions in pgr5 ( Figures 1D , 2D,E and 4 ) and participate in a faster anaerobic induction at the onset of sulfur deprivation ( Figures 1D and 2D,E ). This is in line with the conclusion, that increased oxygen consumption rates allow higher PSII activity under anaerobic –S conditions leading to increased hydrogen production in the pgr mutants. Interestingly, in a recent study using the stm6 and PSII-D1 L159I-N230Y mutant (see also Table 1 ), the authors reported very similar phenotypes under sulfur deprivation as is shown here for the pgr mutants: the mutants reached anaerobiosis faster and increased hydrogen production was mainly achieved by an increased electron supply from PSII toward the hydrogenase due to an increased respiration rate ( Torzillo et al., 2009 ; Scoma et al., 2012 ; Volgusheva et al., 2013 ). The contribution of the indirect pathway of electrons from starch via NAD2 toward the hydrogenase to H 2 production has been reported to be variable, depending on experimental conditions and particularly on the phase of the sulfur deprivation process ( Kosourov et al., 2003 ; Fouchard et al., 2005 ; Hemschemeier et al., 2008 ; Jans et al., 2008 ). It might also contribute to increased hydrogen production in the pgr mutants. Starch breakdown was reported to be faster in the pgrl1 mutant ( Tolleter et al., 2011 ) and from a study with a starch deficient mutant and the inhibitor DCMU and proton uncoupler FCCP, it was concluded that the proton gradient generated by cyclic electron flow around PSI inhibits the indirect pathway of electrons from starch to the hydrogenase ( Chochois et al., 2009 ). Tolleter et al. (2011) reported hydrogen production rates in pgrl1 are enhanced compared to the wild type, and to a similar extent independent of the presence of DCMU. However, H 2 production was 10 times lower than in the absence of DCMU ( Tolleter et al., 2011 ). In our hands, the hydrogen production in the presence of DCMU was even lower ( Figure 1B ), possibly because DCMU was added with a 1-day delay (48 h vs. 24 h of sulfur deprivation). Therefore, since blocking the PSII-dependent direct pathway by DCMU had such a dramatic effect on hydrogen production, the contribution of the indirect pathway to hydrogen production in the pgr mutants should be relatively small compared to the PSII-dependent pathway. Deletion of PGRL1 and PGR5 Sustains PSII Activity under –S Conditions The mechanisms contributing to the inhibition of photosynthetic O 2 evolution in S depleted conditions and their relative importance are still a matter of debate ( Ghysels and Franck, 2010 ). Sulfur deprivation has a general effect on the transcription of the chloroplast ( Irihimovitch and Stern, 2006 ; Irihimovitch and Yehudai-Resheff, 2008 ). It is well described that acclimation to sulfur deficiency is highly controlled and induces a down-regulation of photosynthesis, particularly of PSII ( Davies et al., 1996 ). The absence of down-regulation of photosynthetic activity, as observed in the sac1 mutant ( Davies et al., 1996 ), is deleterious. It has been suggested that sulfur limitation will result in a general decrease in protein synthesis, which in turn would impact the capacity of D1 turn-over and repair cycle ( Wykoff et al., 1998 ; Zhang et al., 2002 ). Hence, under low sulfur and in the absence of efficient PSII-repair, high activity of photosynthetic electron transfer would likely result in damage due to the production of reactive oxygen species, which could not be sustained by D1-repair. In the pgr mutants, however, PSII photochemistry efficiency remained high ( Figure 1C ) and amounts of the D1 protein in the mutants showed some stability compared to the wild type during sulfur starvation ( Figure 3 ), at which PSII activity in the wild type had already dropped to 10% ( Figure 1C ). The inhibitory effect of chloroplastida translation by lincomycin on PSII activity and H 2 production further indicates that a substantial amount of PSII remains functional under sulfur deprivation ( Figures 1B,C ). Thus, the absence of PGRL1 and PGR5 sustained PSII activity under sulfur deprivation. How could this be explained? The increased oxygen consumption capacity in the mutants, and the faster transition to anaerobiosis is, in its own right, sufficient to explain PSII preservation, via at least two mechanisms. First, rapid attainment of anaerobiosis leads to early induction and high activity of the hydrogenase, the final electron acceptor of linear electron flow in –S conditions. This is important because CO 2 fixation is lost during S deprivation, which normally limits LEF by slowing down the reoxidation of its main final electron acceptor NADP + : Rubisco is affected even earlier than D1 and declines by about 80% in the first 24 h of S deprivation, becoming undetectable after 60 h of starvation ( Zhang et al., 2002 ). Loss of CO 2 -fixation has been reported to slow down the D1 repair cycle and was therefore predicted to stimulate the loss of PSII activity during S deprivation ( Takahashi and Murata, 2005 , 2006 ). In summary, early anaerobiosis releases acceptor side limitation of LEF faster due to an earlier activation of the hydrogenase ( Figures 1D and 2D ). Second, faster induction of anaerobiosis at the onset of sulfur starvation and induction of hydrogenase activity preserves existing PSII from photo-oxidative damage, as it reduces the amount of oxygen available to cause potential damage at PSII. All strains show an initial decline in Fv/Fm, which slows after the system becomes anaerobic, but as this occurs much later in the wt, the corresponding damage to PSII is much worse. Under High Light Exposure, pgr5 pgrl1 shows an Increased Hydrogen Burst at the Beginning of Sulfur Depletion and PSI becomes Photo-inhibited in the pgr5 and pgr5 pgrl1 Mutants In this study, we report the generation of a pgr5 pgrl1 double mutant. This double mutant displays a similar phenotype compared to its respective single mutants, confirming that PGR5 and PGRL1 both act on the same pathway as earlier proposed ( DalCorso et al., 2008 ; Johnson et al., 2014 ). However, pgrl1 and pgr5 differ in the severity of one particular phenotype: While PSII remained stable in sulfur limiting conditions, PSI was degraded much faster in the pgr5 and pgr5 pgrl1 mutants when they were exposed to additional stress of high light ( Figure 3 ), reducing the amount of hydrogen produced ( Figure 2 ; Supplementary Figure S2 ). In contrast, pgrl1 remained effective in hydrogen production also in high light and PSI was not degraded at the same rate ( Figures 2 and 3 ). Although PGR5 was previously not detectable by western blot analysis in the pgrl1 mutant ( Johnson et al., 2014 ), the different phenotype of the pgr5 pgrl1 and pgr5 mutant compared to the pgrl1 single mutant suggests that the PGR5 protein is present in the pgrl1 mutant, though below the limit of detection of immunoblot analysis, which could detect about 50% of PGR5 wild type protein amounts ( Johnson et al., 2014 ). High light sensitivity has already been reported for pgr5 under photoautotrophic conditions ( Johnson et al., 2014 ). In pgrl1 , PSI becomes photo-inhibited under high light exposure ( Kukuczka et al., 2014 ), although not degraded as severely as in pgr5 . In the absence of both PGR5 and PGRL1, PSI becomes even more affected, indicating an additive role of these two proteins in terms of PSI protection and suggesting that both proteins operate in the same pathway. PSI photo-inhibition in high light was also described in a Chlamydomonas stt7-1 mutant locked in state 1 ( Bergner et al., 2015 ). However, this phenotype was less severe than in the pgr5 and pgrl1 mutants ( Johnson et al., 2014 ; Bergner et al., 2015 ). Furthermore, the stt7-1 mutant displayed high CEF rates and enhanced formation of a CEF-supercomplex ( Bergner et al., 2015 ). A non-successful acclimation to high light was recently also linked to altered dephosphorylation patterns of LHCII proteins in Arabidopsis pgr5 ( Mekala et al., 2015 ). These findings indicate a regulatory link between the PGR5/PGRL1 pathway, phosphorylation and dephosphorylation dynamics and the susceptibility of PSI to photo-inhibition. In high light, the thylakoid lumen becomes acidified, which down-regulates linear electron flow and induces dissipation of excess energy at PSII in the form of heat. This so called non-photochemical quenching (NPQ) requires the induction of the LHCSR3 protein in algae ( Peers et al., 2009 ) and acidification of the thylakoid lumen ( Petroutsos et al., 2011 ). Both pgr5 and pgrl1 have reduced levels of NPQ due to a reduced proton gradient across the membrane, while LHCSR3 levels are unaltered in the mutants ( Tolleter et al., 2011 ; Johnson et al., 2014 ). Without a sufficient proton gradient, linear electron flow is not down-regulated and the acceptor side of PSI becomes over-reduced. An indication that an increased electron drive toward the hydrogenase occurs under elevated light conditions is the enhanced hydrogen burst at the beginning of the hydrogen production phase in the pgr5 pgrl1 mutant ( Figure 2C ). Hydrogen production at PSI was proposed to act as a safety valve to protect the photosynthetic electron transport chain from over-reduction under natural conditions by safely disposing of excess electrons from PSI ( Melis and Happe, 2001 ; Happe et al., 2002 ; Hemschemeier et al., 2009 ). Now, the deletion of PGR5 and/or PGRL1 is deleterious for PSI under high light conditions. Thus, while PSII stability under sulfur deficiency is more affected in wild type, PSI integrity is most affected in the absence of PGR5 and/or PGRL1; a topic that certainly requires additional work for a more in-depth understanding."
} | 6,151 |
36465113 | PMC9708700 | pmc | 2,745 | {
"abstract": "Summary Chemoautotrophic bacteria play an important role in combating the rise in global CO 2 . However, recently it was found that extracellular free organic carbon (EFOC) generated by chemoautotrophic bacteria inhibits their CO 2 fixation. Although continuous-flow membrane bioreactor can remove EFOC and enrich bacteria, it may also remove beneficial bio-factors for bacterial growth. Finding out the main inhibitory factors and inhibitory mechanisms in EFOC can provide theoretical guidance for the development of targeted inhibitory component removal technology. The results show a significant negative correlation between the increasing proportion of small-molecule EFOC and the decreasing trend of CO 2 fixation efficiency, and simulation experiments confirm that the small molecule organics such as amino acids and organic acids are the main components of EFOC that inhibit CO 2 fixation by inhibiting ribulose bisphosphate carboxylase/oxygenase (RuBisCO) gene ( cbb ) transcription efficiency. Therefore, amino acids and organic acids are suggested to be recovered to promote efficient CO 2 fixation of autotrophic bacteria.",
"conclusion": "Conclusions When EFOC is analyzed in terms of the molecular weight of its components, the proportion and concentration of the small molecules in EFOC increase during chemoautotrophic culture, and the small molecules are the main components of EFOC that inhibit CO 2 fixation by chemoautotrophic bacteria. In terms of the types of compounds in small-molecule EFOC, amino acids and organic acids are the main inhibitory components. The inhibitory effects of amino acids and organic acids on CO 2 fixation are stronger than those of protein and carbohydrate, respectively, because they also inhibit the transcription of the cbb gene. Therefore, nanopore adsorption-culture technology for the adsorption of amino acids and organic acids is suggested to enhance CO 2 fixation by chemoautotrophic bacteria.",
"introduction": "Introduction Carbon capture, utilization, and storage (CCUS) have been regarded as a pivotal technological method for reducing carbon emissions and mitigating climate change. 1 , 2 , 3 The role of chemoautotrophic bacteria in CCUS is too significant to be ignored. 4 Chemoautotrophic bacteria are widespread, and fix CO 2 without light in a variety of harsh habitats (including deep-sea hydrothermal vents, barren desert soil, and large-scale gas treatment reactors). 5 , 6 , 7 , 8 However, the large-scale industrial use of chemoautotrophic bacteria to fix CO 2 is currently limited by their low growth rates and poor ability to fix carbon. 9 , 10 , 11 The internal factors affecting the growth rates of chemoautotrophic bacteria include the copy numbers of their ribosomal RNA (rRNA) genes, the protein synthesis rate of unit ribosomes, and the transcription efficiency of the key genes required for CO 2 assimilation, which provides the substrates for cytoskeleton synthesis. 12 , 13 , 14 , 15 The Calvin-Benson-Bassham (CBB) cycle is considered the main pathway by which most aerobic chemoautotrophic bacteria assimilate CO 2 . 14 , 16 The transcription level of the cbb gene, which encodes the key enzyme of the CBB cycle pathway, ribulose bisphosphate carboxylase/oxygenase (RuBisCO), determines the CO 2 assimilation rate of chemoautotrophic bacteria, which in turn affects their growth rate and CO 2 fixation rate. 17 , 18 In addition to these internal factors, the growth rate of chemoautotrophic bacteria is also affected by environmental factors, such as temperature, the carbon source, pH, O 2 , and the energy source, which play important roles in the bacterial growth and the CO 2 fixation of autotrophic microorganisms. 19 , 20 It has also been demonstrated that external organic compounds generally have a negative impact on the CO 2 assimilation and cell growth of chemoautotrophic bacteria. 21 , 22 , 23 , 24 , 25 Importantly, the self-generated extracellular free organic carbon (EFOC) generated by chemoautotrophic bacteria has an inhibitory feedback effect on bacterial growth and CO 2 fixation. 19 , 26 According to several reports, membrane reactors promote bacterial growth and CO 2 fixation by separating EFOC from the culture environment, 26 , 27 confirming that EFOC has an inhibitory effect on the growth and CO 2 fixation of autotrophic bacteria. Consequently, the inhibition exerted by self-generated EFOC contributes to the relatively low growth rate and low CO 2 fixing efficiency of chemoautotrophic bacteria and may be the most important reason for their inability to sustain efficient autotrophic growth and CO 2 fixation. Although the inhibitory effect of EFOC on bacterial CO 2 fixation has been estimated and a membrane filtration technology has been developed to release the inhibitory effect of EFOC, 26 , 27 there are still several problems with membrane filtration technology, including membrane fouling, high membrane pressures, high energy consumption, and the difficulty of large-scale operation. Moreover, our understanding of the inhibitory components of EFOC that are responsible for low CO 2 fixing efficiency is also limited, and there are no detailed data on the mechanism by which EFOC inhibits the CO 2 fixation process of chemoautotrophic bacteria. Free organic carbon (FOC) is universally generated by chemoautotrophic bacteria during the assimilation of CO 2 inside the cell, and can be secreted to the outside of the cell to form EFOC when too much FOC accumulates inside the cell. 26 , 28 , 29 , 30 , 31 EFOC is a soluble microbial product with a wide distribution of molecular weight (MW), ranging from <5 kDa to >10 kDa, and contains several metabolites, mainly including carbohydrates, peptides, proteins, some lipids, and organic colloids. 26 , 30 , 32 , 33 Large-molecule organic compounds (also known as “biomass-associated products”) are generated during cell lysis and decay. 26 , 33 The organic compounds generated during the assimilation of CO 2 are subsequently involved in the synthesis of the cytoskeleton. When the cytoskeleton synthesis rate cannot keep up with the CO 2 assimilation rate, small-molecule organic compounds accumulate intracellularly and the excess diffuses through the cell membrane into the extracellular environment to form the small-molecule component of EFOC. 26 , 34 Because large and small molecules are generated via different production pathways and have different properties, the components of EFOC with different molecular weights may have different inhibitory effects on CO 2 fixation, and the inhibitory targets of these EFOC components, which affect CO 2 assimilation and protein synthesis, may also differ. Consequently, in this study, we addressed the following questions. (i) What components of EFOC exert the main negative feedback inhibitory effect on bacterial CO 2 fixation? and (ii) What are the main inhibitory targets of the major inhibitory components in EFOC on bacterial CO 2 fixation? We used typical sulfur-oxidizing bacteria (SOB) as the model organism, which have two environmental functions: fixing CO 2 and removing sulfur pollutants in wastewater and gas, 35 , 36 , 37 to (i) analyze the MW distribution characteristics of EFOC and its correlation with CO 2 fixation in a control reactor, and verified the results with a membrane bioreactor, (ii) clarify the main components of small-molecule (or large-molecule) EFOC that are negatively related to bacterial CO 2 fixation efficiency and their possible mechanisms of inhibition, (iii) verify the main inhibitory components of EFOC with simulation experiments and explore the inhibitory targets of their inhibitory effects on CO 2 fixation. In contrast to previous research, which mainly focused on the external factors that inhibit bacterial CO 2 fixation, 11 , 19 , 20 in this study, we investigated the components of self-generated EFOC that inhibit CO 2 fixation by chemoautotrophic bacteria and their mechanisms. Our results provide a theoretical basis for developing targeted measures to remove the main inhibitory components from cultures of chemoautotrophic bacteria to improve their CO 2 fixation, such as the affinity adsorption of small-molecule EFOC, and circulating flow technology with microfiltration membranes for large-molecule EFOC.",
"discussion": "Results and discussion Molecular weight distribution of extracellular free organic carbon and its relationship to CO 2 fixation efficiency in typical sulfur-oxidizing bacteria The large-molecule EFOC is mainly protein and the main component of cells, and its MW is generally more than 10 kDa, while small-molecule EFOC is mainly metabolic products and wastes with MW < 10 kDa. 34 , 38 The MW distributions of the EFOC from the two strains in the control and membrane reactors and in the EFOC filtered from the membrane reactor during its operation were determined with gel filtration chromatography ( Figure 1 ). In the control reactor, the proportion of small-molecule (MW < 10 kDa) EFOC for DSM 505 increased from 0.46% to 5.43% and that for DSM 15147 increased from 49.59% to 74.41% during the operation period, whereas the proportion of large-molecule (MW > 10 kDa) EFOC from both strains in the control reactor continued to decline ( Figure 1 A). For both strains in the control reactor, the concentration of small-molecule EFOC tended to increase, whereas the concentration of large-molecule EFOC increased for DSM 505, but barely changed for DSM 15147 ( Figure 1 B). Figure 1 MW distribution of EFOC (A–D) MW distribution proportion of EFOC in the control reactor (A), EFOC in the control reactor (B), MW distribution proportion of EFOC in the membrane reactor (C) and EFOC filtered from the membrane reactor (D) during operations of DSM 505 and DSM 15147 at early, middle, and late stages. These results show that large-molecule organic matter was the main component of the EFOC from DSM 505, whereas small-molecule organic matter was the main component of the EFOC from DSM 15147. A more important conclusion is that for both strains, the proportion of large-molecule EFOC decreased as the culture time increased, indicating that neither strain in the reactor reached the death phase, because large-molecule organic compounds (also known as “biomass-associated products”) are generated during cell lysis and decay. 26 , 33 The total apparent carbon fixation yields of DSM 505 and DSM 15147 in the control reactor increased rapidly in the early stage of culture and slowly in the late stage ( Figure 2 ). In the control reactor, the CO 2 fixation efficiencies in the early and late stages for DSM 505 were 9.24 and 1.59 mg C/L/day, respectively, and those for DSM 15147 were 6.95 and 0.86 mg C/L/day, respectively. Therefore, the apparent CO 2 fixation efficiencies of both strains in the control reactor decreased with time. Figure 2 Apparent carbon fixation yield and CO 2 fixation efficiency in membrane reactor and control reactor (A and B) Total apparent carbon fixation yield (columnar) and CO 2 fixation efficiency (scatter) of DSM 505 (A) and DSM 15147 (B) in membrane reactor and control reactor (Black error bars represent Standard Deviation.). To determine the main inhibitory factors affecting bacterial CO 2 fixation, the correlation coefficients between the apparent CO 2 fixation efficiency and the proportion of large molecules, the proportion of small molecules, the concentration of large molecules, the concentration of small molecules, or EFOC were calculated ( Figure 3 ). For both strains, there was a significant negative correlation (at the 0.05 level) between the proportion of small-molecule EFOC and the apparent CO 2 fixation efficiency, with a correlation coefficient of −0.999. However, the correlation coefficient between the proportion of large-molecule EFOC and the apparent CO 2 fixation efficiency in the control reactor was positive for both strains ( Figure 3 ). The apparent CO 2 fixation efficiency correlated negatively with the concentration of large-molecule EFOC and correlated significantly negatively with the concentration of small-molecule EFOC for DSM 15147. The negative correlation between the apparent CO 2 fixation efficiency and the three indices was not significant for DSM 505. These results imply that components of small-molecule EFOC are the main inhibitory factors affecting the apparent CO 2 fixation in bacteria. Figure 3 The correlation coefficients of different indexes in control reactor The correlation coefficients of different indexes of DSM 505 (A) and DSM 15147 (B) in control reactor at the early, middle, and late stages (∗. Correlation between apparent CO 2 fixation efficiency and another index is significant at the 0.05 level.). In the membrane reactor, the total apparent carbon fixation yields of DSM 505 and DSM 15147 were 235.19 mg C and 219.14 mg C, respectively, which were 3.72 times and 3.24 times that in the control reactor, respectively ( Figure 2 ). The CO 2 fixation efficiencies of DSM 505 in the early and late stages of culture in the membrane reactor were 23.04 and 4.16 mg C/L/day, respectively, and those of DSM 15147 were 16.51 and 5.05 mg C/L/day, respectively ( Figure 2 ). Therefore, the two typical SOB strains achieved higher CO 2 fixation in the membrane reactor than in the control reactor within the same operation time. The proportion of small-molecule EFOC in the membrane reactor was 0% during the whole operation period for DSM 505 and decreased to 55.67% in the late stage for DSM 15147 ( Figure 1 C). The proportion of small molecules in the filtered EFOC in the membrane reactor decreased to 0% and remained there for DSM 505, and decreased in the late stage of culture for DSM 15147 ( Figure 1 D). This differs from the upward trend of the proportion of small-molecule EFOC in the control reactor and shows that the membrane module changed the MW distribution of the EFOC components in the membrane reactor, and especially reduced the proportion of small-molecule EFOC in the late stage of culture relative to that in the control reactor. Overall, two typical bacteria with different growth characteristics and different MW compositions of EFOC showed some similar results. The proportion of small-molecule EFOC in the control reactor increased with time, with a corresponding downward trend in CO 2 fixation efficiency. The negative correlation between CO 2 fixation efficiency and the proportion of small-molecule EFOC in the control reactor was significant for both strains. The negative correlations between CO 2 fixation efficiency and the concentrations of small molecules, large molecules, and EFOC were not all significant for the two strains. However, the correlation coefficient between CO 2 fixation efficiency and the proportion of large-molecule EFOC in the control reactor was positive for both strains. This indicates that the components of large-molecule EFOC were not the main inhibitory factors of CO 2 fixation. Moreover, the membrane module separated out the EFOC and improved the bacterial CO 2 fixation by changing the MW distribution of the EFOC components in the reactor, especially by reducing the proportion of small-molecule EFOC. These observations indicate that the small molecules in EFOC are the main component inhibiting bacterial CO 2 fixation. The proportion of small-molecule EFOC from DSM 15147 was always >30%, whereas that from DSM 505 was always <6% ( Figure 1 ), which may be related to the different capacities for cytoskeleton synthesis of the two strains. During bacterial autotrophic growth, the small-molecule organic carbon produced via the CO 2 fixation pathway is then used for protein synthesis and the formation of cellular components. 26 The rRNA gene copy numbers of DSM 15147 and DSM 505 are five and seven, respectively, 39 which suggests that the protein synthesis rate of DSM 505 is higher than that of DSM 15147. Therefore, compared with DSM 505, the small molecules synthesized by DSM 15147 are used to synthesize proteins at a lower rate, so more small molecules remain as FOC and EFOC. Furthermore, in the early stage of culture, the proportion of large-molecule EFOC from DSM 505 is high because the cytoskeleton synthesis process is faster than the generation of small molecules, rather than because the bacteria have entered the death phase. Possible inhibitory components of small-molecule extracellular free organic carbon To further clarify the inhibitory components of small-molecule EFOC, the molecular and compositional diversity of the small-molecule EFOC generated by DSM 15147, whose proportion of small-molecule was higher than that of DSM 505, in the four stages of culture were determined with liquid chromatography-mass spectrometry. The composition of small-molecule EFOC and the elemental oxygen-to-carbon and hydrogen-to-carbon ratios were in similar ranges in the early stage (day 2) and late stage (day 8) ( Figures 4 A and 4B). The molecular formulae (MF) of the components of small-molecule EFOC changed only slightly during culture, whereas there was a change in the relative intensities (circle size) of the corresponding mass peaks ( Figures 4 A and 4B). This indicates that the proportions and concentrations of the compound classes in small-molecule EFOC changed during the chemoautotrophic culture. Figure 4 Molecular and compositional properties of small-molecule EFOC (A and B) Molecular formulas (MF) of small-molecule EFOC of DSM 15147 on days 2 (A) and 8 (B) of cultivation with compound classes (colors) and relative intensity (circle size) of the corresponding mass peaks; Data are displayed as molecular hydrogen-to-carbon versus oxygen-to-carbon ratio. (C and D) Proportion of molecule compound classes (C), peak area of molecule compound classes (D) of small-molecule EFOC generated by DSM 15147 in the control reactor in four chemoautotrophic cultivation stages (days 2, 4, 6, and 8). The molecular formulae could be divided into several compound classes according to the results of mass spectrometry. Most (53%-60%) of the molecules in small-molecule EFOC consisted solely of carbon, hydrogen, and nitrogen (CHN), and their proportion decreased with time, whereas the molecules that consisted solely of carbon, hydrogen, oxygen, and nitrogen (CHNO) or carbon, hydrogen, and oxygen (CHO) showed an uptrend in the proportion ranges of 27%-34% and 7%-14%, respectively ( Figure 4 C). This indicates that these molecular compound classes are the main compound classes of small-molecule EFOC. The peak area of the CHN molecules was highest on day 4 and decreased thereafter ( Figure 4 D), whereas the peak areas of the CHNO and CHO classes tended to increase with culture time ( Figure 4 D). The proportions and peak areas of other molecules were very small. The proportions and peak areas of the CHNO and CHO molecules correlated negatively with the CO 2 fixation efficiency, whereas those of the CHN molecules did not correlate negatively with it at all, indicating that CHNO and CHO molecules might be the main inhibitory compound classes in small-molecule EFOC. The CHNO molecules mainly included organic acids, nucleosides, carnitine, and amino acids, and the CHO molecules included fatty acids and amino acids. To clarify the specific compound types that constitute the inhibitory components of EFOC, small-molecular EFOC was divided into six compound types (organic acids, amino acids, nucleosides, fatty acids, lipids, and carnitine), according to the results of typical chromatograms and mass spectrometry ( Figure 5 A), with unrecognizable compound types were not considered. In the four culture stages, amino acids accounted for the largest proportion of these compounds (>53%) and increased with time, followed by organic acids (>23%) ( Figure 5 B). The proportions of nucleosides, fatty acids, and carnitine during the culture period were distributed in the ranges of 3%-11%, 1%-3%, and 0.9%-1%, respectively, and the proportion of lipids was zero in the first three stages ( Figure 5 B). This demonstrates that organic acids and amino acids are the main compounds in small-molecule EFOC. Figure 5 Compound type characterization of small-molecule EFOC (A–C) Typical chromatogram (A), proportion of compound types (B), and peak area of compound types (C) of small-molecule EFOC generated by DSM 15147 in the control reactor in four chemoautotrophic cultivation stages (days 2, 4, 6 and 8). There were also upward trends in the peak areas of both organic acids and amino acids in the small-molecule EFOC during chemoautotrophic culture ( Figure 5 C), and these trends correlated negatively with CO 2 fixation efficiency. The peak area of nucleosides showed an increasing trend at low concentrations, whereas those of the other compound types were almost zero and changed little over time ( Figure 5 C). These results indicate that organic acids and amino acids, as the main compounds in small-molecule EFOC, may be related to the inhibitory effect of EFOC on CO 2 fixation during chemoautotrophic culture. It has been reported that organic acids are toxic to the acidophile Thiobacillus ferrooxidans , because they affect its chemical permeability parameters. 40 A substrate inhibitory effect of organic acids on Thiobacillus acidophilus has been observed in the presence of pyruvate. 41 The addition of exogenous l -amino acids in inorganic culture environments disturbed the normal regulation processes and inhibited the growth of three obligately chemolithotrophic thiobacilli. 25 Moreover, the growth of nitrifying bacteria (nitrosomonas and nitrifying bacteria) is usually inhibited by relatively low concentrations of amino compounds. 42 Exogenous organic acids and amino acids inhibit the growth of chemoautotrophic bacteria. Our experimental results show that organic acids and amino acids are the main compounds in small-molecule EFOC and that small-molecule EFOC is the main fraction of EFOC inhibiting CO 2 fixation. It can be inferred that organic acids and amino acids are the main components of small-molecule EFOC that inhibit CO 2 fixation during chemoautotrophic culture. Verification of the inhibitory effects of self-generated organic acids and amino acids on CO 2 fixation by sulfur-oxidizing bacteria DSM 505 produced a smaller proportion of small-molecule EFOC than DSM 15147, so the use of DSM 505 to verify whether self-generated small molecule organic acids and amino acids inhibit CO 2 fixation could demonstrate both the inhibitory effect of self-generated small molecule organic acids and amino acids and the applicability of the inhibition to different bacteria. The effects of organic acids, amino acids, and their typical metabolites (e.g., carbohydrate, protein) on CO 2 fixation by DSM 505 were studied. Notably, the TOC fixed in the presence of pyruvate, oxaloacetate, fructose, and with no supplementation (control) was −2.47, −1.37, 2.94, and 1.82 mg C/L, respectively, in the early stage of culture ( Figure 6 A). This indicates that organic acids are utilized by bacteria for cytoskeleton synthesis and that the immediate inhibitory effect of organic acids on CO 2 fixation is greater than that of carbohydrates. Furthermore, DSM 505 utilizes organic acids at a higher rate than it utilizes fructose. Figure 6 Fixed TOC under conditions of organic acids, amino acids, and their typical metabolites (A and B) The fixed TOC by DSM 505 under the control group and three conditions of pyruvate, oxaloacetate, and fructose at the concentration of 10 mg C/L (A), and under three conditions of alanine, aspartic acid and protein at the concentration of 10 mg C/L (B) (∗. Difference is significant at the 0.05 level between the control group and the experimental groups separately supplemented with pyruvate, oxaloacetate, fructose, alanine, aspartic acid, and protein. No asterisk means the difference between the control group and the experimental group is not significant. Black error bars represent Standard Deviation.). The TOC fixed in the middle and late stages of culture in the presence of pyruvate or oxaloacetate was still lower than that fixed in the presence of fructose or in the control ( Figure 6 A). This demonstrates that the inhibitory effect of organic acids on CO 2 fixation is greater than that of carbohydrates in all stages of culture, indicating that organic acids are one of the main inhibitory components of EFOC. The effects of amino acids (alanine or aspartic acid) and their typical metabolites (protein) on CO 2 fixation by DSM 505 were investigated, as shown in Figure 6 B. The TOC fixed in the early stage of culture in the presence of alanine, aspartic acid, protein, and in the control was 0.21, 0.03, 0.99, and 1.82 mg C/L, respectively ( Figure 6 B). Simultaneously, the peak areas and proportions of amino acids in small-molecule EFOC in the presence of alanine or aspartic acid were greater than in the presence of protein ( Figures 7 A and 7B). This shows that amino acids and protein inhibited CO 2 fixation relative to that in the control group, and the immediate inhibitory effect of amino acids on CO 2 fixation was more obvious than that of protein. Figure 7 Amino acids in small-molecule EFOC under amino acids and protein conditions (A and B) Amino acids peak area (A) and proportion (B) of small-molecule EFOC generated by DSM 505 under alanine, aspartic acid, and protein conditions (Black error bars represent Standard Deviation.). In the middle stage of culture, the TOC fixed in the presence of alanine, aspartic acid, protein, and in the control was 6.53, 6.33, 4.26, and 8.87 mg C/L, respectively ( Figure 6 B), and the amino acid peak areas and proportions of small-molecule EFOC in the presence of protein increased and exceeded that in the presence of alanine or aspartic acid ( Figure 7 B). This indicates that amino acids and protein inhibit CO 2 fixation in the middle stage compared with the control group, and the inhibitory effect of amino acids on carbon fixation in the middle stage was weaker than that in the early stage, while the inhibitory effect of proteins on carbon fixation in the middle stage was stronger than that in the early stage. This may be related to the decomposition of proteins into amino acids, which has stronger inhibition effect on carbon fixation, in the middle stage. In the late stage of culture, the TOC fixed in the presence of alanine, aspartic acid, protein, and in the control was 8.58, 7.05, 8.45, and 12.04 mg C/L, respectively ( Figure 6 B). The amino acid peak areas and proportion of small-molecule EFOC in the presence of amino acids were slightly smaller than in the presence of protein in the late stage ( Figures 7 A and 7B). These results indicate that the long-term effects of amino acids and protein on CO 2 fixation are similar, because of the metabolic balance between protein and amino acids. Thus, the immediate inhibitory effect of amino acids on CO 2 fixation was greater than that of protein in the early stage of culture, whereas the inhibitory effect of protein was greater in the middle stage. However, the long-term effects of both on CO 2 fixation were similar, which is related to the dynamic balance between protein hydrolysis into amino acids and the synthesis of protein from amino acids. Since bacteria produce amino acids in EFOC continuously, the inhibitory effect of amino acids is always maintained, so amino acids are one of the main inhibitory components of EFOC. The total cbb gene transcription is an important factor influencing bacterial CO 2 fixation by chemoautotrophic bacteria. 12 , 13 , 14 , 15 However, previous studies have mainly focused on the cytotoxicity of organic acids and amino acids to bacteria, 25 , 40 , 41 , 42 and few studies have examined the effects of organic acids and amino acids on the cbb gene transcription. Therefore, the efficiency of the cbb gene transcription in DSM 505 in the middle stage of culture in the presence of different typical metabolites was analyzed to investigate the effects of organic acids and amino acids on cbb gene transcription. The total cbb gene transcription efficiency in the middle stage of culture in the presence of pyruvate, oxaloacetate, or fructose was lower than in the control group, and total cbb gene transcription efficiency in the presence of pyruvate or oxaloacetate was lower than in the presence of fructose ( Figure 8 A). These data show that pyruvate and oxaloacetate inhibit cbb gene transcription more efficiently than fructose. Figure 8 Total cbb gene transcription efficiency under conditions of organic acids, amino acids, and their typical metabolites (A and B) Total cbb gene transcription efficiency of DSM 505 under pyruvate, oxaloacetate, fructose, and control group conditions (A) and under alanine, aspartic acid, protein, and control group conditions (B) in the middle stage and the schematic diagram of inhibition mechanism (∗. Difference is significant at the 0.05 level between the control group and the experimental groups separately supplemented with pyruvate, oxaloacetate, fructose, alanine, aspartic acid, and protein. No asterisk means the difference between the control group and the experimental group is not significant. Black error bars represent Standard Deviation.). The total cbb gene transcription efficiency in the presence of alanine, aspartic acid, or protein was lower than in the control group ( Figure 8 B); and was lower in the presence of protein than in the presence of alanine, although in the presence of protein, it was similar to that in the presence of aspartic acid ( Figure 8 B). These results indicate that the inhibitory intensity of protein on cbb transcription is similar to that of amino acids in the middle stage of culture and even stronger than that of some amino acids, which may be due to the increase in the amino acid peak area and the proportion of amino acids in small-molecule EFOC in the presence of protein. In summary, organic acids exert a stronger inhibitory effect on CO 2 fixation than carbohydrates by inhibiting the cbb gene transcription and by acting as substrates for cytoskeleton synthesis. The inhibitory effects of amino acids on CO 2 fixation are stronger than that of protein in the early stage of culture. However, when protein is decomposed to amino acids in the middle stage of culture, the increases in the concentration and proportion of amino acids in small-molecule EFOC increase the inhibitory effect on CO 2 fixation by inhibiting the cbb gene transcription. The similarity of the long-term inhibitory effects of amino acids and proteins on carbon fixation is related to the metabolic balance maintained between amino acids and proteins. Technical perspectives on the removal of organic small molecules to improve CO 2 fixation by chemoautotrophic bacteria Therefore, to enhance bacterial CO 2 fixation, we recommend the removal of the small molecule EFOC, especially amino acids and organic acids, during their chemoautotrophic culture. The membrane filtration technology is an feasible option to remove EFOC for continuous cultivation of chemoautotrophic bacteria, 26 but the problems of membrane fouling, high membrane pressure, and high energy consumption associated with membrane filtration technology need to be addressed. 43 , 44 The removal of dissolved organic small molecules by adsorption is low-energy and efficient. As a promising adsorption material with nanoscale pore size and designed functional sites, covalent organic frameworks (COFs) may carry out specific intra-pore adsorption for small molecules, 45 , 46 , 47 , 48 , 49 while cells and macromolecules are not adsorbed by COFs, thus isolating small molecules and cells and eliminating feedback inhibition of small molecules. When nanopore adsorption-culture technology with its simple operation is used for batch culture of chemoautotrophic bacteria, there are advantages that the adsorbent can be regenerated and reused, and the adsorbed small-molecule amino acids and organic acids can be utilized, thus making it a cost-controllable technology. Nanopore adsorption-culture technology for the adsorption of small-molecule EFOC can enhance CO 2 fixation by chemoautotrophic bacteria, which broadens the application of adsorbent materials and enables the reuse of fixed organic carbon products. Conclusions When EFOC is analyzed in terms of the molecular weight of its components, the proportion and concentration of the small molecules in EFOC increase during chemoautotrophic culture, and the small molecules are the main components of EFOC that inhibit CO 2 fixation by chemoautotrophic bacteria. In terms of the types of compounds in small-molecule EFOC, amino acids and organic acids are the main inhibitory components. The inhibitory effects of amino acids and organic acids on CO 2 fixation are stronger than those of protein and carbohydrate, respectively, because they also inhibit the transcription of the cbb gene. Therefore, nanopore adsorption-culture technology for the adsorption of amino acids and organic acids is suggested to enhance CO 2 fixation by chemoautotrophic bacteria. Limitations of the study The prospect of nanopore adsorption-culture technology for enhancing CO 2 fixation by chemoautotrophic bacteria needs further experimental testing."
} | 8,348 |
33478163 | PMC7835966 | pmc | 2,746 | {
"abstract": "Networks in the human brain are extremely complex and sophisticated. The abstract model of the human brain has been used in software development, specifically in artificial intelligence. Despite the remarkable outcomes achieved using artificial intelligence, the approach consumes a huge amount of computational resources. A possible solution to this issue is the development of processing circuits that physically resemble an artificial brain, which can offer low-energy loss and high-speed processing. This study demonstrated the synaptic functions of conductive polymer wires linking arbitrary electrodes in solution. By controlling the conductance of the wires, synaptic functions such as long-term potentiation and short-term plasticity were achieved, which are similar to the manner in which a synapse changes the strength of its connections. This novel organic artificial synapse can be used to construct information-processing circuits by wiring from scratch and learning efficiently in response to external stimuli.",
"conclusion": "5. Conclusions This study facilitated conductance changes between two electrodes linked by conductive polymer PEDOT:PSS wires via the continuous application of voltage pulses to the electrode gap, thereby achieving synaptic functions including LTP and STP. The learning efficiency and retention time were dependent on the frequency and magnitude of the input voltage pulses, thus the system may be regarded as an artificial synapse capable of learning efficiently via selective storage of input information. The polymer wire synapse can be used to construct an information-processing network circuit from scratch and learn efficiently in response to external stimuli. This shows promise for future applications in highly integrated brain-type hardware to reproduce the structure and learning mechanism of the human brain.",
"introduction": "1. Introduction The human brain was first described as a complex network of innumerable neurons connected via synapses in the early 20th century. This discovery inspired various mathematical models for information processing, including models by McCulloch and Pitts [ 1 ], which have contributed towards the development of artificial intelligence (AI) technology using conventional von Neumann-type computers. However, the application of AI software leads to issues regarding energy loss and execution processing time due to the structure of these conventional computers, especially as the network models become larger and more complex [ 2 ]. The execution of advanced and highly efficient information processing similar to that performed by the human brain involves complicated and elaborate three-dimensional networks, where a new processor should be developed to replace the conventional computer. New processers based on the concept of neuromorphic hardware have emerged in recent years, where several studies have aimed to reproduce the functions of brain synapses by skillfully utilizing the unique electrical properties of various materials [ 3 , 4 ]. Specifically, synaptic devices have been developed using poly(3,4-ethylenedioxy-thiophene) doped with poly(styrene sulfonate) anions (PEDOT:PSS) [ 5 , 6 ], which is a widely-used conductive polymer that offers high conductivity [ 7 ], thermoelectric conversion [ 8 , 9 ], and chemical sensitivity with bio-adaptability [ 10 , 11 ]. The PEDOT cation and the dopant PSS anion are electrostatically bonded, where PSS pull electrons from the PEDOT chain and inject positive carriers ( Figure 1 a). This property leads to high conductivity. PEDOT:PSS has a hierarchical structure [ 12 ], where the state of each structure has a significant influence on the overall morphology and electrical conductivity. PEDOT:PSS has been previously grown into the shape of a wire during electrodeposition under a bipolar square-wave alternating current (AC) voltage between two electrodes immersed in a precursor solution of 3,4-ethylenedioxythiophene (EDOT) monomer and PSS [ 13 , 14 , 15 , 16 ]. The resulting PEDOT:PSS wire was used to construct information-processing circuits [ 17 ]. The electropolymerization of PEDOT proceeds on the surface of the anode immersed in the precursor solution, where PSS becomes incorporated as a dopant to deposit PEDOT: PSS with a higher-ordered structure. The morphology of the resulting PEDOT:PSS can be controlled by adjusting the growth voltage conditions [ 18 , 19 ]. For example, either dendritic or wire-shaped PEDOT:PSS may be produced, as shown in Figure 1 b,c, respectively (experimental details given in Figure S1 ). In particular, wire-shaped PEDOT:PSS offers excellent wiring properties due to elongation of the polymer wires along the electric field during anisotropic polymerization. This allows for the formation of an electric circuit network with the PEDOT:PSS wire between arbitrary electrodes ( Figure 1 d). The appearance of polymer wires growing and linking between electrodes is similar to the ~100 billion neurons that form independently in a newborn human brain, which comprises axons extending to other neurons to form a neural circuit network [ 20 ]. Therefore, polymer wires show promise for easily and inexpensively producing information-processing circuits with three-dimensional and spatial integration similar to an actual human brain. A previous study by the current authors demonstrated an increase in conductance value between electrodes due to linking of conductive polymer wires to achieve synaptic long-term potentiation (LTP) [ 21 , 22 ], which was applied in a synaptic device [ 17 ]. General electronic synaptic devices reproduce the change in synaptic strength by adjusting the resistance [ 23 , 24 , 25 ]. However, PEDOT:PSS is a stable material, and an effective way to selectively depolymerize the wires after linking has not yet been reported. In other words, previous reports on our synaptic device were able to increase the conductance value between the electrodes, but not decrease it. Therefore, a network constructed by wiring is unable to learn flexibly according to the surrounding environment after network formation. This study aimed to demonstrate the use of bridging wires to control the conductance between two electrodes, thereby emulating the way in which a real synapse changes its strength of connection with another neuron. Conductance modification was achieved by controlling the diameter of the bridging wires and reproducing a synaptic LTP function. Synaptic short-term plasticity (STP) describes a temporary change in synaptic connection strength [ 26 , 27 ], which was also achieved by controlling the conductivity of the bridging wires. This STP characteristic allowed for a temporary reduction in the conductance between the electrodes, which has not yet been reported. The resulting artificial polymer wire synapses show promise for application in highly-integrated information-processing circuits, as they can be formed from scratch, select input information to memorize, and efficiently implement learning.",
"discussion": "4. Discussion LTP expression was affected by the interval of voltage pulse application, as the interval had an effect on the diameter distribution of the polymer wires and the conductivity of the newly-polymerized polymer film on the wire surface. These differences were attributed to the difference in the PSS anion distribution of the solution according to the voltage pulse frequency. The frequent application of voltage pulses led to increased attraction of the PSS anions towards the anode along the line of electric force, which led to a high density of PSS anions near the anode. This facilitated the electrodeposition of PEDOT:PSS with a higher-order structure, where highly conductive PEDOT:PSS containing abundant dopant PSS molecules was deposited on the wire surface near the anode. Further, the insufficient PSS anion density further from the anode led to minimal electrodeposition. This localized increase in wire diameter during electrodeposition led to an asymmetric wire shape and increased conductance, similar to LTP. Less-frequent voltage pulsing allowed for longer diffusion of the migrated PSS anions between the applied pulses, leading to more uniform and less-dense distribution throughout the wire. Consequently, a low-conductivity polymer containing a small amount of dopant was deposited on the entire wire surface, resulting in a low conductance between the electrodes. In the STP expression experiment, the asymmetric wire shape and reversible doping reaction were the driving principles of the doping/dedoping reaction given in Equation (1): (1) P E D O T + : P S S − + e − ⇋ P E D O T 0 + P S S − The application of a voltage to the electrode gap linked by polymer wires led to simultaneous doping near the anode and dedoping near the cathode. These reactions would cancel one another in a typical two-terminal system, thereby minimizing the variation in conductance along the entire wire before and after applying voltage. However, an asymmetric diameter distribution along the longitudinal direction of the wire led to disproportionate doping efficiency at the anode side and dedoping efficiency at the cathode side due to the difference in the wire surface area around the anode and cathode. Consequently, the total number of carriers varied along the wire, leading to a change in conductance before and after applying voltage. The changes in conductance caused by the doping and dedoping reactions decayed and converged to the same value within a period of a few tens of seconds to a few minutes. This may have been attributed to the uniform concentration gradient achieved due to PSS anion diffusion over time. The application of voltage to the electrode gap immersed in the PSS solution generated electrostatic forces that caused PSS anions to aggregate near the anode, which enhanced the doping reaction. When the voltage was halted, the PSS anions that had not engaged in the doping reaction diffused readily, while the PSS anions engaged in doping dissociated more gradually. Consequently, the concentration gradient in the electrode gap became uniform over time. Considering the effect of Le Chatelier’s principle on Equation (1), the change in the concentration gradient of PSS anions led to a new equilibrium state of the wire near the electrodes. The reverse reaction was promoted due to diffusion, thereby returning the conductance to the initial value. The polymer wire synapse allowed for easy LTP expression changes according to the frequency of the applied voltage pulses, thereby reproducing the multistore model of human memory [ 28 ]. This property allows for the selection of frequently encountered input information for long-term memory, thereby facilitating efficient learning. In addition, the polymer wire synapse expressed STP when small voltage pulses were applied after LTP expression ( Figure S3 ). This allowed for rewriting after the formation of a long-term memory, demonstrating the potential for flexible learning according to the surrounding environment. Synaptic devices with STP characteristics have gained recent popularity, and can be applied to information-processing models that reflect past input information for a certain period of time, such as recurrent neural networks (RNNs) [ 29 ]. For example, the implementation of reservoir computing (RC) [ 30 ], which is a type of RNN model, has been previously demonstrated in hardware using memristors with resistance-changing properties, such as STP [ 31 ]."
} | 2,871 |
27242732 | PMC4863899 | pmc | 2,749 | {
"abstract": "Large fraction of mineral nutrients in natural soil environments is recycled from complex and heterogeneously distributed organic sources. These sources are explored by both roots and associated mycorrhizal fungi. However, the mechanisms behind the responses of arbuscular mycorrhizal (AM) hyphal networks to soil organic patches of different qualities remain little understood. Therefore, we conducted a multiple-choice experiment examining hyphal responses to different soil patches within the root-free zone by two AM fungal species ( Rhizophagus irregularis and Claroideoglomus claroideum ) associated with Medicago truncatula , a legume forming nitrogen-fixing root nodules. Hyphal colonization of the patches was assessed microscopically and by quantitative real-time PCR (qPCR) using AM taxon-specific markers, and the prokaryotic and fungal communities in the patches (pooled per organic amendment treatment) were profiled by 454-amplicon sequencing. Specific qPCR markers were then designed and used to quantify the abundance of prokaryotic taxa showing the strongest correlation with the pattern of AM hyphal proliferation in the organic patches as per the 454-sequencing. The hyphal density of both AM fungi increased due to nitrogen (N)-containing organic amendments (i.e., chitin, DNA, albumin, and clover biomass), while no responses as compared to the non-amended soil patch were recorded for cellulose, phytate, or inorganic phosphate amendments. Abundances of several prokaryotes, including Nitrosospira sp. (an ammonium oxidizer) and an unknown prokaryote with affiliation to Acanthamoeba endosymbiont, which were frequently recorded in the 454-sequencing profiles, correlated positively with the hyphal responses of R. irregularis to the soil amendments. Strong correlation between abundance of these two prokaryotes and the hyphal responses to organic soil amendments by both AM fungi was then confirmed by qPCR analyses using all individual replicate patch samples. Further research is warranted to ascertain the causality of these correlations and particularly which direct roles (if any) do these prokaryotes play in the observed AM hyphal responses to organic N amendment, organic N utilization by the AM fungus and its (N-unlimited) host plant. Further, possible trophic dependencies between the different players in the AM hyphosphere needs to be elucidated upon decomposing the organic N sources.",
"conclusion": "Conclusion We found a very strong evidence of AM hyphal proliferation in organic N but not in organic or inorganic P patches, in spite of the fact that our host plant ( M. truncatula ) was not N limited (as inferred from similar experiments using the same soil, AM fungi and plant genotype as here, e.g., Konvalinková et al., 2015 ). Thus, AM hyphal proliferation seems to be caused primarily by the N requirements of the AM fungus rather than the host plant. We consider our results to be particularly robust because we used two independent approaches to quantify the development of AM fungi in soil, namely the traditional microscopy and the qPCR. To our knowledge, this is the first study showing a fair correlation between these two independent methods quantifying the development of AM fungal hyphae in soil. This correlation is particularly important inasmuch as there was a significant stimulation of other soil fungi’s development in certain soil patches and microscopic quantification of the hyphal development could potentially be biased by our inability to discriminate between AM fungal and other hyphae. Using two independent approaches (454-sequencing of pooled soil samples per soil amendment treatment and qPCR using all individual soil samples), we also demonstrated that the hyphal developmental responses of both AM fungi to soil amendments strongly correlated with the abundance of Nitrosospira sp., an ammonium oxidizer, and few other bacterial taxa in the soil including an obligate Acanthamoeba endosymbiont. This last evidence, though only correlative, strongly suggests a possible involvement of soil protists such as Acanthamoeba sp. in mediating the stimulation of AM hyphal development by and possibly hyphal uptake of N from the soil organic-N amendments, possibly through so called microbial loop ( Bonkowski, 2004 ). Both the amoebas and the ammonia oxidizers could be part of such a pathway – with the ammonia oxidizers processing either the NH 4 + ions released directly from the organic matter (scenario 1) or the ammonia accumulating in the soil solution as a by-product of amoebas grazing on soil bacteria (scenario 2). The latter appears much more plausible scenario than the previously suggested primary involvement of ammonia oxidizers in oxidation of ammonia directly released from the organic matter ( Cheng et al., 2012 ). This earlier work established the scenario 1 based on the observed repression of AM-induced organic matter mineralization by nitrification inhibitor dicyandiamine – silently assuming that the AM fungi would be generally very inefficient in taking up the free NH 4 + ions from the soil solution. Such an assertion is, however, not supported by most previous research (e.g., Govindarajulu et al., 2005 ). Either way, complex microbial interactions involved in organic matter mineralization and AM fungal uptake of N released by this process certainly warrant further investigations, addressing possible metabolic/trophic dependencies between the different members of soil microbial communities, as well as consequences of these interactions for the N nutrition of the AM fungus and its host plant. More dedicated efforts with 15 N-labeled sources (be they atmospheric N 2 , soil organic soil amendments, or mineral N fertilizers), specific non-mycorrhizal and non-microbial controls and intensive sampling schemes covering temporal dynamics of microbial communities will be required to quantify precisely the contributions of the various microbial pathways to and involvement of the different players in the plant and AM fungal nutrition.",
"introduction": "Introduction The physical arrangement of soil particles, aggregates, and pores; root and microbial growth; and burrowing activities of animals, together with external inputs of such particulate organic residues as plant litter, animal excreta, and dead bodies, all create a soil environment highly heterogeneous in both space and time and at a range of scales ( Facelli and Facelli, 2002 ; Watt et al., 2006 ). Adaptations of roots to heterogeneously distributed organic and inorganic nutrients in soil have been studied for decades ( Robson et al., 1992 ; Robinson, 1996 ; Hodge, 2004 ). For most plants, however, the root is not the only – and possibly not even the main – organ for primary acquisition of such poorly mobile nutrients as phosphorus (P) from the soil solution. This function is often fulfilled by the plants’ mycorrhizal symbionts. Arbuscular mycorrhizal (AM) fungi, belonging to the Glomeromycota, establish intimate relationships with more than half of currently described plant species ( van der Heijden et al., 2015 ). The nutrients taken up by AM fungal hyphae from the soil solution are then passed on to the host plants at the root–mycorrhizal interface in the root cortex ( Fitter, 1991 ; Smith et al., 2004 ). The importance of AM symbiosis for P acquisition by many plant species is firmly established ( Cox and Tinker, 1976 ; Jakobsen et al., 1992 ), whereas its role in plant nitrogen (N) acquisition, although repeatedly demonstrated ( Mäder et al., 2000 ; Hodge et al., 2001 ; Fellbaum et al., 2012 ), is broadly accepted as being lower than that in plant P acquisition ( Johnson et al., 2015 ). How AM hyphal networks respond to heterogeneously distributed soil resources and what consequences this has for their own as well as for host plant nutrition is much less understood than are root responses. Previously, we and others have shown that at least some AM fungal species establish denser hyphal networks in root-free patches as compared to the rooting zones ( Jansa et al., 2003 ; Thonar et al., 2011 ; Zheng et al., 2015 ) and that root responses to heterogeneously distributed soil nutrients could be negated through the establishment of AM symbiosis ( Felderer et al., 2013 ). This could be caused by positive hyphal developmental response to patches with greater availability of mineral nutrients ( Li et al., 1991 ; Zheng et al., 2015 ), to specific N forms within the patches ( Bago et al., 2004 ), or to variation in such other soil physicochemical properties as clay or organic matter contents ( Jansa et al., 2003 ). Indeed, significant research efforts have been dedicated in the past to deciphering the response of AM fungal networks to soil organic matter. It has previously been shown that extracted soil organic matter, dried plant biomass, dried baker’s yeast, and bovine serum albumin specifically stimulated the development of AM fungal networks in root-free patches and that starch and pure cellulose have depressed it ( St. John et al., 1983 ; Joner and Jakobsen, 1995 ; Ravnskov et al., 1999 ; Gavito and Olsson, 2003 ; Leigh et al., 2009 ). Other experiments, while not specifically distinguishing between root and root-free zones, have recorded significant stimulation of the development of AM extraradical hyphae, root colonization, and sporulation by the addition of crab-shell chitin to the rhizosphere ( Gryndler et al., 2003 ). However, the large biological variation in mycorrhizal experiments and the use of different soils, model plants, and fungal species in the diverse studies testing the various organic amendments one by one preclude direct comparisons as to the effects exerted by different compounds. Surprisingly, to the best of our knowledge, a multidimensional model used previously to examine preferences of AM hyphal networks to colonize spatially discrete patches having different mineral fertilizer amendments in a monoxenic cultivation system ( Bago et al., 2004 ) as well as in planted or unplanted soil patches ( Gavito and Olsson, 2008 ) has not yet been employed to study the effect of different qualities of soil organic patches within the reach of a single AM fungal colony. Since the host plant in most previous experiments has been N-limited (e.g., Leigh et al., 2009 ), AM-induced mineralization of organic N in those experiments might have been driven by the high N demand of both the host plant and the AM fungi. What would be the choice of AM fungi for certain organic patches if the host plant was not N-, but P-limited, has not been sufficiently addressed as yet. It had previously been postulated that at least part of the response of AM fungal hyphae to soil organic amendments is caused and/or modulated by other soil microbes because the saprotrophic potential of AM fungi is thought to be low ( Joner et al., 2000 ; Gryndler et al., 2003 ; Leigh et al., 2009 ). It could be that those microbes live solely upon (derive their energy from) the organic materials, release mineral nutrients (N and/or P) from them, and produce other (secondary) metabolites. In such case, the nutrients or the other metabolites could then be involved in the AM hyphal response. On the other hand, the microbes might also associate directly with the AM hyphal surfaces ( Toljander et al., 2006 ; Jansa and Gryndler, 2010 ), partially or fully derive their carbon/energy from the mycorrhizal hyphae ( Toljander et al., 2007 ; Drigo et al., 2010 ), and carry on degradation of the soil organic materials in order to release the mineral nutrients either for themselves or possibly for their fungal hosts ( Jansa et al., 2013 ). Because knowledge is so fragmented as to the responses of AM fungal hyphae to different soil organic amendments containing or not containing such mineral nutrients as P and/or N, we carried out a multiple-choice experiment whereby the development of hyphal networks was quantified within a number of spatially discrete patches buried in the root-free zone of each pot. The goal was to directly compare the responses of a single AM hyphal network to different soil organic amendments, containing either N or P or both or none of these nutrients. Using the available high-throughput sequencing technology and quantitative real-time PCR (qPCR) we aimed at identification of a possible common denominator of the AM fungal response to the amendments within the soil microbe (prokaryotic and fungal) communities. Because the host plant ( Medicago truncatula ) requirements for N were likely saturated via atmospheric N fixation and because it was grown under P-limiting conditions, we hypothesized (based on the resource limitation theory, Johnson et al., 2015 ) that the AM hyphae would preferentially colonize the P-containing organic materials due to higher demand for P than N by the host plant. This hypothesis is in line with some earlier observations that AM hyphae could contribute a large share of plant P uptake from organic P sources accessible only to the hyphae ( Tarafdar and Marschner, 1994 ; Feng et al., 2003 ; Zhang et al., 2014 ).",
"discussion": "Discussion Here we show, using a multiple-choice experiment, specific and consistent stimulation of AM hyphal development in soil patches amended with N-containing organic compounds, whereas other amendments (particularly the phytate and orthophosphate, but also cellulose) caused no localized hyphal response in the same fungal colony. These results lead to rejection of the original hypothesis predicting preferential colonization of P-containing patches due to high P demand of the N-unlimited host. Although the observed absence of AM hyphal growth stimulation by cellulose just confirms previous observations ( Ravnskov et al., 1999 ; Gryndler et al., 2002 ), no obvious effect of P amendments on the AM hyphae is surprising. It is namely inconsistent with previous studies demonstrating stimulation of AM hyphal growth by phytate and by low levels of inorganic orthophosphate ( Feng et al., 2004 ; Cavagnaro et al., 2005 ; Zheng et al., 2015 ). Thus, our results deserve detailed analysis and possibly further experimentation to find definitive answers about the nature of AM hyphal growth stimulation and mineral nutrient acquisition from root-free patches of different qualities. One possible explanation for the lack of AM hyphal growth stimulation by the P-containing amendments could be that the AM fungal P uptake efficiency from the soil solution is partly or totally uncoupled from the hyphal development. However, based on previous research (e.g., Jansa et al., 2005 ), this is highly unlikely, although we admit that a direct radiophosphorus labeling of the P amendments and including non-mycorrhizal controls would be required to provide a definitive answer to this issue. The most intriguing question is thus why both the AM fungi so clearly and consistently proliferated in the organic N patches in our experiment. Although, half of the stimulatory soil amendments (i.e., DNA and clover biomass) also contained P ( Table 1 ), P alone is unlikely to have caused the observed AM hyphal stimulation. This is because the amounts of DNA or clover biomass used in creating the stimulatory patches contained the same or smaller amounts of added P than did the phytate or inorganic orthophosphate patches, which in turn caused no significant stimulation of AM hyphal development in soil (see above). Given our experimental setup and the (N-fixing) model plant, the requirements of the host plant for soil-derived N should be rather low ( Sulieman et al., 2013 ) compared to previous studies with non-leguminous hosts such as Plantago (e.g., Hodge et al., 2001 ; Gryndler et al., 2003 ). On the other hand, some earlier studies (e.g., Ravnskov et al., 1999 ) using clover (another leguminous and N-fixing plant) as a host also demonstrated stimulation of the AM hyphal networks by localized organic N amendments. Together, these results strongly suggest that it is primarily the AM fungal requirement for N rather than plant N demand, which drives the observed AM hyphal proliferation in organic N patches. This notion is not inconsistent with the previous research clearly demonstrating transfer of the N (but never the C) from the organic matter via AM hyphae to the associated mycorrhizal (and N-limited) host plant ( Hodge et al., 2000 , 2001 ; Hodge and Fitter, 2010 ; Herman et al., 2012 ). However, our results call for a more myco-centric view (e.g., Alberton et al., 2005 ) on the AM symbiosis, where the requirements of both partners and possible competition between the partners for limited resources such as mineral N is considered ( Hodge and Storer, 2015 ). One particularly interesting issue in this regard is whether N-unlimited host would transfer any N to its associated (and N-limited) AM fungus – to the best of our knowledge there is no experimental evidence for this as yet. Although, we did use neither direct 15 N labeling of the organic amendments nor we included non-mycorrhizal controls into our experiment, our 15 N abundance results indicate that the N taken up from soil by Rhizophagus resides mainly in the fungal biomass and is not transferred to the plant host. This is because the 15 N signature of the Rhizophagus- colonized roots was (across both harvests) higher than that of the roots colonized by Claroideoglomus (p = 0.043, see Supplementary Figure S8 for data). On the other hand, the 15 N signature of the shoots of Rhizophagus -colonized plants was much lower than that of the Claroideoglomus- inoculated plants ( p < 0.001, Supplementary Figure S8 ). The latter was most likely due to markedly improved P nutrition of the Rhizophagus -colonized plants having positive feedback on the highly P-dependent symbiotic N fixation ( Scheublin et al., 2004 ). Because positive values usually indicate N taken from the soil and/or fertilizer pool, lower values indicate N derived from biological N fixation ( George et al., 1993 ). Rhizophagus colonized the roots more heavily than did Claroideoglomus in our experiment, and it also provided greater growth and nutritional benefits to its hosts. Thus, our results indirectly indicate greater biological N fixation of the Rhizophagus -inoculated plants, whereas most of the N taken up by the fungus is likely not transferred to the plant tissues. Because the saprotrophic potential of the AM fungi is thought to be low and cannot alone explain the degradation of soil organic amendments ( Joner et al., 2000 ; Koide and Kabir, 2000 ; Leigh et al., 2011 ), the AM hyphae most likely benefit from/depend on the degradatory activities of such other soil microbes as bacteria ( Beier and Bertilsson, 2013 ) to access nutrients such as N in organic forms. Once in the soil solution, the mineral N ions can be directly taken up by AM fungal hyphae ( Bago et al., 1996 ; Govindarajulu et al., 2005 ; Cruz et al., 2007 ). There are always plenty of microbes both in the soil and on the surfaces of AM hyphae, and their communities can also actively be shaped by AM fungal hyphae ( Gryndler et al., 2003 ; Toljander et al., 2006 , 2007 ; Drigo et al., 2010 ; Jansa et al., 2013 ; Nuccio et al., 2013 ). Thus, the possible feedbacks between proliferation of AM hyphae, soil mineral N forms, and the soil microflora in the enriched soil patches need particular attention here. Previously, it has been shown that addition of buffered ammonia could have triggered hyphal branching of R. irregularis in absence of any bacteria in a monoxenic cultivation system ( Bago et al., 2004 ). Open pot experiments (e.g., Gavito and Olsson, 2008 ), could not replicate the same phenomenon, however, and so it remains unclear whether the previous results could be generalized for Claroideoglomus due to substantial functional differences in the hyphal growth traits between the two AM fungal genera ( Thonar et al., 2011 ). Inasmuch as the stimulatory patches in our experiment were all enriched with organic N sources (with the exception of clover biomass, where a small portion of N could have been present as nitrate or ammonia already from the very beginning), release of mineral N forms would have required exoenzyme activity, thus implicating soil saprotrophs as being involved in the AM hyphal response to soil patches. This would be consistent with previous findings showing substantial stimulation of AM hyphal proliferation by some soil bacteria ( Gryndler et al., 2000 ), although other studies have also reported antagonistic interactions between AM fungi and soil bacteria (e.g., Leigh et al., 2011 ). In our study, the dominant OTUs of the 454-sequencing profiles of both bacterial and fungal communities in the different soil patches did not consistently explain the AM hyphal responses to the patches but rather mirrored the different patch qualities in an idiosyncratic manner (see Supplementary Figures S5 and S6 ). On the other hand, several bacterial taxa with lower abundance in the 454-sequencing profiles showed significant positive correlation with the Rhizophagus hyphal responses to the soil amendments, and this was also confirmed by the qPCR analyses including all individual soil samples ( Figure 4 ). First, strong correlation of the hyphal response with some specific Nitrosospira sp. abundance across the different soil amendment types indicates the possibility of a causal (nutritional or signaling) relationship. This is because this bacterial genus has the capacity to oxidize ions of ammonium to nitrite in the first step of nitrification and often dominates this ecosystem function ( Webster et al., 2005 ). It may be important to note that OTUs representing other microbes involved in nitrification (e.g., Nitrobacter, Nitrospira, Nitrosovibrio , as well as other Nitrosospira spp.-like OTUs; Supplementary Table S4 ) did not correlate with AM hyphal response. Second, the correlation with the Acanthamoeba endosymbiont may indicate an involvement of protozoa in the stimulation of AM hyphal growth within the soil patches. Protozoa mobilize N from bacterial biomass and have recently been shown to increase the rates of N translocation from organic fertilizers to plants via the mycorrhizal pathway ( Koller et al., 2013 ). However, the correlation, no matter how strong, does not constitute causal proof, which would need to be obtained by means of specifically focused experiments yet to come. These future experiments should also enable testing as to whether the interactions of AM fungi with protozoa and/or Nitrosospira and possibly other prokaryotes involved in nitrification indeed center upon the N nutrition of the AM fungus or the plant, or whether it could also involve secondary metabolites or one or more of the other degradation products resulting from the enzymatic hydrolysis of the organic amendments. Relatively low AM hyphal densities in the soil samples collected from the different compartments in this study as compared to other, similar studies (e.g., Thonar et al., 2011 ) could possibly be explained by the fact that we used granular zeolite (expanded clay) as a substantial part (45% volume) of our potting substrate. This material is porous and there is evidence that the AM fungi would spread their hyphae inside the pores and cavities of the granules ( Baltruschat, 1987 ; Feldmann and Idczak, 1992 ). Any hyphae inside the granules would not be accessible to mechanical hyphae extraction and microscopy. However, inasmuch as our DNA-based quantification correlates well with the patterns of AM hyphal spread in the different compartments observed microscopically ( Figure 2 ) and the DNA should extract all fungal tissues, both inside and outside the zeolite granules, we are confident as to the validity of the hyphal developmental responses reported in this study."
} | 5,973 |
30242284 | PMC6154969 | pmc | 2,750 | {
"abstract": "Self-organization is the generation of order out of local interactions. It is deeply connected to many fields of science from physics, chemistry to biology, all based on physical interactions. The emergence of collective animal behavior is the result of self-organization processes as well, though they involve abstract interactions arising from sensory inputs, information processing, storage, and feedback. Resulting collective behaviors are found, for example, in crowds of people, flocks of birds, and swarms of bacteria. Here we introduce interactions between active microparticles which are based on the information about other particle positions. A real-time feedback of multiple active particle positions is the information source for the propulsion direction of these particles. The emerging structures require continuous information flows. They reveal frustrated geometries due to confinement to two dimensions and internal dynamical degrees of freedom that are reminiscent of physically bound systems, though they exist only as nonequilibrium structures.",
"introduction": "Introduction Active particles serve as simple microscopic model systems for living objects such as birds, fish, or people and mimic in particular the propulsion of bacteria or cells without the complexity of physical properties and chemical networks in living objects 1 . They consume energy to propel persistently and as such they have given considerable insight into collective behaviors of active materials already 2 – 6 . With their bare function of self-propulsion they are, however, missing the important ingredients of sensing and feedback, which most living objects from cells up to whole organisms have in common. All of their living relatives have signaling inputs which they use to gain information about the environment. Employing external information, organisms interact such that birds and fish are able to self-organize into flocks or schools 7 – 9 and, on a microscopic level, cells may regulate gene expression 10 , 11 . Using sensing, information processing and feedback, living systems may go beyond what is prescribed by physical interactions such as the Coulomb, van der Waals or hydrophobic interactions. For cells/bacteria the information about cellular density (quorum sensing) is, for example, inferred from the concentration of signaling molecules released by the cells leading to a regulation of various physiological activities such as biofilm formation 10 . Concentration gradients of nutrients may serve as sensory input leading to a directed motion termed chemotaxis 12 . For birds, the information used for the formation of flocks is suggested to be the visual perception of the number of neighboring birds 8 . In this respect all information that is processed by the organism is linked to a physical representation 13 (e.g. concentration, number of objects, orientation,…), but the resulting structure and dynamics is disconnected from a direct physical interaction. In the particular examples mentioned above, though, the structure formation depends on the active response of the organism and its ability to steer based on its recognition of the environment 7 , 14 – 16 . While active particles do not have sensory inputs, information processing units and feedback mechanisms built in yet, suitable control mechanisms may introduce this complexity fostering the exploration of new emergent phenomena. An information exchange between active particles has not been tackled so far, but seems to be a natural step towards extending their functionality. Like bacteria, active particles have to break the time symmetry of low Reynolds number hydrodynamics in order to propel. They have to provide asymmetries to generate directed motion. In many cases this asymmetry is built into the structure in form of two hemispheres with different chemical or physical properties, so called Janus particles 1 , 17 . As the propulsion direction is bound to the symmetry axis of the particle, rotational diffusion becomes a relevant process that limits the control of Janus particles 18 , 19 . The self-propulsion mechanism presented below relies on a scheme for generating self-thermophoresis 20 ; unlike past approaches our new scheme utilizes the spatially controlled asymmetric input and release of energy around a symmetric particle. This scheme delivers a precise control of each individual particle in a larger ensemble, and allows us to demonstrate how active particles may form structures just by the exchange of information on other particle positions using a feedback control mechanism for steering the particles.",
"discussion": "Discussion The presented symmetric active particles and the introduced manipulation technique allow us to self-assemble structures by designed feedback controlled interaction rules. Similar structures of passive colloidal particles have been assembled in previous work with the help of external forces in optical tweezers setups 27 , 28 . The structures there are commonly based on a prescribed optical potential energy landscape in which the particles occupy energetic minima and fluctuate according to Boltzmann statistics in equilibrium. The assemblies created in our experiments are conceptually different as there is no external force acting on the particles 17 and more importantly, the structures do only exist in nonequilibrium. The particles need to be propelled continuously to form the assemblies much like living systems need nonequilibrium to maintain their shape and function. The fluctuations of the structures are nonequilibrium even for a vanishing feedback delay and thus they are not tied to a well-defined temperature or Boltzmann statistics. Moreover, as demonstrated for the oscillatory motion, non-zero delay in our setup leads to additional oscillations/fluctuations determined by the propulsion speed and feedback delay. In artificial active particle systems several sources of structure formation and local density enhancement may exist. Active particles are observed, for example, to cluster based on chemical, composition, or temperature gradients mimicking tactic behavior 34 . In the case of chemotaxis of artificial microswimmers the propulsion speed is modulated as direct physical consequence of the local concentration of a substance. In fact, a modulation of the propulsion speed will lead to local enhancement or depletion of active particle concentrations 35 in steady states as it is observed also in motility induced phase transitions with a mutual locking of particles 3 , 36 . With the feedback rule applied in our experiments, we now remove also the modulation of the speed as particles are constantly propelled with the same speed. Yet there is a structure forming contribution resulting from the feedback loop. It stems from the fact that we determine the particle positions in a measurement and decide on the required action, i.e., the velocity direction. It is this particular processing of the information of the particle positions which extracts entropy from the system to form the structures. Analyzing the entropy production in the system, we can identify several contributions (see Supplementary Fig. 6 ). On one side, there is the entropy production rate maintaining the temperature gradients. This “housekeeping” entropy production amounts to the fraction of the incident laser power absorbed by the gold nanoparticles P a divided by the temperature of the surrounding liquid T to which this power is dissipated. It is required for the mobility although only a fraction of it is used during propulsion 37 . This entropy flux is still not sufficient for the structure formation. The propulsion speed and the dissipated power are constant over the trajectory of the particle no matter if it is driven on a random path or is bound in the active particle molecule. Consequently, it is not a spatial modulation of this dissipation which is causing the structure formation. The feedback process is extracting entropy from the system by steering the propulsion direction and not modulating their speed. A theoretical analysis of the entropy fluxes can be obtained by relating the feedback process to the thermodynamics of resetting processes as described by Fuchs et al. 38 (see Supplementary Note 4 ). According to that, it is the continuous loss of structure due to Brownian motion which one has to correct for with the feedback. To form a stable stationary structure, the entropy extracted per time unit in the feedback loop has to compensate at least the increase of entropy per time unit due to Brownian motion. The structure formation is thus the result of the information flow in the feedback loop only. Overall, the structure forming entropy production rate is very small as compared to all other entropy fluxes (see Supplementary Note 4 ) but sufficient to create well-defined active particle assemblies. Returning to the previous example of chemotaxis, one may argue that the response of an artificial active particle to a chemical gradient may be seen as an information based interaction as well even if it is the direct physical consequence of osmotic pressure differences. The local chemical gradient is the information that is available to the particle and causes a response in form of a modulated propulsion speed. While this interpretation of physical interactions may also be valid, we restrict our definition of information based interactions to situations, where the physical interactions are irrelevant. In the well-known thought experiment of Maxwell, for example, a small daemon uses the information on the speed of particles to sort them into two boxes (slow and fast) just by actuating a shutter between the boxes. The physical interactions of the particles are irrelevant there for the appearing temperature difference between the two boxes. A correct physical description not violating the laws of thermodynamics, however, requires to include the entropy of information the daemon uses. In a very similar way, it is the information flow in the feedback loop that is required for a correct physical description of the assembled active particle molecules in our experiment. In conclusion, propulsion speed, Brownian motion as well as the feedback related information flow shape the morphology and dynamics of these artificial self-organized active structures. They require nonequilibrium conditions to exist and contrary to structures assembled by optical tweezers they are not bound to Boltzmann statistics, equipartition or global detailed balance. Nevertheless, the dynamics of the structures carries a number of features, which are found in equilibrium as well. Using the described method, almost any type of interaction can be designed to create large scale interacting assemblies or new self-organized shapes which may not be accessible by conventional interactions, i.e., which do not need to obey the action–reaction principle. Fundamental interaction and signaling rules for emergent complex behavior of cells up to animals may be explored with our model system in the same way as concepts of nonequilibrium thermodynamics. The applied technique provides a direct interface to machine learning algorithms including predictive information or reinforcement learning. The details of information flows in large ensembles may be studied easily and can be connected to different timescales of delayed information processing. Especially the latter type of application including coupled active feedback networks with different inherent timescales shall ignite a vast variety of research on emergent collective and crowd dynamics."
} | 2,903 |
22173357 | null | s2 | 2,752 | {
"abstract": "The function of microbial interactions is to enable microorganisms to survive by establishing a homeostasis between microbial neighbors and local environments. A microorganism can respond to environmental stimuli using metabolic exchange-the transfer of molecular factors, including small molecules and proteins. Microbial interactions not only influence the survival of the microbes but also have roles in morphological and developmental processes of the organisms themselves and their neighbors. This, in turn, shapes the entire habitat of these organisms. Here we highlight our current understanding of metabolic exchange as well as the emergence of new technologies that are allowing us to eavesdrop on microbial conversations comprising dozens to hundreds of secreted metabolites that control the behavior, survival and differentiation of members of the community. The goal of the rapidly advancing field studying multifactorial metabolic exchange is to devise a microbial 'Rosetta stone' in order to understand the language by which microbial interactions are negotiated and, ultimately, to control the outcome of these conversations."
} | 285 |
28585934 | PMC5563964 | pmc | 2,753 | {
"abstract": "Methane-oxidizing bacteria represent a major biological sink for methane and are thus Earth’s natural protection against this potent greenhouse gas. Here we show that in two stratified freshwater lakes a substantial part of upward-diffusing methane was oxidized by filamentous gamma-proteobacteria related to Crenothrix polyspora . These filamentous bacteria have been known as contaminants of drinking water supplies since 1870, but their role in the environmental methane removal has remained unclear. While oxidizing methane, these organisms were assigned an ‘unusual’ methane monooxygenase (MMO), which was only distantly related to ‘classical’ MMO of gamma-proteobacterial methanotrophs. We now correct this assignment and show that Crenothrix encode a typical gamma-proteobacterial PmoA. Stable isotope labeling in combination swith single-cell imaging mass spectrometry revealed methane-dependent growth of the lacustrine Crenothrix with oxygen as well as under oxygen-deficient conditions. Crenothrix genomes encoded pathways for the respiration of oxygen as well as for the reduction of nitrate to N 2 O. The observed abundance and planktonic growth of Crenothrix suggest that these methanotrophs can act as a relevant biological sink for methane in stratified lakes and should be considered in the context of environmental removal of methane.",
"conclusion": "Conclusions Members of the genus Crenothrix are rare methane oxidizers, which are not available in pure or enrichment cultures and will not be readily picked up in environmental samples by the currently available specific FISH probe (Creno445). The ambiguity surrounding their pmoA has further complicated the in situ detection using molecular methods. In the past, this has hampered our understanding of these peculiar organisms and possibly led us to underestimate their role in the biogeochemical nutrient and element cycles. In our study, we could unambiguously demonstrate a key role for these organisms in the mitigation of methane emissions from two stratified lakes. In Lake Rotsee, Crenothrix even contributed more to methane uptake than the ‘classical’ unicellular gamma-MOB. In up to 3 consecutive years Crenothrix was recurrently found throughout the stratification period of Lake Rotsee and Lake Zug, and thus appears to be a stable part of the indigenous microbial community. Our data are also the first to demonstrate that Crenothrix is capable of growing as a planktonic species in the lake water column. Given the capacity of Crenothrix to rapidly grow up into large biomass, its participation in methane cycling also in other relevant habitats should be considered.",
"introduction": "Introduction Freshwater lakes represent large natural sources of methane and contribute more to methane emissions than the oceans despite their comparably smaller area ( Bastviken et al. , 2004 ). Highest rates of methane removal are usually measured at the oxyclines, either in the water column or in the sediment. Lake Rotsee and Lake Zug in Central Switzerland are typical examples of temperate lake systems with methane fluxes across the oxycline of 13±3 mmol and 10±3 mmol m −2 d −1 , respectively ( Oswald et al. , 2015 , 2016 ). Both lakes are stratified, with methane-rich hypolimnia, but whereas the shallow Lake Rotsee overturns annually, the deep Lake Zug remains stratified throughout the year. In both lakes, the vast majority of the upward-diffusing methane is removed at the base of the oxycline at in situ oxygen concentrations in the low micromolar range ( Oswald et al. , 2015 , 2016 ). Methane oxidation at the oxycline was shown to be coupled to the reduction of residual or in situ -produced oxygen, but there were also indications for methane-oxidizing activity under oxygen-deficient conditions ( Oswald et al. , 2015 , 2016 ). Abundant gamma-proteobacterial methane-oxidizing bacteria (gamma-MOB) were shown to be involved in methane removal in both lakes ( Oswald et al. , 2015 , 2016 ). Gamma-MOB are considered aerobes requiring oxygen for methane activation, even though some cultured representatives can perform methane oxidation under denitrifying conditions ( Kits et al. , 2015a , 2015b ). Environmentally relevant representatives of gamma-MOB in lakes and other freshwater habitats belong to the ‘classical’ genera of Methylobacter , Methylomonas , Methylosarcina and Methylomicrobium ( Boschker et al. , 1998 ; Bodelier et al. , 2013 ; Oshkin et al. , 2015 ), and all possess particulate methane monooxygenase (pMMO) as the key methane-oxidizing enzyme ( Bowman, 2005 ). In Lake Rotsee and Lake Zug, unicellular gamma-MOB represented a stable community at the oxycline. The bacteria showed rapid growth on methane as evidenced by the increase in cell abundances and the uptake of 13 C-methane into their biomass ( Oswald et al. , 2015 , 2016 ). In these studies, gamma-MOB were identified by fluorescence in situ hybridization using the 16S rRNA-targeted oligonucleotide probes Mgamma84+705. Interestingly however, these probes do not bind to members of a potentially important subgroup of gamma-proteobacterial MOB, the putative family Crenothrichaceae . Contrary to ‘classical’ MOB, these gamma-MOB are multicellular and filamentous. So far, only two of these bacteria have been documented in literature, Crenothrix polyspora and Clonothrix fusca , and both were retrieved from groundwater ( Stoecker et al. , 2006 ; Vigliotta et al. , 2007 ). Sporadically, environmental occurrence of Crenothrix is reported in literature based on retrieved 16S rRNA or pmoA sequences ( Dörr et al. , 2010 ; Drewniak et al. , 2012 ), but its role in methane cycling has remained unclear. The metabolism of Crenothrix has been a matter of debate since its first description as ‘Brunnenfaden’ (‘a well thread’ Cohn, 1870 ). Initially, Crenothrix/Clonothrix filaments were considered to belong to the ‘iron bacteria’ due to the presence of metal particles in their sheaths ( Roze, 1896 ; Jackson, 1902 ; Molisch, 1910 ). This belief was challenged by studies that failed to observe iron encrustation in Crenothrix/Clonothrix filaments ( Kolk, 1938 ; Wolfe, 1960 ), and the later discovery of membrane invaginations has prompted suggestions for a methanotrophic lifestyle ( Völker et al. , 1977 ). Eventually, the capacity to oxidize methane was experimentally confirmed on filaments retrieved from man-made habitats ( Stoecker et al. , 2006 ; Vigliotta et al. , 2007 ). Interestingly, C. polyspora was reported to possess an ‘unusual’ pMMO, which was only distantly related to ‘classical’ MMO of gamma-proteobacterial methanotrophs ( Stoecker et al. , 2006 ), and has now been recognized to cluster together with the ammonium monooxygenases of completely nitrifying ‘comammox’ bacteria ( Daims et al. , 2015 ; van Kessel et al., 2015 ). Here we investigated the occurrence and involvement of these filamentous bacteria in methane oxidation at and below the oxyclines of Lake Rotsee and Lake Zug. We performed stable isotope labeling experiments followed by single-cell imaging to explore the role of these microorganisms in environmental methane cycling, and metagenomic analyses to investigate their metabolic potential with respect to aerobic and anaerobic respiration. For comparison, we also performed metagenomic analysis of a sample from Wolfenbüttel waterworks sand filter reportedly containing high proportions of C. polyspora .",
"discussion": "Results and discussion Crenothrix in Lake Rotsee and Lake Zug To investigate the potential occurrence of filamentous Crenothrix bacteria in two stratified lakes and their involvement in the lacustrine methane cycle, we first recorded geochemical evidence for methane oxidation in situ . Concentration profiles recorded in Lake Rotsee and Lake Zug over the course of 3 years suggested a zone of methane consumption that persistently coincided with the oxycline (profiles from Lake Rotsee 2013 are shown in Oswald et al. (2015) , from 2014 in Supplementary Figure 1 ; profiles from Lake Zug 2012, 2013 and 2014 are shown in Oswald et al. , 2016 ). Concurrently, incubations with 13 CH 4 confirmed high rates of methane oxidation at the oxycline ( Oswald et al. , 2015 , 2016 ; Supplementary Figure 1 ). These incubations were set up under both oxic and anoxic conditions. In Lake Rotsee, oxic incubation conditions were obtained either by addition of air or solely by incubation of anoxic water in the light. In the latter case, aerobic methane oxidation was presumably sustained by oxygenic photosynthesis ( Milucka et al. , 2015 ; Oswald et al. , 2015 ). In Lake Zug, oxic incubations were solely supplemented with air and incubated in the dark. These different incubation set ups reflected the different nature of the two lakes, Lake Rotsee has a shallow, sun-lit oxycline, whereas the oxycline of Lake Zug is very deep and dark. Additionally, anoxic Lake Zug incubations supplemented with nitrate were also set up as Lake Zug had the appropriate environment to test for methane-dependent denitrification ( Supplementary Table 4 ). We then analyzed the microbial community at the Lake Rotsee oxycline by 16S rRNA gene amplicon sequencing in 2 consecutive years (2013 and 2014; Supplementary Figure 2 ). Along with gamma-proteobacterial Methylococcaceae ( Methylobacter , Methylocaldum , Methylomonas and Methyloglobulus species), CABC2E06 (an uncultured Methylococcales clone; Wang et al. , 2012 ; Quaiser et al. , 2014 ), and the marine methylotrophic group, also sequences belonging to Crenothrix were retrieved. On the basis of the number of recovered sequences, Crenothrix -related organisms were 2–5-fold less abundant than Methylococcaceae and comprised 0.06–0.1% of the total bacterial sequences in situ . However, it is possible that the true abundance of Crenothrix in situ was higher than what the 16S rRNA gene abundances suggest, as, for example, DNA extraction biases might strongly select against these thickly sheathed microorganisms. We could additionally confirm the presence of Crenothrix in both lakes by CARD-FISH with two oligonucleotide probes reported to target Crenothrix , Mgamma669 and Creno445 ( Eller et al. , 2001 ; Stoecker et al. , 2006 ). The more specific oligonucleotide probe Creno445 bound only sporadically, when the hybridization stringency was strongly reduced ( Supplementary Figure 3 ). On the other hand, the Mgamma669 probe hybridized most of the conspicuous filaments in all analyzed samples from both lakes ( in situ water as well as incubations, Figure 1 ; Supplementary Figure 3 ) even though some filaments did not hybridize even with this more general probe (for example, Supplementary Figures 3a, b ). With both probes, we observed two hybridized cell morphotypes—filaments and single round cells ( Figure 1 ; Supplementary Figure 3 ). Both morphotypes have been observed for Crenothrix spp. previously and it has been proposed that the smaller round cells represent reproductive cells that bud from the ends of vegetative cell filaments ( Cohn, 1870 ; Völker et al. , 1977 ). However, given the compromised specificity of the Creno445 probe at low stringency and the broad specificity of the Mgamma669 probe, it is also possible that the hybridized single cells represented other gamma-MOB, reportedly targeted by the Mgamma669 probe (for example, Methylobacter or Methylomonas ; Eller et al. , 2001 ). Therefore, the here-reported Crenothrix cell counts and biovolumes are solely based on counts of Creno445- or Mgamma669-hybridized filaments and thus represent conservative estimates. Overall, in all analyzed incubations from both lakes total Crenothix biovolumes increased over time ( Supplementary Figure 4b ). This confirms that Crenothrix was growing under both oxic and anoxic conditions. Whereas unicellular gamma-MOB had consistently cell sizes of ca. 2 μm, the individual cells in Crenothrix -like filaments reached an average length of ca. 5 μm ( Figure 1 ; Supplementary Figures 3a and 5a ). The average length and width of Lake Rotsee Crenothrix filaments was ca. 45 and ca. 1.5 μm, respectively, with individual filaments reaching >100 μm length ( Supplementary Figure 3 ). Filaments were often intertwined and bunched together, as observed previously ( Cohn, 1870 ; Völker et al. , 1977 ). In Lake Rotsee, the biovolume of Crenothrix was about eight-fold higher than that of unicellular gamma-MOB at depths corresponding to the highest observed methane oxidation rates (in 2012 and 2013; Supplementary Figure 4a ). Only in 2014 unicellular gamma-MOB biomass contribution was higher than that of Crenothrix ( Supplementary Figure 4 ). We speculate that these differences might be connected to the complex life cycle of Crenothrix ( Supplementary Discussion ). In Lake Zug, the filaments were shorter but more consistent in terms of length, reaching an average length and width of ca. 28 and 1.4 μm (in 2013) and ca. 20 and ca. 1.4 μm (in 2014), respectively. Methanotrophic growth of Crenothrix To confirm that the observed cell growth (that is, increase in cell numbers and biovolume over time; Supplementary Figure 4b ) was methane-derived, samples from the 13 CH 4 -supplemented incubations were further analyzed by nanoSIMS. Filamentous bacteria hybridized with the Mgamma669 probe consistently constituted the highest 13 C-enriched population in all three investigated incubations (Lake Rotsee oxic, Lake Zug oxic and Lake Zug anoxic; Figure 1 ; Supplementary Figure 5 ). The 13 C enrichment confirmed that 13 CH 4 was assimilated into cell biomass, such as is common for gamma-proteobacterial methanotrophs ( Trotsenko and Murrell, 2008 ). In some of the images, fragmentation of filaments into single vegetative cells was apparent, even though the uptake of 13 C appeared homogenously spread throughout the whole filament. In both lakes, Crenothrix filaments appeared to be colonized by other non-identified bacteria, which did not show comparably strong enrichment in 13 C and might thus represent heterotrophic epibionts ( Figure 1 ). In contrast, the single round cells (hybridized with Mgamma669 probe) were similarly enriched in 13 C as the Crenothrix filaments ( Figures 1a–c ), supporting the speculation that these cells belong to methanotrophic bacteria and might potentially represent reproductive Crenothrix cells. In the Lake Rotsee oxic incubation, the uptake of methane-derived carbon by Crenothrix filaments was comparable to that of ‘classical’ unicellular gamma-MOB ( 13 C enrichment of 22±4.8 at % and 29±4.1 at %, respectively; Table 1 ; Figure 1 ; Supplementary Figure 2 ). However, due to its larger biovolume Crenothrix assimilated ca. 4–6-fold more methane than the ‘classical’ gamma-MOB in the same incubation (1.73 or 1.18 μmol methane l −1 d −1 and 0.27 μmol methane l −1 d −1 , respectively; Table 1 ). These numbers are based on average filament biovolumes and cell counts determined by CARD-FISH at the beginning of the incubation and do not take into account any increase in cell numbers over time, as the incubation conditions might have differently affected the growth of the different MOB. However, even if we take into account the increase of cell numbers over time, overall contribution of Crenothrix to methane uptake in Lake Rotsee was still higher than that of the unicellular gamma-MOB, even though the difference was not so pronounced (ca. 1.4 higher based on T end cell counts). Crenothrix filaments in Lake Zug oxic incubations were also active and assimilated methane at rates of ca. 0.04 μmol methane l −1 d −1 ( Table 1 ; Figures 1d–f ). This is much lower than the overall contribution of Crenothrix in Lake Rotsee, which is largely due to their lower abundance (1.1E+03 cells per ml) and smaller average biovolume (ca. 30 μm 3 ). Additionally, Crenothrix was also active in our anoxic denitrifying incubations where not enough oxygen was present to account for measured methane oxidation rates (2.7 μmol l −1 d −1 13 CO 2 produced in 15 NO 3 -supplemented incubation, a ca. 10-fold increase compared to control incubation without any added electron acceptor (0.234 μmol l −1 d −1 13 CO 2 produced)). The methane-dependent growth under oxygen-deficient conditions was evidenced as cell biomass enrichment in both 13 C (from 13 C-CH 4 ; Figures 1g–i ) and 15 N (from 15 N-nitrate; Supplementary Figure 6 ), even though the methane uptake rates were somewhat lower (0.03 μmol methane l −1 d −1 ) than those in incubations supplemented with oxygen ( Table 1 ). Metagenomic analyses of Lake Rotsee and Lake Zug Due to the strong dominance of eukaryotic sequences in Lake Rotsee, we were not able to assemble a genomic bin of Crenothrix from any of the sequenced samples ( Supplementary Table 1a ). On the other hand, in the Lake Zug metagenomes eukaryotic sequences were almost completely absent and the relative abundance of Crenothrix -related sequences was considerably higher ( Supplementary Table 1a ). Therefore, a metagenome from a Lake Zug anoxic incubation (sample Z3, Supplementary Table 4 ) was used for the assembly of a Crenothrix genome. The Crenothrix D3 draft genome was binned by exploiting the differential coverage of contigs in metagenomes obtained from the in situ metagenome of Lake Zug and two different incubations (an oxygen-supplemented and an anoxic, nitrate-supplemented; Supplementary Figure 7a ; see also Materials and Methods section and Supplementary Table 4 for sample details). We retrieved several bins representing gamma-MOB from the Lake Zug assembly (data not shown). The metagenomic sequences within these two bins were also present in our Lake Rotsee metagenomes, as indicated by their respective coverage ( Supplementary Figure 7b ). 16S rRNA gene retrieved from one of these bins putatively belonged to a Methylobacter ( Figure 2a ). The other bin contained a partial 16S rRNA gene (909 bp) that clustered closely with C. polyspora ( Figure 2a ), even though the level of similarity (95% identity) suggests that the Lake Zug Crenothrix is a different species. Most closely related environmental sequences were retrieved from groundwater and habitats highlighted primarily for iron richness ( Bruun et al. , 2010 ), yet apparently containing methane ( Kojima et al. , 2009 ; Kato et al. , 2013 ). Retrieval of the Crenothrix D3 16S rRNA gene sequence from the Lake Zug metagenome allowed us to also investigate the reasons behind the poor performance of the Creno445 probe. The comparison of the probe binding region on the 16S rRNA gene sequence revealed that the Creno445 FISH probe (length: 18 nt) had five mismatches with the partial 16S rRNA gene from our metagenomic Crenothrix D3 bin ( Supplementary Table 2 ). Interestingly, out of 47 16S rRNA gene sequences in the SILVA database (NR99, release 123) that were assigned to Crenothrix/Crenothrichaceae, only seven sequences (including four C. polyspora sequences published by Stoecker et al. (2006) ) contained less than five mismatches. Thus it seems that while the Creno445 probe is very specific to C. polyspora , it might not be suitable for environmental detection of other Crenothrix strains and species. In comparison, the lacustrine Crenothrix 16S rRNA gene had only a single mismatch with the Mgamma669 probe, which explains the comparably better performance of this (not Crenothrix -specific) probe on our samples. Interestingly, the clade CABC2E06, which forms an apparent sister group to Crenothrix based on the 16S rRNA tree ( Figure 2a ), had an identical number of mismatches to both probes. As the 16S rRNA gene sequences assigned to this group were retrieved from both Lake Rotsee ( Supplementary Figure 2 ) and Lake Zug (data not shown), it is feasible that the CABC2E06 bacteria in these samples were also hybridized by the Mgamma669 probe. Additionally, if the CABC2E06 bacteria were filamentous, they may have been included in the here-reported cell and biovolume counts. Genome-inferred C1 metabolism of lacustrine Crenothrix D3 In the Crenothrix D3 draft genome from Lake Zug ( Supplementary Table 1b ), we searched for pMMO genes. We found all genes encoding for pMMO, which were organized in the arrangement pmoCAB , such as is common for gamma-proteobacterial type I MOB ( Trotsenko and Murrell, 2008 ). The phylogenetic analysis of the PmoA amino-acid sequence showed that the sequence fell within the PmoA group of other known gamma-MOB, including the PmoA sequence of the other described filamentous methane oxidizer, C. fusca ( Figure 2b ). However, the presence of conventional gamma-proteobacterial pmoA in the lacustrine Crenothrix strain was inconsistent with the findings of ‘unusual’ pmoA previously reported for C. polyspora based on PCR and quantitative PCR ( Stoecker et al. , 2006 ). Our Crenothrix D3 draft genome did not contain any ‘unusual’ pmoA; in fact, no ‘unusual’ pmoA or amoA has been retrieved in any of the other gamma-MOB-assigned bins either. We thus decided to address this discrepancy by obtaining metagenomic data from the original samples used in the Stoecker et al. (2006) study. Two samples obtained in 2004 from the rapid sand filters of the Wolfenbüttel waterworks (Germany) were analyzed and, after differential coverage binning, genomic information of two Crenothrix strains was obtained ( Supplementary Figure 8 ). A partial 16S rRNA sequence retrieved from one sand filter Crenothrix bin (bin 1; 817 bp) was 98% identical to the C. polyspora 16S rRNA sequence. As the sample reportedly contained high proportions of C. polyspora , it is feasible that (at least one of) the sand filter Crenothrix was in fact C. polyspora . However, throughout this manuscript we refer to these organisms as sand filter Crenothrix , without a species name. The sand filter and the lacustrine Crenothrix likely represented different species as indicated by the average sequence identities of their shared genes ( Supplementary Discussion ). Both genomes of the sand filter Crenothrix species contained a pmoCAB operon (gene similarities between both bins 96–99%) and a pxmABC operon (gene similarities between both bins 93–99%). PmoA encoded by the genes from the pmoCAB operon clustered together with other gamma-proteobacterial PmoA sequences ( Figure 2b ) and the affiliation of the pxmABC operon with the sequence-divergent pxm cluster was confirmed by a phylogenetic analysis of pxmA ( Tavormina et al. , 2011 ; Figure 2b ). PxmA has been suggested to play a role in methane oxidation under hypoxic and denitrifying conditions by Methylomonas denitrificans and Methylomicrobium album ( Kits et al. , 2015a , 2015b ). It thus appears that Crenothrix might be another denitrifying methanotroph containing both pmoCAB and pxmABC operons. Importantly, no ‘unusual’ pmoA could be detected in the sand filter Crenothrix bins. However, the ‘unusual’ pmoA sequence previously assigned to C. polyspora was detected in a different bin, clearly belonging to the completely nitrifying Nitrospira , apparently co-occurring with C. polyspora in the sample ( Daims et al. , 2015 ; van Kessel et al., 2015 ; Pinto et al. , 2016 ). This finding is discussed in more detail in the Supplementary Discussion . It is interesting to note that whereas all three Crenothrix PmoA sequences fell within the ‘classical’ gamma-proteobacterial PmoA branch, the lacustrine Crenothrix PmoA clustered separate from the sand filter Crenothrix bins 1 and 2 and C. fusca ( Figure 2b ). Comparison of the 16S rRNA gene and PmoA amino-acid trees suggested that the PmoA of the lacustrine Crenothrix might have been obtained laterally from another gamma-proteobacterial methanotroph. This is supported by the fact that transposase genes were located immediately up- and downstream of the lacustrine Crenothrix pmoCAB operon on the respective contig (data not shown). In addition to the gene cluster encoding for pMMO, we also retrieved a full gene cluster for soluble methane monooxygenase (sMMO; smmoXYBZDC ) in the lacustrine Crenothrix and in one sand filter Crenothrix bin. This enzyme is relatively rare in gamma-proteobacterial methanotrophs ( Murrell, 2010 ) and was not found in C. polyspora previously ( Stoecker et al. , 2006 ), presumably due to the mismatches between the applied PCR primers and the respective target regions in the mmoX gene. We cannot conclusively prove involvement of sMMO in methane oxidation by Crenothrix ; however, as the substrate range of sMMO seems much broader than that of pMMO ( Dalton, 2005 ; Semrau et al. , 2011 ), it is feasible that Crenothrix might have the capacity to utilize other C-compounds as suggested previously ( Stoecker et al. , 2006 ). This could explain the reported occurrence of Crenothrix in, for example, natural bitumen deposits ( Saidi-Mehrabad et al. , 2013 ). All three retrieved genomes (two sand filter Crenothrix genomes as well as the lacustrine Crenothrix D3 genome) further contained all necessary genes for complete oxidation of methane to CO 2 ( Supplementary Discussion ; Figure 3 ). Like many other type I methanotrophs ( Chistoserdova and Lidstrom, 2013 ), Crenothrix might use the RuMP pathway for C1 assimilation from formaldehyde, as genes for all necessary enzymes were found in all three draft genomes ( Figure 3 ). On the other hand, the serine cycle apparently missed genes encoding for hydroxypyruvate reductase and malate thiokinase. Crenothrix had the genomic potential for mixed acid fermentation to succinate and potentially acetate (gene encoding for phosphate acetyltransferase was missing in lacustrine Crenothrix D3 genome and one sand filter Crenothrix , but was putatively present in the other sand filter Crenothrix bin) and hydrogen production (via NAD-reducing hydrogenase, hoxFUYH ; only present in the lacustrine Crenothrix ). Pyruvate, which serves as the starting point for fermentation, could be generated from formaldehyde via enzymes of the RuMP and pyrophosphate-mediated glycolytic pathway that was encoded in all three Crenothrix genomes. Mixed acid fermentation and H 2 production via these pathways has been shown to be a major route of methane-derived carbon respiration in methanotrophs growing under oxygen limitation ( Kalyuzhnaya et al. , 2013 ). Aerobic and anaerobic respiration by Crenothrix In agreement with the demonstrated cell growth and activity in our oxic incubations, all three Crenothrix genomes encoded a multitude of aerobic respiratory chain complexes, such as a sodium-pumping NADH:ubiquinone oxidoreductase (Na + -NQR), the M and L subunits of the NADH:quinone oxidoreductase, the bc1 complex, an A1-type heme copper cytochrome c oxidase, a type B heme copper cytochrome c oxidase (only the sand filter Crenothrix ) and a cytochrome bd oxidase that might potentially act as a high-affinity terminal oxidase ( Figure 3 ). Additionally, the draft genome of the lacustrine Crenothrix D3 as well as one of the sand filter Crenothrix strains also contained a partial pathway for the respiration of nitrate. We retrieved genes encoding for a membrane-bound respiratory nitrate reductase ( narGHI ), a nitrite/nitrate antiporter ( narK ) as well as a periplasmic multi-copper nitrite reductase (nirK). Genes encoding for nitric oxide (NO) and nitrous oxide (N 2 O) reductases ( norBC and q-type nor , and nosZ , respectively) were not found in any of the three bins. Yet, interestingly, all three Crenothrix genomes encoded proteins for alternative pathways of NO detoxification to N 2 O. In the genome of Crenothrix D3, a gene cluster containing hcp and hcr genes was found. The hcp gene encodes for a unique hybrid cluster protein (Hcp), which has recently been shown to act as a high-affinity NO reductase in Escherichia coli , producing N 2 O as the end product ( Wang et al. , 2016 ). The Hcp sequence retrieved from the Crenothrix D3 genome contained the six highly conserved residues involved in 4Fe-2S-2O cluster coordination ( Aragao et al. , 2008 ) as well as a glutamic acid residue (E492 of E . coli Hcp) essential for NO reductase activity ( Wang et al. , 2016 ). Overall, the Crenothrix D3 Hcp shared 49% amino-acid identity with the NO-reducing Hcp of E. coli . The hcr gene, located immediately downstream from hcp , encodes for the Hcr protein and acts as a NADH-dependent Hcp reductase ( van den Berg et al., 2000 ), while simultaneously protecting Hcp from nitrosylation by its substrate, NO ( Wang et al. , 2016 ). The hcp/hcr genes in Crenothrix D3 genome were preceded by norR , a transcriptional regulator of three different enzymes (NO reductase, flavorubredoxin and flavohaemoglobin) that all utilize NO as a substrate ( Rodionov et al. , 2005 ). We thus speculate that, despite being routinely annotated as a hydroxylamine reductase, the Hcp/Hcr system in Crenothrix could in fact act as a NO reductase and substitute Nor-type NO reductases under denitrifying conditions. In the two sand filter Crenothrix genome bins no homologs of Hcp were found. However, both bins (but not the lacustrine Crenothrix genome bin) contained a homolog of cytochrome c '-beta, a member of the cytochrome P460 family found in, for example, gamma-proteobacterial methane oxidizers ( Zahn et al. , 1996 ; Campbell et al. , 2011 ) and gamma- and beta-proteobacterial ammonia oxidizers ( Bergmann and Hooper, 2003 ; Klotz et al., 2006 ). Cytochrome c '-beta can reduce NO to N 2 O ( Elmore et al. , 2007 ). Interestingly, in one of the bins this gene ( cytS ) was located directly downstream of the haoA and haoB genes encoding for hydroxylamine dehydrogenase. As both the Hcp and the cytochrome c ’-beta are predicted to be cytoplasmic proteins and NO is produced in the periplasm (by NirK), it is feasible that their activities are not coupled and (some) NO might escape out of the cell ( Figure 3 ). The experimentally demonstrated and genome analysis-supported metabolic potential for methane-dependent growth under nitrate-reducing conditions cannot serve as a final proof of nitrate reduction by Crenothrix in Lake Zug. However, it is interesting to speculate that such metabolic versatility might expand the habitat of these facultative anaerobic bacteria, potentially enabling them to survive periods of oxygen starvation by switching to using nitrate as an electron acceptor for methane oxidation. Denitrification is an emerging feature of gamma-MOB, which has been supported by genomics and was also experimentally demonstrated ( Hoefman et al. , 2014 ; Kalyuzhnaya et al. , 2015 ; Skennerton et al. , 2015 ; Kits et al. , 2015a , 2015b ). It has been proposed that respiration of nitrate might enable aerobic gamma-MOB to colonize anoxic waters ( Chistoserdova, 2015 ; Knief, 2015 ). In Lake Rotsee and Lake Zug, Crenothrix was indeed found in the anoxic waters below the oxycline in at least 2 consecutive years. Its abundance in anoxic lake waters suggests that it might successfully compete with more obligate anaerobic methane oxidizers, such as archaeal methanotrophs ( Haroon et al. , 2013 ) or ‘ Candidatus Methylomirabilis oxyfera’ ( Ettwig et al. , 2010 )."
} | 7,877 |
21544100 | PMC3197161 | pmc | 2,754 | {
"abstract": "A novel hydrothermal field has been discovered at the base of Lōihi Seamount, Hawaii, at 5000 mbsl. Geochemical analyses demonstrate that ‘FeMO Deep', while only 0.2 °C above ambient seawater temperature, derives from a distal, ultra-diffuse hydrothermal source. FeMO Deep is expressed as regional seafloor seepage of gelatinous iron- and silica-rich deposits, pooling between and over basalt pillows, in places over a meter thick. The system is capped by mm to cm thick hydrothermally derived iron-oxyhydroxide- and manganese-oxide-layered crusts. We use molecular analyses (16S rDNA-based) of extant communities combined with fluorescent in situ hybridizations to demonstrate that FeMO Deep deposits contain living iron-oxidizing Zetaproteobacteria related to the recently isolated strain Mariprofundus ferroxydans . Bioenergetic calculations, based on in-situ electrochemical measurements and cell counts, indicate that reactions between iron and oxygen are important in supporting chemosynthesis in the mats, which we infer forms a trophic base of the mat ecosystem. We suggest that the biogenic FeMO Deep hydrothermal deposit represents a modern analog for one class of geological iron deposits known as ‘umbers' (for example, Troodos ophilolites, Cyprus) because of striking similarities in size, setting and internal structures.",
"introduction": "Introduction Hydrothermal activity associated with deep-sea volcanism occurs at ‘hot-spot' volcanoes such as Lōihi Seamount, Hawaii ( Karl et al. , 1988 ) and Vailuluu Seamount, Samoa ( Staudigel et al. , 2006 ) and at mid-ocean ridge spreading centers such as Axial Seamount ( Johnson and Embely, 1990 ) and Larson's Seamounts ( Alt, 1988 ). Most hydrothermal activity occurs close to volcano summits and ridge axes, where significant thermal and chemical anomalies exist ( Sakai et al. , 1987 ) due to the intrusion and shallow emplacement of magma. However, there is mounting evidence for off-axis venting: examples include Baby Bare seamount on the eastern flank of the Juan de Fuca Ridge ( Wheat and Mottl, 2000 ), Lost City on the Mid-Atlantic Ridge ( Kelley, 2005 ), with consequence for oceanic elemental budgets ( Wheat and Mottl, 2000 ). Lōihi Seamount (summit at ∼1000 m below ocean surface) is a seismically active submarine volcano that represents an emerging Hawaiian Island ( Klein, 1982 ; DeCarlo et al. , 1983 ) (Supplementary Figure S1). Venting fluids near the summit of Lōihi (Pele's Pit) are enriched in CO 2 , NH 4 + , Si, Fe, alkalinity and Mn ( Sedwick et al. , 1992 ; Wheat et al. , 2000 ). Cm-thick Fe-rich microbial mats at Pele's Pit are highly localized, close to focused hydrothermal venting ( Karl et al. , 1988 ; Emerson and Moyer, 2002 ), and are dominated by neutrophilic Fe-oxidizing bacteria related to the genus Mariprofundus , which occurs within a novel class of the Proteobacteria, the Zetaproteobacteria ( Emerson et al. , 2007 ). We report ‘FeMO Deep', a novel hydrothermal field at 5000 mbsl off the southern flank of Lōihi seamount that is characterized by massive Fe-oxyhydroxide deposition. Chemical, mineralogical, morphological and biological evidence indicate that neutrophilic iron-oxidizing bacteria (FeOB) are responsible for the Fe-oxyhydroxide deposition that results in formation of regionally extensive, mineral-rich microbial mat ecosystems. We refer to the type of hydrothermalism observed at FeMO Deep as ultra-diffuse , being characterized by extremely dilute, cooled, distally sourced hydrothermal fluids. Ultra-diffuse hydrothermal fluid venting is expressed as slow leakage on the seafloor, km in size, which is larger than most described hydrothermal vent systems that are characterized by focused flow. The style of Fe-Mn deposition at FeMO Deep (lamination, mineralogy and setting), suggests that the system represents a modern genesis analog for hydrothermal Fe-Mn oxide deposits that are preserved in the geological rock record (for example, umber deposits from the Troodos Ophiolite ( Robertson, 1975 )).",
"discussion": "Results and discussion ROV Jason II dives at 5000 m, southeast of the base of the seamount along the south rift (Supplementary Figure S1) reveal extensive Fe mat deposits overlying and bridging basaltic pillow lavas. A black crust obscures Fe mats before sampling, resulting in a smooth pillow-basalt appearance ( Figures 1a and b ). Except the occasional small patches of rust-colored Fe-staining on the surface of the black crusts, these mats in most places appear indistinguishable from basalt pillows. This Fe mat, named ‘Ula Nui' (meaning ‘big red' in Hawaiian), is characterized by a laminated, mineralized, cohesive crust of alternating Fe- and Mn-rich minerals draping between basalts, and cap the underlying flocculent, gelatinous and Fe-rich mat ( Figure 1 ). Shimmering water is not observed at Ula Nui or at surrounding FeMO Deep mats, and thermal anomalies within mats are not detected in any mat interrogated with Jason II probes (∼3 dozen). Temperature loggers deployed for 1 day in mat indicate a slight, but robust and stable temperature average of 1.72 °C (±0.002 °C), a 0.23 °C thermal anomaly above average ambient bottom water temperature (1.49±0.002 °C) (Supplementary Figure S3). Surveys conducted around Ula Nui between 2006 and 2009 reveal a 15 000 m 2 area where mats occurred from 10's of cm to 1 m depth. Surveys ∼1 km east of Ula Nui reveal Fe mats that present as mounded structures protruding from the seafloor much like pillow basalts (∼1 m) (Supplementary Figure S4). The mounded mats have similarly been monitored with temperature loggers, and again reveal a small but robust temperature anomaly above ambient deep seawater temperatures (Supplementary Figure S5). In-situ voltammetric measurements of O 2 , Fe 2+ , Mn 2+ , H 2 S/HS − , FeS (aq) and Fe 3+ as a function of depth within mats indicate depletion of O 2 and elevated Fe 2+ . These data indicate concomitant upward flow of Fe 2+ and O 2 depletion with depth, implying active redox processes ( Figure 2 ). Fe 2+ concentrations within mats are highly anomalous compared with non-hydrothermal deep-sea environments where Fe 2+ supply (via respiration of Fe(III) oxides) is a diffusive sedimentary process ( D'Hondt et al. , 2004 ). Dissolved Mn and sulfur species are below in-situ detection limits (5 and 0.1 μ, respectively), whereas dissolved Fe species are detected at concentrations up to 150 μ. Similar opposing gradients of O 2 and Fe 2+ are observed repeatedly in mats at FeMO Deep, including the mounded mats (∼3 dozen measurements such as shown in Supplementary Figures S4 and S5). A movie depicting profiling and sampling at Ula Nui shows the neutrally buoyant, gelatinous consistency of the Fe mat that underlies the cohesive Fe-Mn crust (Supplementary Movie S1). Chemical analysis of mats indicate high Si content associated with amorphous opal (∼10 wt% Supplementary Table S1 and SOM methods) ( DeCarlo et al. , 1983 ; Frey and Clauge, 1983 ), which is consistent with a hydrothermal source for the Fe-rich mats at Lōihi's summit ( DeCarlo et al. , 1983 ) (Supplementary Table S1) and other seamounts. Si is generally enriched in diffuse hydrothermal fluids derived from higher-temperature reactions at depth. Ternary diagrams of Mn-Fe-(Co+Ni+Cu) × 10 are also used to distinguish between hydrothermal and other means of forming Fe-Mn-rich deposit formation, such as chemical precipitation from seawater and hydrothermal plume fall out ( Hein et al. , 1994 ). At Ula Nui, the lack of evidence for scavenging of trace metals from seawater (hydrogenous deposition) is consistent with a pure hydrothermal origin ( Figure 3 ), supporting our assessment. Microscopic examination of the mat reveals various filamentous and spherical mineral structures ( Figures 4 and 5 ). Energy dispersive X-ray analysis indicates that spherical structures are Mn-rich, whereas filamentous structures are Fe-rich (Supplementary Figure S2). Extended X-ray absorption fine structure spectroscopy (EXAFS) at the Fe K- and Mn K-edges for the discrete particles, respectively, indicate that the Fe-oxyhydroxides have less short-range structure than the reference mineral 2-line ferrihydrite, and are more consistent with biogenic Fe-oxyhydroxides formed at the Juan de Fuca Ridge ( Toner et al. , 2009 ) ( Figure 6 ); the Mn oxide has a layer-type structure most similar to triclinic birnessite ( Manceau et al. , 2002 ) ( Figure 6 ). The mineralogy and morphologies of the Fe-oxyhydroxide particles in the mat are strikingly consistent with those associated with the activities of known neutrophilic FeOB from Pele's Pit ( Emerson and Moyer, 2002 ), such as the stalks of Mariprofundus ferrooxydans , previously isolated from Pele's Pit ( Emerson et al. , 2007 ). T-RFLP and SSU sequence analysis using rRNA clone libraries to identify the prominent groups of bacteria within the mats reveal some similarities between ‘Ula Nui and Pele's Pit bacterial populations ( Moyer et al. , 1995 ; Figure 7 ). In particular, phylotypes related to the lithotrophic bacterium M. ferrooxydans ( Emerson et al. , 2007 ) are detected. Fluorescent in-situ hybridizations with group-specific probes demonstrate the presence of live Zetaprotebacteria within the Ula Nui mats (Supplementary Figure S6). The prominence of cells that hybridize with probes specific for Zetaprotebacteria , together with the morphological evidence cited above, indicates that Mariprofundus -like lithotrophs are important constituents of the Ula Nui mat community. Microbial cell densities measured in the Ula Nui Fe-Mn crust and underlying flocculent mat material range from 6.7 × 10 8 –1.4 × 10 9 cells per g dry weight, which falls within the range of cell densities detected across mats from the summit region of Lōihi in 2006 (6.5 × 10 8 –2 × 10 9 cells per g dry weight). An important energy source supporting this microbial community that includes live Zetaprotebacteria is Fe 2+ , which is advected into the mat from below ( Figure 2 ). We use bioenergetic calculations based on chemolithoautotrophic growth yields for cells sustaining growth based on redox reactions between Fe 2+ and O 2 to provide estimates of the potential energy that is available for biomass production within FeMO deep mats ( Heijnen and Van Dijken, 1992 ). These calculations indicate that even at very low O 2 activities (1 n and 1 p) the reaction is still strongly exergonic (−256 and −240 kJ mol −1 O 2 , respectively). The electrochemical profiles show that O 2 and not Fe is the limiting reactant, hence, Fe oxidation should be controlled by diffusion of O 2 into the mat. The drop of O 2 , from 145 μ outside the mat to <5 μ 5 cm into the mat, indicates a diffusive flux of at least 9 × 10 6 moles cm −2 a −1 . This O 2 flux equates to an energy flux of around 2.2 J cm −2 a−1 . Assuming that aerobic autotrophs require 292 kJ to fix 1 g of biomass ( Heijnen and Van Dijken, 1992 ) and that one cell weighs 10 −13 g ( Whitman et al. , 1998 ), this energy flux could support growth of 8 × 10 8 cells cm −2 a −1 and maintain a much greater population. With a mat thickness of 1 m, a fraction of Fe hydroxide filaments (dry weight) of 3% in the mat, and a density of solid Fe(OH) 3 of 3.1 g cm −3 , we get an average of 10 7 cells g −1 dry weight a −1 . These numbers are probably higher near the interface with seawater as Fe oxidation there is most rapid. By comparing this conservative number with the actual cell densities, we get turnover times of 20–80 years. These calculations support the inference that cell densities on the order of 10 7 cells per g dry weight can be expected to grow in a year. Hence, the observed cell densities can be attributed to Fe-oxidation entirely, if the turnover time is 20–80 years, or less so if other energy sources also contribute to cell densities observed, as expected in a natural mat ecosystem. A much larger population based on Fe-oxidation can be supported in maintenance mode (∼3 orders of magnitude ( Price and Sowers, 2004 ) by comparison to that which would be sustained during active growth. Our analyses can only resolve that cells are living, and we cannot infer a specific activity level for these populations—this should be an important further area of research for these novel mats. The presence of live bacteria related to a known and phylogenetically distinct clade of FeOB and the presence of Fe-oxyhydroxide minerals, which are morphologically and mineralogically consistent with known bacteriogenic Fe-oxyhydroxides, support our interpretation that these Fe mats are of biological origin. Our chemical measurements further support the hypothesis that this microbial mat is fed by ultra-diffuse advection of hydrothermal fluids, which derive from a higher-temperature source enriched in Fe, Mn and Si that has undergone extensive subsurface cooling. These detectable processes at the seafloor may reflect an abundant deep biosphere harbored in the subsurface, in the cooling and mixing zone beneath the seamount, conditions that would be conducive to supporting microbial life. Bioenergetic calculations further support our inference that Ula Nui represents a Fe-based, lithotrophic ecosystem. Biological, geochemical and mineralogical data suggest that FeMO Deep mats such as Ula Nui represent a distinct style of ultra-diffuse hydrothermalism, hosting a microbial ecosystem that has an integral role in deposition of laminated, regionally extensive Fe-Mn umber deposits. Low-temperature Si- and Fe-rich hydrothermal deposits have previously been observed in modern settings ( Corliss et al. , 1978 ; DeCarlo et al. , 1983 ; Alt, 1988 ; Juniper and Fouquet, 1988 ; Mills et al. , 2001 ). Such vent fields do not generally form Fe-rich deposits at the regional scale of FeMO Deep; rather, they are more similar to the rapidly generated, but highly localized, shallow (a few cm) Fe mats observed at Pele's Pit. However, in contrast to these previous modern-day observations, hydrothermal Si- and Fe-rich deposits that share characteristics with those seen at FeMO Deep are observed in the rock record throughout Earth's history, and some report microfossils of twisted and branching Fe-oxyhydroxide filaments, which are morphologically similar to those found within FeMO Deep mats, and those produced by Mariprofundus spp. ( Juniper and Fouquet, 1988 ; Alt et al. , 1992 ; Little et al. , 2004 ; Emerson et al. , 2007 ; Slack et al. , 2007 ). For example, Si- and Fe-rich deposits up to 20 m thick have been reported at ODP Site 801 ( Alt et al. , 1992 ) within the Jurassic oceanic crust in the Western Pacific, and numerous fossil hydrothermal Si-Fe-deposits (for example, Jasper and Fe formations) report microfossils in ophiolites as old as 1.74 billion years ( Robertson, 1975 ; Little et al. , 2004 ; Slack et al. , 2007 ). The genetic model for one class of Fe oxide deposits, umbers, has previously invoked deposition of iron and manganese oxides via water-column precipitation of hydrothermal fluids, followed by particle fall out and accumulation in local depressions in a ridge flank setting. However, over the course of the past 30 years of research at deep-sea hydrothermal systems, numerous studies and observations made have failed to reveal modern examples of plume fall out forming umber-like deposits. In contrast, the laminated Fe-Mn structures pooled between pillow basalts observed at FeMO Deep offer an alternative interpretation of umber genesis that is consistent with geological observations. To our knowledge, FeMO Deep represents the deepest hydrothermal field reported to date, and is the first modern analog of an umber seafloor iron formation in genesis. The discovery of this microbial community at Lōihi and the potential for Fe-based chemosynthetic ecosystems to exist elsewhere in the deep ocean and subseafloor underscores the importance of geomicrobiological interactions in shaping the planetary ecosystem on Earth today, and in the geological past."
} | 4,021 |
21261829 | PMC3864445 | pmc | 2,755 | {
"abstract": "Summary Hydrogen, the most abundant and lightest element in the universe, has much potential as a future energy source. Hydrogenases catalyse one of the simplest chemical reactions, 2H + + 2e ‐ ↔ H 2 , yet their structure is very complex. Biologically, hydrogen can be produced via photosynthetic or fermentative routes. This review provides an overview of microbial production of hydrogen by fermentation (currently the more favourable route) and focuses on biochemical pathways, theoretical hydrogen yields and hydrogenase structure. In addition, several examples of metabolic engineering to enhance fermentative hydrogen production are presented along with some examples of expression of heterologous hydrogenases for enhanced hydrogen production."
} | 188 |
26601255 | PMC4643789 | pmc | 2,758 | {
"abstract": "A mathematical model helps explain how the complex social systems of ants and bees make collective decisions.",
"introduction": "INTRODUCTION Ants and bees appear to rely on decentralized decision-making for critical choices. For example, in choosing a new nest site—a decision that has huge implications for the survival of the group—decisions must be made without central control and with no single individual evaluating the total available information ( 1 ) or any one individual making direct comparisons of the available options ( 2 – 5 ). Although individual agents follow simple rules that allow them to uncover very limited and local information, the colony as a whole must efficiently integrate the resulting flow of information into a high-quality, final decision ( 6 – 10 ). Consider how a swarm of honey bees, Apis mellifera , chooses a new hive location ( 8 – 14 ). When a swarm abandons the old hive, it temporarily gathers at a tenuous location. About 5% of the bees are scouts, and after exploring the surrounding area for possible hive sites, a scout may perform a waggle dance that indicates the location of the site it discovered ( 15 , 16 ). The likelihood of performing a waggle dance, and its duration, depends on the quality of the site that was investigated ( 9 , 13 , 14 ). The waggle dances serve to recruit additional scouts to further investigate the “advertised” site. The longer the dance, the more likely that the new scouts will investigate the site and bring back independent evaluations. Over time, positive feedback loops are generated ( 8 , 13 ), and once the number of scout bees at a particular site reaches a quorum threshold (of about 30 to 40 bees), those scouts return to the swarm and lead it to the new site ( 12 ). Leptothorax ( Temnothorax ) albipennis ants choosing a new nest site behave similarly ( 1 , 6 , 17 ). When a scout finds a higher-quality site, it quickly returns to the old nest site and recruits a nestmate by tandem running, a tedious process that entails the scout teaching the recruit the route to the new site ( 18 – 21 ). The speed of recruitment is tied to the quality of the site ( 1 , 3 ), with better sites inducing quicker responses. As before, positive feedback arises when recruits become recruiters. Finally, when the number of ants at the new site reaches a quorum threshold, the recruiting ants switch from tandem running to the much faster process of carrying their remaining nestmates from the old to the new site ( 22 ). Other social organisms make collective decisions with mechanisms reminiscent of those of ants and bees. Social spiders coordinate their emigrations to a new nest ( 23 ) with silk draglines, allowing positive reinforcement of existing routes much like the pheromone trails of ants ( 23 – 25 ). Cockroaches are more likely to remain in shelters when other cockroaches are nearby, leading to a collective choice of a single home ( 26 , 27 ). Even bacteria share information and detect quorums, allowing for collective decisions regarding sporulation, virulence, and gene exchange ( 28 ). There is some speculation that primate brains use a similar decentralized decision mechanism. No single neuron is solely responsible for the brain’s decision. In a visual discrimination task, for example, a subset of specialized neurons integrates sensory signals from other neurons and allows the brain to make a decision to trigger other neurons to initiate a motor response ( 29 ). Complex, decentralized information processing can be achieved with a cell assembly, a recurrent circuit of neurons that becomes active when stimulation spreads with positive feedback ( 30 , 31 ). Neurons are generally understood to accumulate information and fire when the stimulus hits a threshold to implement a decision ( 32 – 34 ). In this regard, the primate brain may function analogously to a colony of social insects ( 35 – 37 ). Similar mechanisms may even be at work in large-scale social processes ( 38 ). For example, consider the choice of a personal MP3 player. Consumers who purchase such players “advertise” them when they use them (particularly if the players have some distinctive feature, for example, white ear buds). Moreover, consumers who enjoy their players are more likely to use them. Someone new to the MP3 market may observe the players that others use and purchase on the basis of these observations. This kind of direct marketing is often a major driver of consumer demand for new products, especially when competing brands have not yet established distinct reputations ( 39 ). At some point in the evolution of the market, however, a critical mass of consumers may choose the same product and fundamentally change the market dynamics (say, by adopting a particular technological innovation, by fueling economies of scale in the production process, or by enticing suppliers or producers of complements to enter into exclusive agreements) so that only the leading product can survive in the market ( 40 , 41 ). Similarly, people collaborating to make a group decision also tend to share information that favors options that already have popular support while hesitating to share information that favors unpopular options ( 42 ), and they commonly reach a consensus choice through plurality voting ( 43 ). Any decision mechanism must trade off exploring new options versus exploiting the best option known to date. Although further exploration helps identify superior options, it comes at the cost of not acting on a known option. Thus, too much exploration may lead to indecisiveness and thus harm fitness, whereas too little may imply the acceptance of suboptimal choices. Given the inherent tradeoff between speed and accuracy, the diffusion model of decision-making ( 34 )—which does not incorporate positive feedback in the accumulation of evidence—is efficient for binary choice, but it requires the agents to compare the options or at least to inhibit activity for competing options ( 35 , 44 ). When this is not possible, positive feedback in the process of exploration may prove useful for making reasonably good, quick decentralized decisions. The model We present a model of a two-part process for decentralized decision-making, involving search and recruitment with positive feedback in the first phase and quorum detection to trigger a consensus choice (without centralized processing) in the second phase. First, individual agents randomly search over a set of feasible options, biased by the quality of the options revealed during previous searches. Then, a final consensus choice is triggered when a quorum of agents investigating any one particular option forms. [A consensus choice is a choice that the entire system must abide ( 45 ), but of course, it is not necessarily a unanimous choice via “consensus sensing” ( 10 ).] We capture this process with an urn scheme that runs until hitting a threshold. Search and recruitment Agents (scouts) explore one of many possible options and then return home to recruit additional agents to further explore that option. A Poisson process for each agent governs its trips home, and the recruitment of additional scouts on each trip depends on the quality of the explored option. We assume that there are C possible options and that some number of agents w c t investigate each c ∈ C at any time t . We refer to w c t as the weight on c at time t . All weights are initially set to the same, positive value w 0 and then accumulate over time. Each agent is equally likely to return home at any time, and when it does, it recruits additional agents to join it and further explore the same option. The chance of an agent recruiting for option c at time t is simply proportional to the weight w c t . Each option has a set of immutable attributes that defines its quality, and the extent of recruitment for c depends on its quality. The agents investigating c will recruit v c additional agents to continue exploring it when they return. We think of the number of recruits per return trip home, v c , as an ordinal measure of the quality of c . Quorum detection The search process above generates a distribution of agents investigating each possible option at any given time. Given the decentralized nature of these systems, there must be some feasible trigger that ends the search process and finalizes the consensus choice. One possible solution to this problem would be to have the search probabilities converge to zero or one—that is, have all of the probability concentrated on a single option. Forcing such a unanimous decision on the system is problematic, because it may form extremely slowly, perhaps leading to a serious loss of fitness. Moreover, we have empirical evidence, at least in the case of honey bees ( 46 , 47 ), ants ( 6 ), and stickleback fish ( 48 , 49 ), that unanimity is not what triggers a consensus choice. Instead, a final choice is made once the number of agents in favor of a particular option reaches a quorum ( 45 , 50 ). On the basis of the above arguments, we incorporate into our model a quorum threshold , τ, that triggers—as the final decision—any option that is being investigated by at least that number of agents. This threshold is effectively the finish line in the race for each option to accumulate weight. The decision is determined by the first passage of w c t ≥ τ . The level of the quorum threshold has important implications for the decisions that arise in the system. If the threshold is set too high, then a quorum may not be reached for a long time, resulting in prolonged inaction. If the threshold is set too low, then a quorum might be achieved for a relatively low-quality option. Thus, the optimal quorum threshold depends on a tradeoff between speed and accuracy in the decision-making process. From a normative standpoint, a good threshold allows the system to withstand various transients in the probability distribution while still remaining responsive to the acquired information in a timely manner. The urn scheme We use a simple (Polya) urn process to model this decision mechanism. This process is easy to visualize. Assign to each of the C options a unique color, and place w 0 balls of each color into an urn. The number of balls of a particular color in the urn corresponds to the number of agents investigating the associated option. Each ball has the same rate at which it may be randomly drawn from the urn. When a ball with color c is drawn, it is immediately placed back into the urn along with v c identically colored balls. This process continues until a threshold number of balls τ is reached. We assume that the supply of balls available to enter the urn is large relative to the quorum threshold, because the number of scouts in a swarm of bees or a colony of ants is typically much larger than the number required to achieve a quorum. Analytical results Our analysis aims to characterize the behavior of this decision process, that is, to determine the choice probabilities and the (distribution of) decision times. First, we must understand the accumulation process. We precisely characterize how the composition of balls in the urn evolves over time. Lemma 3.1 of ( 51 ) gives us the distribution of balls of each color at any time t (in the absence of a threshold for stopping the process). Lemma 1. The moment-generating function ϕ c ( t , s ) = E [ e s w c t ] is given by ϕ c ( t , s ) = ( e v c ( s − t ) e v c ( s − t ) − e v c s + 1 ) w 0 v c (where s is the argument of the moment-generating function). Proof. We can describe this Polya process with a diagonal C × C matrix with the v c values along the diagonal and 0’s elsewhere. The evolution of the number of balls of a given color is independent of the evolution of other colors (until the threshold is hit). Thus, Lemma 3.1 of ( 51 ) directly applies. In principle, this moment-generating function fully characterizes the distribution of weights, w c t , where s is the argument of the moment-generating function. In practice, however, calculating the likelihood of hitting a threshold τ at a given time t is complicated. An asymptotic result is simple to obtain. Suppose the threshold τ is infinite so that the Polya process can run forever. Eventually, almost all of the weight converges on the choice with the highest quality. Theorem 1. If there is a unique optimal choice c * = argmax c \n v c , then: lim t → ∞ w c t ∑ j w j t = { 1 if c = c * 0 otherwise . Proof. As t → ∞ , w c t e v c t → D Gamma ( w 0 v c , v c ) [see Theorem 3.1 of ( 51 )]. In the infinite time limit, we recover the unanimity decision rule, and the probability of a mistake (that is, selecting an option other than c *) vanishes. Although the asymptotic properties of the urn process are informative, feasible decentralized systems must make decisions in finite time and require a finite threshold. For any given threshold τ, we would like to describe the probability p c (τ) of selecting each possible choice c as well as the waiting time T (τ) until the decision is made. We can characterize the waiting time T c (τ) until the number of agents exploring option c would hit the threshold τ (independent of recruitment for other options). Let λ denote the intensity of the Poisson process for each agent’s return home. Lemma 2. The waiting time T c (τ) has the Hypoexponential(λ 0 , λ 1 ,…, λ n ) distribution with: λ i = ( w 0 + i v c ) λ for all i , (1) and w 0 + n v c < τ ≤ w 0 + ( n + 1 ) v c , (2) which implies n = ceiling ( τ − w 0 v c ) − 1 . (3) Proof. We have an Exponential(λ) distribution for the time until a given agent returns home, and thus, at any time t , we have an Exponential ( w c t λ ) distribution for the time until additional agents are recruited to explore option c . Thus, the waiting time T c (τ) until the number of agents exploring option c hits the threshold τ is the sum of independent, exponentially distributed variables with arithmetically increasing parameters. The hypoexponential density function is f ( t ) = ∑ i = 0 n C i , n λ i e − λ i t with C i , n = ∏ j ≠ i λ j λ j − λ i . Taking n and λ 1 ,…, λ n to be functions of c and τ (given by Eqs. 1 and 3 ), this gives us the probability density function f c, τ ( t ) for each T c (τ). Modeling the decision process in terms of accumulation to a fixed threshold, we can think of − T c (τ) as a stochastic utility for each alternative c . The system selects the alternative that maximizes this utility. [Of course, because this process is stochastic, maximizing this utility does not necessarily align with maximizing quality. The design of the system reflects dual objectives of aligning utility with quality (to more often choose higher-quality options) and maximizing this utility (to make quicker decisions).] The density functions for the T c (τ) variables thus determine the quantities of interest in the system: the probability p c (τ) of selecting each possible choice c as well as the waiting time T (τ) until the decision is made. Theorem 2. The waiting time T (τ) distribution and the choice probabilities p c (τ) are determined by the T c (τ) distributions given by Lemma 2. The time until a decision is made by the decentralized system is T (τ) = min c \n T c (τ). The probability that the eventual decision is for choice c is p c (τ) = Pr[ T c (τ) < min c ′≠ c T c ′ (τ)]. Proof. The urn process runs until the first time that balls of any one color accumulate to the threshold, and the probability of selecting any given choice is simply the probability that balls of the corresponding color reach that threshold first. Theorem 2 characterizes the (distribution of) time(s) it takes to make a decision and the choice probabilities as functions of the quorum threshold, given any menu of possible choices. We can, of course, calculate the minimum of a set of random variables, as the theorem requires us to do, but there is no simple, closed-form expression for this. Given the lack of a closed-form solution, we use computation to gain additional insight about this process. Computational results To explore the effects of parameter variation and the introduction of noise into the process, we run computational experiments of the proposed mechanism. Although the process described above runs in continuous time, we identify discrete time steps every time an agent returns home to recruit (that is, every time a ball is drawn from the urn). Let the index μ count the number of agents that have returned home for a visit, and denote the time when the μth agent returns home as t μ . When there are w = ∑ c w c t μ agents exploring the set of possible options, the expected time until the next agent returns home is 1 w λ . Setting λ = 1, which normalizes the units of time, we have: E [ t μ + 1 − t μ | w ] = 1 w . We generally set w 0 = 1 for simplicity. Each computation reports the average time until decision T (τ) and the probability of a mistake p ~ c * (τ) = 1 − p c * (τ) as a function of the quorum threshold τ. We see the tradeoff between speed and accuracy across varying quorum thresholds by viewing the expected time until decision and the mistake probability as parametric functions of the threshold. Allowing the threshold to vary, we have a Pareto-efficient frontier along which the speed of the decision mechanism cannot be improved without sacrificing accuracy (and vice versa). Parameter variation Increasing the number of possible options C makes for a less accurate decision, but a slightly quicker one as well. (This is shown in the appendix in fig. S1.) More options provide more opportunities for suboptimal options to accumulate a quorum, leading to more mistakes and less decision time. But then, to reach the same level of accuracy, the system needs a higher threshold, and this increases the time required to make the decision (as shown in Fig. 1 ). Intuitively, more possible options make for a more difficult decision. Fig. 1 Pareto frontiers of mistake probability and expected waiting time with 2 and 4 options. The optimal choice has quality v c* = 2, whereas the suboptimal choices have quality v c = 1 for all c ≠ c* . As an artifact of specifying recruitment (that is, choice quality) so precisely, there are thresholds for which the decision is both slower and less accurate than for a threshold one unit smaller. The corresponding points on the graph are clearly not on the Pareto frontier, but they are shown for completeness. Increasing the quality of the optimal choice, v c * , makes the decision easier (as shown in the left plot in Fig. 2 ). (These Pareto frontiers are derived from simulations shown in figs. S2 and S3.) As the quality of the optimal choice increases, the decision can be made faster and with less chance of error. Recruitment becomes more effective, so the agents accumulate at this option more quickly and the system achieves the quorum sooner. Fig. 2 Pareto frontiers of mistake probability and expected waiting time with varying choice quality. Left: Varying optimal choice quality. There are C = 2 possible choices, and the quality of the suboptimal choice is v ~ c * = 1. Right: Varying suboptimal choice quality. There are C = 2 possible choices, and the quality of the optimal choice is v c * = 4. Increasing the quality of a suboptimal choice, however, does not have such straightforward consequences. It has three effects: (i) it makes the decision process quicker, (ii) it increases the probability of selecting the suboptimal choice, and (iii) it lowers the cost of making a suboptimal decision. Thus, the net impact on the ultimate quality of the decision could go in either direction. The right graph in Fig. 2 shows that increasing the quality of a suboptimal option hurts the accuracy of the decision that can be made within a given time. Noise and risk aversion Assuming that each option has some absolute quality helped keep our model analytically tractable, but we can extend the model by allowing the quality, v c , to be a random variable. At a conceptual level, we might think of the process of search as inherently noisy, because there could be natural variation in the agents’ perceptions of quality. Alternatively, we might think of the process of recruitment as inherently noisy, with variation in the ability of agents to recruit other agents. In either case, the formal treatment is the same. One potential source of noise is in the perception of the quality of options when sampled across multiple attributes. If the individual agents cannot investigate all attributes, they may produce a noisy estimate of overall quality by sampling a single (or a few) attribute(s). We can associate random sampling of attributes with noise in the overall quality of the option. [If agents were to specialize in sampling particular attributes and were also more likely to recruit their own type of specialist, then the dynamics of the decision mechanism would be more complicated. Natural systems (for example, a swarm of bees or a colony of ants) do not (to our knowledge) exhibit such behavior, but it could perhaps arise in human-engineered systems.] The decision mechanism is generally robust to noise (see fig. S4), but capable of distinguishing when one option has noisier quality than another. We can think of an option with noisier quality as riskier. We compare the attractiveness of risky and safe options (that is, options with the same expected quality but more or less noise, respectively) in Fig. 3 , which shows that (for a fixed quorum threshold of 100) the safe option (with v Safe = 2) is more likely to be selected than a risky option ( with v Risky = { 1 , R − 2 R − 1 ; R , 1 R − 1 } ) , and it is increasingly preferred to even riskier options (that is, as R increases). There is nothing special about the threshold of 100, and the result holds for almost all thresholds (possible exceptions being low thresholds that can be reached by a single draw of the risky option, due to recruitment having discrete increments), as shown in fig. S5. The effect persists with high thresholds because noise in the process of search and recruitment does not inevitably balance out; rather, positive feedback in the process makes it more difficult for the risky option to overcome early indications of low quality. [We prove in the appendix (Proposition 1) that the probability of selecting a safe option with quality v Safe = 1 over a risky option with quality v Risky = { 0 , R − 1 R ; R , 1 R } , for a quorum threshold of τ = R + 1 is R B ( R , 1 + 1 R ) (where B is the Euler beta function), which is an increasing function graphed in fig. S6.] Thus, the decision mechanism exhibits a systematic degree of risk aversion. Fig. 3 Probability of selecting the safe option over a risky option with the same expected quality for a fixed quorum threshold of 100. There are C = 2 options. The safe option has quality v Safe = 2. The riskiness of the risky option is indexed by the potential reward R such that the quality of the risky option is v Risky = { 1 , R − 2 R − 1 ; R , 1 R − 1 } , that is, it has expected quality 2 and variance R − 2. Risk aversion is defined for deterministic choice models as a preference against mean-preserving spreads and is conventionally represented with concave utility functions. However, risk aversion can get more complicated for stochastic choice models. A strong condition of risk aversion with stochastic choice would require the entire distribution of the stochastic utility to shift downward for mean-preserving spreads. We do not obtain such universal risk aversion, noting occasional exceptions to this general pattern of preferences at low thresholds (see fig. S5). Instead, we define risk aversion in our context as occurring if a mean-preserving spread of an alternative’s quality decreases the probability of choosing that alternative, for sufficiently high thresholds. We observe this in our computational results. The intuition behind the emergence of risk aversion is that the positive feedback in the search and recruitment process allows small advantages to be self-reinforcing, so an option that consistently appears relatively good fares better than one that occasionally appears either great or lackluster. In a world with natural selection, where an entire population can be decimated if a risky choice turns out badly, it may well be adaptive to use a decision mechanism that inherently favors safer choices ( 52 ). Moreover, our mechanism permits the system’s consensus choice to be risk-averse even when individual agents are risk-neutral. Thus, the system could simultaneously be risk-averse for systemic risk and risk-neutral for idiosyncratic risk (when consensus choice is not required), which would be evolutionarily adaptive ( 53 ). Discovery and disruption We can enrich our model by allowing agents to discover the possible options on their own and to be disrupted from search and recruitment by outside forces. We assume that a Poisson process, with intensity β c , governs the discovery of each possible option c and that each agent has an exponentially distributed lifetime for search and recruitment with hazard rate δ of disruption. With discovery, a natural initial condition is w 0 = 0, that is, the urn is initially empty and agents need to discover an option for recruitment to begin. The index μ for discrete time steps now must count all events, that is, every time an agent discovers a possible choice or falls prey to disruption, as well as when an agent returns home to recruit. The expected time step becomes: E [ t μ + 1 − t μ | w ] = 1 w ( λ + δ ) + ∑ c β c . Figure 4 shows that decision speed and accuracy are fairly robust to the introduction of option discovery and disruption of search and recruitment. Increasing the rate of discovery (for all options) speeds up the decision and reduces mistakes by limiting the sensitivity to initial advantage, making it easier for the optimal choice to catch up when it gets discovered later on. Increasing the rate of recruitment also speeds up the decision, but can lead to more mistakes by reinforcing initial advantages (that is, when a suboptimal option is discovered first). Increasing the rate of disruption slows down the decision, yet it counteracts initial advantages, allowing the optimal option more time to get ahead through stronger recruitment. (Figure S7 shows the distinct effects on mistake probability and expected time until decision.) In all cases, the quorum threshold could be adjusted to efficiently navigate the new speed versus accuracy tradeoff. The net effects, which are shown in Fig. 4 , are that increasing the rate of disruption harms the decision, whereas increasing the rate of discovery or the rate of recruitment improves the decision. Thus, although recruitment introduces positive feedback that can reinforce suboptimal options, it speeds up the decision process enough that at higher thresholds the system can make better, quicker decisions. On the other hand, disruption slows down the decision process so much that despite the opportunities for error correction (by effectively allowing agents to occasionally change their minds), at lower thresholds the system makes slower, worse decisions. The error-correcting feature of disruption could nonetheless be helpful if, contrary to our assumption, a system were constrained to a small number of agents and thus forced to have a low threshold. Fig. 4 Pareto frontiers of mistake probability and expected waiting time with varying rates of discovery, recruitment, and disruption. There are C = 2 options. The optimal choice has quality v c* = 2, and the suboptimal choice has quality v ~c* = 1. The rate of discovery β c is the same for both options, and it varies across the columns. The agents’ hazard rate of disruption δ varies across the rows. The agents’ rate of recruitment λ varies within each panel.",
"discussion": "DISCUSSION From ants ( 54 ) to bees ( 46 ) to neurons in the brain ( 55 ), a variety of systems productively use decentralized decision mechanisms. Our general notion of decentralized decision-making assumes that no single agent has direct access to information across all of the choices or the ability to make, and communicate, a final decision. Although each agent does have the ability to make limited judgments and decisions, it is the system as a whole that must integrate these activities into a final choice and action. Our model of decentralized decision-making abstracts beyond any one of these systems and aims to provide a deeper understanding of how such mechanisms behave. A Polya urn scheme running until it hits a finite threshold parsimoniously captures a decentralized decision mechanism in which agents gather local information about possible options through search and recruitment with positive feedback, and the system then makes a consensus choice when it detects a quorum in support of a particular option. In this approach, we add to the literature that uses the Polya urn process to model positive feedback in firm growth ( 56 , 57 ), technology lock-in ( 40 , 58 ), the common law legal system ( 59 ), the evolution of social and political institutions ( 60 , 61 ), and the design of medical trials ( 62 ). Analytically, we characterized the waiting time to make a decision and the choice probabilities for any quorum threshold, and we identified an inevitable tradeoff between the speed and the accuracy of the decision. Numerical experiments showed that the system’s ability to make reasonably good, quick decisions is robust to parameter variation, noise, and disruption. Moreover, the computation reveals that the decision mechanism naturally exhibits systemic risk aversion. Additional assumptions about the cost of waiting and the relative values of possible options would be necessary to evaluate the exact tradeoff between a decision’s speed and accuracy. At the extremes, an infinite quorum threshold requires infinite waiting time, and a minimal quorum threshold corresponds to uniformly random choice, so the optimal threshold lies in between. The optimal threshold depends on the particular decision context, and there is some evidence that natural systems tune their thresholds in response to the decision context to make better tradeoffs between speed and accuracy ( 17 , 63 ). The fact that many natural systems independently evolved similar decentralized decision mechanisms suggests that such mechanisms may represent a robust solution to the general problem of making good, group-level decisions in the absence of centralized control. Indeed, we suspect that evolutionary forces are sufficient to form such natural systems and, over evolutionary time, tune their performance. The decentralized decision mechanism we described here may also prove useful in the design of new social and artificial systems. Novel applications range from improving human organizations to applying such techniques to artificial systems like algorithmic search and the control of swarms of robots or networked computers."
} | 7,807 |
35028625 | PMC8714767 | pmc | 2,759 | {
"abstract": "The use of intensive non-sustainable agricultural practices for satisfying global food demand is degrading the agro-ecosystems, leading to their inability to produce efficient and equitable sources of calories. Microbial communities play an important role in the improvement of soil fertility and plant development; thus, the genetic and metabolic diversity of microbiota in agro-ecosystems is a promising alternative for designing microbial inoculants to not only produce enough food but also mitigates the economic, health, social, and environmental issues caused by conventional agriculture. This Special Issue has been launched to compile and inspire high-impact recent advancements on bioprospecting beneficial microorganisms as a sustainable strategy to warranty global food security."
} | 197 |
23110251 | PMC3482764 | pmc | 2,760 | {
"abstract": "Major ampullate (MA) dragline silk supports spider orb webs, combining strength and extensibility in the toughest biomaterial. MA silk evolved ~376 MYA and identifying how evolutionary changes in proteins influenced silk mechanics is crucial for biomimetics, but is hindered by high spinning plasticity. We use supercontraction to remove that variation and characterize MA silk across the spider phylogeny. We show that mechanical performance is conserved within, but divergent among, major lineages, evolving in correlation with discrete changes in proteins. Early MA silk tensile strength improved rapidly with the origin of GGX amino acid motifs and increased repetitiveness. Tensile strength then maximized in basal entelegyne spiders, ~230 MYA. Toughness subsequently improved through increased extensibility within orb spiders, coupled with the origin of a novel protein (MaSp2). Key changes in MA silk proteins therefore correlate with the sequential evolution high performance orb spider silk and could aid design of biomimetic fibers.",
"discussion": "Discussion The unrivaled toughness of orb spider MA silk coincides with the appearance of a novel MaSp2 protein, at the base of Orbiculariae. MaSp2 contains a high content of larger side chain amino acids and, in particular proline, which is incorporated into a novel glycine-proline-glycine-glycine-X motif (where X is a small subset of amino acids; GPGGX). Proline generally destabilizes secondary structures in proteins, favoring amorphous protein networks 22 . The GPGGX motif kinks the backbone of the peptide into an extensible β-spiral 24 , which explains in part the up to five fold greater extensibility of the MA silk of orb spiders compared to RTA clade (ε u = 0.8±0.2 vs 0.39±0.09). This high extensibility explains why work to fracture is ~100% greater in orb spiders compared to RTA clade (W f = 290±90 vs 160±50 MJ/m 3 ). The combination of high toughness and extensibility in the MA silk of orb spiders plays a critical role in how webs resist breaking under the impact of flying prey 25 26 . Proline-rich MaSp2 imparted MA silk with a new flexibility at two very different levels. First, Orbiculariae MA silk extensibility in its ground state varies almost three-fold among taxa (ε u = 0.51 to 1.3). The proline content of MA silk in orb spiders correlates closely with interspecific variation in compliance and extensibility 10 27 . In contrast, the tensile behavior of haplogyne and RTA clade silks are homogeneous within each clade. This suggests that shifts in the expression of MaSp1 versus MaSp2 provide a highly evolvable mechanism tailoring the functional properties of MA silk in different species of orb spiders. This is particularly evidenced by Argiope . This genus is ~30MY old, yet the performance of MA silk varies to almost the same degree as exhibited by the rest of the 210MY old Orbiculariae ( Fig. 1 ). This high evolvability results in the MA silk of orb spiders occupying three quarters of the total performance space delimited by all other spider species. Second, MaSp2 facilitates performance plasticity within individual orb spiders 10 . Supercontraction occurs in part as the GPGGX motif interacts strongly with water 23 , mobilizing the amorphous fraction of the silk. By stretching MA fibers in water, the amorphous fraction is increasingly aligned and this structure can be held in place by hydrogen bonds when the silk dries. This stiffens the silk and allows any particular stress-strain curve in the range of accessible properties to be reached in a predictable and reproducible way. This ‘wet stretching’ can therefore make the MA silk from a single orb spider occupy almost any region of the performance space in Fig. 1 from its ground state to the left of the performance space 9 . During natural spinning, shear forces in the duct of the MA gland align silk molecules in still wet fibers, pulling them out of their ground state. Thus, spiders whose silk supercontracts strongly can access a greater range of performance for MA silk 10 . Our results draw a coherent picture of spider dragline silk evolution, relating major innovations in MaSps to changes in silk functional properties. The origin of the orb web is a singular event in the evolutionary history of spiders that played a dominant role in the evolution of silk, but we reveal that the history of MA silk is significantly more complex. MA silk performance is characterized by stability within lineages, punctuated by evolutionary changes that correlate with innovations in molecular composition. Basal MA silk shows poor mechanical properties compared to derived taxa. Increased homogenization and repetitiveness of GA and poly-A motifs, as well as the origin of a new GGX motif, in MA silk proteins correlate with increases in strength and stiffness that pre-date the origin of orb webs. However, the extreme toughness of orb spider MA silk is only reached after the evolution of a novel protein, MaSp2 that greatly improved extensibility. In addition to facilitating the function of aerial orb webs, MaSp2 facilitates inter- and intra-specific variation in the mechanical performance of MA silk. The discovery of these evolutionary correlations between the molecular composition and tensile properties of MA silk should facilitate constructing bioinspired fibers that mimic the outstanding properties of natural orb spider dragline silk 28 ."
} | 1,351 |
31172998 | null | s2 | 2,761 | {
"abstract": "We introduced a new concept to the control of wetting characteristics by modulating the degree of atomic defects of two-dimensional transition metal dichalcogenide nanoassemblies of molybdenum disulfide. This work shed new light on the role of atomic vacancies on wetting characteristic that can be leveraged to develop a new class of superhydrophobic surfaces for various applications without altering their topography."
} | 105 |
34721985 | PMC8520691 | pmc | 2,762 | {
"abstract": "Background Open pit antimony (Sb) mining causes serious soil pollution, and phytoremediation is a low-cost approach to remediate heavy metal contaminated soil. Rhizosphere bacteria play an important role in ecological restoration in mining areas. There is a knowledge gap on how to find suitable rhizosphere microorganisms to improve the phytoremediation effect. Understanding the differences of rhizosphere bacterial diversity in different restoration stages is helpful to find suitable bacteria for ecological restoration. Methods A method of the substitution of “space” for “time” was used to study the effect of natural restoration on rhizosphere bacterial community. According to the dominant vegetation types (herb, shrub, and tree) in the natural restoration area of Sb mining, the early restoration (ER), middle restoration (MR), and later restoration (LR) from the largest Sb mine (Xikuangshan mine) in the world were selected to evaluate the differences in the composition and diversity of rhizosphere bacteria during three natural restoration stages. Each restoration stage had five samples. To determine the relationship between restoration stages and bacterial diversity in the rhizosphere, high throughput sequencing of PCR amplified were used. Results Alpha diversity, as assessed by Chao indices, appeared lowest in ER but this trend was not seen with other diversity metrics, including the Simpson and Shannon. Beta diversity analysis suggested there were differences in rhizobacterial community structure associate with restoration stage. At the phylum level, natural restoration led to a significant increase in the relative abundance of Actinobacteria in the MR, and a significant decrease in the relative abundance of Patescibacteria in the LR. Additionally, Calditrichaeota , Deferribacteres and Epsilonbacteraeota were only found in ER. At the genus level, the relative abundance of RB41 and Haliangium were highest in LR plots, while that of Bacillus and Gaiella were highest in ER plots. Additionally, the Azorhizobium genus was only detected in the ER phase. Overall, our findings suggested that several rhizosphere microbial communities had significant differences among three natural restoration stages (ER, MR, and LR) and the rhizosphere bacterial communities mainly appeared in the early restoration stage can be preferred for remediation of pollution soil in Xikuangshan.",
"conclusion": "Conclusions This is the first report on the diversity and composition of rhizosphere bacterial community in Sb mines with regard to the different natural restoration stages. Sequencing analysis of the rhizosphere soil showed that the alpha and beta diversity of rhizosphere bacteria were significantly different during three natural restoration stages (ER, MR, and LR). RB41 and Bacillus genus had obviously high relative abundance in the ER. Three phyla ( Calditrichaeota , Deferribacteres , and Epsilonbacteraeota ) and Azorhizobium genus only appeared in the ER. These rhizosphere bacteria might be ideal bacteria for improving phytoremediation efficiency in the Sb mining region. In future research, we should try to isolate these rhizosphere bacteria with specific functions and explore their effects on phytoremediation alone or combination. This will provide an important basis for the combined remediation of mine contaminated soil by rhizosphere bacteria and plants.",
"introduction": "Introduction Antimony (Sb) is an important trace element in the world economy with an annual production of about 150,000 t ( Shtangeeva, Bali & Harris, 2011 ). It is widely used in the production of ceramics, glasses, batteries, pyrotechnic materials, paints, ammunition, flame retardants, semiconductors, synthetic fabrics, etc. ( Wilson et al., 2010 ). Open pit Sb mining can damage the landscape and vegetation and contaminate soils with high levels of Sb ( e.g. , Sb(III), Sb(V)) and other heavy metals ( e.g. , As, Zn, Pb, Cd, and Hg) ( Okkenhaug et al., 2011 ; Yang, He & Wang, 2015 ; Tang et al., 2019 ; Zhou, Hursthouse & Chen, 2019 ). Sb contamination of soils inhibits plant growth ( Shtangeeva, Bali & Harris, 2011 ), influences the stucture, of rhizosphere soil microbial community ( Guo et al., 2019 ), and even causing health risk of teratogenesis and carcinogenesis to human body ( WHO, 2003 ). Given the toxicity and biological harm of Sb, Sb and its compounds have long been listed as priority pollutants by the environmental protection structure of European Union and United States ( He et al., 2012 ; Filella, Belzile & Chen, 2002 ). To reduce the ecological and health risks in Sb mining areas, it is critical to study appropriate environmental restoration measures for Sb-contaminated soil. Recently, phytoremediation has been widely used to remediate heavy metal contaminated soil. This low-cost approach has less environmentally impact than chemical remediation technology ( Boyd, 2007 ). However, the success of phytoremediation to pollution soils in metal mining areas depends not only on the selection of heavy metal enrichment-plants ( Orozco-Aceves, Tibbett & Standish, 2017 ; Audino, Louzada & Comita, 2014 ), but also on the selection of soil microorganism, especially rhizosphere soil microorganism, which are not as well studied ( Wei et al., 2015 ). Rhizosphere microorganisms play an important part in the process of plant-soil ecosystem processes, including nutrient cycling, energy transfer, metal resistance, and detoxification, as well as the establishment of sustainable plant communities ( Mishra, Singh & Arora., 2017 ). Many rhizospheric microorganisms, particularly some plant-growth-promoting rhizosphere bacteria, can increase biomass production and/or decrease the accumulation of heavy metal in plants ( Mishra, Singh & Arora., 2017 ; Lebeau, Braud & Jezequel, 2008 ). As one of the most abundant microbial groups in soil microorganism, rhizosphere bacteria actively participate in various biogeochemical reactions in rhizosphere and soil ( Mishra, Singh & Arora., 2017 ). Compared to non-rhizosphere bacteria in soil, rhizosphere bacteria have more direct effects on the growth and development of root ( Mishra, Singh & Arora., 2017 ). In addition, rhizosphere bacteria are more active and have high sensitivity to any small changes in environmental stress, which can be used as an early effective biological indicator to evaluate heavy metal pollution and plant growth status ( Deng et al., 2015 ; Muehe et al., 2015 ). Therefore, the examination of the diversity and community structure of indigenous rhizosphere bacteria under heavy metal stress at different vegetation restoration stages can provide valuable guidance for the remediation of Sb contaminated soil. However, studies on microorganisms in Sb mine area were mainly focused on soil microorganisms and the arbuscular mycorrhizal fungi of heavy metal hyperaccumulators. For example, Wang et al. (2018) found that the relative abundance and alpha diversity of bacterial communities in soil varied along with Sb-contaminated soil gradients. Another study on the molecular diversity and community composition of arbuscular mycorrhizal fungi in the rhizosphere of three heavy metal hyperaccumulators ( Miscanthus anderss, Boehmeria nivea and Cynodon dactylon ) was carried out in an Sb mining area ( Wei et al., 2015 ), but it only described the community structure of arbuscular mycorrhizal fungi in plant rhizosphere in detail, ignoring the rhizosphere bacteria. As we know, only one study investigated the changes of rhizosphere bacterial community during the remediation process of heavy metal enriched plants around Sb mining areas, and found that remediation affected rhizosphere bacterial diversity and most of rhizosphere bacteria belonged to the Acidobacteria , Bacteroidetes , Proteobacteria , and Actinobacteria ( Guo et al., 2019 ). At present, some artificial vegetation restoration measures have been carried out in Sb mines, but due to the cost and technical reasons, natural ecological restoration measure is the main method in maintaining ecosystem health ( Ding, Yang & Deng, 2013 ; Nakamaru & Martín Peinado, 2017 ; Sun, Li & Wu, 2021 ). Despite widespread recognition of natural restoration importance, little information is available on the diversity and community composition of rhizosphere bacterial communities on Sb contaminated soil during natural vegetation restoration progresses. This knowledge gap may hinder the application of rhizosphere bacteria in the remediation of Sb-contaminated soil. China has the most abundant Sb productions and reserves with 617 Sb mines in 18 provinces ( Ding, Yang & Deng, 2013 ). Xikuangshan mine (Hunan, China), the largest Sb mining in the world, known as the “the Sb capital of the world,” has about 1.1 million tons in reserve of Sb, accounting for about 80% of the country ( Tang et al., 2019 ). Mining activities over nearly 100 years have caused serious soil pollution in mining areas ( Okkenhaug et al., 2011 ; Mo et al., 2013 ), and the average content of Sb in soils in Xikuangshan mine was up to 4,368.222 mg/kg, which is far higher than that in other Chinese soil (about 1.06 mg/kg) and in global soil (about one mg/kg) ( Tang et al., 2019 ). The extremely high concentration of Sb makes the Xikuangshan mine an excellent model to study the changes in rhizosphere microbial composition during different natural restoration stages. We hypothesized that vegetation restoration could improve the diversity of rhizosphere microbial community and change the composition of rhizosphere microbial community. The aim of this present work was to observe and compare the changes of composition and diversity in rhizosphere microbial communities during natural restoration. This study is helpful to select the suitable rhizosphere microbial community to accelerate the ecological restoration process in Sb mines.",
"discussion": "Discussion The soil pollution caused by Sb mining in Xikuangshan is from Sb, As, Pb, Cd, Hg, Zn, and other heavy metals. Our investigation confirmed that the content of heavy metals in rhizosphere soil decreased significantly after long-term natural ecological restoration. Another study on the Xikuangshan mine also demonstrated that vegetation restoration is conducive to the reduction of soil heavy metal content ( Guo et al., 2019 ). One possible explanation is that some plants with heavy metal tolerance and enrichment ability are retained during natural recovery, and the reduction of heavy metals in rhizosphere may be due to plant absorption. Rhizosphere bacteria can support natural vegetation restoration by regulating biogeochemical processes, which can further help to improve the physical and chemical properties of rhizosphere soil ( Mishra, Singh & Arora., 2017 ). Our present study observed that the beta diversity index of rhizosphere microbial community varied with different restoration stages, Other studies have also observed that the establishment and development of plant communities could alter rhizosphere bacterial beta diversity ( Sun et al., 2018 ; Shrestha, Gautam & Ashwath, 2019 ). Different from the change of beta diversity, the alpha diversity (the Chao 1 index, Simpson index, Shannon index and Pielou index) under different restoration stages had various changes. Natural recovery significantly increased the Chao 1 index, but had no significant effect on the Simpson index, Shannon index and Pielou index, indicating that the main impact of natural restoration on rhizosphere bacteria was the number of rhizosphere microbes, not abundance or evenness. The alteration of rhizosphere bacterial diversity may be related to the change in plant species and rhizosphere soil habitats ( Guo et al., 2019 ; Shrestha, Gautam & Ashwath, 2019 ). At the phyla level, the ten dominant bacterial populations mainly included Actinobacteria , Proteobacteria , Acidobacteri a, Chloroflexi , Gemmatimonadetes , Bacteroidetes , Rokubacteria , Firmicutes , Patescibacteria , and Planctomycetes , indicating that these bacterial communities had strong adaptability and played an important role in three natural restoration stages (ER, MR, and LR) with Sb contaminated soil. Proteobacteria , the most abundant phyla in our study, exhibited significant survival and reproduction ability in mining area. This result is consistent with the researches in zinc mines ( Luo et al., 2018 ) and copper mines ( Liu et al., 2014 ; Sun et al., 2018 ), but not in gold mines where the predominant phyla is Acidobacteria ( Sibanda et al., 2019 ). In this study, natural vegetation restoration led to a significant decrease in the relative abundance of Patescibacteria in the LR, and an increase in the relative abundance of Actinobacteria . These results were somewhat inconsistent with the observation that vegetation reconstruction in an iron mine could lead to the increases of Patescibacteria and the decreases of Actinobacteria ( Deng et al., 2020 ). The reasons for this distinction may be attributed to the differences of vegetation and mining types. Additionally, three phyla ( Calditrichaeota , Deferribacteres , and Epsilonbacteraeota ) were only found in ER. Calditrichaeota is an independent phylum recently recognized, and it is an anaerobic bacteria with oxygen tolerance and protein fermentation ( Marshall et al., 2017 ). Deferribacteres mainly exists in the rhizosphere of specific restoration plant, such as Phragmites karka and Typha latifolia ( Singh & Singh, 2018 ). In addition, Deferribacteres , often occurring in refinery wastes, can be used for in-situ bioremediation in polluted environments ( Sarkar et al., 2016 ). Epsilonbacteraeota , a new microbial phyla in 2017, has a strong adversity tolerance, and mostly occurs in deep-sea hydrothermal vents and urban sewage ( Wang et al., 2020 ). In general, the bacteria of these three phyla have high stress tolerances and certain preference and capacity to exist in a polluted environment, which can explain the reason that they only appeared in the heavily polluted rhizosphere soil at ER. These findings that Calditrichaeota , Deferribacteres , and Epsilonbacteraeota only appeared in ER, but not in MR and LR, which has an important indicator significance for soil heavy Sb contamination. At the genus level, most of the dominant genera in rhizosphere during natural restoration stages are unclassified genera, which highlights the lack of knowledge about these less studied environments. The most abundant taxa in these environments included (1) RB41 , belongs to Acidobacteria and exhibits a high sensitivity to soil fertility ( Ai et al., 2018 ). RB41 plays a key role in maintaining soil metabolism and biogeochemical function under long-term low nutrient stress conditions ( Ai et al., 2018 ). RB41 is the key genus in cadmium contaminated soil under saline alkali stress ( Wang et al., 2019 ), and it is also the dominant genus in the petroleum-contaminated soil ( Shen et al., 2018 ) and the coal mining areas ( Sun et al., 2020 ). (2) MND1 is also the dominant bacterial communities in the coal mining areas ( Sun et al., 2020 ), the Weihe Terrace soil with lower concentrations of petroleum ( Shen et al., 2018 ) and the whole soil shifts with crop growth ( Hargreaves, Williams & Hofmockel, 2015 ). (3) Ellin6067 , an ammonia-oxidizing bacteria ( Xia et al., 2005 ), has a role in the degradation of xenobiotic and other complex organic compounds ( Lezcano et al., 2017 ). The other unclassified genera included IS-44 , SWB02 , Subgroup10 , CL500-29_marine_group , Mle1-7 were found in rhizosphere soil, though they have a relative low relative abundance. For example, SWB02 , a potential syntrophic bacteria, can establish magnetite-mediated direct electron transfer during the methanogenic degradation of volatile fatty acids ( Lee et al., 2019 ). CL500-29_marine_group , an actinomycete, can effectively utilize a variety of carbon-based compounds ( Lindh et al., 2015 ). Some of these unclassified taxa may originate from the original plants or air and insects and settle in rhizosphere soil through soil-root pathway. Further experiments are needed to better understand the succession of rhizosphere microbial community and their relationship between rhizosphere community and pollutant dynamics. Also, some dominant bacteria in different genera showed a diversity change in the natural recovery process. RB41 had the highest relative abundance in Sb mining, and this result was similar to Sun et al. (2018) , which found that RB41 was a dominant genus in copper mine tailings in Central China. These studies indicated that RB41 could play an important role in maintaining soil metabolism and biogeochemical functions under long-term high pollution stress. From ER-MR-LR, the relative abundance of RB41 and Haliangium increased significantly. RB41 is the key genus in cadmium contaminated soil under saline alkali stress ( Wang et al., 2019 ), and it is also the dominant genus in petroleum-contaminated soil ( Shen et al., 2018 ). RB41 has a high sensitivity to soil fertility ( Ai et al., 2018 ). With the natural restoration, compound pollution of heavy metals decreased, and soil fertility might be gradually enhanced. The improved living environment can increase significantly the abundance of RB41 . Haliangium is a genus of bacteria from the family of Kofleriaceae . It was found to significantly increase in the rhizosphere soil of continuous cropping strawberry and mango ( Li et al., 2019 ). As far as we know, this is the first time found that natural recovery could lead to a significant increase of Haliangium in Sb mining area, and the reasons need to be further studied. In addition, Bacillus decreased significantly from ER-MR-LR. Bacillus has a strong tolerance to heavy metals, and can effectively reduce the absorption of heavy metals by plants through bioaccumulation and bio-transformation involving redox reactions ( Ndeddy Aka & Babalola, 2016 ). Ramírez et al. (2019) have confirmed that the tolerance of Bacillus sp. MH778713 to Cr (VI) and Al can reach 15,000 mg/L and 10,000 mg/L, respectively. Bacillus megaterium H3 has the function of reducing Cr (VI) to Cr (III) with low toxicity, so as to reduce the harm of Cr to plants ( Wang et al., 2018 ). Furthermore, Bacillus thuringiensis X30 can decrease the toxicity of Cu and Pb to plants and increase the biomass of crops ( Han et al., 2018 ). Under Cd stress, Bacillus can promote rice growth, increase its biomass, photosynthetic pigment and micronutrient content, and reduce electrolyte permeability ( Jan et al., 2019 ). Based on these results, we believe that Bacillus has a certain preference for heavy metal pollution, which can reduce the ecological toxicity of heavy metal and promote the growth of plants. Therefore, the Bacillus genus should be preferred in the early stage of vegetation restoration in Sb mining. In addition, our results also showed that Bacillus was a node genus in the dominant seed network, which had a great significance for maintaining the composition and diversity of rhizosphere microorganisms. With natural recovery, the changes in the composition of rhizosphere microorganisms were partly attributed to the significant decrease of Bacillus . Also, with natural recovery, the abundance of Gaiella also showed a decreasing trend. Gaiella genus is sensitive to heavy metal pollution, and it is an important indicator bacteria in the local heavy metal remediation process ( Luciana et al., 2011 ). Furthermore, in this experiment, Azorhizobium was only detected in the ER phase. Azorhizobium can live in the plant rhizosphere in the lead-zinc mine tailings ( Yang et al., 1997 ). Some Azorhizobium ( e.g., Azorhizobium caulinodans ORS571) can colonize in plant roots and play a key role in nitrogen fixation in both symbiotic and aerobic free-living states ( Ryu et al., 2020 ). Based on this, we speculate that Rhizobium genus has high application value in plant recovery in Sb mining for it not only can resist the high concentration of heavy metal, but also can improve the nitrogen fixation ability of plants in poor soil. In sum, the abundance of rhizosphere bacteria were different in three natural restoration stages (ER, MR, and LR). Calditrichaeota , Deferribacteres , Epsilonbacteraeota phyla, and Azorhizobium genus only appeared in ER, and Bacillus and Gaiella genus had obviously high relative abundance in the ER. These rhizosphere bacteria had high tolerance and/or degradation ability to heavy metal pollution in Xikuangshan. The ways of bacteria tolerate and/or degrade heavy metals mainly included absorption, excretion, methylation, oxidation, and reduction of heavy metal, especially Sb ( Filella, Belzile & Lett, 2007 ). Therefore, we can artificially add these rhizosphere bacteria to the rhizosphere of other plants, which will promote remediation of pollution soil in Sb contaminated soil. With natural recovery, the plant rhizosphere bacteria in the mining area reduced the toxic effects of heavy metal on plant roots through their accumulation and transformation ( Mishra, Singh & Arora., 2017 ), resulting in the improvement of plant survival conditions and the change of vegetation types, and then changing the composition and abundance of rhizosphere bacteria ( Deng et al., 2015 ). The change of rhizosphere microbial community could also influence the absorption of heavy metal by plants ( Muehe et al., 2015 )."
} | 5,393 |
33564396 | PMC7848858 | pmc | 2,763 | {
"abstract": "In theory, neurons modelled as single layer perceptrons can implement all linearly separable computations. In practice, however, these computations may require arbitrarily precise synaptic weights. This is a strong constraint since both biological neurons and their artificial counterparts have to cope with limited precision. Here, we explore how non-linear processing in dendrites helps overcome this constraint. We start by finding a class of computations which requires increasing precision with the number of inputs in a perceptron and show that it can be implemented without this constraint in a neuron with sub-linear dendritic subunits. Then, we complement this analytical study by a simulation of a biophysical neuron model with two passive dendrites and a soma, and show that it can implement this computation. This work demonstrates a new role of dendrites in neural computation: by distributing the computation across independent subunits, the same computation can be performed more efficiently with less precise tuning of the synaptic weights. This work not only offers new insight into the importance of dendrites for biological neurons, but also paves the way for new, more efficient architectures of artificial neuromorphic chips.",
"introduction": "Introduction In theoretical studies, scientists typically represent neurons as linear threshold units (LTU; summing up the weighted inputs and comparing the sum to a threshold)\n 1 . Multiple decades ago, theoreticians exactly delimited the computational capacities of LTUs, also known as perceptrons\n 2 . LTUs cannot implement computations like the exclusive or (XOR), but they can implement all possible linearly separable computations and a sufficiently large network of LTUs can approximate all possible computations\n 3 . Research in computer science investigated the synaptic weight resolution required to implement linearly separable computations\n 4 ,\n 5 . Hastad\n et al . studied a computation implementable by an LTU only if its synaptic weight resolution grows exponentially with the number of inputs. We consider, similarly to these studies, the needed resources as the minimal size of integer-valued weights necessary to implement a set of linearly separable computations. Requiring a high synaptic resolution has important consequences. In the nervous system, neurons would need to maintain a large number of synapses or synapses with a large number of stable states. For the same reason, neuromorphic chips based on LTUs have to dedicate a large amount of resources to synapses\n 6 . We demonstrate here that dendrites might be a way to cope with this challenge. Dendrites are the receptive elements of neurons where most of the synapses lie. They turn neurons into a multilayer network\n 7 ,\n 8 because of their non-linear properties\n 9 ,\n 10 . These non-linearities enable neurons to perform linearly inseparable computations like the XOR or the feature binding problem\n 11 ,\n 12 . The non-linear integration also appears to be tuned for efficient integration of\n in vivo presynaptic activity\n 13 . In this study, we investigate whether dendrites can also decrease the synaptic resolution necessary to implement linearly separable computations. We address this question by looking at all the computations of three input variables implementable by an LTU with positive synaptic weights. We then extend the definition of one of these computations to an arbitrarily high number of inputs. Finally, we implement this computation in a biophysical neuron model with two passive dendrites using fewer synapses than an LTU. This work proposes a new role for dendrites in the nervous system, but also paves the way for a new generation of more cost-efficient artificial neural networks and neuromorphic chips composed of neurons with dendrites.",
"discussion": "Discussion In the present work, we extend the linear threshold unit (LTU) to the sub-linear threshold unit (SLTU), a more realistic neuron model that includes non-linear processing in dendrites. We compare these two models on the implementation of a simple computation, the D-AND. We define it for three inputs and then extend it to\n n inputs by keeping its two defining features: a single dominant input that needs to be activated together with at least one of the remaining inputs. In this extension, the synaptic heterogeneity - e.g. the number of distinct binary synapses - grows linearly with\n n in the case of an LTU implementation while all synaptic weights remain equal for an SLTU with two dendrites. For instance, if\n n = 1000 a single pre-synaptic input needs to make 999 synaptic contacts to implement the D-AND with a LTU while a single binary synapse suffices for a SLTU. This example demonstrates that a SLTU can implement the D-AND more efficiently -with less binary synapse - than the LTU. Our denomination of one input as “dominant” and the others as “non-dominant” in the definition of the D-AND relates to the distinction between “driver” and “modulator” inputs\n 20 . This concept, where driver inputs are necessary to activate a neuron, but this activity can be modulated by other inputs, is ubiquitous in the sensory system. For example, neurons in the primary visual cortex require a stimulus in their classical receptive field. Stimuli in the so-called extra-classical receptive field cannot activate the neuron by themselves, but strongly modulate the response if presented together with a stimulus in the classical receptive field\n 21 . This distinction is not entirely applicable for the D-AND, since the dominant input\n X \n 1 is not sufficient to activate the neuron by itself. Nevertheless, both computations rely on making a distinction between synaptic inputs, which can be implemented by placing inputs on different dendrites as we have shown in this study. We show in a previous study that STLUs enable one to robustly implement a computation\n 22 . In that study, an SLTU with eight dendrites implements direction selectivity while being resilient to massive synaptic failure. Alike the present work we exploited the placement of the synapses rather than the magnitude of their weight to implement the computation. Several properties of our biophysical model used here fit with experimental observations. Firstly, synapses at different positions tend to create the same depolarisation at the soma\n 18 . Secondly, while the depolarisation generated at a dendritic tip could be large (>50mV) the depolarisation recorded at the soma never exceeds 10mV. Finally, many experimental studies show examples of sub-linear summation in dendrites\n 8 ,\n 9 , notably in interneurons. How could neurons learn to implement the D-AND in an STLU? Multiple studies have shown that synaptic rewiring can happen at the sub-cellular level in a short time period\n 23 and that such a reorganisation could be used for learning\n 24 . This markedly differs from classic Hebbian learning which uses changes in the total synaptic weight to implement computations. Our biophysical model respects two important experimental observations. First, all synapses taken individually produce the same depolarisation at the soma, the so-called \"synaptic democracy\" like in\n 18 . Second, several experimental studies show examples of sub-linear summationin dendrites\n 10 , notably in interneurons\n 8 ,\n 9 . How could neurons learn to implement the D-AND in an SLTU? Multiple studies have shown that synaptic rewiring can happen at the sub-cellular level in a short time period\n 23 and that such a reorganisation could be used for learning\n 24 . This markedly differs from classic Hebbian learning which uses changes in the total synaptic weight to implement computations, a SLTU friendly learning algorithm would keep the total synaptic weight constant while changing the targeted dendrites. Our findings also have implications beyond neuroscience, in particular for engineering applications. Studies in computer science assert that even problems solvable by an LTU might not have a solution when weights have a limited precision\n 25 . Being able to implement computations with an SLTU is therefore advantageous for hardware with limited resources. In conclusion, dendrites unlock computations inaccessible without them and allow one to more efficiently implement the accessible ones. For instance, to implement the D-AND when n=1001 a SLTU needs a single synaptic contact for the dominant input while a LTU requires a thousand. Dendrites enable us to do more with less."
} | 2,115 |
25538702 | PMC4258642 | pmc | 2,764 | {
"abstract": "Precambrian Banded Iron Formation (BIF) deposition was conventionally attributed to the precipitation of iron-oxides resulting from the abiotic reaction of ferrous iron (Fe(II)) with photosynthetically produced oxygen. Earliest traces of oxygen date from 2.7 Ga, thus raising questions as to what may have caused BIF precipitation before oxygenic photosynthesis evolved. The discovery of anoxygenic phototrophic bacteria thriving through the oxidation of Fe(II) has provided support for a biological origin for some BIFs, but despite reports suggesting that anoxygenic phototrophs may oxidize Fe(II) in the environment, a model ecosystem of an ancient ocean where they are demonstrably active was lacking. Here we show that anoxygenic phototrophic bacteria contribute to Fe(II) oxidation in the water column of the ferruginous sulfate-poor, meromictic lake La Cruz (Spain). We observed in-situ photoferrotrophic activity through stimulation of phototrophic carbon uptake in the presence of Fe(II), and determined light-dependent Fe(II)-oxidation by the natural chemocline microbiota. Moreover, a photoferrotrophic bacterium most closely related to Chlorobium ferrooxidans was enriched from the ferruginous water column. Our study for the first time demonstrates a direct link between anoxygenic photoferrotrophy and the anoxic precipitation of Fe(III)-oxides in a ferruginous water column, providing a plausible mechanism for the bacterial origin of BIFs before the advent of free oxygen. However, photoferrotrophs represent only a minor fraction of the anoxygenic phototrophic community with the majority apparently thriving by sulfur cycling, despite the very low sulfur content in the ferruginous chemocline of Lake La Cruz.",
"conclusion": "Conclusion Chemical profiles of iron with a recurrent secondary peak of Fe(III) in the anoxic, euphotic part of the chemocline, along with the increased inorganic carbon-uptake in presence of Fe(II), and the light-dependent Fe(II) oxidation, provide consistent evidences for photoferrotrophic activity in the La Cruz chemocline, and represent a proof of concept for their possible contribution to ancient BIF formation prior to the evolution of oxygenic photosynthesis. However, we note that photoferrotrophy under the prevailing environmental conditions in La Cruz is a slow process and that most of the Fe(II) at the chemocline is oxidized by molecular oxygen. Moreover, photoferrotrophs represent only a minor fraction of the anoxygenic phototrophic community with the majority apparently thriving by sulfur cycling, despite the very low sulfide and sulfate contents in the ferruginous water column of Lake La Cruz. This observation is also supported by a recent publication that showed that a cryptic sulfur cycle can occur in ferruginous conditions (Crowe et al., 2014 ). We hypothesize therefore that cryptic sulfur cycling, as recently shown for oxygen minimum zones in upwelling areas of the modern Ocean (Canfield et al., 2010 ), was also a feature of the late Archean Ocean where predicted sulfate concentrations were 3–10 times higher than in Lake La Cruz (Canfield, 2005 ; Jamieson et al., 2013 ). 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 The chemistry of the anoxic Archean ocean was characterized by a low sulfate content and high concentrations of ferrous iron (Fe(II)) of probable hydrothermal origin (Holland, 1973 ; Anbar and Knoll, 2002 ; Canfield, 2005 ). From this ferruginous water column, alternating sedimentary deposits of iron oxide minerals and silica precipitated between 3.8 and 1.8 Ga ago (Anbar and Knoll, 2002 ) and became preserved in the geological record as Banded Iron Formations (BIF). The mechanisms of Fe(II) oxidation are still debated and include, in addition to the widely accepted abiotic reaction with photosynthetically produced oxygen (Cloud, 1968 ), photocatalytic oxidation by UV radiation (Braterman et al., 1983 ), and direct oxidation by anoxygenic photosynthesis (Konhauser et al., 2002 ; Kappler et al., 2005 ). Such photoferrotrophic bacteria use light as energy and Fe(II) as an electron source for carbon fixation and biomass formation (Widdel et al., 1993 ; Heising et al., 1999 ) (Equation 1, Table 1 ). Table 1 Simplified stoichiometries of photoferrotrophic primary production (Equation 1) and different anaerobic modes of organic matter degradation (Equation 2: Fe(III)-respiration, Equation 3: sulfate-respiration, Equation 4: methanogenesis) . Equation 1 2 CO 2 + 8 Fe 2+ + 14 H 2 O 2 CH 2 O + 8 FeOOH + 16 H + Equation 2 2 CH 2 O + 8 FeOOH + 8 H + 2 CO 2 + 8 Fe 2+ + 14 H 2 O Equation 3 2 CH 2 O + SO 2− 4 + 2 H + 2 CO 2 + H 2 S + 2 H 2 O Equation 4 2 CH 2 O CO 2 + CH 4 Equation 1 + 3 a 8 Fe 2+ + SO 2− 4 + 12 H 2 O 8 FeOOH + H 2 S + 14 H + Equation 1 + 4 CO 2 + 8 Fe 2+ + 14 H 2 O CH 4 + 8 FeOOH + 16 H + Assuming complete degradation of the biomass produced in Equation 1, only degradation through sulfate-reduction (Equation 1 + 3) and methanogenesis (Equation 1 + 4) will result in a net accumulation of Fe(III)-oxide minerals in the sediment. The fraction of degraded organic matter released as methane into the atmosphere represents a loss of reducing power and will consequently determine the amount of Fe(III) accumulating in the sediment . a Sulfide will react chemically with excess Fe(III)oxides to form elemental sulfur and eventually pyrite, resulting in partially sulfidized Fe(III)-oxide deposits. Additional input of organic matter from anoxygenic photosulfidotrophs and oxygenic cyanobacteria would generally stimulate anaerobic degradation processes (Equations 2, 3, 4) and increase the degree of sulfidization of the sedimentary Fe-pool . While recent experimental work is not in favor of a significant contribution of photochemical processes to BIF formation (Konhauser et al., 2007 ), microbial Fe(II) oxidation remains an appealing possibility (Konhauser et al., 2002 ; Posth et al., 2008 ), especially for periods prior to the evolution of oxygenic photosynthesis. Although the evolution of photosynthesis is complex with horizontal gene transfer playing an important role, it is now accepted that anoxygenic phototrophic bacteria evolved before oxygen-producing cyanobacteria (Xiong et al., 2000 ; Raymond et al., 2003 ). The isolation of phototrophic Fe(II)-oxidizing bacteria (Widdel et al., 1993 ; Heising et al., 1999 ) has allowed the study of the influence of light intensity on iron oxidation and the role of temperature on the alternating precipitation of iron oxides and silica (Posth et al., 2008 ). Yet experimental field work aimed at elucidating the role of phototrophic Fe(II)-oxidation under natural environmental conditions as they may have existed in the chemocline of an Archean ferruginous ocean is still strongly needed (Johnston et al., 2009 ; Severmann and Anbar, 2009 ). Ferruginous water columns are rare, largely unexplored ecosystems of which only freshwater representatives exist today because of the high sulfate concentrations in the modern ocean. Recently, the presence of green anoxygenic phototrophic bacteria in the water column of a late Archean Ocean analog (Lake Matano, Indonesia) has been reported, and it has been suggested that they may be involved in Fe(II)-oxidation (Crowe et al., 2008 ). However, the respective activity could not be shown unambiguously and direct evidence for photoferrotrophic activity in a recent water column and quantitative data on their contribution to Fe(II)-oxidation are lacking to date (Crowe et al., 2014 ). To address this, we investigated microbial iron cycling in the water column of Lake La Cruz (Rodrigo et al., 2001 ) in the Central Iberian Ranges (Spain), a permanently stratified lake ecosystem with a chemocline in the euphotic zone and a water column chemistry matching the putative late Archean conditions (Table 2 ). Using combined microbiological and biogeochemical approaches we assessed whether anoxygenic phototrophic Fe(II)-oxidizing microorganisms (photoferrotrophs) indeed thrive in the chemocline of an Archean ocean analog and contribute to the production of Fe(III) in an anoxic environment. Table 2 Comparison of water chemistry, iron fluxes toward the oxic/anoxic interface, and Fe(II)-oxidation rates of different potential modern Archean Ocean analogs . Lake La Cruz Lake Matano (Crowe et al., 2008 ) Lake Pavin (Bura-Nakic et al., 2009 ) Archean Ocean Mixed layer Anoxic layer Anoxic layer Anoxic layer Anoxic layer Fe(II) (μM) – 230 140 1000 40–120 (Crowe et al., 2008 ) SO 2− 4 (μM) <35 <25 <0.1 <5.0 ≈80 (Jamieson et al., 2013 ) O 2 (μM) 238 0 0 0 <0.03 (Crowe et al., 2008 ) PO 3− 4 (μM) <0.26 <1.6 9 – 0.03–0.29 (Crowe et al., 2008 ) pH 8.60 7.00 7.00 6.08 >6.5 (Crowe et al., 2008 ) T° (°C) 16 6 25–28 4 ≈36 (Crowe et al., 2008 ) Fe(II) flux (μmol cm 2 d −1 ) 0.031–0.244 0.034–0.27 – 12.3 a Estimated In-situ Fe(II) oxidation rate (μmol l −1 d −1 ) 0.174–1.396 b 0.034–0.27 – 14 (Kappler et al., 2005 ) Fe(II) oxidation rate (“ 14 C”) (μmol l −1 d −1 ) 2.56 c Fe(II) oxidation rate (“ ex-situ ”) (μmol l −1 d −1 ) 63.6 d – – Number of cells (cell ml −1 ) 0.5 × 10 5 , 7.0 × 10 5 e 0.3–16 × 10 9 – 10 6 (Kappler et al., 2005 ) a Calculation based on the surface of Hamersley Basin (10 11 m 2 ) and a maximum Fe(III) precipitation rate of 4.5 × 10 12 mol Fe(III) y −1 required to form the Hamersley Basin BIFs (Kappler et al., 2005 ) . b Calculation see Table 3 . c Calculation based on the amount of 14 C-bicarbonate fixed in “Fe(II) + DCMU” treatments (1.36 μg C l −1 h −1 , Figure 3 ) minus the “No addition + DCMU” treatment (0.72 μg C l −1 h −1 ) and assuming a ratio of 4 Fe(II) oxidized per CO 2 assimilated as shown in equation 1 (Table 1 ). Calculated for 12 h illumination per day . d Calculation based on the average iron-oxidation rate of 2.65 μmol l −1 h −1 determined in the ex-situ light incubation with 12 h of illumination per day ( Figure 5 ) and assuming that all Fe(II) was oxidized through photoferrotrophy . e Maxima of GSB and PSB microscopic cell counts during summer stratification (Figure 1B ) .",
"discussion": "Results and discussion Most meromictic lakes and other permanently stratified water bodies are euxinic, i.e., anoxic and sulfidic below the chemocline (Lyons et al., 2009 ). Lake La Cruz is a rare exception as its anoxic bottom water contains little sulfide but is rich in dissolved Fe(II) (Figure 1A ). Sulfate concentrations are low (<35 μM; Figure 1A ) due to the low sulfur content of the surrounding dolomite rocks and marlstones (Rodrigo et al., 2001 ). Dissolved sulfide was detected just below the chemocline, originating from decaying organic matter (Romero-Viana et al., 2010 ) and dissimilatory sulfate reduction as indicated by decreasing sulfate concentrations with depth and 16S rRNA gene sequences of sulfate -reducing bacteria ( Desulfomonile sp.) in a clone library from the anoxic part of the chemocline (Walter, 2011 ). Sulfide concentrations were very low, even though they were determined with the photometric Cline assay, which overestimates free sulfide concentrations as it detects also colloidal and amorphous forms of FeS (Bura-Nakic et al., 2009 ). Most of the iron in the anoxic water column was thus present as dissolved Fe(II). Fe(II) reached the chemocline at a rate of 0.031–0.244 μmol cm 2 d −1 (Tables 1 , 2 ), where it was oxidized as indicated by two separate Fe(III) maxima (Figure 1A ). Figure 1 Summer stratification water column data of Lake La Cruz. (A) Chemical stratification at the time when incubation experiments were performed (13/Oct/2008). The insert depicts a magnified Fe(III) concentration profile around the chemocline. (B) Vertical distribution of oxygenic (chlorophyll a concentration, open circles) and anoxygenic phototrophs. Black symbols stand for microscopic cell counts of the dominant purple sulfur bacterium (PSB) Lamprocystis purpurea) and gray symbols for the dominant green sulfur bacterium (GSB) Chlorobium clathratiforme . The horizontal line at 11.8 m indicates the sampling depth for the in situ radiocarbon-incubations and ex situ Fe(II)-oxidation experiments. An upper broad peak of 2.8 μM Fe(III) was located between 10 and 11.75 m, where picocyanobacteria were most abundant and Fe(III) was presumably formed by direct chemical reaction with O 2 or by microaerophilic chemotrophs (Lehours et al., 2007 ). As nitrate was also present at that depth (2.0 μM), chemotrophic nitrate-dependent iron-oxidation cannot be excluded, but appeared to be of minor importance (Walter, 2011 ), possibly due to the limited supply of this oxidant as well as competition for nitrate with denitrifying bacteria and nitrate-assimilating phototrophs. A second peak of Fe(III) (~2.0 μM, 12–13 m) was typically observed in the anoxic part of the chemocline and coincided with the biomass maxima of the anoxygenic phototrophs Chlorobium clathratiforme and Lamprocystis purpurea (Figures 1B , 8 ). Prevailing light intensities of 0.02–0.002% PAR (Figures 1B , 2 ) and a continuous supply of Fe(II) from the hypolimnion constitute suitable conditions for the development of photoferrotrophs. Figure 2 Depth profiles of physico-chemical conditions in the water column of Lake La Cruz . Profiles were recorded during summer stratification conditions (13/Oct/2008). Photoferrotrophy being an autotrophic metabolism, in-situ \n 14 C-incubation experiments were conducted to detect any Fe(II)-dependent stimulation of carbon uptake in the light. Incubations were performed with water samples from the anoxic part of the chemocline at 11.8 m, where sulfide and Fe(II) concentrations were minimal. The incubations were amended with various electron donors and acceptors, which may fuel autotrophic metabolism, including Fe(II), sulfide, and nitrate. Since both oxygenic and anoxygenic phototrophs were present at this depth (Figures 1B , 8 ), DCMU (3-{3,4-dichlorophenyl}-1,1-dimethylurea) was added to parallel incubations in order to suppress oxygen production by PSII. In the absence of DCMU, none of the additions had a statistically significant influence on carbon uptake (Figure 3 ). The inorganic carbon-uptake rates were generally 35–40% lower in the presence of DCMU, with the only exception of the Fe(II) treatment where a significant increase of 40% was observed ( P < 0.05) and where the highest fixation rates were determined among all assayed conditions (Figure 3 ). Figure 3 Light-dependent inorganic carbon uptake in presence of potential substrates for lithoautotrophic iron- and sulfur-transforming microorganisms . In-situ net photosynthetic carbon fixation experiments showing a statistically significant stimulation upon Fe(II) addition in presence of DCMU (- c -, p < 0.05) that inhibits oxygen production by photosynthesis. The net photosynthetic carbon uptake was obtained by subtracting the dark incubation values from the carbon uptake in the light. Letters represent the statistical groups using generalized linear models. Average values and standard deviations are presented (13/Oct/2008). This result suggests that there was Fe(II)/light-dependent inorganic carbon uptake in the anoxic part of the chemocline, and that oxygenic photosynthesis needed to be inhibited to detect photoferrotrophic autotrophy. It has been proposed that ancestral cyanobacteria could photosynthesize with PS I alone and probably used H 2 , H 2 S, or Fe(II) to reduce CO 2 to organic matter (Pierson, 1994 ). Also some modern cyanobacteria may switch in response to the environmental conditions from oxygenic to anoxygenic photosynthesis with H 2 S (Cohen et al., 1986 ). A contribution of cyanobacteria to the observed stimulation in presence of DCMU cannot be entirely excluded, although such metabolic versatility was never found in picocyanobacteria (Pierson et al., 1999 ) such as those being abundant in Lake La Cruz. Similar incubations done during winter, when the different phototrophic populations were less compact and better separated in the water column, revealed that fuelling of 14 C-uptake by Fe(II) was strongest in water layers where anoxygenic phototrophs were present (15 and 15.5 m). Conversely, only a weak stimulation was observed in samples from the upper cyanobacterial layer (13.5 m) or the aphotic monimolimnion (17 m), respectively (Figure 4 ). Figure 4 In situ light dependent inorganic carbon uptake in presence of potential substrates for iron- and sulfur-transforming microorganisms . Results show highest stimulation of bicarbonate uptake upon Fe(II) addition at 15 m and 15.5 m, the depths where anoxygenic phototrophs were most abundant. Net photosynthetic carbon uptake was obtained by subtracting the dark incubation values from the carbon uptake in the light. N-serve and DCMU were added to inhibit nitrification and PS(II) of oxygenic phototrophs, respectively. Means of duplicate experiments are given, with bar ends indicating highest and lowest values. Winter stratification conditions (11/Feb/2008). No data on rates of anoxygenic phototrophic Fe(II)-oxidation in an Archean Ocean analog existed so far. Anaerobic light dependent oxidation of iron was quantified using ex-situ incubations with Fe(II)-enriched water from the same depth as used for the 14 C-incubations (Figure 5 ). The oxidation of Fe(II) to Fe(III) by the natural chemocline microbiota was light dependent and occurred at a potential rate of 2.7 μmol l −1 h −1 . The rate was similar for incubations with and without DCMU, alluding to the fact that oxygenic phototrophs (e.g., picocyanobacteria) from this depth were photosynthetically not very active, as also demonstrated by parallel studies investigating the annual cycle of inorganic carbon assimilation (Picazo, personal communication). Moreover, this estimate shows that the population size of photoferrotrophs in the chemocline was high enough to reach, under non limiting light conditions, Fe(II) oxidation rates believed to be required for BIFs formation (Kappler et al., 2005 ) (Table 2 ). Alternative Fe(II) oxidation rate estimates based on equation (1) and the amount 14 CO 2 incorporated in situ , under Fe(II)-enriched conditions, amounted only to 0.2 μmol Fe(II) l −1 h −1 (Table 2 ). This was likely due to the lower light intensities available in the lake. Alternatively, it may also indicate that photoferrotrophs in Lake La Cruz are not obligate autotrophic organisms and that they could assimilate additional organic compounds for biomass formation (Heising et al., 1999 ). Our attempts to cultivate phototrophic Fe(II)-oxidizing organisms from chemocline water samples resulted in a co-culture consisting of Chlorobium sp. (Figure 6 , 80%) and as yet uncultivated Acidobacteria (20%). The Chlorobium strain was closely related to Chlorobium ferrooxidans , the only green photoferrotrophic culture known so far, and to Chlorobium clathratiforme , the dominant green phototrophic sulfur bacterium in La Cruz (Rodrigo et al., 2001 ; Romero-Viana et al., 2010 ). The enrichment culture contained both photoferrotrophic and Fe(III)-reducing bacteria (Figure 7 ) suggesting that Fe(II)-oxidizing and reducing processes in the chemocline of Lake Cruz are tightly coupled. The observed Fe(II) oxidation rate of 2.6 μmol l −1 h −1 represents thus a net rate, depending on the relative kinetics of the processes, and falls into the lower range of what has been determined in other cultures of Fe(II)-oxidizing phototrophs (Hegler et al., 2008 ). Figure 5 Light-dependent iron-oxidation by the natural microbiota from the anoxic part of the Lake La Cruz chemocline at 11.8 m depth . Anoxic laboratory ( ex situ ) incubation experiment (11/Feb/2008) (A) without DCMU addition, where oxygenic photosynthesis is active and abiotic Fe(II) oxidation with O 2 may occur; and (B) Fe(II) evolution under anoxic conditions in absence of oxygenic photosynthesis (with DCMU addition). For both experimental settings, light conditions involved consecutive periods of 12 h illumination at 61 μE m −2 s −1 and 12 h darkness. Solid symbols stand for Fe(II) and open symbols for Fe(III) concentrations. The dotted lines represent means of killed controls (stars; n = 3). Figure 6 Phylogeny of the photoferrotrophic enrichment culture based on nearly complete 16S rRNA gene sequences and maximum likelihood analysis . Among the retrieved sequences 80% formed a well-defined cluster within the Chlorobia (25/31 clones), phototrophic green sulfur bacteria; the rest of the sequences were most closely related to uncultivated Acidobacteria (5/31 clones). Numbers in brackets signify the number of clones possessing the same sequence. Bootstrap values >50% for 1000 replications are shown. Figure 7 Net Fe(II)-oxidation by the photoferrotrophic enrichment culture . Results display the dependence of Fe(II)-oxidation on light. Furthermore, Fe(III) added with the inoculum at the beginning of the experiment was reduced during the first 8 h, presumably mediated by Acidobacteria , the second most abundant bacterial group in the enrichment culture. Data are average values of three independent experiments with error bars representing standard deviations ( n = 3). Figure 8 Photomicrographs of (A) green and (B,C) purple anoxygenic phototrophs from the chemocline of Lake La Cruz . They show yellowish sulfur globules inside of PSB cells or deposited around GSB cells (A,B) . PSB cells without apparent internal S 0 deposits are also frequently observed (C) . Scale bars represent 10 μm. Despite complete oxidation of Fe(II) at the chemocline and hence a continuous flux of sedimenting Fe(III) to the bottom of the lake (Figure 1A ), there was no accumulation of iron oxide minerals in the sediments. Only 3 μmol Fe(III) g −1 w/w was detected at the sediment surface (data not shown). Below 0.5 cm depth, HCl extractable iron was reduced and bound to sulfur as suggested by combined iron and sulfur measurements (data not shown). Iron sulfide is produced continuously just below the chemocline and settles down the water column (Ma et al., 2006 ), however, in contrast to Fe(III)-oxides, which can be reduced back to Fe 2+ , FeS is stable and will accumulate in the sediment. Iron sulfide minerals are also produced at the sediment surface from Fe(III)-oxides reacting with sulfide liberated through organic matter degradation and dissimilatory sulfate reduction. An accumulation of Fe(III) in the sediment would only be possible with a strongly reduced primary productivity and significant degradation of organic matter by methanogenesis during sedimentation (Table 1 ). We propose therefore that periods of BIF formation under anoxic Archean conditions were associated with increased methane formation, which is consistent with the current notion of biogenic methane being an important component of the Archean atmosphere (Zerkle et al., 2012 )."
} | 5,782 |
26413297 | null | s2 | 2,765 | {
"abstract": "Growing evidence supports a critical role of dynamic metal-coordination crosslinking in soft biological material properties such as self-healing and underwater adhesion"
} | 42 |
19505144 | null | s2 | 2,766 | {
"abstract": "Natural materials employ many elegant strategies to achieve mechanical properties required for survival under varying environmental conditions. Thus these remarkable biopolymers and nanocomposites often not only have a combination of mechanical properties such as high modulus, toughness, and elasticity, but also exhibit adaptive and stimuli-responsive properties. Inspired by skeletal muscle protein titin, we have synthesized a biomimetic modular polymer that not only closely mimics the modular multidomain structure of titin, but also manifests an exciting combination of mechanical properties, as well as adaptive properties such as self-healing and temperature-responsive shape-memory properties."
} | 175 |
37941766 | PMC10629970 | pmc | 2,767 | {
"abstract": "Wearable thermoregulatory technologies have attracted widespread attention because of their potential for impacting individual physiological comfort and for reducing building energy consumption. Within this context, the study of materials and systems that can merge the advantageous characteristics of both active and passive operating modes has proven particularly attractive. Accordingly, our laboratory has drawn inspiration from the appearance-changing skin of Loliginidae (inshore squids) for the introduction of a unique class of dynamic thermoregulatory composite materials with outstanding figures of merit. Herein, we demonstrate a straightforward approach for experimentally controlling and computationally predicting the adaptive infrared properties of such bioinspired composites, thereby enabling the development and validation of robust structure–function relationships for the composites. Our findings may help unlock the potential of not only the described materials but also comparable systems for applications as varied as thermoregulatory wearables, food packaging, infrared camouflage, soft robotics, and biomedical sensing.",
"introduction": "INTRODUCTION Wearable materials and systems have attracted much attention for applications as varied as fitness tracking, 1,2 medical monitoring, 3,4 safety and security assurance, 5,6 communication and education, 7,8 energy harvesting or storage, 9,10 and personal thermal management. 11–14 Within this context, the development of wearable thermoregulatory technologies has become a major focus in both academia and industry because of their potential for impacting individual physiological comfort and for reducing energy consumption upon widespread adoption. 13–20 Typically, such thermoregulatory technologies have been broadly classified as passive or active based on their mode of operation. 11–14,21–24 For instance, passive technologies are designed to statically regulate heat transfer without any energy input, often making them straightforward to implement, relatively low cost, and quite energy efficient, but they generally exhibit poor adaptability to changes in the external environment. 11–14,21,22 In contrast, active technologies are designed to dynamically regulate heat transfer with a substantial energy input, often making them complex to implement, comparatively higher cost, and less energy efficient, but they generally feature excellent controllable responsiveness to changes in the external environment. 11–14,23,24 As such, there exists powerful motivation for the study of wearable thermoregulatory materials and systems that can merge the advantages of both passive and active operating modes. Recently, our laboratory has introduced a unique new class of thermoregulatory composite materials, 25,26 which were engineered by drawing inspiration from the fascinating appearance-changing capabilities of Loliginidae (inshore squids) [ Fig. 1(a) ]. 25–31 In particular, we considered the natural architecture of squid skin [ Fig. 1(b) ], in which organs called chromatophores are reversibly expanded and contracted via muscle action [ Fig. 1(c) ], 27–29 and we accordingly designed artificial infrared-reflecting metal–polymer composite materials [ Fig. 1(d) ] , for which the overlaid metal layer's microstructure is reversibly reconfigured via mechanical actuation [ Fig. 1(e) ]. 25,26 Excitingly, our bioinspired designer composite materials not only could alter their infrared transmittance by ≳20-fold but also could regulate heat fluxes by ≳30 W m −2 with a minimal mechanical power input. 25,26 Additionally, when integrated into compact wearable sleeve-type devices, the materials could modulate localized body temperature changes by up to ∼10-fold as a result of actuation with applied strain. 25 Moreover, large-area variants of such materials were scalably manufactured via standard industrial techniques at a low estimated cost of ∼US $0.1 m −2 . 26 However, for our composite materials, we did not previously showcase precise control over the surface microstructure, establish detailed general structure–function relationships, or develop computational methods for the prediction of their dynamic infrared properties. 25,26 FIG. 1. Squid skin-inspired design of the adaptive infrared composite materials. (a) Digital camera pictures of an opalescent squid changing its appearance. (b) A simplified cross-sectional illustration of the general natural architecture of squid skin, which shows the epidermis, chromatophore layer, iridophore layer, and musculature. (c) A top-view illustration of organs called chromatophores that are reversibly expanded and contracted via muscle action. 29 (d) A cross-sectional illustration of the composite material, which shows the planar Cu layer, the embedded Cu nanostructures, and the polymer matrix. (e) A top-view illustration of the squid skin-inspired composite material for which the surface microstructure is reversibly reconfigured via mechanical actuation. 25 Note that the pictures in (a) are reproduced with permission from S. Thiebaud, “Opalescent Inshore Squid ( Doryteuthis opalescens ),” iNaturalist, https://www.inaturalist.org/observations/65343592 (2020). Copyright 2020, Authors, licensed under a Creative Commons Attribution license. 31 Herein, we demonstrate a straightforward approach for experimentally controlling and computationally predicting the adaptive infrared properties of our wearable bioinspired composites, thereby enabling the development and validation of robust structure-function relationships for these materials. First, we fabricate composites for which infrared-reflecting planar metal layers with variable thicknesses are overlaid on polymer matrices. Next, we evaluate our composites' strain-reconfigurable microstructural characteristics, i.e., average metal island widths and fractional metal surface coverages. In turn, we characterize our composites' strain-dependent infrared functionalities, i.e., total transmittances and reflectances. Last, we computationally simulate the composites' adaptive infrared properties. Overall, our findings may guide the continued engineering and optimization of both our composite materials and analogous systems for applications as varied as thermoregulatory wearables, food packaging, infrared camouflage, soft robotics, and biomedical sensing. Facile fabrication of the bioinspired composite materials We began our efforts by fabricating composite materials consisting of nanostructured metal films embedded within an elastomeric polymer matrix, as illustrated in the supplementary material, Fig. 1. The scanning electron microscopy (SEM) images of the substrate-bound nanostructured metal films are shown in the supplementary material, Fig. 2, and the digital camera pictures and SEM images of the free-standing composite materials are shown in the supplementary material, Fig. 3. To fabricate the nanostructured metal films, we deposited infrared-reflecting planar copper (Cu) layers with variable thicknesses onto support substrates and then grew tilted columnar Cu nanostructures on top of these planar layers [supplementary material, Fig. 1(a)]. The corresponding SEM images revealed that such films consisted of planar layers with the expected thicknesses of ∼5, ∼10, ∼20, ∼50, and ∼100 nm and arrayed tilted columnar nanostructures with the anticipated heights of ∼90 nm [supplementary material, Fig. 2]. To fabricate the composite materials, we spin-coated infrared-transparent styrene–ethylene–butylene–styrene (SEBS) polymer matrices directly onto the substrate-bound nanostructured films and then delaminated the resulting architectures from the support substrates [supplementary material, Fig. 1(b)]. The corresponding digital camera pictures and SEM images revealed that such composites were globally relatively uniform and featured locally fractured topmost metal layers (supplementary material, Fig. 3). Notably, the composites' planar Cu layers could be removed via chemical treatment, confirming the presence of the embedded columnar Cu nanostructures within the polymer matrices (supplementary material, Fig. 4). 32 Moreover, composites fabricated from planar Cu layers without nanostructures could not be reliably delaminated from the support substrate, resulting in materials with millimeter-scale defects (supplementary material, Fig. 5). Last, SEBS polymer matrices fabricated without planar Cu layers were readily delaminated from the support substrates, resulting in transparent films with no obvious large defects (supplementary material, Fig. 6). Together, our high-yield and versatile process yielded free-standing composite materials with relatively large areas of >160 cm 2 , thereby facilitating subsequent morphological and spectroscopic characterization. Microstructural evaluation of the bioinspired composite materials After fabricating our composite materials, we qualitatively evaluated their strain-reconfigurable surface microstructures, which are illustrated in Fig. 2(a) . The representative SEM images obtained for composite materials with variable planar layer thicknesses of ∼5, ∼10, ∼20, ∼50, or ∼100 nm and subjected to different uniaxial strains of 0%, 30%, 50%, or 100% are shown in Fig. 2(b) . In their relaxed states (i.e., under a strain of 0%), the composites' surfaces consisted of abutted metal domains (islands) that completely covered the underlying polymer matrices, but in their actuated states (i.e., under strains of 30%, 50%, or 100%), the composites' surfaces consisted of separated metal islands that only partially covered the underlying polymer matrices [ Fig. 2(b) ]. Here, the composites with variable planar layer thicknesses analogously featured fractional metal surface coverages that progressively decreased with the applied strain but also did exhibit some noteworthy qualitative differences in their microstructural characteristics [ Fig. 2(b) ]. Specifically, the surfaces of the composites with 5 and 10 nm planar layer thicknesses were covered by small metal islands and some interspersed defects, presumably due to incomplete delamination during fabrication; the surfaces of the composites with 20 nm planar layer thicknesses were covered by intermediate-sized islands and few-to-no defects, presumably due to optimum delamination during fabrication; and the surfaces of the composites with 50 and 100 nm planar layer thicknesses were covered by large metal islands with occasional raised edges, presumably due to partial metal debonding after delamination [ Fig. 2(b) ]. Notably, the different types of composites all featured analogous mechanical properties (i.e., Young's moduli of ∼1–∼2 MPa and elongations to break of >900%), which were seemingly primarily dictated by the rubber-like SEBS polymer matrix and were only somewhat affected by the thicknesses of the overlaid planar Cu layers (supplementary material, Fig. 7). These experiments suggested that the morphological characteristics of our composite materials were primarily determined by a single adjustable parameter, i.e., the thickness of their planar Cu layer. FIG. 2. Surface microstructure of the composite materials. (a) An illustration of the composite materials with planar layer thicknesses of 5, 10, 20, 50, and 100 nm (from left to right) before (top) and after (bottom) mechanical actuation. (b) Representative top-down SEM images of the composite materials with planar layer thicknesses of 5, 10, 20, 50, and 100 nm (from left to right) under applied strains of 0%, 30%, 50%, and 100% (from top to bottom). (c) The average metal island widths for composite materials with planar layer thicknesses of 5, 10, 20, 50, and 100 nm under applied strains of 0%, 30%, 50%, and 100%. The red line corresponds to a linear fit of the data. (d) The average fractional metal surface coverages for composite materials with planar layer thicknesses of 5, 10, 20, 50, and 100 nm under applied strains of 0%, 30%, 50%, and 100%. The error bars in (c) and (d) represent the standard deviations of the mean. We next quantitatively evaluated the strain-reconfigurable surface microstructures of our composite materials, as illustrated in the supplementary material, Fig. 8 (see Methods for further details). The average metal island widths and average fractional metal surface coverages calculated for composite materials featuring variable planar layer thicknesses and subjected to different uniaxial strains are shown in Figs. 2(c) and 2(d) , respectively. First, regardless of the applied strain, the composites' metal islands featured widths of ∼20, ∼22, ∼33, ∼55, and ∼68 μ m for planar layer thicknesses of 5, 10, 20, 50, and 100 nm, respectively [ Fig. 2(c) ]. Notably, the islands' widths increased almost linearly with the planar Cu layer thickness, in excellent agreement with classic theories developed for analyzing the fracture of thin metal films on polymers. 33,34 Second, before actuation with strain, the composites' fractional metal surface coverages featured values of ≳94% for all planar Cu layer thicknesses, but upon actuation with strain, the composites' fractional metal surface coverages all decreased monotonically and reached values of ∼69%, ∼62%, ∼59%, ∼57%, and ∼54% (at strains of 100%) for planar layer thicknesses of 5, 10, 20, 50, and 100 nm, respectively [ Fig. 2(d) ]. Interestingly, the changes in the fractional metal surface coverages were directly dependent on the planar Cu layer thicknesses, in agreement with our qualitative analysis of the SEM images (vide supra). The combined analyses reinforced the notion that the morphological characteristics of our composite materials were primarily determined by a single adjustable parameter. Adaptive infrared functionality of the bioinspired composite materials Having evaluated the surface microstructures of our composite materials, we characterized their mechanically actuated infrared-reflecting properties, which are illustrated in Fig. 3(a) . The representative total infrared reflectance spectra and the average changes in the reflectance measured for unactuated and actuated composite materials with variable planar layer thicknesses are shown in Figs. 3(b) and 3(c) , respectively. For composites with smaller planar layer thicknesses of 5 and 10 nm, the spectra revealed initial average total reflectances of ∼98 ± 4% and ∼101 ± 3%, respectively, at 0% strain, and decreased average total reflectances of ∼74 ± 2% and ∼73 ± 4%, respectively, at 50% strain [ Fig. 3(b) ]. For composites with larger planar layer thicknesses of 20, 50, and 100 nm, the spectra revealed initial average total reflectances of ∼103 ± 1%, ∼104 ± 1%, and ∼104 ± 4%, respectively, at 0% strain, and decreased average total reflectances of ∼72 ± 2%, ∼72 ± 2%, and ∼73 ± 4%, respectively, at 50% strain [ Fig. 3(b) ]. Here, before actuation, the spectra revealed reflectances that occasionally exceeded 100% presumably because of noise associated with environmental scattering and the use of a diffuse gold standard for calibration. 25,35–37 More generally, the average changes in the total reflectance progressively increased as a function of the applied strain, with the most substantial reflectance modulation observed for the composites featuring the thickest planar layers [ Fig. 3(c) ]. Such trends presumably resulted from the relatively larger changes in the fractional metal surface coverage quantified for composites with planar layer thicknesses of ≥20 nm [ Fig. 2(d) ]. Notably, for composites with 5 and 10 nm planar layer thicknesses, the reflectance spectra indicated some performance degradation after repeated mechanical cycling (presumably due to the propagation of fabrication defects), and for composites with 20, 50, and 100 nm planar layer thicknesses, the reflectance spectra remained nearly unchanged after repeated mechanical cycling (presumably due to the presence of fewer defects) [supplementary material, Fig. 9(a)]. Moreover, for composites with their planar Cu layers removed and for polymer matrices without planar Cu layers, the reflectances remained relatively unchanged regardless of the applied strain, underscoring the Cu layers' critical functional roles (supplementary material, Figs. 10 and 11). Taken together, the measurements provided insight into the relationship between our composites' adaptive infrared-reflecting functionalities and reconfigurable surface microstructures. FIG. 3. Measured adaptive infrared properties of the composite materials. (a) An illustration of the reflection and transmission of infrared light by the composite material before (left) and after (right) mechanical actuation. Note that the absorption of infrared light is not depicted for clarity. (b) The representative total infrared reflectance spectra measured for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under applied strains of 0% (solid lines) and 50% (dashed lines). (c) The average changes in the total infrared reflectance for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under different applied strains of ≤100%. (d) The representative total infrared transmittance spectra measured for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under applied strains of 0% (solid lines) and 50% (dashed lines). (e) The average changes in the total infrared transmittance for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under different applied strains of ≤100%. The error bars in (c) and (e) represent the standard deviations of the mean. We next spectroscopically characterized the mechanically actuated infrared-transmitting properties of our composite materials, which are illustrated in Fig. 3(a) . The representative total infrared transmittance spectra and the average changes in the transmittance measured for composite materials with variable planar layer thicknesses are shown in Figs. 3(d) and 3(e) , respectively. For composites with smaller planar layer thicknesses of 5 and 10 nm, the spectra revealed initial average total transmittances of ∼2 ± 1% and ∼2 ± 1%, respectively, at 0% strain, and increased average total transmittances of ∼16 ± 2% and ∼18 ± 4%, respectively, at 50% strain [ Fig. 3(d) ]. For composites with the larger planar layer thicknesses of 20, 50, and 100 nm, the spectra revealed initial average total transmittances of ∼1 ± 1%, ∼2 ± 1%, and ∼3 ± 2%, respectively, at 0% strain, and increased average total transmittances of ∼23 ± 2%, ∼24 ± 1%, and ∼26 ± 2%, respectively, at 50% strain [ Fig. 3(d) ]. Here, after actuation, the spectra revealed peaks at ∼6–∼8 μ m and at ∼12–∼15 μ m corresponding to the chemical functional groups of the partially uncovered SEBS matrices. 25,38,39 More generally, the average changes in the total transmittance progressively increased as a function of the applied strain, with the largest modulation again observed for the composites featuring the thickest planar layers [ Fig. 3(e) ]. Such trends presumably resulted from the relatively larger changes in the fractional metal surface coverage quantified for the composites with planar layer thicknesses of ≥20 nm [ Fig. 2(d) ]. Notably, for composites with 5 and 10 nm planar layer thicknesses, the transmittance spectra indicated some performance degradation after repeated mechanical cycling (presumably due to the propagation of fabrication defects), and for composites with 20, 50, and 100 nm planar layer thicknesses, the transmittance spectra remained almost unchanged after repeated mechanical cycling (presumably due to the presence of fewer defects) [supplementary material, Fig. 9(b)]. Moreover, for composites with their planar Cu layers removed and for matrices without planar Cu layers, the transmittances remained relatively unchanged regardless of the applied strain, further reinforcing the Cu layers' critical functional roles (supplementary material, Figs. 10 and 11). Taken together, the measurements elucidated the relationship between our composites' adaptive infrared-transmitting functionalities and reconfigurable surface microstructures. Computational simulation of the infrared properties of the bioinspired composite materials To better understand the adaptive infrared properties of our composite materials, we computationally simulated their strain-dependent infrared reflectance spectra via a straightforward model, as illustrated in Fig. 4(a) (see Methods for further details). The calculated total infrared reflectance spectra and the calculated changes in the reflectance obtained for the various composites are shown in Figs. 4(b) and 4(c) , respectively. For composites with smaller planar layer thicknesses of 5 and 10 nm, the simulated spectra revealed average total reflectances of ∼99% and ∼99%, respectively, at 0% strain, with the reflectances decreasing to values of ∼79% and ∼75%, respectively, at 50% strain [ Fig. 4(b) ]. For composites with larger planar layer thicknesses of 20, 50, and 100 nm, the simulated spectra revealed average total reflectances of ∼99%, ∼99%, and ∼99%, respectively, at 0% strain, with the reflectances decreasing to values of ∼74%, ∼73%, and ∼72%, respectively, at 50% strain [ Fig. 4(b) ]. In general, the calculated changes in the total reflectance progressively increased with the applied strain and were maximized for the composites featuring the thickest planar layers [ Fig. 4(c) ]. Notably, the simulated total reflectance spectra and calculated total reflectance modulation trends were in close agreement with our experimental measurements [supplementary material, Fig. 12(a)]. The computational simulations thus provided powerful validation for our experimental observations and confirmed the fact that our composites' reconfigurable surface microstructure governed their adaptive infrared-reflecting functionalities. FIG. 4. Simulated adaptive infrared properties of the composite materials. (a) An illustration of the model used to computationally simulate the reflection and transmission of infrared light by the composite material before (left) and after (right) mechanical actuation. Note that the absorption of infrared light is not depicted for clarity. (b) The simulated total infrared reflectance spectra for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under applied strains of 0% (solid lines) and 50% (dashed lines). Note that the spectra simulated for composite materials under applied strains of 0% are overlaid on top of each other. (c) The changes in the simulated total infrared reflectance for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under different applied strains of ≤ 100%. (d) The simulated total infrared transmittance spectra for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses for applied strains of 0% (solid lines) and 50% (dashed lines). (e) The changes in the simulated total infrared transmittance for the composite materials with 5 nm (black), 10 nm (red), 20 nm (blue), 50 nm (green), and 100 nm (purple) planar layer thicknesses under different applied strains of ≤ 100%. Note that the spectra simulated for composite materials under applied strains of 0% are overlaid on top of each other. We next computationally simulated the strain-dependent infrared transmittance spectra via the same model, as illustrated in Fig. 4(a) (see Methods for further details). The calculated total infrared transmittance spectra and the calculated changes in the transmittance for the various composites are shown in Figs. 4(d) and 4(e) , respectively. For composites with the smaller planar layer thicknesses of 5 and 10 nm, the simulated spectra revealed average total transmittances of ∼0% and ∼0%, respectively, at 0% strain, with the transmittances increasing to values of ∼17% and ∼20%, respectively, at 50% strain [ Fig. 4(d) ]. For composites with larger planar layer thicknesses of 20, 50, and 100 nm, the simulated spectra revealed average total transmittances of ∼0%, ∼0%, and ∼0%, respectively, at 0% strain, with the transmittances increasing to values of ∼21%, ∼22%, and ∼23%, respectively, at 50% strain [ Fig. 4(d) ]. In general, the calculated changes in the total transmittance progressively increased with the applied strain and were maximized for the composites featuring the thickest planar layers [ Fig. 4(e) ]. Notably, the simulated total transmittance spectra and calculated total transmittance modulation trends were in close agreement with our experimental measurements [supplementary material, Fig. 12(b)]. The computational simulations again provided powerful validation for our experimental observations and further reinforced the fact that our composites' reconfigurable surface microstructure governed their adaptive infrared-transmitting functionalities.",
"discussion": "DISCUSSION AND CONCLUSION In summary, we have validated a straightforward methodology for experimentally controlling and computationally predicting the adaptive infrared properties of our squid skin-inspired wearable composite materials, and our findings hold significance for multiple reasons. Specifically, we have demonstrated that the surface microstructure of our composite materials can be controlled via modification of a single parameter (i.e., the planar layer thickness) during fabrication. This discovery establishes the planar layer as a critical consideration during the high-throughput scalable manufacturing of our composites. Additionally, we have shown a direct relationship between the reconfigurable surface microstructure of our composite materials and their tunable infrared-reflecting and infrared-transmitting properties. This observation provides additional nuanced fundamental insight into the origins of our composites' adaptive infrared functionalities. Moreover, we have developed a straightforward computational model that precisely predicts the infrared properties of our composite materials. This advance will enable the targeted design of improved variants of our composites from arbitrary combinations of metals and polymers. Importantly, our key reported outcomes with respect to fabrication, structure–function relationships, and computational modeling should prove valuable for the engineering and optimization of other adaptive infrared platforms. As such, the described findings may help unlock the potential of not only our composite materials but also comparable systems for applications as varied as thermoregulatory wearables, food packaging, infrared camouflage, soft robotics, and biomedical sensing."
} | 6,758 |
36262715 | PMC9574570 | pmc | 2,768 | {
"abstract": "Methanol is an attractive C1 feedstock with high abundance and low cost in bio-manufacturing. However, the metabolic construction of cell factories to utilize methanol for chemicals production remains a challenge due to the toxic intermediates and complicated metabolic pathways. The group of Zhou rescued methylotrophic yeast from cell death and achieved high-level production of free fatty acids from methanol through a combination of adaptive laboratory evolution, rational metabolic engineering and multi-omics analysis."
} | 131 |
22363336 | PMC3282481 | pmc | 2,769 | {
"abstract": "On a global scale, crustal fluids fuel a large part of the deep-subseafloor biosphere by providing electron acceptors for microbial respiration. In this study, we examined bacterial cultures from sediments of the Juan de Fuca Ridge, Northeast Pacific (IODP Site U1301). The sediments comprise three distinctive compartments: an upper sulfate-containing zone, formed by bottom-seawater diffusion, a sulfate-depleted zone, and a second (∼140 m thick) sulfate-containing zone influenced by fluid diffusion from the basaltic aquifer. In order to identify and characterize sulfate-reducing bacteria, enrichment cultures from different sediment layers were set up, analyzed by molecular screening, and used for isolating pure cultures. The initial enrichments harbored specific communities of heterotrophic microorganisms. Strains affiliated to Desulfosporosinus lacus , Desulfotomaculum sp., and Desulfovibrio aespoeensis were isolated only from the top layers (1.3–9.1 meters below seafloor, mbsf), while several strains of Desulfovibrio indonesiensis and a relative of Desulfotignum balticum were obtained from near-basement sediments (240–262 mbsf). Physiological tests on three selected strains affiliated to Dv . aespoeensis , Dv . indonesiensis , and Desulfotignum balticum indicated that all reduce sulfate with a limited number of short-chain n -alcohols or fatty acids and were able to ferment either ethanol, pyruvate, or betaine. All three isolates shared the capacity of growing chemolithotrophically with H 2 as sole electron donor. Strain P23, affiliating with Dv . indonesiensis , even grew autotrophically in the absence of any organic compounds. Thus, H 2 might be an essential electron donor in the deep-subseafloor where the availability of organic substrates is limited. The isolation of non-sporeforming sulfate reducers from fluid-influenced layers indicates that they have survived the long-term burial as active populations even after the separation from the seafloor hundreds of meters above.",
"conclusion": "Conclusion Even though cultivation might not cover the whole microbial diversity of a given habitat, we were able to isolate and physiologically characterize indigenous microorganisms that are numerically and metabolically important for the marine deep subsurface. Thus, cultivation-based studies offer the opportunity to complement molecular techniques. In our study, the isolation of SRB from deep sediment layers was the precondition to answer questions concerning specific metabolic adaptations to the conditions at the sediment–basement interface. The isolation of facultatively autotrophic sulfate reducers from near-basement layers strongly suggests that these organisms survive due to their capability of consuming hydrogen after organic compounds have been depleted or become too recalcitrant for microbial degradation. The continuous supply of sulfate from the aquifer below supports their viability within their respective sediment layers even after the separation from organic matter input at the seafloor due to sediment accumulation. When organic substrate availability from the ocean becomes a limiting factor, hydrogen becomes the most important electron donor.",
"introduction": "Introduction The subseafloor biosphere is probably the largest reservoir for prokaryotic life on Earth (Whitman et al., 1998 ; Heberling et al., 2010 ). It extends several hundred meters into deeply buried sediments (Parkes et al., 1994 ; Roussel et al., 2008 ) and even further down into the upper layers of the oceanic crust (Thorseth et al., 1995 ; Furnes and Staudigel, 1999 ; Ehrhardt et al., 2007 ). Recently, it was estimated that the ocean crust contains a similar amount of microorganisms as the entire volume of the world’s oceans (Heberling et al., 2010 ). The continuous circulation of seawater within the upper crust turns these voluminous, porous, and permeable basalts into the largest globally connected aquifer (Johnson and Pruis, 2003 ; Johnson et al., 2006 ). Intense fluid circulation is a consequence of specific geological settings evolved during crust formation at ocean-spreading centers. It is especially pronounced at ocean ridges such as the Juan de Fuca Ridge in the Northeast Pacific (Johnson et al., 2006 ). This area is one of the most intensively studied locations in terms of heat-driven fluid flow (Fisher et al., 2003 ; Hutnak et al., 2006 ). While cold bottom-seawater is recharged at seamounts, it warms up within the oceanic crust beneath the sediments before being discharged again at other rocky outcrops exposed at the seafloor. The chemical composition of these low-temperature hydrothermal fluids [<150°C (Cowen, 2004 )] is altered during long-term circulation through the basalt due to continuous abiotic water–rock interaction (Edwards et al., 2003 ) especially with increasing basement temperature (Wheat and Mottl, 1994 ; Wheat et al., 2000 ), or as a response to volcanic eruption (Butterfield et al., 1997 ). Additionally, microbial activity of crust-hosted communities contributes to changes in fluid composition by removing seawater constituents such as sulfate as indicated by sulfur-isotope measurements (Rouxel et al., 2008 ). However, due to a limitation in electron donors, crustal fluids are not fully reduced and still contain suitable electron acceptors, such as sulfate, for anaerobic respiration (Wheat and Mottl, 1994 ; Wheat et al., 2000 ; Cowen et al., 2003 ; Edwards et al., 2005 ). It was postulated that basement fluids not only supply electron donors and acceptors to microbial life within the crust, but also to the microbial communities in the overlying sediments by diffusion from below (Cowen et al., 2003 ; DeLong, 2004 ; D’Hondt et al., 2004 ). We tested this hypothesis during an expedition to the eastern flank of the Juan de Fuca Ridge (IODP Exp. 301) by analyzing a 265-m-long sediment column of IODP site U1301. Sampling included material taken only two meters above the sediment–basement interface (Expedition 301 Scientists, 2005 ). At this site, sulfate diffuses into the sediments from both the seafloor (∼27 mM) and the underlying basement (∼16 mM). As a precondition for a sound microbiological and geochemical analysis, contamination controls were performed directly onboard the drillship JOIDES Resolution and proved the pristine character of the sediment samples (Lever et al., 2006 ). Our previous work has shown that fluids from the oceanic crust do support microbial life in the overlying sediments (Engelen et al., 2008 ). Exoenzyme activities and sulfate reduction rates were not only elevated near the seafloor but also at the bottom of the sediment column which correlated well with the overall geochemical settings. We detected enhanced microbial abundance in sediment layers above the basement by direct counting and the cultivation-based most probable number (MPN) technique. Microbial growth in anoxic MPN dilution series from sediment layers near the oceanic crust indicated considerable amounts of viable microbial populations. Thus, the detection of a deep sulfate reduction zone and the successful enrichment of anaerobic microorganisms was the motivation for isolating sulfate-reducing bacteria (SRB) especially from fluid-influenced sediment layers. Identifying defined physiological adaptations of indigenous microorganisms to environmental conditions can be achieved best when pure cultures are available. Even though sulfate reduction is supposed to be an important process in deeply buried sediments, only few isolates are available in strain collections. The type strain of Desulfovibrio profundus was isolated from 500 m depth in sediments of the Japan Sea (Parkes et al., 1995 ; Bale et al., 1997 ). Other piezophilic isolates closely related to Dv. profundus were cultivated from 222 m deep sediments of the Cascadia margin of the Pacific Ocean (Barnes et al., 1998 ). However, cultivation-based studies on the marine deep biosphere are still limited to a few sampling sites representing pinpricks in the ocean floor. So far, isolates from the marine subsurface were obtained from sediment samples retrieved from Mediterranean sediments (Süss et al., 2004 ) and from various sites in the Pacific Ocean: The Sea of Okhotsk, north of Japan (Inagaki et al., 2003 ), the Nankai Trough south–east of Japan (Mikucki et al., 2003 ; Toffin et al., 2004a , b , 2005 ; Kendall et al., 2006 ), the Equatorial Pacific, and the Peru Margin (D’Hondt et al., 2004 ; Biddle et al., 2005 ; Lee et al., 2005 ; Batzke et al., 2007 ). Recently, several heterotrophic bacteria and methanogenic Archaea were isolated from up to 106 mbsf deep sediments off Shimokita Peninsula, Japan using a continuous-flow bioreactor (Imachi et al., 2011 ). In this study, we extended our previous investigations on IODP Site U1301 to determine the microbial diversity within different sediment layers of the deep subsurface. We hypothesize, that zones with different sulfate concentrations harbor different populations of SRB due to varying substrate availabilities. A cultivation-based approach in combination with molecular screening tools was chosen to isolate and compare SRB from fluid-influenced sediments and near-surface layers. The metabolic properties of the isolates might provide new insights on the impact of crustal fluids on microbial metabolism in the deep-subseafloor biosphere where substrates are recalcitrant but electron acceptors are still available.",
"discussion": "Discussion Organic matter and sulfate availability generate the three different zones of the sediment column The stratification of the different sediment compartments has an imprint on the life conditions. In both, the seawater- and fluid-influenced layers, the availability of electron acceptors stimulates microbial growth and activity of indigenous microorganisms (Engelen et al., 2008 ). In terms of electron donors, bacteria that thrive in the upper 30 m of the sediments are supported by burial of relatively young organic carbon (Fisher et al., 2003 ; Johnson et al., 2006 ). Therefore, they are used to a higher supply of electron donors and adapt much better to the given cultivation conditions. In deeper sediment horizons, indigenous bacteria have to survive long-term burial by adapting to a minimum supply of substrates and electron acceptors. Their limited availability strongly influences the metabolic activities in the deep marine subsurface. Indeed, based on geochemical porewater profiles, it has been concluded that the metabolic activities of subseafloor prokaryotes are very low (D’Hondt et al., 2002 , 2004 ). They probably have developed different life strategies such as slow growth or survival as spores. The latter were presumably stimulated to germinate during our cultivation experiments since a major part of 16S rRNA gene sequences detected in all enrichment cultures affiliated to sporeforming Firmicutes (Figure 1 ). However, the decreasing number of Firmicutes with depth indicates that not all of them survive the long-term burial as spores as they might have germinated stochastically over geological time scales (Epstein, 2009 ). Other subsurface organisms that are adapted to low organic carbon concentrations might not be able to grow under the given laboratory conditions. Even though the composition of our culture media was designed to provide organic substrates in sub-millimolar concentrations, a substrate shock (Straskrabová, 1983 ) might not have been circumvented. For instance, we were not able to grow any Archaea (data not shown) even though they are proposed to represent a substantial part of the deep biosphere as indicated by intact-lipid analysis (Lipp et al., 2008 ). The supply of electron acceptors into the sediment column by crustal fluid diffusion dramatically changes the situation for microbial life within these deeply buried layers. The large numbers of non-sporeforming Gammaproteobacteria that were enriched from near-basement layers indicate the presence of viable cells. Many Gammaproteobacteria are adapted to elevated substrate concentrations (Lauro et al., 2009 ) and are therefore readily cultivated using our media. Some of them might even be typical for oceanic ridge systems. Halomonas and Marinobacter species were found to be present in hydrothermal fluids collected at the Juan de Fuca Ridge (Kaye et al., 2011 ). They were enriched during in situ colonization experiments on basaltic crust (Smith et al., 2011 ) and have also been detected in basaltic seafloor lavas and overlying seawater at the East Pacific Rise (Santelli et al., 2008 ). The upper and lower sulfate-containing zones harbor different sulfate-reducing bacteria The majority of sequences obtained from upper sediment horizons that were affiliated to SRB have Desulfosporosinus and Desulfotomaculum species as closest relatives, both sporeforming Firmicutes . However, it is unclear if they contribute to the high sulfate reduction rates of up to 8 nmol cm -3 d -1 determined for the upper sulfate-containing zone of IODP Site U1301 (Engelen et al., 2008 ). This would only be the case if these SRB are present as viable cells. It cannot be specified if they are metabolically active or if they only survive as spores within these layers. In contrast, fluid-influenced sediments exclusively harbor sulfate reducers that are members of the Deltaproteobacteria , which are not known to form any resting stages. These viable populations contribute to sulfate reduction rates of up to 3 pmol cm -3 d -1 within the lower sulfate reduction zone (Engelen et al., 2008 ). Due to their high abundance, this activity might derive from sulfate reducers affiliated to Dv. indonesiensis . This is quite surprising since the in situ temperature is around 60°C and most Desulfovibrio species are not active above 40°C (Widdel and Bak, 1992 ). However, a broad temperature range of growth was not only found for our isolates, but also for the Japan Sea isolates of Dv. profundus (Bale et al., 1997 ) and might represent an adaptation to the conditions in the deep biosphere. Thus, one reason for the divergence in the SRB communities detected in both sulfate-containing zones might be the different temperature and pressure regimes present at the top and bottom of the sediment column. Surprisingly, the isolates from the deepest fluid-influenced layers did not grow at in situ temperatures of approximately 60°C. This might be due to the chosen initial incubation conditions at 20°C and ambient hydrostatic pressure instead of the in situ pressure of ∼30 MPa. As temperature and pressure counteract on the cell membrane composition (Mangelsdorf et al., 2005 ), an insufficient combination of both parameters might result in membrane disintegration. This assumption is supported by the fact that no isolates were obtained from enrichment cultures that were incubated under in situ temperatures (data not shown). In future experiments, pressure incubations might help to overcome such problems in cultivation efficiencies. Sulfate-reducing bacteria from the lower zone have relatives in deep terrestrial aquifers Previous microbiological investigations on crustal fluids from the Juan de Fuca Ridge have identified several isolates (Nakagawa et al., 2006 ) and 16S rRNA clones (Cowen et al., 2003 ; Huber et al., 2006 ) that were affiliated to SRB. In general, the overlap between these studies compared with our culture collection from fluid-influenced sediments is quite low. Only relatives of Desulfotomaculum and Desulfonatronovibrio species were detected in two studies on the adjacent ODP Site 1026. One 16S rRNA gene sequence that is affiliated to Desulfobacterium species was found in fluids that discharge at “Baby bare seamount.” A possible explanation for this discrepancy might be that most of our isolates represent typical sediment inhabitants, which do not necessarily occur in the upper oceanic crust. However, our Deltaproteobacteria that were isolated from the lower sulfate-containing zone are facing similar physico-chemical conditions in the highly compacted sediments above the basement as in the crustal aquifer. A close relation of deep marine with terrestrial aquifers is indicated by the cultivation of Dv. aespoeensis strains from the fluid-influenced layers. Dv. aespoeensis is the most abundant sulfate reducer within formation waters of deep terrestrial boreholes at the Aespoe hard rock laboratory in Sweden (Motamedi and Pedersen, 1998 ). Those aquifers are also inhabited by complex microbial communities that are comparable to those thriving within the ocean crust (Pedersen, 2000 ). The energetical constrains are similar and select for, e.g., iron-reducing bacteria, acetogens, methanogens, and sulfate reducers (Pedersen, 1997 ). Our most frequently isolated strains from up to 260 m deep fluid-influenced sediments that are affiliated to Dv. indonesiensis also have close relatives within the deep terrestrial biosphere. Even though the type strain was originally isolated from a biofilm on a corroded ship off the Indonesian coast (Feio et al., 1998 , 2000 ), relatives were obtained from porewater brines of a deep terrestrial gas-reservoir (Sass and Cypionka, 2004 ). Furthermore, these organisms are supposedly involved in iron corrosion as determined during a study on hydrogen-consuming microorganisms in oil facilities from Japan (Mori et al., 2010 ). Biocorrosive capabilities (Feio et al., 1998 ) of Dv. indonesiensis might be an indication for a crustal origin of this species as this process plays an important role in the weathering of basalts (Edwards et al., 2005 ). Under anoxic conditions, SRB, and especially Desulfovibrio species are responsible for the corrosion of metal surfaces in consuming cathodic hydrogen (Pankhania, 1988 ; Dinh et al., 2004 ). This process might occur in the habitat as well as in our metabolic tests. As all isolates deriving from the fluid-influenced zone were capable of using hydrogen as electron donor, they might even exhibit a chemolithoautotrophic life-mode in situ . Chemolithoautotrophy within the deep biosphere Autotrophic, hydrogen-consuming microorganisms were repeatedly detected in deep continental aquifers and can even outnumber heterotrophs (Stevens and McKinley, 1995 ). The assumption that autotrophy is also a common metabolic attribute within the crust at IODP Site U1301, is supported by the isolation of a novel member of the genus Archaeoglobus from a fluid-influenced sample of ODP Site 1226 (Steinsbu et al., 2010 ). Archaeoglobus \n sulfaticallidus sp. nov., is a thermophilic and facultatively lithoautotrophic sulfate reducer and was isolated from black rust formations on top of a leaking borehole seal. Although there is no clear evidence available for lithoautotrophy within the subseafloor (Stevens, 1997 ), there are numerous studies that deal with hydrogen as suitable source for deep subsurface life. In these habitats, hydrogen can originate from many sources (Nealson et al., 2005 ) such as the fermentation of organic matter or mechanochemical processes due to the tectonic action of the Earth (Parkes et al., 2011 ), degassing from the Earth’s mantle during serpentinization of ultramafic rocks (McCollom and Bach, 2009 ), or even by radiolysis of water (Blair et al., 2007 ; D’Hondt et al., 2009 ). Furthermore, in the presence of sulfate, the oxidation of hydrogen is thermodynamically favored at high temperatures (Orcutt et al., 2010 ). Thus, in many deep subsurface habitats, hydrogen might become apparently the biochemically most important electron donor and carbon dioxide is a ubiquitous carbon source. For example, both gases were found in micro-molar concentrations in deep igneous-rock aquifers (Pedersen, 1997 ) and deep aquifers of the Columbia river basalt which is located close to our investigated site (Stevens and McKinley, 1995 ). For both sites, the authors have proposed a model for a hydrogen-driven biosphere. They assume autotrophic acetogens to form acetate from hydrogen and carbon dioxide. Acetoclastic methanogens can utilize acetate to produce methane or hydrogenotrophic methanogens might directly use hydrogen and CO 2 . At relatively young ridge-flank systems, hydrogen-utilizing sulfate reducers will outcompete methanogens as sulfate is still available within the fluids."
} | 5,116 |
35655548 | PMC9093108 | pmc | 2,770 | {
"abstract": "Superwettable materials have attracted much attention due to their fascinating properties and great promise in several fields. Recently, superwettable materials have injected new vitality into electrochemical biosensors. Superwettable electrodes exhibit unique advantages, including large electrochemical active areas, electrochemical dynamics acceleration, and optimized management of mass transfer. In this review, the electrochemical reaction process at electrode/electrolyte interfaces and some fundamental understanding of superwettable materials are discussed. Then progress in different electrodes has been summarized, including superhydrophilic, superhydrophobic, superaerophilic, superaerophobic, and superwettable micropatterned electrodes, electrodes with switchable wettabilities, and electrodes with Janus wettabilities. Moreover, we also discussed the development of superwettable materials for wearable electrochemical sensors. Finally, our perspective for future research is presented.",
"conclusion": "Conclusion and outlook In this review, recent developments in bioinspired electrodes with special wettability have been summarized. The special structure and interfacial properties of superwettable electrodes have injected new vitalities to electrochemical biosensing, which are listed in Table 1 . Research on electrodes with special wettabilities not only has fundamental interest but also promising practical applications in biomarker detection, clinical diagnosis, environment monitoring, food safety, and so on. Superwettable electrodes have exuded unique advantages in promoting the electrochemical reaction such as accelerating component transport, enhancing sensitivity and selectivity, and target solution management. However, most superwettable electrodes are limited to laboratory research and have not yet been applied to practical applications and industrial-scale production. Thus there are some challenges including the cost of biosensor construction and the long-term storage ability of the electrode to be addressed. The properties of superwettable electrodes in electrochemical biosensing Type Interfacial property Sensing Performance Limitation Superhydrophilic electrode Large electroactive area Signal enhancement High sample consumption for bulk electrodes Accelerating recognition dynamics Nonspecific adsorption Reproducibility needs to be improved Expensive Superhydrophobic electrode Self-cleaning Good reproducibility Unstable superhydrophobicity Small contact area Sample enrichment Difficulty in anchoring droplets Superaerophilic electrode High gas affinity Enrichment of gas reactant Only for gas consumption reactions Promoting gas consumption reactions Superaerophobic electrode Gas repelling Accelerating gas product desorption Only for gas evolution reactions Promoting gas evolution reactions Superwettable patterned electrode Patterned hydrophilicity and hydrophobicity Liquid manipulation Reproducibility needs to be improved Sample enrichment Nonspecific adsorption in hydrophilic zone Low sample consumption High-throughput Wettability switchable electrode Stimuli response wettability change High selectivity Not versatile Janus wettability electrode Asymmetric wettability Sample directional transport Reproducibility needs to be improved Nonspecific adsorption in hydrophilic side 1. Cost is the most crucial parameter in industrial manufacture and diagnostic application. Most superwettable electrodes are based on expensive noble metals. Nanostructured conductive polymer materials might be a good alternative in fabricating electrodes with special wettability. Moreover, the fabrication process for superwettable electrodes is limited to laboratory research and not suitable for industrial scale. Therefore, there is an urgent need for a stable and efficient engineering method to lower the fabrication cost. 2. The mechanical stability and durability of superwettable electrodes are significant in practical applications. The collapse of the surface structure may result in instability of wettability. In this case, self-healing and self-replenishing materials (such as hydrogel or slippery surfaces) may be helpful in fabricating superwettable electrodes with long-term stability under different physical conditions. 3. For now, the surface structures of many organisms have been revealed. Bioinspired materials with special wettability ( Fig. 3 ) based on these special surfaces have also been developed and widely used in many ways. However, these surfaces have not been explored for electrochemical sensors. Extending the concept of superwettable electrodes to water/oil/solid systems might be crucial for bio-detection in complex samples. 4. The understanding of wetting and adhesion of reactants/products at superwettable electrodes at the molecular level remains obscure. More effort should be contributed to theoretical calculations and simulations to enhance the fundamental understanding. The relationship between the surface wettability of superwettable electrodes and the biosensor performances should be intensively studied. Although the study about superwettable electrodes in biosensing is still in its infancy, numerous opportunities are emerging ranging from chemistry and materials science to clinical applications. With the rapid development of biomimetic superwettable materials, a variety of surfaces with special wettability have been successfully fabricated. In the future, more electrodes with special wettable interfaces will be developed to meet the requirements of various electrochemical biosensors. Given the continuous efforts devoted to this field, we are confident that superwettable surfaces may not only possess tremendous potential in biosensing but also provide many insights for other multiphase reaction systems.",
"introduction": "Introduction Billions of years of evolution endows many plants and animals with fascinating wetting properties to adapt to complex natural environments, such as the superhydrophobic lotus surface, underwater superoleophobic fish scales, or Nepenthes' slippery surface. 1–4 In the last two decades, much effort has been devoted to revealing the wetting behavior mechanism and fabricating artificial superwettable surfaces. 5–8 In 2021, superwettability was announced as one of the IUPAC Top Ten Emerging Technologies in Chemistry. 9 Superwettability has been employed in many applications, including self-cleaning, water harvesting, anti-icing, anti-fogging, anti-corrosion, printing, sensors, and water-oil separation. 10–18 With the outbreak of coronavirus-2019, a massive number of diagnostic devices or biosensors are needed to fight this global pandemic. Innovation in biosensing technologies can accelerate the early diagnosis of diseases and the control of epidemics. 19 Recently, superwettable materials have injected new vitality into the construction of efficient biosensors. 20–26 Biosensing interfaces with special wettabilities exhibit unique solid–gas–liquid interfacial behavior and could further enhance the efficiency of biorecognition. 27 Special wettable materials have been broadly applied in biosensors and can be combined with different signal output methods, including fluorescence, electrochemistry, surface-enhanced Raman spectroscopy (SERS), and colorimetry assay. 28–36 Among various biosensors, electrochemical biosensors have attracted extensive attention due to their high sensitivity, good selectivity, low cost, simple manipulation, multiplexed detection capabilities, ease of miniaturization, and real-time signal feedback. 37–39 The electrochemical sensing interface is crucial in electrochemical biosensors and directly affects the biosensing sensitivity, specificity, stability, and response dynamics. 40–45 Advantages that superwettable materials bring to electrochemical biosensors include, but are not limited to, behaving as biocompatible substrates for biomolecule immobilization, enlarging the active electrochemical areas, intelligent liquid management, favoring the enrichment of target biomolecules, and accelerating the desorption of products. 46–48 Most superwettable electrodes are nanostructured materials with high roughness, increasing the active electrode area dramatically. Superwettable surfaces also exhibit particular liquid management capability, which will play an important role in sample collection and multiplex biosensing. By optimizing the wettability of electrodes, the contacting behavior between reactants/products and electrodes can be varied, thus affecting the desorption/absorption of reactants or products. 49–51 Hence, the thermodynamics and reaction rates of electrochemical reactions can be well-tuned for improved performance of biosensors. Inspired by natural superwettable interfaces, several superwettable electrodes have been fabricated, including superhydrophilic, superhydrophobic, superaerophilic, superaerophobic, superwettable patterned, Janus wettability, and wettability switchable electrodes as shown in Scheme 1 . 51–57 Due to their remarkable features, these superwettable electrodes inject fresh energy into the development of electrochemical biosensing and exhibit outstanding promise. 58–60 Although superwettable electrodes are still in their infancy stage, it is required to summarize the progress in superwettable electrodes for sensitive and wearable biosensing and predict future studies in this field. First, we discussed the electrochemical reaction process at the electrode/electrolyte interfaces. Then, we introduced some fundamental understanding of superwettable materials. Subsequently, we summarized the recent progress of superwettable material-based electrochemical biosensors. Finally, we also highlighted the perspectives and challenges of superwettable electrodes. We expect this review to promote the further development of superwettable electrodes, intelligent electrochemical biosensors, and wearable biosensors towards routine disease surveillance or early diagnosis of malignant diseases. Scheme 1 Bioinspired superwettable surface for electrochemical biosensors. Due to their intrinsic advantages, including large active areas, intelligent liquid management, and dynamics acceleration, several superwettable materials have been applied in electrochemical biosensors. This review mainly focuses on the recent progress in superhydrophilic, superhydrophobic, superaerophilic, superwettable patterned, and Janus wettability electrodes, and electrodes with switchable wettability. Electrochemical reaction at the interfaces Electrochemical sensing systems consist of a biorecognition element and a transducer. Biorecognition elements refer to the immobilized capture probes such as DNA, aptamer, antibodies, enzymes, or receptors. 40 Electrochemical transducer parts, i.e. electrodes, transfer the biorecognition process to the electrochemical signal. Electrochemical reactions are the core part of bioelectrochemical biosensors, occurring at an “electrified interface” between electrodes and electrolytes. 61,62 According to the Stern model, the electrode/electrolyte interfaces can be divided into the inner Helmholtz layer (IHP) filled with non-solvated ions, the outer Helmholtz layer (OHP) filled with solvated ions, and the diffuse layer where the ion diffusion occurs, as shown in Fig. 1a . 63,64 The electrochemical reaction generally occurs at the IHP. Fig. 1 Schematic illustration for the electrochemical reactions occurring at the interface between the electrode and electrolyte. (a) Stern model of the EDL generated on a negatively charged electrode surface. (b) The key processes consist of mass transport and electron conduction in an electrochemical reaction. In the electrochemical reaction, the reaction process is generally accompanied by reactant transport, product transport, and electron conduction, as shown in Fig. 1b . 65 For a reversible reaction, A + B ⇋ AB, the thermodynamics equilibrium constant can be defined as eqn (1) , 1 where C is the concentration of the reactant and the product. K is constant under a given condition. Increasing the C A or C B or decreasing C AB can promote the forward reaction. Enriching reactants or accelerating the desorption of products by changing the surface structure or interfacial properties can promote the forward reaction. Reactant enrichment and product escape can be realized by tailoring the wettability of electrodes, which will accelerate the thermodynamics of the electrochemical reaction. 59 In addition to mass transfer, enhancing the current intensity is also crucial for improving the electrochemical sensitivity. The diffusion flux was employed further to explain the relationship between the current and interface reaction. In brief, the diffusion flux is the amount of substance that passes through a unit area in a unit of time, according to Fick's first law ( eqn (2) ), 66–68 2 where J is the diffusion flux, D is the diffusion coefficient, C is the concentration of the component and x is the diffusion distance. Current is the amount of charge passing through any cross-section in unit time according to Faraday's law ( eqn (3) ), 69–72 3 where I is the current, n is the reactant valence, F is the Faraday constant, and A is the electrode area. Generally, the reactant valence and Faraday constant are usually constant for a given redox reaction. The current intensity is proportional to the diffusion flux and cross-sectional area. Thus, materials with high conductivity and a high surface area have a significant impact on the signal current. Most superwettable electrodes exhibit high roughness, which further favors the increase of current and improves the sensitivity of superwettable electrode-based biosensors. 73,74 According to the above discussion, the development of electrodes that promote the enrichment of reactants, accelerate the desorption and escape of products, have high electrical conductivity, and high specific surface areas will facilitate the development of sensitive electrochemical biosensors. General concepts about superwettable materials Wettability generally refers to the liquid wetting ability on a solid surface. The wettability of the surface can be described using the contact angle (CA). 75 As shown in Fig. 2a , Young's equation described the relationship between the CA and interfacial energy on an ideal solid surface (smooth and chemically homogeneous surface) using eqn (4) , 76 4 where θ w is the CA of the liquid and γ sv , γ sl, and γ lv are the specific energies of solid–vapor, solid–liquid, liquid–vapor, respectively. Young's equation revealed the relationship between wettability and surface energy. According to Young's equation, surfaces with special wettability can be constructed using a material with different surface energies. 77 The surface with a water CA less than 65° is hydrophilic, and the hydrophobic surface is the one with a water CA larger than 65°. 78,79 In particular, some surfaces have shown extreme wettability. Thus a further definition of superhydrophilic (CA < 5°) and superhydrophobic (CA > 150°) was proposed. 75,80 Surface wettability is determined by the surface chemical composition and morphology. Higher roughness of the surface can drive the wettability to the extreme. 6 Generally, two models, including the Wenzel and Cassie models, explain the wetting state on a rough surface. In the Wenzel model ( Fig. 2b ), the droplet was wetted into the surface structure at the contact area, resulting in high CA hysteresis. 81 In this case, the CA was enlarged by a factor r (related to the surface roughness) according to eqn (5) . 5 Fig. 2 (a) Definition of the contact angle of the droplet on an ideal surface. The wetting state on the rough surface: (b) Wenzel model and (c) Cassie model. \n Eqn (5) shows that the surface roughness will promote either hydrophilicity or hydrophobicity of the surface. In the Cassie model ( Fig. 2c ), the droplet only sits on the top of the surface with a gas layer between the liquid and substrate, which leads to a liquid–gas–solid triphase interface. 82 In this composite state, the CA will be corrected by the area fraction ( f s ) of the solid on the surface according to eqn (6) . 6 As has been discussed, the wettability of the surface is determined by the chemical composition and roughness. For the fabrication of superwettable surfaces, high roughness nanostructures were constructed by different methods, including electrodeposition, self-assembly, etching, and so on. The modulation of the chemical composition can be achieved by direct use or chemical modification of materials with different surface energies. For the superwettable patterned surface, photoirradiation or plasma treatment through a photomask might be employed. Recently, superwettable materials in air have been extended to water and oil systems. Several other types of superwettability are also studied, including underwater superaerophobicity, underwater superaerophilicity, underwater superoleophobicity, underwater superoleophilicity, underoil superhydrophilicity, underoil superhydrophilicity, etc as shown in Fig. 3 . 14,77,83–86 Accordingly, the application of superwettable materials has also been extended. 87–91 Electrochemical reactions at superwettable electrodes may differ markedly from those on traditional electrodes due to the special interactions between solid, liquid, oil, and gas phases. 92,93 For example, the superoleophobic electrode can accelerate the desorption of oil phase products in the Kolbe reaction and greatly improve the working life of the electrode. 94 Highly efficient electrochemical biosensing can be achieved by tailoring the wettability of electrodes. Fig. 3 The superwettability systems. (a) Superhydrophilicity (left) and superhydrophobicity states in air. (b) Underwater superaerophilicity (left) and superaerophobicity (right). (c) Underwater superoleophillicity (left) and superoleophobicity (right). (d) Underoil superhydrophilicity (left) and superhydrophobicity (right). Superhydrophilic electrodes in electrochemical biosensing Superhydrophilic surfaces with CA < 5° have high roughness and can be wetted by electrolyte solutions thoroughly according to the Wenzel model. 95 The Wenzel contact mode results in a large contact area between the electrolyte solution and the electrode, i.e. , higher electroactive surface areas, which improves the sensitivity for biochemical detection ( Fig. 3a , left). 96 Metal electrodes have been broadly applied in electrochemical biosensors due to their excellent electrical conductivity and high surface energy. By fabricating hierarchical structures on metal electrodes, superhydrophilic electrodes with micro/nano roughness can be obtained. Compared with the planar electrode, the electrochemical sensing performance can be improved because of the increased specific surface area. 97 Many superhydrophilic metal electrodes with rough structures have been constructed to detect proteins and nucleic acids. Our group has developed a superhydrophilic fractal gold electrode (FracAu) for electrochemical biosensors ( Fig. 4a ). 52,97 The superhydrophilic FracAu-based biosensor has been employed to detect thrombin and apolipoprotein E4 (APOE4) and achieved a limit of detection (LOD) of 5.7 fM for thrombin and 0.3 ng mL −1 for APOE4, respectively. Compared to the plain gold electrode, the LOD of the superhydrophilic FracAu electrode was reduced by 3 to 10 times, and the electroactive surface area was improved to almost 50 times. In addition, the fractal structure can decrease the steric hindrance, which can promote the bio-recognition efficiency between probes and targets. The FracAu electrode has provided a high sensitivity biosensor and shown great potential in detection. Besides, our group has developed a free-standing superhydrophilic Pt nanowire network electrode (PtNNE) for glucose biosensors ( Fig. 4b ). 73 The PtNNE exhibited high electrocatalytic activity and stability due to the 3D structure of high-index facet surface polycrystalline nanowires. The superhydrophilic PtNNE also provided an excellent sensitivity of 1360 μA mM −1 cm −2 for detecting hydrogen peroxide (H 2 O 2 ) and 114 μA mM −1 cm −2 for detecting glucose. The excellent performance of the superhydrophilic PtNNE-based biosensor is ascribed to the 3D nanowire self-interconnecting network nanostructure induced high surface area and hydrophilicity. The electrochemical active surface area of PtNNE was measured as almost 214 times higher than that of the plane electrode. Moreover, the porous structure of PtNNE provided abundant active sites for rapid electron migration and mass transportation. On the other hand, the substrate-free noble metal electrode has shown better electrochemical properties than the substrate-based electrode. This superhydrophilic PtNNE has provided a reliable electrochemical platform for oxidase-based enzyme biosensors. Notably, the liquid wetting behavior on the superhydrophilic electrode is the Wenzel state, which makes the liquid thoroughly wet the electrode and further increases the sensitivity of the electrochemical biosensor. Fig. 4 The applications of superhydrophilic electrodes in biosensors. (a) The superhydrophilic fractal gold electrode for thrombin detection. Reproduced from ref. 52 with permission. Copyright 2012, Royal Society of Chemistry. (b) The superhydrophilic electrode based on the self-interconnecting Pt nanowire network for an electrochemical amperometric biosensor. Reproduced from ref. 73 with permission. Copyright 2015, Royal Society of Chemistry. (c) The NME array for the multiplexing miRNA detection. The signal of NME hybridized with target miRNA can be exported by differential pulse voltammetry (DPV). Adapted from ref. 101 with permission. Copyright 2009, Wiley-VCH. Bulk superhydrophilic electrodes are generally used in large volumes of solutions, which causes large sample consumption. In addition, bulk superhydrophilic electrodes are not easily integrated into mobile biosensing platforms for portable biosensing. In this case, nanostructured microelectrodes (NMEs) with superhydrophilicity have attracted much attention due to their high electric current density, large specific surface area, and small size. 60,98 Microelectrodes are generally developed by electrochemical deposition. The as-prepared NMEs have a highly fractal nanostructure which further induces a superhydrophilic surface. Shana's group developed a Pd NME-based biosensor for the detection of nucleic. 99 PNA single strands were employed to capture the target with [Ru(NH 3 ) 6 ] 3+ and [Fe(CN) 6 ] 3− as signal molecules. Due to the high surface area of the superhydrophilic nanostructured microelectrodes, the LOD of this nanostructured microelectrode-based biosensor is 1 fM, which is 100 times higher than that of a smooth microelectrode. Superhydrophilic NMEs lead to thorough wetting between the electrode and the electrolyte, increasing the sensitivity. The high curvature of NMEs causes lower steric hindrance, which is conducive to bioconjugation of the biomarkers. 100 Besides single biomarker detection, superhydrophilic NMEs have also shown great potential in the multiplex detection of biomarkers. Shana's group developed an NME array for multiplexing miRNA detection based on these nanostructured superhydrophilic microelectrodes, as shown in Fig. 4c . 101 The NME array was fabricated by electrodeposition of Pd on an Au micropatterned silicon substrate. The sensitive detection of miRNA was achieved with a LOD of 10 aM. Moreover, they have developed a gene-circuit-based sensor based on the NME array. 102 In response to the target, gene circuits generated restriction enzymes and released methylene blue-labeled reporter DNA. The capture DNA modified NMEs were then bioconjugated with the methylene blue-labeled reporter DNA for the output of the signal. The gene-circuit-based sensor has achieved sensitive and simultaneous detection of colistin antibiotic resistance genes (mcr-1, mcr-2, mcr-3, and mcr-4) with LOD of 1 fM for mcr-4. Such an approach has provided a multiplexing NME platform for electrochemical biosensors and shown great potential in high-throughput biosensing for clinical diagnoses. Besides, NMEs were combined with a neutralizer displacement assay (NDA) for simultaneous detection of nucleic, protein, and small molecules. 103 The neutralizer-modified NME exhibited a charge-free state. Displacement of the neutralizer in the presence of the target resulted in a change of charge which further generated an electrochemical signal. Such a NDA-based NME has achieved sensitive detection of cocaine, DNA (LOD 100 aM), E.coli RNA (10 pg mL −1 ), bacteria (0.15 c.f.u. mL −1 ), proteins (lower to 10 fM for thrombin), and adenosine triphosphate. And the NME-based neutralizer displacement assay has shown exciting multiplexing capabilities, which allows simultaneous detection of multiple analytes. Nanostructured microelectrodes can also be combined with a digital microfluidic device to develop portable biosensors. Rackus et al. combined NMEs with digital microfluidics and achieved the sensitive rubella virus diagnosis with a LOD of 0.07 IU mL −1 for rubella virus IgG. 104 And digital microfluidics gives an alternative for automated manipulation sample handling as well as lowering sample consumption. Therefore, NME-based digital microfluidics has provided high-integration, automation, portable, user-friendly, and low-cost biosensors for distributed diagnoses ( e.g. blood glucose meter) in practical applications. According to the above discussion, superhydrophilic electrodes and superhydrophilic NMEs have provided high electrochemically active area and low steric hindrance in biosensors. And the incorporation of superhydrophilic electrodes with digital microfluidics has further provided an efficient approach for practical applications. Superhydrophobic electrode in electrochemical biosensing Inspired by the self-cleaning properties of lotus, many superhydrophobic surfaces have been developed and have various applications in different fields. 105 In electrochemical biosensing, the passivation of the electrode surface has led researchers to develop electrochemical biosensors. Superhydrophobic electrodes (CA > 150°) provide an optional solution for fabricating electrodes with long-term stability. 13 On the superhydrophobic electrode with a Cassie state, air was trapped between droplets and substrates, which led to a nonwet contact mode ( Fig. 3a right). Superhydrophobic electrodes have exhibited particular self-cleaning properties and provided refreshable ability for repetitive use by simply washing. Zhu et al. developed a PDMS@MWCNT modified glassy carbon (GCE) superhydrophobic electrode for refreshable biosensors. 51 The polydimethylsiloxane (PDMS) provided a superhydrophobic surface, and the multi-walled carbon nanotubes (MWCNT) provided high conductivity. Due to the self-cleaning properties of the electrode, biosensors based on superhydrophobic electrodes have presented excellent stability in repeated experiments with a lower RSD (1.4% for dopamine and 5.5% for quercetin) of anodic peak current compared to the bare GCE (10.5% for dopamine and 29% for quercetin). Superhydrophobic electrodes have achieved sensitive detection of dopamine and quercetin with LODs of 0.25 μM and 0.5 μm, respectively. Self-cleaning superhydrophobic electrodes can be combined with the magneto-controlled moveable architecture (MCMA) strategy for detecting carcinoembryonic antigen (CEA). This work achieved linear detection ranges of 0.1–100 ng mL −1 and a LOD of 0.041 ng mL −1 . Due to its self-cleaning properties, the superhydrophobic electrode can be used repetitively by simple washing for the desorption of the MCMA. The self-cleaning superhydrophobic electrode was also combined with a ratiometric strategy for complex detection, as shown in Fig. 5a . Zhang et al. developed a self-cleaning electrode by modifying the zeolite imidazole framework (ZIF) and PDMS on the GCE. 106 The ratiometric strategy was built by introducing methylene blue as a reference molecule into the electrolyte solution, which avoids tedious modification of electrodes. The superhydrophobic electrode has proved to be stable in air for 45 days due to its great self-cleaning ability. And the reproducibility was verified by repeatedly polishing and remanufacturing with an average RSD of 6.67%. Moreover, the integrated biosensor has presented good sensitivity in multiple biomarker detection (LOD of 0.13 μM for Adrenaline, 0.03 μM for Serotonin, and 0.5 μM for Tryptophan). The superhydrophobicity of the fabricated electrodes is beneficial for refresh ability and stability of the electrode, improving the reproducibility of detection results. Fig. 5 Biosensors based on the superhydrophobic electrodes. (a) The combination of the self-cleaning superhydrophobic electrode with the ratiometric strategy for Adrenaline, Serotonin, and Tryptophan detection. Reproduced from ref. 106 with permission. Copyright 2019, American Chemical Society. (b) The droplet evaporation on the nanotextured superhydrophobic electrode for DNA enrichment. Adapted from ref. 108 with permission. Copyright 2013, Royal Society of Chemistry. Biomarker enrichment can effectively improve the detection sensitivity of low-abundance biomarkers. 107 Evaporation of sample droplets is an efficient method to achieve biomarker enrichment. 48 The droplet evaporation process on solid surfaces can be divided into two modes: constant contact area mode and constant contact angle mode. Droplet evaporation on surfaces with high contact angle hysteresis usually adopts constant contact area evaporation. In contrast, evaporation on the surface with low contact angle hysteresis is constant contact angle evaporation. Alam's group has developed a nanotextured superhydrophobic electrode array for ultratrace biomarker analysis, as shown in Fig. 5b . 108 The high roughness of the nanotextured surface provided superhydrophobicity with high contact angle hysteresis. Droplet evaporation on the fabricated superhydrophobic electrode followed the constant area mode, and the three-phase contact line did not move backward. Thus, the enrichment of biological samples and signal enhancement were achieved. In addition, the thermal effect of the electrodes in the detection process accelerates the evaporation process of the droplets. The superhydrophobic electrode has exhibited ultra-high sensitivity towards label-free DNA detection with a LOD of 60 aM by non-faradaic impedance spectroscopy. Superhydrophobic electrodes give insight into ultra-trace sample detection and will have broad application in biosensing. Superaerophilic/superaerophobic electrodes for electrochemical biosensing In some gas involving electrochemical reactions, the contact state between the gas and electrode interface will affect the thermodynamics of the reaction. 53,109 Gas involving electrochemical reactions can be divided into gas consumption reactions and gas evolution reactions. 59 For gas consumption reactions, increasing the gas concentration can promote the forward progress of the reaction. 110 In contrast, accelerating the desorption and escape of the gas is the key process to promote the progress of gas evolution reactions. 111,112 Inspired by bubble bursting on the surface of lotus leaves, many superaerophilic surfaces have been fabricated. 113 Superaerophilic electrodes are electrodes with high gas affinity. The rough superaerophilic electrode can trap air bubbles on the surface when immersed into the electrolyte solution ( Fig. 3b left), and a liquid–gas–solid triphase interface formed. 114 Since the diffusion coefficient of the gas in the liquid phase is much lower than that in the gas phase, the triphase interface on superaerophilic electrodes can provide a stable and abundant concentration of the gas reactant, which favors the forward process for those gas consumption reactions. The electrochemical glucose detection is a typical gas consumption reaction, where oxygen is the main reactant. The stable oxygen concentration at the electrode surface is critical for glucose biosensing. 115 The liquid–gas–solid triphase interface induced by the superaerophilic electrode will provide an abundance of oxygen. Feng's group has developed an oxygen-rich enzyme biosensor for glucose detection based on the superaerophilic electrode. 88 As shown in Fig. 6a (left), the superaerophilic electrode was fabricated by modifying a catalyst ( e.g. , Pt) and oxidase ( e.g. , glucose oxidase) on an underwater superaerophilic ( i.e. superhydrophobic substrate) carbon fiber substrate. The superaerophilic electrode has been used for glucose detection and achieved a large linear detection range (up to156 mM), which is much higher than the linear detection range (up to 5 mM) of the GCE, as shown in Fig. 6a right. Furthermore, the superaerophilic electrode exhibited excellent stability by measuring 20 mM glucose 120 times with a relative standard deviation of 1.1%. Fig. 6 Applications of the underwater superaerophilic electrode for glucose detection. (a) The triphase interface induced by the superaerophilic electrode for gas involving biosensing (left). Superaerophilic electrode (red line) shows a larger linear detection range than a flat electrode (black line) (right). Adapted from ref. 88 with permission. Copyright 2016, Wiley-VCH. (b) The H 2 O 2 cathodic reaction for glucose detection based on the superaerophilic electrode (left). The cathodic reaction (red line) exhibits better interference resistance compared to the anodic reaction (blue line) (right). Adapted from ref. 116 with permission. Copyright 2018, Wiley-VCH. (c) The PEC bioassay system based on the superaerophilic electrode. Compared with the diphase system, the triphase system provided sufficient oxygen content. Adapted from ref. 117 with permission. Copyright 2018, Wiley-VCH. In the detection of glucose, there are many interferences ( e.g. , ascorbic acid (AA), homovanillic acid (HVA), and uric acid (UA)) in the anodization reaction. In contrast, the H 2 O 2 cathodic reduction reaction can avoid these interferences in glucose detection in complex solutions. Since the reduction potential of oxygen is similar to that of H 2 O 2 reduction, the unstable oxygen concentration will result in inconsistent background signal. Due to the triphase interface, the oxygen concentration in the system is stable, resulting in a stable background signal. Song et al. developed a superaerophilic electrode for H 2 O 2 cathodic detection as shown in Fig. 6b (left). 116 Compared with normal electrodes, superaerophilic electrodes showed stable signals in solutions with different oxygen contents. The cathodic reduction reaction has effectively avoided the interference of other substances ( Fig. 6b right). The detection of glucose based on the superaerophilic electrode was achieved in a linear dynamic range of 80 × 10 −3 M, which is significantly higher than that of a standard GCE (0.7 × 10 −3 M). The superaerophilic electrode has provided a novel strategy by introducing a triphase reaction interface that caused sufficient gas supply for gas consumption biosensors. In addition, superaerophilic electrodes have also been used in photoelectrochemical (PEC) bioassays. Wang et al. developed a PEC bioassay system based on the cathode reaction as shown in Fig. 6c . 117 The superaerophilic electrode has provided a stable concentration of O 2 due to the triphase interface. Such superaerophilic electrode-based PEC biosensors have been applied in glucose detection and achieved high sensitivity with a LOD of 1 × 10 −6 M. The linear detecting range of the triphase electrode is 100 times higher than that of the diphase one. The triphase interface based on the superaerophilic electrode can be applied as a versatile platform for other gas consumption reactions and provides an optional way to fabricate highly sensitive biosensors with gas involved. Electrochemical water splitting has been considered an efficient and sustainable strategy to produce clean fuels. In the hydrogen evolution reaction, hydrogen gas bubbles are produced, and generally adhere to the electrodes and desorb when grown to a critical size (about a few hundred μm). These bubbles hinder the contact between the electrolyte and the electrode and reduce the electrochemical catalytic efficiency. 118 The superaerophobic surface has low adhesion to gas bubbles ( Fig. 3b right), and can facilitate the desorption and escape of gas bubbles on the electrode surface. This unique property will promote the forward progress of some gas evolution electrochemical reactions. Sun's group has developed an underwater superaerophobic electrode based on the MoS 2 nanostructured film for a highly efficient hydrogen evolution reaction (HER), as shown in Fig. 7a left. 92 The superaerophobic electrode has exhibited good HER efficiency ( Fig. 7a right), which can be higher than the HER efficiency of commercial Pt/C films at sufficiently high overpotentials. On the superaerophobic electrode, the bubbles desorb and escape before reaching 100 μm in diameter. The superhydrophobic electrode exhibits anti-bubble adhesion performance compared with the flat electrode (bubbles diameter more than 400 μm). They also developed a superaerophobic electrode based on a pine-shaped Pt nanoarray for the HER as shown in Fig. 7b left. 119 The pine-shaped Pt nanoarray electrode exhibited a high electrochemically active surface area, which is twice that of the flat electrode. Compared with spherical nanostructured electrodes with similar electroactive surface areas, the pine-shaped Pt nanoarray electrodes exhibited better HER performance (2.55 times higher) due to better bubble repulsion induced by the unique nanostructures. Furthermore, the pine-shaped Pt nanoarray electrode has showed steady HER performance (about 100% retention) over 36 hours of stability measurements ( Fig. 7b right). This work revealed the relationship between electrode wettability and sensing performance and provided a new approach to developing highly efficient electrodes for gas evolution reactions. Fig. 7 The highly efficient HER on an underwater superaerophobic electrode. (a) Schematic of the adhesion behavior of bubbles on the superaerophobic electrode (left). The superaerophobic electrode (nanostructured film) showed high HER performance compared with flat films and Pt/C films, (right). Adapted from ref. 92 with permission. Copyright 2014, Wiley-VCH. (b) The pine-shaped Pt nanoarray superaerophobic electrode for the HER (left). Pine-shaped Pt nanoarray superaerophobic electrode showed good long-term stability compared with the Pt nanosphere electrode and the Pt flat electrode (right). Adapted from ref. 119 with permission. Copyright 2015, Wiley-VCH. Due to their extraordinary low adhesion to gas bubbles, nanostructured superaerophobic electrodes favor gas evolution reactions, which might be very promising in developing efficient electrochemical assays with gas products, such as H 2 O 2 detection. Superwettable micropatterned electrodes for electrochemical biosensors Inspired by desert beetles, superwettable micropatterns have been fabricated by combining superhydrophobic backgrounds with superhydrophilic microwells. 120 Superwettable patterns have been broadly applied in biosensors due to their several advantages, including: (1) superhydrophilic microwells provide stable droplet anchoring ability; (2) the superhydrophobic background prevents contamination between adjacent droplets; (3) the evaporation of droplet enables ultratrace sample enrichment; (4) the small size of the superhydrophilic microwells allows low sample consumption; (5) the superhydrophilic microwell array allows multiplex detection. 47,55,121–123 Our group has combined the superwettable micropattern microchip with the dual-DNA walker strategy to detect E. coli O157: H7 DNA. 124 The microchip was prepared based on a fractal gold substrate. The superhydrophobic background prevents the liquid from spreading, and the superhydrophilic microwells provide the point of liquid anchoring. Such a synergistic effect has provided good droplet management and significantly reduced the amount of the analytical solution. Moreover, the nanostructured fractal gold has provided a large contact area to enhance the response signal. The microchip has achieved ultrahigh sensitive detection of the E. coli O157: H7 DNA with a LOD of 30 aM. In addition, we also used the superwettable micropatterned microchip for the multiplex detection of prostate cancer biomarkers (miRNA-375, miRNA-141, and prostate-specific antigen), as shown in Fig. 8a . 125 The superhydrophobic background has successfully prevented the inter-contamination of adjacent droplets. And the multi-biomarker sensitive detection has proved that the LODs of miRNA-141 and miRNA-375 are 0.8 nM, and the PSA is 1.0 pM. The superwettable micropatterned electrodes have presented great potential in multiplex electrochemical analysis. Fig. 8 The superwettable micropatterned electrode-based microchip for electrochemical analysis. (a) The miRNAs and PSA electrochemical analysis on the superwettable patterned microchip. Adapted from ref. 125 with permission. Copyright 2018, American Chemical Society. (b) The electrochemical analysis in the droplet microarray based on orthometric gold electrode bands. Reproduced from ref. 126 with permission. Copyright 2017, American Chemical Society. Levkin's group has developed a superwettable droplet array based on orthometric gold electrode bands ( Fig. 8b ). 126 The superwettable droplet array was fabricated by covering superwettable patterned porous polymethacrylate on orthometric gold electrode bands. Each superhydrophilic well can act as an individual electrochemical assay without the inter-contamination of adjacent droplets and ion transport. The individual droplet cell array was obtained by rolling liquid on the surface of the electrode. The electrochemical signal of each droplet in the array can be read out individually. The sensitive detection ability of 1,4-benzoquinone and H 2 O 2 within a single droplet has been confirmed. Electrodes with switchable wettabilities The coupling of the amplification-by-wettability switching concept with the electrochemical method offers great promise in bio-detection. The surface with switchable wettabilities in response to the external stimulus (pH value, irradiation, heat, or some molecules) was explored. 127 Adding some biomolecules (saccharides, nucleic, or protein) can induce hydrogen-boding interactions between the polymer substrate and biomolecules, which incurs wettability switching. The biomarker responded switchable surface could be applied for fabricating electrochemical biosensors. 49,128,129 Ding et al. developed a wettability switchable electrode for chiral sensing of monosaccharide enantiomers. 56 As shown in Fig. 9 , the presence of monosaccharide induces enantiomer conformational transition of the copolymer, which further results in the transformation of wettability and facilitates the diffusion of electroactive probes to the electrode. In the absence of target monosaccharides, the copolymer consisting of poly( N -isopropylacrylamide) (PNI), β-Asp-Phe dipeptide (β-MAP), and bis(trifluoromethyl)-modified phenylthiourea (TP) can form intramolecular hydrogen bonds, which results in the contraction of the copolymer chains and superhydrophobic surface. When different chiral monosaccharide enantiomers were present, the intramolecular hydrogen bonds of the copolymer were broken to different extents, which further resulted in the switching of the wettability from hydrophobicity to hydrophilicity. The change of electrode wettability favors the enrichment of the target on the screen-printed carbon electrode (SPCE) and further improves the sensitivity of electrochemical detection. Such a wettability switchable electrode has achieved sensitive detection of d -glucose as observed by electrochemical impedance spectroscopy with a LOD of 1 nM. In addition, electrodes with switchable wettabilities have been employed to monitor the d -glucose uptake behavior of cancer cells. Fig. 9 Schematic of wettability switchable electrodes for chiral sensing of monosaccharide enantiomers. Reproduced from ref. 56 with permission. Copyright 2016, American Chemical Society. According to the same strategy, an electrode with switchable wettability in response to sialic acid (SA) has been fabricated for SA electrochemical biosensing. 130 The modified SPCE/Au electrode can switch from superhydrophobic to superhydrophilic with the presence of SA, which further allows the enrichment of redox labels and targets on the electrode surface, thus enhancing the electrochemical signal response. Sensitive SA detection (LOD 0.4 pM) was achieved based on the electrode with switchable wettability. The dynamic monitoring of SA in a living mouse brain was also performed by combining with in vivo microdialysis. Electrodes with switchable wettability have shown excellent selectivity due to their inherent specificity molecule recognition system. Such electrodes have indicated broad applications in highly specific biomarker monitoring in complex samples. Superwettable materials for wearable electrochemical sensors Wearable electrochemical sensors have garnered considerable attention due to their tremendous promise in real-time and non-invasive monitoring of chemical markers and physical signals. 131–133 Through the modality of accessories or clothes, sweat components (including glucose, sodium and potassium ions, and pH) can be non-invasively detected by the wearable electrochemical sensor. 134–137 Connecting wearable electrochemical sensors to mobile devices ( e.g. , mobile phones and tablet computers) can achieve real-time, user-friendly, and household/bedsides monitoring. 138 Superwettable materials have been used in wearable biosensors and provide unique sample management capabilities. 139 Eccrine sweat is a significant biofluid containing rich biomarkers ranging from electrolytes, metabolites, hormones to proteins. Sweat collection is critical for wearable biosensors. 138 Current sweat collection relies on absorbent pads adhering to the skin. He et al. recently developed a wearable sweat sensor based on a superhydrophilic carbon textile derived from silk fabrics for six sweat biomarker simultaneous detection ( Fig. 10a ). 140 A superhydrophilic nitrogen-doped carbon textile (SilkNCT) was integrated on a flexible PET substrate as the working electrode. The hydrophobic PET substrate can separate the six working electrodes and avoid inter-contamination. The superhydrophilic SilkNCT has provided excellent sweat collection ability and shown good sensitivity for detecting glucose, lactate, ascorbic acid, uric acid, Na + and K + , and the LODs are 5 μM, 0.5 mM, 0.5 mM, 0.1 μM, 1 mM and 0.5 mM, respectively. The excellent reproducibility was confirmed with relative standard deviations (RSD) for six biomarkers ≤ 8.2%. In addition, the SilkCNT-based wearable biosensor combined with mobile phones enabled real-time monitoring of the glucose concentration. Fig. 10 Superwettable materials in wearable electrochemical biosensors. (a) The wearable biosensor based on the superhydrophilic electrodes for the multiplex sweat analysis. Adapted from ref. 140 with permission. Copyright 2019, AAAS. (b) The superhydrophobic (superaerophilic) electrode integrated wearable biosensor for gas-involving on-body analysis. Adapted from ref. 141 with permission. Copyright 2019, Wiley-VCH. (c) Wearable sweat sensor based on the Janus electrode for sweat transportation and analysis. The Janus electrode transported the sweat from the hydrophobic side to hydrophilic side for sample enrichment. Adapted from ref. 148 with permission. Copyright 2020, American Chemical Society. As has been discussed previously, the triphase interface can provide a stable and abundant concentration of gas reactants. Lei et.al. combined the MXene-based superhydrophobic electrode with wearable biosensors for the detection of glucose and lactate in sweat as shown in Fig. 10b . 141 The MXene-based superhydrophobic electrode exhibited high gas affinity and formed a stable liquid–gas–solid triphase interface. The existence of the triphase interface promised a high abundant concentration of oxygen during the detection process. The superaerophilic electrode presented high electrochemical sensitivity in the detection of glucose (35.3 μA mm −1 cm −2 ) and lactate (11.4 μA mm −1 cm −2 ) with LODs of 0.33 × 10 −6 M and 0.67 × 10 −6 M respectively. And the glucose values during the on-body tests showed good agreement with the reported blood and sweat glucose levels very well. Inspired by the binary synergistic wettability of lotus leaves, Janus films with asymmetric wettability have been developed and are promising in various fields, including water harvesting, fog collection, liquid separation, microfluidics, and wound dressing. 142–146 Incorporating Janus films with wearable biosensors can achieve the directional transport of sweat. The hydrophobic side of Janus films was placed on the skin. When sweat was secreted, the capillary force helped sweat transport from the hydrophobic side to the superhydrophilic side. 147 The collected sweat was retained in the superhydrophilic side for sample enrichment. Compared to the ordinary gauze, the Janus band showed excellent sweat transportation, resulting in an almost dry skin side. He et al. presented a Janus wettability textile strategy to transport sweat from the skin to the embedded electrode surface for the directional collection of sweat ( Fig. 10c ). 148 A Janus band was developed by combining a hydrophobic polyurethane (PU) and a superhydrophilic gauze through the electrospinning process. The PU side was covered on the skin, and the superhydrophilic side was linked to the electrode. The Janus band has shown good sensitivity in glucose (8 nA μM −1 ), lactate (67 nA μM −1 ), Na + (35.0 mV dec −1 ), and K + (45.6 mV dec −1 ). The on-body test can achieve real-time perspiration tracking, and the results were in good agreement with the physiological indicators of healthy people. In addition, Janus materials with sweat management can conduct thermal management on wearable biosensors. They have developed a Janus silk-based wearable sweat electrochemical biosensor with good wet-thermal comfort. 57 A conductive silk yarn electrode was woven into the Janus silk. When sweat was transported, the signal can be directly recorded, processed, and transmitted by the connected printed circuit board. The directional transport of the sweat ensures good thermal and humidity management, which improves the wearing comfort. The silk electrode showed excellent electrochemical stability through 800 cycle bending and a sensitive response to glucose (0.49 nA μM −1 ), urine acid (1.703 nA μM −1 ), pH (62.25 mV/lg(H + )) and K + (67.44 mV/lg(K + )). And the Janus silk showed a higher response rate than the ordinary silk in the on-body measurements due to its good ability of sweat transportation."
} | 12,881 |
26764164 | PMC4725873 | pmc | 2,771 | {
"abstract": "There have been many studies on superwetting surfaces owing to the variety of their potential applications. There are some drawbacks to developing these films for biomedical applications, such as the fragility of the microscopic roughness feature that is vital to ensure superwettability. But, there are still only a few studies that have shown an enhanced durability of nanoscale superwetting films at certain extreme environment. In this study, we fabricated intrinsically stable superwetting films using the organosilicate based layer-by-layer (LbL) self-assembly method in order to control nano-sized roughness of the multilayer structures. In order to develop mechanically and chemically robust surfaces, we successfully introduced polymeric silsesquioxane as a building block for LbL assembly with desired fashion. Even in the case that the superhydrophobic outer layers were damaged, the films maintained their superhydrophobicity because of the hydrophobic nature of their inner layers. As a result, we successfully fabricated superwetting nano-films and evaluated their robustness and stability.",
"discussion": "Results and Discussions Growth of superwetting films BF film assembly was driven by the hydrogen bonding force (amine groups in the BPEI and hydroxyl groups in the F-SiSQ worked as the hydrogen bonding acceptor and the hydrogen bonding donor, respectively) 40 . Figure 2(a,b) depicts the thickness growth and roughness profile of the LbL assembled BF film and the influence of post-treatment through a comparison of BF, BFA and BFFA films. First, BF film thickness and RMS roughness exhibited linear growth as function of the number of deposited bilayers. The thickness and roughness of the BF film show an average increase of 22.04 nm and 1.264 nm per bilayer, respectively. This result demonstrates the fact that a uniform silica particle monolayer is formed in every bilayer of the BF film, where the calculated radius of silica nanoparticles observed in F-SiSQ was 20 nm. In view of the uniform growth of the silica particle monolayer, RMS roughness also gives a roughly linear relationship with respect to number of bilayers. Therefore, by way of changing the number of bilayers, we could precisely adjust the thickness and roughness of the BF multilayer at the nanoscale. Second, the effect of heat treatment on thickness and roughness was also recorded. During heat processing, BFA film thickness and roughness per bilayer decreased by an average of 3.86 nm and 1.70 nm, respectively, until 40 bilayers of film had been deposited. Third, the thickness and roughness of the BFA and BFFA films were compared in order to determine the effect of the FDTS coating process. The thickness and roughness of the BFFA film was higher (0.560 nm, on average) and lower (0.452 nm on average), respectively, than for the BFA film. We observe that the influence of the FDTS monolayer coating is quite small due to its low thickness. As a result, the FDTS coating process has a negligible effect on BFA film morphology, and so the film could maintain its nano structure after post-treatments such as FDTS coating and the annealing process. Topography Figure 3(a–d) shows FE-SEM top view images of BFFA films with 10, 20, 30, 40 bilayers. From this data, we can determine that there exists a micro-nanoscale roughness in the BFFA film. P. -C. Lin et al . studied the effect of controlled dual-scale roughness on superhydrophobic characteristics 41 . As a result, the sliding angle of surfaces with micro-nanoscale roughness decreased more than 3 times compared to those exhibiting single nanoscale roughness. Thus, dual-scale roughness is most suitable when building a superhydrophobic film. In this study, a nanoscale roughness caused by the presence of nano-sized silica particles was observed on the BF, BFA, and BFFA films, and the micro-scale roughness was a result of silica nanoparticle agglomeration. As the number of deposited bilayers increased, the agglomeration of silica nanoparticles also approached the microscale. For BFFA films containing 10, 20, 30 and 40 bilayers, the sizes of the agglomeration were on average 142.0 ± 41.6 nm, 252.77 ± 43.5 nm, 303.8 ± 35.5 nm, and 540.4 ± 56.5 nm respectively. FE-SEM cross-section images of a BFFA film containing 20 bilayers are shown in Fig. 4 . An image with increased magnification, shown in Fig. 4(b) , displays the nano roughness structure of the BFFA film, together with a nanoporous structure. This nano roughness structure is particularly desirable for a superwetting film 42 43 , and a porous structure increases film transmittance by reducing the reflexive index of the film 44 45 . The low-magnification image presented in Fig. 4(a) shows a BFFA film of a homogeneous thickness, due to the LbL assembly molecular level film deposition process, while the BFFA film exhibits a particle-based rough structure. The AFM images of the BF, BFA, BFFA films show topology changes after post-treatment (Described in Fig. S2 , Supporting information ). As previously mentioned, the FDTS monolayer coating and annealing processes decreased the RMS roughness of the films (BF film: 41.9 nm, BFA film: 38.5 nm, BFFA film: 37.8 nm). Any topology changes resulting from heat treatment can be determined by comparison with the BF ((a), (b)) and BFA ((c), (d)) films. The z-direction displacement of the film decreased after the annealing process but the BFA film still exhibited a nano structure similar to that of the BF film. The effect of FDTS monolayer coating can be analyzed by comparing the BFA and BFFA films ((e), (f)). As a thin, uniform coating of the FDTS monolayer was applied to the BF film, only slight morphological differences between the two films could be detected. Wetting properties Several studies on superhydrophilic and superhydrophobic dual-wetting properties have previously been reported 22 23 24 and the critical factors contributing to the superwetting characteristics of a surface are its roughness and surface chemistry. We have confirmed morphology of BF, BFA, and BFFA films. In this section, we mainly focus on the surface chemistry changes resulting from FDTS coating and annealing steps. After the impact of post-treatments, the BF, BFA, and BFFA films were superhydrophilic, intrinsically hydrophobic, and superhydrophobic, respectively. Furthermore, we will compare the surface chemistries of the BF, BFA, and BFFA films. Due to the hydrophobicity of silsesquioxane, a small quantity of PFAS served as a coupling agent for F-SQ. The as-synthesized F-SQ therefore contained a large amount of hydrophilic-hydroxyl groups and hydrophobic-perfluoroalkyl groups. Due to the impact of long perfluoroalkyl chains, the F-SQ coated films displayed hydrophobic wetting characteristics (SCA = 110°). However in the case of F-SiSQ coated films display hydrophilic wetting property, because the majority of the silica nanoparticles coexisted with plenty of hydroxyl groups, and therefore the influence of perfluoroalkyl chains was reduced. F-SiSQ and BPEI containing hydrophilic-amine groups were used as the building blocks for the BF film, which means that the BF film was superhydrophilic, as shown in Fig. 5(a) . Even though only 5 bilayers of BF film had been deposited, the average of SCA values was 3.1°. Therefore, we can readily fabricate a superhydrophilic film containing several bilayers. A superhydrophilic film can be converted into a hydrophobic surface through an annealing process. The BFA film exhibited intrinsically hydrophobic properties due to the presence of hydrophilic functional groups such as amine and hydroxyl groups, which were almost completely removed during an annealing step at a temperature of 200 °C, in vacuum conditions for 4 h. Hence, the effect and also the proportion of perfluoroalkyl chains increased, and thus the BFA film exhibited intrinsically hydrophobic characteristics. The wetting properties of BFA films as a function of number of deposited bilayers are shown in Fig. 5(b) , and changes in the functional groups of both the BF film and the F-SQ were caused by the annealing process, as described in Figs S3,4 , Supporting information . The average contact angle of the BFA films is presented in the range from 5 to 40 bilayers (at intervals of 5 bilayers). Similar to the case of the BF film, only a small number of BFA bilayers are needed in order to create intrinsically hydrophobic surfaces. The sample containing 5 bilayers of BFA exhibited intrinsically hydrophobic properties with an SCA of 127°, while the sample coated with 40 bilayers of BFA exhibited a maximum SCA and lowest contact angle hysteresis (CAH) value of 134° and 52°, respectively. Deviations in the SCA values of the BFA film were not large (with an average of 131 ± 3°), although the CAH values of the BFA films exhibited a trend of decreasing with increasing number of deposited bilayers, which arose from the increase in the RMS roughness. In order to understand the intrinsically-hydrophobic characteristics of the BFA film that were caused by annealing, Fourier Transform Infrared Spectroscopy (FT-IR) was employed. FT-IR spectra of F-SQ depicts the, before ((a), a pristine sample) and after the annealing process ((b), an annealed sample), together with a table giving the corresponding peak positions. Before the annealing process, F-SQ contained a large number of hydroxyl groups, and so the silanol (Si-OH) peaks were observed at 3326 and 927 cm −1 . However, in high-temperature conditions, almost all the hydroxyl groups in the F-SQ condensated and underwent a conversion into siloxane (Si-O-Si) groups. Therefore, B (the annealed F-SQ) sample presented a 13.38% increase in the Si-O-Si band area (asymmetric cyclic Si-O-Si: 1143 cm −1 , linear asymmetric Si-O-Si: 1070 cm −1 , and linear symmetric Si-O-Si : 787 cm −1 ) 46 47 and a 78.00% decrease in the Si-OH bands area, which indicates that a high degree of condensation occurred during the annealing process. These phenomena show not only a decline in the amount of hydrophilic functional groups but also an increase in the coherence between heat-treated BF film components. The FT-IR spectra of the BF, BFA, and BFFA films also indicate the reinforcement of coherence between film components. In the BF film spectrum, the peaks at 2932 and 1582 cm −1 were in correspondence with the secondary amine in the BPEI. However, distinct changes can be observed in the BFA and BFFA films compared with the BF film. The absorption peaks of the secondary amine groups were reduced and converted into primary amine peaks (1655 cm −1 ) 48 49 , and these changes indicate a condensation reaction in the BFA and BFFA films. In order to fabricate superhydrophobic surfaces, an FDTS monolayer was deposited on the dual-scaled BF film using the self-assembly monolayer method to control the surface chemistry of the film. A film containing only 5 deposited bilayers gives a SCA of 162° and a CAH value of 5°, which are located in the superhydrophobic region (SCA > 150°, CAH > 10°). We noted a decrease in CAH and increase in SCA for the BFFA films as a function of the number of deposited layers. The SCA and CAH values for a BFFA film containing 40 bilayers were 171° and 1°, respectively. These phenomena indicate that the BFFA films exhibited a uniform nano structure and their surface chemistry was well controlled by the fluoroalkyl chains in the FDTS monolayer. The presence of fluoroalkyl chains in the FDTS monolayer of a BFFA film was verified using the FTIR spectra. As the FDTS monolayers were very thin (0.462 nm), the intensity of perfluoroalkyl chains was low, although we observed increments in the perfluoroalkyl chains bands, such as the rocking CF 2 peak (647 cm −1 ) and the wagging CF 2 peak (571 cm −1 ), in relation to the BF and BFA films. The contact angle and FTIR results clearly demonstrated that the surface chemistry and wetting properties of the BF, BFA, BFFA films differed. Transmittance Figure 5(d) shows the BFFA 20 bilayer film on quartz glass. Despite its thickness of 341.16 nm, the film displayed antireflection properties (AR), together with a high transmittance. The optical transmittance of superhydrophobic BFFA films containing various numbers of deposited bilayers on quartz glass, that were measured using UV-vis spectroscopy in the visible light range is presented in Fig. S5 , Supporting information . We deposited 10 and 20 bilayers of BFFA film in order to evaluate the effect of varying film thickness on film transmittance. The transmittance graphs confirm several points. First, we may note that the transmittance of BFFA films was higher than that of bare quartz glass substrates at wavelengths beyond 412 nm (10 bilayer sample) and 500 nm (20 bilayer sample). When compared with uncoated quartz, the quartz covered with a superhydrophobic BFFA film exhibited AR properties as a result of two factors. In order to fabricate films with AR properties, many reports have been focused on porous surfaces and materials with a low refractive index 50 . The first factor is the presence of F-SiSQ in the BFFA film, which has a low reflective index, and the second is related to the nanoporous structure of the BFFA film. M. Saito et al . reported on the refractive losses of polymethyl methacrylate (PMMA), polyimide, and fluorocarbon polymer (CYTOP) 39 . In this study, the reflective index of CYTOP was found to be 1.35 ± 0.05, while the minimum refractive index in a homogeneous dielectric material is 1.34, due to their low absorption coefficients. H. Hoerauf et al . also examined the refractive indices of various perfluoroalkyl-substitituded liquids, such as perfluoro-octane, perfluoro-decane, and perfluoro-hexylethane, and their refractive indices ranged from 1.27 to 1.34 51 . Low-refractive-index materials effectively reduce the amount of light loss caused by reflection. Fluoroalkyl chains are superior to the other polymers for use as an AR coating material. The nanoporous structure of a BFFA film was detailed in Fig. 4(a) . There are various approaches used to fabricate nanoporous surface structures, such as the block copolymer film 52 , particle based sol-gel film 53 , and surface etching film 54 methods, which effectively reduce the refractive index of a film. Due to the low refractive index of BFFA films, they are highly transparent and can be used as antireflection coatings. There are distinct differences in the optical transparencies exhibited by BFFA coatings of different thicknesses. V. Kumar et al . investigated the optical properties of films, stating that they depend strongly on film thickness 55 . The transmittance of a film will be decreased with an increasing number of bilayers. The thicknesses of the 10 and 20 bilayer BFFA films in this study were 145.05 nm and 342.16 nm, respectively, as shown in Fig. 2(a) . The optical transmittance of the (BFFA) 10 films was almost 5.48% lower than that of the (BFFA) 20 film, at a wavelength of 550 nm, due to its thickness and the red-shift in the absorption edge. In order to fabricate an ideal AR coating, the thickness of the coating should be equal to /4 45 56 . The optical thickness of the BFFA film was 137.5 nm, almost equal to that of 8 bilayers of BFFA film. However, as discussed in Fig. 5(c) , the CAH of the BFFA film finally reached the value of less than 2° when more than 15 bilayers of film were deposited. As the number of deposited bilayers was increased, the superwetting characteristics of the films were improved, but the film transmittance decreased as a result of increasing film thickness. Also, the high roughness accompanying a large number of BFFA film bilayers caused light scattering that reduced the AR properties of a film 45 . However, 20 bilayers of BFFA film with superhydrophobic properties exhibited a high transmittance in the range from 400 nm to 800 nm, together with a transmittance of approximately 99.99% at 550 nm, whereas the transmittance of untreated quartz glass at this wavelength was only approximately 96.62%. These experimental results indicate that BFFA films are a promising candidate for self-cleaning windows that require superhydrophobicity and high transmittance. Film stability Our proposed film fabrication method mainly employed silsesquioxane as a layer-by-layer building block because of their high stability resulting from the presence of strong siloxane bonds. The stability of our films was revealed by the presence of covalent bonding between the BPEI and the F-SiSQ, and the F-SiSQ with itself, as discussed before, by using the FTIR results. Here, the stability enhancements of superwetting films are discussed, which may be considered to be the strongest point of our research. The mechanical durability of our prepared superwetting films was tested using tape peeling, bending, and phosphate buffered saline (PBS) stability tests, and their thermal stability was verified by heat testing. A quantitative discussion of the physical strength of the films arising from the presence of silsesquioxane is also included. Figure 6(a) shows that the static contact angle for the prepared superhydrophobic surface of the BFFA film, as a function of the number of times the tape was peeled. It could resist the peeling test 50 times, with a slight decline in hydrophobicity. At first, the SCA of the BFFA film was 172°, but after peeling 20 times, the SCA was reduced to 139°, and then this contact angle was maintained up to 50 peeling times. This decrease in the contact angle at the early stages (10 bilayers) can be explained by a delamination of weakly molecular level thickness of FDTS monolayers from the surface of the BF film. During 10 peeling tests, the unbonded FDTS monolayer was partially eliminated, and only few FDTS monolayer remained on the BFFA film. There is also an intrinsically hydrophobic layer (134°) under the FDTS monolayer coating. For these reasons, the BFFA film maintains its hydrophobicity during additional peeling steps. In addition, we also attempted to estimate the stability of BF films using a peeling test. However, it was difficult to measure the changes in the contact angle due to high adhesive forces between the BF film and the scotch tape ( Fig. S6 , Supporting information ). After one step of tape peeling test, the adhesive material of the scotch tape became strongly attached to the BF film without delamination. That indicates that the BF film strongly combines with the silicon wafer through hydrogen bonding. The interfacial fracture toughness of superwetting films has been evaluated using a bending test. As shown in Fig. 6(b) , 20 bilayer BF and BFFA films stacked on overhead projector (OHP) film were bent to various degrees (from 0° to 70°) and the bent structure was maintained for 1 min, after which the SCA and CAH values were measured. After the bending tests that performed drastic transformations, the SCA values of BFFA films were maintained at almost 170°, and the CAH of the BFFA film exhibited a slight change, but still maintained its initial value. As the BFFA and BF films were influenced by the OHP film substrate, the CAH values of the film used in the bending test were different from those of films deposited on the silicon wafer. As could be expected, the BF films maintained their superhydrophilic characteristics (SCA < 5°) after a 70° bending test. These results indicate that our films exhibited high flexibility without changes to their super-wetting properties, due to strong siloxane bonding in the F-SQ and coherence between the layer-by-layer building blocks. An evaluation of the phosphate buffered saline (PBS) stability of BFFA films is given in Fig. 6(c) . A high cohesion force between film components with high proportion of covalent bonding leads to a lower swelling ratio for a film in a good solvent which can dissolve components of film. There is a standard for calculating the degree of crosslinking within a film by determining the amount of swelling resulting from immersion for a day in a good solvent 57 . Due to nano-scale thicknesses of our superwetting films, we measured the change in film thickness instead of measuring the change of in mass or volume. Superhydrophobic and superhydrophilic (BFFA) 20 and (BF) 20 films were immersed in PBS solution for 1 day, and their thicknesses were measured as a function of immersion time in order to indirectly measure the degree of crosslinking. As the driving force of BF film formation is hydrogen bonding, it is the weaker BFFA film that consists of strong covalent bonds. During the PBS solution stability test, the thickness of the BF film was maintained for 2 hours, although after this point, the film thickness began to decrease. ( Fig. S7 , Supporting information ) However, the BFFA film maintained its thickness for more than 24 hours in PBS solution ( Fig. 6(c) ) and its maximum variation in its thickness was only approximately 2.6%. The BF film was weaker than the BFFA film in PBS solution because the hydrogen bonds in the BF film were weaker than the covalent bonds in the BFFA film, These results indicate that the BFFA film was stable in PBS solution because it contained a number of covalent bonds. The contact angles of heat-treated BF and BFFA films were used in order to evaluate the thermal stability of the superwetting films, as shown in Fig. 6(d) . For heat treatments in the temperature range from 100 °C to 300 °C, the SCA values of the BF and BFFA films were below 5° and above 165°, respectively, which indicates that the films maintained their superwetting properties at high temperatures. As the annealing temperature increased, the CAH values of heat-treated BFFA films slightly increased (from 9.17° in 100 °C to 14.96° in 300 °C). This is due to the loss of the FDTS coating (with a boiling point of 240 °C) and BPEI (with a boiling point of 250 °C) during the heat-treatment. When the film was heat-treated at 300 °C, its SCA and CAH values were 166.60° and 14.96°, respectively, and it approached the sufficient conditions for a superhydrophobic film (SCA > 150°, CAH < 10°). These results demonstrated that the prepared superwetting films exhibited a high resistance above the boiling points of the film components. This is due to the fact that the layer-by-layer building blocks and self-assembly monolayer coatings on the BFFA film were covalently bonded, and these bonds suppressed the evaporation of FDTS and BPEI from the film during heat treatment. The silica included in the F-SiSQ also acted as a thermal enhancement compound, improving the heat resistance of the superwetting films. Images of BFFA films after heat treatment are presented in Fig. S8 , Supporting information . We improved the stability of our film through the introduction of silsesquioxane, which consists of several strong siloxane bonds. In order to verify the stability enhancement that may be attributed to silsesquioxane, we fabricated a (BPEI/Si-NP) n film without F-SQ, which was subject to annealing under the same conditions as the BFA film (at 200 °C for 4 h). We then compared its hardness and Young’s modulus with those of the BFA film using nanoindentation (See Table 1 ). In order to facilitate precise comparison, the film thicknesses were of similar values, of approximately 600 nm. The local hardness and Young’s modulus values of the (BPEI/Si-NP) 20 and (BFA) 35 films are shown in Table 1 . The average hardness and Young’s modulus of the BFA film were 2.9 and 1.4 times higher than the values of the (BPEI/Si-NP) n film, respectively. It is clear that the hardness and Young’s modulus increased with the presence of F-SQ. There are two reasons for this increase. First, the BFA film that contained F-SQ was composed of a harder structure compared to the (BPEI/Si-NP) n film, because the silica nanoparticles were well dispersed in the polymer-like F-SQ matrix. A different number of bilayers was required to reach a thickness of approximately 600 nm (BFA: 35 bilayers, (BPEI/Si-NP) n : 20 bilayers). The second reason is that the rigid silica core included in silsesquioxane improved the mechanical properties of the BFA film. Application to large-area process Textured surfaces generated by lithography, etching, and electrophoretic deposition methods are difficult to fabricate for films with a large area. In the case of particle-based layer-by-layer assembled textured surfaces, films can be generated on different types and sizes of substrates using various fabrication methods, such as dipping, spraying, spin coating, and printing. We fabricated a successfully manufactured transparent and superhydrophobic (BFFA) 20 film on a large area (9 cm × 9 cm) of glass using the layer-by-layer dipping method (Described in Fig. 7 ). Taking full advantage of layer-by-layer methods, superwetting films can be employed in large-area processing for use in various commercial applications. In summary, superwetting films were prepared by particle-based LbL assembly using silsesquioxane and silica nanoparticles in order to obtain dual-scale roughness and improved durability. In order to obtain various wetting properties, subsequent post-treatments such as annealing and FDTS coating were employed. Untreated BF films and post-treated BFFA films exhibited superhydrophilic and superhydrophobic wetting properties, respectively. Due to their nanoscale thickness and the presence of the perfluorocarbon compound in the BF and BFFA films, they exhibited a high transparency. The presence of the silica core in F-SiSQ also successfully improved the durability of the superwetting films, as was evident from the nanoindentation results. The BFFA film was more stable than the BF film, due to the condensation caused by heat treatment. There are two differences between these films that may be attributed to heat treatment: the increment of the degree of condensation present and the elimination of hydroxyl groups that induce the intrinsically hydrophobic properties of the BFA layer (the inner layer of the FDTS coating). Therefore, the components of the BFFA film are covalently bonded, and it maintains its hydrophobicity if the FDTS monolayer is eliminated. The durability of the superwetting film was verified using a peeling test, a bending test, a heat test, and a Phosphate Buffer Saline (PBS) stability test. By taking full advantage of LbL assembly, the proposed superhydrophobic and superhydrophilic films can be generated for various compositions and sizes, and easily adopted for use in various industrial applications."
} | 6,650 |
36542271 | PMC9958174 | pmc | 2,772 | {
"abstract": "Quantifying cellular components is a basic and important step for understanding how a cell works, how it responds to environmental changes, and for re-engineering cells to produce valuable metabolites and increased biomass. We quantified proteins in the model cyanobacterium Synechocystis sp. PCC 6803 given the general importance of cyanobacteria for global photosynthesis, for synthetic biology and biotechnology research, and their ancestral relationship to the chloroplasts of plants. Four mass spectrometry methods were used to quantify cellular components involved in the biosynthesis of chlorophyll, carotenoid and bilin pigments, membrane assembly, the light reactions of photosynthesis, fixation of carbon dioxide and nitrogen, and hydrogen and sulfur metabolism. Components of biosynthetic pathways, such as those for chlorophyll or for photosystem II assembly, range between 1000 and 10,000 copies per cell, but can be tenfold higher for CO 2 fixation enzymes. The most abundant subunits are those for photosystem I, with around 100,000 copies per cell, approximately 2 to fivefold higher than for photosystem II and ATP synthase, and 5–20 fold more than for the cytochrome b 6 f complex. Disparities between numbers of pathway enzymes, between components of electron transfer chains, and between subunits within complexes indicate possible control points for biosynthetic processes, bioenergetic reactions and for the assembly of multisubunit complexes. Supplementary Information The online version contains supplementary material available at 10.1007/s11120-022-00990-z.",
"conclusion": "Conclusions We employed four mass spectrometry-based quantification methods, which revealed the cellular levels of 97 proteins involved in photosynthesis, the biosynthesis of carotenoid, chlorophyll and bilin pigments, membrane assembly, the light reactions of photosynthesis, fixation of carbon dioxide and nitrogen, hydrogen, and sulfur metabolism. Figure 5 summarizes the cellular locations and associations of these proteins, also indicating that regulatory and biosynthetic components are generally less abundant than those involved in bioenergetic reactions. We also found disparities in the abundances of subunits relative to their stoichiometries apparent in high-resolution structures, as expected given that both complete complexes and assembly intermediates would contribute to the analysis. Furthermore, we were able to calculate cellular levels for some large complexes such as photosystems, assemblies such as carboxysomes and VIPP1 oligomers, and hexameric FtsH Zn-metalloproteinases. Our quantitative proteomic baseline for wild-type Synechocystis enhances our understanding of this model organism and will be a valuable resource for the photosynthesis community, as well as forming the basis for synthetic biology projects aimed at manipulating biosynthetic, metabolic and energy-transducing pathways. Synechocystis is an important model organism for engineering photosynthetic metabolism and knowing the numbers of the essential components will inform attempts to use this bacterium as a chassis for using sunlight, CO 2 and water to produce valuable metabolites and increased biomass. While the focus of this study has been on the 97 proteins described, cellular levels of 1081 proteins are also presented. Fig. 5 Diagrammatic summary of the proteins quantified in this study, showing the range of biological processes, such as biosynthetic and assembly pathways, covered by the quantitative mass spectrometry analysis. Abbreviations are defined in Figs. 1 , 2 , 3 , 4 . Proteins and subunits of complexes that have been quantified are colored in blue and shaded according their abundance levels; those not quantified are in white. Complexes such as photosystem I, photosystem II and cytochrome b 6 f are drawn as monomers for simplicity. Thylakoids are drawn as elongated tubular structures, which converge on a thylakoid convergence zone that appears to connect plasma and thylakoid membranes (Stengel et al. 2012 ; Heinz et al. 2016 ). The photosystem II assembly intermediates are based on those in Konert et al. ( 2022 ) and Rahimzadeh-Karvansara ( 2022 ), with the exception of the PSII-I assembly complex, the structure of which was determined by Zabret et al. ( 2021 )",
"introduction": "Introduction Cyanobacteria evolved approximately 2.4 billion years ago as the only prokaryotes that utilize solar energy to oxidize water. Electron transport coupled to proton translocation then drives the production of the ATP and NADPH needed for CO 2 fixation and other metabolic processes and reviewed by Lea-Smith et al. ( 2016 ). Cyanobacteria have colonized almost every terrestrial and aquatic habitat, with marine species alone responsible for an estimated carbon capture rate of 4 × 10 12 kg y −1 (Rousseaux and Gregg 2014 ), contributing 3.8% of global net primary production (Field et al. 1998 ). In view of their ecological significance, the effects of the anthropogenic increase in atmospheric CO 2 on cyanobacterial populations and their influence on the entire biosphere are important areas of study (Ullah et al. 2018 ).\n Some avenues of research are specific for cyanobacteria, but others have wider relevance to photosynthesis in algae and plants due to the common ancestry of cyanobacteria and chloroplasts (Yoon et al. 2004 ). Similarities between cyanobacterial and eukaryotic photosystems have led to the adoption of several species of cyanobacteria, often thermophiles, as models for exploring the fundamental mechanisms that underpin oxygenic photosynthesis (Ferreira et al. 2004 ; Umena et al. 2011 ; Suga et al. 2015 ; Gisriel et al. 2019 ; Çoruh et al. 2021 ). The early availability of a complete genome sequence for Synechocystis sp. PCC 6803 (hereafter Synechocystis ) was also an invaluable resource for studies of photosynthesis and many other cellular functions. This was the first such information for any phototroph, and only the third genome sequence for any organism (Kaneko et al. 1996 ). This advance, as well as the amenability of this cyanobacterium to genetic manipulation (Vermaas 1994 ), hastened the adoption of Synechocystis as a model for photosynthesis research. One recent example, which builds upon the evolutionary relationship between chloroplasts and cyanobacteria, is the use of Synechocystis as a platform for the rapid development of genetic diversity by adaptive evolution. Laboratory-evolved mutations in cyanobacteria such as Synechocystis that confer enhanced efficiency for converting solar energy into biomass may then be transferable to crop plants with the aim of increasing yield (Leister 2012 ; Dann and Leister 2017 ; Dann et al. 2021 ). Model cyanobacterial species have also been valuable for designing solar-powered systems for synthetic biology (reviewed by Sengupta et al. 2018 ). In view of its universally recognized importance in photosynthesis research, Synechocystis has been a frequent target for mass spectrometry (MS)-based proteomic analysis, as reviewed by Gao et al. ( 2015 ) and Battchikova et al. ( 2018 ). A principal focus has been on the comparative quantification of protein abundance following adaptation to varying culture conditions (for example see Fulda et al. 2000 ; Hong et al. 2014 ; Angeleri et al. 2019 ) and mutant strains have been used as tools for the dissection of adaptive and regulatory pathways (for example see Tokumaru et al. 2018 ; Krynická et al. 2019 ). This approach has also mapped proteins to their subcellular locations to track the development of thylakoid membranes (TMs), the specialized photosynthetic membranes that cyanobacteria, algae and plants all have in common (Kwon et al. 2010 ; Pisareva et al. 2011 ). More recently, a proteomic catalog of subcellular localization has been produced, comprising 1712 proteins (Baers et al. 2019 ). Photosynthesis research is starting to encompass larger and larger structures, from complexes, to supercomplexes, to membrane organization studied by atomic force microscopy and finally to whole cells (MacGregor-Chatwin et al. 2017 ; Casella et al. 2017 ; Zhao et al. 2020 ). Here, spectacular recent advances in cryo-electron tomography, augmented by super-resolution fluorescence imaging, are starting to reveal the molecular details of cyanobacterial cells, their internal cellular components and their membrane architectures (Rast et al. 2019 ; Huokko et al. 2021 ). With this focus on cells it is appropriate to count cellular components in terms of the copy numbers of proteins per cell, which would appear to be a basic requirement for understanding cellular function, and for manipulations of cells and their pathways for synthetic biology purposes. Yet despite the considerable volume of proteomic information now available for Synechocystis and other cyanobacteria, to our knowledge there has been no MS-based absolute quantification of proteins in terms of copy number per cell (cpc). So far, cpc determination has been confined to photosystems I and II (PSI and PSII) using absorption spectra associated with their bound chlorophylls (Fujita and Murakami 1987 ; Hihara et al. 1998 ; Keren et al. 2004 ; Fraser et al. 2013 ; MacGregor-Chatwin et al. 2017 ). MS-derived absolute quantitative studies of proteins are nevertheless commonplace, having been undertaken in many different organisms and subcellular systems. Some examples, employing both stable isotope labelled (SIL) standards and label-free approaches, are: Escherichia coli (Wiśniewskia and Rakus 2014 ), Leptospira interrogans (Malmström et al. 2009 ), chromatophores in Rhodobacter sphaeroides (Cartron et al. 2014 ), glycolytic pathway enzymes in Saccharomyces cerevisiae (Carroll et al. 2011 ) and xenobiotic metabolizing enzymes in human liver microsomes (Li et al. 2015 ). Two of the SIL-based methods are confined to the absolute quantification of single or low numbers of proteins and involve calibration with either 15 N-labelled synthetic peptide fragments (usually tryptic) mapping to the target protein(s) (Kirkpatrick et al. 2005 ) or full-length SIL proteins produced in E. coli grown in 15 N-containing liquid culture (Brun et al. 2007 ; Singh et al. 2009 ). For larger scale absolute quantification the high-financial cost of the former and long lead-time of the latter are potential obstacles that make label-free quantification (LFQ) methods more attractive and therefore more widely used. Since no single LFQ method has emerged as the ‘gold standard’, numerous performance comparisons can be found in the literature (for example see Arike et al. 2013 ; Fabre et al. 2014 ; Krey et al. 2014 ; Al Shweiki et al. 2017 ; Tang et al. 2019 ). Here, we employed a third SIL-based quantification method that is more feasible for this larger scale study than the two options described above. It is also well characterized and uses artificial 15 N-labelled proteins comprising concatenated tryptic peptides mapping to the target proteins (Pratt et al. 2006 ; Brownridge et al. 2013 ). To provide further validation and extend the range of target proteins, we additionally employed three LFQ methods: (1) iBAQ (intensity-based absolute quantification) (Schwanhäusser et al. 2011 ) based on data-dependent acquisition (DDA) and Top3 (Silva et al. 2006 ), based on both (2) DDA and (3) data-independent acquisition (DIA) (Venable et al. 2004 ). We demonstrate that our MS-based absolute quantification of PSI and PSII complexes is in close agreement with copy numbers determined spectrophotometrically, both here and in earlier studies (Fujita and Murakami 1987 ; Hihara et al. 1998 ; Keren et al. 2004 ; Fraser et al. 2013 ; MacGregor-Chatwin et al. 2017 ). Similarly, we show subunit stoichiometry in the ATP synthase complex aligns with the known structure for the complex in bacteria (Guo et al. 2019 ). Having validated our approach, we present the first report of MS-based absolute quantification of cytochrome b 6 f (cyt b 6 f ) subunits, associated mediators of electron transport and downstream electron-accepting metabolic processes including CO 2 fixation. Also quantified are enzymes and auxiliary proteins in the chlorophyll (Chl), carotenoid and phycobilin biosynthesis pathways, together with assembly factors implicated in TM biogenesis and PSI and PSII assembly. These 97 proteins and their interrelationships are summarized in Figs. 1 , 2 , 3 , 4 , 5 . Finally, we exploit our quantitative analyses to present the cellular abundances of 1081 proteins in Synechocystis .",
"discussion": "Results and discussion Validation of protein identification and quantification Validation of the cell counting, protein identification and absolute quantification methods are described in Supplementary Information. Conversion of solar to chemical energy Photosystem II PSII is a multi-subunit complex that is integrated into the TM and contains Chl cofactors that enable its photochemical function (Ferreira et al. 2004 ). The process of oxygenic photosynthesis in Synechocystis (reviewed by Lea-Smith et al. 2016 ) is initiated when solar energy captured by phycobilisomes is transferred to PSII to drive the abstraction of electrons from water, producing O 2 and releasing protons into the thylakoid lumen. The four core PSII subunits were quantified: PsbA-D, also referred to as D1, CP47, CP43 and D2 respectively, alongside one of the small subunits, PsbH. Figure 1 a shows that the reaction center heterodimer subunits D1 and D2 are closely aligned at 24,000–44,000 cpc, indicated by the two horizontal dashed lines. This range is not only consistent with structural studies establishing the 1:1 stoichiometry of D1 and D2 (Umena et al. 2011 ; Suga et al. 2015 ; Gisriel et al. 2022 ) but also a functional assay of PSII abundance, using flash-induced O 2 yield in the related Synechocystis sp. PCC 6714 (Fujita and Murakami 1987 ), at 18,000–22,000 cpc. At the lower extent of their abundance range, CP47 and CP43 are within 24,000–44,000 cpc, however at the upper extent these subunits occur at approximately 100,000 cpc, substantially in excess of the one subunit per complex stoichiometry in PSII (Umena et al. 2011 ; Suga et al. 2015 ; Gisriel et al. 2022 ). This observation may be explained by evidence for an excess of the CP47m and CP43m assembly modules over D1- and D2-containing intermediates (Tichý et al. 2016 ; Bečková et al. 2022 ), all of which would be included in our whole cell analysis. We therefore suggest that the cellular level of functional PSII complexes in Synechocystis , grown under the conditions employed here, is 24,000–44,000 cpc. Fig. 1 Cellular levels of protein complexes involved in the conversion of solar to chemical energy. Consensus cpc ranges for the subunits are condensed from the different quantification methods, as described in Results and Discussion, with individual data-points shown in boxplots in Supplementary Fig. S2. The minimum and maximum cpc values are rounded to the nearest 10 (< 1000), 50 (1000–10,000) or 500 (> 10,000) and displayed as bars, shaded according to cpc range, for photosystem II (PSII, PsbA-H), photosystem I (PSI, PsaA-F) ( a ), cytochrome b 6 f (cyt b 6 f ), PetP, phycocyanin (PC), photosynthetic NAD(P)H dehydrogenase-like complex type-1 (NDH-1) (b) and ATP synthase (c). Abundance levels corresponding to the largest extent of overlap between subunits, shown by the horizontal dashed lines, are taken to represent probable ranges for the complexes, explained in Results and Discussion: PSII (24,000–44,000 cpc), PSI (86,000–118,500 cpc), cyt b 6 f (3350–8450 or 14,000–28,000 cpc), NDH-1 (7000–13,500 cpc), ATP synthase ( α , β : 67,000–83,500 cpc; γ , δ , ε , a, b, b': 12,000–25,000 cpc) Like CP47 and CP43, the 7 kDa subunit PsbH is shown to occur markedly in excess of D1 and D2, at 155,500–187,000 cpc (Fig. 1 a). Since PsbH has a stabilizing function, as an obligatory component of both complete, active PSII (Komenda et al. 2002 ) and several of its assembly intermediates (Komenda 2005 ), this high level would be expected. Cytochrome b 6 f complex and plastocyanin The cyt b 6 f complex provides the link in the electron transport pathway between PSII and PSI in oxygenic photosynthesis and, like the photosystems, is integral to the TMs (Lea-Smith et al. 2016 ). In oxygenic phototrophs, this complex comprises four major and four minor subunits (reviewed by Malone et al. 2021 ). The four major subunits, PetA, PetB, PetC and PetD (cyt f , cyt b 6 , the Rieske Fe-S protein (ISP) and Subunit IV (subIV) respectively) were quantified at 3500–6250, 14,000–20,000, 19,000–28,000 and 2700–8450 cpc, respectively (Fig. 1 b). Despite the known stoichiometry of 1:1:1:1 in the cyanobacterial and plant cyt b 6 f structures (Malone et al. 2021 ), these ranges unexpectedly fell into lower (cyt b 6 and subIV) and higher (cyt f and ISP) groups. The lower end of the cpc range for subIV aligns with the 2700 cpc determined in Thermosynechococcus elongatus (hereafter T . elongatus ) by Rexroth et al. ( 2014 ) who also reported an abundance of 2500 cpc for ISP, in agreement with the known stoichiometry within the complex. However, the PSII:cyt b 6 f ratio of 1.08–1.14 reported by Fujita and Murakami ( 1987 ) alongside our determination of PSII at 24,000–44,000 cpc suggests that, in Synechocystis , cyt b 6 f may actually align with the higher abundance cyt b 6 and ISP at 14,000–28,000 cpc. The ISP identified in this analysis was the principal PetC1 isoform encoded by the sll1316 gene. The alternative isoform (PetC2, slr1185-encoded) which is produced under low oxygen (Summerfield et. al 2008 ) or high light stress (Tsunoyama et al. 2009 ) was undetectable. The third ISP, PetC3 (sll1182-encoded) was identified in our analyses (Supplementary Data Sets S1 and S2), however, it is localized to the cytoplasmic, not thylakoid membrane (Aldridge et al. 2008 ) where its function appears unrelated to photosynthetic electron transport (Veit et al. 2016a ). The minor subunits, PetG, PetL, PetM and PetN, encoded in Synechocystis by smr0010, ssl3803, smr0003 and sml0004 respectively (Schneider et al. 2007 ) were not identified in our analyses; these proteins are all smaller than 40 residues, hydrophobic and form a single transmembrane helix (TMH). Therefore, their resistance to trypsin digestion and consequent non-detection in these analyses might be expected. The putative regulator of electron transport, PetP (Ssr2998), previously characterized in T. elongatus (Rexroth et al. 2014 ; Veit et al. 2016b ), was quantified at 3350–12,000 cpc (Fig. 1 b). The lower end of this range coincides with the 3300 cpc determined for PetP in T. elongatus by Rexroth et al. ( 2014 ). Plastocyanin (PC, PetE), the copper-containing electron carrier between cyt b 6 f and PSI is found to be highly abundant at 101,500–182,000 cpc (Fig. 1 b). The alternative electron carrier to PC, cyt c 6 (PetJ) which is produced in response to growth conditions with low copper levels (García-Cañas et al. 2021 ) is detected at a low level (Supplementary Data Sets S1 and S2) but, with representation by a single peptide, was not validated for quantification. The dominance of PC over cyt c 6 is predictable under the nutrient-rich culture conditions used here; in the presence of Cu, expression of petJ is repressed while petE expression is induced by a mechanism involving the protease Slr0241 (García-Cañas et al. 2021 ), not detected here. Photosystem I PSI receives electrons from either PC or cyt c 6 (see above) and uses solar energy to drive the reduction of Fd, which functions in a wide range of metabolic pathways including NADPH generation in a reaction catalysed by ferredoxin-NADP + reductase (FNR, see below). NADPH is subsequently involved in, among other processes, the CBB cycle for CO 2 fixation (Lea-Smith et al. 2016 ; Figs. 2 , 3 and 5 ). Four PSI subunits were quantified: PsaA–C and PsaF. Figure 1 a shows abundance levels that cover the majority of data-points over the 86,000–118,500 cpc range. As in the case of the PSII D1 and D2 subunits, the data-point overlap for PsaA, PsaC and PsaF approximately fits a one subunit per complex stoichiometry (Çoruh et al. 2021 ). PsaB, appearing to be underestimated at 38,000–61,500 cpc, exemplifies the potential for inaccuracy with MS-based protein quantification resulting from, for example, idiosyncratic proteolysis and/or peptide ionization properties, as detailed in Supplementary Table S1. Our quantification of PSI aligns, at its lower end, with the 63,000–99,000 and 96,000 cpc determined spectrophotometrically by Fujita and Murakami ( 1987 ) and Keren et al. ( 2004 ) respectively. The latter authors also quoted their measurement of Chl content at 1.4 × 10 7 molecules/cell. On the basis that each PSI complex houses 95 Chl molecules (Malavath et al. 2018 ), 96,000 cpc would account for 9.1 × 10 6 Chls, approximately 65% of the total cellular content. Assuming the same proportion, our determination of 1.57 ± 0.04 × 10 7 Chl molecules/cell (Supplementary Table S4) correlates with 107,500 cpc of PSI, which is within the range determined by our MS-based quantification. Photosynthetic NAD(P)H dehydrogenase-like complex type-1 Under conditions of environmental stress that require an increased ATP:NADPH ratio, for example to meet a higher demand for protein synthesis, the proton gradient across the TM, utilized by the ATP synthase complex, can be elevated by cyclic electron transfer (CET; Kramer et al. 2004 ). One of the CET mechanisms used by cyanobacteria is mediated by photosynthetic NAD(P)H dehydrogenase-like complex type-1 (NDH-1; reviewed by Laughlin et al. 2020 ; Fig. 1 b). In Synechocystis , NDH-1 is composed of 19 subunits, distributed between membrane-intrinsic and peripheral arm regions (Pan et al. 2020 ). In our analysis two subunits located in the latter region, NdhI and NdhK are quantified at 8550–13,500 and 7000–11,500 cpc, respectively (Fig. 1 b), consistent with evidence that all NDH-1 subunits occur at one copy per complex (Pan et al. 2020 ). ATP synthase complex The (cyano) bacterial ATP synthase complex (Fig. 1 c) comprises two main functional components, one membrane extrinsic and the other embedded in the membrane bilayer (Guo et al. 2019 ). The membrane-extrinsic, catalytic F 1 component contains α - and β -subunits at three copies each with single-copy γ -, δ - and ε -subunits. The membrane-intrinsic, proton translocating F O complex contains a single a-subunit and, in Synechocystis , 14 c-subunits (Pogoryelov et al. 2007 ). Forming the peripheral stalk connecting the two sectors are the single-copy b- and b'-subunits. Here, we determined the ranges 67,000–83,500 cpc for α, β and 12,000–25,500 cpc for γ, δ, ε, b, b', giving a ( αβ ):(γδεbb') ratio of 2.6–7.0 (Fig. 1 c). The lower end of this range reflects the known stoichiometry of the ATP synthase complex, as detailed above (Guo et al. 2019 ). A higher (αβ) stoichiometry, approaching 7.0, may be accommodated by our quantification of fully assembled, functional complexes together with nascent complexes composed of partially assembled modules. Indeed several F 1 sub-assemblies have been detected in E. coli : γε (Rodgers and Wilce 2000 ), α 3 β 3 γ (Koebmann et al. 2002 ) and α 3 β 3 γε (Deckers-Hebestreit 2013 ). In E. coli , ATP synthase subunits were quantified at 2700–3700 cpc for α/β and 600–1700 cpc for γ/δ/ε (Wiśniewski and Rakus 2014 ), giving a (αβ):(γδε) ratio of 1.6–6.2, in close agreement with the present study. Based on γ/δ/ε cpc, the cellular abundance of ATP synthase complexes in Synechocystis is therefore approximately 15–40-fold higher than in E. coli. This marked difference in ATP synthase levels between these two organisms may be predictable given their vastly divergent metabolic characteristics. Biosynthesis Carboxysomes Carboxysomes are 100–200 nm icosahedral structures located in the cytoplasm of cyanobacteria (Faulkner et al. 2017 ), each comprising a self-assembling multi-protein shell that houses two of the Calvin–Benson–Bassham (CBB) cycle enzymes: carbonic anhydrase (CcaA) and ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO). Two types of carboxysomes ( α and β ) have been defined based on their associated RuBisCO isoforms, with the β -type occurring in Synechocystis (Badger et al. 2002 ). The shell contains pores that selectively allow HCO 3 − , the substrate for CcaA to enter while preventing the exit of CO 2 and the entry of O 2 (Cai et al. 2009 ). Thus, carboxysomes are nano-compartments in which CO 2 becomes concentrated and O 2 , the competitive inhibitor of RuBisCO with respect to CO 2 is excluded. Six of the seven proteins that comprise the carboxysomal shell are quantified in this study, together with one of the two proteins reported to function in assembly and organization (Cameron et al. 2013 ). The CcmK1–4 proteins first assemble as homohexamers alongside mixed stoichiometry K1–K2 and K3–K4 heterohexamers, each with a central pore. The hexamers then associate as a single-layer forming the shell facets (Faulkner et al. 2017 ). In terms of quantification, CcmK1–4 fall into two abundance groups with K1 and K2 at 79,000–169,000 and 61,000–94,500 cpc respectively and, at 10–20-fold lower abundance, K3 and K4 at 2900–8100 and 9400–19,500 cpc (Fig. 2 a). These groupings align with earlier observations of K1/2 as major and K3/4 as minor shell proteins (Yeates et al. 2011 ). CcmL proteins, quantified here at 3550–8850 cpc, are homopentameric and form the carboxysomal vertices (Tanaka et al. 2008 ). Using the dimensions presented by Faulkner et al. ( 2017 ), based on TEM and cryo-EM imaging and assuming regular icosahedral geometry with 20 facets and 12 vertices, we estimate that our quantification of CcmK1 and CcmL would give 11–23 and 59–148 carboxysomes per cell respectively. The EM-based carboxysome count is six on average (Reinhold et al. 1991 ) therefore the excess shell protein levels are likely due to a substantial population of incomplete carboxysomes, as described in models for both biogenesis (Cameron et al. 2013 ) and degradation (Hill et al. 2020 ). Fig. 2 Cellular levels of carboxysomal proteins and enzymes of the Calvin–Benson–Bassham cycle. Consensus cpc ranges are derived from data-points shown in Supplementary Fig. S3 and displayed as in Fig. 1 . Structural carboxysomal proteins ( a ), Calvin–Benson–Bassham cycle enzymes: carbonic anhydrase (CcaA), ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCO: large, RbcL and small, RbcS subunits), glycerate-3-phosphate kinase (Pgk), glycerate-1,3-phosphate dehydrogenase (Gap1, 2), transketolase (TktA), phosphoketolase (Xfp) ( b ), triose phosphate isomerase (TpiA), fructose-1,6-bisphosphate aldolase (Fba1, 2), fructose-1,6- and sedoheptulose-1,7-bisphosphatase (Fbp1, 2), ribose-5-phosphate isomerase (RpiA), ribulose-3-phosphate epimerase (Rpe), ribulose-5-phosphate kinase (Prk) ( c ). The horizontal dashed lines in ( b ) and ( c ) define a 25,000–60,000 cpc range, explained in Results and Discussion Like CcmK1–4, the shell protein CcmP oligomerizes around a central pore. However, CcmP is more complex than the single-layer CcmK1–4 since it integrates into the shell as a stacked dimer of trimers. This arrangement not only encloses an internal compartment but also provides a mechanism for gating the central pore (Cai et al. 2013 ). Assuming six carboxysomes per cell, alongside our quantification at 930–1900 cpc (Fig. 2 a), there would be 25–50 of these CcmP gated pores per carboxysome. Deletion mutants of the seventh shell protein CcmO have confirmed its absolute requirement for carboxysome assembly and that it associates with CcmK1/2 during the encapsulation phase (Rae et al. 2012 ; Cameron et al. 2013 ). The hypothesis that CcmO occurs at the facet edges (Rae et al. 2012 ) might imply that this protein is at least moderately abundant. Unexpectedly, CcmO was not detected in either this or previous (Faulkner et al. 2017 ) proteomic analyses. In addition to the outer shell proteins, carboxysomes contain CcmM and CcmN, which are involved in structural organization. CcmM occurs as two isoforms (Cot et al. 2008 ): (1) the full-length translation product CcmM73 is located within an inner shell where it functions as a scaffold protein interacting with CcmK1-4, CcmL, CcaA and CcmN (Long et al. 2007 ; Kinney et al. 2012 ) and (2) the truncated product of an alternative downstream initiation codon CcmM52 which crosslinks RuBisCO in paracrystalline arrays for assembly into the carboxysome interior (Cameron et al. 2013 ). Our quantification of CcmM at 21,000–44,500 cpc might be expected, given its dual functionality and multiple interaction partners, including possibly 50,000 RuBisCO hexadecamers (see below). However, without the ability to differentiate CcmM73 and CcmM52 by our methodology, further interpretation is not applicable. CcmN, like CcmO (see above) was not detected either here or by Faulkner et al. ( 2017 ). Calvin-Benson-Bassham cycle The enzymes belonging to the CBB cycle were selected for quantification in this study by referring to https://www.genome.jp/kegg-bin/show_pathway?syn00710 and are shown in Figs. 2 b, c. The first step is the conversion of HCO 3 − to CO 2 by carbonic anhydrase (CcaA), quantified here at 2400–9200 cpc (Fig. 2 b). Its catalytic unit is a homodimer and, with a k cat = 3340 s −1 , CcaA potentially generates 4–15 × 10 6 CO 2 molecules cell −1 s −1 (McGurn et al. 2016 ). This reaction occurs inside carboxysomes (see above) where the second step, in which RuBP is carboxylated by RuBisCO, generates two molecules of glycerate-3-P. RuBisCO is often cited as the most abundant enzyme on Earth and, accordingly the result of our quantification of its large (RbcL) and small (RbcS) subunits is 156,000–204,500 and 136,000–406,500 cpc, respectively (Fig. 2 b). In the majority of photoautotrophs, including cyanobacteria, RuBisCO is hexadecameric with 8 copies each of the RbcL and RbcS subunits (Andersson et al. 2008). Therefore the number of active enzyme complexes might approach 50,000 per cell. This very high abundance would mitigate for its slow carboxylation rate, estimated at ~ 0.7 s −1 per active site in S. elongatus (Flamholz et al. 2019 ). The cpc levels presented here would accommodate the known 1:1 stoichiometry, although the upper extent of the RbcS range at > 400,000 cpc also suggests a two-fold excess over RbcL in vivo. Similarly, a relative LFQ proteomic analysis of Synechocystis by Kwon et al. ( 2013 ) reported a 1.4-fold excess of RbcS. Following its formation by RuBisCO, glycerate-3-P exits the carboxysome and the CBB cycle continues in the cytoplasm, ultimately regenerating RuBP. Intermediates downstream of glycerate-3-P in the CBB cycle additionally feed into numerous biosynthetic and metabolic processes (reviewed by Mills et al. 2020 ). While kinetic parameters for cyanobacterial CBB cycle enzymes are apparently not well represented in the literature (Janasch et al. 2019 ), some patterns do emerge that align with the quantification results shown in Figs. 2 a, b. The ATP- and NADPH-utilizing steps, catalysed by glycerate-3-P kinase (Pgk), glycerate-1,3-phosphate dehydrogenase (Gap2) and ribulose-5-phosphate (Ru5P) kinase (Prk) all occupy an intermediate abundance range of 25,000–60,000 cpc, indicated by horizontal dashed lines in Figs. 2 b, c. The alternative dehydrogenase isoform Gap1 principally catalyses the reverse reaction (Koksharova et al. 1998 ). Its lower abundance, at 4650–6900 cpc may be expected given that Gap1 activity would be counterproductive to flux through the CBB cycle. Metabolomics-based kinetic modelling in Synechocystis has revealed that the steps catalysed by Pgk, Prk and another enzyme within the 25,000–60,000 cpc range, fructose-1,6/sedoheptulose-1,7-bisphosphatase (FBP/SBPase, Fbp2) are strong effector reactions in the control of flux through the CBB cycle (Janasch et al. 2019 ). This role of Fbp2 in flux control was corroborated in Synechocystis by Hing et al. ( 2019 ) who also identified transketolase (TktA), another enzyme in the same intermediate abundance range, as a key player. Therefore, Pgk, Prk, Fbp2 and TktA may all be targets of the same expression control mechanism in response to ambient growth conditions. In their analysis, Janasch et al. ( 2019 ) additionally identified RuBP, FBP and SBP as intermediates that would perturb the overall steady-state stability of the CBB cycle if their cytosolic concentrations were to increase beyond a critical level. Limiting the production of RuBP, FBP and SBP might be mediated by restricting the abundance levels of upstream enzymes. In support of this hypothesis, ribose-5-phosphate isomerase A (RpiA) and ribulose-3-phosphate epimerase (Rpe), which convert ribose-5-phosphate (R5P) and xylulose-5-phosphate (Xu5P) respectively into Ru5P, the RuBP precursor, are both quantified in a relatively low abundance range (4850–12,500 and 6550–9800 cpc respectively; Fig. 2 c). Similarly, production of SBP precursors glycerone-phosphate (also named dihydroxyacetone phosphate) and erythrose-4-phosphate (E4P) may be limited by the relatively low abundance of triose phosphate isomerase (TpiA) and phosphoketolase (Xfp) at 9100–19,000 and 1200–5400 cpc, respectively. Like Gap1/2, other CBB enzymes occur as two isozymes. The reaction catalysed by RpiA, described above, is also potentially mediated by RpiB (see KEGG link above). This apparently uncharacterized protein is not detected in this analysis, prompting the conclusion that RpiA is probably the only R5P isomerase isozyme produced by Synechocystis under the growth conditions used here. In the case of fructose-1,6-bisphosphate (FBP) aldolase which converts G3P + glycerone-P to FBP and E4P + glycerone-P to SBP, both Fba1 and Fba2 isozymes are quantified at widely differing ranges: 3600–17,000 and 75,500–111,500 cpc, respectively (Fig. 2 c), again highlighting the dominance of one isozyme over the other. This abundance pattern is repeated with quantification of Fbp1 and Fbp2 at 980–1900 and 42,000–53,500 cpc respectively. While these isozyme abundance level differences cannot be rationalized without experiments to examine the effects of different growth conditions, there is evidence that deployment of alternative CBB cycle isozymes in Synechocystis provides a mechanism for acclimation to environmental CO 2 levels that extends beyond transcriptional control (Jablonsky et al. 2016 ). Ferredoxin-dependent processes There are nine ferredoxins (Fds) in Synechocystis : Ssl0020 is the isoform that mediates electron transfer from the Fe-S centers on the PsaC subunit of PSI (Yu et al. 1995 ). Like PC, Fd is highly abundant: although its quantification is below the validation threshold with two quantotypic tryptic peptides, it may occur above 200,000 cpc (Fig. 3 a), exceeding the abundance of PSI by factor of 2–3 (cf. Figure 1 a). The 10 5 order of magnitude for the copy number of Fd in Synechocystis cells has been corroborated by quantitiative immunoblotting (Moal and Lagoutte 2012 ). This high cellular level reflects the numerous Fd-dependent metabolic processes that exist in cyanobacteria. Of particular relevance to photosynthesis, the ferredoxin NADP + reductase (FNR), which catalyses the production of NADPH for CO 2 fixation (Cassier-Chauvat and Chauvat 2014 ), occurs in both full-length, and truncated (FNR s , with alternative initiation at M113) isoforms (Thomas et al. 2006 ). We quantified FNR with peptides identified from both N- and C-sides of M113 (Supplementary Data Sets S5 and S6). Although the two isoforms would, therefore, be indistinguishable in this analysis, we assume that FNR s would be absent under the photoautotrophic growth conditions used here since it is only detected during heterotrophic metabolism (Thomas et al. 2006 ). With FNR at 20,500–30,500 cpc (Fig. 3 a), the PSI:FNR ratio would be approximately 4:1, which again aligns closely with a quantitative immunoblot assay by Moal and Lagoutte ( 2012 ). Fig. 3 Cellular levels of enzymes and auxiliary proteins involved in biosynthesis. Consensus cpc ranges are derived from data-points shown in Supplementary Fig. S4 and displayed as in Fig. 1 . Ferredoxin (Fd) and Fd-dependent enzymes: ferredoxin-NADP + reductase (FNR), glutamate synthase 2 (GlsF), flavodiiron protein (Flv1/3), bidirectional hydrogenase (HoxF, U, H), sulphite reductase (Sir), nitrite reductase (NirA) ( a ), enzymes and auxiliary proteins in the Mg-branch of the chlorophyll a biosynthesis pathway: Mg-chelatase (ChlI, ChlD, ChlH, Gun4), Mg-protoporphyrin IX methyltransferase (ChlM), O 2 -dependent Mg-protoporphyrin IX methyl ester cyclase (CycI, Ycf54), light-dependent protochlorophyllide oxidoreductase (LPOR), 8-vinyl reductase (DVR), geranylgeranyl reductase (ChlP), chlorophyll a synthase (ChlG) ( b ), enzymes in the carotenoid biosynthesis pathway: geranylgeranyl pyrophosphate synthase (CrtE), phytoene desaturase (CrtP), prolycopene isomerase/CRTISO (CrtH), ζ-carotene desaturase (CrtQ) ( c ), enzymes in the phycobilin biosynthesis pathway: ferrochelatase (FeCh), heme oxygenase 1 (PbsA1), chromophore lyase (CpcS1, CpcT) ( d ). The horizontal dashed lines, explained in Results and Discussion, define in: ( a) 740–1400 and 3150–6100 cpc, ( b) 1150–3550 cpc, ( c) 380–1100 cpc and ( d) 2300–3800 cpc In addition to these photosynthesis-related processes, Fd is required for at least nine other metabolic processes in cyanobacteria (Lea-Smith et al. 2016 ). Figure 3 a shows quantification results for eight of these including glutamate synthase 2 (GlsF, encoded by gltS ) at 1900–2850 cpc. The remaining seven fall into two abundance groups shown by dashed lines in Fig. 3 a. Three of the five subunits of the bidirectional hydrogenase, HoxF, HoxU and HoxH are quantified here at 740–1550, 1050–1400 and 830–1350 cpc (Fig. 3 a, lower abundance group), consistent with the known subunit stoichiometry (Vignais and Billoud 2007 ). The flavodiiron protein (FDP), comprising Flv1 and Flv3 subunits, functions as a heterodimer (Allahverdiyeva et al. 2011 ). In this case, our analysis is at odds with a 1:1 stoichiometry by revealing levels of 1100–1400 cpc, in the lower abundance group for Flv1 and 4050–6100 cpc in the higher abundance group for Flv3. The observation that Flv3 is substantially more abundant than Flv1 does however align with more recent evidence of FDP activity by Flv3 oligomers in addition to Flv1/3 heterodimers (Mustila et al. 2016 ). Other Fd-dependent enzymes in the higher abundance group are sulfite reductase (Sir) at 3150–4300 cpc and nitrite reductase (NirA) at 3600–6300 cpc (Fig. 3 a),. The enzyme that acts upstream of NirA, nitrate reductase (NarB), is not detected in this analysis despite being essential for growth in a culture medium containing nitrate as the sole nitrogen source (Baebprasert et al. 2011 ), as used in this study. Two subunits of the nitrate transporter complex NrtA (Sll1450) and NrtC (Sll1452) are identified (Supplementary Data Sets S1 and S2), implying the cells’ probable competency in importing nitrate from the medium. In addition, not quantified here is ferredoxin–thioredoxin reductase; the catalytic subunit FtrC is identified with two tryptic peptides, below the validation threshold for LFQ, but the second subunit, FtrV, is not detected. The biosynthetic pathway for chlorophyll a Magnesium-chelatase The biosynthesis of Chl a is carried out by a series of enzymes and auxiliary proteins (Fig. 3 b), starting with the magnesium-chelatase (MgCh) enzyme complex. MgCh catalyses the ATP-dependent insertion of Mg 2+ into protoporphyrin IX, which is also the substrate for ferrochelatase (FeCh) that produces heme. Thus, the insertion of Mg 2+ or Fe 2+ represents a branchpoint in tetrapyrrole biosynthesis (reviewed by Bryant et al. 2020 ). MgCh is associated with the cytoplasmic/stromal surface of TMs (Kopečná et al. 2015 ; Farmer et al. 2019 ), as depicted in Fig. 3 b, and comprises three core subunits: (i) the AAA + ATPase ChlI (Fodje et al. 2001 ) which provides the free energy for Mg 2+ chelation (Reid and Hunter 2004 ), (ii) ChlD, an allosteric regulator (Adams and Reid 2013 ) that also transmits the energy released by ATP hydrolysis by ChlI (Adams et al. 2016 ; Farmer et al. 2019 ) to (iii) ChlH where the active site resides (Karger et al. 2001 ; Sirijovski et al. 2008 ; Adams et al. 2020 ). Quantification reveals that data-points for ChlI cover a 2400–7300 cpc range while ChlD and ChlH levels are confined to 2200–3550 and 2250–2950 cpc, respectively (Fig. 3 b). On the assumption that the abundance range shown by ChlD/H represents a 1:1 molar ratio for these subunits in the active MgCh complex (Farmer et al. 2019 ), then the level of ChlI in relation to ChlD/H is either the same or only higher by a factor of 2–3. Given that the active MgCh probably comprises multiple copies of ChlI, based on structural evidence that ChlI associates into hexamers (Gao et al. 2020 ) or heptamers (Reid et al. 2003 ), it appears that the number of fully-assembled, active MgCh complexes may be limited by the availability of ChlI subunits to 500–1200 per cell, sufficient to produce 7–16 molecules of Mg-protoporphyrin IX (MgP IX ) s −1 cell −1 , based on a k cat of 0.013 s −1 (Reid and Hunter 2004 ; Viney et al. 2007 ). Association of the auxiliary MgCh subunit Gun4 is not an absolute requirement for activity in vitro but does enhance the rate of MgP IX formation by a factor of 3–10 (Larkin et al. 2003 ; Davison et al. 2005 ; Davison and Hunter 2011 ; Adams et al. 2016 ). We quantified Gun4 at 570–2050 cpc (Fig. 3 b), which aligns with the 500–1200 cpc range suggested for the number of active MgCh complexes per cell. If MgCh binds Gun4 in a 1:1 ratio, the rate of MgP IX generation may therefore approach 160 s −1 cell −1 . Mg-protoporphyrin IX methyltransferase Following the insertion of Mg 2+ , MgP IX is converted to Mg-protoporphyrin IX monomethyl ester (MgPME) by Mg-protoporphyrin IX methyltransferase (ChlM). The production of MgPME occurs with k cat = 57 s −1 (Shepherd and Hunter 2004 ) which, alongside our quantification of ChlM at 1750–2950 cpc (Fig. 3 b), equates to 99,750–168,150 s −1 cell −1 thereby exceeding the rate of MgP IX production by a factor of > 1000. This marked difference may represent a mechanism for preventing the accumulation of MgP IX and therefore its potentially cytotoxic effects (Tanaka and Tanaka 2007 ). O 2 -dependent Mg-protoporphyrin IX methyl ester cyclase In the third step of Chl biosynthesis, the C13 methylpropionyl sidechain of MgPME is cyclized by O 2 -dependent Mg-protoporphyrin IX methyl ester cyclase to form the fifth isocyclic (E) ring of 3,8-divinyl protochlorophyllide a (DV-PChlide). The presence of the E ring induces a change in the absorption profile, which transforms the red color of the substrate to a green product (Chen et al. 2021 ). The cyclase has two isoforms in Synechocystis , CycI and CycII, which share 56.7% sequence identity and are encoded by sll1214 and sll1874, respectively. Consistent with the normal culture aeration used here, we identified CycI at 1150–1700 cpc (Fig. 3 b) as the only cyclase isoform present. CycII, which is synthesized in addition to the constitutive CycI under low O 2 (Minamizak et al. 2008 ; Peter et al. 2009 ), was below the limit of detection in all analyses. The k cat of 0.015 s −1 measured by Chen et al. ( 2021 ) for CycI is comparable to that of MgCh at 0.013 s −1 (determined in the absence of Gun4), giving a potential rate of MgPME to DV-PChlide conversion of 17–26 s −1 cell −1 . Like MgCh with Gun4, CycI also associates with an auxiliary protein, Ycf54, in Synechocystis (Hollingshead et al. 2012 ) and plants (Bollivar et al. 2014 ; Herbst et al 2018 ). Although the structural elements in Ycf54 that mediate its association with CycI have been characterized (Hollingshead et al. 2017 ), its precise role in MgPME cyclase activity is not yet defined. While our analysis shows that the Gun4 copy number appears to be in an approximate 1:1 ratio with that of assembled MgCh complexes (see above), Ycf54, at 5350–6200 cpc (Fig. 4 b) is 3–fivefold more abundant than CycI. This higher stoichiometry may reflect evidence that, although CycI is membrane-associated (Tottey et al. 2003 ; Rzeznicka et al. 2005 ; Allen et al. 2008 ; Hollingshead et al. 2012 ), Ycf54 is located in both soluble and membrane fractions (Hollingshead et al. 2012 ). Protochlorophyllide oxidoreductase (POR) Following DV-PChlide formation, the next two steps in the Chl biosynthesis pathway result in the reduction of the C17 = C18 and C8-vinyl double bonds. In cyanobacteria, POR exists as two structurally unrelated versions: light-dependent (LPOR) and light-independent or dark-operative (DPOR) (Reinbothe et al. 2010 ). The latter is composed of three subunits: ChlN, ChlB, and ChlL, and none of these was detectable in our analyses (Supplementary Data Sets S1 and S2). This finding is expected for the growth conditions used here since DPOR activity is inhibited by O 2 at > 3% and the expression of its subunits is induced only under anaerobic conditions (Yamazaki et al. 2006 ). LPOR is a single subunit enzyme that uses the energy from a photon absorbed by its substrate, DV-PChlide, to acquire H (with two electrons) from NADPH and a proton from a conserved Tyr residue (Heyes et al. 2006 ). The analysis presented in Fig. 3 b reveals an abundance range of 3450–4700 cpc. A k cat = 0.027 s −1 (Zhang et al. 2021 ) gives potential production rates for DV-Chlide of 93–127 s −1 cell −1 , thereby exceeding the upstream cyclase step by a factor of 4–7. As suggested for the MgP IX to MgPME conversion, the LPOR substrate DV-PChlide is potentially hydrogenated in the light to form DV-Chlide faster than it can accumulate. 8-Vinyl reductase In the next reduction step, catalysed by DV-(P)Chlide 8-vinyl reductase (DVR), the C8 vinyl group of DV-Chlide is converted to ethyl, producing (monovinyl) chlorophyllide (Chlide). DVR, encoded by slr1923 in Synechocystis (Islam et al. 2008 ; Ito et al. 2008 ), is detectable in both membrane and soluble fractions (Canniffe et al. 2014 ). Despite its wide distribution, DVR was quantified at the lowest abundance of all the Chl biosynthesis-associated proteins at < 500 cpc (Fig. 3 b). To our knowledge, there are as yet no published steady state activity measurements for cyanobacterial DVR, therefore, the potential cellular rate of DV-Chlide to Chlide conversion cannot currently be estimated. Geranylgeranyl reductase and chlorophyll a synthase The final two steps in Chl biosynthesis can occur in either order (Soll et al. 1983 ; Proctor et al. 2022a ). Geranylgeranyl reductase (ChlP) catalyses the hydrogenation of three C = C double bonds on the C 20 isoprenoid geranylgeranyl pyrophosphate (GGPP) to produce phytyl pyrophosphate. The phytyl group is then attached via an ester linkage to the C17 propionate sidechain on Chlide by Chl a synthase (ChlG). Alternatively, ChlG can first attach a geranylgeranyl group to Chlide for subsequent reduction to phytyl by ChlP. ChlP is probably active on the cytosplasmic surface of the TM, whereas ChlG is predicted to be membrane-intrinsic (Gaubier et al. 1995 ) with up to nine TMHs (Proctor et al. 2022a ). ChlP was quantified at 1150–1850 cpc, the sixth Chl biosynthesis pathway component to fall into the shared 1150–3550 cpc range indicated by horizontal dashed lines in Fig. 3 b. ChlG was revealed as a member of the lower abundance group of Chl synthesis pathway components, falling in between CycI and DVR, at 450–1000 cpc (Fig. 3 b). Since, to our knowledge, there are currently no published k cat measurements for either ChlP or ChlG, their potential cellular catalysis rates remain undetermined. The carotenoid biosynthesis pathway Carotenoids are membrane-intrinsic isoprenoids synthesized by all oxygenic photoautotrophic organisms, where they play roles in the function, assembly and stability of complexes including PSII (Umena et al. 2011 ), PSI (Jordan et al. 2001 ), cyt b 6 f (Malone et al. 2019 ) and NDH-1 (Schuller et al. 2019 ). Recently, structures of some of these complexes from Synechocystis have been determined, revealing the positions of the carotenoids (Malavath et al. 2018 ; Gisriel et al. 2022 ; Proctor et al. 2022b ). Carotenoids are also essential for photoprotection, quenching ROS-generating Chl triplet states and superoxide (Cogdell et al. 2000 ), and in cyanobacteria the photoactive orange carotenoid protein (OCP) is involved in thermal dissipation of excess energy from the phycobilisome antenna (Muzzopappa and Kirilovsky 2020 ). Carotenoid biosynthesis (summarized in Fig. 3 c and reviewed by Canniffe and Hitchcock ( 2021 )) commences with the condensation of the products of the 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAP) producing geranyl pyrophosphate (GPP). In oxygenic phototrophs this reaction, together with two subsequent additions of IPP to GPP to produce GGPP, are catalysed by the GGPP synthase (CrtE). Subsequently, two molecules of GGPP are condensed to generate 15- cis -phytoene by phytoene synthase (CrtB), which then undergoes a series of desaturations and isomerizations resulting in the production of all- trans -lycopene, the common precursor of the major carotenoid species utilized by Synechocystis , namely β-carotene, myxoxanthophyll, echinenone, zeaxanthin and synechoxanthin (Lagarde and Vermaas 1999 ; Graham et al. 2008 ). In our analysis, CrtE, which is also required for Chl biosynthesis, is quantified, along with phytoene desaturase (CrtP), ζ-carotene desaturase (CrtQ) and the prolycopene isomerase CRT-ISO (CrtH), all of which occur early in the pathway in the synthesis of all- trans -lycopene. Figure 3 c shows that CrtE, CrtP and CrtH occur at abundance levels in the 380–1100 cpc range (horizontal dashed lines), while CrtQ is higher at 1300–1900 cpc. In addition to CrtE, CrtP, CrtQ and CrtH, the other enzymes required to generate all- trans -lycopene are CrtB and ζ-carotene isomerase (Z-ISO). A Synechocystis crtB mutant cannot synthesize carotenoids, is light sensitive and lacks functional PSII (Sozer et al. 2010 ), suggesting the enzyme is likely to be present just below the detection limit. Given that the cells used here were grown under constant illumination, the known photo-lability of the 15- cis bond of 9,15,9′-tri- cis -ζ-carotene (Li et al. 2007 ) may explain non-detection of the Synechocystis Z-ISO (Slr1599; Proctor et al. 2022c ), although in plants Z-ISO is important in both “dark” and light-exposed tissues (Chen et al. 2010 ). Our detection of CrtH, the other carotenoid isomerase, suggests photoisomerization alone is insufficient in the case of the second carotenoid isomerization step; a S ynechocystis crtH mutant produces normal carotenoids under light conditions due to photoisomerization of the cis -bonds in prolycopene, albeit at different ratios to the wild-type organism (Masamoto et al. 2001 ). The remaining eight enzymes are either not synthesized under the culture conditions used here or occur at < 200–500 cpc. The major lycopene cyclase CruA (Xiong et al. 2017 ) is identified but at levels below the threshold for quantification (Supplementary Data Sets S1 and S2), whereas CruP, the lycopene cyclase function of which is controversial (Maresca et al. 2007 ; Liang et al. 2008 ), is not detected. Similarly, CruF and CruG, specific to myxoxanthophyll biosynthesis (Graham and Bryant 2009 ), CruE and CruH, required for synthesis of synechoxanthin (Graham and Bryant 2008 ), and CrtO, the ketolase for echinenone and 3-hydroxy-echinenone biosynthesis (Fernández-González et al. 1997 ) are not detected, nor is CrtR, which adds the hydroxyl groups to the β -rings of zeaxanthin, myxoxanthophyll and 3-hydroxy-echinenone (Lagarde and Vermaas 1999 ). Levels below the threshold of identification for carotenoid biosynthesis enzymes were also apparent in two previous proteomic studies that employed sub-cellular fractionation to potentially enhance proteomic coverage (Xu et al. 2021 ; Baers et al. 2019 ). This low copy number may be a consequence of the low turnover of carotenoids under moderate illumination conditions and/or extremely efficient enzymes meaning high cellular levels are not required. The phycobilin biosynthesis pathway Bilins, linear tetrapyrroles derived from heme, are light-harvesting chromophores covalently attached to phycobiliproteins, which assemble to form the phycobilisome antenna complex (Dominguez-Martin et al. 2022 ). The biosynthesis pathway of bilins, reviewed by Bryant et al. ( 2020 ), is common with that of Chls up to protoporphyrin IX, where it diverges from the branch initiated by the insertion of Fe 2+ catalysed by ferrochelatase (FeCh, HemH) to produce heme. Heme oxygenase (Hox) then cleaves the heme macrocycle to produce the linear molecule biliverdin IXα, which is subsequently converted to a bilin via a Fd-dependent bilin reductase; Synechocystis produces only phycocyanobilin (PCB). We quantify FeCh (Slr0839; 1300–1600 cpc) and HoxI (PbsA1, 1800–3800 cpc), while the PCB-ferredoxin oxidoreductase (PcyA) is detected but present at < 500 cpc. As expected, Hox2 (PbsA2), which is produced under microoxic conditions (Yilmaz et al. 2010 ), is not identified. The attachment of phycobilins to specific cysteine residues of phycobiliproteins requires specific bilin lyases (Scheer and Zhao 2008 ). In cyanobacteria such as Synechocystis and Synechococcus sp. PCC 7002, three PCBs are attached to a phycocyanin heterodimer, one to CpcA by a heterodimeric lysase comprised of CpcE and CpcF (Fairchild et al. 1992 ), one to CpcB by a CpcS/CpcU family lyase (Saunée et al. 2008 ) and one to CpcB by a CpcT family lyase (Shen et al. 2006 ). CpcS/CpcU also attach PCB to the core allophycocyanin antenna subunits ApcA, ApcB, ApcD and ApcF (Zhao et al. 2007 ). We quantify CpsS1 (CpcU, 2300–3800 cpc) and CpcT (2700–3700 cpc) but not CpcE, CpeF or CpcS2 (CpcS), suggesting these are present at < 500 cpc. An alternative fate of biliverdin Ixα is reduction to bilirubin by the pyridine nucleotide-dependent biliverdin reductase (BvdR); the Synechocystis enzyme is quantifed by only 1–2 peptides at < 500 cpc (Supplementary Data Set S4). Although not a direct component of the phycobiliosme, bilirubin is suggested to act as a ROS scavenger (Hayes and Mantle 2009 ) and BvdR is important for normal phycobiliprotein biosynthesis in Synechocystis (Schluchter and Glazer 1997 ). Photosystem assembly and repair Coordination of chlorophyll and photosystem II biosynthesis ChlG and HliD As stated above, the step in the Chl biosynthesis pathway in which either a geranylgeranyl or phytyl chain is ester-linked to the Chlide macrocycle is catalysed by ChlG. Immunoprecipitation (IP) experiments employing FLAG-tagged ChlG have shown that this membrane-intrinsic enzyme co-isolates with the single-TMH proteins HliC (Niedzwiedzki et al. 2016 ) and HliD (Chidgey et al. 2014 ), two of the four high light-inducible proteins (Hlips) present in Synechocystis (HliA-D; Komenda and Sobotka 2012 ). The ChlG-HliC/D complex also incorporates Chl and carotenoids (Chidgey et al. 2014 ; Niedzwiedzki et al. 2016 ), implicating HliC and HliD in photoprotection specifically during PSII, but not PSI (see below), assembly in which the delivery of Chl is coordinated with the co-translational insertion of nascent apoproteins into the membrane (Chidgey et al. 2014 ; Knoppová et al. 2014 ). According to our analyses, the abundance of ChlG is 450–1000 cpc (Fig. 3 b) with HliD 3–sevenfold higher at 2000–3200 cpc (Fig. 4 a). HliC was also identified in this study (Supplementary Data Sets S1 and S2) however, recovery of its single proteotypic tryptic peptide from artificial SIL standard proteins proved non-reproducible during initial tests (results not shown) and LFQ would be below the validation threshold with < 3 peptides. The greater abundance of HliD over ChlG is supported by the observation that carotenoids only bind to HliC/D dimers, not monomers (Shukla et al. 2018 ), implying that the functional units of these Hlips are dimers. The detection of ChlG- and HliD-containing complexes at > 100 kDa by native-PAGE/immunoblot analysis (Proctor et al. 2020 ) further suggests that PSII assembly centers may comprise multiple copies of ChlG and HliC/D dimers. Fig. 4 Cellular levels of assembly factors and enzymes involved in thylakoid membrane biogenesis and photosystem assembly/repair. Consensus cpc ranges are derived from data-points shown in Supplementary Fig. S5 and displayed as in Fig. 1 . Proteins involved in the coordination of chlorophyll and PSII biosynthesis ( a ), proteins involved in thylakoid membrane biogenesis and PSII assembly ( b ), proteins involved in thylakoid membrane biogenesis and PSI assembly ( c ), and ATP-dependent zinc metalloproteases: membrane protein quality control and PSII repair ( d ) Ycf39 An additional protein co-isolating with Flag-ChlG in IP analysis is Ycf39 (Slr0399; Chidgey et al. 2014 ), although its interaction with the complex is lost under high-light conditions (Proctor et al. 2018 ). Ycf39 is predicted to be hydrophilic and its interaction with the membrane-intrinsic ChlG-HliC/D complex is on the TM cytoplasmic surface (Knoppová et al., 2014 ). IP analysis using FLAG-tagged Ycf39 showed that its direct binding partner is dimeric HliC/D (Staleva et al., 2015 ) and native-PAGE/immunoblot analysis confirmed the association of Ycf39 with early intermediates in the PSII assembly pathway (Knoppová et al., 2014 , 2022 ; Heinz et al. 2016 ; Konert et al. 2022 ). Ycf39, is quantified here at 1200–1800 cpc (Fig. 4 a), therefore HliD alone outnumbers Ycf39 by a factor of two, highlighting the possibility that all copies of Ycf39 are bound to Hlip dimers. YidC and SecY The identification of ChlG-HliC/D-Ycf39 complexes highlights the PSII assembly process in terms of Chl delivery and photoprotection. The additional detection of the membrane insertase YidC in FLAG-ChlG IP analyses (Chidgey et al. 2014 ; Niedzwiedzki et al. 2016 ) establishes the direct link with co-translational integration of PSII apoproteins into the TM. While the ChlG-HliC/D-Ycf39 complex is evidently specific to PSII assembly (Knoppová et al. 2014 , 2022 ; Pascual-Aznar et al. 2021 ), YidC participates in the co-translational insertion of a wide range of membrane-intrinsic proteins (Kudva et al. 2013 ). YidC was quantified here at 1000–3350 cpc (Fig. 4 a), coincident with the 2500–3000 cpc determined in E. coli (Urbanus et al. 2002 ; Kudva et al. 2013 ). SecY, the only subunit of the SecYEG translocon identified here, is quantified at 110–830 cpc. Again, this range is in close agreement with the 200–600 cpc in E. coli (Kudva et al. 2013 ). The perhaps unexpectedly low abundance of these proteins in Synechocystis may be rationalized on the basis that a chaperone-type function is similar to catalysis in that, after membrane insertion of the substrate protein, YidC and SecYEG are released and available to bind a new substrate. Furthermore, a recent fluorescent imaging study of mRNA sequences mapping to PsaA and PsbA has revealed that translation sites for these membrane-integral PS subunits are not widely distributed but instead confined to the interior cytosol-facing surface of the TMs (Mahbub et al. 2020 ). TM-intrinsic protein insertion appears therefore to be localized and the levels of YidC and SecYEG may reflect this. Thylakoid membrane biogenesis and photosystem II assembly CurT The characteristic morphology of thylakoids in cyanobacteria is dependent on CurT, an integral membrane protein that induces membrane curvature (Heinz et al. 2016 ). Accordingly, inactivation of curT results in the development of aberrant TM structure and additionally a 50% reduction in PSII accumulation compared to wild-type. This effect on PSII levels is accompanied by the elimination of the biogenesis centers, more recently referred to as ‘convergence zones’ (Rast et al. 2019 ), implicating CurT in the formation of these features as part of normal TM morphology (Heinz et al. 2016 ). Quantification of CurT in this study indicates the possibility of cellular levels approaching 140,000 cpc (Fig. 4 a). Imaging by both immunofluorescence and immunogold labelling supports this finding since CurT is detectable, not only in association with convergence zones but also throughout the TM on both concave and convex surfaces (Heinz et al. 2016 ). PratA, Pitt and Slr0151 Specifically localized to the convergence zones where they function as PSII assembly factors are three tetratricopeptide repeat (TPR) proteins: PratA (Slr2048; Klinkert et al. 2004 ), Pitt (Slr1644; Schottkowski et al. 2009 ) and Slr0151 (Rast et al. 2016 ). Using high-resolution cryo-electron tomography, it has been demonstrated that, within the convergence zones, membranes continuous with the TM are in close contact with the CM (Rast et al. 2019 ). This juxtaposition enables a mechanism whereby PratA delivers Mn 2+ from the periplasm to the membrane-integrated PsbA precursor (pD1; Stengel et al. 2012 ). The level of PratA was quantified here at 320-630 cpc (Fig. 4 b). Pitt (encoded by slr1644) is anchored in the TM via an N-terminal TMH and inactivation of slr1644 was shown to reduce the accumulation of LPOR (see above) by 70% (Schottkowski et al. 2009 ). These authors suggested that the association of LPOR with the TM might be via its binding to Pitt. With representation by only two tryptic peptides, Pitt was not validated for quantification. However, iBAQ abundance scores are consistent with a level of Pitt at < 500 cpc (Supplementary Data Set S4). Our quantification of LPOR at 3450–4700 cpc (see above), up to tenfold more abundant than Pitt, would align with the further idea of an LPOR-Pitt membrane-attachment complex that directs a relatively small LPOR sub-population to the convergence zone for an, as yet unknown function. Slr0151 has been characterized as functioning in PSII biogenesis (Rast et al. 2016 ) and repair (Yang et al. 2014 ) and slr0151 inactivation results in impaired TM morphology (Rast et al. 2016 ). Unlike the other two TPR proteins PratA and Pitt, and probably consistent with its wider distribution both within convergence zones and throughout the TMs (Rast et al., 2016 ), Slr0151 is quantified in our analysis at 4450–7900 cpc (Fig. 4 b). Furthermore, there is evidence that Slr0151 has a greater range of interaction partners than PratA and Pitt including PSII subunits PsbA/D1 and PsbC/CP43 (Yang et al. 2014 ), and is involved in the modulation of Fd-5 phosphorylation (Angeleri et al. 2018 ). Pam68, Ycf48 and RubA The Pam68 assembly factor, first characterized in Arabidopsis as TM-intrinsic, was identified by homology with Sll0933 in Synechocystis (Armbruster et al. 2010 ). Using FLAG-tagged Pam68, Bučinská et al. ( 2018 ) co-isolated a complex containing PsbB/CP47 as the only PSII subunit represented, alongside YidC, SecY and several riboproteins. These associations reveal that the probable function of Pam68 is facilitating the co-translational insertion of PsbB/CP47 into the TM and possibly also the correct apoprotein configuration for the delivery of Chl (Bučinská et al. 2018 ). Co-isolating with Pam68 in IP analysis is the lumenal protein Ycf48 (Rengstl et al. 2013 ; Bučinská et al. 2018 ). Ycf48 is homologous with Arabidopsis HCF136 (Meurer et al. 1998 ) and was shown to be essential for the accumulation of PsbB/CP47 and PsbC/CP43 in the TM (Rengstl et al. 2013 ). It has been proposed that Ycf48 may, like Pam68, facilitate co-translational Chl delivery to nascent apoproteins (Crawford et al. 2016 ). We quantified Pam68 and Ycf48 at 1250–1600 and 3600–5150 cpc, respectively (Fig. 4 b). We suggest that Ycf48 is approximately threefold more abundant than Pam68 because it associates with a greater number of precursor modules: CP47, CP43 (Rengstl et al. 2013 ), D1 and RCII (Yu et al. 2018 ). Another component of the D1 assembly complex, with evidence of a direct association with Ycf48, is RubA (Slr2033; Kiss et al. 2019 ), quantified here at 2850–7150 cpc. Psb27 and Psb28 Analysis of intermediates in PSII biogenesis by immunoblotting identified Psb27 as a component of assembly modules containing PsbC/CP43 (Komenda et al. 2012 ; Fig. 4 b) and recent structural analyses by cryo-EM revealed not only docking sites on PsbC/CP43 (Zabret et al. 2021 ) but also the induction of conformational changes in PsbB/CP47 and PsbD/D2 (Huang et al. 2021 ). According to IP analysis, Psb28 co-isolates with the RC47 assembly module (Bečková et al. 2017 ), docking with both PsbA/D1 and PsbD/D2 subunits, where it induces temporary conformational changes that may protect nascent PSII from photo-damage until the Mn 4 CaO 5 cluster is assembled and water oxidation activated (Zabret et al. 2021 ). Psb27 and Psb28 (Sll1398) were quantified as relatively high abundance assembly factors at 3500–60,500 and 28,000–100,000 cpc respectively, reflecting their participation in interactions with multiple PSII assembly intermediates (Komenda et al. 2012 ; Bečková et al. 2017 ; Pascual-Aznar et al. 2021 ). A second isoform of Psb28 encoded by slr1739 (Boehm et al. 2012a ) and proposed to have a divergent function (Bečková et al. 2017 ) was identified with only two peptides in the label-free DDA analysis (Supplementary Data Set S4) and therefore not validated for quantification. Thylakoid membrane biogenesis and photosystem I assembly It may be argued that CurT facilitates the accumulation of PSII indirectly via its role in inducing the correct TM architecture and that this activity of CurT also provides the TM environment for PSI biogenesis. Unexpectedly however, inactivation of curT has been shown to have no effect on PSI abundance (Heinz et al. 2016 ), highlighting a link between TM biogenesis and PSI assembly via an alternative mechanism. VIPP1 In addition, referred to as IM30, VIPP1 was first identified in chloroplasts as essential for TM biogenesis (Kroll et al. 2001 ), with membrane insertion of PsaA and PsaB compromised in a Δ vipp1 strain of Synechococcus sp. PCC 7002. This evidence suggests that VIPP1 is actually functional in PSI biogenesis and that the presence of PSI, with CurT, is required for the formation of normal TM architecture (Zhang et al. 2014 ). Structural characterization has established that VIPP1 binds to the membrane surface as homo-oligomers of > 1 MDa (Aseeva et al. 2004 ) forming stacked rings that induce curvature of the membrane into a central hydrophobic channel (Gupta et al. 2021 ). In view of this MDa size, the 40,500–45,000 cpc abundance determined in this analysis (Fig. 4 c) suggests the actual number of functional VIPP1 units within the TM may be only 1000–1200. Unlike CurT, which may be maintained at high abundance because its membrane curvature activity is a constant requirement over the entire TM, VIPP1 is proposed to have a localized function at convergence zones where it may participate in lipid transfer between the CM and nascent TM (Gupta et al. 2021 ). Ycf3, Ycf37 and Ycf4 Although PSI assembly occurs so rapidly that the isolation of intermediate subcomplexes has proved challenging (Schöttler et al. 2011 ), several assembly factors have emerged from the investigation of mutants with defective PSI accumulation (Wilde et al. 1995 ; Boudreau et al. 1997 ; Bartsevich and Pakrasi 1997 ; Wilde et al. 2001 ). Ycf3 is a hydrophilic TPR protein that was found to co-isolate in IP analysis with PsaA and PsaD on the cytoplasmic surface of the TM (Naver et al. 2001 ). A similar strategy demonstrated the association of subunits PsaA-D with Ycf37, a TPR protein anchored in the TM by one TMH, also on the cytoplasmic surface. Furthermore, Ycf37 co-migrated with PSI(1), not PSI(3) after sucrose gradient centrifugation, indicating its possible role in PSI trimerization (Dühring et al. 2006 ). Ycf3 and Ycf37 are quantified here at comparable abundance levels: 2950–4750 and 1750–5300 cpc respectively (Fig. 4 c), suggesting that these proteins may have similar stoichiometric relationships with nascent PSI complexes. Their approximate 20-fold lower abundance than the cellular PSI population (86,000–118,500 cpc; Fig. 1 a) supports the view that Ycf3 and Ycf37 operate as chaperones, interacting only transiently with their respective binding partners. The use of TAP-tagged Ycf4 has revealed interactions with six subunits: PsaA-PsaF, with stability dependent on bound PsaF (Ozawa et al. 2009 ). In a model proposed by Nellaepalli et al. ( 2021 ), Ycf3 is the first to bind to a newly synthesized PsaA/B heterodimer, followed by Ycf4 which stabilizes the complex as more subunits join, followed by Ycf37. Ycf4 is quantified at 6800–14,000 cpc (Fig. 4 c), almost twofold higher than Ycf3 and Ycf37. This 1:1:2 stoichiometry for Ycf3:Ycf37:Ycf4 may be explained if Ycf4 is active as a dimer. FtsH proteases The PsbA/D1 subunit of PSII is highly susceptible to photo-oxidative damage as part of its normal function (Adir et al. 2003 ). To maintain continuity of PSII activity and enable acclimation to changing illumination, a repair mechanism has evolved in oxygenic phototrophs (Nixon et al. 2005 ). Four homologous, membrane intrinsic, ATP-dependent Zn-metalloproteases encoded by ftsH1-4 are involved in PSII repair and a wide range of other cellular processes. Their functional diversity is based on the assembly of different combinations of both homo- and hetero-oligomeric complexes (Mann et al. 2000 ; Boehm et al. 2012b ). All four were quantified in the analysis reported here, ensuring that only unique proteotypic tryptic peptides were used. The Synechocystis FtsH proteases all have similar sequence identities (41–47%) to the single FtsH occurring in E. coli , with a reported abundance of 660 cpc (Wiśniewskia and Rakus, 2014 ). This level is in close agreement with our determination of FtsH1 at < 1000 cpc (Fig. 4 d), suggesting that FtsH1 is the basic isoform (Bittner et al. 2017 ). The 3–sixfold higher levels of FtsH2, FtsH3 and FtsH4 at 1800–5150, 2400–3000 and 2800–5950 cpc, respectively, may reflect functions that extend beyond the remit of a basic FtsH protease, with a greater number of protein targets. Both FtsH2 and FtsH3 are functional in the degradation of photo-damaged PsbA/D1 (Silva et al. 2003 ; Komenda et al. 2006 ), specifically as an FtsH2/3 complex (Boehm et al. 2012b ). An FtsH1/3 complex plays a role in the regulation of gene expression in acclimation to nutrient stress (Krynická et al. 2014 ; 2019 )."
} | 17,867 |
29133886 | PMC5684352 | pmc | 2,773 | {
"abstract": "Species establish different interactions (e.g. antagonistic, mutualistic) with multiple species, forming multilayer ecological networks. Disentangling network co-structure in multilayer networks is crucial to predict how biodiversity loss may affect the persistence of multispecies assemblages. Existing methods to analyse multilayer networks often fail to consider network co-structure. We present a new method to evaluate the modular co-structure of multilayer networks through the assessment of species degree co-distribution and network module composition. We focus on modular structure because of its high prevalence among ecological networks. We apply our method to two Lepidoptera-plant networks, one describing caterpillar-plant herbivory interactions and one representing adult Lepidoptera nectaring on flowers, thereby possibly pollinating them. More than 50% of the species established either herbivory or visitation interactions, but not both. These species were over-represented among plants and lepidopterans, and were present in most modules in both networks. Similarity in module composition between networks was high but not different from random expectations. Our method clearly delineates the importance of interpreting multilayer module composition similarity in the light of the constraints imposed by network structure to predict the potential indirect effects of species loss through interconnected modular networks.",
"introduction": "Introduction Species establish multiple interactions with other species throughout their life cycle. For instance, plants can be attacked by herbivores and seed predators, pollinated by flower visitors and dispersed by birds 1 – 3 . In some cases, a given organism can also behave as a mutualistic and antagonistic partner of the same species (e.g. adult insects of a given species can behave as pollinators or nectar robbers, or adult insects can act as pollinators while their larvae are herbivores 4 , 5 ). These multiple interactions among species can be integrated in multilayer interaction networks, that is, networks encompassing different types of links between species. Interaction networks often show non-random topological structures and properties 3 , 6 – 8 . These properties and structures can affect the ecological and evolutionary dynamics of species assemblages and therefore biodiversity 6 , 7 , 9 – 18 . In this context, one important challenge for network ecologists is to develop tools to analyse the co-structure of multilayer interaction networks 2 , 19 – 23 , because such co-structure properties might be key to understand how perturbations (e.g. species loss) can propagate across and between linked networks. Although studies investigating ecological multilayer networks do exist, they all fail to consider similarities of co-structure through a proper statistical framework. One of the most prevalent patterns found in ecological networks is modular structure. Modular networks emerge when subsets of species interact more among themselves than with other species of the network 24 . Modular structures have been reported in classic food webs 7 , plant-herbivore 10 , 25 , host-parasitoid 25 , 26 , plant-pollinator 27 , 28 , ant-plant 11 , and plant-frugivore networks 29 , 30 . The modularity of ecological networks may be influenced by features such as interaction type (e.g. antagonistic networks may show higher modularity than mutualistic ones) and intimacy, i.e. the degree of biological integration between interacting individuals (e.g. among plant-ant interactions, non-symbiotic and symbiotic interactions are of low and high intimacy and show low and high modularity, respectively) 10 , 11 , 19 , 31 . Modularity may be more frequently observed in networks of species that establish antagonistic interactions of low intimacy (e.g. herbivory) than in low-intimacy mutualistic networks (e.g. pollination) 10 . However, species-rich mutualistic networks describing low-intimacy interactions may frequently show modular structures (e.g. pollination networks of > 150 species) 27 . In this article, we present a new comparative method aimed at disentangling the co-structure of multilayer interaction networks, with emphasis on the analysis of network modularity. Since network structure determines the ecological and evolutionary dynamics of multispecies assemblages, understanding the co-structure of interlinked networks is a first key step to unravel the effects that species loss may have on the maintenance of biodiversity 19 . To disentangle the co-structure of multilayer interaction networks, we first propose to compare their distributions of species degree. This degree co-distribution analysis allows understanding the association between the number of interactions (i.e., the degree) that a species establish in one network with the same species’ degree in the other network, and thus helps hypothesize how species loss can propagate through multilayer networks. In this context, we introduce the use of mosaic plots to represent over- and under-representation of species interaction patterns among plants and lepidopterans, advancing the methods proposed in previous studies 2 , 5 , 19 , 22 , 32 , 33 . Next, we propose a statistical test to characterize the similarity in module composition between multilayer networks, i.e. in subgroups of species that interact more within groups than among them, using the normalized mutual information of the two classifications of species induced by network modules. The two steps of our method straightforwardly apply to bipartite interaction networks that share species. The most classic example would be organisms that establish different interactions at different life stages, such as herbivorous insect larvae and their pollinating adult stages with plant species. Our comparative analysis can also be extended to study the co-structure of other multilayer ecological networks such as those including pollinators, plants and nectar robbers, and of networks describing spatial or temporal variation of interactions. We illustrate our method by comparing the co-structure of two Lepidoptera-plant networks from the state of Baden-Württemberg (Germany), one describing low-intimacy antagonistic interactions (i.e. the herbivory network) and the other describing low-intimacy mutualistic interactions (i.e. the flower visitation network). The larvae of most Lepidoptera species (caterpillars) feed on plant tissues, thus establishing antagonistic interactions with plants, whereas the adult lepidopterans visit flowers to feed on nectar and can pollinate them, and thus can act as mutualistic partners 5 . Caterpillars are often characterised by a very particular host plant range and adult lepidopterans often feed on nectar from only a few key flower species 5 . Indeed, sympatric plant species, even if they are closely related, are visited by different moth species 5 . The specificity of antagonistic and mutualistic interactions among plants and Lepidoptera species could be translated into modular structures in interaction networks 34 . Moreover, caterpillars host breadth may influence the number of nectar sources with which adults interact and adults tend to feed on nectar from plant species on which they have fed as larvae 5 . Therefore, similarities in species degree distribution and overlap in module composition among these antagonistic and mutualistic networks can be expected.",
"discussion": "Discussion Disentangling the ecology and evolution of species immersed in multispecies assemblages implies understanding the organization of multiple types of interactions across complex networks 1 , 15 , 19 . Thus, developing methods to analyse the structure of multilayer networks should be a priority for the research agenda of network ecologists 2 , 19 – 22 . Here we presented a method to evaluate how species interaction patterns change between paired networks by focusing on two key features of network organisation: the degree distribution and the modular structure of ecological networks. The application of our method to Lepidoptera-plant herbivory and visitation networks showed that (1) species that established either herbivory or visitation interactions but not both were prevalent (more than 50% of species) and over-represented among plants and lepidopterans, and were present in most modules in both networks; (2) species whose degree highly increased or decreased from one network to the other were over-represented; (3) similarity in module composition between networks was high but not different from random expectations. Hereafter, we discuss the contributions of our method in relation to existing methodological approaches to multilayer networks and the impact of our results in the light of previous studies on herbivory and visitation networks 2 , 5 , 19 , 22 , 33 , 35 . The likely effects of species loss on multispecies assemblage persistence suggested by our results are also discussed to offer new directions for future studies on multilayer networks. A first step towards understanding how species interact in multispecies assemblages lies in comparing species interaction patterns among ecological networks depicting different interaction types. Most existing studies explored changes in the number of species with which species interact (i.e. their degree) between ecological networks 2 , 5 , 19 , 22 , 33 . The simplest relationship that can be tested is a correlation between species degrees in these different networks 5 , 19 , 22 . Analysing correlations assumes that interactions among organisms may depend on the ability of individuals to detect (mobile organisms) and attract (sessile ones) each other independently of interaction type 19 . For instance, herbivores and nectar robbers may detect plants by recognizing the same phenotypical signals that pollinators do (e.g. floral display), thus plant traits that attract pollinators may also attract herbivores and nectar robbers 19 . Even though correlations may inform about key generalities of species interaction patterns, they may over-simplify them. There are different methods allowing the identification of over- and under-represented interaction patterns. For instance, the distribution of the ratio of degrees in different ecological networks 2 , 32 or the rank-degree curves of the same set of species in networks describing different interaction types 33 can be compared with those obtained in networks constructed under different null models. However, working with degree-ratios may not discriminate between species interacting with few species in both networks (e.g. with 5% of species) from those that interact with several species in both (e.g. with > 90% of species of the network). In the same vein, the visual comparison of rank-degree curves 33 may be difficult when differences are not so evident, and this procedure lacks a proper statistical test. Moreover, to test if combinations of species degrees are over- or under-represented, randomized networks should maintain the degree distribution of observed networks. Constraining the randomization of observed networks only by maintaining connectance may lead to misleading results since network connectance strongly influences variability in species degree distribution 36 . Our proposal advances the analysis of interaction patterns in multilayer ecological networks by identifying combinations of species degrees that are over- and under-represented, i.e. that are more or less frequent than expected under the assumption of independent interactions of different types. In the dataset we used for illustration purpose, we found higher prevalence and over-representation of plants with highly asymmetrical combinations of degrees (i.e. those plants which only established interactions with larval or adult lepidopterans and plants that were eaten by few species of herbivores but visited by several species of adult lepidopterans). Extremely specialized plants (i.e. those interacting with few Lepidoptera species in both networks) were under-represented. As far as we know, there are only two datasets exploring the visitation and herbivory interaction patterns of plant species in multispecies assemblages and using the network approach. Pocock et al . 3 studied multispecies interactions in an agroecosystem in England (hereafter the Norwood dataset) and found that some individual plant species were disproportionately well linked to many visitor and herbivore species, but these plants differed between the visitation and the herbivory networks. Sauve et al . 22 explored plant degree correlations in the Norwood dataset and found that the number of flower visitors was positively correlated to the number of herbivores that interacted with plants. Since the reported degree-correlations are low 22 and different from what can be expected by degree distribution, taken together these results suggest that asymmetrical interaction patterns may be prevalent and over-represented among plants of the Norwood dataset, which is in accordance with our results. Melián et al . 2 studied antagonistic (herbivory) and mutualistic (visitation and frugivory) interactions of the Doñana natural reserve (Spain) and also found that asymmetrical interaction patterns were prevalent among plant species. In the Doñana multispecies assemblage, most plants had low mutualistic-to-antagonistic ratios (i.e. plant species interacted with higher number of herbivores than species of pollinators and seed dispersers) and few had much higher mutualistic-to-antagonistic ratios than expected by chance. The analysis of the degree co-distribution of larval and adult Lepidoptera species showed that highly asymmetrical foraging interaction patterns were prevalent and over-represented. These patterns included Lepidoptera species that only interacted with plants in the adult stage and those that had specialist larvae and generalist adults or vice versa. Past work comparing the plant as herbivore hosts and nectar sources of species that can be herbivores and floral visitors studied a subset of the data analysed in our study 5 . They found that the number of plant species on which Lepidoptera species feed as larvae was positively correlated with the number of plant species on which adults look for nectar, this relationship being stronger for diurnal Lepidoptera species 5 . Similarly, the degree of adult lepidopterans was also higher for oligophagous and polyphagous species than for monophagous and strictly oligophagous ones in diurnal Lepidoptera species 5 . Our results complement these findings by showing the highly asymmetrical nature of the interactions established by larval (herbivores) and adult (flower visitors) lepidopterans. As far as data quality is concerned, the plant-insect interaction dataset we used is virtually complete, and thus highly robust (see also Pearse & Altermatt 37 for an analysis on the robustness of the dataset when removing interactions), but also does not underrepresent the interactions of rare species. Indeed, a correlation between interaction records/degree of interactions and the rarity/commonness of a species would be problematic 38 . However, our dataset is based on the sum of observations of hundreds of entomologists, and rare species often received disproportionate attention (see also 5 ), such that plant-insect interactions are very well resolved for all species, regardless of its rarity. In fact, some species may be rare because they have only few interactions, and thus are limited by their host plant use (see Pearse & Altermatt 39 on that dataset). Differences between insect herbivores and flower visitors in their level of generalism seem to be widespread. Fontaine et al . 40 studied the interactions established by insect herbivores and flower visitors of species belonging to 44 plant–insect networks describing either visitation or herbivory communities. They found that insect flower visitors tend to interact with far more plant species than herbivores 40 . This difference was mainly attributed to differences in the structure of antagonistic (modular) and mutualistic (nested) networks, which may promote, respectively, the evolution of specialization and generalism 40 and system stability 10 . We found that adult lepidopterans interacted with higher number of plants than larval lepidopterans. However, by looking at the degree co-distribution of the larval and adult stages of species, we found that higher generalism in herbivores than in flower visitors can also be prevalent and more frequent than expected. Thus, analysing species interaction among modular networks depicting different interaction types may challenge our current understanding of the ecological and evolutionary mechanisms that modulate the distribution of generalism among species 40 . Understanding the organization of multispecies assemblages also involves analysing similarity of species interaction partners among networks. Similarity analyses at the species level, as performed in previous studies, may be highly informative 2 , 20 , 22 . However, as interaction networks show well-defined structures that modulate the ecological and evolutionary dynamics of species interaction patterns 6 , 7 , 15 , incorporating network structure into similarity analyses may advance our understanding of the functioning of multispecies assemblages 19 , 22 . Based on both the high prevalence of modular structures among multispecies assemblages and the idea that modules may be the functional and evolutionary building blocks of ecological networks 7 , 10 , 11 , 27 , we proposed a method to evaluate the similarity in module composition between multilayer networks. Our analysis involves the characterization of the modular structure of multilayer networks, the assessment of their similarity and the comparison of this similarity with values obtained from randomized networks. Several methods that allow the classification of interacting species in modules (i.e. clustering methods) have been proposed for binary and quantitative networks 41 – 43 . The ability of the different methods to retrieve module composition may depend on network properties 42 , and there is a limit to the resolution of such methods, i.e. modules under a certain size might be undetectable whatever the method used 44 . The use of the eigenvector-based maximizing modularity algorithm 45 in our method relies on results of previous studies showing that it is among the clustering methods that best classify species in modules in binary networks, while requiring the lowest computational time 42 , 45 , 46 . There are also several similarity measures to compare the species composition of modules from multilayer networks 46 – 48 . Measures based on information theory, as the one used in this article, are built on the idea that if species are grouped similarly in two networks, little information is needed to infer the structure of one of the networks given the other 46 . The use of mutual information measures is encouraged because they are not affected by the number and size of modules found in each network as other similarity measures are (e.g. pair counting measures) 48 . As the normalized mutual information index cannot be easily interpreted when it is far from 0 (independent classifications by modules of the two networks) or 1 (same classification by modules of the two networks), network visualization tools may facilitate the analysis of its biological significance, as illustrated by our results. Network features such as size, connectance and degree distribution can impose constraints on network structure 36 , 49 – 51 . Thus similarity in module composition among networks needs to be compared with expectations from random network structure constrained by degree distributions (as performed here), and not with expectations from random network of the same size but with different degree distributions, nor with expectations from random network sharing degree distribution but of different size. Network connectance is negatively correlated with network modularity 10 and herbivory networks have both lower connectance and higher modularity than visitation networks 10 , 52 . Thus, it can be expected that among sets of species establishing visitation and herbivory interactions, species within visitation modules may likely be spread among herbivory modules, which may constrain module similarity between networks. Since species degree constrains the interaction pattern of species 53 , the degree distribution of multilayer networks may also limit their similarity in module composition. Considering the range of similarity values that can be achieved by randomized networks with the same size, connectance and degree distribution, we showed that similarity in module composition between the studied herbivory and visitation networks was high, although not different from random expectations. The importance of degree distribution in predicting the similarity in species interaction partners between herbivory and visitation networks has been little explored in previous studies. In the Norwood dataset, degree distribution predicted that the similarities in flower visitors and in herbivores among pairs of plant species were unrelated 22 . In both the Norwood and the Doñana datasets, the similarity in plant groups between the herbivory and visitation networks was higher than expected by the number of plant groups of each individual network (i.e. groups being subsets of plant species interacting with more similar sets of herbivores and flower visitors) 20 . However, how degree distribution is associated with this similarity remains unknown for the Norwood and Doñana datasets. Indirect evidence of how widespread the role of degree distribution might be in influencing the similarity of herbivory and visitation networks can be found in the results reported by Fontaine et al . 40 . According to this study, herbivore species interacted with plants that were more phylogenetically related than flower visitors did, and plant phylogenetic relatedness was negatively associated with the degree of herbivores, but unrelated to the degree of visitors 40 . In herbivory networks, phylogenetically related plants tend to share modules 25 and in visitation networks they tend to interact with more similar partners, which also was found for flower visitors 54 . Thus, the difference in the phylogenetic relatedness of plants interacting with herbivores and pollinators and its relationship with species degree 40 suggests that degree distribution may modulate the structural similarities in species composition between herbivory and visitation networks."
} | 5,672 |
34591868 | PMC8483398 | pmc | 2,774 | {
"abstract": "Scum is formed by the adsorption of long-chain fatty acids (LCFAs) onto biomass surface in anaerobic digestion of oily substrates. Since scum is a recalcitrant substrate to be digested, it is disposed via landfilling or incineration, which results in biomass washout and a decrease in methane yield. The microbes contributing to scum degradation are unclear. This study aimed to investigate the cardinal microorganisms in anaerobic scum digestion. We pre-incubated a sludge with scum to enrich scum-degrading microbes. Using this sludge, a 1.3-times higher methane conversion rate (73%) and a faster LCFA degradation compared with control sludge were attained. Then, we analyzed the cardinal scum-degrading microbes in this pre-incubated sludge by changing the initial scum-loading rates. Increased 16S rRNA copy numbers for the syntrophic fatty-acid degrader Syntrophomonas and hydrogenotrophic methanogens were observed in scum high-loaded samples. 16S rRNA amplicon sequencing indicated that Syntrophomonas was the most abundant genus in all the samples. The amino-acid degrader Aminobacterium and hydrolytic genera such as Defluviitoga and Sporanaerobacter became more dominant as the scum-loading rate increased. Moreover, phylogenic analysis on Syntrophomonas revealed that Syntrophomonas palmitatica , which is capable of degrading LCFAs, related species became more dominant as the scum-loading rate increased. These results indicate that a variety of microorganisms that degrade LCFAs, proteins, and sugars are involved in effective scum degradation.",
"conclusion": "Conclusions This study aimed to explore the microbial community that played a key role in anaerobic scum digestion. The pre-incubated sludge with scum (Sludge I) showed a 73% ± 3% methane conversion rate, which was 1.3-times higher than that of Sludge II, and temporal VFA accumulation was not observed. The relatively high scum degradation potential of this sludge was confirmed. It was suggested that the degradation efficiency of not only LCFAs but also of other complex substrates such as proteins and polysaccharides affected the methane yield from scum. As the scum-loading rate increased, the larger 16S rRNA copy numbers of Syntrophomonas and hydrogenotrophic methanogens were detected. Aminobacterium , Defluviitoga , and Sporanaerobacter , which degrade protein- or polysaccharide-related substrates, also became more abundant as the scum loading amount increased. Phylogenetic analysis of the Syntrophomonas genus revealed that the predominant OTUs were close to S . palmitatica , which can degrade LCFAs. A higher abundance of S . palmitatica -related species in 17.3 g COD L -1 scum-loaded vials was observed. These likely played an important role in LCFA degradation. Overall, our results indicate that scum degradation is a complex process in which a variety of genera are involved. To promote scum degradation, it should not only be focused on Syntrophomonas but also on the hydrolytic genera suggested in this study.",
"introduction": "Introduction Anaerobic digestion can produce biomethane as an energy resource from a variety of organic wastes. Wastewater from food processing, edible oil producers, and slaughterhouses contains high concentrations of lipids [ 1 ]. Lipids have higher theoretical methane yields (1.01 m 3 kg -1 VS) than that of carbohydrates (e.g., 0.37 m 3 kg -1 VS for glucose) and proteins (0.74 m 3 kg -1 VS) [ 2 ]. However, there are several challenges in the anaerobic digestion of lipidic waste. Scum formation is one of the most severe problems. Scum is formed by the adsorption of floating substrates such as long-chain fatty acids (LCFAs) to the surface of biomass [ 1 , 3 ]. Scum is a recalcitrant substrate to be digested and inhibits the liquidity in the reactor. Accumulated scum must be periodically removed and disposed via landfills or incineration, which leads to the washout of organic waste. Scum formation and following biomass washout disturb biomethane production by reducing substrate utilization rates. The main cause of scum formation is LCFAs, which are produced by the hydrolysis of lipids. LCFAs are degraded to volatile fatty acids (VFAs). This is the rate-limiting step of the anaerobic digestion process, as the reaction proceeds only under very low H 2 pressure [ 4 ]. Moreover, LCFAs have been considered to adsorb onto the microbial surface and inhibit the activities of anaerobic microorganisms [ 5 ]. Previous reports showed that 0.5 mmol L -1 oleic acid (C18:1) or 2–4 mmol L -1 palmitic acid (C16:0) reduced methane generation by more than 50% [ 6 ]. Despite these troublesome characteristics of LCFAs, it has been reported that scum formation leading to biomass washout contributed more to the failure of anaerobic digestion than the toxicity of LCFAs themselves [ 1 ]. Biomass washout severely impairs the high methane convertibility of lipids. To reuse the collected scum as energy resources, several studies have employed scum as a co-digestion substrate with sewage sludge [ 7 ], thickened activated sludge, and primary sludge [ 3 , 8 ]. The development of efficient scum digestion methods enables reduction of the cost of disposal and increase in biomethane yield. If anaerobic scum digestion can be promoted, the application range of anaerobic digestion can be expanded to include more oily substrates. However, the key microbes that play an important role in anaerobic scum digestion are unclear. It has been reported that pre-incubation with lipidic substrates is an effective strategy for improving methane yield. A previous study reported that pre-incubation with soybean oil-based wastewater increased the relative abundance of Synergistales, which was highly correlated with the methane production rate [ 9 ]. Ziels et al., (2016) showed that acclimation with waste restaurant oil increased the relative abundance of the genus Syntrophomo nas, which has the capability of degrading various fatty acids [ 10 ]. The substrate-loading rate also affects the microbial community. Ziels et al., (2016) showed that an increase in the loading rates of lipidic substrates positively affected the abundance of Syntrophomonas 16S rRNA genes [ 10 ]. It is assumed that a higher loading rate of the substrate increases the core microbial population that plays an important role in substrate degradation. Our microbiological findings on anaerobic scum digestion will enable new approaches to solve the problems caused by scum. In this study, we aimed to reveal the microbes contributing to effective scum degradation. We pre-incubated a collected sludge with scum to enrich scum-degrading microbes. Additionally, to investigate the core scum-degrading microbes in this pre-incubated sludge, we evaluated the changes in microbial communities according to various loading rates.",
"discussion": "Results and discussion Experiment 1: Assessment of the digestibility of scum The cumulative methane production of each sludge is shown in Fig 2A . Until day 7, the methane production of Sludge I was 30 ± 1 mL g -1 VS. Then it increased to 121 ± 1 mL g -1 VS on day 11. In contrast, the methane production of Sludge II was only 19 ± 3 mL g -1 VS until day 11. The cumulative methane production on day 11 of Sludge I was 6-fold higher than that of Sludge II. Finally, 582 ± 27 and 457 ± 5 mL g -1 VS methane were produced from Sludge I and Sludge II, respectively. Methane conversion rates of Sludge I and Sludge II calculated from COD were 73% ± 3% and 57% ± 1%, respectively. Sludge I showed relatively high methane productivity from scum. The methane gasification process of scum was fitted with the modified Gompertz model in all samples (RMSE = 21 ± 6). The theoretical methane production of Sludge I and Sludge II calculated by the modified Gompertz model was 630 ± 25 and 470 ± 6 mL g -1 VS, respectively. Fig 2B shows the transition of total VFA concentrations. On day 0, total VFA concentrations were 60 ± 6 and 87 ± 4 mg L -1 in Sludge I and Sludge II, respectively. In Sludge I, VFA concentration then decreased gradually and became undetectable on day 22. In Sludge II, the VFA concentration gradually increased to 216 ± 13 mg L -1 on day 15. Among them, acetate was the dominant VFA (202 ± 13 mg L -1 ). The concentration then decreased gradually and became undetectable on day 30. Considering the low methane production in Sludge II, it was inferred that acetate conversion to methane was inhibited until day 15. Methanogens are more susceptible to LCFA toxicity than acetogens [ 29 ]. In particular, acetoclastic methanogens have been reported to be more sensitive to LCFAs than hydrogenotrophic methanogens [ 4 , 30 ]. It was inferred that LCFAs degradation in Sludge I was faster than that in Sludge II. 10.1371/journal.pone.0257651.g002 Fig 2 Cumulative methane production (A) and VFA concentration (B). Error bars represent the standard deviation of the mean (n = 3). The concentrations of total LCFAs were compared at two time points: day 8, when methane production was verified, and day 30, which was the end of our experimental fermentation ( Fig 3 ). The total LCFA concentrations on day 8 in Sludge I and Sludge II were 1,142 ± 179 and 2,721 ± 198 mg L -1 , respectively. The total LCFA concentration in Sludge I was only 42% of that in Sludge II (T.TEST, p = 0.02). Finally, most LCFAs were degraded. On day 30, LCFA concentrations were 4 ± 0 and 4 ± 1 mg L -1 in Sludge I and Sludge II, respectively. A previous study showed that palmitate inhibited anaerobic microorganisms with IC 50 values of more than 1,100 mg L -1 [ 5 ]. The high concentration of palmitic acid might be one of the causes of the temporal VFA accumulation and the slower methane production in Sludge II ( Fig 2 ). Though LCFAs and VFAs were almost completely digested in both sludges after the 30-day incubation, the methane conversion rate of Sludge I was 26% higher than that of Sludge II. Other substrates than fatty acids may have remained undegraded in Sludge II. The VS decomposition rate was higher in Sludge I (72.7% ± 0.9%) than in Sludge II (64.6% ± 0.4%) (T.TEST, p <0.05) ( S1A Fig ). The dissolved COD concentrations in Sludge I and Sludge II on day 30 were 68 ± 6 and 62 ± 3 mg L -1 , respectively, and there were no significant differences (T.TEST, p >0.05) ( S1B Fig ). These results suggest that non-dissolved, complex substances, such as proteins or polysaccharides, remained undigested in Sludge II. The rate of hydrolysis depends on several parameters such as pH, particle size, and the diffusion barrier between an enzyme and substrate [ 1 ]. LCFAs may have prevented hydrolytic enzymes from reacting with biomass. It was inferred that the degradation efficiency of LCFAs affected the degradation efficiency of other substrates. 10.1371/journal.pone.0257651.g003 Fig 3 The LCFAs concentration in each reactor. Error bars represent the standard deviation of the mean (n = 3). Our results confirmed that Sludge I had a relatively high potential for methane gasification of scum. It was assumed that scum-degrading microorganisms were enriched in Sludge I. To reveal the core scum-digesting community, the differences in microbial communities were evaluated according to the scum loading concentration. Experiment 2: Exploration of core scum-digesting microbes The effect of scum loading concentration on methane productivity Cumulative methane production is shown in Fig 4A . During the 30-day incubation, 582 ± 27, 607 ± 8, and 563 ± 22 mL g -1 VS methane were produced from 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials, respectively. The methane gasification process of scum was well fitted with the modified Gompertz model in all samples (RMSE = 41 ± 13). The theoretical methane productions calculated by the modified Gompertz model were 643 ± 37, 667 ± 7.3, and 630 ± 25 in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials, respectively. The methane conversion rates in 8.6, 12.6, and 17.3 g COD L -1 loaded vials were 73% ± 3%, 77% ± 1%, and 70% ± 3%, respectively. There were no significant differences between these values (TukeyHSD, p >0.05). The total VFAs gradually decreased in all the samples ( Fig 4B ). Temporal VFA accumulation was not observed. The total LCFA concentrations in 8.6, 12.6, and 17.3 g COD L -1 loaded vials were 1,142 ± 179, 4,987 ± 222, and 6,869 ± 585 mg L -1 , respectively ( S2 Fig ). Finally, most LCFAs were degraded in all the samples. Total LCFAs on day 30 in 8.6, 12.6, and 17.3 g COD L -1 loaded vials were 3.3 ± 0.3, 5.7 ± 0.3, and 19.3 ± 8.0 mg L -1 , respectively. Remarkably, the total LCFA concentrations in 12.6 and 17.3 g COD L -1 loaded vials were much higher than that in Sludge II (Experiment I), in which VFA accumulation and lower methane productivity was observed. Even though high concentrations of LCFAs were detected in scum high-loaded vials, inhibitory effects on scum degradation and methane productivity were not observed in comparison with 8.6 g COD L -1 loaded vials. VS decomposition rates were 73% ± 1%, 74% ± 1%, and 71% ± 2% in 8.6, 12.6, and 17.3 g COD L -1 loaded vials, respectively ( S3A Fig ) (TukeyHSD, p >0.05). Dissolved COD in each vial gradually decreased. Finally, the dissolved COD concentrations in 8.6, 12.6, and 17.3 g COD L -1 loaded vials were 68.3 ± 5.9, 94.7 ± 2.0, and 95.3 ± 8.0 mg L -1 , respectively ( S3B Fig ) (TukeyHSD, p >0.05). These results suggested that the scum digestion process was not inhibited in high-loaded samples. The inhibition caused by LCFAs has been reported as a reversible process [ 29 ]. In this study, LCFA-oxidizing bacteria may have degraded the LCFAs adsorbed onto methanogens smoothly. Methanogens and Syntrophomonas were quantified to assess the effect of various loading rates on these microbes. 10.1371/journal.pone.0257651.g004 Fig 4 Cumulative methane production (A) and VFA concentration (B) at the different initial scum loading concentrations. Error bars represent the standard deviation of the mean (n = 3). Quantitative dynamics of methanogens and Syntrophomonas by qPCR 16S rRNA gene copies for group-specific methanogens were quantified using real-time qPCR ( Table 2 ). Hydrogenotrophic methanogens reduce the partial pressure of hydrogen and enable syntrophic bacteria to proceed with fatty acid oxidation [ 1 ]. Methanomicrobiales and Methanobacteriales, which are hydrogenotrophic methanogens, gradually increased in all the samples. At the beginning of the batch experiment, the copy numbers of the 16S rRNA genes belonging to Methanomicrobiales were 7.75 ± 0.10 log copies mL -1 . On day 8, the copy numbers in 8.6, 12.6, and 17.3 g COD L -1 scum loaded vials were 8.63 ± 0.05, 8.67 ± 0.14, and 8.72 ± 0.04 log copies mL -1 , respectively. On day 30, those in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 8.65 ± 0.06, 9.01 ± 0.05, and 9.06 ± 0.07 log copies mL -1 , respectively. The 16S rRNA gene concentration belonging to Methanomicrobiales was significantly higher in 12.6 and 17.3 g COD L -1 scum-loaded vials than those in 8.6 g COD L -1 loaded vials (TukeyHSD, p <0.05). On day 0, the 16S rRNA gene concentration belonging to Methanobacteriales was 5.47 ± 0.01 log copies mL -1 . On day 8, the copy numbers in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 5.88 ± 0.02, 6.10 ± 0.02, and 6.09 ± 0.03 log copies mL -1 , respectively. The 16S rRNA gene concentration belonging to Methanobacteriales in 12.6 and 17.3 g COD L -1 scum-loaded vials was significantly higher than that in 8.6 g COD L -1 loaded vials (TukeyHSD, p <0.05). On day 30, the concentrations in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 6.71 ± 0.09, 6.70 ± 0.03, and 6.76 ± 0.05 log copies mL -1 , respectively. In summary, the copy numbers of the 16S rRNA genes for Methanomicrobiales or Methanobacteriales were significantly higher in scum high-loaded samples on days 8 and 30. It was inferred that hydrogenotrophic methanogenesis was the key process during anaerobic scum degradation. 10.1371/journal.pone.0257651.t002 Table 2 The transition of 16S rRNA copies for acetoclastic and hydrogenotrophic methanogens (log copies mL -1 ). 8.6 g COD L -1 12.6 g COD L -1 17.3 g COD L -1 \n Methanomicrobiales \n Day 0 (seed sludge) 7.75 ± 0.10 Day 8 8.63 ± 0.05 8.67 ± 0.14 8.72 ± 0.04 Day 30 8.65 ± 0.06 a 9.01 ± 0.05 b 9.06 ± 0.07 b \n Methanobacteriales \n Day 0 (seed sludge) 5.47 ± 0.01 Day 8 5.88 ± 0.02 a 6.10 ± 0.02 b 6.09 ± 0.03 b Day 30 6.71 ± 0.09 6.70 ± 0.03 6.76 ± 0.05 \n Methanosarcinaceae \n Day 0 (seed sludge) 5.64 ± 0.05 Day 8 6.38 ± 0.07 6.12 ± 0.03 6.13 ± 0.04 Day 30 9.20 ± 0.05 9.26 ± 0.03 9.37 ± 0.04 \n Methanosaetaceae \n Day 0 (seed sludge) 7.89 ± 0.01 Day 8 7.06 ± 0.02 a 7.38 ± 0.05 b 7.21 ± 0.04 ab Day 30 9.92 ± 0.11 9.95 ± 0.03 10.08 ± 0.04 All values represent the mean of triplicate reactors ± standard deviation. Different letters indicate significant differences on the same day (TukeyHSD, p <0.05). Methanosarcinaceae and Methanosaetaceae, which are acetoclastic methanogens, showed different quantitative transitions. The copy number of the 16S rRNA genes belonging to Methanosarcinaceae was 5.64 ± 0.05 log copies mL -1 on day 0. The copy numbers in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 6.38 ± 0.07, 6.12 ± 0.03, and 6.13 ± 0.04 log copies mL -1 on day 8 and 9.20 ± 0.05, 9.26 ± 0.03, and 9.37 ± 0.04 log copies mL -1 on day 30, respectively. The 16S rRNA gene copy numbers of Methanosarcinaceae increased gradually in all the samples, and there were no significant differences with respect to scum-loading rates (TukeyHSD, p >0.05). Meanwhile, the copy number of the Methanosaetaceae 16S rRNA gene on day 0 was 7.89 ± 0.01 log copies mL -1 . On day 8, the copy numbers in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 6.38 ± 0.07, 6.12 ± 0.03, and 6.13 ± 0.04 log copies mL -1 , respectively. On day 30, those in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 9.92 ± 0.11, 9.95 ± 0.03, and 10.08 ± 0.04, respectively. The 16S rRNA gene copy number of Methanosaetaceae decreased from day 0 to day 8 in all the samples. Methanosaetaceae prefers low acetate concentration [ 31 ]. It was simulated that Methanosaetaceae would compete for acetate utilization with Methanosarcinaceae when the acetate concentration was lower than 522.9 mg L -1 [ 32 ]. Besides, Methanosaeta has been reported to be more sensitive to environmental stress such as ammonia toxicity and overcharging the loading rate than Methanosarcina [ 31 ]. Because the VFA concentration detected in this study was kept low ( S2A Fig ), other stress seemed to inhibit the growth of Methanosaetaceae in the early stage of scum digestion. Through the experiment, the effect of the scum-loading rate on Methanosarcinaceae and Methanosaetaceae was not observed clearly. The high-loaded scum may have restricted the growth of acetoclastic methanogens. In anaerobic scum digestion, LCFA degradation seemed to be an important process that affects the entire digestion speed and methane conversion rate. 16S rRNA gene copies for Syntrophomonas increased in all the samples during the 30-day incubation ( Table 3 ). At the beginning of the batch experiment, the copy number of the 16S rRNA genes for Syntrophomonas was 5.56 ± 0.10 log copies mL -1 . On day 8, the copy numbers in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 6.71 ± 0.06, 6.80 ± 0.03, and 6.59 ± 0.04 log copies mL -1 , respectively. On day 30, 16S rRNA gene copy number for Syntrophomonas in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were 7.29 ± 0.04, 7.55 ± 0.03, and 7.77 ± 0.07 log copies mL -1 , respectively. On day 30, 16S rRNA gene copies for Syntrophomonas observed in 17.3 g COD L -1 loaded vials were 3.15-times higher than those in 8.6 g COD L -1 loaded vials (TukeyHSD, p <0.05). Similar to hydrogenotrophic methanogens, Syntrophomonas abundance increased as the scum-loading rate increased. These results suggest that syntrophic fatty acid degradation by Syntrophomonas and hydrogenotrophic methanogens was a key process in scum degradation. It was reported that the abundance of Syntrophomonas was correlated with the specific mineralization rates of LCFA [ 10 , 16 ]. Even in anaerobic scum digestion, the 16S rRNA concentration of Syntrophomonas may be used as an indicator of the scum-degrading potential. The transition of Syntrophomonas 16S rRNA concentrations was analyzed expecting that LCFA degradation was one of the key processes to determine the efficiency of methane production of scum. The results of Experiment 1 suggested that other complex substances such as protein- or polysaccharide-degrading kinetics, also affected the methane conversion rate. Next, the differences in bacterial communities were analyzed with respect to various scum-loading rates. 10.1371/journal.pone.0257651.t003 Table 3 The transition of 16S rRNA copies for Syntrophomonas (log copies mL -1 ). 8.6 g COD L -1 12.6 g COD L -1 17.3 g COD L -1 \n Syntrophomonas \n Day 0 (seed sludge) 5.56 ± 0.10 Day 8 6.71 ± 0.06 6.80 ± 0.03 6.59 ± 0.04 Day 30 7.29 ± 0.04 a 7.55 ± 0.03 ab 7.77 ± 0.07 b All values represent the mean of triplicate reactors ± standard deviation. Different letters indicate significant differences on the same day (TukeyHSD, p <0.05). The effect of scum-loading concentration on bacterial community The V3-V4 region of the 16S rRNA gene was amplified and sequenced. A total of 48,529–52,393 non-chimeric reads (average 49,957) were obtained from the raw MiSeq data. PCoA suggested that the bacterial communities in 12.6 and 17.3 g COD L -1 loaded vials were relatively similar ( S4 Fig ). Fig 5A shows the top 10 abundant phyla belonging to bacteria. In 8.6 g COD L -1 loaded vials, Firmicutes (22.1%) were most abundant and followed by Proteobacteria (18.9%), Bacteroidetes (15.2%), and Cloacimonetes (9.7%). In 12.6 g COD L -1 loaded vials, Firmicutes (31.4%) were most abundant, followed by Bacteroidetes (19.3%), Cloacimonetes (9.8%), Proteobacteria (9.5%), and Synergistetes (8.9%). In 17.3 g COD L -1 loaded vials, Firmicutes (30.2%) were most abundant, followed by Bacteroidetes (20.7%), Proteobacteria (11.7%), Synergistetes (10.6%), and Cloacimonetes (6.4%). Firmicutes, Bacteroidetes, and Synergistetes were more predominant in scum high-loaded vials. These phyla contain hydrolytic and fermentative bacteria [ 1 ] and have been reported to play an important role in the co-digestion of fats, oil, and grease [ 1 , 33 ]. Firmicutes include several syntrophic bacteria which degrade various substrates and produce VFAs [ 34 ]. Bacteroidetes include various microbial genera that secrete various hydrolytic enzymes and disintegrate complex organic matter [ 35 ]. Synergistetes contain only Synergistaceae, which can ferment glucose and organic acids [ 36 ]. It was inferred that these phyla contributed to the hydrolysis of complex biomass and degrade the hydrolysate to VFAs. 10.1371/journal.pone.0257651.g005 Fig 5 Relative abundance of bacterial groups based on 16S rRNA genes amplicon sequencing on day 8. The 10 most abundant phyla (A) and the 12 dominant genera (B) are shown. The top 12 most abundant genera are shown in Fig 5B . In 8.6 g COD L -1 loaded vials, Syntrophomonas was the most abundant genus (8.3%) and followed by Macellibacteroides (4.5%), Mesotoga (2.9%), and Aminobacterium (1.7%). Macelibacteroides use the hydrolysate of polysaccharides as a substrate to produce acetic acid [ 37 ]. Mesotoga is reported to degrade a wide range of sugars [ 38 ]. In 12.6 g COD L -1 and 17.3 g COD L -1 loaded vials, Syntrophomonas was the most abundant genus (12.7% and 10.5%) and followed by Aminobacterium (4.6% and 6.2%), Macellibacteroides (3.1% and 3.4%), and Mesotoga (3.0% and 2.4%). Among the top 12 abundant genera in 12.6 g COD L -1 and 17.3 g COD L -1 loaded vials, the relative abundance of Aminobacterium , Defluviitoga , and Sporanaerobacter was more than twice as high as that in 8.6 g COD L -1 loaded vials ( Fig 6 ). Aminobacterium utilizes several amino acids to produce VFAs and ammonia [ 39 ]. Defluviitoga , which has often been detected in thermophilic biogas plants as the key hydrolytic bacterium, can utilize a large diversity of monosaccharides, disaccharides, and polysaccharides including cellulose and xylan [ 40 ]. Sporanaerobacter is reported to play a key role in the hydrolysis of proteins and polysaccharides [ 41 ]. Other hydrolytic genera also showed higher relative abundance in scum high-loaded samples. Proteiniclasticum , which hydrolyze proteins [ 42 ], showed a 1.7- and 2.3-times higher abundance in 12.6 and 17.3 g COD L -1 loaded vials, respectively. A previous report indicated that Proteiniclasticum became dominant by acclimation to fats, oil, and grease [ 34 ]. Bacteroides, which play a key role in the initial hydrolysis of protein, fat, cellulose, and other polysaccharides [ 43 ], showed a 1.6- and 2.4-times higher abundance in 12.6 and 17.3 g COD L -1 loaded vials, respectively. These results suggested that not only LCFA degradation but also the hydrolysis of complex organic matter such as proteins and polysaccharides was an important process in anaerobic scum digestion. Focusing on the abundance of Syntrophomonas may not be enough to promote methane gasification of scum. It is also important to increase the abundance of genera shown in this study. 10.1371/journal.pone.0257651.g006 Fig 6 Comparison of the relative abundance among the 12 dominant genera between 8.6 g COD L -1 and 12.6 or 17.3 g COD L -1 scum-loaded vials. The predominance of Syntrophomonas indicated the importance of syntrophic LCFA oxidation. It has been reported that five Syntrophomonas species can degrade LCFAs when they are co-cultured with hydrogenotrophic methanogens [ 44 ]. Ziels et al (2017) showed that the 16S rRNA gene concentration of several OTUs classified as Syntrophomonas indicated a closer relationship with oleate degradation kinetics than total Syntrophomonas [ 45 ]. Species-level analysis would enable us to evaluate the capacity of LCFA degradation more precisely. The sequences assigned to the genus Syntrophomonas were aligned into OTUs, and a phylogenetic tree was constructed. Fig 7 shows the top three dominant OTUs of each sample. The dominant OTUs in 8.6, 12.6, and 17.3 g COD L -1 scum-loaded vials were closest to S . palmitatica , which can oxidize straight-chain saturated fatty acids with carbon chain lengths of C4-C18 [ 46 ]. Including uncultured bacteria, OTU1 and OTU9 in 8.6 g COD L -1 , OTU12 and OTU67 in 12.6 g COD L -1 , and OTU3 in 17.3 g COD L -1 were closest to the clone PM63 (DQ459214) (Identity >97%), which was detected predominantly by DGGE analysis in the oleate and palmitate enrichment cultures [ 47 ]. As shown in Table 4 , it is noteworthy that the sum of the top 3 abundant OTUs, which were the most closely related to S . palmitatica , accounted for 67% of the reads in 17.3 g COD L -1 loaded vials. These OTUs possibly played an important role in LCFA degradation. Our results showed that the higher loading rates of scum induced a higher abundance of syntrophically LCFA-oxidizing and complex substrate-hydrolyzing microorganisms. 10.1371/journal.pone.0257651.g007 Fig 7 Phylogenetic tree based on 16S rRNA gene representing the top 3 abundant sequences in the taxon classified as Syntrophomonas . The significance of each branch is indicated at the nodes by bootstrap values (%) based on 100 replications. Only values greater than 50% are shown. The numbers following the scum loading concentration indicate the size of OTUs. 10.1371/journal.pone.0257651.t004 Table 4 Top 3 abundant taxa in each sample. OTU ID Closest species of 16S rRNA gene Oxidizable fatty acids length a Accession no. Identity % of sequences b OTU5 \n Syntrophomonas \n C4-18 AB274040 462/466 26.3 8.6 g COD L -1 \n palmitatica \n (99%) OTU1 \n Syntrophomonas \n C4-18 AB274040 443/466 15.8 8.6 g COD L -1 \n palmitatica \n (95%) OTU9 \n Syntrophomonas \n C4-18 AB274040 439/466 5.6 8.6 g COD L -1 \n palmitatica \n (94%) OTU3 \n Syntrophomonas \n C4-18 AB274040 462/466 30.9 12.6 g COD L -1 \n palmitatica \n (99%) OTU12 \n Syntrophomonas \n C4-18 AB274040 442/466 7.1 12.6 g COD L -1 \n palmitatica \n (95%) OTU67 \n Syntrophomonas \n C4-18 AB274040 439/466 7.0 12.6 g COD L -1 \n palmitatica \n (94%) OTU1 \n Syntrophomonas \n C4-18 AB274040 464/466 38.9 17.3 g COD L -1 \n palmitatica \n (99%) OTU3 \n Syntrophomonas \n C4-18 AB274040 446/466 12.3 17.3 g COD L -1 \n palmitatica \n (96%) OTU24 \n Syntrophomonas \n C4-18 AB274040 460/466 6.1 17.3 g COD L -1 \n palmitatica \n (99%) The best hits of the cultured microorganisms are given. a The number of oxidizable carbon chain lengths of straight-chain saturated fatty acids when co-cultured with hydrogenotrophic methanogens. b % of sequences assigned to the Syntrophomonas genus."
} | 7,232 |
22675595 | PMC3368408 | pmc | 2,775 | {
"abstract": "Marinithermus hydrothermalis Sako et al . 2003 is the type species of the monotypic genus Marinithermus . M. hydrothermalis T1 T was the first isolate within the phylum “ Thermus-Deinococcus ” to exhibit optimal growth under a salinity equivalent to that of sea water and to have an absolute requirement for NaCl for growth. M. hydrothermalis T1 T is of interest because it may provide a new insight into the ecological significance of the aerobic, thermophilic decomposers in the circulation of organic compounds in deep-sea hydrothermal vent ecosystems. This is the first completed genome sequence of a member of the genus Marinithermus and the seventh sequence from the family Thermaceae . Here we describe the features of this organism, together with the complete genome sequence and annotation. The 2,269,167 bp long genome with its 2,251 protein-coding and 59 RNA genes is a part of the G enomic \n E ncyclopedia of \n B acteria and \n A rchaea project.",
"introduction": "Introduction Strain T1 T (= DSM 14884 = JCM 11576) is the type strain of the species M. hydrothermalis , which is the type species of the monotypic genus Marinithermus [ 1 , 2 ]. The genus name is derived from the Latin word 'marinus' meaning 'of the sea' and the latinized Greek word 'thermos' meaning 'hot', yielding the Neo-Latin word 'Marinithermus' meaning 'an organism living in hot marine places' [ 1 ]. The species epithet is derived from the Neo-Latin word 'hydrothermalis' (pertaining to a hydrothermal vent) [ 1 ]. Strain T1 T was isolated in November 2000 from the surface zone of a deep-sea hydrothermal vent chimney at Suiyo Seamount in the Izu-Bonin Arc, Japan, at a depth of 1,385 m [ 1 ]. M. hydrothermalis was the first isolate within the phylum “ Thermus-Deinococcus ” that grew optimally under a salinity equivalent to that of sea water [ 1 ]. The absolute requirement of NaCl for growth distinguishes M. hydrothermalis from members of the genera Thermus and Meiothermus [ 1 , 3 ]. No further isolates have been reported for M. hydrothermalis . Here we present a summary classification and a set of features for M. hydrothermalis T1 T , together with the description of the complete genomic sequencing and annotation."
} | 556 |
36794201 | PMC9906648 | pmc | 2,776 | {
"abstract": "In the quest for stimuli-responsive materials with specific, controllable functions, coacervate hydrogels have become a promising candidate, featuring sensitive responsiveness to environmental signals enabling control over sol–gel transitions. However, conventional coacervation-based materials are regulated by relatively non-specific signals, such as temperature, pH or salt concentration, which limits their possible applications. In this work, we constructed a coacervate hydrogel with a Michael addition-based chemical reaction network (CRN) as a platform, where the state of coacervate materials can be easily tuned by specific chemical signals. We designed a pyridine-based ABA triblock copolymer, whose quaternization can be regulated by an allyl acetate electrophile and an amine nucleophile, leading to gel construction and collapse in the presence of polyanions. Our coacervate gels showed not only highly tunable stiffness and gelation times, but excellent self-healing ability and injectability with different sized needles, and accelerated degradation resulting from chemical signal-induced coacervation disruption. This work is expected to be a first step in the realization of a new class of signal-responsive injectable materials.",
"conclusion": "Conclusions This work shows the development and potential application of a chemical signal responsive coacervate hydrogel based on a Michael addition-based CRN. Allyl acetate electrophiles can trigger the cationization of pyridine groups located in the A blocks of an ABA triblock copolymer. In the presence of a polyanion, the electrophile addition triggers the formation of coacervate gels. The gels can be disassociated to solutions by reaction with a competing nucleophile, regenerating the starting neutral pyridine-based polymers. Further electrophile additions can recover the gel state, allowing for repeated and reversible sol–gel transitions. Moreover, varying the electrophile dose can tune the mechanical strength and gelation time of the coacervate gels. We further demonstrated promising self-healing properties of these coacervate gels, with application as injectable materials. Importantly, in cell culture media-based (rich in amino acids) environments, gel degradation can be highly accelerated due to the nucleophile-responsiveness of the coacervate gels. Extrapolating these findings, the concept of nucleophile-responsive coacervate gels, being injectable and selectively degradable, shows potential for application as therapeutical injectable materials or degradable scaffold in 3D printing.",
"introduction": "Introduction Soft materials that can respond to signals from the environment, display promising applications in therapeutic delivery. 1–3 Programmable sol–gel transitions are required for many therapeutic applications, e.g. , where materials could be injected into the body as a gel, but then degrade in response to external triggers. 4,5 Among various materials which are broadly applied, hydrogels formed by strong electrostatic interactions (coacervation assembly), have been of great interest due to their inherent responsiveness to external signals. 6–8 These signal-sensitive (pH, temperature or salt) coacervate materials do, however, show limited responsiveness under physiological conditions. 9–12 To overcome the restrictions above, it is of great significance to develop materials which are responsive to specific signals, enabling precise regulation of material properties. Inspired by biological systems, where biochemical signals play an important role to control the growth and behavior of supramolecular structures, chemical reaction networks (CRNs) have been developed as a useful tool to control the properties of synthetic supramolecular materials. 13–17 Typically, non-interacting building blocks are activated into an assembling structure by an activator. 18,19 The initiated products are later deactivated by a competing reaction, leading to disassembly and recovery of the starting non-interacting building blocks. 20,21 Through delicate design, temporary sol–gel transition processes have been realized, regulated by methylating agents, 22,23 glucose, 24 enzymatic reaction, 25–27 and other chemical fuels. 28,29 Recently, CRNs have also been used to control coacervate assembly with applications in drug delivery, nanoreactors and protocell simulation. 30–33 These have typically been dispersed nano- and micro-scale assemblies and there are no studies on CRN-regulated coacervate assembly for macroscopic gels. In the current work, we start from a Michael addition-based CRN to create chemical signal responsive coacervate hydrogels. Specifically, allyl acetates (electrophile) act as an activator which can react with tertiary amines, yielding a cationic quaternary ammonium species. This species can subsequently react with a competing nucleophile (deactivator), re-forming the initial tertiary amine, completing the reaction cycle. 34 For example, we selected a pyridine-based polymer as the poly(tertiary amine), whose ionization could be reversibly regulated with the electrophiles (to ionize the polyamine) and competing nucleophiles (amines or thiols, to deionize the charged polyamine). Accordingly, the hydrophobicity and charge state of the pyridine-based polymer could be controlled. 35 With this strategy, we have controlled the assembly of hydrophobic surfactant micelles and complex coacervate-core micelles (C3Ms) as well as the degree of swelling in cross-linked gels. 35,36 At the base of the current work is the ABA triblock copolymer PVP1, which has pyridine groups incorporated into the A block that can be charged by treatment with an electrophile – methyl-2-(acetoxymethyl)acrylate (ME, activator). The copolymer can then revert to a neutral state after addition of a nucleophile – pyrrolidine (P, a secondary amine, deactivator). Importantly, the positively charged A blocks can interact with a polyanion, spontaneously forming a network of coacervate domains interconnected by the hydrophilic B block chains. Above a threshold polymer concentration, network formation leads to gelation. Addition of P can then revert the positively charged pyridinium units to their initial neutral state, destroying the gel. Gel properties like mechanical strength and gelation time are controllable by means of the amount of activator (ME). We then investigated the self-healing and injectability of these coacervate gels, which demonstrated their potential use as injectable materials. Moreover, our coacervate gels show selective nucleophile-responsiveness and therefore exhibit excellent degradability in cell culture media-based solutions or gels, acting as simulated interstitial fluid or tissue. We propose that these properties make CRN controlled coacervate hydrogels promising candidates for therapeutic biomedical materials. Design concept of chemical signal regulated coacervate hydrogels Inspired by reported polymer architectures, 37 where the designed ABA triblock copolymer usually requires each block in moderate length for adequate interaction sites and moderate water-solubility, we synthesized the ABA triblock copolymer PVP1 ( Scheme 1A ) by a two-step RAFT polymerization using a symmetrical RAFT agent. A flexible water-soluble DMA ( N , N -dimethylacrylamide) ‘B’ block of 410 units was prepared, followed by symmetrical extension with a random mixture 17 DMA and 41 VP (4-vinyl pyridine) units on either side as the ‘A’ blocks (further details and characterization were shown in Fig. S1–S2, S4–S5, S7 and Tables S1, S3 † ). Herein, DMA units in the ‘A’ blocks serve to weaken hydrophobic interactions and suppress gel formation in the neutral state. The pendent polyvinyl-pyridine units can be activated to quaternary pyridinium groups by ME, forming the cationic copolymer PVP1 + can interact with anionic polymer PAMPS 236 (poly(2-acrylamido-2-methylpropane sulfonic acid sodium salt)), forming coacervate domains. 36 When the polymers are of sufficient concentration, these domains are linked through the hydrophilic ‘B’ blocks, forming a coacervate gel network. Upon reaction with P, PVP1 + returns to its neutral state (PVP1), thus the electrostatic interactions disappear and the gel disassociates to a solution ( Scheme 1B ). Scheme 1 (A) Chemical structures of polymers and reagents involved. (B) Scheme of the responsive coacervate hydrogel formation regulated by chemical signals: ME (activator, an electrophile) activates the pyridine groups (on the polymer chain) from neutral to positive charged state, which can interact with anionic polymer PAMPS 236 , leading to a coacervate gel. With addition of P (deactivator, a nucleophile), the charged pyridine groups are deactivated, leading to a gel to solution transition.",
"discussion": "Results and discussion Chemical reaction process and assembly mechanism We began by studying the (de)ionization processes of our system in solution state based on NMR tests (Fig. S8 † ). Since the p K a of PVP1 is 4.1 (Fig. S9 † ), we performed the experiments with 0.5 wt% PVP1 in phosphate buffer (100 mM, pH 7.4), where the polymer is in a neutral state. After adding 1.0 eq. ME into the PVP1 solution at t = 0 h, we noticed that the conversion (from PVP1 to PVP1 + ) increased rapidly to approximately 75% in the first 3 hours and reached to its peak (78%) at t = 5 h (see Fig. 1A and B ). Next, we added 1.0 eq. P into the solution at t = 30 h, where we observed a dramatic recovery of PVP1 from PVP1 + over the following 2 hours (from 78% to 20%). After a further 3 hours, the PVP1 + contents stabilized at 10%. From the NMR data for the reaction VP + ME ⇄ VP + + acetate with an equilibrium conversion as 78%, the K eq for the reaction can be calculated as [VP + ][acetate]/[VP][ME] = 0.78 × 0.78/0.22 × 0.22 = 13. This indicates that when an equivalent amount of ME is used (ME/VP = 1), 22% of the pyridine groups remain uncharged. With this K eq , increasing the amount of ME will lead to an increase in the ratio of charged over uncharged pyridines, thereby regulating the strength of polyelectrolyte complexation. This provides the possibility to modulate the material states by adjusting polymer concentration and quaternization degree. Fig. 1 (A) Scheme of chemical reaction network (CRN) for reversible cationization of pyridine groups. Specifically, ME can induce the ionization of tertiary amines via a nucleophilic substitution reaction, yielding the cationic amines and acetate. The starting neutral tertiary amines can then be regenerated by addition of a competing nucleophile P followed by the production of waste. (B) Reversible formation of PVP1 + following sequential 1.0 eq. additions of ME and P to PVP1 (0.5 wt% PVP1, in 100 mM pH 7.4 phosphate buffer). The reaction cycle was also studied by dynamic light scattering for 2-cycles of 1.0 eq. ME and P sequential additions followed with (C) scatter count and (D) Z-average diameter (0.1 wt% PVP1 and 1.0 eq. PAMPS 236 , in 100 mM pH 7.4 phosphate buffer). After NMR experiments demonstrated signal responsive control over the ionization state of PVP1, we next investigated whether this induced polycation could be used to trigger coacervate formation with an anionic polymer. For this we synthesized an AMPS (2-acrylamido-2-methylpropane sulfonic acid sodium salt) polymer of approximately 236 units (PAMPS 236 , for synthesis details see Fig. S3, S6 and Tables S2, S3 † ). We studied their interactions via dynamic light scattering (DLS) at a relatively low PVP1 concentration (0.1 wt% PVP1, in 100 mM pH 7.4 phosphate buffer) in the presence of 1.0 eq. PAMPS 236 (meaning the ionic molarity for AMPS and VP are equal). Fig. 1C and D show that, although the neutral PVP1 cannot interact with anionic PAMPS 236 , it can still form micellar structures by itself in an aqueous environment based on hydrophobic interactions (the presence of PAMPS 236 did not influence the result, see Fig. S11A and B † ). Thus, the initial system had a high scatter count and large Z-average size (15 Mcps, 175 nm). Charging the pyridine groups with 1.0 eq. ME led to disassociation of hydrophobic micelles. As a result, in first 2 hours, both the scatter count and size decreased. Interestingly, this was followed by an increase in both parameters for the next 11 hours, reaching stable values around 12 Mcps, 150 nm. This implies that once sufficient ionization of PVP1 + is achieved, its coacervate interactions with PAMPS 236 start to dominate, leading to the formation of coacervate micellar structures (for the dilute 0.1 wt% PVP1 case). Then, 1.0 eq. P was added to neutralize the charged pyridine groups, triggering a reduction in the scatter count and size of our system due to dissociation of the coacervate micelles. Interestingly, the scatter count and corresponding particle size stabilized here (1.5 Mcps and 50 nm) and did not return to the values observed before ME addition. This implies that the hydrophobic association is not reversible in our system, probably because of a small amount of residual PVP1 + . A second cycle was then conducted which demonstrated the reversibility of the system, with 1.0 eq. ME triggering the appearance of coacervate micelles and 1.0 eq. P destroying them. Since there was no hydrophobic micelles disassembly process during the second cycle, the time for reaching the maximum count rate (∼7 hours) was much shorter than in the first cycle (∼13 hours). Promisingly, the peak values for light scatter intensity and particle size (13 Mcps, 145 nm) were approximately the same as in the first cycle. To further study why the hydrophobic micelles did not form again, we repeated the experiment without PAMPS 236 . At the start of the process, again hydrophobic micelles of similar size and scatter count as the system with PAMPS 236 were observed. Addition of 1.0 eq. ME similarly decreased these values, indicating micelle disassociation (Fig. S11A and B † ). After this point, further additions of P were unable to recover the hydrophobic micelles, even when treated with 3.0 eq. excess of P (Fig. S12A and B † ). Based on these experiments, we speculate that the incomplete recovery of PVP1 from PVP1 + (as indicated in NMR experiments) leads to weaker hydrophobicity of recovered PVP1, blocking the re-formation of hydrophobic micelles. At the same time, the presence of a small amount of residual cationic pyridine groups after P addition appears to be insufficient to allow coacervate formation, opening the door to repeated transitions between coacervate states. Coacervate hydrogels formation controlled by chemical signals We next sought to use such coacervate assembly to construct macroscopic hydrogels and study their chemical signal-triggered sol–gel transitions. Herein, 5 wt% PVP1 and 1.0 eq. PAMPS 236 were dissolved in 100 mM pH 7.4 phosphate buffer. Then, the solution was mixed with 1.0 eq. ME, and a gel formed in 30 min ( Fig. 2A , for detailed gel preparation see ESI † ). With addition of 1.0 eq. P, the obtained gel can revert to a solution in 5 min with similar rheological properties as the starting solution ( Fig. 2A and B ). To clearly illustrate the (de)generation of the coacervate gel, we conducted time sweep rheological measurements. As shown in Fig. 2B , at the beginning the storage modulus ( G ′) of the polymer solution was much lower than loss modulus ( G ′′), indicating a solution state. After treatment with 1.0 eq. ME, both G ′ and G ′′ increased rapidly. Within 30 min, G ′ surpassed G ′′, suggesting the gelation of polymer solutions. After further 60 min, G ′ stabilized at about 1000 Pa, showing a stable gel state. Then, 1.0 eq. P was added to neutralize the charged pyridine groups, which should lead to disassociation of the coacervation domains in the gel. After 3 min shaking, the gel converted back to solution ( G ′′ > G ′) (shaking does not induce sol–gel transitions in this system, see Fig. S13 † ). To demonstrate the reversibility of the chemical signal-induced coacervate gel, we completed two more cycles with sequential additions of 1.0 eq. ME and 1.0 eq. P. Note that the gelation time for cycles 2 and 3 were about 15 min (much shorter than the gelation time (∼30 min) of cycle 1), which is in good agreement with our DLS test results. Moreover, this sol–gel cycle can bear at least 8 cycles, otherwise the materials will keep as a solution state, possibly due to the accumulation of reagents waste. Fig. 2 Chemical signal-programmed sol–gel transition process. (A) Photographs showing the gelation of PVP1 solution (in the presence of 1.0 eq. PAMPS 236 ) driven by 1.0 eq. ME and the disassembly of coacervate gels induced by 1.0 eq. P. Rhodamine B was added for coloring (1 μM). Mixing the PVP1 solution with 1.0 eq. ME at 0 min led to gelation after 30 min. After adding 1.0 eq. P, the formed coacervate gel converted to the solution state in 5 min. (B) Time sweep rheological tests of the ME-driven gelation and P-induced disassembly process of the coacervate gels (3 cycles). (C) Time sweep rheological tests monitoring autonomous recovery of coacervate gels (triggered by 4.0 eq. ME) with re-additions of 1.0 eq. P. For the time sweep oscillatory tests, the strain ( γ ) and frequency ( ω ) were set as 5% and 10 rad s −1 , respectively. All samples: 5 wt% PVP1 and 1.0 eq. PAMPS 236 (overall polymer content: 6.8 wt%), in 100 mM pH 7.4 phosphate buffer. As a further test, we tried to achieve an ‘autonomous recovery’ by preparing a gel with excess ME (4.0 eq.) and using successive 1.0 eq. additions of P to transiently modulate gel properties. We observed that the gelation of polymer solutions (triggered by 4.0 eq. ME) quickly happened in 10 min (see Fig. 2C ), which is much faster than the 1.0 eq. ME-triggered gelation. It seems plausible that a larger amount of ME leads to a higher ionization rate for PVP1, resulting in faster gelation. After the storage modulus stabilized around 1000 Pa, 1.0 eq. P was added to weaken the gel (as there is still some excess ME inside, 1.0 eq. P cannot fully convert the gel to a solution). Immediately after adding P, the moduli dropped sharply, but they then recovered to the starting value in minutes. Apparently, the added P initially removes (part of) the charges, but reaction with the remaining excess of ME restores these charges nearly entirely. In the second recovery cycle, an additional 1.0 more eq. P was placed onto the gel surface. Although this time there was a recovery process for gel mechanical strength, the maximum attained storage modulus (∼300 Pa) was lower than the first cycle (∼900 Pa). Furthermore, we noticed that after the third P addition, a solution formed that was unable to recover to a gel. We believe this is due to insufficient amount of remaining ME to achieve the required ionization on PVP1 for coacervate gel formation. Overall, the experiments illustrate how sol–gel transitions can be achieved by modulating coacervate formation through reaction with an activator electrophile (ME) and a deactivator secondary amine (P). We found that the ‘forward’ reaction to PVP1 + is slower than the reverse reaction. In addition, with excess ME in the solution, a non-monotonic autonomous recovery, namely an initial drop in stiffness followed by a recovery, can be realized. Effects of chemical signals on the mechanical performance of coacervate hydrogels and diagram of material states Next, we characterized the impact of different ME equivalents on the mechanical properties and gelation times of the coacervate gels. For this purpose, we first prepared the polymer solutions (5 wt% PVP1 and 1.0 eq. PAMPS 236 , in 100 mM pH 7.4 phosphate buffer) and activated them with different ME equivalents. The gelation time and mechanical strength of gels were determined by oscillatory rheological measurements (Fig. S14 † ). The results shown in Fig. 3A suggest that the maximum storage modulus of coacervate gels increased from ∼25 to ∼1000 Pa with increasing ME equivalents (from 0.5 to 1.0 eq.). Clearly, a larger ratio of ME to pyridine groups leads to more charged pyridine groups, leading to stronger coacervation interactions (higher ). However, further increasing the ME equivalents beyond 1.0 eq. does not unlimitedly lead to enhanced mechanical properties such as G ′ (Fig. S14 † ), likely due to reaching a plateau in pyridinium formation on the copolymers. Moreover, the gelation time (first time point where G ′ > G ′′) can also be regulated by varying ME equivalents. As shown in Fig. 3B , we found that the gelation time decreased from ∼1.5 h to ∼0.3 h with increasing ME equivalents (from 0.5 to 1.5 eq.). This indicates that higher ME concentration increases the ionization reaction rate and therefore shortens the gelation time of polymer solutions. Fig. 3 Effect of ME equivalents on coacervate gelation process. (A) Maximum storage modulus and (B) gelation time of the coacervate hydrogels as a function of ME equivalents. Error bars in (A) and (B) are calculated from three independent measurements. All samples: 5 wt% PVP1 and 1.0 eq. PAMPS 236 (overall polymer content: 6.8 wt%), in 100 mM pH 7.4 phosphate buffer. Note that a sol–gel mixture (5 wt% PVP1, 0.5 eq. ME) naturally gives variable results. In three measurements, we found one gel-like curve (from which we got one crossover point/gelation time, the open circle) and two solution-like curves. (C) Diagram of material states by varying both PVP1 concentrations and ME equivalents. The diagram is a compilation of the results from rheological measurements and visual observation (red circle and line – hydrophobic micelle gel state; blue circles and area – coacervate gel state; orange square – sol–gel mixture state; white area – potential sol–gel intermediate state; and green triangles and area – solution state). For 1 wt%, 2 wt%, 5 wt% and 10 wt% samples (overall polymer content: 1.4 wt%, 2.7 wt%, 6.8 wt% and 13.6 wt%): PVP1 and 1.0 eq. PAMPS 236 , in 100 mM pH 7.4 phosphate buffer; for 15 wt% and 20 wt% samples: PVP1, in 100 mM pH 7.4 phosphate buffer. To show how chemical signals may regulate these coacervate materials, we investigated the relationship between PVP1 concentrations, ME equivalents and material states (see Fig. 3C ). We firstly prepared polymer solutions of different concentrations (1 wt%, 2 wt%, 5 wt%, 10 wt% PVP1 with 1.0 eq. PAMPS 236 , 15 wt% and 20 wt% PVP1) and initiated their gelation by addition of different ME equivalents. All samples were characterized by the vial inversion method and frequency sweep rheology experiments to identify the material states (solution or gel) (see Fig. S15 and S16 † ). We found that 20 wt% PVP1 can form a hydrophobic micelle gel by itself. In agreement with earlier dilute solution DLS analysis, these hydrophobic micelle gels showed poor reversibility, even after we attempted recovery using excess P (3.0 eq.) (see Fig. S11C, S12C and D † ). The samples with ≤1 wt% PVP1 never formed a gel with ME (even with >1.0 eq. ME, see Fig. S15A and S16 † ). Samples with a PVP1 concentration between 1 wt% and 10 wt% were able to form coacervate gels, with less ME required for gel formation as polymer concentration increased (1.0 eq. ME for 2 wt% PVP1 and 0.5 eq. ME for 10 wt% PVP1). Interestingly, for 5 wt% PVP1 treated with 0.5 eq. ME (in the presence of 1.0 eq. PAMPS 236 ), we observed the intermediate case of a sol–gel mixture. It seems likely that such intermediate states can also occur at other combinations of PVP1 concentrations and ME equivalents, just below a threshold. Based on these observations, a ‘diagram of material states’ can be constructed ( Fig. 3C ). To clarify the diagram, datapoints referring to different states were given different colors. In the diagram, the red circle and line represents the material states which can assemble into hydrophobic micelle gels by themselves without any reagents added. The blue circles and area represent the coacervate gel state, where the gelation of polymer solutions can be activated by ME. The green triangles and area represent the liquid state, in which ME cannot trigger the gelation owing to insufficient polymer concentration. For the white area (containing the orange square) between the green and blue areas, we defined it as potential sol–gel intermediate state. Combined, the diagram clearly shows that as polymer concentration increases, the critical degree of PVP1 ionization (or eq. ME added) required to form a gel decreases. Self-healing and injectability of coacervate gels We then took the 5 wt% coacervate gels (initiated by 1.0 eq. ME) as a basis to investigate their self-healing properties. Strain sweep rheological tests showed that the gel networks are temporarily disrupted upon application of over 200% strain (see Fig. 4A ). In the fixed-frequency measurement (with repeated strain), the gels were subjected to cycles of 5% and 500% strain. At 500% strain, G ′ decreased from 800 Pa to 80 Pa and G ′′ from 300 Pa to 200 Pa ( Fig. 4B ). In other words, under a large amplitude oscillatory force, G ′′ was higher than G ′, indicating the disruption of gel networks and conversion to a viscous fluid. When the strain returned to 5%, G ′ and G ′′ rapidly recovered to starting value (800 Pa and 300 Pa, respectively), suggesting the recovery of coacervate gels. Moreover, the gels can be subjected to at least 5 cycles without any loss of mechanical strength. In macroscopic self-healing experiments ( Fig. 4C ), we prepared two gels (one gel was colored by rhodamine B, and another one was non-dyed) and hand-pressed them together into a whole gel. After 5 min, the two separate gels reconnected and could be easily lifted by a tweezer. After 2 further days, the dye diffused into the non-dyed section, demonstrating the initially separate gels have completely self-healed to an integrated gel. These results provide clear evidence that these coacervate gels are capable of self-healing, as expected for gels with reversible, physical bonds. Fig. 4 Self-healing and injectability of coacervate gels. (A) Strain sweep and of the coacervate hydrogels. Strain sweep measurements were performed from 0.1% to 1000% at fixed frequency ( ω = 10 rad s −1 ). (B) Fixed-frequency ( ω = 10 rad s −1 ) measurement with repeated strain jumps from 5% to 500% and back. (C) Self-healing of coacervate hydrogels. Two separate gels (one gel was dyed by rhodamine B, another gel was non-dyed) were pressed together. After 5 min, the two gels had become a single gel; after 2 days, the dye diffused from the stained gel to non-stained gel. (D) Coacervate hydrogels injection in air or into a PAAm gel (the coacervate gels can be extruded from 20G, 21G and 26G needles by hand-pressing). The injected gels (which were dyed by rhodamine B) allowed for printing ‘TUD’ letters (‘T’, ‘U’ and ‘D’ were extruded from 20G, 21G and 26G needles, respectively). Also, the coacervate gels (in red circles, non-dyed gels) can be injected into a PAAm gel via a 26G needle. All samples: 5 wt% PVP1 and 1.0 eq. PAMPS 236 (overall polymer content: 6.8 wt%), in 100 mM pH 7.4 phosphate buffer (initiated by 1.0 eq. ME). Since the gels were self-healable and responded quickly to the strain changes, we investigated whether our coacervate gels were also injectable. 5,38–40 As shown in Fig. 4D , a rhodamine B dyed gel was prepared and put into a syringe, then we extruded it from needles with various diameters (20G – 0.9 mm, 21G – 0.8 mm and 26G – 0.45 mm). Using hand pressure, these gels could be extruded through all investigated needles; after extruding from the needles (the high shear force was removed), the gel state was recovered instantly. As a proof of injectability, we wrote three letters (‘T’, ‘U’ and ‘D’) through different sized needles ( Fig. 4D ), a video showing the gel injection can be found as Movie S1. † In future therapeutic applications, such gels would have to injected into various tissues. Thus, we synthesized a polyacrylamide (PAAm) gel as a simulated tissue (the synthesis procedure of the PAAm gel can be found in the ESI † ). With a 26G needle, these gels can be easily injected into the simulated tissue (Movie S2 † ), leading to gel-in-gel hybrid materials. Accelerated degradation of coacervate gels in cell culture media-based environment One important requirement for injectable materials is whether they can degrade in human tissue or interstitial fluid. Conventional coacervate hydrogels are rendered degradable because of their salt, pH or temperature responsiveness, however the degradation time can be exceedingly long under physiological conditions. 9,11,12,39,41 Previous work in our group on related pyridine-based polymeric systems has shown that (in addition to secondary amines) primary amines and thiols can be used to trigger deionization. 35,36 Treating our coacervate gels with two representative chemical signals – glycine (Gly, a primary amine) and sodium-3-mercaptopropane-sulfonate (SH-Na, a thiol), we indeed observed gel–sol transitions at a similar rate to that achieved with P (within 5 min). Interestingly, the recovered loss modulus depended on the nucleophilicity of the applied nucleophile, 42,43 with stronger nucleophiles giving a lower recovered loss modulus (SH-Na ∼0.3 Pa, P ∼2 Pa, Gly ∼4 Pa) (Fig. S17 † ). A stronger nucleophile implies a larger reaction free energy, and, hence, a larger conversion of cationic into neutral pyridine groups. This means that after treatment with a stronger nucleophile, a smaller number of physical bonds would have to be broken under flow, observed as solutions with lower viscosity (lower G ′′). Since the disassembly of these coacervate gels is responsive to these chemical signals, we anticipate that the degradation for such gels could be accelerated by physiological signals such as amino acids and thiols present in the cytosol or extracellular fluids. We first selected a cell culture medium (Ham's-F12, no phenol red, CM) as the simulated amino acids-rich interstitial fluid, which was anticipated to accelerate gel degradation. 150 mM phosphate buffer (PB) was used to mimic the physiological salt environment without nucleophiles present, isolating the effects of high salt concentration (which is known to degrade non-responsive coacervate gels). 6,9,11,44,45 Pure water was applied as a blank culture environment (see Fig. S18 and Table S4 † ). At 25 °C, the coacervate gels completely disappeared within 4 h when incubated in CM, while those in PB or water did not disassemble even after 24 hours ( Fig. 5A ). When the degradation temperature was elevated to 37 °C, the coacervate gels in CM degraded in 2 h, again significantly faster than those in PB (24 h) and water (gel fragments still observable after 24 hours, Fig. 5B ). These results suggest that the gels are sensitive to nucleophilic biological signals but not to physiological salt or temperature in liquid environments. Fig. 5 Gel degradation tests in various liquid environments. Degradation of the coacervate gels in water, or 150 mM pH 7.4 phosphate buffer (PB) or cell culture medium (CM) at (A) 25 °C and (B) 37 °C over time. All samples: 5 wt% PVP1 and 1.0 eq. PAMPS 236 (overall polymer content: 6.8 wt%), in 100 mM pH 7.4 phosphate buffer (initiated by 1.0 eq. ME). We then evaluated cell cytotoxicity of these gels by a viability test (CCK-8 assay) on NIH-3T3 mouse fibroblast cells (Fig. S19A † ). These tests showed that the ME-activated gels show negligible toxicity at low concentration (gel concentration ≤1000 μg mL −1 ) but substantial cytotoxicity at high concentration (gel concentration >1000 μg mL −1 ). To investigate the source of toxicity within the ME-activated gels, we also evaluated the gel components separately. This showed that PAMPS 236 and PVP1 are non-toxic even at high concentrations (viability above 80% up to 350 μg mL −1 and above 95% up to 500 μg mL −1 polymer concentration, respectively), while PVP1 + and ME decreased cell viability below 65% at 250 and 10 μg mL −1 , respectively (Fig. S19B–E † ). To increase biocompatibility, we replaced ME with a more biocompatible electrophile, 35,36 for which we have found diethyl(α-acetoxymethyl) vinylphosphonate (DVP) retains ∼80% cell viability at high concentrations (236 μg mL −1 , Fig. S20A † ). Gels prepared using DVP were found to be significantly less toxic, with cell viability remaining above 85% even at high gel concentrations (10 mg mL −1 , Fig. S20B † ). Alternatively, the hydrogel may be applied as degradable scaffold or mold in 3D printing, because of its proven extrudability and controllable degradability. 46 For example, the coacervate gel could be printed as a mold or scaffold to support the printing of a second material, after which the mold or scaffold can be easily removed by immersing it in an amine–rich medium. Overall, we conclude that this chemical-signal-regulated coacervate hydrogel could be used as injectable material (preferably with DVP as the activator for obtaining a more biocompatible material) or as scaffold in 3D printing. In future in vivo applications, such gel materials can be introduced in the body by intradermal, subcutaneous or intramuscular injection, which is a reason to examine the gel degradation process in viscoelastic solid environments. Hence, we used swollen PAAm gels as tissue mimics. In brief, PAAm hydrogels were swollen in different media (water, PB and CM) for 18 hours, allowing for solvent exchange (further details shown in ESI and Table S5 † ). Frequency sweep rheological measurements on these swollen PAAm gels showed that they possess similar viscoelastic properties as many tissues (500 Pa < G ′, G ′′ < 10 000 Pa, see Fig. S21 † ). 47 The coacervate gels were then injected into the different swollen PAAm gels and kept at 25 °C or 37 °C, and photos of gel degradation were taken over time (Fig. S22 and S23 † ). At 25 °C, the coacervate gels decomposed completely in CM-swollen PAAm gels within 2 h ( Fig. 6A ). In contrast, the coacervate gels were still observable in PB-swollen PAAm gels even after 48 hours (blue pixels retain 0.6%, Fig. S22 and S24 † ). At 37 °C, coacervate gels that were injected into a CM-swollen PAAm gel degraded completely within 1 h ( Fig. 6B ). In PB-swollen PAAm gels the coacervate gels completely disappeared in 36 h (Fig. S23 † ). However, in water swollen PAAm gels, there was almost no observable degradation within 24 h at 25 °C or 37 °C (blue pixels retain 8.1% and 11.7%, respectively; Fig. S24 † ), and even after 36 hours the coacervate gel residues can still be observed (blue pixels retain 7.4% and 7.8%, respectively; Fig. S23 and S24 † ). Fig. 6 Degradation tests in the swollen PAAm gels. Degradation of the coacervate gels in water, or 150 mM pH 7.4 phosphate buffer (PB) or cell culture media (CM) swollen PAAm gels at (A) 25 °C and (B) 37 °C over time. Red arrows are added to point out the gel residue. (C) Degradation time (time needed for no visible gel remaining or below 0.5% blue pixels remaining, see ESI † ) of coacervate gels (in the swollen PAAm gels). All samples: 5 wt% PVP1 and 1.0 eq. PAMPS 236 (overall polymer content: 6.8 wt%), in 100 mM pH 7.4 phosphate buffer (initiated by 1.0 eq. ME). On the basis of these results, we concluded that the degradation of coacervate gels can be highly accelerated in both simulated interstitial fluid and simulated tissue. While relatively insensitive to the non-specific signals of physiological temperature (37 °C) and salt concentration (150 mM), gel degradation is then triggered by specific nucleophilic biochemical signals, enabling a tissue-responsive degradation process while maintaining gel integrity ex vivo."
} | 8,943 |
36498973 | PMC9741481 | pmc | 2,777 | {
"abstract": "Marine biofouling is a natural process often associated with biofilm formation on submerged surfaces, creating a massive economic and ecological burden. Although several antifouling paints have been used to prevent biofouling, growing ecological concerns emphasize the need to develop new and environmentally friendly antifouling approaches such as bio-based coatings. Chitosan (CS) is a natural polymer that has been widely used due to its outstanding biological properties, including non-toxicity and antimicrobial activity. This work aims to produce and characterize poly (lactic acid) (PLA)-CS surfaces with CS of different molecular weight (Mw) at different concentrations for application in marine paints. Loligo opalescens pens, a waste from the fishery industry, were used as a CS source. The antimicrobial activity of the CS and CS-functionalized surfaces was assessed against Cobetia marina , a model proteobacterium for marine biofouling. Results demonstrate that CS targets the bacterial cell membrane, and PLA-CS surfaces were able to reduce the number of culturable cells up to 68% compared to control, with this activity dependent on CS Mw. The antifouling performance was corroborated by Optical Coherence Tomography since PLA-CS surfaces reduced the biofilm thickness by up to 36%, as well as the percentage and size of biofilm empty spaces. Overall, CS coatings showed to be a promising approach to reducing biofouling in marine environments mimicked in this work, contributing to the valorization of fishing waste and encouraging further research on this topic.",
"conclusion": "5. Conclusions In this study, the long-term antimicrobial and antifouling performance of CS-based surfaces against C. marina biofilms under conditions that mimic some marine environments were demonstrated. Although the Mw and concentration of chitosan did not impact the characteristics of the produced surfaces, the most effective antibiofilm surfaces were those coated with the lowest Mw CS, regardless of the concentration. The antimicrobial activity of the CS studied in this work was demonstrated to be linked to cell membrane depolarization with consequent loss of membrane integrity. The results obtained in this study suggest that the incorporation of CS in marine paints may be a promising eco-friendly antifouling approach to reduce the biofilm formation on ship hulls and consequently fight biofouling in this environment.",
"introduction": "1. Introduction Marine biofouling is a spontaneous and complex process by which natural and artificial submerged structures are colonized by marine organisms [ 1 ]. This undesirable attachment of molecules and fouling organisms has been recognized as a concern in the marine industry since it is responsible for several economic, industrial, environmental, and health-related implications [ 2 , 3 ]. The presence of organisms on marine vessels increases the weight of ships and their drag resistance, resulting in higher fuel consumption and environmental pollution [ 4 ]. Moreover, biofouling changes the physicochemical properties of marine surfaces, promoting their fast deterioration and corrosion, and can also contribute to species invasion, causing negative effects on global biodiversity [ 5 ]. Marine biofouling can also affect partially submerged equipment used for monitoring dissolved oxygen, turbidity, and pH, resulting in incorrect measurements [ 6 ]. All these consequences create a massive economic and ecological burden, stressing the need to develop new approaches to protect submerged surfaces from biofouling organisms. Biofouling in the marine environment is a dynamic process that usually involves three steps: conditioning film formation, microfouling, and macrofouling [ 7 ]. Conditioning film is formed by the adsorption of organic molecules on submerged surfaces and promotes microfoulers (e.g., bacteria and diatoms) adhesion and consequently biofilm formation (microfouling). Biofilms established on surfaces promote the settlement of macrofoulers (such as sponges, mussels, and algae) and, within days to weeks, macrofouling communities are completely established over the submerged surfaces [ 8 ]. Since biofilm formation is one of the first steps of this natural process, a potential strategy to delay macrofouling is to prevent adhesion and biofilm formation by marine bacteria, which are early marine surface colonizers. To date, several antifouling paints have been used to prevent biofouling on ship hulls, mainly by the gradual erosion and release of biocides and toxic chemicals [ 9 ]. However, as these antifouling agents can persist in the environment and pose a threat to marine organisms, the International Maritime Organization (IMO) has banned their use in the production of antifouling paints [ 10 , 11 ]. Therefore, the development of novel, non-toxic and eco-friendly approaches to prevent marine biofouling in ship hulls, such as bio-based coatings, is urgently required. Among different biopolymers, chitosan (CS) has received significant attention from academia and industry for its many applications. CS is a cationic polysaccharide obtained by the deacetylation of chitin, which is the second most abundant polymer on Earth and is commonly sourced from crustacean shells, mollusks, insects, and fungi [ 12 , 13 ]. The use of chitin and CS can be advantageous in solving some environmental problems, and in the last few decades, squid pens have been increasingly explored as a source of chitin. Since the average yield of edible flesh in squid is around 70% [ 14 ], squid processing produces a substantial amount of waste that ranges from 0.8 to 1.6 million tonnes per year [ 15 ]. Therefore, to avoid the costly disposal of this waste and enhance the potential of chitin and CS valorization, integration into a biorefinery and a circular economy strategy were suggested. These aim to benefit both the economy and the environment through the sustainable conversion of chitin and CS into nitrogen-rich chemicals for various applications (e.g., pharmaceuticals, cosmetics, and water treatment) [ 16 ]. Besides the use of CS enabling the valorization of fish processing industry discards, CS has been widely used due to its interesting intrinsic properties including non-toxicity, biocompatibility, film-forming ability, chemical stability, low cost, and antimicrobial activity against a broad spectrum of microorganisms [ 17 ]. Although the CS mechanism of action is not entirely known, three main mechanisms have been proposed for the inhibition of microbial growth: (i) cell membrane disruption, as a result of electrostatic interactions between the positively charged CS molecules and the negatively charged cell membranes, which can lead to loss of intracellular content and cell death [ 18 , 19 ]; (ii) inhibition of protein synthesis that can occur when CS molecules penetrate microbial cells, complex with DNA and inhibit mRNA synthesis [ 20 ]; and (iii) chelation of CS molecules with some metals ions, which damages the microorganism cell wall [ 18 , 21 , 22 ]. Although the antifouling activity of CS-based coatings has been reported by our group for other applications (food packaging and medical settings) [ 23 , 24 ] and short-term applications in the marine field [ 25 ], in vitro studies to test the long-term performance of CS coatings under operational conditions that simulate marine environments remain scarce. In addition, since the antimicrobial activity of CS and its derivatives depends on a set of structural properties such as molecular weight (Mw), degree of deacetylation (DD), concentration, and source [ 20 , 26 ], studies based on the physical and chemical properties of CS and their influence on biofilm formation are required for the development of more effective antifouling surfaces. The present study aims to (i) produce and characterize poly (lactic acid) (PLA) surfaces coated with CS of different Mw and concentrations obtained from the Loligo opalescens pen, and (ii) evaluate the antifouling activity of these surfaces against Cobetia marina biofilm formation. Besides, the mechanism of action of this type of CS was clarified. To the best of our knowledge, this is the first study that encompasses the crucial steps for the synthesis and characterization of PLA-CS films for application on marine surfaces, with CS recovered from marine by-products. Moreover, this is the first study that reveals the potential of PLA-CS surfaces to reduce C. marina fouling on underwater surfaces under nutritional conditions, temperature, and hydrodynamics that mimics the conditions typically found in marine environments. C. marina DSMZ 4741 is a ubiquitous bacterium isolated from coastal seawater [ 27 ] and was chosen as a microfouler model [ 28 ]. Considering the goal of developing antifouling paints for ship hulls, PLA was the substrate chosen for this proof of concept since it has been used in several environmental-friendly antifouling approaches, including the production of marine coatings [ 29 , 30 , 31 ]. Furthermore, it is described that PLA does not biodegrade in normal ambient conditions or marine environments, and offers mechanical stability, with no changes in mechanical properties after submersion tests in the sea [ 32 ].",
"discussion": "3. Discussion Given the economic and ecological effects of marine biofouling, the development of antifouling strategies for marine environments is imperative. Although some antifouling paints, such as biocide-containing paints, have been used to reduce the propensity of biofouling, the rigid international regulations and environmental concerns call for sustainable and environmentally friendly antifouling approaches, such as bio-based coatings [ 35 , 36 ]. In this study, CS-based surfaces with different concentrations and Mw were produced and characterized, and their long-term performance in preventing biofilm formation of C. marina was evaluated through an analysis of biofilm cell amount and architecture. In order to increase the predictive value of this work, this analysis was performed under laboratory conditions that mimic real marine environments. C. marina biofilms were developed at 25 °C for 49 days [ 37 ] under a shear rate of 40 s −1 , close to the shear rate reported for a ship hull anchored in a port (50 s −1 ) [ 38 ]. In a previous work of the group, by testing an innovative multifunctional coating [ 39 ], this methodology was shown to provide similar results when compared to surface immersion in a real marine environment for 2.5 years. The first part of this study consists of extracting chitin from by-products of the fishery industry and producing native CS (Mw of 294 kDa) and its derivates with different Mw (CS1, CS2, and CS3 with 186, 129, and 61 kDa, respectively). Although CS is commercially available, its extraction from L. opalescens pens enables the valorization of the fish processing industry discards and is an economically and environmentally sustainable strategy. This species of squid, together with Illex argentinus , is the most captured around the world [ 40 , 41 ]. To the best of our knowledge, this is the first work dealing with the application of β-chitosan isolated from squid pen against marine bacteria. Moreover, most of the chitin commercially available is in the form of α-chitin, which can be extracted from crustaceans shells and is characterized by its antiparallel polymeric chains [ 12 ]. Conversely, β-chitin isolated from squid pens has parallel polymer chains connected by hydrogen bonds which, due to its alignment, create inter- and intra-molecular forces weaker than those found in α-chitin, increasing its water-absorbing capacity and its solubility [ 40 ]. Considering that β-chitosan isolated from squid pen was tested against C. marina for the first time, we sought to clarify its mechanism of action. Flow cytometric analysis indicates that β-CS targets the bacterial cell membrane inducing its depolarization and pore formation. Since the β-CS effect was more measurable at the membrane potential level than at the membrane integrity level after 24 h exposure, these two events likely occur in cascade. Moreover, the β-CS mode of action was independent of the tested concentrations. In fact, several authors have postulated that CS disrupts cell membranes as a result of electrostatic interactions between the positively charged CS molecules and the negatively charged cell membranes, leading to loss of intracellular content and cell death [ 18 , 19 , 22 , 23 , 24 , 42 ]. After the functionalization of the PLA-CS surfaces by dip coating, they were characterized regarding their wettability and roughness as these properties can affect their antimicrobial activity [ 43 , 44 ]. Water contact angles reveal that all surfaces exhibited a hydrophilic behaviour, and the CS immobilization decreased the water contact values. Similarly, previous studies demonstrate that the incorporation of CS molecules increased the hydrophilicity of the surfaces due to the hydrophilic nature of the polymer and the increase of polar groups in the coatings [ 45 , 46 , 47 , 48 ]. Since bacterial adhesion is favored by the hydrophobic character of surfaces, the immobilization of CS on PLA films may reduce microbial attachment and subsequent biofilm formation [ 49 ]. Additionally, the wettability of the PLA-CS surfaces was not dependent on CS concentration and Mw. Similar results were obtained by Ururahy et al. [ 50 ], which show that different CS concentrations did not influence the wettability of the substrate. Moreover, Stoleru et al. [ 51 ] revealed that the wettability of PLA films functionalized with different CS was not affected by CS Mw. Concerning profilometry analysis, the PLA and CS-based surfaces display similar roughness values, regardless of the CS concentration and Mw. Although some studies have reported that the surface roughness increases with the deposition of CS [ 52 , 53 ], this effect is highly dependent on CS properties (Mw and degree of deacetylation) and the coating method [ 51 ]. Overall, although the characterization of the PLA-CS surfaces indicates that CS was successfully incorporated onto PLA films, the effect of CS Mw and concentration on the surface properties was not significant. Lately, several researchers have studied the antifouling performance of CS combined with other compounds, such as zinc oxide and copper oxide, to improve its antifouling properties and stability [ 17 , 54 , 55 ]. To the best of our knowledge, this is the first study that evaluates the antifouling performance of PLA surfaces coated with native CS and its derivatives without adding any other compound against C. marina . Results from biofilm cell culturability indicate that C. marina biofilms formed on 0.5% and 1% ( w / v ) CS-based surfaces presented a significantly lower number of culturable cells compared to those grown on PLA films, revealing the antimicrobial performance of the functionalized surfaces. Although the efficacy of an antifouling coating may be dependent on a wide range of environmental factors, such as salinity, availability of nutrients, hydrodynamics, and organisms [ 7 , 56 ], the results obtained are supported by the literature. The antimicrobial activity of CS-based coatings with CS concentrations below 2% ( w / v ) has already been reported against some fouling microorganisms, including Bacillus sp., Pseudomonas sp., and Vibrio [ 57 , 58 , 59 ]. Jena et al. [ 60 ] investigated the effect of CS-based coatings on biofilm density and revealed that coated surfaces allowed to reduce Pseudomonas sp. density (CFU·cm −2 ) by 84% compared to uncoated specimens. Moreover, previous laboratory and mesocosm experiments performed by Al-Naamani et al. [ 54 ] revealed that CS paints were able to significantly reduce the density of diatom Navicula incerta and marine fouling bacteria Pseudoalteromonas nigrifaciens . Elshaarawy et al. [ 61 ] evaluated the antibacterial effect of CS against a range of significant biofilm-inducing bacterial strains such as Escherichia coli , Staphylococcus aureus , Aeromonas hydrophila and Vibrio , revealing that CS presents a higher bactericidal activity compared to a standard antifoulant Diuron. Moreover, Al-Naamani et al. [ 62 ] show that plastic films coated with 2.5% CS significantly reduced the settlement of Bugula neritina compared to uncoated plastic films. After being incorporated into a marine paint and applied to plastic substrates, CS-based coatings were found to inhibit bacterial fouling over one week, and to significantly reduce the cell density of fouling bacteria after two weeks of immersion in a natural seawater environment. These results were corroborated by Dobretsov et al. [ 63 ], who disclosed that CS paints significantly reduced the biofouling on surfaces exposed to the environmental conditions in the Sea of Oman, emphasizing the potential of CS to be applied in protective paints. The antimicrobial activity of CS-based coatings is dependent on a range of intrinsic and extrinsic factors, including microorganism species, surface wettability and roughness, CS degree of deacetylation, concentration, and Mw [ 20 , 26 ]. In the present work, surfaces with different CS concentrations and Mw were produced and show different antifouling performances. Comparing the results obtained for 0.5% and 1% ( w / v ) CS surfaces, no significant differences in the number of C. marina biofilm culturable cells were observed, corroborating a previous study published by Al-Belushi et al. [ 64 ], where the effect of 1% and 2% ( w / v ) CS coatings against a Gram-negative bacteria was assessed. Since all CS-based surfaces presented similar values of water contact angles and roughness, it is not expected that the physicochemical properties and morphology of the surfaces directly impact their antimicrobial activity. In addition, flow cytometric experiments indicate that 0.5 and 1% β-CS treatments yield a similar antimicrobial effect. Therefore, CS Mw seems to be the main parameter to influence the bactericidal performance of the functionalized surfaces. Indeed, regardless of CS concentration, the PLA-CS surfaces show different bactericidal performances; the reduction of C. marina culturability is higher on the PLA films coated with CS of lower Mw. There is no general agreement on the relationship between the CS Mw and its antimicrobial activity. No et al. [ 59 ] demonstrated that oligo-CS with an Mw of 1–10 kDa had a lower antimicrobial activity compared to CS of higher Mw (22–1671 kDa). Moreover, no significant differences were found between the antibacterial performance of low (22 and 59 kDa), medium (224 kDa), and high (470, 746, 1106, and 1671 kDa) CS Mw [ 59 ]. The effect of CS Mw (2–16 kDa) was also assessed in a study developed by Simunek et al. [ 65 ] which revealed that CS antimicrobial activity increased with the increase of CS Mw. On the other hand, some studies have shown that lower Mw CS is more effective. Tayel et al. [ 66 ] evaluated the effect of Mw on the antimicrobial activity of CS with 21, 27, 140, and 190 kDa and showed that, in general, decreasing the Mw of CS slightly increased its antimicrobial activity. Likewise, Benhabiles et al. [ 67 ] demonstrated that a native CS extracted from shrimp shell waste presented a reduced antimicrobial activity compared to its derivates of lower Mw. Moreover, Zheng et al. [ 68 ] revealed that the antimicrobial performance of CS was higher with lower Mw against Gram-negative bacteria, but not against Gram-positive bacteria, which corroborates the results obtained in the present study. Since the biofilm structure can impact its resistance to mechanical and chemical agents, such as fluid shear and antifouling compounds [ 69 ], the effect of immobilized CS and its derivates against biofilm formation was also analyzed by OCT imaging. Both quantitative data of biofilm thickness and 3D biofilm structures highlighted the effect of CS on C. marina biofilm growth, demonstrating that all functionalized surfaces had thinner and more compact biofilms than PLA films. A less compact structure combined with the presence of streamers on biofilms developed on PLA surfaces can enhance biofilm growth by promoting the transfer of nutrients to the inner layers and the capture of new cells and other components to the biofilm [ 70 ]. These results corroborate the biofilm culturable cell analysis and are in accordance with previous studies where the effect of CS-based paints on biofilm structure was evaluated. El-Saied et al. [ 71 ] investigated the antifouling performance of a CS-based marine paint by immersing coated PVC panels in the Mediterranean Sea Eastern Harbor of Alexandria. The findings revealed the long-term antifouling activity of CS since the coatings inhibited the development of tubeworms and barnacles on panels submerged for more than two months. Similarly, Elshaarawy et al. [ 72 ] showed that CS-coated panels were highly efficient against tube worms, barnacles, and macroalgae settlement, even when compared to a standard antifoulant. Within the biofilm, microorganisms are commonly organized in specialized niches, where heterogeneous microenvironments exist as a result of nutrient transport and chemical gradients [ 34 ]. Quorum sensing, intracellular communication, and consequently biofilm formation can be affected by the spatial distribution of microorganisms, which is impacted by microenvironmental factors [ 69 ]. The presence of empty spaces in biofilms can impact the structure of microbial communities and consequently their diversity, activity, and synergism [ 34 , 73 ]. In the present work, OCT analysis revealed that PLA-CS surfaces show lower percentage and mean size values of empty spaces compared to uncoated PLA films. The empty spaces could be beneficial for mass transport within the biofilm, enhancing the distribution of nutrients and oxygen and providing a way for the removal of metabolic end-products [ 74 ]. Thus, biofilms with a higher percentage and larger size of empty spaces, such as the biofilms developed on PLA surfaces, are more prone to the flow of medium throughout the biofilm, which may result in the establishment and expansion of a channel network, relieving nutrient limitations and promoting biofilm growth [ 74 , 75 ]. Altogether, these findings suggest that PLA coated with CS extracted from fishery industry discards may provide an efficient and environmentally friendly approach to retard biofouling in submerged surfaces. Moreover, the present work reveals that the effect of CS Mw strongly affects biofilm cell number and architecture, and the underlying mechanism of the antimicrobial effect of studied CS was cell membrane depolarization and consequent integrity loss."
} | 5,699 |
25915636 | null | s2 | 2,780 | {
"abstract": "High-throughput DNA sequencing has proven invaluable for investigating diverse environmental and host-associated microbial communities. In this Review, we discuss emerging strategies for microbial community analysis that complement and expand traditional metagenomic profiling. These include novel DNA sequencing strategies for identifying strain-level microbial variation and community temporal dynamics; measuring multiple 'omic' data types that better capture community functional activity, such as transcriptomics, proteomics and metabolomics; and combining multiple forms of omic data in an integrated framework. We highlight studies in which the 'multi-omics' approach has led to improved mechanistic models of microbial community structure and function."
} | 190 |
32168387 | null | s2 | 2,781 | {
"abstract": "Microalgae and cyanobacteria contribute roughly half of the global photosynthetic carbon assimilation. Faced with limited access to CO"
} | 33 |
29318200 | PMC5655347 | pmc | 2,784 | {
"abstract": "Microbial production of chemicals and proteins from biomass-derived and waste sugar streams is a rapidly growing area of research and development. While the model yeast Saccharomyces cerevisia e is an excellent host for the conversion of glucose to ethanol, production of other chemicals from alternative substrates often requires extensive strain engineering. To avoid complex and intensive engineering of S. cerevisiae, other yeasts are often selected as hosts for bioprocessing based on their natural capacity to produce a desired product: for example, the efficient production and secretion of proteins, lipids, and primary metabolites that have value as commodity chemicals. Even when using yeasts with beneficial native phenotypes, metabolic engineering to increase yield, titer, and production rate is essential. The non-conventional yeasts Kluyveromyces lactis, K. marxianus, Scheffersomyces stipitis, Yarrowia lipolytica, Hansenula polymorpha and Pichia pastoris have been developed as eukaryotic hosts because of their desirable phenotypes, including thermotolerance, assimilation of diverse carbon sources, and high protein secretion. However, advanced metabolic engineering in these yeasts has been limited. This review outlines the challenges of using non-conventional yeasts for strain and pathway engineering, and discusses the developed solutions to these problems and the resulting applications in industrial biotechnology.",
"introduction": "1 Introduction The microbial production of fuels and chemicals from biomass and other renewable carbon sources is an attractive alternative to petroleum-derived products. One of the largest scale example of this is ethanol production by the yeast Saccharomyces cerevisiae — in 2015, over 25 billion gallons were produced worldwide from starch, waste sugar streams, and biomass-derived sugars ( www.afdc.energy.gov/data/10331 ). S. cerevisiae is the organism of choice because of its high rate of production and tolerance to ethanol titers upwards of 120 g L −1 \n [1] , [2] . These phenotypes, among others, have led to the widespread study of S. cerevisiae and its development as a model eukaryotic host for chemical biosynthesis. A valuable approach to metabolic engineering is identifying organisms with desirable phenotypes and developing new synthetic biology tools to enhance these phenotypes. Bioethanol production in S. cerevisiae is a good example of this, and illustrates the potential of identifying other hosts and phenotypes to synthesize bioproducts other than ethanol. A number of examples of this strategy already exist in industry, where non-conventional yeasts with unique and advantageous phenotypes are used to produce proteins, lipids, and commodity chemicals. Metabolic engineering in these yeasts is, however, more challenging in comparison with S. cerevisiae , because less is known about their metabolism and genomics, and advanced genetic engineering tools are limited. In this review, we focus on six non-conventional yeasts ( Table 1 ): Kluyveromyces lactis, K. marxianus, Scheffersomyces (Pichia) stipitis, Yarrowia lipolytica, Hansenula polymorpha, and Pichia pastoris. In contrast to S. cerevisiae, these yeasts are Crabtree negative and favor respiration over fermentation; phenotypes that are particularly useful for protein production as well as the biosynthesis of chemicals other than ethanol [3] . K. lactis is discussed here because of its capacity to metabolize inexpensive substrates such as waste whey and because of its use as a host for heterologous protein production in the food, feed, and pharmaceutical industries [4] . The Kluyveromyces species K. marxianus is also industrially relevant because of its wide substrate spectrum, fast growth characteristics, and thermotolerance to ∼50 °C [5] , [6] . Native strains of K. marxianus are also known to synthesize ethyl acetate at rates above 2 g L −1 h −1 in aerated bioreactors [7] , [8] . S. stipitis is capable of fermenting xylose at high rates compared to other yeasts and has been widely studied for ethanol production from biomass-derived sugars [9] , [10] . Y. lipolytica is a well-studied oleaginous yeast and has attracted interest due to its ability to synthesize and accumulate high levels of intracellular lipids [11] , [12] , [13] . The methylotrophic yeast H. polymorpha has been studied as a model system for peroxisome function as well as for its methanol and nitrate assimilation pathways [14] , [15] . Significant efforts have gone into heterologous protein production in H. polymorpha due to its efficient secretion pathways, effective glycosylation machinery, and tightly controlled expression systems [16] . H. polymorpha is also thermotolerant to temperatures comparable to K. marxianus and can assimilates various substrates, thus making it a potential alternative host for ethanol production [17] . The methylotrophic yeast P. pastoris has similar protein secretion and glycosylation capabilities to H. polymorpha and has been widely used for heterologous protein production [18] . Its capacity to grow to extremely high cell densities and high capacity for membrane protein expression also provide inherent advantages over other yeast hosts [19] , [20] . Table 1 Overview of non-conventional yeast species, their industrially-relevant phenotypes, common uses in biotechnology, and comparison with S. cerevisiae . Table 1 Yeast Beneficial Phenotype Products Reference K. lactis High protein secretion Growth on lactose Proteins for food and feed industry Pharmaceutical enzymes [4] K. marxianus Thermotolerance Fast growth characteristics High ethyl acetate production Growth on a range of sugars Ethanol and volatile acetate esters [5] S. stipitis High ethanol production from xylose Ethanol fermentation from biomass derived carbohydrates [21] Y. lipolytica Efficient production of lipids Growth on glycerol and alkanes Lipids and oleochemicals [12] H. polymorpha Thermotolerance Tightly regulated expression system Beneficial glycosylation for therapeutics Heterologous protein High temperature ethanol fermentation [17] , [18] P. pastoris Tightly regulated expression system High cell density on minimal media Beneficial glycosylation for therapeutics Efficient production of membrane proteins Pharmaceuticals and industrial enzymes [18] S. cerevisiae High ethanol production High HR capacity Well known genomics and physiology Advanced synthetic biology tools Ethanol in fermented beverages and as biofuel Commodity and specialty chemicals Pharmaceuticals [2] , [22] Despite these many advantages, metabolic engineering of non-conventional yeasts is limited by a lack of sophisticated genome editing tools and an incomplete understanding of their genetics, metabolism, and cellular physiology. In this review, we discuss the challenges and solutions that have arisen in engineering non-conventional yeasts for metabolic engineering and synthetic biology applications. We begin our review with a discussion of the challenges to genetic engineering, followed by a discussion of strategies for improving genome and pathway engineering. Finally, we discuss representative examples of metabolic engineering in each of the selected yeasts. While the presented examples are not exhaustive, they are exemplative of current and past research efforts that exploit the yeasts' advantageous phenotypes. Reviews that provide comprehensive discussions on engineering each of the non-conventional yeasts described here are available elsewhere [4] , [12] , [17] , [21] , [23] , [24] ."
} | 1,900 |
24022336 | PMC3778718 | pmc | 2,785 | {
"abstract": "Understanding the compatibility between spider silk and conducting materials is essential to advance the use of spider silk in electronic applications. Spider silk is tough, but becomes soft when exposed to water. Here we report a strong affinity of amine-functionalised multi-walled carbon nanotubes for spider silk, with coating assisted by a water and mechanical shear method. The nanotubes adhere uniformly and bond to the silk fibre surface to produce tough, custom-shaped, flexible and electrically conducting fibres after drying and contraction. The conductivity of coated silk fibres is reversibly sensitive to strain and humidity, leading to proof-of-concept sensor and actuator demonstrations.",
"discussion": "Discussion The essential aspects leading to the realization of the f-CNT-SS material are as follows. The f-CNTs are polar, with positive charge at the amine sites. The SS is a protein-based polymer where the amino acid groups vary along the backbone, some are neutral and some polar 27 . By mechanical mixing, the dry f-CNT powder is partially dispersed and adheres to the SS fibres due to polar interaction ( Supplementary Fig. S3 ). When water is applied to the mixture, the f-CNTs disperse further and the SS fibres experience hydrogen bond breaking 8 , resulting in fibre swelling and softening. As a result, the surface area of the fibre is increased, allowing more f-CNTs to adhere to the fibre. Applications of shear and pressure bring the f-CNTs in closer proximity to the surface of the fibre, promoting both physical and chemical interactions between them. Upon drying, the SS fibre matrix shrinks further as hydrogen bonding is re-established 8 28 , concentrating the CNT array and making it electrically conducting. The FTIR spectra in Fig. 3 indicate a change in the nature of the SS carboxylic acids, consistent with the aqueous chemical reaction between the NH 2 side groups of the f-CNT and the COOH component of aspartic and glutamic acids in the SS. As our water-based method is performed at room temperature, amide formation is unlikely because this type of reaction normally occurs at high temperatures 22 . However, ionic and hydrogen bonding are both likely to occur. In an aqueous solution, for example with pH of 4–7, some amount of side-chain carboxylic acids and amines are typically ionized. By applying shear and pressure, ionic bonding between them is promoted. In parallel, the NH 2 side groups may form hydrogen bonding with the non-ionized aspartic and glutamic acids ( Supplementary Fig. S9 ). The formation of hydrogen bonding at room temperature has been observed in the fabrication of buckypaper from O-H-functionalised CNTs 29 . Even though the f-CNT anchoring to the SS fibre is minimal due to the small abundance of aspartic and glutamic acids, it generates a significant grafting, such that when combined with the van der Waals interaction and the natural tendency of the f-CNTs to entangle, a new hybrid functional silk fibre is produced ( Supplementary Fig. S9 ). The intimate adherence of the f-CNTs on the SS fibre is the key factor that results in many of the observed phenomena such as the improved extensibility and toughness of the f-CNT-SS fibre. The toughness of f-CNT-SS fibres can be attributed first to the supercontraction that occurs during the coating process and second to the distribution of SS radial deformation by the f-CNT coating. SS fibres generally become tougher after being supercontracted. The release of internal pre-stress 8 in neat SS fibres during supercontraction accounts for the additional energy they can absorb before rupturing. The f-CNT coating may further improve the toughness by effectively distributing the SS radial deformations associated with the rupture point during the extension process. A SS fibre experiences a very large radial deformation when strained longitudinally, with a Poisson ratio (a measure of how much the diameter of the fibre shrinks when the fibre is extended) of ~1.5 (ref. 30 ). In contrast, a CNT network, such as the buckypaper, experiences significantly less deformation (Poisson ratio up to 0.3), except in some special cases 31 . If intimately connected to the SS fibre, this can reduce the radial deformation of the silk fibre at the highest rupture point, for example, in the middle of the fibre, allowing further extension before it ruptures. The uniformity of the f-CNT coating is demonstrated by the absence of sudden jumps in the resistance versus strain curve up to at least 50% strain ( Fig. 5a ). The uniform adherence strength between the f-CNT to SS and f-CNT to f-CNT contacts allows homogenous strain distribution during extension, similar to the mechanism observed in CNT-elastomer or gold-elastomer systems 32 33 , where extensions of >100 or 20% are observed, respectively. Additional flexibility may be provided by the fibre shrinkage during the water-based processing, which effectively concentrates the f-CNT network. We emphasize that no f-CNT crosslinkers are used in our water-based coating method. Without crosslinkers, it has been reported that the f-CNT network is typically very brittle 23 . The 3D VRH transport of the f-CNT-SS reveals that the conductivity is dominated by inter-tube charge carrier hopping between the f-CNTs 34 , which may be related to the submicron length of the f-CNTs used (therefore resulting in a larger number of inter-tube contacts). The T 0 and conductivity of the f-CNT-SS and the benzoquinone-crosslinked buckypaper are similar. This suggests that the simple water-based coating method is effective. The slightly higher T 0 could mean that the f-CNTs were not as tightly entangled as in the case of crosslinked buckypaper, resulting in a slightly higher contact resistance between the f-CNTs. The application of an annealing current under ambient conditions indicates a reduction in the contact resistance of the f-CNT network. A similar effect is observed in carbon nanofibre (CNF)–gold interconnects, where the application of an annealing current at ambient condition reduces the contact resistance between the CNF–gold interconnects 35 . As the f-CNT coating on the SS surface is very thin, ~80–100 nm, a 100-μA annealing current may generate a considerable current density that heats the CNT–CNT joints, producing better contacts. Owing to the thin, flexible and porous nature of the f-CNT network, external stimuli such as varying strain and humidity levels will affect the SS fibre. We emphasize that for gold-coated (~20-nm thick) or thicker f-CNT coated SS (several micrometre thick; Methods section), and neat buckypaper (~30 μm thick), the humidity response is not observed. As a local heating element, the f-CNT coating allows us to drive the SS-based actuator reported earlier 11 using electrical currents in contraction and extension mode by exploiting the swelling/de-swelling and thermal expansion/contraction of the silk fibre, respectively. The f-CNT-SS contraction of ~1% in our proof-of-concept demonstration (performed at 55% RH) is comparable to the reported value in the previous study 11 . In their case, the average contraction for lifting a 9.5-mg mass is ~1.7% in the full-humidity range of 90–10% RH. Assuming a linear correlation between the contraction length and RH, we expect that a variation of RH from 55 to 10% will generate an average contraction of 0.95%, which agrees very well with our result. This suggests that our coating approach does not degrade the actuating properties of the silk fibre. In conclusion, we have developed a simple and effective water-based and shear-assisted method of fabricating tough, versatile, flexible and multi-functional f-CNT-SS fibres. Amine-functionalised MWCNT adheres effectively to the SS fibre, as revealed by SEM and TEM images and by structural changes in the carboxylic acid of the SS as observed in the FTIR spectra. The uniformity of the coating is further confirmed from the Raman spectra, strain-dependent resistance and electrical conductivity estimation. The charge carrier transport is primarily driven by inter-tube charge hopping, as revealed by the 3D VRH temperature-dependent transport. The combination of a thin, flexible and porous CNT network with SS fibres is synergistic, resulting in polar, custom-shapeable, self-monitoring and actuating devices."
} | 2,081 |
30457843 | null | s2 | 2,786 | {
"abstract": "Microbiomes impact nearly every environment on Earth by modulating the molecular composition of the environment. Temporally changing environmental stimuli and spatial organization are major variables shaping the structure and function of microbiomes. The web of interactions among members of these communities and between the organisms and the environment dictates microbiome functions. Microbial interactions are major drivers of microbiomes and are modulated by spatiotemporal parameters. A mechanistic and quantitative understanding of ecological, molecular, and environmental forces shaping microbiomes could inform strategies to control microbiome dynamics and functions. Major challenges for harnessing the potential of microbiomes for diverse applications include the development of predictive modeling frameworks and tools for precise manipulation of microbiome behaviors."
} | 220 |
29062953 | PMC5625740 | pmc | 2,787 | {
"abstract": "The UK Synthetic Biology Research Centre, SYNBIOCHEM, hosted by the Manchester Institute of Biotechnology at the University of Manchester is delivering innovative technology platforms to facilitate the predictable engineering of microbial bio-factories for fine and speciality chemicals production. We provide an overview of our foundry activities that are being applied to grand challenge projects to deliver innovation in bio-based chemicals production for industrial biotechnology."
} | 121 |
35527906 | PMC9069832 | pmc | 2,788 | {
"abstract": "Zwitterionic hydrogels have promising potential as a result of their anti-fouling and biocompatible properties, but they have recently also gained further attention due to their controllable stimuli responses. We successfully synthesized two zwitterionic polymers, poly(2-methacryloyloxyethyl phosphorylcholine) (poly-MPC) and poly(2-(methacryloyloxy)ethyl dimethyl-(3-sulfopropyl)ammonium hydroxide) (poly-DMAPS), which have complementary ionic sequences in their respective zwitterionic side groups and likely form an interpenetrating double network to improve their mechanical strength. The synthesized poly-MPC was blended in a poly-DMAPS matrix (MD gel) and showed high viscosity, while poly-DMAPS was blended in a poly-MPC hydrogel (DM gel) and revealed UCST behavior as the temperature increased. In addition, cross-section images of the MD hydrogel exhibited its compact and uniform structure, while the DM gel was found to exhibit a porous micro-structure with clear boundaries. The results explained the low viscosity of the DM gel, which was also confirmed via 3D Raman mapping. To sum up, the preliminary data demonstrated that binary zwitterionic hydrogels have thermosensitive mechanical properties, promoting further bio-applications in the future, such as in wound healing.",
"conclusion": "Conclusions Zwitterionic hydrogels with promising properties, such as anti-fouling functionality, can be currently used in implanted devices or wound patches. However, their weak mechanical properties are still an unsolved issue for these bio-applications. Herein, hydrogels have been successfully prepared with an interpenetrating double network via the mixing of two zwitterionic polymers, poly-MPC and poly-DMAPS, to improve mechanical strength. The MD gel presented DN network-boosted mechanical strength and a homogeneous microstructure but lacked a thermosensitive response. On the other hand, unbound UCST poly-DMAPS in the DM hydrogel demonstrated a reduction in viscosity as the temperature increased. Although poly-DMAPS immersed in the DM gel lost its clear and instantaneous thermo-response, the DN network still improves the mechanical strength of the DM gel and promotes further applications, such as thermosensitive steady drug release via a warm compress for wound healing care or pain patches.",
"introduction": "Introduction Synthesized hydrophilic polymers can offer wide-ranging designs for gelation via physical or chemical cross-linking, resulting in the construction of three-dimensional polymer network structures and good water swelling properties. 1 Of these, in particular, zwitterionic polymers forming hydrogels show anti-fouling properties and have indirect pro-healing effects for use in implantable medical devices, tissue scaffolds and, especially, wound care. 2–4 However, hydrogels composed of chemically crosslinked polyzwitterions have issues with respect to their mechanical limitations. 5 In the literatures, several methods have been used to improve the mechanical strength of hydrogels, such as the formation of non-covalent bond (ionic interactions 6 and hydrogen bonds 7 ) hydrogels with highly stretchable networks, 8 and double-network (DN) hydrogels. 9–11 The way in which DN hydrogels improve mechanical strength is through the two polymer matrices, as one can provide a rigid but brittle network and the other can present a soft and ductile matrix, preventing crack propagation and resisting hydrogel breaks. 11,12 The most typical DN gel pairs consist of poly(2-acrylamido-2-methylpropanesulfonic acid) gel and polyacrylamide gel, with upgraded strength 100–1000 times higher than the individual gels. 10,11,13 The difference here is that we do not prepare an ultra-high toughness DN hydrogel in this study but design an excellent anti-biofouling hydrogel with controlled mechanical strength using a dual-polyzwitterionic double network. Two polyzwitterions are individually prepared via the polymerization of either 2-methacryloyloxyethyl phosphorylcholine (MPC) or [2-(methacryloyloxy) ethyl]dimethyl-(3-sulfopropyl)ammonium hydroxide (DMAPS). There are complementary ionic sequences on the zwitterionic side groups of MPC and DMAPS, likely providing potential interpenetration interaction sites via coulombic interactions. 10,14,15 Poly-MPC is composed of a cell-membrane structure with phosphorylcholine as a side chain and it has hemocompatible properties and cell-adhesion resistance, 16–18 while poly-DMAPS is a well-known zwitterionic polymer having upper critical solution temperature (UCST) behavior. 19,20 When the temperature is lower than the UCST, poly-DMAPS chains aggregate, enhancing chain entanglements and exhibiting higher viscosity. 19,20 Once above the UCST, the polymers preferably change to the sol state, relieving the polymer tension and reducing the viscosity of the hydrogel. 21 Herein, we expect that this dual-polyzwitterionic matrix could not only bolster the mechanical strength of a bulk-type hydrogel but it could also lead to control of the thermoresponsive mechanical properties for controlled sustained release. In future applications, such as wound healing care, thermosensitive hydrogels have promising potential to encapsulate painkiller medicines in wound patches and thus accelerate drug release via a warming compress, promptly achieving pain relief.",
"discussion": "Results and discussion The components of the synthetic zwitterionic polymers and the formation of truly independent-DN gels (no covalent bonds between the two polymeric networks) are depicted in Fig. 1 . The synthesis processes can be divided into two parts: one involves homo-polymer synthesis and the other involves the prepared homo-polymers being mixed in secondary crosslinked matrices to form truly independent DN gels. 10 The combination of poly-MPC and poly-DMAPS to improve the intermolecular interactions forms the basis of the thermosensitive mechanical strength of the hydrogels in this study. Fig. 1 A schematic illustration of polymerization and double network hydrogel formation. (a) The preparation of the initial polymers. (b) Polymerization through mixing the initial polymers (either poly-MPC or poly-DMAPS) with the respective secondary monomers and a crosslinker to form double network hydrogels. Note: MD gel represents the initial polymer poly-MPC mixed in chemically crosslinked poly-DMAPS hydrogel. In contrast, DM gel represents the initial polymer poly-DMAPS mixed in chemically crosslinked poly-MPC hydrogel. Firstly, zwitterionic homopolymers (either poly-MPC or poly-DMAPS) were synthesized via UV irradiated free radical polymerization and characterized via 1 H-NMR. The proton peaks corresponding to the side chains of poly-MPC are located at δ 3.8 ppm (–C H 2 N + –), 4.1 ppm (–POC H 2 CH 2 O–), and 4.3 ppm (–NCH 2 C H 2 OP–). On the other hand, the peaks from the poly-DMAPS chains appear at 2.2 ppm (–NCH 2 C H 2 CH 2 SO 3 ) and 3.0 ppm (–CH 2 C H 2 SO 3 ) in the 1 H-NMR spectrum (ESI, Fig. S2 † ). In addition, the double peaks from vinyl groups (at 5.8 and 6.3 ppm) disappear, indicating that the radical polymerization reaction has been completed. Following the structural characterization, it is practical to tune the critical temperature around the physiological temperature range; for this, the molecular weight and concentration of the polymers are two pivotal parameters to control the thermal response. The long poly-DMAPS chains preferably form hydrophobic associations, leading to precipitation in aqueous solvent. On the other hand, the low molecular weight of the polymer makes it difficult to form such associations between the polymer chains. 19 Herein, we adjusted the ratio of photo-initiator to monomer to different values (0.5 mol% and 1 mol%) to prepare poly-DMAPS with two molecular weights ( M n 21k Da and 46k Da), which were characterized by GPC as shown in Fig. S3. † In order to develop a hydrogel with UCST functionality, the USCT properties of the poly-DMAPS should be well understood. Turbidity measurements are a simple and common method to evaluate UCST behavior. The UCST transition temperature is defined as the temperature at the cloud point, which is taken as the 50% transmittance of visible light in UV-visible absorption spectra. 22 When the solution temperature is below the transition temperature, the polymers aggregate and present a cloudy appearance in solution. On the contrary, the aggregated polymer chains untangle and the solution turns transparent once the temperature becomes higher than the UCST. During turbidity measurements ( Fig. 2 ), the low molecular weight poly-DMAPS shows a relatively low cloud point at three concentrations, compared to the high molecular weight polymer. The short polymer segments may cause aggregation between polymer chains, but only a small amount of heat is required for dissolution. 23 In addition, the different polymer concentrations presented different turbidity profiles. When the concentration of low molecular weight poly-DMAPS increased from 1 wt% to 10 wt%, the cloud point shifted from 35 °C to 51 °C, while the high molecular weight polymer presented a high UCST above 50 °C at all concentrations. As the result, the low molecular weight ( M n 21k Da) homopolymer poly-DMAPS was chosen for the following studies because its UCST range (1 wt% to 10 wt%) covered the physiological temperature range. Although the cloud point of lower molecular weight ( M n ≤ 20k Da) poly-DMAPS was also measured via turbidity testing, the poly-DMAPS chains are too short to aggregate, even at room temperature. Fig. 2 Turbidity cooling curves of poly(DMAPS) in aqueous solution: (a) lower molecular weight ( M n 21k Da) and (b) higher molecular weight ( M n 46k Da) synthesized homopolymer poly-DMAPS. A cloudy solution of poly-DMAPS is found at temperatures below the UCST (c, left), while the solution becomes transparent at temperatures above the UCST (c, right). Following the preparation of the initial zwitterionic homopolymers (either poly-MPC or poly-DMAPS), one prepared polymer was mixed with the other monomer and a crosslinker before polymerization to form bulk hydrogels in home-made cells (18 mm × 18 mm × 5 mm, Fig. S1 † ). In this study, there are two DN gel prototypes, as shown in Fig. 1 . One is the synthesized homopolymer poly-MPC mixed in a crosslinked poly-DMAPS hydrogel, which is called MD gel. The other is poly-DMAPS immersed in a crosslinked poly-MPC hydrogel (DM gel), as shown in Fig. S4. † We expect that both gels will have different characteristics due to the thermally responsive poly-DMAPS existing in different matrices, either incorporated as linear polymers unbound in a crosslinked matrix or bound in a chemically crosslinked network; therefore it will likely present varying UCST behavior in the hydrogels. For the characterization of bulk-type samples, Raman spectroscopy is a useful and direct method and it takes advantage of low sensitivity to hydrated samples. 24 It can provide a quick characterization of bulk gels to obtain the composition ratio and homogeneity without breaking samples. In Fig. 3 , the interactions between incident photons and different functional groups of poly-MPC and poly-DMAPS have been qualified and quantified. The peaks at 795 cm −1 correspond to the phosphate group of poly-MPC (labelled as peak “p”) 25 and the peaks at 1034 cm −1 are attributed to the sulfonate group of poly-DMAPS (labelled as peak “s”). 26 In addition, the composition of the bulk hydrogels was firstly determined from the intensity ratios of the selected peaks. The spectra can present the intensity ratios of the selected functional groups at different monomer concentrations ( C [MPC] : C [DMAPS] ). However, an increasing concentration of the DMAPS monomer being present during polymerization did not result in a proportional composition in the hydrogel. Herein, a reasonable explanation is that it is like that MPC has been diluted by DMAPS monomers. Moving from 1D Raman spectra to 2D Raman mapping, we chose the “p” and “s” peaks to investigate the distribution of the two polymers in a cross-section of MD gel. From this we can see that MD gel presents a clearly homogeneous distribution of both polymers, as shown in Fig. 4a . Fig. 3 Raman spectra analysis of hydrogels formed via the random polymerization of various monomer concentration ratios ( C [MPC] : C [DMAPS] = 1 : 1 to 1 : 4). The peaks marked “p” represent the phosphate groups of poly-MPC chains and the peaks marked “s” represent the sulfonate groups of poly-DMAPS chain. Fig. 4 Raman mapping images of bulk hydrogels with C [MPC] : C [DMAPS] = 1 : 3. (a) 2D mapping of the intensity distribution of poly-MPC and poly-DMAPS in the MD gel, in which the scale bar is the intensity ratio of DMAPS to MPC side chains. 3D mapping images of (b) the MD gel, (c) the DM gel, and (d) a random gel. The selected area for Raman scanning is 10 × 10 μm 2 with each layer at 10 μm intervals. The characteristic scale bar represents the intensity of the sulfonate group derived from poly-DMAPS. Furthermore, bulk hydrogels (MD gel, DM gel, random gel) were characterized via 3D-mapping Raman scanning to investigate the structural homogeneity. 27 Based on the selected peak of the sulfonate group of poly-DMAPS, tomographic Raman mapping was performed from the intensity distribution in a 3D micro-zone (10 × 10 × 10 μm 3 ). A high Raman intensity indicated potential scenarios, such as self-chain interactions, which are likely caused by UCST-related aggregation. More interestingly, X – Z cross-section images of the DM gel showed noticeable differences in intensity signals, as shown in Fig. 4c , demonstrating more inhomogeneity in this dual-polymeric matrix compared to MD gel. The phase separation of the DM gel is likely caused because thermosensitive poly-DMAPS preferably demonstrates intramolecular and intermolecular attractions between the DMAPS side chains at temperatures below the UCST. In addition, the X – Z cross-section mapping images of a random gel as a control presented the slight aggregation of either poly-MPC- or poly-DMAPS-dominated phases, distributed in a random pattern through the gel ( Fig. 4d ). Since the random gel was prepared via the one-step polymerization of DMAPS and MPC monomers with a chemical crosslinker, the zwitterionic side chains of both monomers undergo interactions with each other before polymerization. As the result, there is no inherent UCST behavior between the polymer chains, because once polymer chains are bound by the chemically crosslinked matrix, the UCST polymers do not have enough freedom to undergo sufficient intra-molecular interactions and the MPC side chains can even interfere with the micro-structural aggregation of poly-DMAPS segments during polymerization. To study the double network rheology and thermosensitive response of the zwitterionic hydrogels, the viscosities of the DM gel, MD gel and random gel were measured; they are ranked in the following sequence: MD gel > random gel > DM gel > poly-DMAPS-only gel > poly-MPC-only gel, at slow shear rates (10 −3 to 10 −2 s −1 ), as shown in Fig. 5a . All two-component hydrogels (MD gel, DM gel, random gel) were found to exhibit higher viscosities than the single-component hydrogels made from either poly-DMAPS or poly-MPC, proving the concept that the double network boosts the hydrogel mechanical properties. This concept was similarly confirmed through mixing the linear polyzwitterions in the same polyzwitterionic chemical networks. 28 Fig. 5 Mechanical measurements of various zwitterionic hydrogels via a rheometer at an MPC : DMAPS molar ratio of 1 : 3. The hydrogels were studied (a) at different shear rates at a constant temperature of 25 °C or (b) at different temperatures at the same shear rate (0.01 s −1 ) to investigate thermal sensitivity. (c) Photographs of the DM gel at room temperature or at 55 °C. In our binary polyzwitterion hydrogel systems, the properties of the highest strength gel, the MD gel, were attributed to both the physical and chemical crosslinked networks, which co-facilitated the formation of a uniform double network structure. 14,15 Poly-DMAPS in the MD gel has, however, been bound by covalent crosslinking, resulting in the absence of UCST behavior in the MD gel ( Fig. 5b ). In contrast, the DM gel presented thermosensitive viscosity at a constant shear rate. More interestingly, compared to the UCST behavior of linear poly-DMAPS-only solution (no chemical-crosslinking) shown in Fig. 2 , the bulk DM gel presented a steady reduction in rheological strength when the temperature increased from 25 °C to 60 °C. This continuous gradual decrease in viscosity might be attributed to the chemically crosslinked poly-MPC matrix. Although the aggregated poly-DMAPS can be untangled once the temperature is above the UCST, poly-DMAPS is still confined in the hydrogel matrix. As shown in Fig. 5c , photos of the DM gel reveal a noticeable difference in transparency at low and high temperatures, but even at 55 °C there are still some thin but opaque reticulation patterns, probably caused by chemically-bound poly-DMAPS. However, the mechanical strength of the DM gel is weaker than the random gel, which is likely related to the inhomogeneity of its UCST component during the gelation process ( Fig. 4c ). We hypothesized that poly-DMAPS, as the free polymer, was immersed in the second polymeric matrix yet formed UCST aggregations at room temperature, reducing the interactions with MPC side chains during the UV-initiated polymerization. Given the weaker mechanical strength of the DM gel compared to the other DN gels, the thermo-reduced viscosity DM gel has high potential for future drug release applications, such as in heat-stimulated pain relief patches. 29 Regarding mechanical failure, crack propagation is one of the important factors undermining the usability of hydrogels. 30 To deeply understand the microstructures of the hydrogel, we investigated surface sections and cross-sections of the dual-zwitterionic DN hydrogels via SEM imaging. As shown in Fig. 6 , the MD gel was constructed of a dense and uniform interior network. On the contrary, a cross-section of the random gel showed clear porosity and elliptical pores, but these were still distributed homogeneously throughout the bulk gel, while the DM gel image revealed clear boundaries and uneven porosity across the cross-section. This also provides additional evidence to explain why the rheological viscosity of the hydrogels is in the order: MD gel > random gel > DM gel. Fig. 6 Morphological images of different hydrogels on the surface and in cross-sections from SEM imaging. (a) The surface and (b) cross-section of the MD gel; (c) the surface and (d) cross-section of the DM gel; and (e) the surface and (f) cross-section of the random gel."
} | 4,726 |
35077540 | PMC8848325 | pmc | 2,789 | {
"abstract": "Abstract Background Mitigating the effects of global warming has become the main challenge for humanity in recent decades. Livestock farming contributes to greenhouse gas emissions, with an important output of methane from enteric fermentation processes, mostly in ruminants. Because ruminal microbiota is directly involved in digestive fermentation processes and methane biosynthesis, understanding the ecological relationships between rumen microorganisms and their active metabolic pathways is essential for reducing emissions. This study analysed whole rumen metagenome using long reads and considering its compositional nature in order to disentangle the role of rumen microbes in methane emissions. Results The β-diversity analyses suggested a subtle association between methane production and overall microbiota composition (0.01 < R 2 < 0.02). Differential abundance analysis identified 36 genera and 279 KEGGs as significantly associated with methane production ( P adj < 0.05). Those genera associated with high methane production were Eukaryota from Alveolata and Fungi clades, while Bacteria were associated with low methane emissions. The genus-level association network showed 2 clusters grouping Eukaryota and Bacteria, respectively. Regarding microbial gene functions, 41 KEGGs were found to be differentially abundant between low- and high-emission animals and were mainly involved in metabolic pathways. No KEGGs included in the methane metabolism pathway (ko00680) were detected as associated with high methane emissions. The KEGG network showed 3 clusters grouping KEGGs associated with high emissions, low emissions, and not differentially abundant in either. A deeper analysis of the differentially abundant KEGGs revealed that genes related with anaerobic respiration through nitrate degradation were more abundant in low-emission animals. Conclusions Methane emissions are largely associated with the relative abundance of ciliates and fungi. The role of nitrate electron acceptors can be particularly important because this respiration mechanism directly competes with methanogenesis. Whole metagenome sequencing is necessary to jointly consider the relative abundance of Bacteria, Archaea, and Eukaryota in the statistical analyses. Nutritional and genetic strategies to reduce CH 4 emissions should focus on reducing the relative abundance of Alveolata and Fungi in the rumen. This experiment has generated the largest ONT ruminal metagenomic dataset currently available.",
"conclusion": "Conclusions The full metagenome compositional analysis used in this study provided novel insights in the association between the microbiota and CH 4 emissions through differential abundance analysis, pairwise correlation, and interaction networks. Our approach evidenced a phenotypic association between microbiome composition and methane production, regardless of the challenges posed by the microbiome complexity and the compositional nature of the data. This association is mainly driven by the relative abundance of ciliates and fungi, which carry host-specific genetic functions providing substrate to the methanogenic archaea. On the other side, we detected some bacterial groups that performed a more efficient feed digestion, leaving less hydrogen available to archaea and hence associated with lower methane emissions. This study generated the largest ruminal metagenomic dataset sequenced using ONT and grants free access to a publicly available dataset. The complexity of the rumen microbiome and the compositional nature of their sequencing data require proper statistical methods to allow disentangling the role of microbes and their genes in host complex traits such as methane emissions. Future nutritional and genetic strategies to reduce CH 4 emissions should focus on reducing the relative abundance of Alveolata and Fungi in the rumen, without impairing other important metabolic processes for an efficient feed digestion in ruminants.",
"introduction": "Introduction Next-generation sequencing technologies have provided special relevance to microbial communities from different niches because they let their taxonomic and functional profile be identified. This has made it possible to unravel the relationships between host and microbiota, as well as the complex interactions between microbes, focusing on the special contribution of the role of digestive microbiome in complex traits both in humans [ 1 ] (e.g., Type 2 diabetes, cancer, mental diseases) and in domestic animals [ 2 , 3 ] (e.g., feed efficiency, methane emissions, animal health). Microbial communities are of special relevance in livestock. In ruminants, one of the main microbial communities is found in the rumen, owing to its high diversity and large microbial mass [ 4 ] and its main role in feed fermentation to provide substrate to the animal, which is then transformed into product. Additionally, enteric methane is produced in the rumen by methanogenic microorganisms during feed fermentation [ 5 ] and is the main contributor of greenhouse gases from livestock, with 2.8–3.5 gigatonnes of CO 2 -equivalent per year [ 6 , 7 ]. The ongoing climate emergency urgently calls for efficient strategies to mitigate the carbon footprint from all sectors, including agriculture and livestock farming. Previous studies have proven that complex traits in ruminants are usually influenced by global changes in ruminal microbial communities, more than by fluctuations in the abundance of specific microorganisms [ 8 , 9 ]. These global changes are usually due to the intricate interactions between different species in these communities (i.e., predation, competition of ecological niche, or co-dependency). Consequently, a better understanding of the interactions between microbial genes during methanogenesis is needed to propose strategies for reducing methane emissions. Promising strategies have been proposed to modulate the metagenome, nutrition, and genetics [ 10 ]. Classic statistical approaches do not allow the results of microbiome studies to be accurately assessed. The high sparsity of these data and their compositional nature generate multiple problems in statistical analysis, including subcompositional incoherence, increased false-positive rates in differential abundance analyses, and detection of spurious correlations [ 11 ]. As a consequence, new approaches considering both compositionality and multiple correlations are needed. It is also important to point out the advantages of whole-metagenome sequencing over metataxonomic studies because the latter cannot be used to determine functionality and because they pose some difficulties at simultaneously analysing different superkingdoms [ 12 ], which is necessary to account for the total variability of microbiomes and the interactions among their components. Different amplicons must be used to correctly classify Bacteria, Archaea, Protozoa, and Fungi, increasing the cost of the studies and involving additional bias due to PCR [ 13 ]. They pose the additional difficulty of a proper comparison between communities sequenced in different reactions with different primers. Nanopore sequencing offers a cost-efficient sequencing strategy for metagenomics studies, providing both taxonomical and functional information simultaneously and for microbes from all superkingdoms. This technology has been improved in recent years, allowing taxonomic and functional assignments to be performed with an accuracy comparable to Illumina [ 14 ]. The objective of this study was to characterize the taxonomical and functional composition of rumen microbiota using long sequence reads obtained with Nanopore technology, and their relationship with enteric methane emission.",
"discussion": "Discussion In this study we assessed the composition of the ruminal microbiota using long reads from Nanopore sequencing technology. We observed predominance of Bacteroidetes, Firmicutes, and Fibrobacteres, as reported in previous studies [ 8 , 20 ]. Bacteroidetes and Firmicutes are common bacteria in all kind of ecosystems, including gut microbiota of multiple animals. The fraction of Bacteroidetes was mainly composed by Prevotella , a group of anaerobic gram-negative bacteria involved in saccharolytic processes [ 21 ]. Their large abundance in the digestive microbiota has been previously reported in ruminant [ 22–26 ] and monogastric species [ 27 , 28 ]. Firmicutes were less abundant, with a more diverse distribution of genera. Fibrobacteres, a small group of cellulose-degrading bacteria usually present in ruminant digestive systems [ 29 ], was mainly represented by the Fibrobacter genus. Eukaryotes also represented a relevant amount of the rumen core metagenome. This group has been reported to contribute up to 50% of total ruminal biomass [ 30 ]. The SAR supergroup and Fungi were the most relevant ones, which are found in a wide variety of ruminants and pseudoruminants [ 15 , 31 ]. Other eukaryotes included Stentor and Paramecium ; the former are aquatic free-living heterotricheans that can be particle filtrators or predators of other protozoa and live symbiotically with some algae species [ 32 , 33 ], whereas the latter are well-known ciliates that predate bacteria and other microorganisms, including protozoa [ 34 ]. Archaeal fraction was mostly composed of strict methanogenic organisms from Methanomicrobia and Methanobacteria clades [ 35 ] but also included Thermoplasmata, which are methylotrophic-methanogenic acidophilic organisms [ 36 ]. The DA analysis showed that ciliates, fungi, and pseudo-fungi were more abundant in cows with higher levels of methane emissions. Microbes associated with lower methane emissions were saccharolytic members of class Gammaproteobacteria ( Anaerobiospirillum [ 37 ], Vibrio [ 38 ], or Pseudoalteromonas [ 39 ]), as well as Negativicutes genera from Veillonellaceae ( Dialister, Megasphaera ) and Selenomonadaceae ( Mitsuokella ). Dialister produce succinate decarboxylation, and Megasphaera ferment carbohydrate and lactate [ 40 ], while Mitsuokella are saccharolytic bacteria [ 41 ]. The low-emissions ruminotype had larger abundance of Proteobacteria and Firmicutes genera. Other authors also reported higher abundances of these bacterial phyla in low methane emissions animals [ 8 ]. Lactate and succinate producers have been reported to be more abundant in low-emitters as well [ 42 ], supporting the higher abundance of Anaerobiospirillum or Megasphaera in LOW animals. Despite this association between methane and large taxonomic groups, it is of interest to find out which specific clades and microbial genes are participating directly or indirectly in methanogenesis. The genera co-abundance network showed a clear cluster of eukaryotes, with many of them being significantly more abundant in the high-emissions group. Other authors have already established a positive correlation between fungi abundance and methane emissions [ 8 ], as well as a close interdependence of protists and fungi. Although correlation between methane emissions and protozoa abundances is still under discussion [ 43 , 44 ], current meta-analyses point to a linear relationship between protozoa abundance and methane emissions ( r = 0.96) [ 45 ]. Interestingly, no taxonomic group of methanogenic archaea showed association with methane emissions. The relationship between Archaea and methane production in rumen is not consistent in the literature. Some authors reported either individual relationships between methane emissions and some archaeal species [ 46 , 47 ] or correlations between overall archaeal gene abundance and methane emissions level [ 43 , 48 ]. However, other studies showed no relationship between methanogenic Archaea and methane [ 47 , 49 ]. All studies to date showed a low relative abundance of archaea in the rumen, compared to eukaryotes and bacteria [ 50 ]. However, the association between the abundance of rumen eukaryotes and methane emissions has been demonstrated through defaunation experiments, both in vitro [ 51 , 52 ] and in vivo [ 44 , 53 ], with lower emissions in defaunated animals [ 54 ]. This has been attributed to the tight link existing between methanogenic archaea abundance and some fungi and protozoa [ 50 ]. Specifically, ciliates and some Chytridiomycota (e.g., Neocallimastix sp.) are known to symbiotically engulf a variety of methanogenic archaea. They provide the archaea with substrate for methane production from H 2 produced in their hydrogenosomes, as well as protection against oxygen toxicity [ 30 , 55 , 56 ]. Thus, free-living methanogens might represent a low fraction of microbial population [ 45 ], and CH 4 biosynthesis might be more influenced by endosymbiotic methanogens [ 55 ]. Hence, a larger methanogenesis activity is expected to be correlated with a larger abundance of eukaryotes, especially ciliates, which are more abundant and better represented. Another partial explanation for the low abundance of free archaea, and thereby for the lack of association between Archaea and methane emissions in previous studies [ 10 ], is that lysis of archaea cell walls often requires specific protocols during DNA extraction, and they might be under-represented in metagenomics studies [ 57 ]. In terms of Gene Ontology, the KEGGs were associated with several metabolic functions and cellular processes (nutrient metabolism and biosynthesis, cellular transport, cell growth, or genetic information processing). Pathways related to pathogenic activity were also found, in agreement with the RA of several genera that include known pathogenic species (e.g., Vibrio, Haemophilus, Trypanosoma , or Staphylococcus ), although not every species from these genera is pathogenic, but opportunistic or commensal organisms. Besides, pathogenic activity presence in our dataset might be biased owing to a larger representation of human-related diseases in the databases. The KEGGs were classified according to their presence or absence in ko00680 pathway (methane metabolism), as a way to evaluate their direct involvement in methanogenesis or an indirect involvement in pathways leading to biosynthesis of precursor compounds. Although we found several ko00680 KEGGs, which are presumably involved in the biosynthesis of methanogenesis precursors, most of them were not associated with methane emissions (i.e., not differentially abundant between methane groups). Most of these KEGGs were mainly present in bacteria or eukaryotes and might be functioning in metabolic pathways not related to methanogenesis. For instance, some of the KEGGs inside the methane metabolism pathway can also be involved in glycine, serine, and threonine metabolism (e.g., K00058, K00831, K01079, and K00600), pyruvate and propanoate metabolism (e.g., K00625 and K13788), glycolysis (e.g., K01689, K15633, K01624, and K02446), or anaerobic carbon fixation (e.g., K00198) [ 16–18 ]. Another group of ko00680 KEGGs is exclusive from Archaea, but the under-representation of this clade in our dataset might obscure statistical significance. Other detected KEGGs could be indirectly related to methanogenesis through biosynthesis of precursor compounds. For instance, K00209 and K13788 are involved in butyrate and propanoate biosynthesis, being essentially carried by primary fermentative bacteria [ 58 ]. Then the volatile fatty acids can be used by secondary fermenters to produce methanogenesis precursors such as H 2 , CO 2 , acetate, and formate [ 59 , 60 ]. In fact, K13788 is a phosphate acetyltransferase (EC:2.3.1.8) that can be involved in the biosynthesis of acetate from acetyl-CoA [ 61 ]. Also, K09251 is involved in biosynthesis of GABA and 2-oxoglutarate. GABA has been related to a volatile fatty acid concentration increment [ 62 ], while 2-oxoacid compounds can be used by Archaea to synthesize coenzyme M and coenzyme B, which are essential in methane production [ 63 ]. However, all these KEGGs were observed as overabundant in the LOW methane group, suggesting a strong presence of fermentative bacteria in these animals, not directly correlated with methane production. Other KEGGs that were overabundant in LOW emitters might offer an explanation of the lower presence of active methanogenesis processes through competence mechanisms (e.g., LOW-OA KEGGs K01682, K01902, and K13788 are involved in citrate cycle and pyruvate metabolism, related to respiration). The K00370 and K00371 are nitrate oxidoreductase subunits playing a role in anaerobic respiration using nitrate as electron acceptor. This enzyme uses nitrate as electron acceptor, a process that has been reported as a competitive inhibitor of methanogenesis [ 64 , 65 ]. Nitrate supplementation has proven to be an useful strategy to mitigate methane emissions [ 66 ]. Nitrite produced by the nitrate-reductases has a known antimicrobial effect and toxicity to animal cells [ 67–69 ], which might also reduce the proportion of free archaea in LOW animals, although toxicity to archaea must be further studied [ 70 ]. However, the role of ciliates and fungi must be clarified because their abundance is also lower in LOW emitters. We hypothesize that the predatory nature of these eukaryotes might be a control mechanism for bacterial populations, and their lower relative abundance in LOW animals might allow overgrowth of related bacteria. Nevertheless, there is the possibility that a higher proportion of facultative anaerobes using nitrate as acceptor might affect ciliate populations by toxicity, thus reducing the presence of endosymbiotic methanogenic archaea. The SqueezeMeta software [ 71 ] uses a last common ancestor (LCA) algorithm, which assigns to 1 read the lowest-level taxon common to all hits, using a stringent cut-off identity value for each taxonomic rank. On its part, functional assignments are done with the fun3 algorithm, which by default assigns the hit with the highest mean bitscore compared to the n first hits passing the e-value, identity, and coverage filters. This LCA approach ensures that reads have a large probability of being correctly classified, at the expense of a large number of reads remaining unclassified, which explains the larger number of reads assigned to a known KEGG than to taxa. Despite this strict requirement, this composition is consistent with other populations reported before [ 2 , 3 , 20 ]. Most studies to date report large abundance of Bacteroidetes and Firmicutes, with Prevotella spp . as the most prevalent genus. Some minor discrepancies with other studies were observed in the RA of the core subcomposition. For example, Wallace et al . [ 20 ] showed a higher presence of Proteobacteria and Euryarchaeota, although using amplicons instead of whole-metagenome sequencing. Our statistical approach evidenced the difficulty of inferring a phenotypic association between microbiome composition and methane production, with an important role of environmental factors that mask the statistical signal. However, a meaningful relationship between the microbiome composition and methane emissions could be uncovered yet, emphasizing the role of the different phyla, with the Eukaryota superkingdom being of particular relevance. Previous studies also revealed a link between ruminal microbiota and methane production. Difford et al . [ 3 ] showed different clusters of high and low methane emitters according to their bacterial and archaeal subcomposition. Danielsson et al . [ 46 ] also found clustering for low and high methane emitters within prokaryotic rumen subcompositions. Wallace et al . [ 20 ] found that a core set of rumen microbiome was capable of explaining up to 30% of methane emissions variability, mostly formed by prokaryotes. The aforementioned studies used different methodologies, like amplicon analysis and operational taxonomic unit clustering, contrasting with our full-metagenome genus-clustering protocol, which increases the information entropy. Stewart et al . [ 72 ] used Nanopore sequencing and found significant differences between low and high methane emitter sheep, with clear clustering between groups, but using a lower number of microbial groups and animals in the same farm with similar management practices."
} | 5,055 |
23441102 | null | s2 | 2,790 | {
"abstract": "Transient network hydrogels cross-linked through histidine-divalent cation coordination bonds were studied by conventional rheologic methods using histidine-modified star poly(ethylene glycol) (PEG) polymers. These materials were inspired by the mussel, which is thought to use histidine-metal coordination bonds to impart self-healing properties in the mussel byssal thread. Hydrogel viscoelastic mechanical properties were studied as a function of metal, pH, concentration, and ionic strength. The equilibrium metal-binding constants were determined by dilute solution potentiometric titration of monofunctional histidine-modified methoxy-PEG and were found to be consistent with binding constants of small molecule analogs previously studied. pH-dependent speciation curves were then calculated using the equilibrium constants determined by potentiometric titration, providing insight into the pH dependence of histidine-metal ion coordination and guiding the design of metal coordination hydrogels. Gel relaxation dynamics were found to be uncorrelated with the equilibrium constants measured, but were correlated to the expected coordination bond dissociation rate constants."
} | 295 |
37352836 | null | s2 | 2,791 | {
"abstract": "The gut microbiome is complex, raising questions about the role of individual strains in the community. Here, we address this question by constructing variants of a complex defined community in which we eliminate strains that occupy the bile acid 7α-dehydroxylation niche. Omitting Clostridium scindens (Cs) and Clostridium hylemonae (Ch) eliminates secondary bile acid production and reshapes the community in a highly specific manner: eight strains change in relative abundance by >100-fold. In single-strain dropout communities, Cs and Ch reach the same relative abundance and dehydroxylate bile acids to a similar extent. However, Clostridium sporogenes increases >1,000-fold in the ΔCs but not ΔCh dropout, reshaping the pool of microbiome-derived phenylalanine metabolites. Thus, strains that are functionally redundant within a niche can have widely varying impacts outside the niche, and a strain swap can ripple through the community in an unpredictable manner, resulting in a large impact on an unrelated community-level phenotype."
} | 260 |
37352836 | null | s2 | 2,792 | {
"abstract": "The gut microbiome is complex, raising questions about the role of individual strains in the community. Here, we address this question by constructing variants of a complex defined community in which we eliminate strains that occupy the bile acid 7α-dehydroxylation niche. Omitting Clostridium scindens (Cs) and Clostridium hylemonae (Ch) eliminates secondary bile acid production and reshapes the community in a highly specific manner: eight strains change in relative abundance by >100-fold. In single-strain dropout communities, Cs and Ch reach the same relative abundance and dehydroxylate bile acids to a similar extent. However, Clostridium sporogenes increases >1,000-fold in the ΔCs but not ΔCh dropout, reshaping the pool of microbiome-derived phenylalanine metabolites. Thus, strains that are functionally redundant within a niche can have widely varying impacts outside the niche, and a strain swap can ripple through the community in an unpredictable manner, resulting in a large impact on an unrelated community-level phenotype."
} | 260 |
39987198 | PMC11846464 | pmc | 2,793 | {
"abstract": "Background Methane emission from enteric rumen fermentation is a main source of greenhouse gas (GHG) emission and a major concern for global warming. Results In this study, we isolated methanotroph-methylotroph consortium NC52PC from the rumen after a series of sub-culture and repetitive streaking on an agar plate and polycarbonate membrane filter. The NC52PC comprises methanotroph species ( Methylocystis sp.) and methylotroph species ( Methylobacterium sp.), forming a consortium capable of growing solely on methane as a carbon source. Their morphology, growth, and genome sequence were characterized. We assessed its effectiveness in mitigating methane emissions through both in vitro and in vivo experiments. During the in vitro trial, the introduction of NC52PC (at a concentration of 5.1 × 10 7 CFUs/ml) demonstrated a reduction in methane production exceeding 40% and 50% after 12 and 24 h, respectively. Also, NC52PC did not significantly alter other aspects of the in vitro rumen fermentation parameters such as pH, total gas production, and digestibility. Further investigation involved testing NC52PC as a dietary supplement in 12 young Hanwoo steers over three 30-day test periods. The steers received a diet comprising 70.8% concentrate and 29.2% bluegrass on a dry matter basis, with variations including 3 × 10 7 CFUs/ml of NC52PC ( LOW ) and 3 × 10 8 CFUs/ml ( HIGH ) of NC52PC, and without NC52PC as a control ( CON ). Steers administered with HIGH and LOW concentrations of NC52PC exhibited reduced enteric methane emission (g/day) by 14.4% and 12.0%, respectively. Conclusion Feeding methanotroph-methylotroph consortium NC52PC significantly reduced methane emissions in Korean beef cattle without any adverse effects on animal health. These findings suggest that this probiotic could serve as a promising feed additive to effectively mitigate methane emissions from ruminants. However, further research is needed to evaluate the long-term effects of NC52PC on animal health, and on meat and milk quality. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-025-00385-0.",
"conclusion": "Conclusion \nThis research aims to provide a novel approach by utilizing methanotrophs as potential probiotics to primarily reduce enteric methane emissions without negatively impacting the ruminal ecosystem. Our results show that methane emission was reduced by over 14% when 12 Hanwoo steers were administered with 3 × 10 8 CFUs/mL of methanotroph-based probiotics for two weeks without adversely impacting overall animal health. To the best of our knowledge, this is the world’s first study on the isolation of methanotrophs from the rumen, and the successful application of methanotroph-based probiotics to reduce methane emission in cattle. The methanotroph-based probiotics hold tremendous potential to mitigate methane emissions from ruminants and could serve as a promising feed additive to combat climate change. Despite a significant methane reduction, further study is required to evaluate the long-term effect of methanotroph-based probiotics on methane emission and overall animal productivity.",
"discussion": "Discussion \nRuminal microorganisms play an important role in the metabolic processes of ruminants by breaking down complex feedstuffs into volatile fatty acids, which provide up to 70% of the ruminant’s energy requirements [ 32 ]. Methane is generated as a byproduct of this microbial fermentation process that not only contribute to anthropogenic greenhouse gas emissions and enlarge the carbon footprint of dairy or beef production but also deplete nutritional energy [ 8 , 33 ]. Various strategies have been explored to reduce enteric methane emission. Here, we exploited the potential of methane metabolizing microbes to mitigate methane emission in ruminants. Methanotrophs are ubiquitous in either anoxic or aerobic environments and have been previously enriched but were never applied in vitro or in vivo rumen fermentation systems. The cannulated Holstein Friesian cows were used in the in vitro setting of this study, as rumen cannulation is widely recognized as the reference method for obtaining representative samples of rumen digesta from donor animals [ 34 , 35 ]. For the in vivo experiments, the oral stomach tube technique was employed on Hanwoo steers, as this method was suitable for collecting liquid fractions only, whereas sampling via rumen cannula allows for the collection of both solid and liquid digesta fractions [ 36 ]. In this study, aerobic methanotrophs were isolated from a rumen sample. The aerobic methanotrophs are likely present due to oxygen entering the rumen via diffusion across the epithelium [ 37 ]. After a series of sub-culture and repetitive re-streaking on agar plates, colonies were transferred to a polycarbonate membrane to ensure purity and minimize heterotrophic contamination. Among three isolates, NC52PC robustly grew at 39 °C, making it the best candidate for further characterization and testing under in vitro and in vivo rumen fermentation setup. \nFurther morphological and genomic analysis revealed that NC52PC consisted of two bacterial strains; one belonging to methanotrophic group ( Methylocystis sp. NC52PC) and other belonging to methylotrophic group ( Methylobacterium organophilum NC52PC). Methanotrophs and methylotrophs have often coexist in nature [ 38 ]. Methylotrophs can metabolize the excess methanol formed from the methane oxidation of methanotrophs, thereby reducing methanol toxicity and enhancing the growth of methanotrophs [ 39 ]. There is also a possibility of essential nutrient exchange between Methylocystis and Methylobacterium species that can drive the overall growth performance of this consortium [ 40 ]. \nUnder anaerobic in vitro rumen fermentation, NC52PC decreased methane production by approximately 50% after 24 h of incubation. This substantial suppression of generated methane in vitro makes NC52PC a potential candidate for in vivo testing. However, since NC52PC primarily requires oxygen for growth and methane oxidation, methane reduction was tested in NMS-Cu using 20% methane under anaerobic conditions to confirm further NC52PC’s ability to grow and consume methane without oxygen. Similar to the in vitro fermentation test, we inoculated NC52PC in NMS-Cu media with the final concentration of 5 × 10 7 CFUs/mL. Results show that methane concentration reduced from 20 to 18% after 48 h, indicating NC52PC’s ability to oxidize approximately 2% (equivalent to 20,000 ppm) methane in 48 h under anaerobic conditions (Figure S4 ). This highlights the versatility of aerobic methanotrophic NC52PC, such as its ability to oxidize methane despite the steady lack of oxygen supply. Oxidation of methane by aerobic methanotrophs under an anaerobic environment is possible by exploiting other alternative electron acceptors in the rumen content. Members of the Methylomonadaceae and Methylocystaceae family have been shown to utilize nitrate/nitrite- or mineral oxide-dependent methane oxidation under oxygen limitation [ 41 – 44 ]. NC52PC may have evolved to utilize denitrification or mineral reduction processes in an anoxic environment such as rumen. Finally, we assessed the efficacy of NC52PC in reducing methane emission in Hanwoo steers for a 3-cycle 30-day period. NC52PC when fed as a methanotroph-based probiotic at a concentration of 3 × 10 8 CFUs/ml significantly lowered methane emission by 14.4% compared to the control group without negatively impacting animal growth. Although methane reduction exceeded 50% during in vitro rumen fermentation, the in vivo experiment showed only about a 14% reduction. This discrepancy may be due to the amount of methanotrophs supplied in vivo, which was roughly 1,000 times less, considering the rumen size and the final methanotroph concentration. The pmoA gene copy number observed in vitro (Fig. 4 E) compared to in vivo rumen fluid samples (Fig. 7 D) further highlights the significant difference in methanotroph concentration. We hypothesize that matching in vivo concentrations to in vitro levels could significantly boost methane consumption. Future studies will focus on optimizing delivery methods and dosages to achieve these higher in vivo concentrations and investigate the kinetics of NC52PC in the complex rumen environment, including factors such as passage rate and competition with other microbial populations. This methanotroph-based probiotic holds immense potential as a sustainable feed additive to effectively reduce methane emissions from ruminants. However, the evaluation of long-term effects of NC52PC on animal health and productivity will be our future goal."
} | 2,176 |
36769958 | PMC9918016 | pmc | 2,794 | {
"abstract": "Skin is the largest organ of many animals. Its protective function against hostile environments and predatorial attack makes high mechanical strength a vital characteristic. Here, we measured the mechanical properties of bass fish skins and found that fish skins are highly ductile with a rupture strain of up to 30–40% and a rupture strength of 10–15 MPa. The fish skins exhibit a strain-stiffening behavior. Stretching can effectively eliminate the stress concentrations near the pre-existing holes and edge notches, suggesting that the skins are highly damage tolerant. Our measurement determined a flaw-insensitivity length that exceeds those of most engineering materials. The strain-stiffening and damage tolerance of fish skins are explained by an agent-based model of a collagen network in which the load-bearing collagen microfibers assembled from nanofibrils undergo straightening and reorientation upon stretching. Our study inspires the development of artificial skins that are thin, flexible, but highly fracture-resistant and widely applicable in soft robots.",
"conclusion": "4. Conclusions In summary, the mechanical properties of fish skins were characterized by testing skin strips in tension. We found that the fish skins become stiffer with stretching. More importantly, the fish skins are highly damage tolerant. This was demonstrated by the phenomena that stretching can eliminate the stress concentration near the holes and edge notches but cannot propagate or expand these defects. In particular, the flaw-sensitivity length for the fish skin was calculated to be in the order of 10 mm, superior to most engineering materials. We attribute strain stiffening and the superior damage tolerance to the elementary collagen network movements, namely, microfiber straightening and reorientation, as verified by our agent-based modeling. The fundamental understanding of our study also provide guidance to the synthesis of multifunctional artificial skins [ 27 ] for soft robots that are mechanically robust and damage tolerant.",
"introduction": "1. Introduction Skin is a multifunctional organ that provides protection to animals from their living environment by regulating body temperature and sensing external stimuli [ 1 ]. Skin is made of three layers: epidermis, the top layer; dermis, the middle layer; and hypodermis, the bottom layer [ 1 ]. The epidermis is the water-resistant outer layer of skin and the body’s first line of defense against environmental attacks, ultraviolet radiation, bacteria, and other pathogens. It is also responsible for cell renewal. The dermis is responsible for the structure and mechanical properties of skin and mostly composed of the proteins collagen and elastin. Collagen is the major load-bearing structure [ 2 , 3 ], while elastin provides flexibility for the skin [ 4 ]. The lower layer of the dermis, the stratum compactum, consists of an orthogonal cross-ply arrangement of collagen microfibers with well-defined angles (~40–60°) relative to the long axis (length) of the fish [ 5 ]. For scaled skins, each scale is typically embedded in the dermis and projected out through the epidermis and contributes to the puncture resistance of the skin [ 6 , 7 ]. The hypodermis is the layer of skin where fat is deposited and stored. The presence of multiple layers allows the skin to achieve many functions. The robust mechanical properties of skin are essential to ensure its function as a self-healing protective layer after being subject to mechanical loads such as abrasion and tearing. The skins of different animals have been studied [ 8 , 9 , 10 , 11 ], which are often considered to be nonlinearly elastic. Far less is known about the plasticity and damage tolerance of skin. Because skin is a polymeric structure with a hierarchical order [ 4 , 10 ], plastic deformation likely originates from the rearrangement of the collagen microfiber network. Indeed, Diamant et al. [ 12 ] and Fratzl et al. [ 13 ] noted the uncrimping of rats’ tail collagen microfibers with increasing tensile loading at small tensile strains of around 0–3%. How collagen microfibers in skin rearrange at high strains of tens or hundreds of percent, and their effect on mechanical properties at these strains, however, has been previously unexplored. In this study, we characterize the mechanical properties of fish skins under tensile loading. Fish skin must endure the harsh aquatic environment in very dynamic motions and processing damage tolerant skins is essential for survival. How collagen microfibers in skin rearrange at high strains, and these rearrangement leads to superior mechanical properties and damage tolerance, have been previously unexplored. It is well characterized that the mechanical properties of fish skins are location dependent, as experimentally observed for tensile stress–strain curves of striped bass skin [ 5 ] and Chinese sturgeon skin [ 14 ]. However, our study here focuses on the damage tolerance mechanism of the skin in general, not on specific locations. We hypothesize that though the mechanical properties are location dependent, the deformation mechanism and damage tolerance strategy are location invariant. Our measurements show that fish skin is highly ductile, nonlinear, and exhibits a strain-stiffening behavior. Loading and unloading cycles show that fish skin undergoes large plastic deformation with significant hysteresis. Stretching can hardly lead to the expansion of pre-cut holes and edge notches on the fish skin; instead, stretching effectively eliminates the stress concentration near the defects, demonstrating that the skin is highly damage tolerant. To explore the microscale mechanism behind the macroscale mechanical properties, an agent-based model was developed. The out-of-plane crimping and the in-plane reorientation of collagen microfibers are widely regarded as the origin of the nonlinear elastic response of collagen network [ 3 , 10 , 12 , 13 , 15 ]. Although molecular dynamics simulations provide the mechanical response of straightening for a single collagen fibril [ 16 ], the tissue scale microfiber movement under tensile test has not been investigated. Considering that the collagen network is the major load-bearing structure in the skin, we develop an agent-based model of the collagen network. Our agent-based simulations show that the superior mechanical properties of fish skins can be attributed to the rearrangement of the collagen microfibers network upon stretching."
} | 1,614 |
39443794 | PMC11655356 | pmc | 2,795 | {
"abstract": "The ability of stony corals to thrive in the oligotrophic (low-nutrient, low-productivity) surface waters of the tropical ocean is commonly attributed to their symbiotic relationship with photosynthetic dinoflagellates 1 , 2 . The evolutionary history of this symbiosis might clarify its organismal and environmental roles 3 , but its prevalence through time, and across taxa, morphologies and oceanic settings, is currently unclear 4 – 6 . Here we report measurements of the nitrogen isotope ( 15 N/ 14 N) ratio of coral-bound organic matter (CB-δ 15 N) in samples from Mid-Devonian reefs (Givetian, around 385 million years ago), which represent a constraint on the evolution of coral photosymbiosis. Colonial tabulate and fasciculate (dendroid) rugose corals have low CB-δ 15 N values (2.51 ± 0.97‰) in comparison with co-occurring solitary and (pseudo)colonial (cerioid or phaceloid) rugose corals (5.52 ± 1.63‰). The average of the isotopic difference per deposit (3.01 ± 0.58‰) is statistically indistinguishable from that observed between modern symbiont-barren and symbiont-bearing corals (3.38 ± 1.05‰). On the basis of this evidence, we infer that Mid-Devonian tabulate and some fasciculate (dendroid) rugose corals hosted active photosymbionts, while solitary and some (pseudo)colonial (cerioid or phaceloid) rugose corals did not. The low CB-δ 15 N values of the Devonian tabulate and fasciculate rugose corals relative to the modern range suggest that Mid-Devonian reefs formed in biogeochemical regimes analogous to the modern oligotrophic subtropical gyres. Widespread oligotrophy during the Devonian may have promoted coral photosymbiosis, the occurrence of which may explain why Devonian reefs were the most productive reef ecosystems of the Phanerozoic."
} | 442 |
26690249 | PMC4687082 | pmc | 2,796 | {
"abstract": "Background Genomic uptake of DNA by prokaryotes often encompasses more than a single gene. In many cases, several horizontally transferred genes may be acquired together. Accordingly, we expect that horizontally transferred genes cluster spatially in the genome more often than expected if transfers were independent. Further, genes that depend on each other functionally may be unlikely to have beneficial fitness effects when taken up individually by a foreign genome. Hence, we also expect the co-acquisition of functionally related genes, resulting in the clustering of horizontally transferred genes in functional networks. Results Analysing spatial and metabolic clustering of recent horizontal (or lateral) gene transfers among 21 γ-proteobacteria, we confirm both predictions. When comparing two datasets of predicted transfers that differ in their expected false-positive rate, we find that the more stringent dataset shows a stronger enrichment of clustered pairs. Conclusions The enrichment of interdependent metabolic genes among predicted transfers supports a biologically significant role of horizontally transferred genes in metabolic adaptation. Our results further suggest that spatial and metabolic clustering may be used as a benchmark for methods that predict recent horizontal gene transfers. Reviewers This article was reviewed by Peter Gogarten in collaboration with Luiz Thiberio Rangel, and by Yuri Wolf.",
"conclusion": "Conclusions The observed degree of clustering in general and its dependence on the likely false-positive rate within a candidate set suggests that a certain degree of clustering is a typical feature of reliable candidate sets of recently transferred genes. The spatial clustering is consistent with the horizontal co-transfer of neighbouring genes on a continuous stretch of DNA; the mutational process of HGT does not preferentially transfer individual genes [ 11 ], and thus co-transfer is likely if the transferred piece of DNA is large enough. The enrichment of functionally related gene pairs in HGT candidate sets supports the role of HGT in bacterial adaptation; however, because functionally related genes often reside in the same operon, functional and spatial clustering may not be fully independent. The observed patterns suggest that a quantification of (genomic or functional) clustering may be used as a quality measure for methods that aim to identify horizontally transferred genes: candidate sets produced by different methods (or parameter settings) can be tested against each other. This benchmarking approach may be most useful for methods that detect relatively recent transfers. Furthermore, its application requires that the expected age distributions of HGT events are similar between compared methods, as the strength of clustering likely decreases with increasing age. Note that this method is only suited to compare false discovery rates, but does not inform about the sensitivity; a dataset constructed using more stringent parameters will likely have fewer false positive predictions of HGT, but may also have more false negatives. For a meaningful comparison of two methods, one should thus also take the absolute numbers of HGT predictions into account. All HGT detection methods that involve tuneable parameters are expected to show a trade-off between sensitivity and specificity, resulting in a negative correlation between size and positive predictive value of HGT candidate sets, just as observed between our terminal.pen1 and terminal.pen2 datasets. This phenomenon might be used to select appropriate parameters for a given study.",
"discussion": "Results and discussion To identify recent horizontal transfers, we used previously published sets of HGT candidates [ 22 ]. In this previous work, gene presence/absence data for orthologous gene families was projected onto the terminal nodes of a well-resolved phylogenetic tree representing vertical inheritance from an ancestral species to 21 extant γ-proteobacteria. Horizontal gene transfers were identified based on the most parsimonious explanation of gene presence/absence [ 4 ]. Of 42,677 examined genes, 2,020 (equal penalties for gene gains and losses, terminal.pen1 ; see [ 4 , 22 ] for details) and 961 (higher penalties for gene gains than for losses, terminal.pen2 ) putative horizontal gene transfers were mapped to the terminal branches of the phylogeny. Based on our null model of no association between HGT and the distance of genes along the chromosome, we would expect to find 540 and 145 pairs of genes that are genomic neighbours, respectively, in the two datasets. We actually observe 882 and 401, corresponding to a 1.6- and 2.8-fold enrichment, respectively. Both enrichment values are statistically highly significant (Table 1 ). Table 1 Clustering scores ( CC ) for spatial clustering of horizontally transferred genes across all examined genomes. Two horizontally transferred (HGT) genes were considered genomic neighbours if they had at most 2 intervening genes between them Result set # HGT candidates Expected pairs Observed pairs \n CC \n \n p \n terminal.pen1 2020 540.17 882 1.633 <0.001 terminal.pen2 961 144.88 401 2.768 <0.001 The Escherichia coli K12 genome contains 205 ( terminal.pen1 ) and 85 ( terminal.pen2 ) genes predicted to be the result of recent horizontal transfers. If metabolic interactions were independent of HGT status, we would expect these candidates to contain about 5 and 1 pairs of interacting genes, respectively. We actually observe 24 and 0, respectively (Table 2 ). For the first, larger set of HGT predictions, this corresponds to a 5-fold enrichment, which is statistically highly significant ( p < 0.001). The second dataset, with a null expectation of 1 pair, is too small to assess statistical significance. Table 2 Clustering scores ( CC ) for metabolic clustering of genes recently transferred horizontally into E. coli K12. Two horizontally transferred (HGT) genes were considered metabolic neighbours if they encode reactions that catalyse tightly correlated fluxes in the E. coli K12 metabolic network Method # HGT candidates Expected pairs Observed pairs \n CC \n \n p \n terminal.pen1 205 4.68 24 5.128 <0.001 terminal.pen2 85 0.80 0 0.000 0.76 Our results show clear evidence for a clustering of horizontally transferred genes, both spatially and functionally. The observed clustering scores, quantifying the degree of departure from our null models of no association, range from 1.6 to 5.1. This analysis is based on rather conservative assumptions: only chromosomally encoded genes are included, only the most recent (therefore most reliably inferred) transfer events are taken into account, and genes without orthologs among the examined genomes are discarded. Finally, our statistical model biases the inferred clustering scores downwards and the inferred p values upwards, as it does not account for the gaps between potential HGT candidates induced by excluded genes. We have chosen a method of generalised parsimony for the detection of HGT events, as this method does not rely on local sequence features, which are known to vary systematically with chromosomal position [ 23 ]. As the only information used for our classification is the presence or absence of orthologs in other genomes, it seems very unlikely that two neighbouring genes’ probabilities to be classified as horizontally transferred are correlated due to methodological biases. We found that the clustering score for spatial clustering of the more ‘conservative’ and smaller data set ( terminal.pen2 ) is, at 2.8, substantially increased compared to the clustering score of the larger set ( terminal.pen1 ), at 1.6. This increase is consistent with a true biological relationship between horizontal gene transfer and spatial gene clustering. The terminal.pen2 dataset is expected to contain less false-positive HGT candidates than terminal.pen1 (due to the increased gain/loss penalty ratio, stronger support is required before a gene is considered to result from HGT): as the fraction of false positives decreases, the strength of clustering grows. Our results can be interpreted as strong evidence for horizontal co-transfer, e.g., by the uptake of a complete or partial operon by the host genome. Alternatively, it is conceivable that certain genomic regions (“islands”) may be more suitable for the integration of foreign DNA [ 8 ], either due to a mutational bias such as a specific nucleotide composition, or due to a selective bias."
} | 2,125 |
21255333 | PMC3158429 | pmc | 2,798 | {
"abstract": "Summary The global regulator H‐NS of Escherichia coli controls genes related to stress response, biofilm formation and virulence by recognizing curved DNA and by silencing acquired genes. Here, we rewired H‐NS to control biofilm formation using protein engineering; H‐NS variant K57N was obtained that reduces biofilm formation 10‐fold compared with wild‐type H‐NS (wild‐type H‐NS increases biofilm formation whereas H‐NS K57N reduces it). Whole‐transcriptome analysis revealed that H‐NS K57N represses biofilm formation through its interaction with the nucleoid‐associated proteins Cnu and StpA and in the absence of these proteins, H‐NS K57N was unable to reduce biofilm formation. Significantly, H‐NS K57N enhanced the excision of defective prophage Rac while wild‐type H‐NS represses excision, and H‐NS controlled only Rac excision among the nine resident E. coli K‐12 prophages. Rac prophage excision not only led to the change in biofilm formation but also resulted in cell lysis through the expression of toxin HokD. Hence, the H‐NS regulatory system may be evolved through a single‐amino‐acid change in its N‐terminal oligomerization domain to control biofilm formation, prophage excision and apoptosis.",
"introduction": "Introduction Biofilm formation converts single cells into a complex heterogeneous community ( Stewart and Franklin, 2008 ) attached to a surface and requires precise regulation of many genes ( Karatan and Watnick, 2009 ). For example, genes related to stress response, quorum sensing (QS), motility, fimbriae, metabolism and transport are differentially regulated in Escherichia coli biofilms ( Domka et al. , 2007 ). It is important to control biofilm formation for engineering and medical applications such as reducing corrosion ( Jayaraman et al. , 1999 ), facilitating remediation ( Wood, 2008 ) and reducing disease ( Jayaraman and Wood, 2008 ). The histone‐like nucleoid structuring protein H‐NS is widely conserved in Gram‐negative bacteria ( Tendeng and Bertin, 2003 ) and is a global regulator that represses transcription ( Dorman, 2004 ) by recognizing intrinsically curved DNA sequences ( Rimsky, 2004 ). H‐NS is a small protein (137 amino acids) that is very abundant with more than 20 000 copies per cell ( Rimsky, 2004 ). H‐NS regulates the transcription of many environmental responsive genes; for example, H‐NS represses the locus of enterocyte effacement (LEE) in enteropathogenic and enterohemorrhagic E. coli (EHEC) by binding LEE regulatory DNA ( Mellies et al. , 2007 ) and decreases resistance to high osmolarity and low pH in E. coli K‐12 ( Hommais et al. , 2001 ). H‐NS consists of an N‐terminal oligomerization domain (1–64 aa), a C‐terminal DNA‐binding domain (90–137 aa), and a flexible linker (65–89 aa) between both domains ( Dorman, 2004 ). Interactions between the N‐terminus of H‐NS and other nucleoid‐associated proteins enhance their activities ( Fang and Rimsky, 2008 ) as H‐NS increases repression by Hha of the haemolysin operon ( Madrid et al. , 2007 ) and protects StpA from Lon‐mediated proteolysis, which results in increased viability of stationary phase cells ( Johansson and Uhlin, 1999 ). To date, little is known about H‐NS and its effect on biofilm formation; however, H‐NS regulates genes related to biofilm formation in a temperature‐dependent manner ( White‐Ziegler and Davis, 2009 ), and the deletion of hns decreases biofilm formation ( Belik et al. , 2008 ). H‐NS also silences genes acquired from lateral transfer in that H‐NS recognizes foreign DNA with AT‐rich content compared with the resident genome ( Navarre et al. , 2007 ). Prophage genes are obtained from lateral transfer and are common in most bacterial genomes contributing as much as 10–20% of a bacterium's genome ( Casjens, 2003 ). Escherichia coli K‐12 has six cryptic prophage and three prophage‐like elements ( http://www.ecogene.org/ ), which have lost some functions essential for lytic growth such as excision, tail formation and the production of phage particles, yet these loci retain some functional genes ( Blattner et al. , 1997 ). H‐NS completely or partially binds prophage and prophage‐like DNA in E. coli ( Oshima et al. , 2006 ), and the H‐NS–Hha complex tightly silences foreign DNA ( Baños et al. , 2009 ). However, the specific function of H‐NS for prophage gene regulation remains unclear. Previously, we discovered that the global regulator Hha decreases biofilm formation and regulates the cryptic prophages CP4‐57 and DLP12 ( García‐Contreras et al. , 2008 ). Deletions of single genes of these prophages increased or decreased biofilm formation significantly, and Hha induced excision of CP4‐57 ( Wang et al. , 2009 ). Although Hha decreases biofilm formation ( García‐Contreras et al. , 2008 ) and H‐NS increases biofilm formation ( Belik et al. , 2008 ), Hha can be functionally equivalent to the N‐terminal domain of H‐NS as demonstrated by a chimeric Hha–H‐NS protein with Hha fused to the N‐terminus of H‐NS, which complements some of the hns ‐induced phenotypes ( Rodríguez et al. , 2005 ). Thus, both Hha and H‐NS control prophage excision as well as influence biofilm formation. Directed evolution is a useful tool to engineer proteins for industrial applications as well as to explore natural evolutionary processes ( Otten and Quax, 2005 ). Since cell communication and biofilm formation are important for bacterial survival in microbial consortia ( Jayaraman and Wood, 2008 ), bacteria may readily evolve global regulators for enhanced fitness. Illustrating this concept, we evolved the E. coli QS regulator SdiA to control biofilm formation via the extracellular signals indole and N ‐acylhomoserine lactone ( Lee et al. , 2009 ). The aims of this study were to investigate how the global regulator H‐NS influences biofilm formation and to determine whether H‐NS may be evolved to reduce biofilm formation as well as to derive new insights for how the structure of H‐NS impacts its function. We identified an H‐NS variant with a single‐amino‐acid substitution in the N‐terminal oligomerization domain that reduces biofilm formation as a result of its interaction with nucleoid‐associated proteins Cnu and StpA, that induces Rac prophage excision, and that induces toxin HokD which leads to cell lysis.",
"discussion": "Discussion In this study, we demonstrate that the global regulator H‐NS of E. coli may be evolved to control three important cellular phenotypes: biofilm formation, Rac prophage excision and cell lysis. We screened biofilm variants in the absence of endogenous Hha and H‐NS and identified an H‐NS variant with a single‐amino‐acid change, H‐NS K57N, which reduces biofilm formation by 10‐fold ( Fig. 1 ). The mechanism by which H‐NS K57N reduces biofilm formation is through its interaction with Cnu and StpA (Table S2 and Fig. 2A and B ). Cnu is an Hha paralogue, and StpA is an H‐NS paralogue, and both are nucleoid‐associated proteins. Here, we also show that Cnu reduces biofilm formation like Hha. Both Cnu and StpA have the ability to associate with H‐NS by interacting with the N‐terminus of H‐NS ( Bae et al. , 2008 ; Leonard et al. , 2009 ). Therefore, our results imply that the altered H‐NS may interact closely with Cnu and StpA to reduce biofilm reduction. The N‐terminus of H‐NS includes three helical segments, H1 (1–8 aa), H2 (12–19 aa) and H3 (23–47 aa) ( Fang and Rimsky, 2008 ), and plays an important role in forming homo‐ or hetero‐oligomers of H‐NS with Hha, Cnu and StpA. In this study, the K57N substitution of H‐NS occurred at a filament‐like region in the N‐terminus which is near a flexible linker, but not in the helical segments. This substitution replaces the positively charged, long side‐chain of lysine with a neutrally charged, short side‐chain of asparagine. The K57N substitution of H‐NS may influence oligomerization of H‐NS, cause a conformational change of the flexible linker, or cause a conformational change of the C‐terminus, all of which would affect the DNA binding properties of H‐NS. It is interesting that only a single‐amino‐acid substitution causes such dramatic physiological changes in the cell (biofilm formation, Rac prophage excision and cell lysis). The role of H‐NS is to silence foreign DNA ( Navarre et al. , 2007 ) by binding to curved DNA which is commonly found at promoters ( Dorman, 2004 ). Prophages are genes laterally acquired that play an important role in the diversification of the genome ( Oshima et al. , 2006 ), and repression of the Rac prophage excision by H‐NS is the first report of this activity by H‐NS (Fig. S1A). Moreover, we found that the reconfigured H‐NS K57N protein induces Rac prophage excision, which is opposite to wild‐type H‐NS ( Fig. 3B ). When Rac was deleted from the chromosome, it influenced biofilm formation by increasing biofilms as early as 13 h (BW25113 Δ rac versus wild‐type BW25113) and by decreasing biofilms with the deletion of hha and hns (BW25113 hha hns Δ rac versus BW25113 hha hns ) (Fig. S2). Also, deletion of Rac enhanced cell lysis by H‐NS K57N ( Fig. 5 ). Note Rac is one of the most excisable prophages in E. coli K‐12 (Fig. S1A). By regulating excision and integration of Rac prophage via H‐NS, E. coli cells may enhance fitness in response to the environmental changes in terms of biofilm formation and cell lysis. H‐NS K57N induces cell lysis by regulating a small toxic membrane protein HokD ( Fig. 5 ). Previously, the function of HokD in E. coli was unclear, since the regulation element of the upstream part of hokD seems to be missing compared with hok in the R1 plasmid ( Pedersen and Gerdes, 1999 ). However, our whole‐transcriptome study and qRT‐PCR results show that HokD is induced by H‐NS K57N (Table S2), that HokD is required for H‐NS K57N to increase cell lysis ( Fig. 5 ), and that hokD expression is related to Rac excision. Rac prophage may repress cell lysis by inhibiting HokD when Rac is in the chromosome, but if E. coil experiences some environmental stress, cells may kill themselves by expressing toxin proteins such as HokD that is induced due to Rac excision. The protein engineering of H‐NS to form variant K57N thus represents one of the first examples of creating a global regulator to control prokaryotic apoptosis. H‐NS represses virulence genes in EHEC ( Mellies et al. , 2007 ), and several virulence genes are located in prophages; for example, Shiga toxin 1 ( stx1 ) and 2 ( stx2 ) genes are in cryptic prophage CP‐933V and bacteriophage BP‐933W, respectively, in E. coli O157:H7 EDL933 ( Perna et al. , 2001 ). CP‐933R in O157:H7 EDL933, Sp10 in O157 Sakai and Rac in K‐12 are located at identical positions in the three genomes and have identical attachment sites ( Casjens, 2003 ). Our preliminary data show that transcription of virulence genes, stx1A in CP‐933V, stx2A in BP‐933W and espB in LEE, is repressed up to eightfold by producing wild‐type H‐NS, while stx1A is induced by producing H‐NS K57N in EHEC. These results indicate that H‐NS and the evolved H‐NS influence expression of virulence genes in EHEC. Since wild‐type H‐NS represses the excision of Rac prophage (Fig. S1A) and H‐NS K57N increases excision (Fig. S1B), it is possible that H‐NS may affect virulence by controlling prophage excision in EHEC. In this study, we demonstrate that the global regulator H‐NS may be evolved readily, which is similar to our results in which we evolved the QS signal regulator SdiA of E. coli to respond to homoserine lactones and to control indole concentrations ( Lee et al. , 2009 ). Evolution of global regulators may cause widespread changes in the regulatory system in bacteria, including apoptosis, prophage excision and biofilm formation as we show here; hence, this study implies that bacteria may evolve global regulators for the beneficial use of foreign genes that were originally introduced by prophage."
} | 2,980 |
32218956 | PMC7029940 | pmc | 2,799 | {
"abstract": "A computational model has been developed to predict the role of environment in the forms and functions of termite mounds. The proposed model considers the most relevant forces involved in the heat transfer process of termite mounds, while also reflecting their gas-exchange function. The method adopts a system configuration procedure to determine thermally optimized mound structures. The model successfully predicts the main architectural characteristics of typical Macrotermes michaelseni mounds for the environmental conditions they live in. The results indicate that the mound superstructure and internal condition strongly depend on the combined effect of environmental forces. It is noted that mounds being exposed to higher solar irradiances develop intricate lateral channels, inside, and taller and more pronounced spire tilt towards the Sun, outside. It is also found that the mounds' spire tilt angle depends on the geographical location, following the local average solar zenith angle for strong irradiances. Although wind does not influence the overall over-ground mound shape, it significantly affects the mound internal condition. The results of this study resonate with what is seen in nature. The proposed approach provides a broader view of the factors that are effective in the form and function of a naturally made structure.",
"conclusion": "6. Conclusion In this work, a computational model has been developed to predict the effect of environmental conditions on the form and function of a natural structure—termite mound. The proposed model considers the most relevant forces involved in the heat transfer process of termite mounds and adopts a system configuration procedure to determine an optimized mound structure. Due to the availability of the information, the study used the structural characteristics of the mounds of M. michaelseni termites. However, the proposed methodology can be applied to any natural structure, where thermodynamics and heat transfer play an important role in their functions. A series of simulations were conducted to confirm the accuracy of the model under the typical environmental conditions of M. michaelseni termite habitat in northern Namibia. The optimal configuration that arose from the model was found to be strikingly similar to those observed in nature, including the mound spire being inclined towards the Sun—as noted in Turner's observation [ 34 ]. The study was then extended to examine the mound configurations under different values of wind, solar irradiance and zenith angle (representative of geographical latitude). The results indicate that the mound superstructure and its internal condition strongly depend on the combined effect of all environmental forces. In general, different amounts of exposure to direct insolation lead to dissimilar internal and external mound features. Mounds being exposed to higher solar irradiances develop wider lateral channels, inside, and taller and more pronounced spire tilt, outside. Moreover, for extremely high solar irradiances, the mound points exactly to the average zenith angle, indicating that the spire inclination is a consequence of thermal processes within the mound. This is in agreement with the hypothesis proposed by Turner [ 34 ] and with reported field observations. In nature, M. michaelseni mounds located in northern Namibia (19°S) tend to have a prominent spire tilt of 19.6° [ 18 ], while mounds studied in Kajiado, Kenya (1°S), do not show any significant tilt northwise [ 4 ]. Additionally, we determined that mounds located closer to the equator line tend to be taller and narrower in their upper structure, which matches field observations. For example, Macrotermes jeaneli mounds in southern Ethiopia (3°N) [ 35 ], reach, in average, 2–5 m in height, presenting a distinctive upright, narrow spire and a large base. By contrast, mounds that are located further away from the equator line were found to have more compact configurations, with shorter and wider structures. For example, this is the case for the M. michaelseni in northern Namibia (19°S) [ 18 ] (around 2 m high), as well as the Conitermes cumulans and Syntermes dirus termites from central Brazil (18°S) (around 1 m high) [ 36 ]. From our investigation, it was found that while wind does not influence the overall over-ground mound shape, it affects the mound internal structure and temperature. Increasing wind speeds led to wider lateral channels and narrower central chimney, which reduced the nest temperature significantly. In nature, mounds under unfavourable environmental conditions, such as strong winds, can be found with a nest temperature below a comfortable range [ 16 ]. To reverse this effect, termites reshape their mound to a more compact shape, with less internal channels, to increase the nest temperature [ 16 ]. However, this was not observed in this study since the reshaping process occurred due to a different objective function. The configuration that arose from the gas-exchange analysis was like those obtained from the thermal study, except in the spire alignment, which was found to be oriented vertically. The similarity between configurations shows that the scalar diffusion did not affect the mound configuration, while the contrast in spire behaviour provided substantial evidence of the mound inclination being a consequence of thermal processes within the mound, rather than gas exchange. The effect of the porosity of the mound body was also investigated. It was found that a porous body mound led to configurations analogous to those of a solid body mound. These similarities in the architectures show that the consideration of a solid body with air-filled internal channels is adequate for the analyses presented in this work. While optimal mound configurations did not exhibit a funnelled spire, further investigations showed that the funnelling in the spire did not have a significant effect on the heat exchange process of the mound. This observation suggests that the presence of this feature in mounds might not be due to thermal processes. Overall, the mound structure and its internal channels are arranged in such a way as to promote homeostasis of the mound atmosphere [ 5 ]. This is seen throughout the different analyses of this work, where each environmental factor displayed a specific effect on the mound structure. From a thermal point of view, the optimal configuration of the mound is the one that facilitates access to the currents that flow through the system, such as the metabolic and solar heats. While the current research has a focus on the shape and function of termite mounds under relevant thermal forces, it is important to note the importance of other factors, including erosion due to water dynamics and material stability. In addition, mounds play much more intricate roles in creating suitable microclimates for the termites, for example, through water retention [ 37 – 39 ]. The effects of these important factors are left for future studies. While the focus of the current research has been on one particular species, the methodology could be applied to predict the effect of environmental forces on any natural structure for which structural and environmental information can be obtained. As such, this practice provides a broader view of the factors that are effective in the form and function of a naturally made structure.",
"introduction": "1. Introduction In nature, some animals build structures to shelter themselves against environmental and physical threats. In some species, the function of these natural structures goes beyond a roof-over-head, as they are means for regulating temperature and moisture, and for gas exchange. The odd-shaped structures that mound-building termites create are famously known for providing both sheltering and environmental regulating functions [ 1 – 6 ]. Historically, the internal environment regulation and ventilation aspect of termite mounds have received more attention than their morphology and inspired most of the past studies in the field. These studies led to several theories relating the mound gas-exchange process to the metabolism-induced buoyant forces [ 7 – 9 ], forces due to the interactions between the mound superstructure and the atmospheric turbulent wind flow [ 5 ], or forces induced by the diurnally variable mound surface temperature [ 10 – 14 ]. It is noted that the forms and physical characteristics of the mounds, ranging from small domes to massive cone-, cathedral- and wedge-shaped structures, depend on the local climate and available materials. However, no definite correlation between termite species and mound shapes has been identified [ 15 ]. As an example, Korb & Linsenmair [ 16 ] studied two mounds of Macrotermes bellicosus termites with different structural characteristics in one geographical location. The one built in hot, open savannah was described as a more complex mound with thinner walls and a larger surface area compared to that in a cooler shaded area. Furthermore, Kooyman & Onck [ 17 ] reported that mounds of Pseudacanthotermes spiniger termites in a hot environment can be as twice as tall as those built in a nearby colder condition, while Turner [ 18 ] noted that mounds of Macrotermes michaelseni termites located under tree shadows tend to be more upright than those directly exposed to the Sun. The large diversities/similarities seen among termite mounds in one/different geographical locations and species have made a comparative study challenging. Recently, Claggett et al . [ 15 , 19 ] began an effort to use existing databases to extract the relationship between the environmental conditions and the structural form of termite mounds—a correlation that has, historically, been understudied [ 15 ]. Their study concluded that, in addition to the environmental conditions, the soil composition and property also play a significant part in the resulting mound shape. In this study, we use an engineering point of view to address the connection between the structural form of termite mounds and environmental forces. The question to be answered is ‘how does the mound's structure adapt to its local environment to provide a favourable living condition for its termites?’ The study implements a system configuration procedure, supported by the Constructal Law, to determine an optimized mound structure. The Constructal Law states that for a finite-size flow system to persist in time (to live), its configuration must evolve in a way that provides easier access to the currents that flow through it [ 20 , 21 ]. As long as a system has the freedom to alter its form, it will evolve its configuration in time to allow for better access to the fluxes that flow through it, while minimizing the holistic resistance and losses in the system. The study uses the structural characteristics of the mounds of M. michaelseni termites as the base since available information and literature on the physical features of these mounds are far richer compared to the other termite species. The aim is to investigate whether environmental (wind and solar irradiance) and metabolism-induced forces involved in the heat transfer process of termite mounds affect their structural configurations and if yes, in which direction. The methodology developed here could, in principle, be applied to any termite mound and the overall approach could be extended to other animal-built structures, where thermodynamics and heat transfer play important roles in their functions. The paper is structured as follows: Section 2 reviews the characteristics of the M. michaelseni termite mounds, focusing on the information used in the study to create the heat transfer model described in §3. Section 4 validates the model for different structural set-ups. Section 5 discusses the effect of different environmental conditions (including solar irradiance, solar zenith angle and wind), geometric features and porosity on the mound architecture. Finally, §6 elucidates the conclusions of the analysis based on the results of the current study and previous observations.",
"discussion": "5. Results and discussion The optimal configuration for the mound with no internal channels was found to show negligible responses to the environmental conditions. Therefore, only the results for the mound with internal channels are reported below. 5.1. Influence of environmental conditions on the mound architecture and thermal performance 5.1.1. Solar irradiance To study the effect of solar irradiance on the mound geometry and thermal performance, the irradiance ratio ( I ~ ) is changed between 0 and 150. This range accounts for cases with no or small ( I c = 0 − 175 W m −2 ), mild ( I c = 350 W m −2 ) and strong ( I c = 700 W m −2 ) solar irradiances typical at the region of northern Namibia [ 30 ] throughout the year, and for mature and small mounds that produce 50 and 4.7 W of metabolic heat [ 23 ]. Values of wind speed and zenith angle were fixed to the typical values of Outjo, Namibia ( u 0 = 1.38 (m s −1 ), χ = 19°). The results indicate that the spire inclination angle ( γ ) is strongly influenced by the amount of solar irradiance. Under the extreme solar irradiance ratio ( I ~ = 150 ) , the mound architecture tends to be taller and has a pronounced spire inclination, with a value equal to the imposed solar zenith angle, while mounds located in less exposed regions (experiencing weaker solar irradiance) are shorter and more vertically oriented. As shown in figure 5 , for I ~ = 0 − 150 , the spire inclination angle ranges between γ = 1 ∘ and γ = 19 ∘ . This finding is strongly in agreement with Turner's observation, where he reports that mounds located under tree shadows are more upright, while those exposed to the Sun exhibit a spire tilt of 19° (same angle as the local latitude) [ 18 , 34 ]. In open areas, mounds become more sensible to solar heat, especially over the base surface on the right (north) side (note that in the Southern Hemisphere, northward surfaces receive more direct irradiance). Hence, these mounds respond in a way to shade the north base surface, while also distancing the nest from the external heat transferred through the central chimney. This manifests through a more prominent spire inclination.\n Figure 5. Effect of the solar irradiance ratio ( I ~ ) on the spire inclination ( γ ) for a wind speed of u 0 = 1.38 m s −1 and zenith angle of χ = 19 ∘ . The red line and the error bars, respectively, represent the overall trend and the numerical uncertainty of the results within the resolution of the optimization method. In addition, the results indicate that being exposed to higher solar irradiances leads to the presence of wider lateral channels, representing a more complex internal structure and, in turn, a narrower central chimney ( figure 6 ), as they are related in the model owing to the area constraint. In these regions, wider lateral channels provide a much more effective way to disperse the heat from the mound surfaces. While the central chimney experiences a high-intensity solar irradiance coming from the top, narrowing the central channel creates more resistance for the heat, making a difficult path for it to reach the nest. By consequence, the heat is directed to the wider lateral channels to be dissipated through convective cooling.\n Figure 6. Effect of the solar irradiance ratio ( I ~ ) on the width of the central chimney ( D 0 / L ), for a wind speed of u 0 = 1.38 m s −1 and zenith angle of χ = 19 ∘ . The red line and the error bar, respectively, represent the overall trend and the numerical uncertainty of the results within the resolution of the optimization method. 5.1.2. Solar zenith angle (the effect of geographical location) The spires of M. michaelseni mounds located 22 km north of Outjo, in northern Namibia (16°4.50′ E, 19°59.05′ S) have a northward tilt angle of 19°, on average [ 18 ]. The similarity between the spire tilt angle and the geographical latitude led to the hypothesis that the Sun's position in the sky might be a determinant factor for the spire tilt [ 34 ]. This premise is also noted in the study of Darlington [ 4 ], who observed that the mounds in Kajiado, Kenya (36°48′ E, 1°50′ S) were only slightly tilted to the west (away from the prevailing wind), but not to the north. To investigate this effect, different zenith angles ( χ ) ranging between 0° and 25° are analysed with fixed wind speed ( u 0 = 1.38 m s −1 ) and irradiance ratio ( I ~ = 14 ) . The irradiance ratio of 14 represents a yearly average irradiance of 700 W m −2 for a mound of 50 W. Results indicate that the inclination of the spire decreases with decreasing zenith angle, showing that mounds closer to the equator ( χ = 0°) are almost upright ( γ = 2°), while for χ ≥ 15°, mounds tend to have a more prominent spire inclination ( γ = 12°). These results support the theory proposed by Turner [ 34 ]. To strengthen this claim, the tilt angle behaviour was investigated further by analysing different zenith angles for the extreme value of irradiance ( I ~ = 150 ) . Results showed that the tilt angle of the spire consistently follows the amount of zenith angle. Thermally, the inclination in the spire allows the structure to partially shade the northward face of the mound and block a large portion of the direct solar radiation over its surfaces. This allows the nest temperature to be regulated for different geographical locations, as the dimensionless temperature changed 7% within the zenith angle range analysed. The effect of solar zenith angle over the spire inclination is shown in figure 7 .\n Figure 7. Influence of solar zenith angle ( χ ) over the optimal inclination of the mound spire ( γ ) for irradiance ratio I ~ = 14 and wind speed u 0 = 1.38 (m s −1 ). The red line and the error bar, respectively, represent the overall trend and the numerical uncertainty of the results within the resolution of the optimization method. Mounds located closer to the equator line receive more prominent heat at their top surface and, therefore, they tend to be taller and, consequently, narrower in the spire ( H / L = 1.3 , L ′ / L = 0.35 for χ = 2 ) to create more resistance for the heat to reach the nest. On the other hand, mounds that experience solar heat more prominently on the lateral surfaces are prone to present a compact architecture with a wider spire ( H / L = 1 , L ′ / L = 0.45 for χ = 12 ). This arrangement distances the lateral surfaces from the nest and, given that these surfaces experience a higher solar irradiance, allows the nest temperature to be more effectively regulated. The effect of different solar zenith angles on the mound shape is shown in figure 8 .\n Figure 8. Effect of solar zenith angle on ( a ) the optimal mound height, represented by H / L , and ( b ) spire width, represented by L ′ / L . The red line and the error bar, respectively, represent the overall trend and the numerical uncertainty of the results within the resolution of the optimization method. 5.1.3. Wind Wind energy is known to be one of the driving forces for the nest ventilation (e.g. [ 5 ]). To understand the wind effect on the mound geometry, different wind speeds are considered, ranging from 0 to 5 m s −1 [ 5 ], while assuming fixed values for I ~ ( = 14 ) and χ ( = 19 ∘ ) . For these analyses, logarithmic and constant wind velocity profiles were investigated, with results showing no noteable difference between them (not shown). Therefore, the results are presented for the constant wind profile. For the range of values investigated, the results indicated that the optimal mound structure remained the same for different wind speeds, indicating that the mound is capable of convectively diffusing the heat from the mound surfaces, without a need for altering the overall mound architecture. Although wind does not modify the external structure of the mound, it has a significant influence on the size of the internal channels. Mounds that experience stronger winds tend to have more capacious lateral channels, while also presenting a reduced central chimney. This increase in the lateral channels allows them to better harness the wind energy. At the same time, mounds located in regions with no or weak wind speeds display a larger central chimney to allow the metabolic heat to be diffused through a larger channel area ( figure 9 ).\n Figure 9. Effect of wind speed on the width of the mound central chimney, represented by D 0 / L . The red line and the error bar, respectively, represent the overall trend and the numerical uncertainty of the results within the resolution of the optimization method. Besides its effect on the internal channels, wind can play a significant role in the nest temperature. Strong winds around the mound can cool down the mound body and, by consequence, the nest. The nest temperature of a mound that experiences a wind of u 0 = 5 (m s −1 ) is around 14% lower than that of a mound exposed to no wind. This reduction in temperature might be an adverse effect for the termites, as they reverse this temperature reduction by building their mound in a more compact, dome-like shape [ 16 ]. Such a compact configuration capitalizes on the metabolic heat from the nest to increase its temperature and, therefore, it has a different objective function that is not considered in this work. Figure 10 summarizes the optimal mound configurations for different environmental conditions explained above.\n Figure 10. Dimensionless temperature distribution of the optimal termite mound configurations for different solar irradiances ( a ), solar zenith angles ( b ) and wind speeds ( c ) (the same colour bar was used for all configurations). It is noted that in all cases analysed above, the funnelling angle of the optimized configurations remained unaffected. Additional simulations were set up to investigate whether the funnelled spire of termite mounds have any contribution to the thermal processes within the mound, and whether the lateral channels that contribute to heat dissipation midway between the nest and the top surface affect this result. 5.1.4. Effect of the lateral channels A new set of simulations were performed for a mound with only a central chimney to investigate the effect of the removal of the lateral channels on the mound geometry. It is found that the mound optimal configuration remained unaffected to the wind and zenith angle parameters, while significantly influenced by the amount of the irradiance ratio. In contrast with the mound with chimney and lateral channels, the mound without lateral channels is much more sensitive to the intensity of the external heat. That is because, in this case, the chimney is the main passageway to balance the effect of the two heat sources (solar and metabolic), and removal of the lateral channels removes additional effective paths for these heats. As a result, unlike the case with lateral channels, the mound geometry revealed a significant sensitivity to the amount of the irradiance, reflected in the quantities of mound aspect ratio, spire tilt and funnelling angles. The mound aspect ratio varied between 0.6 (for I ~ = 1 ) to 1.1 (for I ~ > 50 ), and the spire tilt angle changed from γ = 1 ∘ \n ( I ~ = 1 ) to γ = 31 ∘ \n ( I ~ > 75 ) . These large variations appear because, under stronger solar irradiances, taller mounds create greater distances, thus larger resistances, between the nest and the top surface. For higher irradiance ratios, the external heat overwhelms the metabolic heat and becomes more important in increasing the mound temperature. Thus, to lessen the effects of this heat, the mound distances the top surface by becoming taller and more tilted, as there are no lateral channels to effectively diffuse the external heat. In contrast with the previous configurations, the mound without lateral channels exhibited funnelling in the spire for I ~ > 50 , increasing until it stabilized at α = 12 ∘ for I ~ > 100 . The reason behind the appearance of funnelling angle, in this case, is the mound's need for becoming taller under strong solar heats. By increasing the funnelling angle, the mound uses the last resource of its constrained body to become taller and separate the top surface further away from the nest. In contrast with the tilting angle, this effect manifests only at high irradiance ratios, indicating that it is more thermally efficient for the mound to tilt its spire than funnel it. In addition to studying the effects of the removal of the lateral channels, the influence of having several lateral channels with independent heights was also investigated. The inclusion of several lateral channels did not affect the mound architecture and altered the nest temperature compared to those with one set of channels by only Δ θ = 0.01. Considering the small sensitivity of the model to this additional complexity and the amount of additional computational efforts, mounds with more than one set of lateral channels were not analysed further. 5.1.5. Effect of forcing funnelled spire on the thermal process of the mound To further investigate the absence of spire funnelling in the mound optimum configurations, the area constraint was relaxed for the mound with lateral channels and all the aspects of the optimal configuration (§4.2), except the funnelling angle, were kept unchanged. The results of these investigations indicated that the nest temperature did not alter (Δ θ = 10 −4 ) in response to the forced funnelling. This shows that the configurations previously obtained are not just thermally optimized, but they also have a certain resilience in maintaining their thermal performance. It could be concluded that the spire funnelling may not arise due to thermal processes, but rather due to different mechanisms, such as mechanical stability and erosion (not considered in this work). 5.2. Influence of porosity on the mound architecture and thermal performance The weak convective process in the mound conduits is driven by an airflow of 0.03–0.06 m s −1 [ 14 ], while the rest of the mound body is impervious to bulk flow. The mound body is porous on a microscopic scale that only allows for diffusive exchanges (personal communication with Dr Hunter King, 2019). This evidence led the study to consider the mound model to be composed of a solid body with a conductivity of clay and internal channels with higher conductivity to represent the higher capacity of the conduits to transfer heat and mass (see §3). However, to investigate the effect of the porosity of the mound soil-based body on its thermal performance and architecture, the study was extended as described below. The effect of porosity is considered by using porosity-based effective thermal conductivity ( k ~ eff ( ρ ) ) . The effective thermal conductivity is calculated based on a linear interpolation between the thermal conductivity of the soil and the air, and in its dimensionless form, it is defined as k ~ eff ( ρ ) = 1 + ( k ~ − 1 ) ρ , where ρ is the local porosity and k ~ is the normalized thermal conductivity of the soil. The performance of the model was, initially, tested for the typical environmental conditions in Outjo, Namibia, for a mound with half of the total porosity (e.g. 18%) distributed in the body and the other half concentrated in the internal channels. The optimum configuration obtained in this case (i.e. H / L = 1 , L ′ / L = 0.45 , D 0 / L = 0.04 , γ = 12 ∘ , α = 0 ∘ , with a minimized dimensionless nest temperature ( θ nest ) of 0.21) was similar to that found for the mound with a solid body (§4.2). The investigation was extended for different values of environmental forces. It was found that the optimum mound structure remained insensitive to different values of wind speed and zenith angle. However, the spire tilt increased significantly ( γ = 29 ∘ ) for higher values of irradiance ratios ( I ~ ≥ 50 ) , deviating from those of the solid mound ( 12 ∘ < γ < 19 ∘ for I ~ ≥ 100 ). This change in the behaviour of spire tilt is due to the lower effective conductivity of the mound body in the porous case. The lower conductivity results in a concentration of heat around the nest area, especially on the north side, where the Sun heats the base. Therefore, the mound becomes more inclined to better shade the north face of the base."
} | 7,109 |
22629571 | null | s2 | 2,800 | {
"abstract": "Industrial biotechnology promises to revolutionize conventional chemical manufacturing in the years ahead, largely owing to the excellent progress in our ability to re-engineer cellular metabolism. However, most successes of metabolic engineering have been confined to over-producing natively synthesized metabolites in E. coli and S. cerevisiae. A major reason for this development has been the descent of metabolic engineering, particularly secondary metabolic engineering, to a collection of demonstrations rather than a systematic practice with generalizable tools. Synthetic biology, a more recent development, faces similar criticisms. Herein, we attempt to lay down a framework around which bioreaction engineering can systematize itself just like chemical reaction engineering. Central to this undertaking is a new approach to engineering secondary metabolism known as 'multivariate modular metabolic engineering' (MMME), whose novelty lies in its assessment and elimination of regulatory and pathway bottlenecks by re-defining the metabolic network as a collection of distinct modules. After introducing the core principles of MMME, we shall then present a number of recent developments in secondary metabolic engineering that could potentially serve as its facilitators. It is hoped that the ever-declining costs of de novo gene synthesis; the improved use of bioinformatic tools to mine, sort and analyze biological data; and the increasing sensitivity and sophistication of investigational tools will make the maturation of microbial metabolic engineering an autocatalytic process. Encouraged by these advances, research groups across the world would take up the challenge of secondary metabolite production in simple hosts with renewed vigor, thereby adding to the range of products synthesized using metabolic engineering."
} | 459 |
30675014 | PMC6344598 | pmc | 2,802 | {
"abstract": "We demonstrate for the first time the direct stereolithographic 3D printing of an extrinsically self-healing composite, comprised of commercial photocurable resin modified with anisole and PMMA-filled microcapsules. The composites demonstrate solvent-welding based autonomous self-healing to afford 87% recovery of the initial critical toughness. This work illustrates the potential of stereolithographic printing to fabricate self-healing composites with user-defined structures, avoiding the need for extensive rheological optimization of printing inks, like in direct-write 3D printing. Importantly, this work also demonstrates the inclusion of microcapsules into 3D printing resins to incorporate additional functionality into printed composites, which could be adapted for applications beyond self-healing materials.",
"introduction": "Introduction The lifetimes of composite materials are typically limited by fatigue or other material failure mechanisms due to damage encountered during service. However, in nature, plants and animals overcome this limitation by utilizing self-healing as a crucial survival strategy to repair damage to their tissues. Taking lessons from nature, scientists have extensively researched the incorporation of self-healing capabilities into synthetic polymeric materials to prolong their operational lifetimes 1 – 5 . Self-healing materials are classified into two categories – intrinsic and extrinsic. Intrinsic self-healing materials rely on reversible bonds such as metal-ligand bonds 6 or hydrogen bonds 7 to facilitate healing and are therefore typically limited to gels or elastomeric materials which allow for molecular diffusion 1 . Extrinsic self-healing, however, utilizes healing components sequestered from the main matrix within microcapsules or vascular networks 2 . During composite fracture, the capsules or network are broken, releasing the healing agents which react with each other or interact with the matrix to seal the fracture. In this case, stiff polymer matrices can be used, as molecular diffusion of the matrix is not a requirement for healing. Therefore, extrinsic self-healing is desirable for many practical applications requiring hard polymeric structures. In parallel, the use of 3D printing (3DP) has become increasingly ubiquitous in different fields due to the ability to generate user-defined 3D objects with a variety of materials 8 . 3DP encompasses a family of additive manufacturing techniques that allow rapid yet flexible fabrication of complex 3D structures with features from the sub-micron to the multi-meter scale. Materials such as ceramics 9 , resins 10 and even novel nanocomposites 11 can be precisely structured through 3DP. Direct 3DP of elastomers and hydrogels exhibiting intrinsically self-healing properties has also been demonstrated 12 – 14 . Researchers have also 3D printed sacrificial scaffolds, which are then utilized for templating vascular self-healing systems. The printed scaffolds are embedded into a polymeric matrix which is cured, then the scaffold is removed via washing or heating under vacuum, and replaced with healing agents. The vascular ends are then sealed to afford the resulting self-healing composite 15 , 16 . However, to our knowledge, the direct 3D printing of an extrinsically self-healing system has not been reported.",
"discussion": "Results and Discussion Here, we demonstrate a technique of combining UV-curable resin embedded with solvent-containing microcapsules in conjunction with stereolithographic (SL) 3DP to construct user-defined 3D structures, whereby a laser (405 nm) spatioselectively polymerizes/crosslinks the resins according to a computer aided design. The self-healing employed in this work follows a solvent welding mechanism, as illustrated in Fig. 1 . When a crack occurs and ruptures a capsule along the propagation pathway, the solvent within the capsule is released into the matrix. Solvent release promotes polymer diffusion and entanglement across cracks formed in the matrix, leading to crack healing 17 , 18 . Such a method is advantageous in its simplicity and cost-effectiveness, with no need for expensive metal catalysts 19 or the preparation of multiple types of microcapsules containing different healing reagents 20 . Figure 1 Schematic illustration of the solvent welding based self-healing mechanism: ( a ) The virgin material with intact microcapsules embedded within the polymer matrix; ( b ) crack propagation and rupture of the microcapsule shell. The encapsulated solvent anisole diffuses into the surrounding polymer matrix. This enhances polymer diffusion across the crack and polymer chain entanglement; ( c ) polymer chain entanglement heals the crack. Anisole, which is widely used in both the fragrance industry and as a food additive, was selected as the solvent for encapsulation due to its low toxicity 17 . The high boiling point and immiscibility of anisole with water also allows it to be easily encapsulated using in situ polymerization of urea-formaldehyde in an oil-in-water emulsion 21 . Importantly, anisole has been shown to be a suitable solvent for solvent welding based self-healing in PMMA 22 , 23 which contains methacrylate functionalities commonly found in most commercially available SL 3DP resins. Anisole was therefore expected to also be a good candidate for healing photocured SL 3DP resins and preliminary testing showed that anisole could soften and increase tackiness of surfaces in such photocured samples. Anisole-containing capsules were prepared using a technique modified from that described by Brown 21 . This afforded urea-formaldehyde microcapsules with an anisole and 5 wt% PMMA core, which had an average diameter of 130 ± 15 µm (Figs S1 and S2 ). PMMA was incorporated into the microcapsules together with anisole, as the work by Gladman et al . showed that the inclusion of PMMA improved healing efficiencies 23 . Scanning electron microscopy (SEM) (Fig. 2 ) showed that the microcapsule walls possessed a rough surface, similar to observations reported by others 21 , 24 . The capsule roughness has been attributed to the precipitation of polymerized urea-formaldehyde from the water phase and its deposition onto the capsule wall at the oil-water interface during in situ polymerization. This shell roughness is desirable as it promotes capsule adhesion to the polymer matrix and provides a greater possibility for microcapsule rupture in the event of crack propagation 21 . Figure 2 Representative SEM images showing ( a ) the rough surface of a urea-formaldehyde microcapsule shell; ( b ) a ruptured capsule showing the cross-section of its shell wall. The microcapsules were mixed into the UV-curable resin to achieve 2.5, 5 and 10 wt% capsule concentrations. SEM showed that the capsules were successfully embedded into the cured polymer matrix (Fig. S3 ). FTIR spectroscopy was also performed on the cured composites (Figs S4 , S5 and S6 ), whereby only the composites with capsules showed the presence of the urea stretching mode at 3410 cm −1 . Thermogravimetric analysis (TGA) of anisole/PMMA-filled urea-formaldehyde microcapsules showed a sudden rupture of the capsules at approximately 260 °C. Repeating this experiment twice more, we again observed this phenomenon each time. This was attributed to the explosion of the capsules upon sufficient thermal degradation of the urea-formaldehyde shell, which occurs between 220–300 °C 25 , 26 and internal capsule pressure arising from vaporization of anisole and PMMA degradation products. However, TGA of microcapsules in cured resin mixtures did not show a similar rupture event, presumably as the presence of the resin matrix prevented explosion of microcapsules (Fig. S7 ). To investigate their self-healing properties, mode 1 fracture testing was performed on tapered double cantilever beam (TDCB) test samples comprising of these mixtures (Fig. S8 ), which were generated using a molding technique. Degassed microcapsule-resin mixtures were poured into silicone molds and UV cured. The self-healing efficiency of anisole/PMMA microcapsule composites was quantified using a protocol for extrinsically self-healing materials first utilized by White et al 27 . This approach defines the self-healing efficiency of the material as a ratio of the fracture toughness, K c , of the virgin material versus that of the healed material. K c is linked to the critical load, P c (the load at which crack propagation occurs) as shown in equation ( 1 ). 1 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${K}_{c}=\\alpha {P}_{c}$$\\end{document} K c = α P c Here, α is a geometric constant specific to the host matrix. However, this is complicated by the heavy reliance of P c on initial crack length, as the initial crack lengths of the virgin samples may differ from that of the healed samples. The use of samples with TDCB geometry allows us to overcome this complication. In samples with this geometry, the crack length of the healed and virgin samples can be ignored as P c remains constant along the length of the sample 28 . Mechanical testing of the virgin and healed materials was carried out to obtain values for P c and the values inserted into the following equation: 2 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$Self-healing\\,efficiency=(\\frac{\\alpha {P}_{{c}_{healed}}}{\\alpha {P}_{{c}_{virgin}}})\\times 100 \\% $$\\end{document} S e l f − h e a l i n g e f f i c i e n c y = ( α P c h e a l e d α P c v i r g i n ) × 100 % Assuming that the geometric constant α remains the same for the virgin and healed material, it can be cancelled from equation ( 2 ) to give: 3 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$Self-healing\\,efficiency=(\\frac{{P}_{{c}_{healed}}}{{P}_{{c}_{virgin}}})\\times 100 \\% $$\\end{document} S e l f − h e a l i n g e f f i c i e n c y = ( P c h e a l e d P c v i r g i n ) × 100 % The molded TDCB samples were pre-cracked and loaded into a universal testing instrument to perform mode I tensile fracture testing (Figs S9 ). Although the exact value of α is unknown for this material, we can study the relative effect of capsule loading on composite K c at different microcapsule loadings by plotting the experimentally obtained values of P c , which is equivalent to \\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}$$\\frac{{K}_{c}}{\\alpha }$$\\end{document} K c α , against capsule loading. We observe that the presence of capsules within the matrix increases the composite fracture toughness, although the effect plateaus and no increase in fracture toughness was observed by increasing the capsule loadings beyond 2.5 wt% (Fig. 3a ). The presence of tail-like structures in the wake of the microcapsules in the fracture plane (Fig. S3 ) suggests that crack pinning contributes to the fracture toughening 29 . Hackle markings, which tend to form during violent fracture when both plastic deformation and branching of the crack front occur 30 , can also be observed. Both the tail and hackle markings increase the surface area of the crack plane, and thus the energy absorbed by the composite during crack growth, thereby increasing the resulting fracture toughness. This fracture toughening mechanism as a result of incorporation of urea-formaldehyde capsules is supported in the literature for a number of materials including epoxy resins 29 and thermoplastics such as PMMA 22 . Figure 3 Bar charts showing ( a ) the effect of capsule loading on the critical loadings, \\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}$$\\frac{{K}_{c}}{\\alpha }$$\\end{document} K c α . Tests were performed in triplicate – error bars denote the standard deviation; ( b ) healing efficiencies of 5 wt% capsule loaded samples with different healing times. Samples were healed at 25 °C for 24, 72 and 120 hours; ( c ) healing efficiencies of samples with different capsule concentrations. Samples were healed at 25 °C for 72 hours. The tests all were performed in triplicate – error bars denote the standard deviation. As outlined in Equation 2 , the healing efficiency could be calculated through mechanical testing of the TDCB specimens. After the initial fracture event, samples were allowed to heal in a temperature regulated environment at 25 °C for 24, 72 and 120 hours to determine the optimum healing time. The healing efficiencies of samples with 5 wt% capsule loading were investigated (Fig. 3b ) and on average, the healing efficiency appears to reach a maximum after 72 hours.Keeping the healing time at 72 hours, the capsule loading was then increased in order to ascertain whether increasing the capsule concentration could achieve higher healing efficiencies. Capsule loadings of 2.5, 5 and 10 wt% were investigated; all samples were healed for 72 hours at 25 °C (Fig. 3c ). In all samples some healing was observed, with higher healing efficiencies achieved at higher loading rates. In this work, a maximum healing efficiency of 87% was achieved at a capsule concentration of 10 wt%. However, at 10 wt% capsule loadings, we observe a drop in the fracture toughness of the composite (Fig. 3a ) from that at 5 wt% capsule loading. Therefore, the healing efficiencies for composites with higher loadings of microcapsules were not investigated. To demonstrate 3DP of extrinsically self-healing composites, resins with 0, 5 and 10 wt% loadings of microcapsules were poured into printer resin trays and printed by an SL 3D printer (Form 1+ , Formlabs, USA). Addition of the microcapsules to the resins caused the originally clear resin to appear cloudy. However, we observed no noticeable effect on print quality due to light scattering by the capsules (Fig. 4 ). This observation is supported by the fact that many commercial SL 3DP resins contain light-scattering particulates and pigments, giving them high turbidity, whilst still affording high quality prints. To test the healing of 3D printed samples, damage was inflicted onto the samples and the damage monitored. We found that the release of microcapsule contents onto the fracture planes allowed the healing of cracks and breakages (Fig. 5 ). Therefore, by incorporating anisole/PMMA containing microcapsules into the resins, objects with self-healing ability can be 3D printed with these resins. Figure 4 Photograph of 3D printed objects with various capsule loading values. From left to right the microcapsule loading increases from 0 to 5 and then 10 wt%. Figure 5 Photograph of ( a ) A 3D printed sample which contains 5 wt% anisole with PMMA capsules. The cut is highlighted in the red circle; ( b ) the 3D printed sample after the two fracture planes were pushed back together and allowed to heal for 3 days at 25 °C. The healed section is highlighted by the red circle. The ability to directly 3D print structures with extrinsic self-healing characteristics has, to our knowledge, not been reported prior to this work. Previous reports on 3D printed self-healing materials are based on intrinsically self-healing materials and their direct-ink writing. However, direct-ink writing based 3DP requires significant optimization of ink rheology, which will differ for each formulation, thereby presenting significant complications 13 , 14 . In summary, we have demonstrated SL 3DP of a solvent-welding based self-healing material through addition of self-healing capsules to commercially available resins. The healing efficiency of the work showed a maximum recovery of 87% with respect to the critical load, tested by mode 1 fracture toughness. Further investigation could improve the self-healing efficiency of this material through exploration of different solvents and encapsulated polymers to enhance the solvent-welding mechanism. These results are promising for applications requiring materials with bespoke structures as well as extended structural integrity, such as in personalized medicine. For example, researchers have started to explore the use of solvent and PMMA containing urea-formaldehyde microcapsules to improve the lifetime expectancy of bone cement 23 . The ability to combine this with 3DP would further improve the prospect of such materials being utilized within this field, particularly with the rapid development of commercial biocompatible resins. Further, our approach of adding microcapsules to rapidly incorporate functionality to readily available commercial inks is attractive due to its ease of adoption and flexibility. This promising approach has widespread applications that can be easily modified to incorporate alternative functionalities to 3D printed materials, such as for hollow glass sphere containing light-weight composites 31 , or for flame retardant composite materials; we will therefore investigate such alternative applications in future work."
} | 4,467 |
25753826 | null | s2 | 2,804 | {
"abstract": "Metabolically engineered strains of the hyperthermophile Pyrococcus furiosus (T(opt) 95-100°C), designed to produce 3-hydroxypropionate (3HP) from maltose and CO2 using enzymes from the Metallosphaera sedula (T(opt) 73°C) carbon fixation cycle, were examined with respect to the impact of heterologous gene expression on metabolic activity, fitness at optimal and sub-optimal temperatures, gas-liquid mass transfer in gas-intensive bioreactors, and potential bottlenecks arising from product formation. Transcriptomic comparisons of wild-type P. furiosus, a genetically-tractable, naturally-competent mutant (COM1), and COM1-based strains engineered for 3HP production revealed numerous differences after being shifted from 95°C to 72°C, where product formation catalyzed by the heterologously-produced M. sedula enzymes occurred. At 72°C, significantly higher levels of metabolic activity and a stress response were evident in 3HP-forming strains compared to the non-producing parent strain (COM1). Gas-liquid mass transfer limitations were apparent, given that 3HP titers and volumetric productivity in stirred bioreactors could be increased over 10-fold by increased agitation and higher CO2 sparging rates, from 18 mg/L to 276 mg/L and from 0.7 mg/L/h to 11 mg/L/h, respectively. 3HP formation triggered transcription of genes for protein stabilization and turnover, RNA degradation, and reactive oxygen species detoxification. The results here support the prospects of using thermally diverse sources of pathways and enzymes in metabolically engineered strains designed for product formation at sub-optimal growth temperatures."
} | 408 |
35848307 | PMC9397406 | pmc | 2,805 | {
"abstract": "The global expansion of biomanufacturing is currently\nlimited by\nthe availability of sugar-based microbial feedstocks, which require\nfarmland for cultivation and therefore cannot support large increases\nin production without impacting the human food supply. One-carbon\nfeedstocks, such as methanol, present an enticing alternative to sugar\nbecause they can be produced independently of arable farmland from\norganic waste, atmospheric carbon dioxide, and hydrocarbons such as\nbiomethane, natural gas, and coal. The development of efficient industrial\nmicroorganisms that can convert one-carbon feedstocks into valuable\nproducts is an ongoing challenge. This review discusses progress in\nthe field of synthetic methylotrophy with a focus on how it pertains\nto the important industrial yeast, Saccharomyces cerevisiae . Recent insights generated from engineering synthetic methylotrophic\nxylulose- and ribulose-monophosphate cycles, reductive glycine pathways,\nand adaptive laboratory evolution studies are critically assessed\nto generate novel strategies for the future engineering of methylotrophy\nin S. cerevisiae .",
"conclusion": "5 Concluding Remarks The field of synthetic\nmethylotrophy in S. cerevisiae is still in its\ninfancy. To date, there have been four research\ngroups that have reported work on synthetic methylotrophy in S. cerevisiae , with only a handful of publication and\npreprints available ( Table 1 ). These efforts have provided a valuable foundation of knowledge.\nIt has been shown that synthetic methanol assimilation in S. cerevisiae can be achieved through a synthetic XuMP\ncycle, 39 a synthetic RuMP cycle, and a\nhybrid RuMP/XuMP cycle. 40 , 42 A chimeric RuMP/XuMP\ncycle for dual formaldehyde assimilation has also been demonstrated\nin Y. lipolytica , which is relevant to S. cerevisiae . 44 Additionally,\nthe core module of the rGly pathway has been engineered in S. cerevisiae to convert formate to glycine. 43 From these works, it has been established that\na synthetic methanol assimilation cycle on its own is not sufficient\nto enable full synthetic methylotrophy with growth that is industrially\nrelevant in S. cerevisiae . 40 Therefore, any engineered strain will require further modification\nand/or ALE to achieve this. This review has assessed several\nstrategies to enhance synthetic\nmethanol assimilation that are worth considering for the future engineering\nof industrially relevant methylotrophy in S. cerevisiae . These include the following: careful screening and selection of\nappropriate methylotrophy genes; PEX gene modifications\nand peroxisomal compartmentalization to enhance XuMP cycle methylotrophy;\nhybrid methanol oxidation through mixed expression of an Aox and/or\nNAD-Mdh; dual formaldehyde assimilation through combined expression\nof XuMP and RuMP cycles; increasing available formaldehyde through\ntargeted knockout of dissimilatory enzymes; speeding up regeneration\nof the formaldehyde acceptor through engineering of the PPP; ALE;\nand engineering the rGly pathway. Despite the progress made so far, 39 , 40 , 42 , 44 , 43 only limited growth of S. cerevisiae on methanol as the sole carbon source has been achieved. This demonstrates\njust how difficult it is to engineer synthetic methylotrophy in S. cerevisiae , even after exhaustive modification and\nALE. To overcome this challenge, it may be beneficial to look\ntoward\nthe recent success in enabling effective synthetic methylotrophy in E. coli . 25 This was\nachieved by first engineering synthetic methanol auxotrophy, making\ngrowth on xylose dependent on a functioning RuMP cycle that requires\nmethanol, and increasing the selective pressure on cells to assimilate\nmethanol during co-carbon substrate ALE. 91 By imposing selection pressure for methanol assimilation, the authors\nwere able to set the appropriate conditions for the methylotrophic\nphenotype to arise after being slowly weaned off xylose. 25 Importantly, this breakthrough in E. coli demonstrates that rational engineering\ncan be effectively employed as a tool to enhance the power of ALE\nfrom the start of a project. It is important to note that previous\nexperiments in S. cerevisiae that have employed\nstandard modes of ALE have achieved limited success. 41 , 42 Therefore, the next logical step is to implement synthetic methanol\nauxotrophy, which will then be followed by ALE on xylose and methanol.\nAs well as using a RuMP cycle, it may also be possible to use a XuMP\ncycle to engineer synthetic methanol auxotrophy in S. cerevisiae , which would provide additional engineering options for future experiments.\nThe successful engineering of the core module of the rGly pathway\nin S. cerevisiae also presents a possible route\ntoward synthetic methylotrophy. 43 The next\nsteps in this strategy would require additional modifications to allow\nfor the efficient oxidation of methanol to formate to feed the pathway,\nand the efficient conversion of glycine into pyruvate for biomass\ngeneration. Engineering synthetic methylotrophy in S. cerevisiae remains an ongoing challenge, yet it is one worthy of accepting.\nUtilizing methanol as a fermentation feedstock has the potential to\nunlock waste carbon utilization on a grand scale, with no organism\nbetter suited than S. cerevisiae to maximize\nits conversion into useful products.",
"introduction": "1 Introduction The global expansion of\nbiomanufacturing is currently limited by\nthe availability of hexose sugar, which is the primary microbial feedstock.\nThis limitation arises because sugar requires arable farmland for\ncultivation of sugar cane or corn, and therefore cannot support broad\nscale biomanufacturing without impacting the human food supply. 1 One-carbon (C1) feedstocks such as methanol have\nattracted intense interest as an alternative to sugar because they\ncan be produced as byproducts from other activities, and their availability\nis not limited by the scarcity of arable farmland. 2 For example, methanol can be derived from methane, which\nis generated from natural gas deposits or biogas from municipal waste.\nMethanol can also be derived from synthesis gas, which can be obtained\nthrough gasification of coal or waste organic material ( Figure 1 ). 1 Interestingly, methanol can also be produced from atmospheric CO 2 , which is reduced to methanol using hydrogen generated from\nelectrolysis. 3 Work also continues on the\ndevelopment of electrochemical reduction technologies for the direct\nconversion of atmospheric CO 2 into other C1 compounds. 4 , 5 As these technologies develop further, it is likely that atmospheric\nCO 2 will become an abundant source of methanol in the future.\nFurther, compared to other gaseous C1 compounds (carbon monoxide,\nCO 2 , methane, etc.), methanol provides a convenient liquid\nfeedstock for large-scale transport and industrial fermentation. 6 Therefore, the development of industrial production\nhosts that can utilize methanol as their sole carbon source will enable\nbioproduction from hydrocarbons, greenhouse gases, or biomass, and\nreduce our dependence on arable farmland for the production of microbial\nfeedstocks. 7 However, this remains a difficult\ntask because methanol is a toxic substrate that cannot be utilized\nby most common industrial microorganisms. Figure 1 The future of methanol-based\nbiomanufacturing. Methanol presents\nan enticing alternative to sugar-based microbial feedstocks for biomanufacturing.\nUnlike sugar, methanol does not require agricultural land for cultivation\nof sugar cane or corn and can be produced in large quantities as a\nbyproduct from other processes. Yeast presents several advantages\nfor use in biomanufacturing, and there is a pressing need to develop\nplatform industrial yeast species for methanol-based biomanufacturing. S. cerevisiae is the most widely used, well developed,\nand versatile industrial yeast species for biomanufacturing. Therefore,\nit would be most convenient for industry if robust growth on methanol\nas the sole carbon source could be achieved in S. cerevisiae . However, because methanol is a toxic metabolite for most industrially\nrelevant microbes, significant challenges remain. This image was created\nusing Biorender.com . Native methylotrophs with the capacity to utilize\nC1 compounds\nas their sole carbon source are common in nature, where they subsist\non a range of different substrates such as methane, formate, methyl\namine, and methanol. 8 , 9 However, native methylotrophs\nare not currently suitable for broad use in biomanufacturing due to\ntheir poor genetic characterization and limited genetic tractability. 10 This results in limited product spectra, low\nyields, and slow and expensive strain engineering. Comparatively,\nnon-methylotrophic model industrial species have benefited from decades\nof research and development and have extremely well characterized\ngenetics, established synthetic biology tools, and high genetic tractability. 11 This has resulted in many established high-yield\nbioproduction pathways and flexibility for the development of new\nproducts, as demonstrated in the two most utilized model species, Saccharomyces cerevisiae and Escherichia\ncoli . 7 S. cerevisiae is now used widely to produce a broad spectrum of value-added products 12 ( Figure 1 ). Some of these include the following: organic acids, 13 fatty acids, 14 advanced\nbiofuels, 15 pharmaceuticals, 16 , 17 biopharmaceuticals, 18 alkaloids, 19 carotenoids, 20 and\nflavors. 21 Two tracks have emerged\nfor developing methanol-based biomanufacturing.\nOne track involves the engineering of native methylotrophs to improve\ntheir capacity for bioproduction to match model species. The other\ntrack involves engineering synthetic methylotrophy in established\nmodel species. Both tracks have their own merits and challenges, and\nit remains to be seen which will be most successful. 22 , 23 In recent years, considerable progress in engineering synthetic\nmethylotrophy in model bacteria has been made, and an excellent summary\nof the work thus far in bacteria can be found in a review from Gregory\net al. 24 This has culminated in the development\nof a strain of E. coli that is\ncapable of robust growth on methanol as a sole source of carbon. 25 However, there remains a significant need to\ndevelop methanol-based biomanufacturing in yeast, as yeast provides\na number of advantages over bacteria for use in industry ( Figure 1 ). Yeast has a greater\ntolerance for high-speed centrifugation and mixing, solvent exposure,\nand variations in temperature and pH, making it a more robust production\nhost that is suitable to a broader range of industrial conditions. 26 Industrial yeast fermentation is not susceptible\nto bacteriophage contamination, which can significantly reduce the\nyield of industrial bacterial fermentations. 27 Eukaryotic protein folding mechanisms and post-translational modifications\nalso mean that yeast is able to produce a broader range of complex\nrecombinant proteins. 28 Further, yeast’s\neukaryotic cell structure provides additional options for metabolic\nengineering through organelle targeted expression of biosynthetic\npathways that can benefit from organelle specific metabolic processes, 29 although it is possible to localize metabolism\nin bacteria using bacterial microcompartments. 30 Some progress in harnessing yeast for methanol-based\nbioprocessing\nhas been made in the facultative methylotroph, Pichia pastoris (reclassified as Komagataella phaffii ), which has\nalready proven to be an efficient industrial species for recombinant\nprotein production. 31 In addition to these\nproteins, a handful of valuable metabolites have now been synthesized\nin P. pastoris using methanol as the sole carbon\nsource. 32 These include the following:\norganic acids, d -lactic acid, 33 malic acid, 34 and 6-methylsalicylic acid; 35 the polysaccharide hyaluronic acid; 36 and the important antihypertensive compound\nlovastatin and its precursor monacolin. 37 Despite this progress, P. pastoris and other\nnative methylotrophic yeasts remain less amenable than S. cerevisiae to cutting-edge synthetic biology, metabolic engineering, and systems\nbiology approaches, and therefore have potentially narrower bioproduction\ncapabilities. 38 Accordingly, it would be\nmost convenient if efficient synthetic methylotrophy could be enabled\nin S. cerevisiae . The field of engineering\nsynthetic methylotrophy in S. cerevisiae is\nstill in its infancy, and only a handful of primary publications\nand preprints are available on the topic. 39 − 43 Some progress toward synthetic methylotrophy in S. cerevisiae has been made in recent years using synthetic\nversions of the xylulose monophosphate (XuMP) cycle from methylotrophic\nyeast, 39 and the ribulose monophosphate\n(RuMP) cycle from methylotrophic bacteria. 40 Synthetic methanol assimilation has also recently been demonstrated\nin the oleaginous industrial yeast, Yarrowia lipolytica , using a combined RuMP/XuMP synthetic methanol assimilation pathway. 44 Additionally, the core module of a synthetic\nreductive glycine (rGly) pathway has been engineered in S. cerevisiae to convert C1 formate into glycine. This review summarizes the molecular\nmechanisms behind methylotrophy with a focus on the RuMP and XuMP\ncycles, and the reductive glycine pathway, and how they may be applied\nin S. cerevisiae . Despite the recent progress,\nrobust growth of S. cerevisiae using methanol\nas the sole carbon source is yet to be achieved. We therefore discuss\ndifferent strategies to enhance methanol assimilation and metabolic\ncycling through the RuMP/XuMP cycles and the rGly pathway. These include\nthe following: screening candidate methylotrophy genes; hybrid methanol\noxidation; dual formaldehyde assimilation; peroxisome proliferation\nand compartmentalization; and adaptive laboratory evolution (ALE).\nIt is our intent that the strategies discussed in this review will\naid in the design of future synthetic methylotrophy projects in S. cerevisiae and facilitate further progress in the\nfield. Table 1 Summary of the Work so Far in Engineering\nSynthetic Methylotrophy in Yeast a a Including studies in S. cerevisiae by four separate research groups and\none group in Y. lipolytica ."
} | 3,559 |
21559391 | PMC3084812 | pmc | 2,806 | {
"abstract": "Lignin is often the most difficult portion of plant biomass to degrade, with fungi generally thought to dominate during late stage decomposition. Lignin in feedstock plant material represents a barrier to more efficient plant biomass conversion and can also hinder enzymatic access to cellulose, which is critical for biofuels production. Tropical rain forest soils in Puerto Rico are characterized by frequent anoxic conditions and fluctuating redox, suggesting the presence of lignin-degrading organisms and mechanisms that are different from known fungal decomposers and oxygen-dependent enzyme activities. We explored microbial lignin-degraders by burying bio-traps containing lignin-amended and unamended biosep beads in the soil for 1, 4, 13 and 30 weeks. At each time point, phenol oxidase and peroxidase enzyme activity was found to be elevated in the lignin-amended versus the unamended beads, while cellulolytic enzyme activities were significantly depressed in lignin-amended beads. Quantitative PCR of bacterial communities showed more bacterial colonization in the lignin-amended compared to the unamended beads after one and four weeks, suggesting that the lignin supported increased bacterial abundance. The microbial community was analyzed by small subunit 16S ribosomal RNA genes using microarray (PhyloChip) and by high-throughput amplicon pyrosequencing based on universal primers targeting bacterial, archaeal, and eukaryotic communities. Community trends were significantly affected by time and the presence of lignin on the beads. Lignin-amended beads have higher relative abundances of representatives from the phyla Actinobacteria, Firmicutes, Acidobacteria and Proteobacteria compared to unamended beads. This study suggests that in low and fluctuating redox soils, bacteria could play a role in anaerobic lignin decomposition.",
"introduction": "Introduction There is a strong impetus both nationally and internationally for devising new, non-fossil based fuels that are generated in a sustainable way with minimum greenhouse gas production [1] . Plant biomass derived from either crop waste or dedicated feedstocks such as switchgrass ( Panicum virgatum ) could potentially provide energy via biofuels if a system for unlocking this energy were devised that was robust, efficient and inexpensive [2] . One hurdle in cellulosic biofuels engineering is the presence of lignin, which can comprise up to 25% of plant biomass in herbaceous plants [3] . While pretreatment eliminates most of the lignin during biofuels production, lignin can pose a challenge due to its ability to inhibit cellulosic enzymes and as a potentially viable waste feedstock [4] , [5] . Lignin is a complex heteropolymer linked to cellulose, giving plants structural integrity. The deconstruction of lignin and its dissociation from cellulose presents a challenge for soil microbes and biofuels engineers alike. The repeating units of phenolic monomers, p -coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol, are synthesized in different ratios and combinations depending upon the type of plant, and so conferring its structural characteristics. The best understood mechanism for breaking open the rings in the lignin phenols belongs to fungi, specifically via oxygen free radical attached by the enzymes dioxygenases [6] , generally requiring oxic conditions. The known potential lignin-degrading bacteria are mostly derived from guts of wood-eating insects and include Alphaproteobacteria, Gammaproteobacteria and Actinomycetes [7] , with the best-characterized being Streptomyces viridosporus \n [8] . Phenol-degrading bacteria such as Kocuria and Staphylococcus \n [9] , peroxidase-producing Flavobacterium meningosepticum \n [10] , and bacterial degraders of polyaromatic hydrocarbons [6] may also have a natural ability for degrading lignin derived from decomposing plant biomass. Discovery of novel anaerobic bacterial lignin-degrading enzymes would be beneficial to the industrial production of next-generation biofules, due to their potential application to microbial engineered biofuels-producing organisms, lack of requirement of oxygen, and range of specificity or environmental conditions. Plant litter quality is a key controller of decomposition rates in soils, and lignin and the lignin∶N ratio play a particularly important role in late stage decomposition [11] , [12] . Humid tropical forest soils have the fastest rates of above- and belowground plant litter decomposition globally [12] . Near complete decomposition of a wide range of plant tissues has been recorded over 1–2 years in these ecosystems [12] , [13] , [14] , [15] . This rapid and complete decomposition belowground is surprising given the low and variable redox conditions typical of humid tropical forest soils [16] . The combination of fast decomposition and low and fluctuating redox suggests the presence of efficient anaerobic or facultative lignin-degrading microorganisms in the soils. While generally it is believed that fungi dominate plant decomposition and lignin degradation [17] , few fungi are able to tolerate anoxic conditions [18] , [19] . Thus, humid tropical forest soils are ideal sites to explore the potential for bacterial lignin degraders. Humid tropical forest soils house an immense and unexplored microbial diversity [20] , extremely high biomass [21] , and a microbial community that is very productive and uniquely fueled by the high iron present in these strongly weathered soils [22] , [23] . They present an attractive target for discovery of novel enzymes and pathways for deconstruction of plant material and improvement of efficiency of biofuels production derived from cellulosic feedstocks. In this study we used lignin-baited ‘bio-traps’ to investigate the microbes and enzymes responsible for lignin decomposition in Puerto Rico tropical forest soils.",
"discussion": "Discussion This study demonstrates that the lignin-amended biosep beads are an effective method for trapping soil populations with the specific capability of decomposing lignin. Substantial phenol oxidase and peroxidase accompanied by depressed carbohydrate-active enzyme activity and low microbial community richness after one week suggests the capture of a fairly specialized group of microorganisms adapted to the lignin-amended bead environment. There were a number of taxa that were dominant early on in the experiment and more abundant in lignin-amended than unamended bead communities, which presumably play a role in lignin decomposition in the soil. Bacteria known to break down lignin are concentrated in the Alphaproteobacteria, Gammaproteobacteria, and Actinomycetes [7] . Taxa in the class Alphaproteobacteria were the most dominant taxa from the earliest sampling time point, and significantly enriched in lignin beads compared to unamended beads. The Alphaproteobacteria picked up by the PhyloChip were closely related to Caulobacter intermedius and Brevundimonas diminuta , and these taxa are known catalase producers. Caulobacter crescentis is an obligate aerobe that produces catalase likely as protection from oxidative stress in late-stationary phase in culture [40] . Rhodomicrobium is an Alphaproteobacteria in the family Rhizobiales that was detected by the pyrosequencing analysis, and also a known purple non-sulfur bacterium. While taxa in this genus are able to link iron reduction and denitrification to photosynthesis [41] , [42] , their role in below-ground lignin decomposition likely involves their ability to fix nitrogen [19] . The Gammaproteobacteria we detected were in the Enterobacteraceae, closely related to the Escherichia spp. observed as lignin-degrading from the guts of wood-boring beetles [43] . Likewise the Actinomyces we observed were only distantly related to the well-characterized Streptomyces viridosporus and Rhodococcus spp. demonstrated to have lignin degrading activity [8] , [44] . This departure is likely due to the many differences between tropical forest soils and the wood-eating insect gut environment where bacterial lignin degradation is well-documented. The lignin beads are going to pick up lignin-degrading bacteria as well as bacteria able to live on little to no carbon and also tolerate the presence of lignin and potentially toxic lignin byproducts of decomposition. However, the scarce availability of oxygen in these soils [16] accompanied by high amounts of iron and iron cycling [45] suggests potential non-oxidative mechanisms of lignin decomposition. Frequent episodes of soil anoxia have been observed in these soils and are known to affect the microbial community [22] , [46] . This fluctuating redox may facilitate the development of lignin-amended bead microbial communities with a diversity of mechanisms for decomposition. Fermentation is likely to play a role in anaerobic metabolism of complex carbon, evidenced by dominance of Bacilli in the phylum Firmicutes in the lignin-bead populations. Fermenters like Enterobacteriaceae has been observed to out-compete obligate anaerobes under similar conditions [47] . The Bacteroidetes bacterium Flavobacterium meningosepticum was isolated from soil and shown to not demonstrate catalase activity, though it has the ability to grow on phenolic, model lignin compounds as sole C and energy source [10] . Because labile carbon is limiting to soil microbes, we might expect lignin decomposition and assimilation to also be linked to denitrification, sulfate reduction, iron reduction, and methanogenesis, encompassing the range of metabolisms previously observed in these soils [21] , [22] . Fungi are generally considered the main microbial decomposers of plant material [18] , [19] , [48] , though we hypothesize that their role in tropical forest soils is diminished because of frequent anaerobic soil conditions [16] . Fungi were detected in the pyrotag data, but comprised a relatively small portion of the richness (<5.5%) and evenness (<5.8%), with more fungi in the unamended compared to the lignin-amended beads ( Table S4 , S5 ); phylogenetic information from metagenomic analysis supports this hypothesis [20] . We detected an abundance of acid-tolerant strains such as from the phylum Acidobacteria, which were enriched in the lignin-amended beads relative to the controls and have been found with decomposing fungi where perhaps their acid tolerance confers a competitive environment in the decomposing litter [49] . In anaerobic systems Actinobacteria and filamentous bacteria may play the role of fungi, producing phenol oxidases and peroxidases [19] . Members of the genus Kocuria (Actinobacteria) and Staphylococcus (Firmicutes) were previously described as phenol-degraders in soil [9] and were also detected in our beads. There was a strong effect of time on the microbial community structure and function ( Table 1 , Figure 2 ) suggestive of microbial community succession. On a natural plant substrate, the initial community would have grown utilizing the more accessible cellulose and hemicellulose components before leaving the more resilient lignin compounds for the later stage communities [50] . In this respect, the beads are selecting directly for the organisms that are able to access the complex plant biomass that is characteristic of late stage decomposition. There was no cellulose substrate on the beads when they were buried, but these enzyme activities represent potential activities likely due to colonization of microbes. The increase in bacterial richness as well as cellulase activity towards the end of the experiment suggests that the lignin created relatively unfavorable conditions for the majority of the soil microbial community. While the pyrosequencing and PhyloChip microarray microbial community profiles agreed well with each other, there were some differences at the species-level identification of lignin bead-associated microbial taxa due to the fact that these two methods assay microbial communities in very different ways. Though both begin with PCR amplification, the primers used are tailored to each method; for pyrosequencing, the universal primers are designed to capture as much of the bacteria, archaea, and eukaryota in any environmental sample [32] , while the PhyloChip was designed around primers that capture as much of the 16S rRNA gene (bacteria and archaea only) as possible [51] . Though PCR amplification will distort relative abundances in mixed communities, pyrosequencing has the potential to more faithfully maintain relative abundances, while PhyloChip is sensitive enough to amplify and detect even quite rare members of the microbial community [52] , [53] . Both methods are intended to provide a microbial community profile of specific environments, where the association with lignin beads suggests tolerance or utilization of lignin, though further studies are required to understand which taxa are responsible. The lignin-baited biosep beads appear to be efficient bio-traps for capturing lignin-degrading microbial populations, baited with commercial alkali lignin and tested with L-DOPA, phenolic model lignin compounds which bear structural similarities to carbon compounds found in the environment like humics, lignin breakdown products and contaminants. The aromatic compounds benzoate, phenylpropionate and phenylacetate are produced as natural by-products in the anaerobic rhizosphere of rice field soil [54] . Some of these same compounds are formed in anaerobic fermentation reactions and can inhibit cell growth, biofuels production, or both [55] . So although we cannot directly assay the microbes active in late-stage decomposition through this method, we are able to identify and measure the activity of phenol-oxidase producing populations. The data taken together suggest that the lignin had an adverse effect on all but a specific subset of the microbial community, and this select population is likely able to enzymatically access and assimilate carbon derived from the lignin. Phylogenetic analysis also demonstrated a significant increase in the diversity and clustering of the community on lignin-amended beads compared to unamended beads, suggesting that the lignin either directly created an chemical environment unfavorable to all but a small population of bacteria, or the taxa initially able to utilize the lignin had a competitive advantage and out competed later immigrant populations. The molecular mechanisms of this largely anaerobic lignin-degrading population are of interest and under investigation."
} | 3,646 |
26474966 | PMC4802824 | pmc | 2,807 | {
"abstract": "Microbial electrochemical systems exploit the metabolism of microorganisms to generate electrical energy or a useful product. In the past couple of decades, the application of microbial electrochemical systems has increased from the use of wastewaters to produce electricity to a versatile technology that can use numerous sources for the extraction of electrons on the one hand, while on the other hand these electrons can be used to serve an ever increasing number of functions. Extremophilic microorganisms grow in environments that are hostile to most forms of life and their utilization in microbial electrochemical systems has opened new possibilities to oxidize substrates in the anode and produce novel products in the cathode. For example, extremophiles can be used to oxidize sulfur compounds in acidic pH to remediate wastewaters, generate electrical energy from marine sediment microbial fuel cells at low temperatures, desalinate wastewaters and act as biosensors of low amounts of organic carbon. In this review, we will discuss the recent advances that have been made in using microbial catalysts under extreme conditions and show possible new routes that extremophilic microorganisms open for microbial electrochemical systems.",
"introduction": "INTRODUCTION Microbial electrochemical systems (MESs) describe all electrochemical devices that make use of microbial catalysts to drive or accelerate electrochemical reactions at the anode, the cathode or both electrodes (Rabaey and Rozendal 2010 ; Logan and Rabaey 2012 ). An MES in which energy is recovered is a microbial fuel cell (MFC), while a system which requires the input of electrical energy to drive a reduction reaction is a microbial electrolysis cell (MEC) (Logan et al. 2006 , 2008 ). The discovery that an electrical current can be derived directly from organic electron donors (e.g. acetate) led to the testing of a nearly endless number of substrates for microbial-assisted production of electrons (Pant et al. 2010 ). At first, this production was mainly used to (i) extract the energy available from wastewaters (Logan et al. 2006 ), (ii) remove recalcitrant organic and inorganic compounds (Kim et al. 2008 ; Catal, Bermek and Liu 2009 ; Mu et al. 2009 ; Ter Heijne et al. 2010 ; Virdis et al. 2010 ), (iii) extract energy from plant rhizodeposits (Strik et al. 2011 ) or (iv) power small-scale off-grid devices and sensors (Shantaram et al. 2005 ). These applications have evolved into chemical or microbial catalyzed generation of products with a higher added value from the produced electrons, such as hydrogen (Logan et al. 2008 ), methane (Cheng et al. 2009 ), hydrogen peroxide (Rozendal et al. 2009 ), and short- and medium-chain fatty acids (Steinbusch et al. 2011 ; van Eerten-Jansen et al. 2013 ). The more recent discovery that these electrons can be used to drive microbial metabolism opens up a whole new world of possible products of which we cannot yet see the full impact (Rabaey and Rozendal 2010 ; Nevin et al. 2011 ). Nowadays, even the ionic current generated with the production of the electrical current is used to drive separation processes such as the recovery of ammonia (Kuntke et al. 2012 ), the production of alkalinity (Sleutels, Hamelers and Buisman 2010 ; Modin et al. 2011 ) or to provide additional energy to drive microbial activity (Logan and Elimelech 2012 )."
} | 845 |
39966672 | PMC11836136 | pmc | 2,809 | {
"abstract": "Heat stress and other factors cause the loss of endosymbiotic dinoflagellates by corals, and is known as coral bleaching. Coral reef bleaching is a global environmental problem. To better understand corals’ responses and adaptability to stressful conditions, we applied a lipidomic approach in combination with cytometry and microscopy to study the coral bleaching of Acropora aspera under heat stress (32 °C) and subsequent recovery. For eight days of bleaching, the coral lost 50% of its symbiont population and 100% after a week of recovery. It took 126 days to fully recover the symbiont population, content of chlorophyll a and reserve lipids. There were degradations in symbionts’ thylakoids and disruption of thylakoid lipid homeostasis. Variations in the content of phosphatidylinositols involved in apoptosis and autophagy and changes in the molecular profile of glycosylceramides that may be involved in the sphingosine rheostat were observed. However, upon A. aspera bleaching, the loss of symbionts was compensated by increased mucociliary nutrition. An increase in the content of hydroxylated ceramideaminoethylphosphonates for membrane stabilization and a decrease in ether phosphatidylethanolamines for providing protection from oxidative stress may have been used as adaptation mechanisms by the coral host. Thus, the coral undergoes physiological and biochemical changes during heat stress that are aimed at mitigating the adverse destructive effects, which may be key to successful recovery. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-025-90484-4.",
"introduction": "Introduction Corals establish complex interactions with a wide range of microorganisms that constitute an important part of the entire metaorganism referred to as holobiont 1 . The coral symbionts are intracellular microalgae (dinoflagellates of the family Symbiodiniaceae). These obligate symbionts are found at high densities in the gastrodermis of corals 1 where they photosynthesize nutrients for the coral host 2 . The mutualistic endosymbiosis with dinoflagellates promotes active coral growth and formation of coral reefs that are the most important ecosystems in the world’s oceans 3 . As sea temperatures rise due to global warming, corals are losing their symbiotic dinoflagellates 4 . This phenomenon is known as coral bleaching 5 . The large-scale coral reef bleaching leading to mass mortality, associated with heat stress and other environmental factors, is now a critical global environmental problem 6 – 8 . Coral responses and adaptability to thermal stress can vary within and among species depending on the degree of accumulated thermal stress 9 , as well as the ability of the coral host to associate with various Symbiodiniaceae endosymbionts 10 – 13 , being influenced by several factors such as habitat conditions that may act simultaneously. The bleaching process most likely begins in chloroplasts of symbiotic dinoflagellates when the accumulation of reactive oxygen species (ROS) leads to a coral bleaching cascade 14 , 15 . Bleaching can be considered as a process of detection and destruction of opportunistic symbionts by the coral host 16 , which may trigger the loss of symbionts through exocytosis 17 , 18 . Another mechanism for the loss of symbionts including apoptosis and autophagy has also been described 16 , 19 – 22 . Not only the innate immunity of the coral host but also the inter-partner nutrient dynamics is involved in the onset, ongoing maintenance, and dysregulation of symbiosis 23 . To better understand the coral’s response and adaptability to heat stress, large-scale studies are needed that would include not only the cell biology but also biochemical aspects. Lipids play an important role in maintaining the health and metabolic balance of the coral organism 24 , 25 . Neutral lipids (triacylglycerides (TG), monoalkyldiacylglycerides (MADAG) and wax esters (WE)) serve as the major depot of reserves and source of energy, while phospholipids (PL), glycolipids (GL), sphingolipids (ceramideaminoethylphosphonate (CAEP) and glycosylceramides (GlcCer)) and sterols perform structural functions and form the basis of cell membranes 26 , 27 . The lipid composition of each taxon largely depends on its genetic capability of synthesizing certain lipid molecules. Coral holobiont’s lipids are a mixture of lipids from Symbiodiniaceae, the host coral, and other members of the coral microbiome. Several classes of GL such as sulfoquinosyldiacylglycerol (SQDG), mono- and digalactosyldiacylglycerol (MGDG and DGDG), as well as phosphatidylglycerol (PG), are the most important components in membranes of the plant photosynthetic apparatus and constitute the major part of lipid content in Symbiodiniaceae 28 , 29 . Extraplastidal membranes of coral symbionts, in addition to diacyl forms such as phosphatidylcholine (PC), phosphatidylethanolamine (PE) and phosphatidlinositol (PI) and sphingolipids GlcCer 29 , also contain betaine lipids (BL), e.g., diacylglycerylcarboxymethylcholine (DGCC) 28 – 31 . The major classes of PL in corals are PC, PE, phosphatidylserine (PS), and phosphatidlinositol (PI). Typical classes of cnidarian lipids, the ether (alkyl- or alkenylacyl) PC, PE, PS, PI, and the sphingophosphonolipid CAEP, are absent from lipids of Symbiodiniaceae and can be considered as markers of coral host tissues 29 , 32 , 33 . To elucidate the role of lipids in the mechanisms of bleaching and adaptation after stress, we set up an experiment to study the bleaching responses of the reef-building coral, Acropora aspera (Cnidaria: Anthozoa: Hexacorallia: Scleractinia: Acroporidae), exposed to heat stress (8 days, 32 °C) and the subsequent recovery (126 days, 27 °C). The dynamics of molecular profiles of storage lipids and membrane lipids of the coral, photosynthetic pigments in symbiotic dinoflagellates, and morphological and physiological changes for both the coral host and symbionts were performed by different analytical techniques in combination with the microscopy and cytometry.",
"discussion": "Discussion For the reef-building coral A. aspera , the bleaching period was eight days with exposure to a temperature of 32 °C, which resulted in only a twofold decrease in symbiont density. The density of symbionts decreased to almost 0% a week after, when the temperature was decreased to 27 °C. A similar pattern of symbiont loss following a bleaching period was shown in a study on the reef-building coral Stylophora pistillata . The density of symbionts per unit surface area of the S. pistillata colony decreased by approximately 20% during the bleaching period at a temperature of 32 °C and subsequently dropped to zero values five days after, when the temperature was decreased to 28 °C 34 . In our study, it took 126 days for A. aspera to recover the symbiont density values to the initial numbers, the content of chlorophyll a as an indicator of the photosynthetic activity of the symbionts, and the TG content as an indicator of the energy reserves of the organism to the levels before the bleaching period. The reef-building coral A. aspera was relatively resistant to bleaching under heat stress and recovered faster than the soft coral S. heterospiculata , which lost most of its symbionts after 2 days at 32 °C and was recovering during 205 days 35 , 36 . For other reef-building corals, Agaricia agaricites and Orbicella franksi , bleaching also appeared to occur more slowly (19–31 and 24–38 days, respectively), which can be related with Symbiodiniaceae species 37 . Associations with symbionts whose thylakoid lipidome has thermotolerant properties may be an explanation for the better resistance of the coral A. aspera to heat stress. In our previous study, the coral A. aspera hosted one Symbiodiniaceae species, the acroporide-specific Cladocopium 29 , that showed lipidome features characteristic of the thermotolerant dinoflagellate Durusdinium trenchii 30 , 38 . Response of the A. aspera symbiotic dinoflagellate to heat stress The response to heat stress occurred in both the host coral A. aspera and its symbionts at the physiological and biochemical levels. As the microscopy analysis showed, during the bleaching period and in the first half of the recovery period, disturbances occurred in the structure of thylakoids in the symbionts of A. aspera . In an organism under stable environmental conditions, the membrane lipidome in thylakoids is in a state of homeostasis, which is manifested as a constant ratio of thylakoid lipids 39 . The SQDG/PG ratio and the ratio of bilayer-forming DGDG to non-bilayer MGDG are crucial for proper physiological functioning of the thylakoid membrane and have constant values for a particular photosynthetic organism 40 . The ratios of these lipids changed during the experiment: the SQDG/PG and the DGDG/MGDG ratio decreased after the bleaching period. The DGDG molecule is formed through addition of galactose unit to the MGDG molecule by DGDG synthase 41 . If its action is disrupted, the biosynthesis of DGDG cannot continue, which leads to a change in the DGDG/MGDG ratio. Furthermore, not only the action of synthases but also lipid transport can be disrupted. In A. aspera symbionts, the profile of GL molecular species was subject to restructuring, especially in the example of MGDG and DGDG. Their molecular species with 22:6 and 20:5 PUFA decreased. These PUFA are formed by elongation and desaturation reactions on the endoplasmic reticulum (ER) and then imported to plastids 41 . During the bleaching period, we observed the appearance of apoptotic bodies in the A. aspera symbionts and an increased content of dead symbiont cells. At the beginning of the recovery period, when symbionts were nearly absent, almost all of them were in an apoptotic and/or necrotic state. Such changes at the biochemical level should be manifested as variations in plasma membrane lipids. In lipids of the A. aspera symbionts, we observed a decrease in the PC content relative to the proportion of symbionts, while, in the case of MGDG and DGDG, there was a decrease in the molecular species with 22:6 PUFA (PC 22:6/22:6). PC is a source of the 22:6 PUFA moiety for biosynthesis of galactolipids 41 , and, thus, a decrease in PC molecular species with such PUFA may lead to a decrease in the molecular species of the respective GL molecular species. While the molecular species of GL and PC with 22:6 PUFA decreased, the profile of the TG molecular species changed with an increase in the TG 22:6/16:0/18:4 content. TG is biosynthetically closely associated with structural lipids. In higher plants, there is an acyl-CoA independent pathway involving PC as an acyl donor 42 . TG, in turn, can act as an acyl donor for GL. Such acyls are transferred from the ER to chloroplast by trigalactosyldiacylglycerol (TGD) transport protein. As was shown in the example of a Chlamydomonas TGD mutant, MGDG synthase is strongly stimulated but with TG accumulation due to the defective lipid trafficking 43 . In our case, there might also be a disruption in the transport of lipids from the ER to chloroplast, which led to the observed restructuring in the profile of the molecular species of GL, PC and TG lipids. In addition to the possible disruption of lipid biosynthesis and transport, an attempt to adapt to stress was, nevertheless, detected in symbionts. The main chlorophyll a , which is located in the reaction centers of photosystem I (PSI) and photosystem II (PSII) and is also the major component of light-harvesting complexes 44 , decreased compared to chlorophyll b . The latter was detected initially during the bleaching period, and its content became comparable to that of the main chlorophyll a . Previously, we observed similar dynamics of chlorophyll b content in S. heterospiculata 35 . This chlorophyll b dynamics is likely a common feature of symbiotic corals during bleaching events. The bleaching process begins in the PS II reaction center of coral symbionts 14 , 15 , and chlorophyll b is contained as an accessory pigment of PSII 44 where repackaging of chlorophyll b may take place as a photoprotective mechanism 44 , 45 . Response of the A. aspera host to heat stress exposure The innate immunity of the coral host plays a crucial role in the symbiosis of cnidarians and dinoflagellates 23 . Elevated temperatures that cause oxidative stress in the symbionts also result in a stress response resembling the immune response in the coral host, which causes the loss of symbionts. As shown by microscopy examinations, the symbionts of A. aspera that remained after the bleaching period were digested in symbiophagasomes. At the biochemical level, changes were also observed, potentially related to the immune system involved in the bleaching process A. aspera . The decrease in symbiont density also occurred simultaneously with a decrease in PL, including PI, which were involved in apoptosis and autophagy 22 , 46 . Sphingolipids also play a role in the mutualistic cnidarian–Symbiodinium symbiosis 23 , which we also showed in our study. The accumulation of signaling sphingolipids that comprise the sphingosine rheostat, pro-apoptotic sphingosine (Sph) and pro-survival sphingosine-1-phosphate (S1P), is a key to determining the cell fate 47 . Stressful conditions such as heat stress result in the disruption of the photosynthate transport from symbionts to the host and in the re-engagement of the innate immune system. The sphingosine rheostat is pushed toward pro-apoptoptic sphingosine, and the host mounts an immune response to eliminate the stressed symbiont. This dysregulation causes dysbiosis and bleaching 23 . During the bleaching period, we observed a threefold increase in the number of host’s apoptotic cells and, simultaneously, a decrease in GlcCer of the host A. aspera (19:2b/20:0OH and 19:2b/22:0OH). Furthermore, during the recovery period, the content of these GlcCer molecules increased simultaneously with a decrease in apoptotic host’s cells. There might be an increase in the content of pro-apoptotic sphingosine during the A. aspera bleaching, which, in turn, could lead to a decrease in biosynthetically bound GlcCer. During the recovery, we observed the opposite effect. Thus, the sphingosine rheostat can play a key role in determining the fate of cells of the host A. aspera during bleaching and recovery. In addition to the stress and immunity cascades, metabolic dysregulation is a critical component of dysbiosis and bleaching processes 23 . We observed a simultaneous decrease in the density of A. aspera symbionts and the content of TG reserve lipids. As we showed earlier, the symbiotic coral Junceella fragilis had a 30-fold higher supply of nutrients in the form of TG than that in asymbiotic corals 48 . When the coral was kept at a temperature of 29 °C, there was an increase in the number of lipid droplets in the gastroderm cells and symbionts. Small-diameter lipid droplets were found in symbiosomal membranes, in particular under the plasma membrane of dinoflagellates, in the periplast, and in the cytoplasm of the host cell near these structures. Thus, lipids were probably transferred from the symbionts to the host cells. The symbiosome membrane complex is critical component of the cnidarian–dinoflagellate symbiosis 49 where nutrients are transported from the symbionts to the coral host 23 . Damaged symbionts, incapable of normal photosynthesis, cease to be a source of nutrients for the coral. During the period of bleaching, at 32 °C, the reserve substances of the symbionts such as lipid droplets, starch grains, vacuoles with crystalline structures and the storage body’s components were consumed. The coral host did not obtain the necessary nutrients, as was confirmed by a decrease in storage lipids (TG). There was a disruption of the interpartner dynamics of nutrients. Under the conditions of symbiosis is breakdown, the host organism attempts to adapt to the lack of nutrients resulting from the absence of a sufficient density of symbionts. At 29 °C and during the bleaching period at 32 °C, we observed an intensive mucus secretion and an increase in the number of phagosomes, probably with captured non-symbiotic algae. This was especially noticeable in the epidermis, where normally phagosomes were few and the secretion was not so intense. The increased mucus secretion by epidermal cells (myoepithelial cells of the pharynx and epithelial cells of the coenosarc wall) and mucocytes caused a large number of areas with “empty” cells to appear, in which most of the cytoplasm was occupied by a large vacuole. Hypertrophied and empty mucocytes of the epidermis are characteristic of bleached corals 50 – 52 . Mucus helps to capture food particles and phyto- and zooplankton. It is likely that in response to the decrease in autotrophic nutrition (due to lost symbionts), the contribution of heterotrophic mucociliary nutrition grows. An increase in the role of heterotrophic nutrition with the seasonal increase in temperature has been described from some corals, e.g., Montastraea annularis 53 and Millepora platyphylla 54 . The heterotrophic plasticity of corals compensates for the reduction in carbon supply from symbiotic dinoflagellates 55 , 56 . The high density of mucocytes in the coenosarc epidermis at the beginning of the recovery period, when the density of symbionts decreases, is likely associated with increased mucociliary nutrition. Changes were also identified at the biochemical level, which were most likely aimed at adaptation of the coral. The exposure of A. aspera to heat stress led to a rearrangement of the profile of PE molecular species. After the bleaching period, we observed an increase in diacyl molecular species of PE in A. aspera . The observed a decrease in their level in A. aspera lipids (16:1alk/20:5, 16:0alk/20:4, 16:1alk/20:4, 18:1alk/20:5, 18:1alk/20:4, 37:5alk, 19:1alk/20:4, and 20:2alk/20:4), which led to an increase in the relative content of diacyl molecular species of PE. Ether lipids (lipids with an ether bond at the sn -1 position) are endogenous antioxidants 57 . It is likely that the observed decrease in their levels was caused by oxidative stress during the A. aspera bleaching. Molecules of ether PE may act as a defense against oxidative stress by scavenging ROS. We also observed changes in the sphingolipid CAEP molecular profile. The was an increase in the content of hydroxylated CAEP, previously shown for the reef-building coral A. cerealis 58 . These sphingolipids, like PL, are cell-membrane structural components and possibly play a specific role in membrane permeability and stabilization by protecting the membrane from the effect of hydrolytic enzymes such as phospholipases and phosphatases 47 . Thus, the increased content of CAEP with hydroxyl groups can be considered as an additional mechanism to stabilize the membrane through the formation of intermolecular hydrogen bonds. Thus, during the A. aspera bleaching, the symbiosis was broken down within 8 days of heat stress exposure. One of the causes is known to be the coral host’s immune response aimed at expulsion and in situ degradation of symbionts. This is manifested at the biochemical level, as we have demonstrated: a decrease in PI and a change in the GlcCer molecular species profile. Another probable cause of the bleaching and symbiosis breakdown lies in the disruption of general metabolism. Serious degradation occurs in thylakoids with probable disruption of lipid transport and biosynthesis. As a result, the symbionts stop supplying nutrients in abundance which is evidenced by a decrease in not only reserve TG but also structural PL. Under such conditions, the coral makes attempts to adapt, primarily, to compensate for the loss by increasing the heterotrophic nutrition. Changes also occur in membranes at the lipid level: an increase in the content of hydroxylated CAEP for membrane stabilization and a decrease in ether PE as protection from oxidative stress. Symbionts also demonstrate photoprotective mechanism by varying the content of photosynthetic pigments. These adaptive mechanisms may be a crucial mechanism to maintain the coral’s viability until environmental conditions become favorable to recover. As we demonstrated, the density of A. aspera symbionts recovered and TG returned to a normal level. Coral species, which contain symbiotic dinoflagellates with different thermotolerances, are expected to have different duration of bleaching and recovery, however, general patterns of changes in lipidomic profiles are observed. These changes are believed to be associated both with the action of the host innate immune response and also with a disruption of the general metabolism of the symbiotic coral organism. In this work, probable adaptive mechanisms aimed at stabilizing and protecting membranes from oxidative stress are supposed. These issues require further study and involvement of more coral species."
} | 5,271 |
29239113 | PMC6011928 | pmc | 2,811 | {
"abstract": "Summary To enrich syntrophic acetate‐oxidizing bacteria ( SAOB ), duplicate chemostats were inoculated with sludge from syntrophic acetate oxidation ( SAO )‐dominated systems and continuously supplied with acetate (0.4 or 7.5 g l −1 ) at high‐ammonia levels. The chemostats were operated under mesophilic (37°C) or thermophilic (52°C) temperature for about six hydraulic retention times ( HRT 28 days) and were sampled over time. Irrespective of temperature, a methane content of 64–69% and effluent acetate level of 0.4–1.0 g l −1 were recorded in chemostats fed high acetate. Low methane production in the low‐acetate chemostats indicated that the substrate supply was below the threshold for methanization of acetate via SAO . Novel representatives within the family Clostridiales and genus Syntrophaceticus (class Clostridia) were identified to represent putative SAOB candidates in mesophilic and thermophilic conditions respectively. Known SAOB persisted at low relative abundance in all chemostats. The hydrogenotrophic methanogens Methanoculleus bourgensis (mesophilic) and Methanothermobacter thermautotrophicus (thermophilic) dominated archaeal communities in the high‐acetate chemostats. In line with the restricted methane production in the low‐acetate chemostats, methanogens persisted at considerably lower abundance in these chemostats. These findings strongly indicate involvement in SAO and tolerance to high ammonia levels of the species identified here, and have implications for understanding community function in stressed anaerobic processes.",
"introduction": "Introduction Syntrophic acetate‐oxidizing bacteria (SAOB) drive anaerobic conversion of acetate to methane in high‐ammonia biogas processes and thus play a major role in many commercial‐scale production systems (Karakashev et al ., 2006 ; Sun et al ., 2014 ; Frank et al ., 2016 ; Mosbaek et al ., 2016 ). SAOB work in close association with hydrogenotrophic methanogens, performing a two‐step reaction in which SAOB convert acetate to H 2 /format and CO 2 , which are then used by the methanogens for production of CH 4 and CO 2 . Under conditions such as high ammonia, the syntrophic acetate oxidizers outnumber their competitors for acetate, the aceticlastic methanogens (Sun et al ., 2014 ). This poses challenges regarding biogas digester operation, as low‐acetate conversion rates by the syntrophs may limit the overall efficiency and stability of the process. However, tailoring operation to underpin the syntrophic interactions, such as allowance of microbial adaptation during start‐up and changed operating conditions, long retention times and addition of trace elements has been shown to be effective in improving performance and mitigates the effect of ammonia toxicity in SAO processes (Westerholm et al ., 2016 ). To date, only a few SAOB have been isolated and characterized. These are the thermophilic Thermacetogenium phaeum (Hattori et al ., 2000 ) and Pseudothermotoga lettingae (Balk et al ., 2002 ; Bhandari and Gupta, 2014 ), the thermotolerant Tepidanaerobacter acetatoxydans (Westerholm et al ., 2011a , b ) and the mesophilic [ Clostridium ] ultunense (Schnürer et al ., 1996 ) and Syntrophaceticus schinkii (Westerholm et al ., 2010 ). Presence and abundance of these SAOB during changed operating conditions in anaerobic systems have been indicated using species‐specific 16S rRNA gene‐targeting approaches (Westerholm et al ., 2011a , b , 2012a , b , 2015 ). However, known SAOB are often low in abundance in relation to the overall microbial community (Westerholm et al ., 2016 ). Their relative low abundance obstructs their detection, but also imposes limitations for identification of new potential SAOB using high‐throughput sequencing of complex anaerobic digester communities. Nevertheless, altering process operation towards SAO‐inducing conditions (such as increased ammonia level, injection of H 2 or elevated temperature) may affect the microbial community enough for detection of potential SAOB using sequencing approaches. Through such approaches, SAOB candidates have been suggested within the families Thermoanaerobacteraceae (which also includes the known SAOB T. phaeum and S. schinkii ) and Thermodesulfobiaceae (Ho et al ., 2014 ; Yamada et al ., 2014 ; Bassani et al ., 2015 ; Müller et al ., 2016 ) and the phylum Spirochaetes (Lee et al ., 2015 ). However, the inconceivable numbers of microbes and high microbial diversity in anaerobic digesters pose a major challenge when seeking to establish reliable links between abundant species and SAO‐function and further research is required to confirm species within these groups as SAOB. Most known SAOB are affiliated to the physiological group of acetogens, a feature that has been used to reveal further information about potential SAOB by targeting the fhs gene, encoding a key enzyme of both acetogenic and SAO metabolism (Müller et al ., 2016 ). Through this method, acetogenic groups unique to high‐ammonia biogas processes have been identified and are suggested to be involved in SAO (Müller et al ., 2016 ). Results of other techniques to track down potential SAOB, such as stable isotope‐based functional probing and meta‐omics, suggest members of the orders Clostridiales and/or Thermoanaerobacterales (Zakrzewski et al ., 2012 ; Lü et al ., 2014 ; Müller et al ., 2016 ), uncultured phylotypes affiliated with the Firmicutes (Frank et al ., 2016 ), the Thermotogae (Zakrzewski et al ., 2012 ; Nubo et al ., 2015 ) and the phylum Synergistes (Ito et al ., 2011 ) as candidates for SAO capacity. Taken together, the few SAOB isolates and the wide taxonomic diversity of the proposed SAOB currently pose an obstacle to predicting their function and behaviours in complex microbial communities. Identification of novel key players would thus be highly beneficial and would increase knowledge of SAO and help manage ammonia‐stressed anaerobic digesters and develop innovative operating guidance. The aim of this study was to enrich acetate‐degrading microbial communities using a continuous cultivation approach to preserve and enrich core acetate‐utilizing communities occurring in high‐ammonia biogas systems. Continuous feeding with acetate for a long period was expected to enable microbial enrichment and facilitate identification of prominent SAOB, which due to their relatively low abundance are difficult to detect in more complex environments. The enrichments were initiated with inocula taken from anaerobic digesters previously demonstrated to be dominated by SAO. Differing factors between the enrichment chemostats included acetate influent concentration (0.4 or 7.5 g l −1 ), temperature (37°C or 52°C) and inoculum source, which were selected with the objective of enriching SAO populations, occupying different niches with regard to acetate concentration and optimal temperature conditions. Other operating parameters of the parallel acetate enrichments were set to mimic the continuous biogas system that was the source of the inoculum, that is high free ammonia level (0.6–0.9 g NH 3 l −1 ) and ~30 day retention time. A combination of molecular methods, including Illumina sequencing of 16S rRNA genes, quantitative polymerase chain reaction (qPCR) and terminal restriction fragment length polymorphism (T‐RFLP) analysis, was used to identify microbial structure patterns over time and to quantify abundant species. This approach made it possible to focus on the metabolic group restricted to acetate degradation.",
"discussion": "Discussion Possible acetate dependencies of SAOB yet to be revealed A continuous enrichment approach was employed in the present study to select for syntrophic acetate oxidizers, which due to their relatively low abundance are difficult to detect in more complex environments. The experimental set‐up was designed to distinguish mesophilic and thermophilic key populations during continuous feeding of two different levels of acetate. Acetate‐dependent growth rate of SAOB has been indicated in co‐cultivation and metagenomic studies (Oehler et al ., 2012 ; Manzoor et al ., 2015 ; Müller et al ., 2015 ; Westerholm et al ., 2016 ). Estimation of the degree to which the acetate level shapes syntrophic acetate‐degrading communities in biogas digesters is complex, particularly as high acetate levels often co‐occur with high levels of ammonia, which is a strong driver of development of the microbial community (Werner et al ., 2014 ; De Vrieze et al ., 2015 ; Müller et al ., 2016 ). The set‐up of the present study was designed to shed more light on this subject. However, absence of methane formation in the mesophilic and thermophilic low‐acetate chemostats indicated that the influent acetate level was below the threshold for methane production via SAO. Aceticlastic methanogens are known to degrade acetate to concentrations below that level (Smith and Ingram‐Smith, 2007 ); however, in present chemostats, they were likely inhibited by the high ammonia level. Nevertheless, acetate was apparently consumed, judging by the lower acetate level in effluent than in influent. Consistent with this, Illumina sequencing analyses indicated presence of relatively diverse bacterial communities in M L and T L . These microorganisms might degrade acetate and/or remain in the chemostats by consuming compounds included in the medium for growth support or as reducing agents (i.e. yeast extract, cysteine). However, the microbial community profiles obtained for the low‐acetate chemostats were still useful as references to distinguish SAOB candidates prevalent in M H and T H . Given the declining trends in S. schinkii and T. acetatoxydans abundance in the M L and T L chemostats, the acetate level in the chemostats was below the threshold for SAO activity of these known SAOB. However, the acetate level did not affect the abundance of C. ultunense , indicating that this species was able to stay syntrophically active at low acetate levels or used cysteine supplied in the medium for its growth (Schnürer et al ., 1996 ). In particular at thermophilic temperature, the overall community sequencing and the T‐RFLP profiling revealed surprisingly high structural variations between duplicate chemostats (representing biological replicates). Quantification of known SAOB, however, revealed quite similar trends in the duplicate chemostats. The explanation for this and possible impact by microbial variations on community functions and performances are unclear at this point. Similar results have been shown in previous microbial studies of parallel digesters under high‐stress, showing diverse structures at genus and species level OTUs (Goux et al ., 2015 ; De Vrieze et al ., 2016 ). In line with present result, these studies report on similar and stable digester performances despite microbial divergences. This highlights the challenges in interpreting links between microbial dynamics, operating conditions and process performance and emphasizes the need for increased understanding within this area to search for potential sources of variability and significances for process performances. Differences in richness and dynamics between mesophilic and thermophilic enrichment communities mimic trends seen in complex anaerobic communities The initial strain richness and evenness were significantly lower in the thermophilic than in the mesophilic microbial communities, which are in line with other studies (Levén et al ., 2007 ; Guo et al ., 2014 ; Jang et al ., 2016 ). We observed declining trends for these indices in all mesophilic and thermophilic high‐acetate chemostats throughout operation, which was expected as feeding a restricted number of substrates limits the potential metabolic pathways used for microbial growth and thus likely affects the phylogenetic distribution. Notably, the M H communities were still significantly higher in richness at the end of the operating trial, indicating that mesophilic temperatures support a more diverse community than thermophilic conditions during a restricted feeding strategy. Alpha diversity of microbial communities displayed significantly decreased richness and evenness in the M L chemostats over time but, contradicting expectations, these indices remained similar at all T L sampling points. However, due to the higher initial level at mesophilic temperature, similar ranges were recorded at both temperatures by the end of the trial, indicating that the low content of yeast extract and/or cysteine included in the medium was enough to support growth of a relatively diverse microbiota at both temperatures. Highly abundant OTUs suggest that mesophilic Clostridiaceae sp. and thermophilic Syntrophaceticus sp. are drivers of syntrophic acetate degradation Despite the SAO‐selective conditions applied in the chemostats, previously known SAOB did not represent dominant populations. Thus, even though their abundances were higher in high‐acetate than in low‐acetate conditions (which suggests SAO activity), these SAOB were probably not the major acetate degraders in the systems investigated. However, the T‐RFLP profiling and partial fhs gene sequencing strongly indicated presence of bacteria previously found in high‐ammonia SAO‐dominated digesters (Westerholm et al ., 2015 ; Moestedt et al ., 2016 ; Müller et al ., 2016 ). Unfortunately, it is currently not possible to phylogenetically position these bacteria based on the fhs gene. However, the Illumina results highlighted ubiquity and high relative abundance of a Clostridiaceae sp. (OTU_M H C) and a Syntrophaceticus sp. (OTU_T H S), which strongly suggests them as novel SAOB candidates. The closest relatives of Clostridiaceae sp. OTU_M H C ( A. halophilus and A. oremlandii ) are moderately halophilic (optimum growth at pH 8) and fermentative bacteria (Fisher et al ., 2008 ; Wu et al ., 2010 ). Some characteristics of these Alkaliphilus species resemble the features of SAOB when grown in pure culture, such as similar substrate patterns, formation of formate and acetate as main fermentation products ( A. halophilus ) and respiratory capabilities (Schnürer et al ., 1996 ; Hattori et al ., 2000 ; Fisher et al ., 2008 ; Westerholm et al ., 2010 , 2011a , b ; Wu et al ., 2010 ). A. oremlandii also oxidizes acetate, with reduction of thiosulfate as the electron acceptor (Fisher et al ., 2008 ), which is a capability it shares with the SAOB T. phaeum (Hattori et al ., 2000 ). Alkaliphilus species have also been detected in high‐ammonia biogas digesters (> 0.3 g NH 3 ‐N l −1 /8–10 g NH + \n 4 ‐N l −1 , Kovács et al ., 2013 ; Müller et al ., 2016 ; Tsapekos et al ., 2017 ; Ziganshina et al ., 2017 ), suggesting that they play a critical role in biogas digesters operating under such conditions. Their presence has been attributed to the ability of certain members to encode crucial peptidases for proteolysis of proteins (Stolze et al ., 2015 ). Moreover, the levels of these bacteria have been shown to correlate with an ammonia‐induced shift from aceticlastic methanogenesis to SAO (Müller et al ., 2016 ). Phylogenetic analyses of the deduced FTHFS amino acid sequences positioned retrieved sequences (represented by the T‐RFs 58 and 108 bp) close to fhs genes of Alkaliphilus sp., which indicates Clostridiaceae sp. OTU_M H C. Other sequences assigned to the T‐RF 406 bp significantly increased in abundance during operation of both M H chemostats, positioned close to the known SAOB P. lettingae in the phylogenetic analyses and are thus considered interesting SAOB candidates. The taxonomic survey of the OTU_T H S, which represented a significant proportion of the bacterial community in the thermophilic T H chemostats, revealed phylogenetic relatedness (97% identity) with the cultured representative of S. schinkii . The mesophilic S. schinkii oxidizes acetate in association with the hydrogenotrophic M. bourgensis , tolerates high ammonium levels and has a growth temperature ranging from 25 to 40°C (Westerholm et al ., 2010 ). Despite the narrow temperature range for growth of this type strain, relatives to this species have been found in mesophilic biogas digesters (37–40°C; Westerholm et al ., 2011a , b , 2012a , b , 2015 ; Karlsson et al ., 2012 ; Moestedt et al ., 2014 ; Sun et al ., 2014 ; Müller et al ., 2016 ) and in digesters operating at moderate (42–45°C; Moestedt et al ., 2014 ; Westerholm et al ., 2015 ) and thermophilic (49–60°C; Weiss et al ., 2008 ; Sun et al ., 2014 ; Lebuhn et al ., 2015 ; Müller et al ., 2016 ) temperatures. Together, these findings indicate that, apart from the wide‐ranging temperature span, relatives to S. schinkii are able to remain active under varying operating conditions in terms of ammonia concentration, HRT, substrate feed and digester configuration. Despite the lower phylogenetic relationship between OTU_T H S and T. phaeum , the thermophilic growth condition was a shared capability of these species. Searches in the Blast database furthermore revealed that sequences with identical identity to OTU_T H S have been discovered in syntrophic propionate‐degrading communities (Sugihara et al ., 2007 ), indicating ability to degrade both propionate and acetate or involvement in acetate degradation in association with syntrophic propionate‐degrading bacteria. 16S rRNA gene sequences identical to OTU_T H S have previously also been found in thermophilic (53°C) dry anaerobic digestion of waste paper‐based medium and suggested to be involved in SAO (OTU 1‐1B‐29 in Tang et al ., 2011 ) The fhs gene sequencing indicated that the gene from OTU_T H S was not targeted by the current primers. Hence, further development of the fhs‐primers targeting thermophilic species is needed for a more complete coverage of potential SAOB. However, the significantly higher relative abundance of OTU_T H S in T H than in T L , the relatively close identity to the SAOB S. schinkii and T. phaeum and the detection of genes with high identity in systems with a high possibility of being SAO‐dominated indicate that OTU_T H S may be a hitherto undescribed species in the genus Syntrophaceticus that is capable of SAO. The strict hydrogenotrophs M. bourgensis and Methanothermobacter are likely SAO partners, but may compete with formate‐utilizing bacteria The dominance of the strictly hydrogenotrophic M. bourgensis and Methanothermobacter in the high‐acetate chemostats strongly suggests SAOB partnerships. The mcrA gene analysis for M H also suggested dominance of a M. bourgensis strain previously found in a biogas digester supplied with trace elements (OTU15 in Westerholm et al ., 2015 ) that was the inoculum source for our mesophilic enrichment chemostats. This parental biogas digester had relatively high richness of mcrA genes (~14 dominant OTUs with relatively even distribution, Westerholm et al ., 2015 ), and the dominance of one particular strain in the M H chemostats was thus somewhat surprising. A relevant finding in the previous study is that OTU15 was not detected in a digester operating under corresponding conditions, which did not receive trace elements (Westerholm et al ., 2015 ). This indicates that continuous feeding of the trace element‐rich medium in the M H chemostats may have been particularly advantageous for this M. bourgensis strain. Relatively low partial hydrogen partial pressure was another differing operating parameter between the parental digester and other digesters in the previous study (Westerholm et al ., 2015 ). While the influence of this particular strain on hydrogen removal remains to be investigated, the ability for hydrogen removal to low levels (due to its high affinity for hydrogen) and the tolerance to high ammonia levels have been suggested as key drivers for M. bourgensis suitability as a SAO methanogenic partner (Westerholm et al ., 2015 ; Neubeck et al ., 2016 ). This is also stressed by the frequent detection of M. bourgensis in SAO‐dominated biogas digesters (reviewed in Westerholm et al ., 2016 ). In thermophilic SAO digesters, Methanomicrobiales and Methanobacteriales ( Methanothermobacter ) are frequently reported as the dominant hydrogenotrophic methanogens (reviewed in Westerholm et al ., 2016 ). Similarly, the Illumina results in the present study demonstrated high relative abundance of Methanothermobacter . Methanobacteriales was also detected by qPCR analyses of the T H chemostats, but corresponding analyses for Methanomicrobiales indicated that it was present in considerably higher abundance in these chemostats. These conflicting findings obtained in Illumina sequencing and specific qPCR analyses of ratios between Methanomicrobiales and Methanobacteriales can be related to differences in specificity in the primers used for these analyses. Due to the need for high sensitivity, encompassing as many 16S rRNA variants as possible, the Illumina primers will likely not have the same specificity for archaeal methanogens as the qPCR primers. The relatively high abundances of Sulfurospirillum sp. in M H and Bacillus sp. in the T H chemostats were puzzling, as SAOB candidates have not previously been suggested to belong to these genera. However, negligible detection of these OTUs in M L and T L indicates involvement in acetate degradation. The strict anaerobic requirements and temperature ranges for growth reported for the isolated type strains of the closest relatives, that is S. alkalitolerans and B. infernus , are in agreement with the conditions in M H and T H2 . Furthermore, the substrate patterns of these bacteria resemble what has been observed in pure cultures of known SAOB (e.g. lactate, pyruvate, fumarate or glucose, Boone et al ., 1995 ; Schnürer et al ., 1996 ; Hattori et al ., 2000 ; Balk et al ., 2002 ; Westerholm et al ., 2010 , 2011a , b ; Sorokin et al ., 2013 ). However, S. alkalitolerans is also capable of utilizing formate and H 2 (with acetate as carbon source, Sorokin et al ., 2013 ), and all components for growth of this bacterium would thus be provided in the present acetate‐enriched chemostats, with formate or H 2 produced by SAOB (Müller et al ., 2015 ) and acetate present in the medium. B. infernus can also use formate with Fe(III), MnO 2 , trimethylamine oxide or nitrate as an electron acceptor, but cannot grow on acetate (Boone et al ., 1995 ). Hence, these data indicate that growth of B. infernus and Sulfurospirillum sp. may rely on supply of formate and/or H 2 from acetate oxidation by SAOB. This would thus lead to competition between B. infernus and Sulfurospirillum sp. and the hydrogenotrophic methanogens for substrate, although further investigation is required to establish such a scenario. However, despite the uncertain metabolic roles of Sulfurospirillum OTU_M H Ep and B. infernus in our enrichment chemostats, their detection confirms them to be ammonia‐tolerant. To conclude, the continuous enrichment approach applied in the present study permitted analysis of a community specialized in acetate conversion, with a limited confounding effect of other metabolic groups. As anticipated, the populations identified as dominant SAOB were present at very low abundances or below detection in the initial operating period. The continuous acetate feeding then strongly restricted the microbial community structure and decreased phylotype richness, which enabled detection of potential SAOB after a few HRT. Based on our results, we posit SAO capability of the highly abundant OTU_M H C (GenBank accession number MG356789 ) and OTU_T H S (GenBank accession number MG356790 ) in mesophilic and thermophilic conditions respectively. However, this requires further investigation. Considering the importance of acetate removal for the overall biogas production system, identifying acetate‐degrading strains capable of remaining active in the presence of high ammonia concentrations is of the utmost importance. Whether bioaugmentation of key SAO populations or/and altered operating conditions to support their activity is a suitable approach to improve biogas yield remains to be determined. Nonetheless, this is an area in which increased insights into how to predict behaviours of key microbial populations and support their activity shows great promise for maintaining robust performance under stressed conditions."
} | 6,150 |
36134226 | PMC9416888 | pmc | 2,813 | {
"abstract": "The past few decades have witnessed significant development in the field of artificially biomimicking extremely water repellent interfaces, developed mostly through tedious synthetic processes using synthetic/non-biodegradable polymers and fluorinated derivatives rendering health and environment related hazards. Only a few approaches furnish superhydrophobic materials that can withstand different harsh environments. Here, in this current design, naturally abundant and biodegradable bovine serum albumin (BSA) protein nanoparticles and cotton fibers are rationally selected for environment-friendly green synthesis of a highly sustainable and deformable artificial superhydrophobic material through strategic association of facile and rapid Michael addition reactions between amine and acrylate moieties under ambient conditions without the aid of any catalyst. This protein based nature-inspired interface can endure severe repetitive physical manipulations, abrasions and prolonged (30 days) chemical exposure i.e. extremes of pH, artificial sea water, river water and surfactant contaminated water. This highly durable and compressible superhydrophobic material was successfully exploited for efficient (above 2000 wt%), selective and repetitive removal of contaminating oils from aqueous phases under harsh chemical conditions. Such a durable biomimicking interface derived directly from serum protein following a facile synthetic approach would be useful for developing various other functional materials.",
"conclusion": "Conclusion To summarize, for the first time, a facile and scalable process for developing a nature-inspired water-repellent interface on an eco-friendly and naturally abundant fibrous substrate with a biodegradable protein macromolecule has been developed. The naturally synthesized bovine serum protein was rationally integrated for developing the essential hierarchical topography, and further, covalent modulation of appropriate chemistry conferred artificial superhydrophobicity. The BSA protein derived artificial superhydrophobic material was capable of sustaining repetitive (1000 times) and high (80%) compressive strain, different physical manipulations ( e.g. , bending, creasing, twisting), various abrasive (adhesive tape test, abrasive sand paper abrasion etc. ) physical challenges, prolonged exposure (30 days) to UV radiation and various severe chemical exposures, i.e. , extremes of pH, surfactant contaminated aqueous phases, artificial sea water, river water, etc. Moreover, the as-synthesized superhydrophobic material was further exploited for selective absorption (efficiency above 2000 wt%) based eco-friendly remediation of oil spills repetitively, irrespective of the density and viscosity of the contaminating oil/oily phases under various practically relevant challenges. This current approach has immense potential for developing different functional materials for various applications in practically relevant scenarios.",
"introduction": "Introduction Extremely water repellent interfaces are artificially synthesized by mimicking the essential features and chemistry present in the ‘lotus-leaf’, 1–4 and have immense potential for resolving various health and environment related issues, including controlled and triggered drug delivery, tissue engineering, antibacterial coatings, eco-friendly remediation of oil spills etc. 5–15 However, mostly synthetic polymers, inorganic components and fluorinated compounds are used to develop these artificial superhydrophobic materials, 1–18 which are known to have adverse effects on both the environment and health. 19 The combination of appropriate hierarchical (micro/nano) structures and essential low surface energy coatings is the essential criterion for developing artificial superhydrophobic surfaces. 20,21 Most of the hierarchical structures are inherently fragile to various mechanical forces including physical abrasion, stretching, compression etc. , which results in the loss of superhydrophobicity. 22–26 Furthermore, appropriate low surface energy coatings are generally optimized in artificially fabricated superhydrophobic surfaces by associating with weak chemical interactions, including metal–thiol interaction, 27,28 metal–ion interaction 29,30 and silane chemistry, 31,32 which are known to be labile 33 and unsustainable in practically relevant harsh aqueous chemical conditions. Hence, many of the reported superhydrophobic interfaces are inappropriate for performing in practically relevant severe and diverse conditions. As a consequence, the demonstrations of oil/water separation under practically relevant harsh conditions are rare in the literature. 34–36 In the recent past, few early attempts were made to develop artificial superhydrophobicity using naturally existing and biodegradable components including cellulose, chitosan etc. However, the durability of these synthesized materials under challenging conditions is a major concern, where superhydrophobicity is compromised after washing the chitosan derived superhydrophobic interface. 37,38 Thus, the further design of a durable superhydrophobic substrate from naturally abundant ingredients is essential for sustainable and safe application of this nature-inspired wettability in practically relevant severe scenarios. Here, in this current design, a naturally existing biodegradable biomacromolecule, bovine serum albumin (BSA) protein, 39,40 has been strategically and unprecedentedly exploited for the facile and scalable synthesis of a durable and highly deformable superhydrophobic coating for eco-friendly remediation of various oil-spills in real world scenarios. In the recent past, BSA protein was successfully deposited on various nanomaterials for surfactant-free stabilization of colloidal dispersion, improving therapeutic properties, synergistic therapy of tumors, exfoliation of transition metal dichalcogenides, photothermal therapy of cancer cells etc. 41–47 Furthermore, nanoparticles of BSA have been widely exploited for therapeutic applications. 45–47 This economic and naturally abundant serum protein was successfully exploited for the synthesis of a covalently cross-linked and chemically reactive coating (referred to as the BSA coating) on the fibrous water absorbent cotton for developing a highly tolerant superhydrophobic material. In this current approach, the Michael addition reaction that allowed a catalyst free and rapid chemical reaction, between amine and acrylate groups under ambient conditions, 48–50 is used for covalently integrating the BSA protein nanoparticles, and the residual acrylate groups in the BSA-coating provided a facile basis for post-covalent modification with primary amine containing small molecules. This simple process eventually yielded durable superhydrophobicity, and the synthesized material is efficient in sustaining repetitive physical deformations, prolonged (30 days) exposure to UV radiation and various other practically relevant chemically severe challenges without compromising the embedded extreme water repellency.",
"discussion": "Results and discussion Synthesis of chemically reactive BSA protein nanoparticles In the past, the nanoparticles of serum proteins were mostly synthesized following the standard desolvation technique, where the protein nanoparticles are desolvated by the addition of ethanol followed by covalent cross-linking with glutaraldehyde molecules. 46,47 In our current approach, the BSA protein nanoparticles (10 mg mL −1 ) that were synthesized by adding ethanol were strategically exposed to dipentaerythritol penta-acrylate (5Acl), instead of glutaraldehyde as shown in Fig. 1A , to develop the chemically reactive and covalently cross-linked protein nanoparticles. The amine groups in the BSA protein readily react with acrylate groups through a Michael addition reaction under ambient conditions, 48–50 without the aid of any catalyst as shown in Fig. 1B , and this strategic cross-linking of BSA nanoparticles using multifunctional small molecules (5Acl) provided residual chemical reactivity to the granular BSA nanoparticles (average diameter: 487 nm ± 16.85). The appearance of the IR peaks due to (a) the asymmetric C–H stretching (at 1410 cm −1 ) of β carbon of the vinyl group and (b) stretching of the carbonyl group (at 1736 cm −1 ) together revealed the presence of unreacted acrylate groups in the BSA nanoparticles as shown in Fig. 1D . 49 Hence, these residual acrylate groups provided further scope for post-covalent modification with primary amine containing small molecule i.e. ODA. The treatment of these reactive BSA nanoparticles with octadecylamine (ODA) resulted in the change of zeta potential from −10.9 mV to −5.03 mV at pH 7. The long hydrocarbon tail of attached ODA molecules is likely to screen the surface charge of the BSA protein nanoparticles. Furthermore, the significant depletion in the IR peak intensity for the asymmetric C–H stretching (at 1410 cm −1 ) of β carbon of the vinyl group with respect to the carbonyl stretching strongly suggested the covalent reaction between the chemically reactive protein nanoparticles and ODA through a Michael addition reaction under ambient conditions. The stretching vibration of carbonyl groups provided internal reference for this Michael addition reaction as the carbonyl moiety remained unaltered during the course of this mutual reaction between acrylate and amine groups; however, the vinyl moiety of the acrylate groups is compromised. This FTIR characterization and zeta potential analysis unambiguously revealed the unprecedented synthesis of primary amine reactive BSA nanoparticles. Fig. 1 (A) Schematic illustrating the formation of the chemically ‘reactive’ bovine serum albumin nanocomplex by desolvating BSA protein from aqueous media using ethanol followed by covalent cross-linking through a Michael addition reaction (B). (C) DLS data showing the size of the ‘reactive’ BSA nanocomplex. Inset is the FESEM image of the covalently cross-linked BSA nanocomplex. (D) FTIR spectra of a BSA nanoparticle (violet) after covalent cross-linking (black) with 5Acl molecules, and after post-modification with octadecylamine (red). The peaks at 1736 cm −1 and 1410 cm −1 correspond to the carbonyl stretching and symmetric deformation of the C–H bond for the β carbon of the vinyl group, respectively. Development of protein-derived superhydrophobicity This simple synthetic approach of catalyst-free covalently cross-linked and chemically reactive protein nanoparticles was extended for developing nature-inspired durable superhydrophobicity. In the presence of naturally abundant fibrous medical cotton, the same desolvation process was repeated for directly depositing the BSA nanoparticles on the selected fibrous substrate. Upon addition of ethanol, (a) BSA nanoparticles were immediately deposited on the fibrous substrate, and (b) upon treatment with 5Acl molecules, a covalently cross-linked uniform coating of BSA nanocomplexes on the selected fibrous substrate was formed as characterized by FESEM imaging ( Fig. 2A and B ). The granular domains randomly aggregated and provided the essential hierarchical topography for achieving extreme water repellency. The uncoated. i.e. , pristine cotton otherwise has a smooth and uniform surface as shown in Fig. S1G. † On the other hand, the 5Acl treated BSA coating that consists of residual acrylate groups provided a facile basis for covalent modulation of essential chemistry in the deposited BSA coating on the fibrous substrate through appropriate primary amine containing small molecules. Octadecylamine (ODA) molecules that consist of a long hydrocarbon tail (–C 18 H 39 N) and primary amine group were strategically and covalently integrated with this hierarchically featured and chemically reactive BSA coating through a 1,4-conjugate addition reaction for adopting the appropriate chemistry to confer extreme water repellency. An optimum concentration of BSA, i.e. , 10 mg mL −1 is required to ensure uniform formation and deposition of the nanoparticles. Concentrations less than the optimized value, i.e. , 3 mg mL −1 , 5 mg mL −1 and 7 mg mL −1 failed to provide a superhydrophobic interface as the percentage of nanoparticle deposition is exceedingly reduced. Only 1.56 wt%, 3.42 wt%, and 4.31 wt% of BSA nanoparticles were deposited for 3 mg mL −1 , 5 mg mL −1 and 7 mg mL −1 concentrations of BSA solution respectively, whereas 7% deposition was calculated for 10 mg mL −1 concentration of BSA. After post-chemical modification with octadecylamine (ODA), the BSA derived coating that was prepared with a lower concentration (3 mg mL −1 ) of BSA is hydrophilic (water contact angle (WCA) of 80°) and higher concentrations of BSA provided hydrophobic coatings with a WCA of 115° (5 mg mL −1 ) and 140° (7 mg mL −1 ) as shown in Fig. S1A–F. † The FESEM images of the BSA derived coating that was prepared using 7 mg mL −1 concentration of BSA revealed the lack of an appropriate topography that is essential to display superhydrophobicity as shown in Fig. S1H. † Next, the presence of residual acrylate groups and the post-covalent modification of the chemically reactive BSA coating with ODA molecules were characterized with standard and widely recognized FTIR analysis as shown in Fig. 2C . 49 The appearance of IR peaks at 1410 cm −1 and 1710 cm −1 and the reduction in the intensity of the IR peak at 1410 cm −1 with respect to the IR peak at 1710 cm −1 upon treatment with ODA unambiguously suggested the existence of residual acrylate functionalities in the BSA coating and the mutual covalent reaction between the primary amine groups of ODA and the residual acrylate groups of the BSA coating respectively. The water wettability was examined on this synthesized material wherein a dyed (aids visual inspection) aqueous droplet (15 μl) was extremely repelled by this ODA treated BSA coating with a static contact angle of ∼157° as shown in Fig. 2D and E . The shiny interface ( Fig. 2F ) of the synthesized material underwater suggested the presence of an external phase, i.e. , metastable trapped air, which contributed to its extreme heterogeneous water wettability. A stream of aqueous phase immediately bounced away upon touching the BSA protein derived superhydrophobic cotton as shown in Fig. 2G and Movie 1. † This simple demonstration revealed the existence of nonadhesive superhydrophobicity. Fig. 2 (A and B) FESEM images of the BSA nanocomplex coated cotton fibres at low (A) and high (B) magnifications. (C) FTIR spectra depicting the BSA nanocomplex coated cotton before (red) and after (black) treatment with octadecylamine. (D and E) Digital image (D) and static water contact angle (E) of a beaded water droplet on the BSA protein derived superhydrophobic cotton. (F) Digital image showing a shiny interface when the superhydrophobic cotton was submerged under water, and the shiny interface revealed the existence of metastable trapped air. (G) Digital image illustrating the bouncing of a stream of water from the superhydrophobic cotton. Physical and chemical durability of the synthesized superhydrophobic interface Various practically relevant standard physical and chemical challenges were adopted for investigating the durability of the embedded extreme water repellency in the BSA protein derived superhydrophobic material in severe scenarios. First, different physical manipulations, i.e. , bending, creasing, twisting and winding, were performed on this nature-inspired protein derived fibrous substrate and it was found that the extremely water repellent property remained intact with a static contact angle of ∼156° and a contact angle hysteresis of ∼7° as shown in Fig. 3A–F and S2. † Furthermore, severe physical abrasion tests, i.e. , adhesive tape test, sand paper abrasion and finger wiping tests, were also carried out on the synthesized superhydrophobic material; however, the anti-wetting property remained intact with a static water contact angle above 155° and a contact angle hysteresis of ∼8° as shown in Fig. 3G–L and S3. † The FESEM images acquired after performing the adhesive tape test and sand paper abrasion revealed the presence of essential hierarchical topography which attributed to the intact water repellency as shown in Fig. S4A and B. † These demonstrations unambiguously suggested the existence of highly tolerant superhydrophobicity in the BSA protein derived coated medical cotton. Next, this superhydrophobic cotton was manually separated into four individual pieces arbitrarily and the water wettability was examined on these freshly exposed interfaces. The interiors of the synthesized material were also capable of displaying extreme water repellency with a water contact angle above 155° and a contact angle hysteresis of ∼8° as shown in Fig. S5A. † These simple demonstrations revalidated the uniform deposition of the BSA derived coating with essential topography and chemistry, which allowed it to exhibit nature-inspired superhydrophobicity. Next, the as-synthesized superhydrophobic cotton was physically deformed applying different compressive strains, and the impact of the severe physical deformation on the nature-inspired water wettability was examined in detail as shown in Fig. 4A . The applied compressive strain on the synthesized material was increased from 0% to 80% without compromising the embedded nonadhesive superhydrophobicity, and the red colored aqueous droplet beaded on the highly compressed superhydrophobic cotton with a water contact angle above 155° and a contact angle hysteresis below 10° as shown in Fig. 4A . On releasing the applied load, the material restored its initial shape and size, and its water repellency remained unperturbed with a static contact angle of ∼153° as shown in Fig. S3A–C. † This BSA protein derived superhydrophobic and spongy cotton was successively squeezed with 80% compressive strain 1000 times; however, the material remained capable of displaying uninterrupted superhydrophobicity with an advancing water contact angle above 155° and a contact angle hysteresis below 10° as shown in Fig. 4B . This highly compressible superhydrophobic material that is capable of sustaining repetitive physical deformation is of potential interest for absorption based eco-friendly remediation of oil spills under severe practical conditions. This aspect will be discussed later in more detail. Another important durability test was performed on this synthesized material, where the superhydrophobic cotton was kept under UV radiation (at λ max = 254 nm and 365 nm) for 30 days, and the water wettability was examined at regular time intervals for examining the stability of the embedded superhydrophobicity. The prolonged UV light irradiated superhydrophobic cotton continued to display extreme water repellency and the water droplets (inset of Fig. 4C ) beaded with an advancing water contact angle above 158° and a contact angle hysteresis below 10° as shown in Fig. 4C . The superhydrophobic cotton was also exposed to various practically relevant chemically challenging conditions, i.e. , extremes of pH, artificial sea water, river water and even surfactant contaminated aqueous phases for 30 days; however, the embedded superhydrophobicity remained intact with an advancing water contact angle of above 155° and a contact angle hysteresis below 10° as elucidated in Fig. 4D . This unperturbed biomimicked wettability indicating the coexistence of appropriate topography and essential chemistry in the treated material. Fig. 3 Digital images (A, B, D and E) and contact angle images (C and F) depicting various physical manipulations, i.e. , twisting (A–C) and winding (D–F) of the BSA derived superhydrophobic cotton. Digital images (H and K) and contact angle images (I and L) showing the impact of different abrasive physical durability tests—adhesive tape test (G) and sand paper abrasion test (J)—on the protein derived superhydrophobicity. Fig. 4 (A) The plot showing the advancing contact angle (black) and the contact angle hysteresis (red) of a beaded water droplet on the superhydrophobic cotton that was manually and gradually deformed up to 80% compressive strain. The inset digital images show the superhydrophobic cotton with various compressive strains—including 0% (left), 40% (middle) and 80% (right). (B) The plot illustrates the advancing contact angle (black) and contact angle hysteresis (red) of the protein coated superhydrophobic cotton after repetitively squeezing the material with 80% deformation for 1000 cycles. (C) Plot showing the advancing contact angle (black) and contact angle hysteresis (red) of the superhydrophobic cotton exposed to UV radiation for 30 days. Inset images of beaded water droplets on the substrate before and after prolonged UV irradiation. (D) Plot showing the impact of various severe chemical exposures including acidic (pH 1, grey), alkaline (pH 12, orange), anionic (SDS, black) and cationic (DTAB, red) surfactant contaminated water, artificial sea water (blue) and river water (green) on the extremely water repellent property of the superhydrophobic cotton for 30 days. Absorption based selective removal of various oil spills under severe conditions These extremely water repellent interfaces were noticed to be inherently superoleophilic with an oil contact angle of 0° as shown in Fig. S5 and Table S1. † This selective and extreme affinity (oil/oily phase) and repellency (aqueous phase) of the synthesized material towards two distinct liquids phases (oil and water) provided a facile basis to extend this durable superhydrophobic material for the eco-friendly remediation of oil/oily pollution from aqueous phases. As a proof of concept demonstration, a droplet of floating oil, i.e. , motor oil at an air/water interface was selectively removed using the BSA derived superhydrophobic cotton as shown in Fig. 5A–C , S6A–D and Movie 2, † where the oil phase was selectively and rapidly absorbed by the as-synthesized material, and the absorbed oil was collected back by squeezing the superhydrophobic cotton. Furthermore, this material is capable of removing selectively sediment oil under water as shown in Fig. 5D–F , S7A–E and Movie 3. † The oil/water separation for silicone oil and crude oil has been illustrated in Fig. S6E–L. † The oil absorption capacity was found to be more than 2000 wt%, irrespective of the density (heavy and light oils; from 0.83 g cm −3 to 1.49 g cm −3 ) and viscosity (from 0.428 cP to 244 cP; see Fig. S8B † ) of used oils ( i.e. motor oil, silicone oil, ethyl acetate, crude oil, dichloroethane and chloroform) as shown in Fig. 5G . Next, the oil absorption capacity of the protein derived superhydrophobic cotton was investigated in the presence of various severe chemical conditions including extremes of pH, sea and river water and surfactant contaminated water, where oil/aqueous mixtures were prepared using both heavy (chloroform) and light (ethylacetate) oils. The oil absorption capacity remained above 2000 wt%, in the presence of the various severe practically relevant challenges as shown in Fig. S8A. † The repetitive use of the BSA derived superhydrophobic cotton for the removal of oil spillages was also examined wherein the cotton was used for the separation of oil from water followed by squeezing out the oil. The same material was reused for repetitive separation of oil/water mixtures and interestingly, the oil separation efficiency remained above 95% even after 50 cycles of subsequent reuse of the protein derived material in selective collection of either light or heavy oils as shown in Fig. 5H . After successive (50 times) uses of the protein based superhydrophobic coating, it continued to display extreme water repellency with an advancing contact angle above 150° and a contact angle hysteresis below 10° as shown in Fig. S9. † The oil separation efficiency was also calculated under various harsh chemical conditions, i.e. , extremes of pH, seawater, river water and surfactant contaminated water using both heavy (chloroform) and light (ethylacetate) oils, and the oil separation efficiency remained above 95% irrespective of the imposed chemical challenges as shown in Fig. 5I . Hence, this naturally derived superhydrophobic cotton can be used under harsh practical conditions without compromising the embedded water repellent property. The oil absorption capacity was noticed to be significantly higher compared to recently developed superhydrophobic materials. 50,51 Furthermore, this BSA protein derived functional material can be strategically used for post-loading, followed by sustained release of selected bioactive drug molecules 8 for various biomedical applications. Fig. 5 Digital images (A–C) illustrating the absorption based separation of light oil (motor oil) and heavy oil (DCE, D–F) from aqueous phases using superhydrophobic cotton, and the absorbed oils were collected back by simply squeezing the material. (G) The plot showing the capacity of the superhydrophobic cotton to absorb oils having a wide range of densities and viscosities. (H) The plot illustrating the oil separation efficiency of the superhydrophobic cotton for the removal of both model heavy (chloroform, black) and light (ethylacetate, red) oils from oil/water mixture 50 times. (I) The plot showing the oil separation efficiency of the BSA derived superhydrophobic cotton in the separation of both light (red) and heavy (black) oils under various practically relevant and severe conditions—including pH 1, pH 12, surfactant contaminated water, river water and sea water."
} | 6,428 |
24603697 | PMC3945109 | pmc | 2,814 | {
"abstract": "Chemosensory systems (CSS) are complex regulatory pathways capable of perceiving external signals and translating them into different cellular behaviors such as motility and development. In the δ-proteobacterium Myxococcus xanthus , chemosensing allows groups of cells to orient themselves and aggregate into specialized multicellular biofilms termed fruiting bodies. M. xanthus contains eight predicted CSS and 21 chemoreceptors. In this work, we systematically deleted genes encoding components of each CSS and chemoreceptors and determined their effects on M. xanthus social behaviors. Then, to understand how the 21 chemoreceptors are distributed among the eight CSS, we examined their phylogenetic distribution, genomic organization and subcellular localization. We found that, in vivo , receptors belonging to the same phylogenetic group colocalize and interact with CSS components of the respective phylogenetic group. Finally, we identified a large chemosensory module formed by three interconnected CSS and multiple chemoreceptors and showed that complex behaviors such as cell group motility and biofilm formation require regulatory apparatus composed of multiple interconnected Che-like systems.",
"conclusion": "Conclusions In this study we sought to understand the partitioning of M. xanthus chemoreceptors among eight CSS to constitute sensory modules. We hypothesized that Che modules might attract multiple receptors and Che proteins as observed in other bacterial species and that the analysis of their cellular organization would help us to understand the role of these proteins in the M. xanthus life cycle. For this purpose, we first compiled a full list of the putative M. xanthus Che proteins and chemoreceptors. We were not surprised to find a total of 67 proteins as, in most cases, the number of one- and two-component systems present in a bacterial genome directly relates to the complexity of the life cycle [56] . The same might be true for CSS. Our systematic deletion of the 21 M. xanthus chemoreceptors and CheA encoding genes revealed that two thirds of them are involved in the temporal regulation of fruiting body formation, a multi-step differentiation process requiring the perception of numerous signals for the activation of key regulation check-points [20] , [57] . Based on an integrated approach, we found that MCPs and CSS show comparable phylogenetic distributions in three main groups and that MCPs belonging to the same phylogenetic group colocalize. In particular, MCPs of Group 3 seemed to constitute a large chemosensory module together with three CSS, namely Che4, Che5 and Che6 ( Figure 8 ). The presence of such a complex array of chemosensory proteins suggests that social behaviors such as cell group motility and biofilm formation might require interwoven regulatory systems composed by multiple Che-like systems and that the final cellular responses are generated following both the integration of signals transduced by different MCPs at the CheA level and the interaction among different Che systems. Also, cross-regulation between different Che systems can add an additional layer of complexity, as suggested by previous work showing inter-dependence between the Frz and the Dif pathway [58] . Once the composition of each module has been dissected, it will be possible to identify their signals and outputs to clarify their precise function in the M. xanthus life cycle. 10.1371/journal.pgen.1004164.g008 Figure 8 Schematic organization of M. xanthus Che modules as depicted from phylogenetic, cell biology and protein interaction analyses. For clarity, we omitted CheR and CheB proteins and do not specify the MCP-CheW interactions. MCPs in light green are the ones for which interactions with a CSS have not been demonstrated. The different color backgrounds indicate taxonomic Group 1 (green), Group 2 (blue) and Group 3 (pink). Group 1 was further divided in two subgroups labelled with light and dark green, based on the localization analysis. It has recently been reported that multiple chemosensory systems occur as frequently as single ones, highlighting the importance of investigating model microbes that encode multiple chemosensory systems [1] . By providing a broad perspective on how a complex multicomponent chemosensory apparatus is arranged within cells, this work establishes a basis for a deeper analysis on how signals are perceived, integrated and translated in cell behaviors at the level of each chemosensory module. Analogous approaches could be applied to bacterial systems with similarly complex regulatory networks.",
"introduction": "Introduction Perceiving and responding to external stimuli allows living organisms to adapt to changes in their environment and thus enhance their survival fitness. Perception universally occurs through the aid of receptors coupled to signaling pathways that translate an initial signal into the appropriate cellular behaviors. Perception of stimuli in bacteria is largely mediated by one-component, two-component and chemosensory systems (CSS). CSS are modified two-component systems in which the histidine kinase, CheA, does not directly perceive the chemical signal [1] . Instead, this function is delegated to specialized chemoreceptors, known as Methyl-accepting Chemotaxis Proteins (MCPs) for the presence of a methyl-accepting domain in their C-terminal cytoplasmic region [2] . An adaptor protein, CheW, facilitates the interaction between the MCP and the CheA proteins. MCPs are methylated and demethylated on glutamate residues by a methyltransferase (CheR) and a methylesterase (CheB), respectively [2] . These enzymatic activities allow adaptation of the receptor to persistent stimuli [3] . The best-studied CSS are specialized for chemotaxis. In this case, the output response regulator CheY has the function of directly communicating with the flagellar motor proteins, FliM and FliN, in order to adjust the cell swimming behavior [4] . Interestingly, over the past years, many CSS have been identified that regulate behavioral responses other than taxis [5] . Examples are the Myxococcus xanthus Che3 system that regulates gene expression during development [6] , the Pseudomonas aeruginosa Wsp system that regulates c-di-GMP production and biofilm formation [7] and the Rhodospirillum centenum Che3 system involved in cyst formation. When multiple receptors mediate signal reception and stimulate kinase activity, the various signals must be integrated to generate a single response. For example, in the E. coli Che system that contains a single chemosensory pathway, five receptors of different ligand specificity signal to the same kinase, CheA [8] , [9] . However, in bacteria with multiple chemosensory pathways, the recruitment of chemoreceptors to the different Che systems depends on protein specificity and the physical location of the Che modules [10] , [11] . Structural studies have shown that receptor clusters are formed by interconnected heterotrimers of homodimers, which are associated with two CheWs and a dimer of CheA. Receptor homodimers can in turn form heterotrimers if they share common structural features and belong to the same class [12] , [13] . The spatial segregation of MCPs to distinct cellular compartments also plays a role in the partitioning of MCPs among multiple CSS. For example, in Rhodobacter sphaeroides , membrane-associated and soluble MCPs are partitioned between polar and cytoplasmic clusters [10] , [14] . We have been studying the multiple CSS of the Gram negative δ-proteobacterium Myxococcus xanthus . M. xanthus carries up to eight predicted chemosensory systems with 21 chemoreceptors [15] , [16] . We speculate that the large number of CSS reflects the complexity of the M. xanthus life cycle, in which cells swarm as large groups to prey on other micro-organisms or build multicellular fruiting bodies [17] – [20] . Movement on surfaces does not employ flagella but instead requires two distinct motility machineries: polar retractile Type IV pili required for social (S) motility [21] , [22] and distributed Agl-Glt complexes that form periodic foci and generate thrust for adventurous (A) motility [23] – [27] . Evidence suggests that M. xanthus motility behaviors are controlled by CSS. The Frz pathway, the first characterized Che-like system from M. xanthus , controls both motility systems by triggering periodic cellular reversals. This allows the bacteria to periodically reorient themselves, and may be similar to periodic switches in flagellar rotation, which allow the enteric bacteria to move along a chemotactic gradient by following a biased random swim. Motility is also regulated by Dif (Che2), a second sensory system that controls the production of surface exopolysaccharides in response to pilus activity [28] . Che4, a third CSS, also appears to be involved in the regulation of motility, although the specific mechanism remains unclear [29] . However, CSS are not exclusively dedicated to motility regulation in M. xanthus . In fact, the Che3 system regulates gene expression during fruiting body development [6] , [29] – [32] . While future M. xanthus research on the exact contribution of each Che-like system to its life cycle will yield considerable biological insights, this task is complicated by the occurrence of 21 MCPs encoded in its chromosome, 13 of which are orphans. Furthermore, the activity of each CSS might be modulated by multiple MCPs as shown in other bacterial species. Cross-regulation and redundancies between additional pathways may also occur and thus further complicate the picture. In this work, we set out to characterize each M. xanthus CSS and MCP and combine phylogenetic and cell biology analyses to examine their organization within cells to constitute functional modules. With this approach, we were able to show that MCPs belonging to the same phylogenetic group colocalize in cells and interact in vivo with components of CSS of their respective phylogenetic group. Protein-protein interaction analyses also suggest that colocalizing CSS belonging to same phylogenetic group constitute a unique large sensory module. Such organization is likely required to regulate complex cell behaviors such as biofilm and fruiting body formation. This analysis provides a broad perspective as to the function and organization of complex multicomponent chemosensory systems within bacterial cells and could be applicable to bacterial systems with similarly complex regulatory networks.",
"discussion": "Results/Discussion Identification of M. xanthus chemosensory modules Four Che pathways have been characterized in M. xanthus : Frz, Dif, Che3 and Che4 [6] , [29] , [30] , [33] . We used the conserved protein domain sequences from these pathways ( Tables 1 , 2 and 3 ) as queries to search for all M. xanthus Che homologues. Most che genes are organized in eight che operons, as previously described ( Figure 1 ) [31] . Their predicted organization is depicted in Figure S1 . The M. xanthus genome does not contain homologs of CheD and CheX or of CheZ, which are usually found in genomes of β- and γ-proteobacteria [1] , [34] . None of the eight che clusters is located near known motility genes or other genes encoding cellular functions known to be controlled by CSS [6] , [35] , [13] , [36] . 10.1371/journal.pgen.1004164.g001 Figure 1 Genetic clusters carrying che genes in M. xanthus . Genetic organization of the genes composing the eight che clusters encoding the putative components of the chemosensory apparatus in Myxococcus xanthus . Predicted genes are indicated with their locus_tag, and their annotations and assigned names. The color code indicates homologous genes. 10.1371/journal.pgen.1004164.t001 Table 1 List of M. xanthus MCPs. b \n Protein name Locus tag (MXAN_) Refseq Length (aa) Che operon Functional domains (Pfam) Transmembrane domain (TMHMM) HAMP a domain (position) MCP b signal domain (position) Others domains (position) FrzCD 4141 ABF87130 417 \n Frz \n - (195 to 417) - - Mcp3B 5148 ABF90272 605 \n che3 \n (246 to 315) (376 to 596) - 6 Mcp3A 5149 ABF91174 595 \n che3 \n (240 to 309) (377 to 590) - 6 Mcp4 2683 ABF86464 538 \n che4 \n (189 to 258) (322 to 538) - 2 Mcp5 6027 ABF86588 531 \n che5 \n (180 to 250) (313 to 531) - 2 Mcp6 6950 ABF87794 547 \n che6 \n (184 to 253) (335 to 546) CHASE3 c (39 to 172) 2 Mcp7 6962 ABF86943 821 \n che7 \n - (317 to 534) - - DifA 6696 ABF91646 413 \n dif \n (33 to 103) (186 to 411) - 2 McpA 174 ABF91199 530 - (180 to 250) (318 to 530) - 2 McpB 227 ABF88952 553 - (202 to 271) (340 to 552) CACHE d (45 to 139) 1 McpC 878 ABF90878 631 - (280 to 351) (416 to 631) - 2 McpD 1248 ABF89240 599 - - (386 to 598) - 5 McpE 1414 ABF92510 609 - (259 to 329) (392 to 609) - 2 McpF 2244 ABF92314 729 - - (502 to 720) 7TM diverse intracellular signalling 7 McpG 2735 ABF92828 523 - (170 to 240) (299 to 521) - 2 or 1 McpH 3668 ABF87603 510 - (160 to 230) (293 to 510) - 2 McpI 3754 ABF87204 519 - - (184 to 497) - 3 McpJ 5907 ABF91268 723 - (203 to 275) (431 to 640) (636 to 723) - 2 McpK 6456 ABF92748 1048 - - (246 to 463) Bacterial extracellular solute-binding proteins domain 3 McpL 6938 ABF90672 591 - - (380 to 591) - 5 McpM 3453 ABF92699 737 - - (503 to 722) 7TM diverse intracellular signalling 7 a = Histidine kinase, Adenylate cyclase, Mcp, Phosphatase b = Methyl accepting Chemotaxis Protein c = Cyclase/Histidine kinases Associated Sensory Extracellular d = CAlcium channels and CHEmotaxis receptors 10.1371/journal.pgen.1004164.t002 Table 2 List of M. xanthus CheA, CheW and CheY. Protein name Locus tag (MXAN_) Refseq Length (aa) Che operon Functional domains (Pfam) Histidine-containing phosphotransfer (HPt) domain Response regulator binding domain Signal transducing histidine kinase, homodimeric domain Histidine kinase domain CheW-like domain Response regulator receiver domain \n CheA \n FrzE 4140 ABF89991 777 \n frz \n (9 to 108) - - (365 to 508) (513 to 644) (661 to 774) DifE 6692 ABF92648 857 \n dif \n (7 to 107) (366 to 450) (474 to 537) (582 to 724) (729 to 857) - CheA3 5147 ABF90293 798 \n che3 \n (6 to 99) - - (376 to 522) (527 to 658) (681 to 794) CheA4 2686 ABF90689 848 \n che4 \n (8 to 106) - - (433 to 575) (580 to 714) (729 to 842) CheA5 6029 ABF88344 714 \n che5 \n (9 to 114) - - (303 to 445) (450 to 581) (591 to 704) CheA6 6951 ABF88527 726 \n che6 \n (9 to 112) - - (316 to 455) (460 to 590) (610 to 723) CheA7 6964 ABF92123 683 \n che7 \n (5 to 110) - (279 to 344) (385 to 525) (530 to 663) - CheA8 4758 ABF85910 862 \n che8 \n (6 to 110) - (463 to 534) (579 to 721) (726 to 855) - \n CheW \n FrzA 4143 ABF93093 160 \n frz \n \n - \n \n - \n \n - \n \n - \n (16 to 154) \n - \n FrzB 4142 ABF91972 112 \n frz \n \n - \n \n - \n \n - \n \n - \n (26 to 109) \n - \n DifC 6694 ABF92146 140 \n dif \n \n - \n \n - \n \n - \n \n - \n (2 to 133) \n - \n CheW3 5151 ABF89050 145 \n che3 \n \n - \n \n - \n \n - \n \n - \n (12 to 142) \n - \n CheW4a 2685 ABF89623 194 \n che4 \n \n - \n \n - \n \n - \n \n - \n (59 to 186) \n - \n CheW4b 2681 ABF90781 188 \n che4 \n \n - \n \n - \n \n - \n \n - \n (48 to 186) \n - \n CheW5a 6030 ABF88929 182 \n che5 \n \n - \n \n - \n \n - \n \n - \n (42 to 174) \n - \n CheW5b 6032 ABF86894 288 \n che5 \n \n - \n \n - \n \n - \n \n - \n (131 to 270) (6 to 115) CheW6a 6947 ABF88743 159 \n che6 \n \n - \n \n - \n \n - \n \n - \n (7 to 142) \n - \n CheW6b 6949 ABF87567 200 \n che6 \n \n - \n \n - \n \n - \n \n - \n (57 to 189) \n - \n CheW7 6963 ABF85986 148 \n che7 \n \n - \n \n - \n \n - \n \n - \n (4 to 141) \n - \n CheW8a 4756 ABF88564 174 \n che8 \n \n - \n \n - \n \n - \n \n - \n (28 to 165) \n - \n CheW8b 4757 ABF90229 188 \n che8 \n \n - \n \n - \n \n - \n \n - \n (43 to 187) \n - \n CheWa 4462 ABF92129 271 - \n - \n \n - \n \n - \n \n - \n (10 to 130) (151 to 265) \n - \n \n CheY \n FrzZ 4144 ABF89606 290 \n frz \n - - - - - (4 to 112)(170 to 282) DifD 6693 ABF87390 122 \n dif \n - - - - - (4 to 116) CheY4 2684 ABF88007 127 \n che4 \n - - - - - (5 to 117) CheY5 6033 ABF92419 128 \n che5 \n - - - - - (8 to 122) CheY7 6965 ABF88516 125 \n che7 \n - - - - - (4 to 117) CheY8a 4751 ABF88585 126 \n che8 \n - - - - - (7 to 120) CheY8b 4759 ABF90183 124 \n che8 \n - - - - - (5 to 118) 10.1371/journal.pgen.1004164.t003 Table 3 List of M. xanthus CheB and CheR. Protein name Locus tag (MXAN_) Refseq Length (aa) Che operon Functional domains (Pfam) Response regulator receiver domain CheB methylesterase CheR methyltransferase, SAM binding domain Tetratricopeptide repeat \n CheB \n FrzG 4139 ABF92096 334 \n frz \n - (150 to 330) - - CheB3 5145 ABF91845 352 \n che3 \n (7 to 118) (162 to 342) - - CheB5 6028 ABF91242 355 \n che5 \n (9 to 122) (160 to 338) - - CheB6 6952 ABF89213 353 \n che6 \n (8 to 115) (166 to 346) - CheB7 6959 ABF87336 348 \n che7 \n (9 to 121) (163 to 346) - - CheB8 4752 ABF87312 345 \n che8 \n (9 to 120) (157 to 340) - - CheBa 714 ABF90117 185 - - (6 to 164) - - \n CheR \n FrzF 4138 ABF86899 593 \n frz \n - - (1 to 61) (74 to 270) ** \n (427 to 491) (496 to 556) CheR3 5144 ABF87122 440 \n che3 \n - - (69 to 262) (358 to 406) CheR4 2682 ABF89636 485 \n che4 \n - - (74 to 256) (357 to 419) CheR5 6031 ABF88773 413 \n che5 \n - - (72 to 250) - CheR6 6948 ABF86317 569 \n che6 \n - - (73 to 255) (467 to 535) CheR7 6960 ABF89444 273 \n che7 \n - - (77 to 267) - CheR8 4753 ABF90193 290 \n che8 \n (82 to 278) - CheRa 7103 ABF86029 346 - - - (144 to 334) - CheRb 713 ABF91568 278 - - - (13 to 68) (81 to 270) ** \n - CheRc 2243 ABF89894 242 - - - (47 to 141) * \n - CheRd 2245 ABF91524 401 - - - (72 to 182) * \n - * incomplete CheR domain. ** CheR short Nter domain + CheR S-adenosyl-L-methionine binding domain. In addition to che operon encoded proteins, we identified several orphan che genes and 13 mcp genes dispersed throughout the chromosome ( Tables 1 – 3 ). The other M. xanthus Che proteins with their respective locus tags, protein lengths and specific domains are listed in Tables 2 and 3 . We did not conduct a thorough analysis of CheY homologs as the M. xanthus genome encodes 260 predicted response regulator domains (data not shown). In addition, it is impossible to distinguish if these proteins retain CheY function based on the sequence alone [1] . We reasonably assume that the response regulator domains encoded within the eight che operons constitute the minimum set of M. xanthus CheYs ( Table 2 ). Deletions of cheA and mcp genes affect motility and fruiting body formation In order to determine the function of the different MCPs and CSS during vegetative and developmental behaviors, we constructed a set of in-frame deletion strains in which all of the mcp and cheA genes were systematically deleted, with the exception of those for which an in-frame deletion in the wild-type strain DZ2 already existed ( frzCD , frzE , mcp3A , mcp3B and mcp4 ) [6] , [29] , [37] . Deletions in cheA3 , cheA7 , cheA4 , mcp6 , mcpA , mcpH , mcpL and mcpM caused S-motility defects, which significantly reduced or enhanced colony spreading compared to wild-type ( p <0.05) ( Figure 2A ). This was also true for ΔdifE , ΔfrzE , ΔdifA and ΔfrzCD , for which a S-motility defect has already been described [33] , [38] . 10.1371/journal.pgen.1004164.g002 Figure 2 Motility and fruiting body formation defects of Δmcp and ΔcheA mutants. (A) Motility was measured after 48 h. Colony spreading of each mutant was normalized with that of a ΔpilA strain [72] completely incapable of S motility, to exclude cell growth effects. Error bars indicate standard deviations. One star corresponds to p <0.05; two stars correspond to p <0.005. (B) ΔcheA fruiting body formation images at 48 h and 72 h are shown. We identified classes of mutants developing earlier or later than wild type. Pictures of Δmcp fruiting bodies are shown in Figure S2 . The blue color indicates ΔcheA mutants, green Δmcp . At least 13 Δmcp and all ΔcheA strains were defective in fruiting body formation, showing altered developmental timing or displayed a complete absence of development ( Figures 2B and S2 ). M. xanthus fruiting body formation requires a functional motility apparatus. Therefore, in ΔcheA4 , ΔcheA7 , ΔmcpH , ΔmcpM and Δmcp6 strains, the developmental defects might result from the motility defects also shown by these mutants ( Figure 2A and B ). However, in most cases the two phenotypes are unrelated, suggesting that most Che proteins either regulate motility exclusively during development or are involved in functions other than motility in M. xanthus . In order to check whether the Δmcp and ΔcheA strains were capable of A motility, we systematically deleted the pilA gene in each Δmcp and ΔcheA strain to exclude an effect of S motility, as this motility system is active on the substrate commonly used to test A motility (1.5% agar plates) [37] . All double mutants displayed individual cells at the colony edges suggesting the presence of a functional A-motility system ( Figure S3 ). Notably, we were unsuccessful at deleting mcpC , suggesting that this gene might be essential in M. xanthus . In our assays, mutants lacking McpG, McpI, McpJ, McpL and Mcp4 did not display any defects ( Figures 2 and S2 ). Among these MCPs, McpI and McpL are not expressed in cells (see below), similar to that observed in the R. sphaeroides cheOp1 operon [39] . In the case of Mcp4, McpG and McpJ, which were clearly expressed in cells (see below), the corresponding mutants might display insignificant defects or have functions masked by the presence of another MCP. Interestingly, most cheA deletions caused more severe defects than deletions of mcp genes from the same operons. These results support the hypothesis that each CSS is activated by multiple receptors, as CheAs are core components of CSS. Thus, phenotypic analyses can be ambiguous for the purpose of clustering M. xanthus MCPs into functional modules. Indeed, MCPs showing opposing functions may still signal to the same Che pathway and contribute differently to the final response. For example, it has been recently shown that the Tar and Tsr E. coli chemoreceptors, both signaling to the same CheA, show opposite pH-taxis responses [40] . \n M. xanthus MCP and Che proteins show similar phylogenetic distributions To obtain additional insights on MCP-CSS associations in M. xanthus , we compared the phylogeny of the MCPs to the phylogeny of the CSS, reasoning that MCPs and CSS that share the same phylogenetic distribution might be functionally associated. We started by determining the phylogenetic associations among the eight M. xanthus Che clusters. First, we obtained the individual phylogenies of the MCP, CheA, CheW, CheR and CheB proteins from those clusters. The five individual phylogenies showed similar topologies ( Figure S4 ). However, as these phylogenies were based on a limited number of unambiguously aligned positions and the nodes of the inferred trees were often weekly supported (PP<0.5), we concatenated the MCP, CheA, CheR and CheB sequences from each locus into a super-sequence and used the resulting supermatrix to obtain phylogenetic trees with a higher resolution ( Figure 3A ). Whenever a Che cluster contained two homologues of uncertain orthologous relationship, we excluded them from the concatenation. This was the case for CheY-like and CheW-like proteins ( Table 2 ). In the case of the Che3 system, Mcp3A and Mcp3B derive from a recent duplication in the Cystobacterineae (unpublished) and thus Mcp3B was included in the supermatrix. The tree obtained from the concatenated data sets was significantly more resolved than the individual trees ( Figure 3A and Figure S4 ). Figure 3A shows that the Che clusters may be categorized in three main groups: Group 1 containing Dif, Che7 and Che8; Group 2, FrzCD and Che3; Group 3, Che4, Che5 and Che6. 10.1371/journal.pgen.1004164.g003 Figure 3 \n M. xanthus MCPs and CSS are organized in three taxonomic groups. (A) Concatamers of M. xanthus Che protein sequences were generated as described in Methods. Based on PP values, the eight concatamers can be divided into Group 1 (green background), Group 2 (blue background) and Group 3 (pink background). (B) The tree generated for the 21 M. xanthus MCP homologs shows a similar partition in three groups. The MCPs in black belong to che operons, while the MCPs in color are the orphans. (C) A tree generated with the MCP conserved protein sequences involved in the MCP-CheW interaction (Vu et al., 2012) gives rise to the same distribution as in (B). The alignment of the protein sequences involved in the MCP-CheW interaction from T. maritime \n [41] and M. xanthus MCPs is shown. Colors indicate residues with the same properties. Numbers at nodes in (A) and (B) indicate posterior probabilities (PP) computed by MrBayes and bootstrap values (BV) computed by PhyML. Only PP and BV above 0.5 and 50% are shown. The scale bars represent the average number of substitutions per site. Next, in order to assign the 13 orphan MCPs of M. xanthus to a Che system, we performed a phylogenetic analysis of the 21 MCPs. The resulting MCP tree was strongly correlated. Specifically, the 21 MCPs formed three major monophyletic groups (PP = 0.99, Figure 3B ) with the first group containing five MCPs (McpB, McpJ, McpK, DifA and Mcp7). Phylogenetic analyses suggest that Mcp7 and McpJ emerged upon a recent gene duplication event and that McpB is closely related to DifA (data not shown). Group 2 contains FrzCD, McpG, Mcp3A and Mcp3B and Group 3, the largest group, contains all of the remaining MCPs. In Groups 2 and 3, the MCPs are strongly associated and therefore may have emerged by recent gene duplication events in the δ-proteobacteria. The congruence between the MCP and CSS distributions suggests that phylogenetic relationships may be useful in predicting MCP-Che associations. These associations should be reflected in binding specificities such that MCPs that interact with the same downstream Che module should have similar CheW-binding motifs. It has been recently shown that a short peptide sequence is involved in MCP-CheW binding in T. maritima \n [41] , [42] . All M. xanthus MCPs contained a conserved predicted CheW-binding motif ( Figure 3C ). Such motifs were aligned and the alignment was used to construct a phylogenetic tree. Although the nodes were poorly supported due to the short sequences, the resulting tree presented the same topology observed in Figure 3B . This analysis further suggests that MCPs belonging to the same group have similar binding specificities and are associated with the same CheW and, therefore, Che system. The MCP C-terminal methylated domain is constituted by a repetition of several heptamers. MCPs can be classified depending on the number of these heptamers and, therefore, on the length of the C-terminal region [12] . It appears that MCPs of different lengths cannot form trimers of dimers as has been described for the P. aeruginosa McpB and WspA belonging to class 36H and 40H [43] , [44] , or R. sphaeroides McpG and TlpT belonging to classes 34H and 36H [45] , [46] . Sequence analyses based on the Alexander and Zhulin classification show that all M. xanthus MCPs belong to class 40H, with the exception of DifA and McpK which belong to class 44H [15] . This result suggests that DifA and McpK interact with each other and signal to the Dif system forming a separate module. Taken together, the phylogenetic and sequence analyses suggest the following associations: DifA and McpK linked to Dif; Mcp7, McpB and McpJ linked to Che7 and Che8; FrzCD, McpG, Mcp3a/Mcp3B linked to Frz and Che3; all remaining Mcps linked to Che4, Che5 and Che6. Subcellular localization of M. xanthus MCPs The R. sphaeroides chemosensory network is composed of two sensory modules each including multiple receptors [13] , [31] , [47] , [48] . The two modules are physically separated in cells, as one constitutes a transmembrane polar cluster and the other one a cytoplasmic cluster [11] , [46] , [49] . We hypothesized that in M. xanthus , much like in R. sphaeroides , MCPs belonging to the same sensory module should have similar localization patterns [45] , [46] . To test this hypothesis, we constructed strains that expressed the C terminus of each MCP fused to the green fluorescent protein (eGFP). Each gene fusion was placed at the respective endogenous locus and were shown not to interfere with cellular functions, with the exception of FrzCD-GFP and DifA-GFP which only partially complemented the motility defects observed in the respective deletion mutants ( Figure S5 ) [50] . The strains were then examined in vivo by live-fluorescence microscopy. Ten of the 21 MCP-GFP fusions were highly expressed in cells, showing bright fluorescent foci and clear localization. Conversely, the remaining fusions showed only weak and diffused fluorescent signal in vegetative conditions ( Figures 4A , S6 and S7 ). Mcp3B, McpE and McpG that we could not detect by fluorescence microscopy during vegetative growth, were instead expressed during development ( Figure S7 ). MCPs that we could not detect in any condition, but that clearly play a role during development, were probably expressed at low levels, which is also the case with the very low-abundance receptors Trg and Tap of E. coli \n [51] , [52] . 10.1371/journal.pgen.1004164.g004 Figure 4 MCP-GFP fusions localize in multiple dynamic clusters in cells. (A) In the first row, fluorescence (left) and overlay between fluorescence and phase contrast images (right) are shown for each MCP-GFP. In the bottom row, n clusters (numbers indicated above the histograms) were analyzed for each mcp-gfp strain and their relative position in cells in the y -axis is shown (0.0 indicate the center of the cell along the y -axis). Bars indicate the fraction of clusters localizing in the corresponding position in the y -axis. (B) Average number of clusters for each MCP-GFP. ( C ) Box plots indicate the medians of the product of the relative cell length and the total distance covered by the MCP-GFP clusters * = p <0.05; ** = p <0. 5E-04 (refer also to Methods, Table S2 and Figure S4 ). We proceeded with the analysis of the ten fusions that were localized in all conditions: DifA, FrzCD, Mcp7, McpJ, Mcp4, Mcp5, Mcp6, McpA, McpH and McpM. These MCP-GFP strains all showed multiple fluorescent clusters at the cell poles, the cell periphery or the cytosol ( Figure 4A and B ). To analyze these patterns we designed a procedure allowing large-scale automated image acquisition and analysis of the clusters formed by each MCP-GFP fusion (refer to Methods). With this approach, we obtained a localization map of each MCP by assigning the detected clusters to their relative cellular position and identified three main localization patterns. Mcp7 and McpJ formed only one or two clusters at the subpolar cell regions ( Figure 4A and 4B ). FrzCD showed a unique localization pattern with cytoplasmic foci excluded from the poles and occupying the central region of the cell body, as previously described [50] , [53] . The remaining MCPs (DifA, Mcp4, Mcp5, Mcp6, McpA, McpH, McpM) formed foci distributed all along the periphery of cells, as predicted by the presence of transmembrane domains in their sequence ( Table 1 , Figure 4A and 4B ). As all of the MCP foci appeared to be dynamic, we systematically analyzed the dynamics of these foci in single cells ( Movie S1 – S3 , Figure S8 ). In order to exclude any interference from cellular movements, MCP foci were tracked in non-motile cells. Our analyses revealed that Mcp7 was significantly more mobile than all the other MCPs ( p <0.005) ( Figure 4C , Figure S8C and Table S2 ). Also, DifA and Mcp4 clusters were significantly more mobile than Mcp5, Mcp6, McpA and McpH fusions which were more static while McpM which showed little mobility if any ( Figure 4C , S8B, S8C and Table S2 ). Interestingly, while the more static McpM carries the highest number of transmembrane domains, the faster Mcp7 is a cytoplasmic protein ( Table 1 ). However, while FrzCD also lacks transmembrane domains, it shows slower movement rates compared to Mcp7. This might be explained by the anchoring of the FrzCD clusters to some intracellular structures [53] , [54] . Based on localization and phylogeny, we can postulate that (i) McpJ is linked to Mcp7, (ii) FrzCD constitutes a sensory module by itself and (iii) MCPs of Group 3 are linked to the Che4, Che5 or Che6 pathways ( Figures 3 and 4 ). Although the localization and dynamics of DifA suggest that it interacts with the receptors of Group 3, this appears unlikely from the divergence of their respective C-terminal domains based on phylogenetic and sequence analyses (see above) [12] . This cellular localization and dynamics of the chemoreceptors is largely consistent with the functional groups suggested by the phylogenomic analysis described above. To verify that Mcp4, Mcp5, Mcp6, McpH, McpA and McpM, predicted to be associated in the same functional module, are colocalized in cells, we constructed M. xanthus strains expressing two fluorescently labeled MCPs, with either the GFP or the mCherry. Fluorescence micrographs of each strain were taken and colocalizations were quantified for 20 cells per double-labelled strain. Quantifications were determined by calculating the Pearson's coefficient that measures the degree of linear dependence between the localization of a red signal and the localization of a green signal in the same cell [55] . Our analyses showed that Mcp5-Mcp6, Mcp5-McpH, Mcp5-McpM, Mcp6-Mcp4 and Mcp6-McpA significantly colocalized in cells (Pearson's coefficient >0.7) ( Figure 5 ). We used frzCD-gfp/aglZ-mCherry cells as negative control because FrzCD and AglZ have previously been shown to be exclusively localized in cells ( Figure 5 ) [53] . 10.1371/journal.pgen.1004164.g005 Figure 5 MCPs colocalization analysis. (A) Fluorescence micrographs of mcp5-mCherry mcpM-gfp and frzCD-gfp aglZ-mCherry cells are shown as examples. From (B) to (G) scatterplots of individual red and green pixel intensities of double-labeled cells are shown. (H) Average Pearson's correlation coefficients (PCCs) each calculated from ten scatterplots per strain. Che4, Che5, Che6 and multiple MCPs might constitute a large chemosensory module The phylogenetic and localization studies suggest that a large number of MCPs are recruited by the Che4, Che5 and Che6 pathways. To directly assess this hypothesis, we tested the interactions between Mcp4, Mcp5, Mcp6, McpH, McpA and McpM with all CheW-like proteins from the Che4, Che5 and Che6 pathways in a bacterial two-hybrid assay. Since these chemoreceptors were not predicted to interact with the Che7 and Che8 pathways, we included CheW homologs from these pathways as specificity controls. The interaction between Mcp7 and CheW7 was also used as a positive control. Except for McpA for which no interaction was detected with any of the tested CheW homologs, all tested MCPs interacted with at least one CheW from Che4, Che5 or Che6 ( Figure 6 ). Remarkably a high level of specificity was observed in some cases: for example, McpM only interacted with CheW5b and McpH only interacted with CheW4b. In other cases, one MCP could interact with several CheW proteins: specifically, Mcp4 interacted with all of the CheWs except for CheW4a and the negative controls; Mcp5 interacted with CheW4b and CheW5b; and Mcp6 interacted with CheW5b and CheW6a. As expected, none of these receptors interacted with CheW7, which specifically interacted with Mcp7, nor with CheW8a and CheW8b, which are phylogenetically distant ( Figure 6 ). Together, these results raise the possibility that M. xanthus Che utilizes higher order chemosensory modules comprised by several MCPs and Che pathways. The two-hybrid analysis suggests that each CheW has binding specificities that can be used to recruit multiple specific MCPs to a given signaling complex. 10.1371/journal.pgen.1004164.g006 Figure 6 \n In vivo MCP-CheW interactions. Bacterial two-hydrid assays on plates. Interactions between MCPs and CheWs are shown. +++, ++ and + indicate bacterial colonies turning red within 24 h, 48 h and 72 h respectively. “NS” (not significant) means that the colony color was as the negative control. We only show interactions resulting positive for both the pUT18C mcp /pKT25 cheW and pKT25 mcp /pUT18C cheW combinations and reproducible in two experiments performed in triplicate. Examples of colonies from negative control (empty plasmids); positive control (pUT18C mcp7 /pKT25 cheW7 ); + (pUT18C mcpM /pKT25 cheW4b ); ++ (pUT18C mcp4 /pKT25 cheW4b ); +++ (pKT25C mcpM /pUT18C cheW4b ) are shown. To further test the existence of a module comprised by the Che4, Che5 and Che6 systems and receptors, we combined deletions of cheA4 , cheA5 and cheA6 and analyzed motility and developmental phenotypes. Interestingly, ΔcheA4 , ΔcheA5 and ΔcheA6 double mutants are significantly more affected in S motility and fruiting body formation than single mutants ( Figure 7 ). However, these phenotypes are restored to wild type in a ΔcheA4ΔcheA5ΔcheA6 triple mutant. While this analysis does not reveal the precise biological function of the Che4, 5 and 6 pathways, it shows that the lack of two CheAs from this module deregulates the remaining CheA. This result strongly suggests that CheA4, CheA5 and CheA6 are part of the same regulatory module. 10.1371/journal.pgen.1004164.g007 Figure 7 \n ΔcheA triple mutants have restored phenotypes as compared to single and double mutants. (A) Motility was measured after 48 h. The colony spreading of each mutant was normalized with the one of a ΔpilA strain [68] completely incapable of S motility, to exclude cell growth effects. Error bars indicate standard deviations. The star corresponds to p <0.005. (B) ΔcheA fruiting body formation images at 72 h are shown. Conclusions In this study we sought to understand the partitioning of M. xanthus chemoreceptors among eight CSS to constitute sensory modules. We hypothesized that Che modules might attract multiple receptors and Che proteins as observed in other bacterial species and that the analysis of their cellular organization would help us to understand the role of these proteins in the M. xanthus life cycle. For this purpose, we first compiled a full list of the putative M. xanthus Che proteins and chemoreceptors. We were not surprised to find a total of 67 proteins as, in most cases, the number of one- and two-component systems present in a bacterial genome directly relates to the complexity of the life cycle [56] . The same might be true for CSS. Our systematic deletion of the 21 M. xanthus chemoreceptors and CheA encoding genes revealed that two thirds of them are involved in the temporal regulation of fruiting body formation, a multi-step differentiation process requiring the perception of numerous signals for the activation of key regulation check-points [20] , [57] . Based on an integrated approach, we found that MCPs and CSS show comparable phylogenetic distributions in three main groups and that MCPs belonging to the same phylogenetic group colocalize. In particular, MCPs of Group 3 seemed to constitute a large chemosensory module together with three CSS, namely Che4, Che5 and Che6 ( Figure 8 ). The presence of such a complex array of chemosensory proteins suggests that social behaviors such as cell group motility and biofilm formation might require interwoven regulatory systems composed by multiple Che-like systems and that the final cellular responses are generated following both the integration of signals transduced by different MCPs at the CheA level and the interaction among different Che systems. Also, cross-regulation between different Che systems can add an additional layer of complexity, as suggested by previous work showing inter-dependence between the Frz and the Dif pathway [58] . Once the composition of each module has been dissected, it will be possible to identify their signals and outputs to clarify their precise function in the M. xanthus life cycle. 10.1371/journal.pgen.1004164.g008 Figure 8 Schematic organization of M. xanthus Che modules as depicted from phylogenetic, cell biology and protein interaction analyses. For clarity, we omitted CheR and CheB proteins and do not specify the MCP-CheW interactions. MCPs in light green are the ones for which interactions with a CSS have not been demonstrated. The different color backgrounds indicate taxonomic Group 1 (green), Group 2 (blue) and Group 3 (pink). Group 1 was further divided in two subgroups labelled with light and dark green, based on the localization analysis. It has recently been reported that multiple chemosensory systems occur as frequently as single ones, highlighting the importance of investigating model microbes that encode multiple chemosensory systems [1] . By providing a broad perspective on how a complex multicomponent chemosensory apparatus is arranged within cells, this work establishes a basis for a deeper analysis on how signals are perceived, integrated and translated in cell behaviors at the level of each chemosensory module. Analogous approaches could be applied to bacterial systems with similarly complex regulatory networks."
} | 10,339 |
25027246 | PMC4100016 | pmc | 2,815 | {
"abstract": "Anaerobic methanotrophic archaea (ANME) play a significant role in global carbon cycles. These organisms consume more than 90% of ocean-derived methane and influence the landscape of the seafloor by stimulating the formation of carbonates. ANME frequently form cell consortia with sulfate-reducing bacteria (SRB) of the family Deltaproteobacteria. We investigated the mechanistic link between ANME and the natural consortium by examining anaerobic oxidation of methane (AOM) metabolism and the deposition of biogenetic minerals through high-resolution imaging analysis. All of the cell consortia found in a sample of marine sediment were encrusted by a thick siliceous envelope consisting of laminated and cementing substances, whereas carbonate minerals were not found attached to cells. Beside SRB cells, other bacteria (such as Betaproteobacteria) were found to link with the consortia by adhering to the siliceous crusts. Given the properties of siliceous minerals, we hypothesize that ANME cell consortia can interact with other microorganisms and their substrates via their siliceous envelope, and this mechanism of silicon accumulation may serve in clay mineral formation in marine sedimentary environments. A mechanism for biomineralization mediated by AOM consortia was suggested based on the above observations.",
"discussion": "Discussion The formation of a siliceous envelope is a bacteria-induced mineralization of clay that has been found to be a common eco-physiological process 24 . Clay-encrusted bacteria have been identified from varied geochemical environments, such as the sediment of iron-rich rivers or lakes and geothermal environments 24 25 . Generally clayey minerals can be formed via three pathways: Si(OH) 4 interacts with positively charged R–NH 3 + groups on the cell surface and induces silica to nucleate and grow 26 , dissolved silicate and aluminum species adhere indirectly to negatively charged COOH − or PO 4 3− groups on the cell wall and associated exopolymer via a metal cation bridge 24 , or colloidal species of (Fe, Al)-silicate react directly with either cellular polymers or adsorbed iron that eventually transform into clay minerals 24 . It has been reported that the physiological activity of cyanobacteria can shift the local pH from 3.4 to 9 (or higher) to induce the dissolution of quartz 27 . AOM consortia may similarly cause a release of silicon from the solid phase. Silica has high solubility and dissolution rates in alkaline solutions. In water with a neutral pH, amorphous and crystalline SiO 2 converts to Si(OH) 4 following a slow hydrolysis step that produces hydroxyl ions to stimulate the polarization and break-up of the Si-O 28 . In cold seep environment, HCO 3 − and HS − produced from AOM cause an increase in alkalinity. Experimentally, ANME and associated bacteria in a high-pressure continuous bioreactor resulted in an increase of pH from 7.0 to 8.5 after 40 days incubation 29 . A recent study on silicification in seeps suggested that the increase in pH resulting from AOM was the main factor for the dissolution of diatom silica skeletons 30 . However, high-resolution in situ measurements of pH in sediment with high AOM activity showed that the pH value varied between 7.7 and 7.9, which is a typical pH range for marine sediment 31 . It is possible that the pH increase resulting from the eco-physiological activity of AOM is compensated by protons released during silica dissolution and carbonate precipitation. We suggest that AOM may accelerate the process of silica leaching and that the free silica will redeposit on the surface of the AOM consortia via any of the above-mentioned pathways. The absence of carbonate minerals on the surface of AOM consortia suggests that the microenvironment of the consortia is not favorable for the precipitation of carbonate. Organic components secreted by microorganisms either favor or inhibit the precipitation of carbonate, depending on their intrinsic characteristics 20 32 ; however, this inhibition does not affect carbonate precipitation elsewhere in the bulk system. We compared a stored marine sediment sample and an AOM enrichment culture to exclude potential bias resulting from laboratory manipulations. The same extracellular structures were identified in all of the AOM consortia, which strongly suggests active biomineralization mediated by AOM consortia. This AOM-mediated biomineralization is summarized in the model shown in Figure 7 . AOM metabolism increases the alkalinity of the microenvironment, accelerating the leaching and dissolution of silica, which re-deposits on the surface of the consortia by biological adsorption. The primary poorly crystallized silica gradually ages to more stable crystalline phases over time. The precipitation of carbonates on the surface of the consortia is, however, prohibited due to the high solubility product values of carbonates when compared with those of clay minerals. The development of a siliceous envelope may benefit the microbial consortia in at least three aspects. First, it will enhance the structural integrity of the syntrophic members. The association between ANME and SRB is thought to be the result of an ancient type of multicellular symbiosis 33 . AOM consortia grow by increasing the cell numbers and size of the consortia, and larger consortia may eventually break into two or more aggregates 19 . Second, as enclosure by the siliceous envelope may keep ANME and SRB together, it has also been suggested that clay-like minerals may provide a source of exchangeable nutrients 22 . Experimental studies have demonstrated that clays, such as montmorillonite and kaolinite, can host methane hydrate 34 35 . Finally, it would be very interesting to test in the future whether the clay envelope surrounding the AOM cell consortia is capable of methane storage, thus serving as a cell-adhering material and an exchangeable matrix of nutrients, as described above. The discovery of AOM consortia encrusted with clayey minerals also reveals a mechanism involved in the preservation and fossilization of AOM in modern and ancient marine sediments. AOM consortia in hydrocarbon seeps are common in the sediments 1 14 where methane-derived 13 C-depleted carbonate precipitates 36 . However, no direct fossilization of AOM cells has been discovered yet. Recent studies on the microstructures in methane seep carbonates interpreted these bacteriomorphs to be pseudofossils resulting from the diagenetic alteration of euhedral Fe-sulfide framboids 37 . On the basis of that work, we argue that microstructures with a particular Si-Al composition in cold-seep 13 C-depleted carbonates may be fossilized AOM consortia. This mechanism of mineralization could be used in searching for fossilized AOM in diagenetic marine carbonaceous sediments or ancient marine sedimentary rocks. Considering that AOM may have represented an important biological process in the ancient anoxic biosphere, this siliceous structure may have played historically significant roles in the oceanic silicon cycle and AOM burial, similar to that played by the siliceous cell walls of diatoms and cyanobacteria in modern oceans."
} | 1,795 |
31289699 | PMC6598746 | pmc | 2,816 | {
"abstract": "Coral reefs rely on their intracellular dinoflagellate symbionts (family Symbiodiniaceae) for nutritional provision in nutrient-poor waters, yet this association is threatened by thermally stressful conditions. Despite this, the evolutionary potential of these symbionts remains poorly characterised. In this study, we tested the potential for divergent Symbiodiniaceae types to sexually reproduce (i.e. hybridise) within Cladocopium , the most ecologically prevalent genus in this family. With sequence data from three organelles ( cob gene, mitochondrion; psbA ncr region, chloroplast; and ITS2 region, nucleus), we utilised the Incongruence Length Difference test, Approximately Unbiased test, tree hybridisation analyses and visual inspection of raw data in stepwise fashion to highlight incongruences between organelles, and thus provide evidence of reticulate evolution. Using this approach, we identified three putative hybrid Cladocopium samples among the 158 analysed, at two of the seven sites sampled. These samples were identified as the common Cladocopium types C40 or C1 with respect to the mitochondria and chloroplasts, but the rarer types C3z, C3u and C1# with respect to their nuclear identity. These five Cladocopium types have previously been confirmed as evolutionarily distinct and were also recovered in non-incongruent samples multiple times, which is strongly suggestive that they sexually reproduced to produce the incongruent samples. A concomitant inspection of next generation sequencing data for these samples suggests that other plausible explanations, such as incomplete lineage sorting or the presence of co-dominance, are much less likely. The approach taken in this study allows incongruences between gene regions to be identified with confidence, and brings new light to the evolutionary potential within Symbiodiniaceae.",
"conclusion": "Conclusions This study cannot be considered unequivocal proof of Cladocopium hybridisation. However, the unambiguous evidence for incongruence between nuclear and organellar gene regions shows the value of the stepwise approach taken here, and conforms to the hypothesis of hybridisation between divergent taxa. While ILS remains a possibility, it is a less intuitive explanation, especially in the light of incongruent samples having clearly distinct, predefined types which were recovered in non-incongruent samples, and the failures of background populations to consistently align to its predictions. Therefore, hybridisation appears to be a credible, if infrequent, mechanism for adaptive change in Cladocopium , and potentially for Symbiodiniaceae in general, though multiple sources of intragenomic variation remain analytically problematic. Ascertaining the frequency and extent of this may be vital to predicting the fate of coral reefs in an environmentally unpredictable future.",
"introduction": "Introduction Coral reefs are a highly diverse and important ecosystem, yet are significantly threatened by anthropogenically-driven climate change ( Hughes et al., 2017 ). In order for coral reefs to survive the stresses of a changing climate, genetic adaptation over rapid evolutionary timescales has to occur. Adaptation in the coral itself may go some way to provisioning for the environmentally challenging conditions predicted to come ( Rodriguez et al., 2009 ). However, given that the response of corals to environmental conditions is inextricably linked to the diversity and performance of their intracellular symbionts (dinoflagellates of the family Symbiodiniaceae, LaJeunesse et al., 2018 ), increasing attention is being focused on the evolutionary potential within this family. Coral symbionts have been thought to be exclusively asexual in hospite ( Trench, 1997 ; LaJeunesse, 2005 ), thanks to their isolated position sequestered inside host cells, and the hypothesis that endosymbiotic sex would encourage exploitation of the host ( Law & Lewis, 1983 ). However, previous work in other taxa has shown that intracellular symbionts can sexually reproduce ( Chesnick & Cox, 1987 ). In general, it is thought that many such organisms may have cryptic sexual cycles that have previously been unappreciated, in addition to the production of clonal populations via asexual reproduction ( Heitman, 2010 ). Now, there is significant evidence that Symbiodiniaceae also displays a mixed reproductive strategy, with periods of asexuality interspersed with occasional to frequent sex ( Thornhill et al., 2017 ). While it has never been explicitly observed, there are distinct and observable traces of sex in their genomes ( Baillie et al., 2000 ; LaJeunesse, 2001 ; Santos & Coffroth, 2003 ; Santos et al., 2004 ; Pettay et al., 2011 ; Baums, Devlin-Durante & LaJeunesse, 2014 ; Chi, Parrow & Dunthorn, 2014 ; LaJeunesse et al., 2014 ; Thornhill et al., 2014 ; Levin et al., 2016 ). However, these studies have been largely focused on a micro-scale, population level (i.e. intraspecific sex). By contrast, sex between diverse symbiont lineages (‘hybridisation’) has received little attention in the literature (but see Wilkinson et al., 2015 ). Given the highly thermally stressful conditions predicted by the end of the century ( Kirtman et al., 2013 ), the mechanism of hybridisation could potentially have significant and vital adaptive value. By mixing diverse pools of genetic material, hybridisation can allow for rapid adaptation, facilitating macro-evolutionary jumps ( Willis et al., 2006 ; Dittrich-Reed & Fitzpatrick, 2013 ). Introgressive hybridisation, where the F1 hybrids subsequently mate with one or both parent populations, can transfer a large quantity of genetic material between the two parent lineages in the space of a few generations. In addition, hybridisation can also produce offspring with elevated fitness (‘hybrid vigour’), which can even outcompete the parent species ( Ellstrand & Hoffman, 1990 ; Rhymer & Simberloff, 1996 ). Importantly, instances of hybridisation have also been shown to increase in taxing conditions ( Rhymer & Simberloff, 1996 ; Moran & Alexander, 2014 ). Therefore, the possibility of hybridisation in coral symbionts raises the potential for adaptation at the required pace and scale for survival. Research on taxa with similar life-histories suggests that hybridisation is plausible. Hybridisation has previously been reported in a range of dinoflagellate genera, including Dinophysis, Protoperidinium, Preperidinium and Diplopsalis ( Edvardsen et al., 2003 ; Gribble & Anderson, 2007 ; Hart et al., 2007 ). There is also evidence from plant-fungi relationships that endosymbionts can successfully hybridise. In particular, the endophytes Epichloë spp. are pathogenic or mutualistic fungi that inhabit a wide range of grasses. Hybridisation appears to be a major mechanism for diversification in this genus, and has been reported to occur inside the grasses Lolium perenne ( Schardl et al., 1994 ), Festuca arundinacea ( Tsai et al., 1994 ), Bromus laevipes ( Charlton et al., 2014 ) and Poa alsodes ( Shymanovich et al., 2017 ). In several instances, multiple cases of hybridisation have been recorded, and evidence put forward that those hybrids are fitter than non-hybrids ( Schardl et al., 1994 ; Moon et al., 2004 ). While Symbiodiniaceae in hospite are generally sequestered inside host cells ( Davy, Allemand & Weis, 2012 ), the extensive presence of background symbiont populations inside hosts ( Santos, Taylor & Coffroth, 2001 ; Kemp et al., 2015 ), the observation that corals themselves hybridise ( Willis et al., 2006 ; Combosch & Vollmer, 2015 ), and the existence of a free-living state ( Coffroth et al., 2006 ; Nitschke, Davy & Ward, 2016 ) mean that it is highly possible that at some point diverse symbiont communities may interact, with the possibility for sexual reproduction. The evolutionary potential of hybridisation has not been targeted within Symbiodiniaceae. However, several indirect observations are suggestive of its occurrence, all within Cladocopium , the most prevalent genus. LaJeunesse et al. (2003) reported an ITS2 sequence variant they called C1c and treated as an intragenomic variant, as it was only observed in Denaturing Gradient Gel Electrophoresis (DGGE) profiles associated with type C1. However, it was then discovered to be an independent type and called C45 ( LaJeunesse, 2005 ). Therefore, the additive DGGE pattern shown in LaJeunesse et al. (2003) could have in fact resulted from the hybridisation of C1 and C45. LaJeunesse (2005) also defined type C3m using the ITS2 region, which has co-dominant characteristics of both C1 and C3, a pattern attributed to either sexual recombination or homoplasy. A similar scenario was also recorded in symbiont type C3h, an apparent intermediary between C3 and C21 ( LaJeunesse et al., 2004 ). This time, the pattern was hypothesised to be due to incomplete lineage sorting (ILS) or sexual recombination between the two different types. Indeed, given the unambiguous existence of ‘pure’ C3 and C21 in the samples, sexual recombination is a credible explanation. Finally, Wilkinson et al. (2015) reported two symbiont types but three distinct symbiont populations inside a single Pocillopora colony: C100 symbionts, C109 symbionts and symbionts having co-dominant C100 and C109 repeats in the same cell. Again, the extensive presence of the two ‘pure’ populations means ILS is a less parsimonious explanation than hybridisation. However, it cannot be completely eliminated as a possibility. In addition, this study took place at Lord Howe Island, the world’s southern-most coral reef, and therefore may not be widely applicable across less marginal, low-latitude sites. Hence, there is a body of indirect evidence for sexual recombination between diverse symbiont types (hybridisation sensu lato ), and this warrants further study. The current study aimed to gather further defendable evidence as to whether hybridisation occurs in coral symbionts. Because it is very difficult to observe hybridisation directly, it is generally inferred through genetic signals. One of the most common of these is incongruence between gene regions. Because nuclear genes are typically inherited biparentally, while organelle genes are inherited uniparentally, sexual reproduction between different species will result in organelle genes resembling one parent only, while the nuclear genome will have clear traces of both parents ( Rieseberg, Whitton & Linder, 1996 ). In extreme cases, repeated backcrosses with a parent type can result in organelle capture, where novel, discordant nuclear-organellar combinations are observed ( Folk, Mandel & Freudenstein, 2017 ). Following a hybridisation event, selection can also act to produce incongruence between gene regions: there may be elevated (or reduced) fitness of certain nuclear-cytoplasmic combinations, or selection pressure may be different for nuclear and cytoplasmic genomes (e.g. a greater selection pressure acting on nuclear genes) ( Rieseberg, Whitton & Linder, 1996 ). Therefore, identifying incongruence between gene regions is a common method for assessing potential hybridisation ( Planet, 2006 ; Govindarajulu et al., 2015 ), and was utilised in the current study. The chosen location for this study, Atauro Island and the north coast of Timor, is in the Coral Triangle and therefore widely applicable to other important reef systems. The hypothesis tested was that hybridisation between distinct Cladocopium genotypes has occurred at these sites, as evidenced by gene regions in separate organelles ( cob , mitochondrion; ITS2, nucleus; psbA ncr , chloroplast) having experienced different evolutionary histories. Defendable evidence of hybridisation would be a significant step towards understanding the evolution of Symbiodiniaceae and potential coral reef persistence in the future.",
"discussion": "Discussion Methodological approach taken There are many factors, such as character sampling and bias due to differential gene length, which can give false signals of incongruence ( Som, 2014 ). However, the approach taken in this study has been able to clearly display incongruence between organellar and nuclear regions in Cladocopium . In isolation, it is true that there are issues with the tests utilised. For example, the AU test presented an issue with most trees being incongruent for the psbA ncr region. The psbA ncr region is highly variable ( LaJeunesse & Thornhill, 2011 ; Thornhill et al., 2014 ), and hence a more complex tree is required to explain it. The cob and ITS2 trees with multiple polytomies could not do this as effectively, and hence a result of incongruence was returned. Therefore, the results from the cob and ITS2 datasets are likely more reliable, and were the focus of the Results. Further, the ILD test has been criticised for being overly sensitive, especially when comparing partitions of different resolutions ( Barker & Lutzoni, 2002 ). The refutation of this is simple: in all cases, it found congruence between the psbA ncr and cob regions, the two most different in terms of resolution ( Table 2 ), so this is clearly not contributing to the positive results between the organellar and nuclear partitions observed here. Indeed, it failed to reject congruence between the cob and psbA ncr regions for sites BSP and BHB despite the tree hybridisation analyses finding potential incongruence ( Figs. 2C and 4C ), and so appears to be reasonably conservative in this case. The results of the AU and ILD tests are also compelling because they are differential: they show consistently different patterns between datasets and are therefore likely responding to genuine phylogenetic signals. This was confirmed by looking at the raw sequence data, and shows the efficacy of the approach taken here. With such a wide range of samples, initially searching for incongruences in sequence data would be functionally impossible, as it would require comparing all possible combinations of sequences (in this study, this would require 1.17 × 10 278 comparisons). However, the stepwise use of analyses allowed the initial identification of which sites may host incongruent samples, and then visualisation on phylogenetic trees allowed simple alignments of appropriate samples to be generated, where incongruence could clearly be refuted or confirmed. In addition, given the issues with tests in isolation, the multiplicity of analyses used generates a far more convincing picture of reticulate evolution. Hybridisation in Cladocopium ? Incongruence was comprehensively established for the samples BHB146, BSP343 and BSP364. However, this does not necessarily translate to hybridisation, as there are a range of analytical or biological factors that can cause incongruence in phylogenetic data. For example, one hypothesised to be quite common but insidious in its undetectable nature is heterotachy, shifts in site-specific evolutionary rates through time ( Som, 2014 ). While there is no particular way to identify heterotachy or exclude it as a cause, except with a very large number of sequences, ML methods in particular have been shown to be robust to even intermediate levels of heterotachy ( Som, 2014 ). A more plausible explanation is ILS, often considered the most common cause of incongruence ( Degnan & Rosenberg, 2009 ). This is due to polymorphisms not segregating fully during speciation events, leading to phylogenetic signals in gene trees that conflict with the overall species tree. This has been shown to be quite common in the ITS2 region, thanks to its multiple-copy nature ( Thornhill, LaJeunesse & Santos, 2007 ). Through this mechanism, ancestral polymorphisms may persist at low levels in the genome. Therefore, it is possible that the divergent sequences recovered actually represent a single symbiont population, which has multiple ancestral polymorphisms present via ILS (i.e. intragenomic variation). Through stochastic DNA processes such as unequal crossing over, slipped-strand mispairing and transposition, these intragenomic variants may be eliminated or promoted in the multiple-copy array (i.e. concerted evolution, see Nei & Rooney, 2005 ). Hence, in the samples from a single reproductively isolated population, one ancestral polymorphism may be dominant in the ITS2 region of some, while a different ancestral polymorphism may be dominant in others. This would cause the patterns observed in this study, with the ITS2 region being occasionally incongruent with the organellar regions. Ideally, a statistical test would be carried out to differentiate between hybridisation and ILS, and such tests do exist. However, they require inputs of information which are not currently available for Cladocopium , such as: (a) An understanding of the effective population size N e ( Pelser et al., 2010 ); (b) a large number of genes, at least some of which must be adjacent ( Pollard et al., 2006 ; Meng & Kubatko, 2009 ); or (c) strictly bifurcating trees and clearly defined species ( Sang & Zhong, 2000 ; Joly, McLenachan & Lockhart, 2009 ). Therefore, ILS as a cause of the observed incongruence cannot be statistically refuted. However, there is good evidence that the patterns observed here are more likely to be caused by symbiont hybridisation. First, the pattern of incongruence observed, with organellar cytoplasmic genes being different to nuclear genes, accords with a large body of prior theory on hybridisation. Nuclear genes are largely inherited biparentally, and the ITS2 region is no exception ( Baldwin et al., 1995 ; Rybalka et al., 2013 ). However, the cytoplasm tends to be inherited maternally ( Rieseberg, Whitton & Linder, 1996 ). This difference is largely due to gametogenesis and fertilisation, where the male gamete typically only contains nuclear information, while the female gamete (egg) contains the cytoplasm that will be passed on to the zygote. Therefore, if an organism encounters a population of another species and produces viable hybrids, theory predicts that over time, repeated backcrossing with the more common species (introgression) will produce hybrids with divergent organellar and nuclear signals. While the nature of the sexual life cycle has yet to be fully elucidated in the Symbiodiniaceae, previous evidence has shown that other unicellular dinoflagellates produce gametes ( Brawley & Johnson, 1992 ). In addition, the presence of ‘plus’ and ‘minus’ mating types, analogous to gender, has been shown in the dinoflagellate Alexandrium tamarense ( Brosnahan, 2011 ). Therefore, it is reasonable to assume that Symbiodiniaceae also produce distinct gametes (as opposed to conducting sex via fusion, for example), making this mechanism eminently plausible. The documentation of functional meiotic genes in Symbiodiniaceae ( Chi, Parrow & Dunthorn, 2014 ; Levin et al., 2016 ) supports this assertion. Such a pattern of discordance between cytoplasmic and nuclear genes caused by hybridisation has been recorded for taxa as diverse as plants ( Rieseberg, Whitton & Linder, 1996 ; Pelser et al., 2010 ; Sun et al., 2015 ), beetles ( Sota & Vogler, 2001 ) and indeed corals ( Van Oppen et al., 2001 ). In general, hybridisation is predicted to cause incongruence between nuclear and cytoplasmic markers in both multicellular and unicellular taxa ( Bull et al., 1993 ). Other factors due to hybridisation, such as semigamy or differential fitness of nuclear-cytoplasmic combinations, can also cause incongruence between nuclear and cytoplasmic gene trees ( Rieseberg, Whitton & Linder, 1996 ). Therefore, the fact that this was the pattern observed in this study is strong circumstantial evidence that hybridisation is the explanation. In addition, hybridisation is made more likely in comparison to ILS by the fact that all of the incongruent ITS2 sequences were previously defined types (i.e. not unique sequences), that were also present in non-incongruent relationships in the analyses. For example, BSP364 had a generic Cladocopium type C3 sequence for the cob gene, was a C40 type for the psbA ncr region, and C3z for the ITS2 region. Significantly, there were also samples recovered which were type C40 for both the psbA ncr and ITS2 regions (samples BSP319–BSP375, Fig. 4B ), and samples which were type C3z for both regions (samples BSP373–BSP386, Fig. 4B ). This confirms that they are clearly separate types, supported by the fact that they differ by four base pairs in the ITS2 sequence and 64 base pairs in the psbA ncr region (including a 49 base pair deletion in the C40 sequences), indicating that this is not just a non-diagnostic polymorphism ( Wilkinson et al., 2015 ). The implications for this being caused by ILS are given in Fig. 8 . Only the psbA ncr and ITS2 genes are presented, as the cob gene was invariant in this case. 10.7717/peerj.7178/fig-8 Figure 8 Predictions under incomplete lineage sorting. (A) General pattern expected for ILS. A single ancestral population with polymorphism in both the psbA ncr and ITS2 regions is present before a speciation event. After speciation, the ITS2 polymorphism fails to segregate, while through stochastic processes the C40 polymorphism is eliminated and leads to incongruence between nuclear and chloroplast genes. (B) The process of ILS that would be required for this example. The ITS2 region fails to segregate after speciation; despite the extensive presence of C40 alleles, a small subpopulation of symbionts with dominant C3z alleles is maintained (weak dashed blue line) in the C40 population and both are recovered in present-day sampling, at the same site, as pure C3z populations. Figure 8B graphically represents the process that would be required for the observed patterns to be due to ILS. Given that symbiont sex is now strongly supported (though in low frequency; Thornhill et al., 2017 ), it seems unlikely that a divergent ancestral polymorphism could be maintained as the dominant sequence in some samples within type C40, as it would be expected that repeated recombination would eventually remove C3z traces from the C40 genome, or vice versa ( Fig. 8A ). It is more parsimonious that a hybridisation event has occurred between symbiont types C40 and C3z, with backcrossing leading to incongruence between organellar and nuclear genes. This is strongly supported by the analysis of the background symbiont populations ( Fig. 7 ). The results show that there is little evidence of C40 and C3z sequences being shared within samples. The C3z population had almost no C40 sequences present at all, with just one sample having an extremely low background abundance of C40 ( Fig. 7B ). C3z sequences were slightly more common in C40 samples ( Fig. 7D ). However, the low proportion of background sequences in this population ( Fig. 7C ) meant that overall the presence of C3z in the C40 population was negligible (mean = 1.61%, median = 0). This reveals essentially pure populations of C40 and C3z at site BSP, something which strongly favours hybridisation vs. ILS as causing the mixed pattern in BSP364 ( Wilkinson et al., 2015 ). While the other two putative hybrid ITS2 types (C3u, C1#) do not have large populations to compare, the same basic pattern was also observed for BSP343, which was identified as Cladocopium type C40 for the organelle regions, and type C3u for ITS2. If this was to be caused by ILS, then both variants would be expected to occur in the ITS2 region, (with one at low frequency), but the NGS data revealed no trace of ITS2 type C40 in that sample. Further, the divergences observed (i.e. C40/C3u, C40/C3z, C1/C1#) all coalesce at the ‘ancestral’ types C1 or C3, rather than one representing an intermediate evolutionary step to the other. Therefore, ILS would also predict these ancestral sequences to be in the ITS2 genome in low frequencies. However, this was only observed in BSP343 (as the fourth most common sequence); neither BHB146 nor BSP364 showed any evidence of these ancestral sequences. While it is acknowledged that hybridisation and ILS are not mutually exclusive and the incongruences observed could be caused by a combination of both, the weight of evidence suggests that these results are more likely a result of interspecific hybridisation between distinct symbiont types. Potentially, the two competing hypotheses could be distinguished by sequencing another nuclear gene, less susceptible to intragenomic variation, for both putative hybrid samples and closely related sequences. If the patterns were due to hybridisation, it would be expected that the additional nuclear gene would support the ITS2 identity, and cluster the sample with the same group as presented in the ITS2 trees ( Figs. 2 and 4 ). In contrast, if the incongruence was caused by ILS, the additional marker would cluster the putative hybrid with the same samples as the organellar gene regions. This was attempted using the actin gene. Unfortunately, low resolution (and difficulties in amplification leading to short usable sequences) meant that neither scenario was supported, as the sequences were not variable enough to recover the groups observed in Figs. 2 and 4 . The other currently-available Symbiodiniaceae nuclear gene markers either suffer from the same issue of significant intragenomic variation (ITS1), or are lower-resolution than actin (SSU, LSU, 5.8S, elf2 ), and therefore the patterns observed cannot currently be independently verified. The further development of highly-variable, reliably amplifiable nuclear gene markers should be a priority for Symbiodiniaceae systematics. However, ILS (and indeed all analytical factors), are random or would be expected to affect all sites. The results obtained, however, are anything but random, with two sites consistently being recovered as incongruent in contrast to all others, despite those incongruences coming from a range of host species that were present at all sites. In addition, both these sites have been shown to be rich in Symbiodiniaceae diversity, when compared with the Timor sites ( Brian, Davy & Wilkinson, 2019 ). This suggests that putative hybridisation may be limited to high-quality sites that maintain high levels of symbiont diversity. Intragenomic variation within the ITS2 region could lead to the incongruences observed via ILS, though the discussion above suggests that hybridisation should be favoured as an explanation. However, the psbA ncr region can also be intragenomically variable ( LaJeunesse & Thornhill, 2011 ), with intragenomic sequence ratios that may fluctuate within a single species. Combined with the intragenomic variation of the ITS2 region, this generates the potential for a wide range of ITS2-psbA ncr combinations within a single genome. For example, one member of a population may have a 9:1 ratio of variant A:variant B within its multiple-copy ITS2 sequences, and a 9:1 ratio of variant A:variant B in its multiple-copy psbA ncr sequences. A second member of the same population could plausibly have an 8:2 ratio of variant A:variant B within its ITS2 sequences, and a 4:6 ratio of variant A:variant B in its psbA ncr sequences. Therefore, there is a possible difference between the most common sequences in total, and the most common associations between ITS2 and psbA ncr ratio types. If only the most common sequences are studied, there is the potential for some natural associations (i.e. not caused by hybridisation) to appear as incongruences. The present study attempts to draw conclusions based on common associations between nuclear and organellar genes, but is only able to utilise common sequences. This is explicit for the psbA ncr region (as Sanger sequencing only amplifies the most common overall sequence), and implicit for the ITS2 region (as the nature of the tests necessitated the selection of the most common overall sequence from NGS data). Within a single genome, a solution would be to sequence multiple markers from the same DNA strand through long-read sequencing, which would preserve the ratio of intragenomic variants. However, this does not work for markers across multiple organellar and nuclear genomes, and currently there is no other acceptable solution to this problem for Symbiodiniaceae. While perhaps unlikely, this issue could potentially explain the patterns seen, and should be acknowledged. Previous tests of incongruence No previous study on Symbiodiniaceae seriously considers symbiont hybridisation, except that of Wilkinson et al. (2015) , which also finds evidence for its existence. However, aside from the potential examples of hybridisation mentioned in the Introduction of this study ( LaJeunesse et al., 2003 , 2004 ; LaJeunesse, 2005 ), three other studies bear mention. Sampayo, Dove & LaJeunesse (2009) also focused on the basis that hybridisation can cause incongruence between genes from different organelles, and built trees from mitochondrial, chloroplast and rDNA nuclear gene regions to test this. Based on visual inspection of these trees, they concluded that different symbiont lineages (types) within Cladocopium are reproductively isolated. Interestingly, they did also use the ILD test to formally test incongruence, which returned a p -value of 0.01, though this result was not explored further. Pochon, Putnam & Gates (2014) assessed six genes from three different organelles (mitochondrion, nucleus and chloroplast). In all cases, they found evidence of incongruence between pairwise comparisons of genes, using the AU test. While they go on to discuss the implications for concatenation in some detail, the cause of these incongruences was likewise not explored further. Another study from Pochon et al. (2006) found the surprising result of incongruence between whole genera rendered from nr28S and cp23S data, using the SH test. However, when they removed all but two members of each clade, the test then showed congruence between datasets. This indicated incongruence was being caused by the accumulation of occasional within-clade mismatches between the nucleus and chloroplasts, something which is also broadly agreeable with a hypothesis of hybridisation in low frequency. These studies certainly do not provide conclusive evidence of hybridisation. However, it is reasonably striking that four studies conduct an explicit statistical test of incongruence within Symbiodiniaceae ( Pochon et al., 2006 , Pochon, Putnam & Gates, 2014 ; Sampayo, Dove & LaJeunesse, 2009 ; this study), and all four find evidence for its existence. At the very least, these add to the body of evidence that the family Symbiodiniaceae has not evolved in a simple linear fashion, and justifies a more careful consideration of patterns of incongruence within this family."
} | 7,705 |
31038122 | PMC6491038 | pmc | 2,817 | {
"abstract": "Bacterial swarming and biofilm formation are collective multicellular phenomena through which diverse microbial species colonize and spread over water-permeable tissue. During both modes of surface translocation, fluid uptake and transport play a key role in shaping the overall morphology and spreading dynamics. Here we develop a generalized two-phase thin-film model that couples bacterial growth, extracellular matrix swelling, fluid flow, and nutrient transport to describe the expansion of both highly motile bacterial swarms, and sessile bacterial biofilms. We show that swarm expansion corresponds to steady-state solutions in a nutrient-rich, capillarity dominated regime. In contrast, biofilm colony growth is described by transient solutions associated with a nutrient-limited, extracellular polymer stress driven limit. We apply our unified framework to explain a range of recent experimental observations of steady and unsteady expansion of microbial swarms and biofilms. Our results demonstrate how the physics of flow and transport in slender geometries serve to constrain biological organization in microbial communities.",
"introduction": "Introduction Bacteria employ sophisticated surface translocation machinery to actively swarm, twitch, glide or slide over solid surfaces ( Kearns, 2010 ; Mattick, 2002 ; Spormann, 1999 ; Hölscher and Kovács, 2017 ). Collectively, they also aggregate into multicellular communities on hydrated surfaces and exhibit large-scale coordinated movement ( Verstraeten et al., 2008 ). Surface motility in macroscopic colonies on hydrated surfaces such as gels occurs primarily via two distinct modes: either by rapid flagella-mediated swarming expansion ( Harshey, 1994 ; Harshey, 2003 ), or alternatively by slow biofilm expansion driven by extracellular polymer matrix production ( Hall-Stoodley et al., 2004 ). In both cases, an interplay between mechanical constraints and biological organization sets limits on the overall colony morphology and expansion dynamics ( Persat et al., 2015 ). The forces driving colony expansion are generated by non-homogeneous patterns of biological activity, originating from spatial localizations in cell growth and division ( Hamouche et al., 2017 ), extracellular polymer matrix production ( Seminara et al., 2012 ; Yan et al., 2017 ; Srinivasan et al., 2018 ), osmolyte secretion ( Ping et al., 2014 ) and active stresses ( Farrell et al., 2013 ; Delarue et al., 2016 ). Conversely, the formation of localized biologically active zones is tightly coupled to the heterogeneity of the environment, including the diffusion and transport of nutrients ( Wang et al., 2017 ), accumulation of metabolic by-products ( Liu et al., 2015 ; Gozzi et al., 2017 ) and presence of quorum sensing and signaling agents that regulate cell-differentiation and development. Consequently, the dynamics of colony growth requires a mechanistic description that accounts for spatiotemporal inhomogeneities in biological activity, emergent forces, and flows that transport metabolic agents. In bacterial swarming, cells within the colony are actively propelled by the rotation of flagella in a thin layer of fluid extracted from the underlying soft tissue or gel ( Kearns, 2010 ). In contrast, bacterial biofilms are surface aggregates of sessile bacteria embedded in a self-generated extracellular polymer matrix ( Flemming and Wingender, 2010 ). Despite marked differences in regulatory genetic pathways, morphology and cell function ( Verstraeten et al., 2008 ), physical characteristics such as the fluidization of the substrate/tissue, gradients in nutrient availability, the low-aspect-ratio geometry and the existence of multiple phases (i.e. cells, biopolymer and fluid) are common to both bacterial film and swarm colonies. Motivated by these similarities, we present a unified multiphase framework that couples mechanics, hydrodynamics and transport to explain the dynamics of bacterial swarm and film expansion.",
"discussion": "Discussion Analysis of collective microbial expansion in thin film geometries often prioritizes biological mechanisms, such as genetic regulation, developmental programs and cellular signaling/competition, over the role of the heterogeneous physical micro-environments. Here we have presented a multi-phase theory that quantitatively describes the expansion dynamics of microbial swarms and biofilms and considers variations in the colony thickness, an aspect of colony expansion that has often been overlooked in many theories ( Korolev et al., 2012 ; Ghosh et al., 2015 ; Wang et al., 2017 ). The resulting unified description of both steady-state swarms and transient biofilm spreading leads to simple estimates and scaling laws for the colony expansion rate that are validated via comparison with experimental measurements for different systems. In swarms, exudation of water from the permeable substrate via bacterial osmolyte secretion facilitates steady state colony expansion. Numerical solutions of our model demonstrate that the shape of the swarm front is determined by capillarity, and its expansion speed by cell-division and growth, leading to scaling laws validated by comparison with previous experiments. In contrast, transient biofilm macrocolony expansion on agar is driven by osmotic polymer stresses generated via EPS matrix production in a spatially localized zone at the periphery. Nutrient transport and depletion leads to the formation of these heterogenous zones, and results in two regimes in biofilm expansion. However our depth-integrated theory also has certain limitations. For example, we are unable to capture discrete thickness variations of the order of a few cells, which might require an agent-based approach. For bacterial swarms, our model is unable to quantitatively account for the region of enhanced thickness (i.e., the multilayer region in Figure 1C and E ), likely because the multilayer width is difficult to experimentally ascertain, owing to the large tail distribution seen in the mean intensity trace in Figure 1E , and the arbitrariness in the choice of threshold in Appendix 3—figure 5 . Similarly, in the context of biofilm colony expansion, our model does not account for sliding and frictional contact between the cells/EPS matrix and the substrate ( Farrell et al., 2013 ). More generally, our mean-field picture neglects fluctuation-driven effects during colony expansion, such as the formation multicellular raft structures ( Kearns, 2010 ) and synchronized long-range interactions ( Chen et al., 2017 ). Natural next steps of our approach include (i) adding three-dimensional effects by allowing for spatial variations in the mechanical stresses, flows and nutrient fields in the vertical direction, (ii) accounting for orientational order in the bacterial swarms and films, and (iii) accounting for interfacial tension on the stability of the growing swarm/biofilm-fluid interface, especially in the context of fingering instabilities in microbial colonies Trinschek et al. (2018) . A rigorous multi-phase approach may also be relevant in revisiting pattern formation phenomena in microbial colony expansion ( Matsushita et al., 1999 ), that so far been addressed primarily using various non-linear diffusion models ( Golding et al., 1998 ; Allen and Waclaw, 2019 ) that ignore the third dimension. Finally, from an experimental and theoretical perspective, our results naturally raise the question of controlling biofilm and swarm expansion by manipulating water and nutrient availability, complementing the better studied approaches of manipulating colonies by the genetic regulation of EPS production, cell division, and chemical signaling in microbial colonies."
} | 1,926 |
23185499 | PMC3502180 | pmc | 2,818 | {
"abstract": "Quorum sensing in Burkholderia cenocepacia H111 involves two signalling systems that depend on different signal molecules, namely N -acyl homoserine lactones (AHLs) and the diffusible signal factor cis -2-dodecenoic acid (BDSF). Previous studies have shown that AHLs and BDSF control similar phenotypic traits, including biofilm formation, proteolytic activity and pathogenicity. In this study we mapped the BDSF stimulon by RNA-Seq and shotgun proteomics analysis. We demonstrate that a set of the identified BDSF-regulated genes or proteins are also controlled by AHLs, suggesting that the two regulons partially overlap. The detailed analysis of two mutually regulated operons, one encoding three lectins and the other one encoding the large surface protein BapA and its type I secretion machinery, revealed that both AHLs and BDSF are required for full expression, suggesting that the two signalling systems operate in parallel. In accordance with this, we show that both AHLs and BDSF are required for biofilm formation and protease production.",
"introduction": "Introduction Many bacteria are capable of coordinating gene expression in a cell density-dependent manner, a phenomenon commonly referred to as quorum sensing (QS) [1] . QS systems rely on the production and release of small signal molecules into the environment. Bacteria respond to these signals when their concentration has reached a certain threshold (and thus the bacterial population has attained a critical density), upon which expression of target genes is activated or repressed. Among the various QS signal molecules identified to date, the two most thoroughly investigated classes are the N -acyl-homoserine lactones (AHLs), which are produced by many Gram-negative bacteria, and small peptides, which are produced by many Gram-positive species [2] , [3] . \n Burkholderia cenocepacia is a Gram-negative opportunistic pathogen belonging to the Burkholderia cepacia complex (Bcc), a group of 17 closely related bacterial species [4] . B. cenocepacia can cause airway infections in susceptible individuals, particularly in persons suffering from cystic fibrosis [5] . All members of the Bcc investigated so far utilize the AHL-dependent CepIR QS system [6] . CepI was shown to catalyze the synthesis of N -octanoyl homoserine lactone (C8-HSL) along with minor amounts of N -hexanoyl homoserine lactone (C6-HSL) [7] . At quorate population densities C8-HSL binds to its cognate transcriptional regulator CepR and in the AHL-bound form CepR binds to specific DNA sequences (so-called cep boxes) in the promoter region of target genes, thereby inducing or repressing gene expression [8] . Previous work has shown that the CepIR system regulates multiple functions, including virulence, biofilm formation, swarming motility, and the production of proteases, siderophores and antifungal compounds (reviewed in [9] ). The CepR regulons of two B. cenocepacia , K56-2 and H111, have previously been determined using functional genomics approaches [8] , [10] , [11] . These investigations not only identified many genes encoding virulence factors [12] but has also shown that in strain H111 AHL-dependent expression of a large surface protein ( bapA , BCAM2143) is critical for biofilm formation on abiotic surfaces [11] . Recent work has identified an additional QS system in B. cenocepacia that relies on BDSF ( B urkholderia \n d iffusible s ignal f actor, cis -2-dodecenoic acid), which belongs to a rapidly growing family of fatty acid signal molecules [13] , [14] . The biosynthesis of BDSF is driven by the product of rpfF \n Bc (BCAM0581), an enoyl CoA hydratase [15] . RpfF Bc is the first protein described to possess both dehydratase and thioesterase activity, which enables the direct conversion of the acyl carrier protein thioester of 3-hydroxydodecanoic acid into cis -2-dodecenoic acid [15] . More recently it has been demonstrated that the gene adjacent to rpfF \n Bc encodes the BDSF receptor protein RpfR, which contains PAS-GGDEF-EAL domains [16] . It has been shown that upon binding of BDSF to RpfR the c-di-GMP phosphodiesterase activity of the protein is stimulated and as a consequence the intracellular c-di-GMP level is lowered. Hence, RpfR is the first example of a c-di-GMP metabolic enzyme that is directly activated by a cell-cell communication signal [16] . Disruption of either rpfR or rpfF Bc was shown to result in reduced motility, impaired biofilm formation, lowered proteolytic activity, and attenuated virulence [16] . All these phenotypes are also known to be AHL-regulated and we were therefore interested to investigate whether the two regulatory circuits regulate the same set of genes and whether they are interconnected or operate independently of each other. In this study the BDSF stimulon of B. cenocepacia H111 was defined both at the transcript and protein level using RNA-Seq and shotgun proteomics. To determine the overlap of the AHL- and BDSF-dependent QS systems we compared the BDSF stimulon to the previously published CepR regulon [11] . In addition, we constructed a cepI rpfF \n Bc double mutant and used it to assess the influence of the two signal molecules individually and in combination on biofilm formation and the production of proteases and on transcription of QS-regulated target genes. Our data demonstrate that, in spite of the observed decrease in AHL production in the rpfF \n Bc mutant, the two QS systems regulate the tested phenotypes and genes independently, suggesting that they are not hierarchically arranged but operate in parallel under the experimental conditions used in this study.",
"discussion": "Discussion Our combined RNA-Seq and proteome analysis revealed that the set of genes regulated by BDSF in B. cenocepacia H111 shows a substantial overlap with the set of genes recently shown to be CepR-regulated ( Figure S1 ). However, we also identified some genes that were almost exclusively regulated by one of the two QS systems. For example, the gene aidA , which encodes a protein required for killing of the nematode Caenorhabditis elegans \n [22] , is stringently regulated by C8-HSL (>100-fold at the transcript level, Table 3 , Figure 2C and D ) whereas the effect of BDSF is marginal (4-fold to 6-fold at the transcript level, Tables 2 and 3 , Figure 2C and D ). It is important to note that the aidA promoter region contains a cep box, i.e. a CepR binding site, which is required for AHL-dependent transcriptional activation of this gene [22] . Likewise, we observed that genes containing a bona fide cep box in their upstream regions are more strongly affected by the CepI/R than the RpfR/F system ( Table S3 ). At the other extreme, expression of the lectin BclB, which is encoded by the last gene of the bclACB operon [11] , in the cepI rpfF Bc double mutant was found to be strongly dependent on BDSF while C8-HSL showed little effect ( Figure 2D , Table 2 ). We have previously shown that in a cepI mutant strain transcription of the bclACB operon is approximately 6-fold down-regulated relative to the wild type and that this defect is reversed by the addition of C8-HSL to the medium [11] . The evidence presented in this study suggests that AHL-dependent regulation of the bclACB operon only occurs when the BDSF-dependent QS system is intact ( Figure 2D , Table 3 ). In summary, these data suggests that the CepIR system synergistically enhances BDSF-dependent activation of bclACB expression. The underlying molecular mechanism of how the two QS systems interact in the expression of the lectin operon remains to be elucidated. It has not escaped our attention that the rpfF Bc mutant produces significantly reduced amounts of AHL signal molecules, most likely due to lowered transcription of cepI ( Table 3 , Figure S2 ). This result suggests a hierarchical arrangement of the two QS systems with the RpfF/RpfR system being on top of the CepI/CepR system. However, several of our results do not support this conclusion. Expression of some of the well-characterized AHL-regulated genes, including aidA , was only marginally affected in an rpfF Bc mutant background. In the case of bapA , which was shown to be regulated by both systems, addition of AHLs to the BDSF-deficient mutant did not restore expression of this gene to the level of the wild type, which would have been expected if the AHL-dependent circuitry operated downstream of the BDSF system. Likewise, biofilm formation and proteolytic activity was dependent on both signal molecules ( Figure 4 ) and could not be rescued to wild type levels when the BDSF mutant was grown in the presence of AHLs ( Figures S4 and S5 ). Only in the case of proteolytic activity a partial complementation was observed in the presence of AHLs ( Figure S5 ) and this effect may be attributed to the lowered level of the signal molecule. Extracellular proteolytic activity of B. cenocepacia is mainly conferred by two metalloproteases, ZmpA and ZmpB [25] , [26] . The expression of zmpA in a rpfF Bc mutant of B. cenocepacia J2315 was previously shown to be restored when the medium was supplemented with either AHLs or BSDF [27] , which may in part explain the slight increase in proteolytic activity when the cepI rpfF Bc double mutant was grown in the presence of C8-HSL. Importantly, it has been demonstrated that the ZmpB protease has greater activity against casein [26] and thus our measurements mainly reflect the activity of this protease, the expression of which appears to be dependent on both QS systems ( Table 2 ). In conclusion, our data suggest that the reduced AHL levels of the BDSF-deficient mutant are not, at least not solely, responsible for the observed phenotypic defects. It is noteworthy that the amount of C8-HSL produced by different B. cenocepacia strains varies dramatically, with concentrations ranging from 1 nM to 0.2 µM [7] . The CF isolate used in this study, strain H111 [7] , [28] produces very high levels (0.2 µM) of C8-HSL. Although the BDSF mutant produces less (50%) AHLs relative to the H111 wild type it still produces much higher amounts of the signal molecule than many other B. cenocepacia wild type strains, including the frequently studied strains K56-2 and J2315 [7] . We therefore cannot exclude the possibility that in strains producing very low amounts of C8-HSL inactivation of rpfF Bc has a more pronounced effect on expression of AHL-regulated functions. In a recent study evidence was presented that inactivation of rpfF Bc in B. cenocepacia J2315 also resulted in a lowered AHL level, albeit this reduction was found to be insignificant when the AHL concentration was normalized against the cell density [27] .Given that the amount of AHLs produced by strain J2315 is very low, a further reduction is difficult to quantify with standard methods and this may be the reason for the statistically insignificant results. In addition, the difference between the studies may be due to different growth conditions (Anwar minimal medium versus LB) or the fact that B. cenocepacia J2315 harbors an additional QS system [29] . In conclusion, our data support a model in which the two QS systems operate in parallel to control specific as well as overlapping sets of genes ( Figure 5 ). This model is also in accordance with the finding that the AHL and BDSF stimulons do not completely overlap and some genes are almost exclusively regulated by just one of the QS systems. It has recently been shown that binding of BDSF to its cognate receptor RpfR activates the c-di-GMP phosphodiesterase activity of this protein, which leads to a lowered intracellular c-di-GMP level [16] . At present it is unknown how this change in c-di-GMP level affects transcription of target genes. 10.1371/journal.pone.0049966.g005 Figure 5 Schematic presentation of the two B. cenocepacia H111 QS circuitries. The CepI/CepR system consists of the AHL synthase CepI directing the synthesis of C8-HSL, and of the transcriptional regulator CepR. The RpfF/RpfR system consists of RpfF which directs the synthesis of BDSF, and of its cognate receptor RpfR. Upon binding of BDSF to RpfR the c-di-GMP phosphodiesterase activity of the protein is stimulated and as a consequence the intracellular c-di-GMP level is lowered. The two QS systems operate in parallel to control specific as well as overlapping sets of genes. Our working model assumes an unknown c-di-GMP receptor protein × that stimulates transcription of target genes. Alternatively, the two QS cascades converge and control the expression or the activity status of an unknown common regulator Y, which in turn regulates expression of target genes. C-di-GMP has a negative regulatory effect on AHL levels via an unknown mechanism (depicted by the dashed line). In our working model we assume an unknown c-di-GMP receptor protein × that activates transcription of target genes either directly or via a regulatory cascade. Given that most genes that were found to be regulated by both QS systems are not directly regulated by CepR/C8-HSL, it is also possible that the two QS cascades converge and control the expression or the activity status of an unknown common regulator Y, which in turn regulates expression of target genes ( Figure 5 ). Work currently under way aims at distinguishing between the two possibilities and at identifying the c-di-GMP effector."
} | 3,363 |
29522743 | null | s2 | 2,820 | {
"abstract": "The composite members of the microbiota face a range of selective pressures and must adapt to persist in the host. We highlight recent work characterizing the evolution and transfer of genetic information across nested scales of host-associated microbiota, which enable resilience to biotic and abiotic perturbations. At the strain level, we consider the preservation and diversification of adaptive information in progeny lineages. At the community level, we consider genetic exchange between distinct microbes in the ecosystem. Finally, we frame microbiomes as open systems subject to acquisition of novel information from foreign ecosystems through invasion by outsider microbes."
} | 170 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.