pmid stringlengths 8 8 | pmcid stringlengths 8 11 ⌀ | source stringclasses 2
values | rank int64 1 9.78k | sections unknown | tokens int64 3 46.7k |
|---|---|---|---|---|---|
39775870 | PMC11774123 | pmc | 2,267 | {
"abstract": "Abstract Land use and agricultural soil management affect soil fungal communities that ultimately influence soil health. Subsoils harbor nutrient reservoir for plants and can play a significant role in plant growth and soil carbon sequestration. Typically, microbial analyses are restricted to topsoil (0–30 cm) leaving subsoil fungal communities underexplored. To address this knowledge gap, we analyzed fungal communities in the vertical profile of four boreal soil treatments: long-term (24 years) organic and conventional crop rotation, meadow, and forest. Internal transcribed spacer (ITS2) amplicon sequencing revealed soil-layer-specific land use or agricultural soil management effects on fungal communities down to the deepest measured soil layer (40–80 cm). Compared to other treatments, higher proportion of symbiotrophs, saprotrophs, and pathotrophs + plant pathogens were found in forest, meadow and crop rotations, respectively. The proportion of arbuscular mycorrhizal fungi was higher in deeper (>20 cm) soil than in topsoil. Forest soil below 20 cm was dominated by fungal functional groups with proposed interactions with plants or other soil biota, whether symbiotrophic or pathotrophic. Ferrous oxide was an important factor shaping fungal communities throughout the vertical profile of meadow and cropping systems. Our results emphasize the importance of including subsoil in microbial community analyses in differently managed soils.",
"conclusion": "Conclusions Our experimental set-up made it possible to study the long-term impacts of land use and soil management intensity on fungal communities. We showed that the effects of land use and soil management intensity on fungal communities persisted throughout the soil vertical profile down to 40–80 cm. In accordance with our hypothesis, the less intensively managed meadow was more associated with the potentially beneficial fungal groups than the more intensively managed organic and conventional cropping systems by having the highest AMF richness and saprotroph proportion. However, the management intensity differences between organically and conventionally managed soils were not reflected in significant differences in the potentially beneficial fungal groups. Organic and conventional treatments were distinguished by having the highest pathotroph richness and pathotroph and plant pathogen proportion, and forest by having the highest symbiotroph proportion. Similarly, as in forest but on a smaller scale, the mycorrhizal mode of requiring nutrients and energy became proportionally more important in deeper soil layers of meadow, organic, and conventional treatments indicating that subsoil nutrient reservoir could potentially be better utilized and the environmental impacts of farming reduced by optimizing agricultural soil management toward AMF favoring practices. We showed several fungal taxa to be proportionally more prominent in certain soil layers. Topsoil-associated taxa in meadow, organic, and conventional treatments, included the fungal classes Dothideomycetes, Sordariomycetes, and Tremellomycetes. Subsoil-associated taxa included the fungal classes Mortierellomycetes in all treatments and Leotiomycetes in meadow, organic, and conventional treatments. This study showed that the only soil property consistently significantly related to fungal communities throughout the soil vertical profile and with strong positive correlation with AMF richness was Fe-ox, which should be further studied. Additionally, this study indicated that sampling depth should be extended at least to 30 cm deep to better describe the diversity of AMF. Vertical profiles of agricultural soils that deploy more extensive regenerative agricultural practices, such as cover-cropping with deep-rooting plants and minimal-tillage, should in the future be explored for their microbial communities to better understand soil management effects in subsoils.",
"introduction": "Introduction Fungi play a key role in agricultural soil health by affecting soil structure through aggregation, nutrient cycling, plant health, and soil organic carbon (SOC) formation and decomposition (Powell and Rillig 2018 , Toju et al. 2018 , Bhattacharyya et al. 2022 , Xiong and Lu 2022 ). High fungal diversity has been linked to improved soil multifunctionality and crop production (Wagg et al. 2011 , Delgado-Baquerizo et al. 2016 ), and fungal abundance and activity to increased carbon sequestration into soil (Kallenbach et al. 2016 , Bhattacharyya et al. 2022 , Hannula and Morriën 2022 ). Soil management practices that promote diverse fungal communities in agricultural soil can potentially increase stable soil carbon formation and ultimately crop yield (Hannula and Morriën 2022 ), contributing to the UN sustainability goals for sustainable agriculture (United Nations 2015 ). Fungi can be divided into functional groups according to their main mode of acquiring energy and nutrients (Nguyen et al. 2016 ). It has been proposed that, rather than an overall fungal community, certain fungal functional groups would better describe soil ecosystem functioning in agricultural soils (Ferris and Tuomisto 2015 ). Symbiotrophic fungi, especially arbuscular mycorrhizal fungi (AMF), which form interactions with plants and contribute to plant nutrient and water uptake (Smith and Read 2008 ), and saprotrophic fungi, which promote nutrient cycling in soil by decomposing organic material (Deacon et al. 2006 ), are potentially beneficial fungal groups for crop production. AMF have been linked to increased plant phosphorus uptake and ultimately higher plant productivity (van der Heijden et al. 1998 , Fall et al. 2022 ) as well as to pathogen suppression in agricultural soils (Fall et al. 2022 , Hannula and Morriën 2022 ). AMF can increase SOC by promoting plant photosynthate translocation into the soil matrix and by forming hyphal biomass (Jeewani et al. 2020 , Parihar et al. 2020 ). Similarly, saprotrophs have been linked to higher soil fertility (Ning et al. 2021 ) and plant pathogen suppression (van der Wal et al. 2013 ). Saprotrophic fungi have been shown to increase SOC in forest ecosystems (Klink et al. 2022 ), and a similar effect in arable soils was recently proposed (Hannula and Morriën 2022 ). Investigating AMF and saprotrophs, as well as other fungal functional groups such as plant pathogens, that can have harmful effects on crop plant production (Corredor-Moreno and Saunders 2020 ), brings important knowledge on the health and functionality of agricultural soils. Agricultural management intensity, which is a measure of the fertilizer and biocide use, irrigation, and mechanization level (Foley et al. 2011 ), is known to affect soil microbial communities. For instance, high management intensity can decrease fungal biomass and the abundance of both AMF and saprotrophic fungi, likely due to their sensitivity to soil disturbance (Strickland and Rousk 2010 , Thiele-Bruhn et al. 2012 , Hydbom et al. 2017 , Banerjee et al. 2019 ). Long-term organic management, representing a lowered management intensity in which chemical biocides and chemical fertilizers are not used, can promote fungal richness and abundance over more intensive conventional management (Martínez-García et al. 2018 , Peltoniemi et al. 2021 ). Similarly, lowered management intensity in extensively managed grasslands or lands with permanent, predominantly herbaceous plant cover has been shown to promote fungal richness over arable lands (Banerjee et al. 2024 ). Yet, more knowledge of the effects of soil management on fungal communities across the soil vertical profile is needed as studies have mostly focused on topsoil, typically reaching to 20–30 cm depth at maximum (Henneron et al. 2022 ). Agricultural soil carbon is decreasing globally (Lal et al. 2004 ). In Finland, agricultural mineral soils lose carbon at a yearly rate of 0.4% (Heikkinen et al. 2013 ). Globally, several agricultural management practices have been shown to promote SOC, including diverse crop rotation, organic amendments, no-tillage systems in some climate and soil type conditions, and organic farming, although the latter was recently revised to need additional actions, such as cover cropping and enhanced plant residue recycling (Francaviglia et al. 2017 , Yang et al. 2019 , Ogle et al. 2019 , Zhang et al. 2021 , Gaudaré et al. 2023 ). Soil management practices enabling the formation of extensive root systems and deep rooting plants can potentially increase SOC not only in topsoil but also in subsoil layers (below 30 cm), where half of the soil carbon of agricultural fields is stored (Balesdent et al. 2018 , Hirte et al. 2021 , Nguyen 2009 , Paustian et al. 2016 ). In addition to roots, fungal hyphae contribute to the translocation of SOC deeper in the soil (Witzgall et al. 2021 ). The important role of deep soil layers in SOC sequestration (Button et al. 2022 ) and fungi in SOC dynamics further emphasizes the need to investigate fungal communities in the soil vertical profile to better understand the fate of SOC in agricultural soils. Here we used amplicon sequencing of ribosomal RNA gene internal transcribed spacer (ITS2) to study fungal communities in the vertical profile of four soil treatments, organic and conventional cropping systems, unmanaged meadow, and forest, down to 80 cm deep after 24 years of field experiment. The comparison of organic and conventional treatments enables the assessment of the long-term combined effects of fertilizer type and herbicide usage on fungal communities. Organic treatment represents a less intensively managed system compared to conventional treatment, whereas meadow treatment represents the least intensive, natural-grassland-like management, creating a management intensity gradient from least intense to most intense: meadow–organic–conventional. Forest is included as a reference and represents the land use type that prevailed in the experiment area before conversion to agricultural system (Salonen et al. 2023 ), providing insight on how the transition into agricultural or meadow land use changes the fungal community over time. Our overall aim was to study how depth within the land use types (forest, meadow, and organic and conventional cropping systems) and soil management intensity within meadow–organic–conventional soil management intensity gradient influence fungal communities. In addition, we aimed to address which soil properties are the drivers of fungal community differences within the soil vertical profile. We hypothesize that lower soil management intensity in meadow and organic soils increase fungal diversity and promote the potentially beneficial AMF and saprotroph communities compared to more intensively managed conventional soil.",
"discussion": "Discussion In a recent meta-analysis, it was shown that in the deeper soil layers, there are on average 47% of soil organic C stocks of agricultural fields (Balesdent et al. 2018 ). Similarly in the forest soils, it has been shown that the total soil C stock under 20 cm may be up to 50% of the total (Jobbágy and Jackson 2000 ) and up to 75% of SOM can be found in subsoil (B and C horizons) (Rumpel et al. 2002 ). Considering deeper soil layers as reservoirs for C, different agricultural management practices can have a significant role both as enhancing fresh C input into deeper layers (Lessmann et al. 2022 , Gaudaré et al. 2023 ), as well as modifying the microbial communities responsible for SOC decomposition and plant nutrient uptake (Morugan-Coronado et al. 2022 ). However, we still lack a comprehensive view of how land use or soil management influences microbial communities in the soil vertical profile and how this ultimately affects the fate of SOC. Depth together with land use and agricultural soil management affected fungal community composition The analysis of the vertical soil profile of the four treatments showed that fungal communities were affected by soil layer and treatment and the treatment effect varied between the studied five soil layers. Overall, we found soil layer to have a bigger effect on fungal community differences compared to treatment. Fungal community composition and diversity have previously been shown to be influenced by depth in cropping systems (Schlatter et al. 2018 , Yin et al. 2021 ) and forest (Baldrian et al. 2012 ). Similarly, there are numerous studies showing how agricultural soil management intensity shapes fungal communities in topsoil (Sun et al. 2016 , Gottshall et al. 2017 , Vahter et al. 2022 , Wu et al. 2022 ). However, previously the comparison of organic and conventional treatment effects on the fungal community has been done down to 30 cm, but we lack studies where below 30 cm layers are analysed (Epp Schmidt et al. 2022 ). Here, we show that the treatment effect between organic and conventional cropping systems can be seen down to the deepest measured soil layer 40–80 cm ( Table S4B–F ). Conventional and organic plots had the same 5-year crop rotation and three different crops growing during the sampling year, indicating that agricultural management affects fungal communities regardless of the crop. Fungal richness was not negatively associated with soil management intensity In line with a previous study by Schlatter et al. ( 2018 ), our results on fungal richness showed a consistent decrease in relation to depth in soil layers between 10–80 cm in all treatments. Fungal richness in meadow, organic, and conventional treatments differed in topsoil (0–10 cm) where organic and conventional had more diverse fungal community compared to meadow and in the deepest soil layer (40–80 cm) where organic treatment had more diverse fungal community compared to conventional. Interestingly, contrary to what we hypothesized and what has been found in multiple previous studies (Martínez-García et al. 2018 , Peltoniemi et al. 2021 , Banerjee et al. 2024 ), low management intensity did not promote higher fungal richness in topsoil. This, however, follows the somewhat surprising fungal diversity pattern found in a Europe-wide study across land-use intensity gradients (woodland–grassland–cropland), where higher land use intensity correlated with higher fungal diversity (Labouyrie et al. 2023 ). Similarly, in grasslands, the intensification of land management practices has been found to have either neutral or positive effects on belowground fungal diversity (Allan et al. 2014 , Gossner et al. 2016 ). In diverse environments such as the meadow, organic, and conventional soils in our study, the common understanding in ecology that a higher species richness contributes to higher ecosystem functioning (Loreau et al. 2001 ) has been disputed (Nielsen et al. 2011 ). Ecosystem functions have rather been linked to succession of fungal communities than to high OTU richness (Hoppe et al. 2016 ). We do not have data for temporal succession in our soils, but we know that fungal communities were more specialized vertically in meadow ( Table S4G ), probably due to higher litter input and the lack of interruption by periodic ploughing. This spatial specialization in meadow could possibly lower fungal diversity in individual soil layers. In addition, the lower topsoil pH in the 0–10 cm soil layer of meadow compared to organic treatment and marginally compared to conventional management may have attributed to the lower fungal diversity in meadow (Zheng et al. 2019 ). The over two-fold higher DOC in the 0–10 cm soil layer of meadow, most probably caused by the high litter input, may also have lowered topsoil fungal richness in meadow similarly as in a previous study where higher arable soil DOC and lower fungal richness were found in straw mulch soil compared to soil without mulch (Huang et al. 2019 ). We did not find difference in fungal richness between organic and conventional in the first four soil layers (0–40 cm). Similarly, in a study with organically fertilized (pig manure) and chemically fertilized crop field, and in long-term organic and conventional cereal crop systems, no significant differences in fungal Shannon diversity (Suleiman et al. 2019 ) or OTU richness (Peltoniemi et al. 2021 ) between the management types were found, but rather in the fungal ITS2 copy numbers (Peltoniemi et al. 2021 ), indicating that management effect on fungal diversity could be more subtle compared to fungal abundance, which was not measured in this study. However, our study provides only a single time point view of fungal diversity which can change during the growing season and between years (Degrune et al. 2017 ). Considering our findings and the literature, the overall effect of management intensity on fungal richness remains somewhat unclear. Management intensity affected AMF richness below the surface soil Previously, it has been shown that rather than the overall fungal community, specialized microbial groups are linked to soil ecosystem functioning and may better describe the effects of land use or soil management intensity (Wang et al. 2022 ). Symbiotrophic fungi in general and specifically AMF can benefit plant productivity and soil fertility (van der Heijden et al. 1998 , Smith and Read 2008 et al. 2008 , Jeewani et al. 2020 , Parihar et al. 2020 , Fall et al. 2022 , Hannula and Morriën 2022 ). Although lower agricultural soil management intensity is shown to positively affect AMF (Hydbom et al. 2017 , Banerjee et al. 2019 ), we did not find a significant effect of treatment on AMF or symbiotroph proportion between meadow, organic, and conventional treatments. AMF richness, however, differed between the low-intensity meadow and the highest intensity conventional treatment in the 20–30 and 30–40 cm soil layers. Organic soil which represents a lowered management intensity fell between the intensity extremes and could not be statistically differentiated from either. The management intensity effect on AMF richness can be attributed to different management practices. For instance, AMF are shown to be negatively affected by fertilization overall (Hannula et al. 2021 , Luo et al. 2021 ), and the use of mineral fertilization over manure can further suppress AMF (Wang et al. 2018 ), which could explain higher AMF richness in unfertilized meadow compared to mineral-fertilized conventional treatment. The differences in root biomass between treatments which followed the management intensity gradient (higher root biomass in lower management intensity; Fig. S5 ; Table S8 ) and the lack of disturbance related to tillage operations in meadow may have promoted higher AMF richness in meadow (Hiiesalu et al. 2014 , Schmidt et al. 2019 ). Plant diversity was not measured from the treatment plots in the sampling year (2019), so we cannot fully assess the effect of plant diversity on fungal communities. However, plant richness and Shannon diversity were recorded 7 and 8 years before the experiment (in 2011 and 2012) ( Fig. S5 ) and showed no differences between meadow and the cropping systems (organic and conventional treatments) but higher plant richness in organic compared to conventional treatment in 2012 ( Fig. S5 ). Plant diversity has previously been positively linked to AMF diversity (Hiiesalu et al. 2014 ), indicating that high plant richness in organic treatment might partly explain why organic treatment did not differ from meadow in AMF richness whereas conventional treatment did. Higher plant richness in organic treatment is most probably a consequence of the lack of herbicide usage and is thus part of the management intensity effect. Organic and conventional treatment in this study already had a moderately diversified cropping system with 5-year rotation which included grass and crop mixtures (Salonen et al. 2023 ). However, decreasing management intensity by incorporating reduced tillage and increasing plant diversity by, for instance, cover-cropping, where noncommercial plants are grown together or after the main crop, could potentially further promote AMF richness in organic and conventional treatments (Thapa et al. 2021 ). Arbuscular mycorrhizal fungal communities were affected by treatment and depth, but no treatment-specific taxa were found We took a closer look at the AMF communities since the beneficial functions associated with AMF, such as induced nutrient uptake and protection against pathogens, can differ between AMF taxa (Sikes et al. 2010 ). We found AMF communities to be affected by treatment and depth but AMF taxa-specific differences between meadow, organic, and conventional treatments were not found. Based on patterns of fungal biomass allocation, AMF taxa can be grouped into rhizophilic guild, that have high biomass in roots and may protect host plants from pathogen colonization, edaphophilic guild, that have high extradical hyphae biomass and improve plant nutrient uptake (Weber et al. 2019 ), and ancestral guild, that produce low biomass both within and outside the root (Treseder et al. 2018 , Phillips et al. 2019 ). In our study, the rhizophilic AMF guild was most pronounced, followed by the ancestral AMF guild. High proportion of rhizophilic AMF guild indicates an improved protection over plant pathogens. Edaphophilic guild was the least represented AMF guild in the studied soils, although the only edaphophilic genus, Diversispora , was found in all treatments. The abundance of many edaphophilic AMF taxa, but not Diversispora , has been linked to a higher C-N-ratio than what was present in our soils (Treseder et al. 2018 , Fig. S4 ). Yet, the presence of a plant-nutrient-uptake improving AMF taxa such as Diversispora in organic and conventional treatment is an encouraging finding as it could benefit crop plants by scavenging large volume of soil, including deep soil, for nutrients. In addition to depth, fungal trophic modes were affected by land use and agricultural management In all soil treatments, the proportion of symbiotrophic fungi increased toward deeper soil layers, and in meadow, organic, and conventional treatment this was shown as an increase of AMF proportion in subsoil in comparison to topsoil. Since AMF benefit plant nutrient uptake (Smith and Smith 2011 ), and subsoil can harbour more than two-thirds of the nutrients in arable fields (Kautz et al. 2013 ), this subsoil association of AMF could indicate an important role of subsoil as a nutrient pool in the studied meadow, organic, and conventional treatments. Regarding forest treatment, our results support the previously proposed hypothesis that symbiotrophic mycorrhizal fungi in boreal forests are more competitive than saprotrophs in deeper layers where litter is more decomposed and C:N ratio is lower (Lindahl et al. 2007 , van der Wal et al. 2013 , Santalahti et al. 2016 , Carteron et al. 2021 ), as both the highest symbiotrophic proportion and the lowest C:N ratios coincided in the same deep forest soil layers (30–40 cm and 40–80 cm) (Fig. 3 ; Fig. S4 ). Our results further suggest that the direct fungal interactions with plants, whether symbiotrophic or pathotrophic, are emphasized in deep forest soil (40–80 cm), where symbiotroph and pathotroph-saprotroph fungi represented the majority (75% and 18%) of fungal functional community and pure saprotrophs only a marginal (2%) (Fig. 3C ). This indicates that the role of aboveground vegetation in shaping fungal communities in subsoil of boreal forest may be substantial. Pathotrophic fungi were affected by treatment and depth. Out of all fungal functional groups, pathotrophic fungi correlated most strongly and negatively with depth (Fig. 3 ; Table S5 ), which may be explained by lower host interactions due to lower plant input and the typically lower richness of protist and soil animals in deeper soil layers (Du et al. 2022 , Islam et al. 2022 ). Yet, contrary results were previously observed in a study with wheat-cropping system, where pathotrophic fungi were either unaffected or positively affected by depth (Schlatter et al. 2018 ). Among the pathotrophic fungi, plant pathogens were strongly influenced by treatment. We found organic and conventional treatments to increase plant pathogen proportion compared to forest and meadow ( Table S6 ). Our results do not agree with a previous study where plant pathogen richness and proportion were shown to increase according to SOC (Du et al. 2022 ). Rather, our results are in line with the plant pathogen-inducing effect of arable soils over grasslands (French et al. 2017 ). Saprotrophic fungi have been gaining attention as a potentially beneficial fungal group in agricultural soils contributing to nutrient cycling, soil fertility, plant pathogen suppression and SOC (Deacon et al. 2006 , van der Wal et al. 2013 , Ning et al. 2021 , Hannula and Morriën 2022 ). We found the proportion of saprotrophic fungi to be more associated with low-intensity meadow treatment than with the cropping systems, organic and conventional treatments. Although the decomposing function of saprotrophic fungi can increase soil respiration and loss of carbon from soil in some cases (Newsham et al. 2018 ), a positive link between saprotroph biomass and SOC is frequently observed in agricultural soil (Six et al. 2006 ). In our study, higher SOC (Salonen et al. 2023 ) and higher saprotroph proportion coincided in meadow ( Tables S6 and S8 ), further supporting the role of saprotrophs in SOC accrual. Fe-ox were positively related to fungal communities down to 40–80 cm soil layer with strong correlation with arbuscular mycorrhizal fungal richness Several soil properties contributed to fungal communities in meadow, organic and conventional treatment (Fig. 2 ; Tables 1 and 2 ). Fungal community differences (Bray–Curtis) were influenced by soil properties commonly observed in previous studies, C, N, DOC, C/N, P-tot, and pH (Francioli et al. 2016 , Khan et al. 2016 , Muneer et al. 2021 , Rousk et al. 2010 , 2011 , Tedersoo et al. 2014 , 2020 , Zheng et al. 2019 ), as well as by root biomass, P-org, and Fe-ox, and less by P-inorg and P-H 2 O. Our results confirm the less commonly reported role of root biomass along the soil vertical profile in shaping fungal communities as well as the positive correlation of root biomass with fungal and AMF richness (Broeckling et al. 2008 , Eisenhauer et al. 2017 , López-Angulo et al. 2020 ). Previously, the role of P in shaping fungal communities has been emphasized, especially in agricultural soil (Francioli et al. 2016 , He et al. 2016 , Wu et al. 2022 ). Here, we consistently found P-org out of the different P forms (P-org, P-tot, P-inorg and P-H 2 O) to best explain variations in fungal community differences and fungal richness. Additionally, P-org explained fungal community differences better than any other soil property when the whole soil profile was considered. Total and available P has been shown to correlate negatively with fungal diversity (Wu et al. 2022 ), and in general, soil P is believed to negatively affect fungal richness (Tedersoo et al. 2014 ). In contrast, we did not find a negative link between fungal richness and any P form measured, and only a weak negative link with P-inorg and AMF richness was observed. Although the negative effects of soil P on AMF richness and abundance are well documented (Abbott et al. 1984 , Camenzind et al. 2014 , Chen et al. 2014 , Jasper et al. 1979 , Mosse 1973 , Olsson et al. 1997 ), recently opposing effects of P in topsoil (negative) compared to subsoil (positive) were found (Luo et al. 2021 ), indicating the effects of P on AMF richness to vary within the soil vertical profile. This could explain why we did not find a negative link with most P forms and AMF richness when assessing the whole soil profile. However, our results do support the strong adverse role of different P forms on AMF proportion (Table 2 ) and suggest that AMF diversity and proportions may be differently affected by P in the soil vertical profile. The strong role of Fe-ox in fungal communities is not commonly reported, making it a novel and interesting finding (Brandt et al. 2024 , Jeewani et al. 2020 ). Fe-ox was the only soil property that associated with fungal communities consistently throughout the soil vertical profile (0–80 cm) and additionally correlated strongly with fungal and AMF richness (Tables 1 , 2 ). This is supported by a previous study, where AMF were reported to preferentially associate with iron oxide surfaces in rhizosphere soil (Whitman et al. 2018 ). Fe-ox is important in soil aggregate formation and is associated with SOC (Jeewani et al. 2020 , Pronk et al. 2011 , Salonen et al. 2023 ), which can at least partly explain the role of Fe-ox as both soil aggregates and SOC are known to shape fungal communities (Fan and Wu 2021 , Upton et al. 2019 ; Yang et al. 2019 ). Additionally, soil P availability is negatively affected by Fe-ox, as well as by Al-ox, which adsorb phosphate ions through ligand exchange reaction (Hingston et al. 1967 ). Thus, Fe-ox may have affected fungal communities by controlling the amount of available P. Further studies are needed to better understand the role and function of Fe-ox in shaping fungal, especially AMF, communities. Meadow treatment had the most distinct soil properties among meadow–organic–conventional soil management intensity gradient As such, pH influences fungal community structure (Hannula et al. 2021 , Tedersoo et al. 2020 ) and it was overall higher in the cropping systems (organic and conventional treatment) compared to meadow ( Table S8 ). Lower pH may also have led to higher Fe-ox content in meadow (Thompson et al. 2006 ). In addition to differences in pH and Fe-ox, meadow treatment was associated with higher C, N, C/N, DOC, Al-ox and root biomass in most soil layers, and higher P-org in the topsoil (0–10 cm) compared to the cropping systems, indicating their role in fungal community differences between meadow and conventional/organic treatment. Root biomass, C, N, P-org, and P-H2O were the only soil properties that significantly differed between organic and conventional treatments at least in some of the soil layers, indicating a link between these soil properties and variations in the fungal communities ( Table S8 ). In deep soil, the role of root biomass may have been important as it was the only significantly different soil property between organic and conventional treatments in 30–40 cm and 40–80 cm soil layers. As Fe-ox differed significantly only between meadow and the cropping systems (organic and conventional), its role in the fungal community differences between organic and conventional treatments remains unclear. However, soil layer and treatment alone better explained the observed fungal community differences than all the measured soil properties together (PERMANOVA; R2 = 0.41 vs. R2 = 0.34). This indicates that soil management and depth may influence fungal communities beyond these commonly measured soil properties."
} | 7,774 |
36869057 | PMC9984465 | pmc | 2,268 | {
"abstract": "Chronically high levels of inorganic nutrients have been documented in Florida’s coral reefs and are linked to increased prevalence and severity of coral bleaching and disease. Naturally disease-resistant genotypes of the staghorn coral Acropora cervicornis are rare, and it is unknown whether prolonged exposure to acute or chronic high nutrient levels will reduce the disease tolerance of these genotypes. Recently, the relative abundance of the bacterial genus Aquarickettsia was identified as a significant indicator of disease susceptibility in A. cervicornis , and the abundance of this bacterial species was previously found to increase under chronic and acute nutrient enrichment. We therefore examined the impact of common constituents of nutrient pollution (phosphate, nitrate, and ammonium) on microbial community structure in a disease-resistant genotype with naturally low abundances of Aquarickettsia. We found that although this putative parasite responded positively to nutrient enrichment in a disease-resistant host, relative abundances remained low (< 0.5%). Further, while microbial diversity was not altered significantly after 3 weeks of nutrient enrichment, 6 weeks of enrichment was sufficient to shift microbiome diversity and composition. Coral growth rates were also reduced by 6 weeks of nitrate treatment compared to untreated conditions. Together these data suggest that the microbiomes of disease-resistant A. cervicornis may be initially resistant to shifts in microbial community structure, but succumb to compositional and diversity alterations after more sustained environmental pressure. As the maintenance of disease-resistant genotypes is critical for coral population management and restoration, a complete understanding of how these genotypes respond to environmental stressors is necessary to predict their longevity.",
"introduction": "Introduction Populations of the Caribbean staghorn coral Acropora cervicornis have markedly declined since the 1980s due to a combination of stressors including infectious disease, poor water quality, high sea surface temperatures, and overfishing 1 . The decline of Acropora cervicornis has contributed to a reduction in shallow Caribbean coral reef cover from about 55% to less than 10% over the last 40 years 2 , 3 . To improve recovery of this species, asexual propagation and outplanting of A. cervicornis to in situ reef environments is employed in Florida’s coral reefs 4 , 5 and throughout the greater Caribbean 6 , 7 . Outplanting of A. cervicornis demonstrably increases local coral coverage 8 , and nursery-grown corals may reach sexual maturity within 2 years after outplanting 4 , 8 . However, long-term survival of restored corals is predicted to be low if environmental stressors are not alleviated 9 . Outplanting efforts that do not consider integrating coral resistance and resilience to future stressors will therefore produce communities with low likelihood of future survival 10 , 11 . A significant increase in nutrient concentrations due to fertilizer, top soil, and sewage runoff has been documented in the Florida Keys over the past three decades, with a peak in nutrient levels occurring around 2014 12 , 13 . The additive negative effect of nutrient enrichment and increased thermal stress leads to increased coral mortality 14 – 17 . A recent study indicated that acroporid species may be particularly susceptible to nutrient enrichment: nutrient-exposed A. cervicornis exhibited population mortality of 84–100% when exposed to a subsequent thermal stress event, while thermal responses of other coral species were not altered by nutrient exposure 14 . These impacts of nutrient exposure on coral health may be regulated by the coral holobiont: nitrogen enrichment leads to phosphate starvation in the symbiotic algae Symbiodiniaceae 18 , and disrupts nutrient cycling in coral-associated bacterial communities 19 . Nutrient enrichment can also increase microbial opportunism, decreasing microbial community richness and evenness, which may be tied to increased stressor-related mortality 16 . The impacts of nutrient enrichment on coral health, however, are not always negative and depend on nutrient source 20 , 21 , exposure time 22 , and ratios of available nitrogen to phosphorus 18 , 23 . Across numerous studies, the coral microbiome has been identified as an important factor in both resistance to and recovery from environmental stressors 24 – 26 . It has been proposed that a highly diverse coral microbiome may provide a greater arsenal of antimicrobial defenses 27 , 28 and that an evenly distributed, highly diverse microbiome may occlude niche space that could otherwise be filled by opportunistic bacterial species 28 – 30 . We previously found that microbiomes of A. cervicornis genotypes known to be disease-susceptible were characterized by an overwhelming dominance of the putative bacterial parasite “ Candidatus aquarickettsia rohweri ” (hereafter, A. rohweri ), while disease-resistant genotypes were characterized by low relative abundance of A. rohweri and a more even and diverse microbiome 31 – 33 . We proposed that the high microbiome diversity of disease-resistant genotypes may provide greater antimicrobial defenses, and furthermore, that a coral host with low A. rohweri may have better immune capacity to combat future infections. In contrast, dominance of a bacterial parasite such as that observed in disease-susceptible genotypes of A. cervicornis may pose a significant nutritional burden on the coral and lead to reduced or altered immune capacity. The dominance of Aquarickettsia in microbiomes of A. cervicornis has been observed across many studies and genotypes 34 – 37 and a high abundance of its members is associated with increased disease prevalence, reduced coral growth, and increased tissue loss 16 , 38 . Despite the apparent disadvantages to high parasite abundance, A. cervicornis genotypes lacking this parasite are rare, suggesting that few genotypes are naturally resistant to Aquarickettsia infection 31 , 32 . As Aquarickettsia abundance responds positively to nutrient enrichment 33 , 38 , further exploration of the stability of identified genotypic associations of this parasitic genus is necessary considering the comparatively eutrophic conditions of Florida’s coral reef compared to other regions 12 . Restoration efforts targeted towards coral genotypes with the highest stress resilience and long-term survival rates may increase the overall success of Caribbean coral restoration. Notably, disease-resistant genotypes are exceedingly rare in restoration broodstock, with only two genotypes from Mote Marine Laboratory’s collections in the Lower Keys exhibiting sustained disease resistance 39 and comparably few resistant genets found in other Florida Keys nurseries 40 . Although disease-resistant genets of A. cervicornis possess diverse microbiomes with low abundances of the presumed parasite Aquarickettsia , a better understanding of the microbiome and phenotypic stability of these genets under shifting environmental regimes is necessary. We exposed fragments of the disease-resistant A. cervicornis genotype ML-7 to elevated nutrient levels to assess the individual and combined effects of ammonium, phosphate, and nitrate on A. cervicornis holobiont health, Aquarickettsia abundances, and changes in coral microbiome dynamics.",
"discussion": "Discussion The role of microbial diversity in the face of environmental stressors The coral microbiome likely plays a significant role in both resistance to and recovery from stressors. As bacterial symbionts of coral are proposed to perform numerous functions including pathogen defense, nitrogen and sulfur cycling, and nutrient translocation to the host, a diverse microbiome containing organisms fulfilling each of these services may allow for optimal holobiont response to environmental shifts 27 , 28 . While microbiome responses to short-term stressors (such as thermal stress) may be reversed with the removal of the stressor 45 – 47 , cumulative and long-term stressors (including nutrient enrichment) can shift the coral microbiome from mutualistic to pathogenic 24 , 48 . In this study, microbiome community structure in genotype ML-7 of the coral Acropora cervicornis (previously characterized as disease-resistant 39 ) was not altered by short-term exposure (i.e., 3 weeks) to significantly elevated levels of nutrient enrichment. There were no significant changes in diversity and few taxa responded to 3 weeks of nutrient enrichment, suggesting that microbial communities of genotype ML-7 samples were initially resistant to nutrient exposure and did not experience community shifts. Chronic (6 weeks) exposure to nutrients, however, led to changes in microbiome diversity and variability, with a significant decrease in overall species richness and evenness. We additionally found that although genotype ML-7 harbored the putative coral parasite, genus Aquarickettsia 49 , relative abundances remained low or undetectable in many samples even after prolonged nutrient enrichment, which increased abundances of this genus in other studies 16 , 33 , 38 . Impacts of nitrate on corals in this study may have been compounded by high ambient levels of nitrate in untreated aquarium source water in comparison to Looe Key Reef (Supp. Fig. 1 ). It is therefore surprising that this treatment had minimal impact on microbial community structure compared to other treatments. Nitrate enrichment did, however, demonstrably impact growth rates, with 6 weeks of nitrate enrichment significantly reducing growth rates compared to untreated corals while no other group responded significantly. The reduction of growth with nitrate enrichment is consistent with previous findings that nitrogen enrichment leads to an overgrowth of the algal symbiont Symbiodiniaceae , which subsequently fixes carbon so rapidly that coral calcification rates are limited 20 , 50 . In a previous study, we found that phosphate, rather than nitrate, reduced growth rates of genotype ML-50, which we hypothesized was linked to the positive response of the putative parasite Aquarickettsia to phosphate enrichment 33 . Potential roles of dominant bacterial taxa from the order Campylobacterales Microbiomes of genotype ML-7 were dominated by two sequence variants classified as Campylobacterales, and both of these ASVs were also found in samples of Panamanian A. cervicornis 41 . The dominance of two taxa classified as Campylobacterales in samples from both Florida and Panama merit further study of the relationship of this order with Acroporid species. Furthermore, a sequence variant identical to ASV1 was found in high abundance in genotype ML-7 as well as in several other genotypes sourced from throughout the Florida Keys sampled in 2019 in Ref. 32 . These unclassified Campylobacterales were not found in high abundance in samples of genotype ML-7 taken from Mote’s in situ nursery in 2015 31 , which suggests that the dominance of these taxa may be a recent development. The order Campylobacterales is primarily comprised of microaerobic chemolithotrophic species, including many that oxidize sulfur 51 , 52 . Sequence variants from Campylobacterales have been found in association with Stony Coral Tissue Loss Disease (SCTLD) and may thrive on decaying and anoxic tissue 53 . These SCTLD-associated taxa, however, were from the genus Sulfurimonas and the family Arcobacteraceae 53 , while the taxa from our study (Unclassified Campylobacterales ASVs 1 and 9) were not closely related to any named group in Campylobacterales. It is therefore difficult to predict the function of the unclassified Campylobacterales harbored by genotype ML-7, though the absence of tissue necrosis suggests that these taxa are more likely inhabiting a different ecological niche than those associated with SCTLD. The methodology of the present study included the sampling of coral skeleton, tissue, and mucus simultaneously. As coral skeleton has been previously found to harbor anaerobic bacteria 54 , 55 , it is possible that bacterial biomass in this genotype is relatively low in the mucus and tissue layers, such that overall microbial community composition is dominated by microaerobic bacterial living in skeletal niches. Anaerobic sulfur-oxidizing species have been found in high abundance in skeletons of other coral genera and have been proposed to play a beneficial role by detoxifying sulfide and supplying their host with organic nutrients 55 , 56 . Further study of genotypes dominated by these unclassified Campylobacterales should aim to identify metabolic pathways enriched in these samples and determine whether these taxa are most dominant in skeletal compartments. Chronic ammonium exposure led to decreased microbiome diversity in genotype ML-7 corals Ammonium is often naturally-derived in reef systems from fish excretion and enhances coral growth 20 , 57 . In this study, ammonium treatment did not significantly increase coral growth compared to untreated conditions, though mean linear extension at 6 weeks was higher for corals treated with ammonium than any other treatment. Shannon diversity declined in response to ammonium and combined (N, P, and A) treatment but no single taxon decreased significantly in ammonium-treated samples. The loss of minor taxa that varied by individual fragment may therefore lead to declines in alpha diversity with ammonium treatment. Indeed, mean per-sample species richness of ammonium-treated samples halved between 0 and 6 weeks of exposure, decreasing from 505 taxa to 250. Ammonium treatment significantly altered beta diversity compared to untreated samples by 6 weeks of nutrient enrichment, and dispersion of ammonia-treated samples decreased significantly over time while no other treatment led to reduced dispersion. The reduced diversity and dispersion of ammonium-treated samples at the conclusion of the experiment may reflect the loss of rare taxa as a result of increased competition from dominant taxa (Unclassified Campylobacterales ASV1, Unclassified Helicobacteraceae ASV2, Cysteiniphilum ). Interestingly, ammonium treatment also induced the proliferation of family P30B-42, a member of Myxococcales, which has been suggested to play a commensal or beneficial role based on its relatively high abundances in disease-resistant genotypes of Acropora 34 . Notably, significant changes in overall microbial community structure and diversity were not significant until 6 weeks of ammonium exposure, suggesting that microbiomes of genotype ML-7 may withstand short-term environmental shifts without alteration in structure, but long-term exposure to elevated nutrients may stimulate the growth of certain microbial taxa and/or the loss of rare taxa. Individual taxa responded to nutrient enrichment without significantly altering microbiome structure While ammonium enrichment alone appeared to reduce rare microbial taxa and decrease community dispersion, other forms of nutrient enrichment, as well as ammonium combined with other forms of enrichment, reduced dominant taxa and stimulated rare taxa. While the dominant taxa genus Cysteiniphilum and Unclassified Bacteria ASV5 declined in response to nutrient treatment, the genus Aquarickettsia increased significantly in these samples, despite a low average abundance of this genus in genotype ML-7 samples (0.191 ± 0.842%). The positive response of Aquarickettsia to nutrient-enriched conditions has been previously documented 33 , 38 . The increase in Aquarickettsia in samples of genotype ML-7 is notable as disease-susceptible genotypes of A. cervicornis are characterized by high abundance of this putative parasite 31 . Although relative abundance remained low across the experiment, the increase over the course of the experiment was considerable, increasing from 0.021 ± 0.101% at week 0 to 0.143 ± 2.67% at week 6 in nutrient-enriched samples, equating to an increase of about sevenfold. Individual nutrient treatments did not increase the abundance of Aquarickettsia in samples from this study, though phosphate enrichment had been previously found to stimulate abundances of this putative parasite 33 . The lack of clear response of this genus to a single treatment, despite a statistically significant increase in nutrient-treated samples as a group compared to untreated samples, suggests that changes in this taxon may have been stochastic and highly dependent on abundance of this taxon at the start of enrichment: only 29 of 48 timepoint 0 samples had detectable Aquarickettsia . The observed loss of Cysteiniphilum in response to nutrient treatment may be a result of its strictly aerobic respiration: nutrient enrichment can induce localized hypoxic conditions that may affect bacteria living in coral mucus or tissue 58 . Unclassified Bacteria ASV5 responded significantly to nutrient enrichment and with a large magnitude of change, though the lack of closely-related reference sequences for this organism makes it difficult to predict its role in the microbiome of genotype ML-7 or why it is lost with nutrient enrichment. The genus Ruegeria , which responded positively to nutrient enrichment, has been proposed to play a beneficial role in coral microbiomes by producing antimicrobial compounds that inhibit growth of Vibrio coralliilyticus 59 . The increase of this taxon may represent preemptive defensive activity in response to nutrient enrichment to prevent the growth of opportunistic pathogens, found to respond positively to nutrients 15 , 60 . The only taxon that changed under untreated conditions was the genus Ferrimonas , which declined slightly with time. As members of this genus are found in microbiomes of corals exposed to thermal stress 61 , 62 , this taxon might have been slightly elevated on the reef due to fragment collection occurring in early summer, while aquarium temperatures were maintained at 27.19 ± 0.6 °C (consistent with late spring coastal water temperatures in the Florida Keys). Microbiomes of a disease-resistant Acropora cervicornis genet do not respond to stress with dysbiosis It has recently been proposed that the ability to tolerate significant changes in microbiome structure without negative outcomes is an important predictor in stress resistance 63 . In seven coral species exposed to white plague disease, coral species that were susceptible to disease experienced minimal microbiome changes between diseased and non-diseased states, while corals that were resistant to disease experienced much greater microbiome shifts without disease development 63 . Genotype ML-7 exhibited significant shifts in diversity over the course of this experiment in response to combined (N, P, and A) and ammonium treatments, and the relative abundance of certain key taxa was demonstrated to respond to nutrient enrichment. Yet no obvious negative outcomes were observed in response to these treatments, as growth rates for ammonium and combined treated corals were not significantly different from untreated corals by 6 weeks of exposure. Furthermore, microbiomes of genotype ML-7 did not undergo a complete community restructuring in response to nutrient enrichment despite shifts in the abundance of individual microbial members. This, coupled with previous data showing that genotype ML-7 is more resistant to white band disease development 39 , suggests that genotype ML-7 may exhibit microbiome flexibility in response to changing environmental conditions. Microbiome flexibility has been proposed as a strategy to adapt to stress, but may risk the loss of essential symbiotic partners or allow the infiltration of opportunistic pathogens 47 . While genotype ML-50 responded to nutrient enrichment rapidly, reaching a new stable state before 3 weeks of enrichment 33 , genotype ML-7 did not exhibit significant shifts in microbiome community structure or diversity until 6 weeks of exposure to tank conditions, likely driven by the high amount of microbiome evenness exhibited by this genotype. This may suggest that while short periods of stress may be tolerated by genotype ML-7 without microbiome restructuring, sustained environmental stress may alter microbiome composition. As our nutrient exposure experiment was limited to 6 weeks, it is possible that microbial community composition in this genotype may shift further after greater periods of nutrient enrichment. Nonetheless, the same duration of exposure induced significant changes in microbial community composition in genotype ML-50 33 . In contrast to the results from the present study, a recent study including genotype ML-7 found that this genotype had reduced survival rates in response to experimental ammonium and phosphate enrichment and to a combination of nutrient enrichment and thermal stress in comparison to other genotypes harboring higher abundance of the putative parasite Aquarickettsia 64 . Corals in this previous study were allowed to acclimate to low-nutrient tank conditions for 4 months prior to enrichment, and this longer acclimation period to ex situ conditions may have created a greater shock upon exposure to nutrient enrichment, as no enrichment-related mortality was observed under the conditions of the present study. Furthermore, ammonium enrichment levels in the previous study were double the highest (4× ambient) concentrations used in this study. Further work is needed to elucidate the effects of nutrient enrichment on restoration genotypes after outplanting, the role the microbiome plays in regulating these effects, and whether removal or reversal of these stressors allows the microbiome to recover. Microbiome diversity may be a biomarker for stress tolerance In this study, we found that the disease-resistant Acropora cervicornis genet “ML-7” harbored high microbial diversity and that microbiomes were largely able to withstand nutrient enrichment, experiencing few significant changes in composition or diversity. While this study was limited to the use of only one genet, only a very small number of disease-resistant A. cervicornis genotypes have been identified 39 , 40 . Indeed, only two of sixteen Florida Keys genotypes screened by Miller et al . in 2016 were disease resistant, and no genotype was disease resistant when fourteen different genotypes were screened in 2017 40 . An even lower prevalence of naturally disease resistant genotypes (6% of 49 genotypes) was found in Panama during wild coral surveillance 65 . As we had limited access to disease-resistant genets, we elected to instead increase replication within this genotype, with an average of ~ 5.9 replicates per treatment (other than T0, Supp. Table 2 ), and found consistent patterns between replicates. Disease resistant A. cervicornis genotypes within Mote’s nurseries exhibit consistent signatures of high microbial diversity, and are distinct in composition from Aquarickettsia -dominated disease susceptible genets 31 , 32 . We therefore contend that patterns observed in this study can be used to make predictions for microbiome responses of other disease-resistant A. cervicornis genets to nutrient stress. Microbiomes of Acropora cervicornis genotype ML-7, previously found to be disease-resistant, were more stable in response to nutrient enrichment when compared to microbiomes of genotype ML-50, identified as disease-receptive. Results from this study parallel those from a separate study, in which exposure of individuals of genotype ML-7 to thermal stress did not significantly alter alpha diversity, community dispersion, or the abundance of any single bacterial taxon 31 . In contrast, many other genotypes of A. cervicornis respond to stressors by experiencing dramatic shifts in abundance of the genus Aquarickettsia . Bleaching of genotype ML-50 induced a near-complete loss of Aquarickettsia 31 , normally high abundance in this genotype. The loss of this dominant taxon allowed for the infiltration of putative pathogens including members of the order Alteromonadales, that appeared to occupy the niche space abandoned by Aquarickettsia . Nutrient enrichment also significantly increased abundances of Aquarickettsia in disease-susceptible genotypes of A. cervicornis with as little as 3 weeks of nutrient exposure 33 . In this study, we found that even with 6 weeks of nutrient enrichment, microbiomes of these two genotypes remained distinct, and genotype ML-7 experienced minimal changes in community structure, in contrast with genotype ML-50, which was lower diversity and more susceptible to microbiome shifts in response to nutrient enrichment. Microbiome dysbiosis, in the form of significant and rapid changes in abundance of Aquarickettsia with corresponding effects on microbiome structure, may therefore contribute to the disease susceptibility of genotypes including ML-50, especially as bleaching and nutrient enrichment, and their co-occurrence, becomes increasingly common 12 , 66 ."
} | 6,267 |
36543800 | PMC9772184 | pmc | 2,270 | {
"abstract": "Silk is a unique, remarkably strong biomaterial made of simple protein building blocks. To date, no synthetic method has come close to reproducing the properties of natural silk, due to the complexity and insufficient understanding of the mechanism of the silk fiber formation. Here, we use a combination of bulk analytical techniques and nanoscale analytical methods, including nano-infrared spectroscopy coupled with atomic force microscopy, to probe the structural characteristics directly, transitions, and evolution of the associated mechanical properties of silk protein species corresponding to the supramolecular phase states inside the silkworm’s silk gland. We found that the key step in silk-fiber production is the formation of nanoscale compartments that guide the structural transition of proteins from their native fold into crystalline β -sheets. Remarkably, this process is reversible. Such reversibility enables the remodeling of the final mechanical characteristics of silk materials. These results open a new route for tailoring silk processing for a wide range of new material formats by controlling the structural transitions and self-assembly of the silk protein’s supramolecular phases.",
"introduction": "Introduction Silk is a natural protein-based biopolymer with unique mechanical properties and exquisite biocompatibility and biodegradability 1 – 3 . Natural silks are produced by many insects and arthropods 4 , 5 . Notably, spider silk and silkworm silk have been the subject of much interest due to the former’s incredible toughness and the latter’s potential for genetic modification and commercial production 6 – 10 . The superior physical and biological properties of silk fibers (produced by spiders or silkworms) are achieved by fine control over the phase transitions and the formation of supramolecular assemblies inside the silk gland. These processes span from the protein synthesis region to the spinning duct 4 , 11 . Generally, silk protein is stored in a silk gland in liquid form as a highly viscous pulp 12 , 13 . During the spinning process, the protein is transformed from a soluble, largely disordered random coil structure into solid fibers containing ordered intermolecular hydrogen-bonded, β -sheet-rich conformations 14 , 15 . Interestingly, the generation of the fiber from a soluble silk protein feedstock is based solely on the underlying structural (secondary structure) transformations, while the final fiber structure requires external processing (spinning in the duct) 16 – 23 . The final composition of the natural silkworm silk fiber comprises a silk fibroin core and a sericin (glycoprotein gum) coating. The coating layer, which is added at the final stages of the spinning process, glues two fibroin fibers to form the final composite material 16 . The structural transformations involve the formation of multiple supramolecular phase states of silk fibroin, from soluble monomers, through microscale spherical assemblies, to liquid crystals and nanofibrils 4 , 17 – 21 . The microscale spherical assemblies (hereafter referred to as micro compartments) play a dual regulatory role: they decrease local viscosity, which enables the storage of a highly unstable and aggregation-prone protein solution, and they regulate protein crystallization under applied shear during the spinning process 22 , 23 . To date, the structural characteristics and conformational transitions of silk protein inside the compartments remain largely unknown. In particular, we have yet to elucidate the true structural form of the protein inside the supramolecular assemblies when stored in the gland, and the process by which this precursor solution is transformed into a hierarchical polycrystalline structure when spun through the duct. By following the key morphological changes in the supramolecular assemblies of silkworm silk, we were able to determine that the initial steps of silk secretion and storage inside the silk gland do indeed follow the micelle theory 4 , 17 , 24 of silk assembly. This theory suggests that silk protein spontaneously forms spherical micelle-like structures at high protein concentrations and high viscosities. These micelles are then further spun into micronscale silk fibroin fiber 24 . We observed that a phase rearrangement occurs inside the microscale spherical structures, accompanied by the appearance of nanoscale spherical assemblies. In order to clearly delineate the different types of spherical organization of the silk protein in the different assembly stages, we classify these structures as compartments: the micron-scale spherical structures as microcompartments, and the nanoscale spherical assemblies as nanocompartments. Our structural analysis reveals that protein confined in microcompartments preserves its native secondary structure (initial secondary structure of the protein when stored inside the silk gland), as postulated by the micelle theory 4 , 17 , 24 . In contrast, the nanocompartments act as a reversible regulatory step that induces the transformation of natively folded protein into a highly ordered structure, rich in β -sheets, in which the formation of nanofibrils is an irreversible part of this structural transformation. These results expand our current knowledge of the morphological nature of silk supramolecular structures and the critical role played by the supramolecular phase states in the structural transformations of fiber-forming proteins in generating fibrillar materials with superior physical properties.",
"discussion": "Results and discussion Mechanism underlying silk transformation from soluble protein into solid nanofilaments To characterize the supramolecular phase states of the fibroin protein inside the silkworm silk gland, we analyzed small (up to 1 cm) sections of silkworm glands from B. mori silkworms (see Methods and Fig. 1a ) with light microscopy, atomic force microscopy (AFM), and scanning electron microscopy (SEM). The analysis revealed a morphological hierarchy, summarized in Fig. 1 . We found that the initial assembly of silk protein, in general, follows the classical micellar theory 4 , 17 , 24 , which postulates that when the protein (fibroin) concentration inside the silk gland becomes very high, it triggers the spontaneous formation of the spherical (micelle) micron-scale assemblies that reduce the local viscosity. Namely, protein monomers (Fig. 1b ), which are secreted in the posterior part of the gland located close to the silkworm tail (Fig. 1a (i) ), first undergo microscale compartmentalization (Fig. 1c ) in the posterior-middle part of the silk gland (see silk gland sections in Fig. 1a (ii) ). This micron-scale compartmentalization stabilizes the soluble protein pulp and prevents premature aggregation (see AFM, optical images and schematics of the compartments’ composition in Supplementary Figs. 1a, b and 2a ). Previous studies, examined the silk fibroin sequence and the spatial alignment of the specific hydrophobic regions (with the repetitive sequence GAGAGAGS) and hydrophilic regions (domains) inside the compartments. They have suggested that larger hydrophilic domains at both the amino and carboxy termini form the outer edges of the microcompartment, sequestering the hydrophobic domains within the compartment’s core 4 , 25 – 27 . The smaller hydrophilic domains in between twelve hydrophobic repetitive sequences in the fibroin chain 25 – 27 ensure the compartments’ solubility in water, thus preventing premature β -sheet formation (see the schematics in Fig. 1 and Supplementary Fig. 2b ). As the concentration of fibroin increases, along with small changes in the chemical environment (decrease in pH and changes in ion concentration), the inter-molecular interaction increases, leading to the formation of colloidosomal spherical assemblies (Fig. 1d ) and, later on, to their coalescence. These processes take place in the middle-middle (Fig. 1a (iii) ) and anterior-middle part of the gland (see Fig. 1a (iv) ). The next stage is microcompartment disassembly, which is associated with the conversion of soluble protein into nanofibrillar β -sheets rich solid structures (Fig. 1a (v) , f ). The solid nanofibrils are then aligned with the elongational flow field and spun into microscale fibers (Fig. 1a (vi), g ) through the spinneret in the anterior part of the silkworm silk gland 28 . Fig. 1 Mechanism underlying silk transformation from soluble protein into solid nanofilaments. a Schematic representation of the five regions of the silkworm silk gland (top panel) and their corresponding silk protein supramolecular phases (bottom panel): (i) posterior region, where protein monomers are synthesized, (ii) posterior-middle region, where protein microcompartments were detected, (iii) middle-middle, where appearance of colloidosomal microcompartments was observed, (iv) anterior-middle, where the presence of nanocompartments was detected, (v) anterior region, filled with silk nanofibrils, and (vi) spun microfiber via spinneret. b AFM image of protein monomers. c Optical microscopy image of microcompartments. d Optical microscopy image of colloidosomal microcompartments. e AFM image of nanocompartments, f AFM image of silk nanofibrils. g SEM image of the spun silk microfiber, composed of a fibroin core and a sericin coating layer. The scale bars are shown at the bottom of each image. However, our analysis reveals that, in addition to the expected above-described microscale compartmentalization, the proteins undergo a secondary compartmentalization by forming nanoscale spherical structures inside the bigger microscale compartment (Fig. 1a (iii), (iv) , Fig. 1d, e ; Supplementary Movies 1 and 2 ). These nanoscale formations are consistent with literature reports 29 , 30 . The morphological transition takes place in the second half of the middle part of the gland, which is close to the anterior (see Fig. 1a (iv) ). Our examination of the stability of these nanoscale compartments (see Methods for details) further revealed a reversibility in the morphological organization. Namely, the nanoscale compartments tend to disassemble and re-assemble. The formation and dissociation of a nanocompartment (with sizes ranging from 20 to 200 nm) is regulated by the protein concentration and/or the amount of water. Thus, an increase in protein concentration above the minimal threshold leads to spontaneous nanocompartment formation (see Supplementary Fig. 3 ). When the protein concentration decreases, the nanocompartment disassembles into soluble monomeric protein with a native molecular weight of ~400 kDa, composed of a ~390 kDa heavy chain and a ~25 kDa light chain linked by a single disulfide bond 31 – 33 (Supplementary Fig. 2c ), thus restoring the native protein monomer’s conformation found in the posterior part of the gland, as resolved by Fourier Transform Infrared spectroscopy)FTIR(structural analysis (Fig. 2a ). Fig. 2 Bulk Fourier Transform Infrared spectroscopy (FTIR) analysis of supramolecular silk fibroin assembly states. a Amide I region IR-spectra showing overlap of monomers released from nanocompartments (blue) and monomers within the nanocompartments (orange). The analysis confirms the re-folding event in protein monomers upon nanospherical assembly disintegration. b Amide I FTIR spectra of silk fibroin monomers, nanocompartments, and nanofibrils. c Comparative analysis of the secondary structure of ( b ) with the band positions of the β -sheets at 1610–1635 cm –1 , anti-parallel β -sheets at 1690–1705 cm –1 , random coil and α -helixes at 1635–1665 cm –1 , and β -turns at 1665–1690 cm –1 . Source data are provided as a Source Data file. We next probed the role of protein concentration in spontaneous formation of silk-rich compartments. To this end, we tracked changes in the critical micelle concentration (CMC) using a pyrene-based staining assay 34 , 35 . Pyrene is a hydrophobic dye that is sensitive to changes in polarity in the environment and is characterized by the presence of several emission peaks in its fluorescence spectra. Changes in the ratio between these peaks occur in response to shifts in the polarity of the surrounding media. We performed a series of silk fibroin dilutions and measured the associated changes in the ratio between I 3 /I 1 (see Methods). A high ratio of I 3 /I 1 points to the presence of micellar structures in the solution, thereby confirming the presence of silk compartments. The measured CMC in fibroin nanocompartments was 0.4 mg/ml (Supplementary Fig. 3a ). The presence of silk nanocompartments was also confirmed by AFM imaging (Supplementary Fig. 3b ). A sigmoid fit confirmed that silk nanocompartments assemble and disassemble as standard micelles at CMC. However, we observed a relatively low slope in the silk nanocompartments 36 , 37 (for concentrations between 0.2 mg/ml and 20 mg/ml), indicating a polydispersed compartment size distribution, which is comparable to the behavior of micellar protein β -casein 36 , 37 . Moreover, the low value of silk fibroin CMC suggests that nanocompartments are relatively energetically stable 36 , 37 . To evaluate the reversibility of spontaneous silk compartmentalization, we diluted highly concentrated compartments containing silk solution (20 mg/ml) down to 0.2 mg/ml and then concentrated them up to 2 mg/ml. The presence and disintegration of the compartments were tracked by measuring changes in I 3 /I 1 (Supplementary Fig. 3c ) and verified by AFM (Supplementary Fig. 3d ). The results confirm that the assembly and disassembly of nanocompartments are reversible and concentration-dependent, as has been theorized in previous literature reports 4 , 24 . Considering that there is a pH gradient inside the silk gland, as well as shear forces (elongational flow), we further probed the effect of pH and shear on the stability of silk nanocompartments. This was done by exposing fibroin compartments at a concentration of 20 mg/ml to pH values varying between 5.5 and 10 (see Methods; Supplementary Fig. 3e, f ). Our observations show that at basic pH values (>pH 8), nanocompartments are stable, while lowering the pH to below pH 7 triggers compartment disintegration, accompanied by structural transitions in the silk protein from the native fold into a highly ordered β -sheet-rich conformation. Surprisingly, the exposure of the compartmentalized silk to shear forces (see Methods) did not alter the compartments’ integrity (Supplementary Fig. 3g ) or protein fold, an observation consistent with our previous report 12 . Structural characteristics of silk phases Our findings regarding the nature of the silk protein inside the nanocompartments raise three questions: What are the structural characteristics of the protein inside the nanocompartments? What is the role of the formation of nanocompartments in the overall process of protein phase transitions? And what is the impact of the formation of nanoscale assemblies on the mechanics of the silk-based phase states? To address these questions, we performed a conformational study of silk protein assemblies, namely, monomers, nanocompartments, and nanofibrils. This entailed a FTIR analysis, both of bulk assemblies (Fig. 2a, b ) and of individual fibrils and monomers by nano-IR imaging and nano-FTIR spectroscopy. In general, the conformational changes in protein structure are characterized by spectral shifts in two major vibrational bands, the amide I (1600–1700 cm −1 ) and the amide II band (1480–1600 cm −1 ), which correspond mainly to C = O stretching vibrations and NH bending/CN stretching vibrations, respectively. The amide I region is commonly used to characterize the intermolecular β -sheets (1610–1635 cm −1 ), random coil/ α-helix (1635–1665 cm −1 ), β -turn (1665–1690 cm −1 ), and anti-parallel β -sheet (1690–1705 cm −1 ) conformations 38 , 39 . The comparative FTIR analysis revealed differences between the protein secondary structure of the monomers, microscale and nanoscale compartments and nanofibrils. An increasing β -sheet conformation presence was observed in the following order: monomers <nanocompartments <nanofibrils (Fig. 2c ). The protein monomers exhibited the lowest fraction of β -sheet, and the nanofibrillar assemblies the highest. This is confirmed by the amide I band’s position, which decreases from 1650 cm −1 for monomers, via 1630 cm −1 for nanomicelles, to 1620 cm −1 for nanofibers. This observation indicates that the nanocompartments play a substantial role as structural intermediates in the conformational transition of proteins from a predominantly disordered state to a highly ordered β -sheet structure. Interestingly, our additional analysis revealed that the structural transformations between the supramolecular assembly states are accompanied by changes in the pH and surface charge (ζ-potential) 40 of the protein complexes. As previously shown 41 , the sequence of silk protein can be roughly classified into three different sections: (1) a repetitive hydrophobic domain with a GAGAGS repetitive motif, which has an isoelectric point of pI = 3.8, (2) a negatively charged hydrophilic N-terminal domain with pI = 4.6, and (3) positively charged domains at the C-terminus with pI = 10.5. Our observations show that the structural transformations of fibroin protein from its monomeric state into nanoscale spherical assemblies as well as into nanofibrils are accompanied by a reduction in the surface charge, driven by a decrease in pH (see Supplementary Fig. 4 ). We, therefore, performed a series of experiments in which we varied the pH and measured the resultant changes in surface charges (ζ -potential) and in protein folding. The monomers (pH ~9) carry a highly negative surface charge of −20 meV, which creates a repulsive interaction between the residues 42 , 43 . This, in turn, forces the entire protein to adopt an elongated, molecular conformation. Nanocompartmentalization leads to a reduction in the surface charge (to between −15 and −8 meV), which occurs simultaneously with the change in pH (~6–7), which is cited in the literature 42 , 43 , and is indicative of the formation of a more compact conformation. Finally, fibrillation, which is caused by lowering the pH, leads to the suppression of dominant repulsive interactions, a phenomenon that promotes a less extended, more compact, β -sheet conformation. This observation is in good agreement with the above FTIR analysis, which indicates the presence of a larger fraction of β -sheet protein conformations in the silk fibroin nanofibrillar constructs. Nano-FTIR analysis of silk assembly states Even though the conformational FTIR analysis under bulk conditions revealed the general trend of the structural transitions of silk protein assemblies, a major limitation of bulk characterization techniques lies in their inability to differentiate between the protein conformational variability inside nanoscale objects. In particular, as protein self-assembly is a complex, dynamic process, bulk samples of protein fibrils might contain a small fraction of monomers and even nanosized spherical structures. This complicates the interpretation of the results. We, therefore, performed an AFM-based nano-FTIR analysis to determine the local protein conformation in each nm-scale construct: monomers, single nanocompartments and single nanofibrils. Nano-FTIR spectroscopy exploits the enhanced field at the sharp metal tip of the AFM to excite and detect FTIR spectral characteristics simultaneously with nanometer-scale topography. This is a very promising tool for characterizing protein complexes and, as shown recently, fibrillar protein aggregation stages, due to its ability to correlate the information from AFM morphological analyses with FTIR-resolved secondary structures 44 , 45 . In our analysis and band assignment, we took into account the fact that the bands of the nano-FTIR spectra are upshifted (~10 cm −1 ) compared to the bulk FTIR analysis, due to the technical differences between the two methods of spectra collection (see detailed explanation in the nano-FTIR spectra analysis section in the Supplementary Note 1 ). Even though we used the standard ranges, we observed the same trend as in the bulk FTIR analysis. We first performed a qualitative test for nano-FTIR spectra verification on three types of protein assemblies: monomers, nanocompartments, and nanofibrils. As depicted in Fig. 3a (and Supplementary Fig. 5 ), three wavenumbers have been chosen for this analysis: 1600, 1629, and 1641 cm −1 . The first is in the “valley” between the amide I and amide II absorption bands, whereas the two other wavenumbers approach the amide I maximum in the nano-FTIR spectra (see below). The phase images are related to the IR absorption and show the expected increase in protein absorption from 1600 to 1641 cm −1 , associated with a disordered random coil and α -helix (Fig. 3a (ii), (iii) and 3b (ii), (iii) ). As expected, we observed no signal at 1482 cm −1 in any of the samples (see Supplementary Fig. 5 ), namely, monomers, compartments, and fibrils. The absorption signals at 1600 cm −1 confirmed that the structures are proteinaceous. Further analysis elucidated the differences in the fraction of disordered (random coil, 1641 cm −1 ) and ordered ( β -sheet, 1629 cm −1 ) regions for different protein assemblies (see Fig. 3a, b (iv), (v) ). Fig. 3 Nano-FTIR analysis of silk protein assemblies. FTIR maps collected from samples containing a a mixture of protein nanocompartments and protein monomers and b silk protein nanofibrils. (i) FTIR- reflection maps collected at 1482 cm −1 , to distinguish general protein content from organic contaminants (negative control); (ii) absorption maps collected at 1482 cm −1 ; (iii) absorption maps collected at 1600 cm −1 ; (iv) absorption maps collected at 1629 cm −1 showing the β-sheet content; (v) FTIR absorption maps collected at 1641 cm −1 showing a signal from a disordered random coil and an α-helix. c AFM topography image of silk monomers (marked with blue dots) and silk compartments (marked with red dots), from which IR spectra were recorded and depicted in d nano-FTIR spectra of silk protein monomers, and e nano-FTIR spectra of protein compartments. f AFM topography image depicting silk nanofibrils (marked with green dots), from which FTIR spectra were recorded and depicted in g . h Bar chart of the relative amounts of the β-sheets at 1610–1635 cm –1 , the anti-parallel β-sheets at 1690–1705 cm –1 , random coil and α-helixes at 1635–1665 cm –1 , and β-turns at 1665–1690 cm –1 in ( d ), ( e ) and ( g ). Source data are provided as a Source Data file. Next, we conducted a more detailed conformational study using localized nano-FTIR spectra. The nano-FTIR spectra for nanocompartments and monomers were collected from five different locations, and for nanofibrils from ten different locations (Fig. 3c, f , Supplementary Figs. 5 and 6 ). The spectra were normalized, averaged, and subjected to curve fitting (see Methods). The nano-FTIR analysis revealed a conformational trend akin to bulk FTIR (Fig. 3d, e, g and h ). The β -sheet content increased from monomers (Fig. 3d ) to nanocompartments (Fig. 3e ) and from nanocompartments to nanofibrils (Fig. 3g, h ). The mechanism underlying the silk protein structural transitions To gain further insight into the link between the molecular level mechanisms of protein fibrillation and the supramolecular structural organization, we investigated the changes in the physical characteristics of the protein in the different silk protein phases. The freeze-fracture cryo-scanning electron microscopy (see Methods at Cryo-SEM,) analysis of the anterior part of the silk gland, as well as an AFM analysis of the silk protein extracted from the posterior part of the silk gland 46 , 47 , revealed that the nanocompartments contracted (Fig. 4a ). Nanocompartments in their liquid state (with no exposure to shear forces), whose measured average volume is 1.2*10 7 nm 3 and average diameter is 257 nm, shrink under the action of elongational flow stress (Fig. 4a, d and Supplementary Fig. 1a, b ). The “prefibrillar” nanocompartments are smaller in volume, featuring an average value of 60,160 nm 3 and an average diameter of 47 nm (Fig. 4a, d ). At the final stage of silk fibrillar assembly, the nanocompartments shrink even further under the action of shear stress and align along the nanofibril axis to form a highly ordered pattern, with repetitive distances of ~30–50 nm (see Supplementary Fig. 7 ) and an average volume and diameter of 16,750 nm 3 and 31 nm, respectively (Fig. 4b–d ). A volumetric analysis confirmed that nanocompartments tend to shrink in size under applied shear (elongational flow field), accompanied by water expulsion (from ~7% to ~70% of total weight 48 – 50 ) (Fig. 4d ). Fig. 4 The mechanism underlying the silk protein structural transitions. a Freeze-fractured cryo-electron microscopy image of the anterior part of the silkworm silk gland revealing an event of linear ordering of the fibroin nanocompartments (denoted by a black rectangle). b AFM image of silk fibroin protein self - assembled into nanoscale fibrils. c Enlarged image of ( b ) revealing the presence of spherical assemblies in nanofibrillar structures. d Volumetric calculation of unordered silk fibroin nanocompartments not subjected to elongational flow stress (average volume of 1.2*10 7 nm 3 ) vs. nanocompartments that have aligned due to shear produced by elongational flow stress (average of 60,160 nm 3 ) vs. nanocompartments subjected to shear stress (average volume of 16,750 nm 3 ). The standard error and the mean are indicated in the graph. e Molecular dynamics simulations of silk fibroin conformational transitions from a relaxed state (representative of an unfolded state) (i) to intermolecular interactions (ii), which gradually induce the formation of β -sheets (iii and iv) due to the pulling of the tetrameric silk fibroin hydrophobic domain’s GAGAGS sequence. (v) MD simulation of a 12-meric hydrophobic silk fibroin domain, formed by the pulling of β -sheets. f SAXS analysis of silk fibroin monomers, nanocompartments and silk fibroin nanofibrils. Azimuthally-interacted background-subtracted solution X-ray scattering absolute intensity, I, as a function of the magnitude of the scattering vector, q, from monomer, nanocompartment and nanofibril- containing samples (open blue symbols). Solid red curves correspond to a linear combination of g silk monomers (left illustration) with either elongated (nanofibrils) and spherical (right illustration) structures. Source data are provided as a Source Data file. To gain further insight into the mechanism of silk structural transitions from a relatively disordered state to a more ordered β -sheet-rich conformation under the action of the elongational flow field, we performed molecular dynamics (MD) simulations. According to literature reports, the mechanism of silk protein fibrillation is quite complex and is initiated by N-to-N terminus dimerization, followed by hydrophobic interactions between repetitive hydrophobic motifs (GAGAGS) 25 – 27 . These interactions are believed to be the major driving force that dictates the rate of structural transitions in silk and the kinetics of the protein’s self-assembly. Thus, we performed our computational simulations on the above-mentioned repetitive motifs, which entailed isolating the hydrophobic motif of the silk sequence, known to adopt β -sheet folds, and tracking the structural transition. We observed that the conformational transition in silk takes place under tension (Fig. 4e ). This is portrayed in Fig. 4e (ii)–(v) , which shows how silk protein undergoes various transitions that eventually lead to a long-range β -sheet formation from silk hydrophobic motifs. Seeking to elucidate further the structural composition of silk monomers and silk nanocompartments, we also performed a small-angle X-ray scattering analysis (SAXS) (see Methods). The results are summarized in Fig. 4f, g , and explained in the Methods. The SAXS analysis revealed that silk monomers adopted a disk-like shape (see Fig. 4g ) with a radius of 3.7 ± 0.2 nm and polydispersity of 1.2 nm and a height of ~1 nm, which is consistent with our AFM analysis (Supplementary Fig. 1c ). Interestingly, we found that for nanocompartment samples, monomers coexisted with spheres (nanocompartments) and fibrillar-like assemblies at a very low mass fraction (see Methods and Supplementary Fig. 8 ). Whereas the measured dimensions and disk-like shape of the monomer component are identical to the sample containing only silk monomers, the radius of the nanocompartments varied in size. Mechanical properties of silk phase states To better understand the relation between structure and mechanics in silk materials, we studied the nanomechanical properties of the different silk assembly states. The results are shown in Fig. 5 to elucidate the resistance to stress of the monomers, nanocompartments, nanofibrils, and silk microfibers. The Derjaguin-Muller-Toporov (DMT) modulus 51 was calculated from force curves made at each pixel while keeping the deformation below 10% of the sample thickness, to eliminate the contribution from the mica substrate (see Methods). No significant difference in the elastic modulus was found between the microfiber and nanofibril samples (19.3 ± 2.8 GPa and 21.8 ± 3.2 GPa, respectively). Furthermore, no significant modulus difference was found between monomers and nanocompartments (4.2 ± 0.7 GPa and 3.1 ± 0.7 GPa, respectively). However, there is a pronounced difference in the stiffness of microfibers and nanofibrils compared to the monomers and nanocompartments. Biological samples with a predominant β -sheet composition are known for their robust mechanical properties, reflected in their large modulus values 52 . These results, thus, support the findings reported above with respect to the high β -sheet content of the microfibers and nanofibrils. In contrast, the markedly lower modulus of monomers and nanocompartments is indicative of protein content with a native fold. Overall, these results suggest that the nanofibers and microfibrils have a similar composition, consisting predominantly of β -sheet-conFig.d protein, whereas the monomers and nanocompartments express the mechanically weaker native fold. Fig. 5 Nanomechanical analysis of silk fibroin assembly states. a Comparison of the height, deformation and elastic modulus of silk protein monomers, compartments, nanofibrils and microfibers. b Quantitative analysis of the elastic modulus (GPa) of the different assemblies. The elastic modulus rises as the β-sheet content increases. The error bars indicate the standard deviation of the elastic modulus sampled from several AFM images. Source data are provided as a Source Data file. In summary, the physical form of proteins within the silkworm silk gland permits their storage at very high concentrations rather than as precipitated material. It also enables their precisely controlled structural transformation from a native form (original conformation) into a β -sheet -rich conformation in solution, prior to being formed into solid fibers. We report here important regulatory steps in the silk protein’s transition from a soluble monomeric state, marked by a native, predominantly disordered fold, into insoluble β -sheet rich nanofibrils, which results in the formation of nanoscale compartments. Importantly, the nanocompartmentalization step is reversible and enables the un-folding of the protein conformation back to the disordered state. Interestingly, such reversibility can occur before the nanocompartment structural organization is subjected to the shear forces created by elongational flow during the spinning process. Thus, our findings might have implications for the construction of artificial routes for mimicking the exceptional intrinsic mechanical properties of silk fibers."
} | 8,045 |
34361010 | PMC8347015 | pmc | 2,271 | {
"abstract": "Biofilms are complex structures formed by a community of microbes adhering to a surface and/or to each other through the secretion of an adhesive and protective matrix. The establishment of these structures requires a coordination of action between microorganisms through powerful communication systems such as quorum-sensing. Therefore, auxiliary bacteria capable of interfering with these means of communication could be used to prevent biofilm formation and development. The phytopathogen Rhizobium rhizogenes , which causes hairy root disease and forms large biofilms in hydroponic crops, and the biocontrol agent Rhodococcus erythropolis R138 were used for this study. Changes in biofilm biovolume and structure, as well as interactions between rhizobia and rhodococci, were monitored by confocal laser scanning microscopy with appropriate fluorescent biosensors. We obtained direct visual evidence of an exchange of signals between rhizobia and the jamming of this communication by Rhodococcus within the biofilm. Signaling molecules were characterized as long chain (C 14 ) N -acyl-homoserine lactones. The role of the Qsd quorum-quenching pathway in biofilm alteration was confirmed with an R. erythropolis mutant unable to produce the QsdA lactonase, and by expression of the qsdA gene in a heterologous host, Escherichia coli . Finally, Rhizobium biofilm formation was similarly inhibited by a purified extract of QsdA enzyme.",
"introduction": "1. Introduction Bacteria are rarely encountered as single dispersed organisms in the environment. They generally live in communities and may colonize the surfaces of minerals and living tissues by forming biofilms [ 1 , 2 , 3 ]. This process involves the formation of aggregates with distinct sessile cells, followed by cell division to form small clusters, microcolonies, and larger aggregates. During the maturation step, the microbial community synthesizes a hydrated matrix of polysaccharides, proteins, and nucleic acids, in which bacterial cells are embedded. Finally, only the underside of the biofilm remains in direct contact with the substrate, the rest of the biofilm forming a multilayer heterogeneous microbial mat [ 1 , 2 , 4 , 5 , 6 ]. Switching to this lifestyle provides the microbes with more favorable environmental conditions, favoring their survival in an otherwise hostile environment (i.e., dehydration, oligotrophy, and the presence of antibiotics and predators), leading to a sustainable colonization of the coveted niche, in addition to avoiding their washing out (e.g., in a flowing stream or an infected gut) [ 1 , 3 , 7 ]. The biofilm phenotype may be seen as a form of collective behavior, in which all the members of the community work together to ensure the persistence of the group within the environment [ 8 , 9 , 10 , 11 ]. For example, bacteria express different phenotypes according to their location within the structure of the biofilm. The bacteria at the periphery can be in an active metabolic state, enabling them to shield the ‘core bacteria’ at the center of the biofilm, including the dormant persister cells [ 9 ]. This social behavior requires finely tuned coordination between members of the biofilm. The switching of bacteria between motile behavior and a sessile lifestyle, including biofilm formation, maturation, and/or dispersion, therefore requires a plethora of molecular tools, including environmental sensors (two-component systems), quorum-sensing (QS) networks, and intracellular messengers such as 3′,5′-cyclic diguanylic acid (cyclic di-GMP) [ 12 , 13 , 14 , 15 , 16 ]. N -acyl- l -homoserine lactone (AHL)-based QS networks are strongly suspected to function as a master regulator of biofilm development in many Gram-negative bacteria [ 13 , 17 , 18 , 19 ]. In this communication system, each individual constitutively produces diffusible AHL signals, the environmental concentration of which is, therefore, directly linked to the density of the emitting population, but also its location, as in a host wound, which favors signal accumulation. Bacterial populations capable of detecting AHLs are, therefore, informed about the number of cells (quorum) and the degree of diffusion from the microenvironment, between bacterial sites within and outside the biofilm, for example. QS also enables the bacteria to synchronize the expression of genes involved in biofilm synthesis, the production of ‘public goods’ and other cooperative traits with benefits available to all the cells of the population at an individual cost to the responding cell [ 8 , 11 , 19 , 20 ]. As QS is essential for biofilm installation and development, disrupting the flow of signals between biofilm members by a quorum-quenching (QQ) mechanism appears to be a promising method of control [ 21 , 22 , 23 , 24 , 25 , 26 , 27 ]. This possible application of QQ first emerged in the clinical context [ 21 , 22 , 23 , 24 ] and in the field of anti-biofouling [ 25 , 26 , 27 , 28 ]. It has since been more explicitly proposed as a biofilm control method in more recent reviews [ 7 , 29 , 30 ]. QQ encompasses all the processes involved in the disturbance of QS, mediated by QS inhibitors, and by enzymes capable of inactivating QS signals, known as QQ enzymes [ 31 , 32 ]. Some environmental microorganisms can synthesize acylases, oxidoreductases, and/or lactonases, which then modify the various parts of AHLs [ 27 , 33 , 34 , 35 , 36 ]. In contexts in which lifestyle switches and biofilm formation are determinant traits for plant disease, the development of an anti-biofilm strategy essentially amounts to the development of an anti-virulence strategy, mediated by beneficial microorganisms acting as biocontrol agents [ 27 , 37 , 38 , 39 ]. Recent studies have shown that AHL-based QS disruption by both QQ inhibitors and enzymes can effectively reduce biofilm formation [ 21 , 40 , 41 ], but our knowledge remains limited by the paucity of direct visual evidence of the jamming (i.e., disturbance) of QS communication and its effects on biofilm formation. We aimed to bridge this gap by studying the biofilm formed by the α -Proteobacterium Rhizobium rhizogenes (formerly Agrobacterium rhizogenes ). This bacterium is a soilborne pathogen that causes hairy root disease in susceptible dicotyledonous plants [ 42 , 43 , 44 , 45 ]. It induces anarchic growth of the root system after wounding and infection. R. rhizogenes introduces its root growth-inducing transfer DNA (Ri T-DNA) into the host plant genome, bypassing the plant defense system, and then modulates the endogenous levels of two hormones, auxin and abscisic acid, leading to the initiation and proliferation of hairy roots [ 45 , 46 , 47 ]. Hydroponically grown tomato ( Solanum lycopersicum L.) and cucumber ( Cucumis sativus ) plants are greatly affected by this disease, which results in large economic losses [ 43 , 44 , 48 ]. In these cropping conditions, pathogenic strains of R. rhizogenes hijack the metabolism of the plant, directing it towards root production at the expense of aerial apical growth and fruit production. They also produce massive biofilms in the greenhouse irrigation system, which may obstruct the supply circuits with saline solutions (biofouling), thereby affecting crop yields [ 42 , 43 , 44 , 48 ]. AHL-based-QS systems are common in rhizobia [ 49 , 50 , 51 , 52 ]. We therefore hypothesized that AHL molecules control biofilm formation in R. rhizogenes . We validated this hypothesis by using the Gram-positive bacterium Rhodococcus erythropolis R138 to counteract biofilm formation by R. rhizogenes [ 53 , 54 ]. This biocontrol agent effectively degrades diverse AHLs, reducing levels of potato ( Solanum tuberosum ) blackleg and soft rot in hydroponic and field conditions [ 54 , 55 ]. In this strain, full QQ activity requires expression of the QS signal degradation (Qsd) pathway, leading to the production of the QsdA and QsdC intracellular enzymes, involved in the lactone ring and acyl chain catabolism of AHLs, respectively [ 53 ]. Confocal laser scanning microscopy (CLSM) can be used to observe the structure of biofilms and to quantify their biovolumes and thicknesses. It can also be used to monitor the QS and QQ phenomena occurring within the biofilm through the use of biosensors carrying promoter-probe vectors fused to reporter genes encoding fluorescent proteins. We used a dual-color strain of R. erythropolis , R138, for simultaneous observation of the disruption of AHL communication and its impact on the biofilm. This strain constitutively produces the mCherry fluorophore as a cell tag, as well as a reporter fusion based on the green fluorescent protein (GFP) gene, making it possible to detect AHL signaling molecules and their degradation simultaneously [ 38 , 53 ]. Using these biosensors, we provide the first images of QS and QQ molecular interactions within biofilm. Finally, we demonstrate the close connection between biofilm development and QS communication by highlighting the capacity of the lactonase QsdA both to silence AHL communication and to prevent biofilm formation.",
"discussion": "3. Discussion QS systems control diverse functions requiring the concerted actions of numerous individuals. They are cell-to-cell communication systems based on both the synthesis and perception of signaling molecules, the best known of which belong to the AHL family [ 17 , 18 ]. QS and biofilm formation are closely interconnected features of the social life of bacteria, particularly during the switch from a planktonic to a sessile lifestyle [ 8 , 11 ]. This synchronization appears to be essential at all stages of biofilm development, for metabolic reasons in particular. For example, swarming motility, which underlies surface motility within the developing biofilm at early stages of its development, has been shown to be under the control of QS regulation and nutrient conditions [ 62 , 63 ]. In addition, throughout biofilm construction, the secretion of a high density of extracellular polymeric substances (EPS) enables the cells to activate EPS synthesis selectively in the biofilm, thereby decreasing the costs of EPS production relative to the planktonic phase [ 64 ]. A downregulation of EPS synthesis at high cell density can also allow attached cells to redirect energy from EPS production to growth and cell division before a dispersal event [ 64 ]. Finally, QS triggers biofilm dispersion in a coordinated manner, probably because biofilm disassembly remains essential to allow bacteria to escape and colonize new niches in conditions of nutrient limitation [ 16 ]. AHLs are signaling molecules produced by various Gram-negative bacteria [ 17 , 18 ]. Among them, many phytopathogens use AHL-based communication [ 50 , 65 ], including a large number of rhizobial strains described as biofilm formers, in which QS plays a crucial role in the switch in lifestyle towards biofilms. This switch, governed by QS, involves modifications to the bacterial phenotype, including bacterial motility [ 66 , 67 ] and EPS production [ 68 , 69 , 70 ]. R. rhizogenes produces bulky biofilms responsible for economically significant damage to vegetables, such as tomatoes [ 42 , 44 , 48 ]. This bacterium was therefore considered a relevant model for this study. HPLC–MS characterization of the AHLs produced by R. rhizogenes 5520 T showed the production principally of 3-OH-C 14:1 -HSL and, to a lesser extent, 3-OH-C 14 -HSL ( Figure 1 ). On the basis of a previous analysis, we suggest that 3-OH-C 14:1 -HSL is the only one of these molecules acting as a QS signal, and that 3-OH-C 14 -HSL is a less specific product of AHL synthase or a catabolite that appears during AHL turnover [ 58 , 71 , 72 ]. Interestingly, 3-OH-C 14:1 -HSL also appears to be the master AHL in the symbiotic nodule-forming Rhizobium leguminosarum species [ 49 , 51 , 52 , 73 ]. This apparent incongruity has been explained, in particular, by Velazquez et al. [ 74 , 75 ], who showed that the coexistence of symbiosis- and pathogenicity-determining genes in R. rhizogenes strains enables these bacteria to induce either beneficial (nodules) or deleterious (tumors or hairy roots) effects in plants. AHL communication has been studied in R. rhizogenes in the context of biofilm establishment in vitro . Confocal microscopy and promoter-probe vectors carrying fluorescent protein-reporter genes are classical tools that have been used to monitor biofilm formation, structure, and development [ 76 , 77 , 78 , 79 , 80 , 81 , 82 ]. They are also powerful tools for localizing and quantifying the AHL-based communication of various plant-associated Gram-negative bacteria [ 57 , 83 , 84 , 85 , 86 , 87 , 88 ]. We used the fine detection capacity of the P. atrosepticum 6276-EI AHL biosensor to show that R. rhizogenes strain 5520 T produces AHLs in the biofilm context ( Figure 3 ). This pectobacterial biosensor colonized the rhizobial biofilm only weakly, but its presence was uniform within biofilm, and the green fluorescence associated with AHL production was detected both in the center and periphery of the biofilm, demonstrating the diffusion of these signaling molecules through the exopolysaccharide matrix ( Figure 3 ). In the plant rhizosphere, the triggering of disease by QS is disrupted by the presence of both bacteria producing QS signals and quencher bacteria capable of interfering with signal exchanges [ 89 ]. Diverse bacteria from the α-, β- and γ-Proteobacteria; Firmicutes; and Actinobacteria have been shown to inactivate QS signals [ 27 , 90 , 91 , 92 ]. Actinobacteria from the genus Rhodococcus have been shown to degrade various QS signals, including AHLs [ 89 , 91 , 93 ]. In R. erythropolis strain R138, recent advances in our understanding of the mechanisms regulating expression of the qsd operon expression made it possible to remove a fundamental technological lock and to construct the first biosensor of AHL-based QQ activity [ 53 ]. A dual-color rhodococcal reporter, encoding the mCherry fluorophore as a cell tag and GFP as a regulator-based reporter system, has been used for the simultaneous visualization in planta of bacterial spread and of the QQ activity, leading to the biocontrol of potato soft rot [ 59 , 94 ]. This work revealed the chronology of events leading to tuber maceration and QS communication, in addition to the disruption of QS communication and concomitant protection of the plant by the rhodococcal biocontrol agent. Here, we used the same biosensor to sense both QS and QQ activities within the rhizobial biofilm ( Figure 4 ). Rhodococcal cells may adopt two different shapes, rods and cocci, depending on their metabolic state. The rod-shaped cells reflect an active metabolic state, whereas the coccal cells correspond to an inactive metabolic state [ 95 , 96 ]. In our study, most of the yellow rhodococcal cells (quenchers) were rod-shaped, whereas most of the red rhodococcal cells were coccoid, this difference potentially reflecting the catabolism of AHLs in the quencher cells. Finally, such AHL catabolism was observed, particularly in regions containing a mixture of large numbers of rhizobial and rhodococcal cells. The proximity of partners is probably required for the entry of the AHL into the rhodococci destined for degradation. These are, to our knowledge, the first images of QQ activity within a mixed biofilm. The presence of the R138 biocontrol agent led to changes in the characteristics of the biofilm formed by the R. rhizogenes phytopathogen ( Figure 2 ). These modifications included changes in both the biovolume and structure of the biofilm. They are thought to be related to disruption of the QS communication between rhizobia in the biofilm. One possible interpretation of this phenomenon is that the jamming of AHL communication has a severe effect on the metabolism of rhizobial EPS, as previously observed in the Pseudomonas aeruginosa model [ 64 ]. This assertion is also based on the demonstration that EPS metabolism is essential for cell attachment and normal biofilm formation in another strain of R. rhizogenes [ 97 ]. The role of the Qsd pathway in modifying biofilm traits was demonstrated by studying the activity of the AHL-lactonase QsdA. The functions of QsdA, the key enzyme in this catabolic pathway, were previously investigated by transferring the qsdA gene to a heterologous host ( E. coli ), which transformed this bacterium into a powerful AHL quencher [ 61 ]. We also constructed an R. erythropolis R138 qsdA deletion mutant with an impaired ability to break down AHL [ 61 ]. As expected, the mutated strain completely lost its ability to alter biofilm formation by R. rhizogenes ; conversely, the transformed E. coli strain acquired this ability following insertion of the R138 qsdA gene ( Figure 5 ). Interestingly, deletion of the qsdA gene did not completely abolish the biofilm-inhibiting effect of R. erythropolis R138. This probably reflects the activity, in strain R138, of other enzymes involved in AHL catabolism, such as an AHL-amidohydrolase and at least one type of AHL lactonase from the α/β hydrolase family [ 36 , 53 , 98 , 99 , 100 , 101 ]. Nevertheless, QsdA lactonase, even following isolation from its progenitor, had strong efficacy against biofilm formation, similar to that recorded with the R138 strain or the heterologous host E. coli , under our assay conditions. These findings confirm the significant potential for QQ already observed for other AHL lactonases of the phosphotriesterase family, to which QsdA belongs [ 24 , 102 , 103 ]. In conclusion, this work shows that biofilm formation by R. rhizogenes is partly QS-dependent and that a biocontrol strategy based on another organism with QQ ability could be used to target this species. Future studies are nevertheless required to determine the physicochemical characteristics of QsdA (e.g., thermal stability, catalytic temperature range, resistance to detergents and organic solvents, radiation and proteases), to evaluate its true potential, in freeze-dried formulations for example [ 41 ]. Finally, additional support for the use of strain R138 as a biocontrol agent [ 54 , 55 , 99 ] is provided by the demonstration of its ability to prevent biofilm formation and biofouling in this study."
} | 4,580 |
31485382 | null | s2 | 2,272 | {
"abstract": "Synthetic biology combines engineering and biology to produce artificial systems with programmable features. Specifically, engineered microenvironments have advanced immensely over the past few decades, owing in part to the merging of materials with biological mimetic structures. In this review, we adapt a traditional definition of community ecology to describe \"cellular ecology\", or the study of the distribution of cell populations and interactions within their microenvironment. We discuss two exemplar hydrogel platforms: (1) self-assembling peptide (SAP) hydrogels and (2) Poly(ethylene) glycol (PEG) hydrogels and describe future opportunities for merging smart material design and synthetic biology within the scope of multicellular platforms."
} | 188 |
30913215 | PMC6435174 | pmc | 2,273 | {
"abstract": "Fe(III)-rich deposits referred to as “iron mounds” develop when Fe(II)-rich acid mine drainage (AMD) emerges at the terrestrial surface, and aeration of the fluids induces oxidation of Fe(II), with subsequent precipitation of Fe(III) phases. As Fe(III) phases accumulate in these systems, O 2 gradients may develop in the sediments and influence the distributions and extents of aerobic and anaerobic microbiological Fe metabolism, and in turn the solubility of Fe. To determine how intrusion of O 2 into iron mound sediments influences microbial community composition and Fe metabolism, we incubated samples of these sediments in a column format. O 2 was only supplied through the top of the columns, and microbiological, geochemical, and electrochemical changes at discrete depths were determined with time. Despite the development of dramatic gradients in dissolved Fe(II) concentrations, indicating Fe(II) oxidation in shallower portions and Fe(III) reduction in the deeper portions, microbial communities varied little with depth, suggesting the metabolic versatility of organisms in the sediments with respect to Fe metabolism. Additionally, the availability of O 2 in shallow portions of the sediments influenced Fe metabolism in deeper, O 2 -free sediments. Total potential ( E H + self-potential) measurements at discrete depths in the columns indicated that Fe transformations and electron transfer processes were occurring through the sediments and could explain the impact of O 2 on Fe metabolism past where it penetrates into the sediments. This work shows that O 2 availability (or lack of it) minimally influences microbial communities, but influences microbial activities beyond its penetration depth in AMD-derived Fe(III) rich sediments. Our results indicate that O 2 can modulate Fe redox state and solubility in larger volumes of iron mound sediments than only those directly exposed to O 2 .",
"introduction": "Introduction Centuries of coal extraction in the Appalachian region of the United States has left a legacy of acid mine drainage (AMD), which remains the region’s greatest threat to surface water quality [ 1 ]. The major environmental damage caused by AMD occurs when the anoxic, acidic, and Fe(II)-rich fluid enters oxic, circumneutral streams, whereupon the higher pH enhances the oxidation of Fe(II) and precipitation of Fe(III) phases, which smother stream substrates and limit the development of robust stream ecosystems (e.g. algae, macroinvertebrates, fish; [ 2 – 4 ]). As such, removal of dissolved Fe(II) is the most pressing objective in AMD treatment and stream restoration activities [ 5 ]. In some cases, AMD flows as a 0.5–1 cm sheet over the terrestrial surface, resulting in aeration of the fluid and enhanced activities of Fe(II) oxidizing bacteria (FeOB; [ 6 – 11 ]). Continuous flow of AMD and sustained Fe(II) oxidation gives rise to massive Fe(III) (hydr)oxide deposits that are referred to as “iron mounds” or “iron terraces,” and can grow to thicknesses of meters [ 6 , 7 , 12 ]. While these iron mounds damage the soil and surficial systems that they cover, they may also be exploited for the treatment of AMD [ 6 – 11 ]. Under this scenario, the iron mounds represent iron removal systems, whereby the activities of FeOB induce oxidative removal of Fe from the AMD before the water enters nearby streams [ 6 – 11 ]. Notably, the iron mounds that we have encountered arise with little or no human intervention, suggesting that constructed iron mounds could serve as inexpensive and sustainable approaches to AMD treatment. As these iron mounds grow upward, FeOB are buried in the Fe(III) (hydr)oxide phases that they produce. The dynamics of iron mound development lead to the continuous upward movement of the air-water interface, and the potential development of anoxic portions of the iron mound [ 12 , 13 ]. We have noted unusual dynamics of Fe(II) oxidation and Fe(III) reduction in iron mounds, where Fe(III) reduction appears to occur in the presence of abundant O 2 , while Fe(II) oxidation might occur under conditions of severe O 2 depletion [ 12 , 14 ]. This observation may be at least partially attributable to the metabolic versatility of the acidophilic Fe-metabolizing microorganisms that inhabit the iron mounds, which are capable of Fe(II) oxidation and Fe(III) reduction (e.g. [ 15 – 18 ]). Indeed, the microbial communities associated with the iron mounds are remarkably uniform with depth [ 12 ]. Anaerobic activities in these iron mounds are important, because they represent a mechanism for remobilization of Fe that had been previously oxidatively precipitated—an undesirable process in the context of AMD treatment [ 13 ]. However, the distributions of anaerobic activities have proven to be difficult to predict, given their (at least partial) independence from O 2 availability. To assess relationships between O 2 availability and Fe(III) reduction and microbial community dynamics associated with aerobic and anaerobic processes in an iron mound setting, we incubated initially homogenized sediments from an iron mound in North Lima, OH (referred to as the Mushroom Farm) in a column format. During these incubations we assessed Fe(II) oxidation and Fe(III) reduction, as well as the associated electrochemical signatures at discrete depths over the course of incubation in the columns. At the conclusion of the incubations, the extents of O 2 penetration into the columns was assessed, and the microbial community composition at various depths within the columns was determined.",
"discussion": "Discussion O 2 availability, controlled by depth in the columns, minimally influenced the composition of microbial communities in iron mound sediments, but profoundly influenced their activities. The only phylotypes that exhibited a substantial change with depth at the conclusion of the column incubations were unassigned sequences that were similar to planktonic phylotypes observed in acidic (pH approximately 2) and high redox potential (approximately 470 mV) Rio Tinto, indicating that these organisms metabolize optimally under mostly oxic conditions [ 35 , 38 ]. Otherwise, microbial communities throughout the remainder of the sediments were nearly identical ( Fig 2E and S2 Fig ). 16S rRNA gene-based surveys can still detect inactive organisms, which could explain the compositional similarities we have observed. In previous experiments at the Mushroom Farm, we have observed discernable shifts in microbial communities over shorter incubation times [ 9 ]. Additionally, a similar pattern of microbial community composition was observed in intact iron mound microbial communities [ 12 ]. In that case, relative abundances of phylotypes attributable to photosynthetic microeukaryotes and obligately aerobic, Fe(II) oxidizing Gallionella sp. diminished in deeper portions of the sediments, but other components of the microbial communities retained similar relative abundances [ 12 ]. The most notable constants in situ were Gammaproteobacteria assignable to Fe-metabolizing Xanthomonadaceae [ 12 ], which also remained abundant at the conclusion of our column incubations ( Fig 2E ). These observations illustrate the metabolic versatility of microorganisms with respect to Fe metabolism in AMD and AMD-impacted systems. They are often capable of Fe(II) oxidation and Fe(III) reduction, depending to some extent (but not completely) on the availability of O 2 [ 14 , 37 , 39 – 41 ]. In the current work, we started with a homogenized microbial community from the upper 6 cm of an iron mound and challenged that community to adjust to limitations on O 2 delivery. The communities did not vary dramatically from a compositional perspective, but exhibited dramatic differences in their activities. Despite the consistency in community composition, the microbial activities over the course of the incubations were dramatically different at different depths, with extensive Fe(III) reduction in the deeper portions of the columns (Figs 1 and 2 ). O 2 was completely depleted from the sediments at depth where Fe(III) reduction did not occur to its maximal extent ( Fig 2A and 2C ). In other words, the extent of Fe(II) accumulation (indicative of Fe(III) reduction) followed a gradient that was not dependent on O 2 availability: less Fe(III) reduction was apparent at 57 mm than at 97 mm, despite complete depletion of O 2 at 33 mm (Figs 1B and 2A and 2C ). Similarly, addition of O 2 to the initially anoxic incubations arrested Fe(III) reduction in a depth-dependent manner, and not exclusively in the shallower sediments ( Fig 1C ). It is not clear if Fe(II) oxidation was occurring in the anoxic sediments or if extremely low O 2 concentrations (i.e. below the detection limit of 0.3 μM) are supporting extensive Fe(II) oxidation [ 42 , 43 ]. It appears unlikely Fe(III) reduction was simply partially inhibited in the shallower sediments, because addition of air to the initially anoxic incubations led to Fe(II) oxidation in deeper sediments ( Fig 1C ). Some insight into the conditions that could allow O 2 to influence Fe(III) reduction or Fe(II) oxidation despite separation of these two species can be gained from examination of our electrochemical measurements. Electrochemical or geophysical approaches are increasingly deployed to interrogate (bio)geochemical processes in field settings and evaluate spatial distributions of microbiologically-induced redox processes [ 44 – 47 ]. The redox potential ( E H ) of a given solution is the potential between a non-polarizable reference electrode and a polarizable electrode in close proximity to each other and is indicative of the capacity for a solution to accept or donate electrons relative to the standard hydrogen electrode (SHE [ 48 ]). The self-potential (SP), which is widely used in geophysical surveys, represents the potential difference between two spatially-separated non-polarizable electrodes (one stationary, and one movable) and is indicative of electric current between relatively reducing and oxidizing regions or an electrochemical gradient [ 44 , 48 – 52 ]. If a stationary non-polarizable electrode is deployed with a movable polarizable electrode, the resulting potential is referred to as total potential (TP), and represents the sum of the E H and SP between the two electrodes [ 48 ]. By deploying PtIr electrodes along the column coupled with a Ag/AgCl reference electrode in the overlying SAMD, our measurements constitute the TP at various depths within the columns. Values shown in Fig 1G–1I represent the theoretical E H for the sediments based on the Fe 2+ /Fe 3+ couple using Fe(II) concentrations from the respective experiments. It is likely that O 2 also contributes to the E H component of the TP. For instance, O 2 was relatively high throughout the deactivated column sediments ( Fig 2C ), so the H 2 O/O 2 redox couple could have influenced TP in addition to the Fe 2+ /Fe 3+ couple. Indeed, TP were higher than predicted based on the Fe 2+ /Fe 3+ redox couple ( Fig 1D and 1G ). However, at the DO in the deactivated columns at the conclusion of the incubations, the theoretical E H was 0.96 V. Therefore, while O 2 clearly contributed to TP in regions where it was present, it appears that the Fe 2+ /Fe 3+ redox couple exerted the most control on the E H component of TP throughout the columns. Additionally, since we could not detect evidence of sulfate reduction in these incubations, the Fe 2+ /Fe 3+ redox couple would predominantly drive E H in the sediments. The measured TP do not necessarily constitute E H , since the non-polarizable and polarizable electrodes are spatially separated from each other, but the calculated E H (based on Fe(II) concentration) and measured TP of the formaldehyde-deactivated incubations match reasonably well ( Fig 1D and 1G ), as do the E H and TP during anoxic incubations (gray-shaded part of Fig 1F and 1I ). Qualitatively, in both of the non-sterile incubations, shallower sediments, with lower Fe(II) concentrations and greater rates and extents of Fe(II) oxidation exhibited higher TP and E H ( Fig 1B, 1C, 1E, 1F, 1H and 1I and Fig 2A and 2B ). The higher E H is consistent with lower Fe(II) concentration, and perhaps higher dissolved Fe 3+ concentration. For instance, the higher-than-predicted TP in overlying AMD and shallow sediments ( Fig 1E and 1H ) could be attributable to accumulation of Fe 3+ exceeding its maximum solubility in the oxic portions of the columns where Fe(II) oxidation is most robust. Overall, these observations illustrate the contribution of redox potential, as controlled by the Fe 2+ /Fe 3+ redox couple, to the TP measured in these incubations. E H calculations did not predict the continuous increase in TP at all depths in the incubations after approximately 20 d ( Fig 1E and 1H ). They also did not predict the increase in TP upon addition of O 2 to initially anoxic incubations ( Fig 1F and 1I ). If based exclusively on Fe(II) concentration, these observed TP would predict a decrease in Fe(II) concentration, which was not the case. In fact, Fe(II) concentrations segregated further with depth as the TP increased ( Fig 1B and 1E ). An explanation for the observed increase in TP in anoxic incubations could be opposing gradients of Fe 2+ and Fe 3+ , where a high rate of Fe(II) oxidation in shallow, oxic portions of the column induced the Fe(II) gradients that we observed, while downward diffusion of Fe 3+ from the oxic zone to the oxic zone induced the increase in TP. However, the rate of Fe(III) reduction during the anoxic period of the incubations (3.7 mM/d, based on Fe(II) accumulation) exceeded the rate of Fe(II) oxidation in the shallowest sediments (0.97 mM/d, based on Fe(II) depletion) after O 2 was allowed into the columns ( Fig 1F ). This pattern of Fe(II) oxidation and Fe(III) reduction rates would result in a steep Fe(II) gradient near the oxic-anoxic interface, and not the gradual Fe(II) gradient from the top to the bottom of the column that we observed here ( Fig 2 ). While TP and predicted E H closely matched during the anoxic period of the short-term incubations, upon introduction of O 2 to the headspace, the TP increased at all depths, and Fe(II) concentrations segregated based on depth in the columns ( Fig 1C, 1F and 1I ). These inconsistencies between the TP and E H when O 2 is available at the top of the columns can be attributed to the SP contribution to TP [ 48 ], and suggest an electron transfer process occurring in the sediments due to the electrochemical pull of O 2 overlying the sediments. Both field- and laboratory-scale electrochemical/geophysical surveys of SP signals have illustrated the development of SP signals across regions that connect high and low E H regions as we have observed here [ 45 , 51 , 52 ]. In order to facilitate the electron transfer that gives rise to SP signals in sediments, it has been suggested that a perhaps disorderly, but integrated network of microorganisms, extracellular material, and redox-active solid phases gives rise electron transfer [ 45 , 52 , 53 ]. Such a model could function quite well in iron mound settings, as the sediments are composed almost exclusively of Fe(III) (hydr)oxide phases [ 6 , 7 , 10 , 12 ], and these phases could facilitate the electron transfer processes [ 54 – 64 ], with opposing Fe(II) and O 2 concentration gradients providing the driving force for electron transfer. Our results are consistent with previous field and laboratory observations of a gap between intrusion of O 2 into the sediments and the zone of Fe(II) oxidation, where an O 2 intrusion front and Fe(II) oxidation zone were spatially separated [ 12 , 14 ]. This work has allowed us to visualize the microbially-mediated development of these gradients in the iron mound sediments and apply electrochemical approaches to assess biogeochemical processes within the sediments. Our results indicate that the chemical and microbiological influence of O 2 in iron mound sediments exceeds its actual penetration into the sediments. Notably, Fe(II) accumulation in deeper sediments was suppressed despite no O 2 availability. If engineered iron mounds are to be used for oxidative precipitation and removal of Fe(II) from AMD [ 8 – 10 ], our results indicate that the longer range influence of O 2 into the sediments could minimize reductive re-release of Fe(II) from the sediments."
} | 4,127 |
37684613 | PMC10492371 | pmc | 2,274 | {
"abstract": "Background Cyanobacteria are emerging as green cell factories for sustainable biofuel and chemical production, due to their photosynthetic ability to use solar energy, carbon dioxide and water in a direct process. The model cyanobacterial strain Synechocystis sp. PCC 6803 has been engineered for the isobutanol and 3-methyl-1-butanol production by introducing a synthetic 2-keto acid pathway. However, the achieved productions still remained low. In the present study, diverse metabolic engineering strategies were implemented in Synechocystis sp. PCC 6803 for further enhanced photosynthetic isobutanol and 3-methyl-1-butanol production. Results Long-term cultivation was performed on two selected strains resulting in maximum cumulative isobutanol and 3-methyl-1-butanol titers of 1247 mg L −1 and 389 mg L −1 , on day 58 and day 48, respectively. Novel Synechocystis strain integrated with a native 2-keto acid pathway was generated and showed a production of 98 mg isobutanol L −1 in short-term screening experiments. Enhanced isobutanol and 3-methyl-1-butanol production was observed when increasing the kivd S286T copy number from three to four. Isobutanol and 3-methyl-1-butanol production was effectively improved when overexpressing selected genes of the central carbon metabolism. Identified genes are potential metabolic engineering targets to further enhance productivity of pyruvate-derived bioproducts in cyanobacteria. Conclusions Enhanced isobutanol and 3-methyl-1-butanol production was successfully achieved in Synechocystis sp. PCC 6803 strains through diverse metabolic engineering strategies. The maximum cumulative isobutanol and 3-methyl-1-butanol titers, 1247 mg L −1 and 389 mg L −1 , respectively, represent the current highest value reported. The significantly enhanced isobutanol and 3-methyl-1-butanol production in this study further pave the way for an industrial application of photosynthetic cyanobacteria-based biofuel and chemical synthesis from CO 2 . Supplementary Information The online version contains supplementary material available at 10.1186/s13068-023-02385-1.",
"conclusion": "Conclusions This study explicitly explored the 2-keto acid pathway for photosynthetic isobutanol (IB) and 3-methyl-1-butanol (3M1B) production in Synechocystis sp. PCC 6803. Enhanced IB and 3M1B production was observed after increasing kivd S286T copy number, indicating α-ketoisovalerate decarboxylase as a rate-limiting enzyme. Moreover, overexpression of five gene targets of the central carbon metabolism effectively increased IB and 3M1B production, which are potential targets for overexpression to enhance any pyruvate-derived bioproduction. In the end, the maximum cumulative IB and 3M1B titers, 1247 mg L −1 and 389 mg L −1 , obtained by strains HX29 and HX42, respectively, represent the currently highest reported.",
"introduction": "Introduction In 2020, fossil resources supplied approximately 81% of total energy, whereas renewable resources accounted for approximately 15% of the total energy [ 1 ]. By 2050, the global energy demand is projected to increase by 47% [ 1 ]. In face of the rapid climate change and increasing energy demand, it is urgent to gradually replace traditional fossil resources with renewable energy, such as biofuels produced, e.g., by metabolically engineered microorganisms feeding on renewable carbon sources [ 2 , 3 ]. Being generated from renewable resources, biofuels are cleaner energy as they release lower amounts of sulfates and black carbon particulates after burning [ 4 ]. Currently, bioethanol, mainly produced from biomass fermentation using sugarcane and corn as feedstocks, is the main biofuel used as gasoline additive. However, the energy density of ethanol is only 66% of gasoline, making it less favorable as gasoline additive compared to advanced alcohols. Isobutanol (IB), a four-carbon advanced alcohol, is recognized as a superior substitution as drop-in fuel, due to the following advantages: higher energy density, lower water solubility, lower vapor pressure and lower hygroscopicity compared to ethanol [ 5 ]. The boiling point and melting point of IB are + 108 °C and -108 °C. Moreover, IB and water form a heterogeneous azeotrope and protocols for separation by distillation are available [ 6 ]. Biological IB production was first demonstrated in Escherichia coli ( E. coli ) by introduction of a synthetic 2-keto acid pathway [ 7 ]. The 2-keto acid pathway involves five enzymes for IB biosynthesis from the central metabolite pyruvate (Fig. 1 ). Within the 2-keto acid pathway, the first involved enzyme, acetolactate synthase (AlsS), condenses two pyruvate molecules into a 2-acetolactate molecule. The 2-acetolactate is further converted to 2-ketoisovalerate by sequential enzymatic reactions catalyzed by acetohydroxy-acid isomeroreductase (IlvC) and dihydroxy-acid dehydratase (IlvD). As an intermediate for valine and leucine biosynthesis, 2-ketoisovalerate is decarboxylated by a heterologously expressed broad-substrate-range α-ketoisovalerate decarboxylase (Kivd) to isobutyaldehyde, and subsequently reduced into IB by an alcohol dehydrogenase (Adh). On the basis of the first report, the same strategy was applied in various microorganisms for IB biosynthesis [ 8 , 9 ]. Meanwhile, due to the existence of native leucine biosynthesis pathway, 2-ketoisovalerate is converted into ketoisocaproate by sequential enzymes, encoded by leuABCD . The resulting ketoisocaproate is decarboxylated and reduced into 3-methyl-1-butanol (3M1B), by Kivd and Adh (Fig. 1 ). Similar to IB, 3M1B is a superior candidate for gasoline additive and is widely used as a precursor for various chemical synthesis [ 10 ]. Fig. 1 Isobutanol (IB) and 3-methyl-1-butanol (3M1B) biosynthesis pathway. Carbon dioxide is fixed by the Calvin–Benson–Bassham (CBB) cycle, and the fixed carbon flows into the 2-keto acid pathway for IB and 3M1B biosynthesis. Endogenous enzymes are written in black, while heterologous enzymes are written in red. Abbreviations of enzymes: Sll0065, small subunit of native acetolactate synthase (AlsS); Slr2088, large subunit of native AlsS; Sll1363, native acetohydroxy-acid isomeroreductase (IlvC); Slr0452, native dihydroxy-acid dehydratase (IlvD); LeuA, 2-isopropylmalate synthase; LeuCD, 3-isopropylmalate dehydratase; LeuB, 3-isopropylmalate dehydrogenase; Kivd S286T , α-ketoisovalerate decarboxylase ( Lactococcus lactis ); Slr1192 OP , codon-optimized native alcohol dehydrogenase. Dotted lines indicate multiple reactions Different from heterotrophic microorganisms feeding on substrates generated from plant biomass, photosynthetic microorganisms, including cyanobacteria, are capable to use sunlight and carbon dioxide for biofuel synthesis in a direct process. In that regard, the 2-keto acid pathway was successfully introduced into cyanobacteria for IB biosynthesis, first reported in Synechococcus elongatus PCC 7942 [ 11 ]. Thereafter, another model cyanobacterial strain, Synechocystis sp. PCC 6803 ( Synechocystis ) was demonstrated to have the ability to produce IB after a single heterologous expression of Kivd, originating from Lactococcus lactis [ 12 , 13 ]. Furthermore, a by-product 3M1B was produced simultaneously with Kivd expression [ 12 ]. Protein engineering was performed on the key enzyme Kivd and a single replacement of Serine286 with Threonine significantly improved the Kivd activity further contributing towards an improved IB and 3M1B production [ 14 ]. This engineered Kivd S286T has been used throughout following studies. In a more recent study, photosynthetic IB production was further enhanced by either increased Kivd S286T expression level or integration of a complete 2-keto acid pathway [ 15 ]. Even with substantial progress reported on 2-keto acid pathway for photosynthetic IB and 3M1B production, the achieved production is still far behind to that of heterotrophic microorganisms [ 10 ] or cell-free system using a synthetic biochemistry approach [ 16 ]. Due to the low IB concentration in the cultivation broth and its azeotropic nature, downstream IB separation and purification require specific equipment with high energy consumption. Rectification is currently a widely used method for IB separation and purification [ 17 ]. Additional methods are available for separation of IB from the cultivation broth, such as gas stripping, pervaporation, vacuum evaporation, absorption, solvent extraction, salting-out and salting-out extraction [ 17 ]. In the present work, selected approaches were employed on the cyanobacterial strain Synechocystis to extensively explore the 2-keto acid pathway for IB and 3M1B biosynthesis. Firstly, two selected strains, HX29 and HX42, were cultivated continuously for 60 days in long-term milking experiments to explore their full capacities of IB and 3M1B biosynthesis. Secondly, additional efforts were invested to address if Kivd S286T is still the bottleneck restraining further improvement of IB and 3M1B production by modifying the kivd S286T copy number. Thirdly, instead of overexpressing heterologous enzymes, selected native enzymes involved in the valine/leucine biosynthesis were overexpressed to explore their effects on IB and 3M1B biosynthesis. Lastly, selective overexpression of genes involving in central carbon metabolism was experimentally verified to have positive contributions towards IB and 3M1B biosynthesis through the 2-keto acid pathway in Synechocystis . The collectively acquired information in this study further guides metabolic engineering strategies towards photosynthetic pyruvate-derived bioproduction.",
"discussion": "Results and discussion Long-term milking experiments of Synechocystis sp. PCC 6803 strains HX29 and HX42 HX29 [ 15 ] is an engineered Synechocystis sp. PCC 6803 ( Synechocystis ) strain containing three copies of kivd S286T : one copy integrated into the ddh site of Synechocystis chromosome; the second copy integrated into the sll1564 site; and the third copy placed on a self-replicating plasmid (Fig. 2 A). In our previous study, strain HX29 showed the highest isobutanol (IB) production per cell among numerous IB-producing Synechocystis strains [ 15 ]. Therefore, long-term cultivation was performed on strain HX29 to characterize its full capacity of IB and 3-methyl-1-butanol (3M1B) production. The whole cultivation period lasted for sixty days, and the maximum optical density (OD 750 ) of the experimental culture reached 5.95 on day 13 (Fig. 2 B). The highest in-flask IB and 3M1B titers obtained of strain HX29 were 536.9 mg L −1 and 138.7 mg L −1 , respectively, on day 48 (Fig. 2 B). After day 48, the measured in-flask titers for IB and 3M1B started to decrease (Fig. 2 B). In the end of the 60-day cultivation, the cumulative titers of using HX29 were 1247 mg L −1 and 326 mg L −1 for IB and 3M1B, respectively (Fig. 2 B). By dividing the whole cultivation into six stages, the IB and 3M1B production rate is summarized in Table 2 . In consistence with previous report [ 20 ], the highest in-flask production rate was observed in Stage I, corresponding to exponential phase, which is significantly higher than the other stages (Table 2 ). The highest cumulative production rate observed was in Stage II for both IB and 3M1B (Table 2 ). Fig. 2 Long-term milking experiments of engineered Synechocystis sp. PCC 6803 strains HX29 and HX42. A Schematic diagram of plasmids used for generating strain HX29. kivd S286T : encodes α-ketoisovalerate decarboxylase ( Lactococcus lactis ). B Growth profile, isobutanol (IB) and 3-methyl-1-butanol (3M1B) in-flask and cumulative titers of strain HX29. C Schematic diagram of plasmids used for generating strain HX42. kivd S286T , encodes α-ketoisovalerate decarboxylase ( L. lactis ); alsS , encodes acetolactate synthase ( Bacillus subtilis ); ilvC , encodes acetohydroxy-acid isomeroreductase ( Escherichia coli ); ilvD , encodes dihydroxy-acid dehydratase ( E. coli ); slr1192 OP , encodes codon-optimized alcohol dehydrogenase ( Synechocystis ). D Growth profile, IB and 3M1B in-flask and cumulative titers of strain HX42. Results are the mean of four biological replicates, each with three technical replicates. Error bars represent standard deviation Table 2 Isobutanol (IB) and 3-methyl-1-butanol (3M1B) production rates of engineered Synechocystis sp. PCC 6803 strains HX29 and HX42 HX29 HX42 Growth stage IB (mg L −1 day −1 ) 3M1B (mg L −1 day −1 ) IB (mg L −1 day −1 ) 3M1B (mg L −1 day −1 ) In-flask Cumulative In-flask Cumulative In-flask Cumulative In-flask Cumulative Stage I (day 0–10) 25.1 ± 0.6 28.0 ± 0.6 6.4 ± 0.1 7.2 ± 0.1 24.1 ± 1.0 27.1 ± 1.0 6.8 ± 0.1 7.7 ± 0.1 Stage II (day 10–20) 16.6 ± 0.8 29.8 ± 1.0 3.9 ± 0.2 7.3 ± 0.2 14.0 ± 2.2 25.6 ± 2.6 5.8 ± 0.6 9.5 ± 0.5 Stage III (day 20–30) 3.1 ± 0.3 20.5 ± 0.4 0.8 ± 0.1 5.1 ± 0.1 3.2 ± 2.4 18.9 ± 4.2 1.1 ± 1.3 6.4 ± 1.8 Stage IV (day 30–40) 1.5 ± 0.7 20.3 ± 0.8 0.9 ± 0.2 5.6 ± 0.2 3.5 ± 1.3 21.3 ± 3.3 1.4 ± 0.4 7.3 ± 1.1 Stage V (day 40–50) 2.4 ± 2.4 21.2 ± 2.9 0.9 ± 0.6 5.9 ± 0.8 1.4 ± 6.3 20.8 ± 10.1 0.4 ± 2.0 6.8 ± 3.2 Stage VI (day 50–60) − 12.8 ± 0.6 4.2 ± 0.9 3.2 ± 0.1 1.3 ± 0.3 − 18.7 ± 5.9 − 3.3 ± 1.7 − 6.3 ± 1.9 − 1.1 ± 0.5 The highest IB and 3M1B in-flask and cumulative production rate of each strain is shown in bold. Results are the mean of four biological replicates, each with three technical replicates. Errors represent standard deviation In parallel, HX42 [ 15 ], a strain with a complete 2-keto acid pathway integrated, was cultivated under the same condition as strain HX29. The engineered strain HX42 contains the following genetic modifications: slr1192 OP and alsS integrated into the ddh site of Synechocystis chromosome; ilvC and ilvD integrated into the slr0168 site; and kivd S286T placed on a self-replicating plasmid (Fig. 2 C). The growth curve showed a maximum OD 750 of 6.77 on day 10 (Fig. 2 D). The highest in-flask titers of IB and 3M1B of strain HX42 observed were 515.4 mg L −1 and 168.2 mg L −1 , respectively (Fig. 2 D). The final resulting cumulative IB and 3M1B titers of strain HX42 were 1155 mg L −1 and 389 mg L −1 , respectively (Fig. 2 D). Both strains HX29 and HX42 achieved significantly higher cumulative IB and 3M1B titers compared to the previously best-performing strain [ 20 ] under the same cultivation conditions (Fig. 2 B, D). The new records of IB and 3M1B cumulative titers were improved 1.4-fold and 1.7-fold by HX29 and HX42, respectively. A comprehensive comparison was performed between strains HX29 and HX42 from growth pattern to IB and 3M1B production. Strain HX42 grew faster in Stage I, while the OD 750 declined faster after day 10 (Fig. 2 B, D). Interestingly, strain HX42 showed lower in-flask/cumulative IB titer but higher in-flask/cumulative 3M1B titer (Fig. 2 B, D). Different from HX29, the highest IB cumulative production rate of HX42 was observed in Stage I (Table 2 ), which may result from the different growth pattern between the two strains. As noted, a relatively large error bar of production curve indicates a relatively large variation among biological replicates of strain HX42. Among the four biological replicates, one of them grew much better than the rest and kept higher optical density (OD 750 ) for a longer time-period, resulting in final cumulative titers of IB and 3M1B up to 1449 mg L −1 and 469 mg L −1 . One possible variation source leading to the varied growth and IB and 3M1B production may be the unavoidable variation in the HCl titration procedures, as culture pH was controlled manually by acid titration and monitored through color indication of MColorpHast™ pH-indicator strips, making it difficult to maintain a precisely controlled culture pH among biological replicates. For further improvement, a controlled cultivation system is an approach to achieve even higher IB and 3M1B titers. With a photobioreactor system equipped a pH controller, it will be possible to achieve maximum carbon assimilation efficiency and optimal growth rate. Moreover, considering that the highest production rate for IB and 3M1B was observed between day 0–20 during the long-term cultivation (Table 2 ), the second approach to further enhance products titer is to employ a re-inoculation strategy [ 21 ] with a cycle of e.g., 20 days, aiming to maintain in highest production rate throughout the cultivation period. In addition, final optimization could be achieved by testing various cultivation parameters, such as light intensity and quality, CO 2 feeding and amount. Short-term screening experiments of newly constructed engineered Synechocystis sp. PCC 6803 strains Generating engineered Synechocystis sp. PCC 6803 strains containing a complete native 2-keto acid pathway In our previous report [ 15 ], we successfully constructed the engineered Synechocystis strain HX42 containing a complete 2-keto acid pathway consisting of four foreign enzymes and one native enzyme: acetolactate synthase (AlsS) from Bacillus Subtilis , acetohydroxy-acid isomeroreductase (IlvC) and dihydroxy-acid dehydratase (IlvD) from Escherichia coli ( E. coli ), α-ketoisovalerate decarboxylase (Kivd S286T ) from Lactococcus lactis , and a codon-optimized alcohol dehydrogenase (Slr1192 OP ) from Synechocystis . In Synechocystis , it is still not settled which gene(s) encode(s) the native AlsS [ 20 ], though it was reported that three endogenous genes are potential candidates: sll0065 encodes the regulatory subunit, while slr2088 and sll1981 encode the catalytic subunits [ 22 ]. The protein sequence of Sll1981 shares about 60% homology to AlsS from B. subtilis [ 23 ], while slr2088 - and sll0065 -encoded proteins are homologous to acetohydroxy-acid synthase (AHAS) [ 24 ]. Native IlvC and IlvD are encoded by sll1363 and slr0452 , respectively. Initially, different engineered strains with different combinations of native AlsS subunits were planned (data not shown), however, it was challenging to obtain correct transformants for most of them even after several attempts. In the end, only three engineered Synechocystis strains with a complete native 2-keto acid pathway (Fig. 1 ) were generated using four integrative plasmids and one self-replicating plasmid (Fig. 3 A). Plasmid P1 was used to overexpress slr1192 OP and sll0065 under the P trc promoter, while simultaneously knocking out the ddh gene. Likewise, plasmid P2 was used to overexpress slr1192 OP , slr2088 , and sll0065 . Plasmids P3 and P4 were used to integrate sll1363 and slr0452 into the slr0168 site under the control of P trc and P psbA2 promoters, respectively. Plasmid P5 was a broad-host-range self-replicating plasmid for kivd S286T expression. Fig. 3 Generation and analysis of engineered Synechocystis sp. PCC 6803 strains with a complete native 2-keto acid pathway integration. A Schematic diagram of plasmids used to generate strains in Fig. 3B-E. P1 and P2 are integrative plasmids targeting ddh ( slr1556 ) site of Synechocystis chromosome. P3 and P4 are integrative plasmids targeting slr0168 site. P5 is a self-replicating plasmid. B The IB titer and IB production per cell on day 10 of engineered Synechocystis strains HX88, HX89 and HX91. Strain HX88 was generated by transformation with plasmids P1, P4 and P5; strain HX89 was generated by transformation with plasmids P1, P3 and P5; strain HX91 was generated by transformation with plasmids P2, P3 and P5. C The IB and 3M1B titers on day 10 of engineered Synechocystis strains HX88, HX89 and HX91. D SDS-PAGE (top) and Western-immunoblot (bottom). L, ladder (in kDa). For SDS-PAGE, 10 μg of total soluble proteins were loaded for each strain. For Western-immunoblot, 10 μg, 20 μg, and 20 μg of total soluble proteins were loaded for each strain to detect Strep-tagged, Flag-tagged, and His-tagged proteins, respectively. Protein size: Kivd S286T , 61 kDa; Sll1363, 40 kDa; Slr0452, 59 kDa; Slr1192 OP , 36 kDa. E Growth profile of engineered Synechocystis strains HX88, HX89 and HX91. Results are the mean of three biological replicates, each with three technical replicates. Error bars represent standard deviation. Asterisk represents significant difference between strains HX88 and HX89, or strains HX89 and HX91 (one-way ANOVA, * p < 0.05, ** p < 0.005) Strain HX88, with a complete native 2-keto acid pathway integrated, produced 98 mg IB L −1 on day 10 (Fig. 3 B, C), which is comparable to the IB titer obtained when using strain HX42 [ 15 ], indicating the native enzymes are as effective as the foreign enzymes for IB biosynthesis. All overexpressed proteins in HX88 were confirmed by SDS-PAGE/Western-immunoblot, except for Sll0065 (Fig. 3 D). It was observed that co-overexpressing IlvC and IlvD is a potential approach to channel more carbon flux into 2-keto acid pathway for IB production [ 20 ]. To further increase protein expression level, the promoter used to drive Sll1363 and Slr0452 expression was changed from P psbA2 to the stronger P trc [ 25 , 26 ], resulting in the engineered strain HX89 (Fig. 3 A, B). As expected, the expression levels of Sll1363 and Slr0452 in HX89 increased compared to in the control strain HX88 (Fig. 3 D). However, both IB and 3M1B titers were significantly lower in HX89 (Fig. 3 C). As one possible explanation, the decreased titers may be due to the slower growth rate between days 0–5 (Fig. 3 E), the time-period when majority of IB and 3M1B are produced. On the other hand, the observed IB and 3M1B titer difference demonstrates that the expression levels of Sll1363 and Slr0452 in strain HX88 are enough and not bottlenecks of the 2-keto acid pathway for IB and 3M1B biosynthesis. Furthermore, as recently commented [ 27 ], the large catalytic subunit Slr2088 may form a complex with the small regulatory subunit Sll0065 to function as native AlsS in Synechocystis . However, this needs further validation. After several attempts, an engineered Synechocystis strain with co-overexpression of both the regulatory subunit and the catalytic subunit was generated (HX91). Compared to strain HX89, strain HX91 had a distinct growth pattern with a longer lag phase, barely any growth between days 0–2 (Fig. 3 E). Thereafter, it caught up and reached a higher optical density (OD 750 = 4.6) on day 8 (Fig. 3 E). Unfortunately, the IB and 3M1B titers of HX91 were lower than that of HX89 (Fig. 3 C), indicating that overexpressing Slr2088 has reverse effects on IB and 3M1B production. The mechanism of Slr2088 overexpression negatively affecting IB and 3M1B production is currently unknown. High-throughput-omics approaches [ 28 , 29 ] may provide new insights to reveal this observation. As shown in Fig. 3 D, similar to HX88, the native AlsS was not detected by SDS-PAGE/ Western-immunoblot in strains HX89 and HX91. Several hypothesize could be made: sll0065 and slr2088 were successfully transcribed into mRNAs and were further translated to functional proteins, but their expression was too low to be detected by Western-immunoblot; or sll0065 and slr2088 were successfully transcribed into mRNAs, which were not translated into functional proteins due to some unknown native regulation; or the native alsS genes were expressed in specific growth phase and specific growth conditions, and unfortunately at the time of cell harvesting, there was no expression of native AlsS [ 23 ]. RT-PCR on the overexpressed genes encoding native AlsS will be the first step to validate the above hypotheses, and for further characterization, native AlsS could be defined by using a newly developed technique [ 30 ], followed by kinetic determinations. kivd S286T copy number makes a significant difference for isobutanol and 3-methyl-1-butanol biosynthesis kivd S286T , encoding α-ketoisovalerate decarboxylase, is a verified critical enzyme for IB biosynthesis using Synechocystis as cell factory [ 14 ], and the IB titer improved in a stepwise manner with varied kivd S286T copy number ranging from one copy to three copies [ 15 ]. In the current study, further attempt was pursued to increase kivd S286T copy number. Initially, integrating plasmids P5, P6, P7 and P13 into wild-type Synechocystis resulted in a control strain HX61, containing three copies of kivd S286T (Fig. 4 A, B). P5 was constructed to express the first copy of kivd S286T on self-replicating plasmid; P6 was constructed to integrate a second copy of kivd S286T into the ddh site of Synechocystis chromosome; P7 was constructed to integrate a third copy of kivd S286T into the sll1564 site; P13 was constructed to integrate an erythromycin resistance cassette into the slr0168 site (Fig. 4 A). Then a base strain was generated by integrating plasmids P5, P6 and P12 (Fig. 4 A), which was used to further construct Synechocystis strains with four copies of kivd S286T . Plasmid P12 was constructed to introduce one copy of kivd S286T in the slr0168 site (Fig. 4 A). Engineered strains HX56, HX62, HX63, HX78 and HX80, containing four copies of kivd S286T , were generated by integrating plasmids P7, P9, P8, P10 and P11 into the base strain, respectively (Fig. 4 A, B). Plasmids P7-11 were designed to integrate a fourth copy of kivd S286T into the selected sites of Synechocystis chromosome (Fig. 4 A, C). The rationale of the integration sites selection has been detailed previously [ 15 ]. Fig. 4 Engineered Synechocystis sp. PCC 6803 strains with four copies of kivd S286T significantly improved isobutanol (IB) and 3-methyl-1-butanol (3M1B) titers compared to control strain with three copies of kivd S286T . A Schematic diagram of plasmids used to generate strains in Fig. 4. P5 is a self-replicating plasmid; P6-P13 are integrative plasmids targeting various sites of Synechocystis chromosome. B The relative IB titer of engineered Synechocystis strains HX56, HX61-63, HX78 and HX80. C Simplified pathway for IB and 3M1B biosynthesis. Endogenous enzymatic reactions are written in black; heterologous enzymatic reactions are written in red; knock-out/knock-down enzymatic reactions are written in grey. Abbreviations of enzymes: Kivd S286T , α-ketoisovalerate decarboxylase ( Lactococcus lactis ); PEPc (encoded by sll0920 ), phosphoenolpyruvate carboxylase; PDH (encoded by slr1934 and sll1721 ), pyruvate dehydrogenase E1 component; Ddh (encoded by slr1556 ), D-lactate dehydrogenase; LeuA (encoded by slr0186 and sll1564 ), 2-isopropylmalate synthase. Abbreviations of intermediates: PEP, phosphoenolpyruvate. D The relative 3M1B titer of engineered Synechocystis strains HX56, HX61-63, HX78 and HX80. E Western-immunoblot analysis of all expressed enzymes. L, ladder (in kDa). Ten micrograms, 20 μg, and 20 μg of total soluble proteins were loaded for each strain to detect Strep-tagged, Flag-tagged, and His-tagged proteins, respectively. Protein size: Kivd S286T , 61 kDa. F Growth profile of engineered Synechocystis strains HX56, HX61-63, HX78 and HX80. Results are the mean of three biological replicates, each with three technical replicates. Error bars represent standard deviation. Asterisk represents significant difference between engineered strains and control strain (one-way ANOVA, *p < 0.05, **p < 0.005) Strain HX56, containing one more copy of kivd S286T in the slr0168 site, produced a 1.7-fold increased IB titer compared to the control strain HX61 (Fig. 4 B). Similarly, the 3M1B titer was increased by 1.5-fold (Fig. 4 D). The above results serve as evidence that the IB and 3M1B titers are positively correlated with kivd S286T copy number, in consistence to the facts observed in a previous study [ 15 ]. Kivd S286T expression of both strains was confirmed by Western-immunoblot (Fig. 4 E) and growth profile was generated through measuring optical density (OD 750 ) every day (Fig. 4 F). Strain HX56 grew slower between days 0–4, while maintained a higher OD 750 from day 5 and reached a higher maximum OD 750 at 3.2 on day 6 (Fig. 4 F). To explore if the different integration sites of Synechocystis chromosome will make differences on growth and IB and 3M1B titers, four more strains were generated, named HX62, HX63, HX78 and HX80. As expected, all strains produced significantly higher IB and 3M1B titers than that produced by control strain HX61, except for strain HX80 (Fig. 4 B, D). The fact that the IB and 3M1B titers of HX80 were not significantly improved may be caused by its deficient growth profile (Fig. 4 F). As a continuation study of our previous report [ 15 ], we successfully generated metabolically engineered Synechocystis strains containing four copies of kivd S286T . Taken together, Kivd S286T is still the critical enzyme catalyzing the rate-limiting step of 2-keto acid pathway for IB and 3M1B biosynthesis. Currently, four antibiotics were used to screen for positive transformants containing four copies of kivd S286T , making it infeasible to further increase kivd S286T copy number using traditional transformation approaches, since there is no report using more than four antibiotics for Synechocystis transformants screening and cultivation. As one of the solutions, marker-less genome editing strategy [ 31 – 33 ] may make it possible to generate engineered Synechocystis strains with higher kivd S286T copy number. The marker-less-based CRISPR (clustered regularly interspaced short palindromic repeats) editing has been successfully applied in cyanobacteria for succinate production [ 34 ]. On the other hand, considering the time and efforts required for multiple transformation and selection procedures in marker-less genome editing approaches, instead of generating strains with multiple kivd S286T copies, protein engineering will be a powerful alternative to improve the performance of Kivd S286T enzyme on IB and 3M1B biosynthesis. Currently, Kivd S286T , an engineered version of wild-type Kivd after site-directed mutagenesis [ 14 ], was used throughout this study. Starting from Kivd S286T , directed evolution [ 35 ] may be employed to screen for superior Kivd variants, with further enhanced catalytic activity and/or specificity. Identification of targets in central carbon metabolism for enhanced isobutanol and 3-methyl-1-butanol production Pyruvate is one of the central carbon compounds used as substrate for many cellular metabolite biosynthesis. IB and 3M1B are synthesized through pyruvate-derived 2-keto acid pathway. Apart from focusing on optimizing the 2-keto acid pathway itself, for the first time, various targets of the central carbon metabolism (Fig. 5 A) were systematically evaluated for the effects on IB and 3M1B production in Synechocystis . The detailed information of the engineered strains is shown in Fig. 5 and Table 1 . The genetic constructs designed to generate engineered Synechocystis strains are listed in Fig. 5 B. Strains HX75, HX79 and HX81 serve as control strains. In detail, strain HX75 contains a complete 2-keto acid pathway consisting of four foreign enzymes and one native enzyme, while strains HX79 and HX81 contain three copies of kivd S286T . Fig. 5 Schematic overview of metabolic engineering strategies adopted for isobutanol (IB) and 3-methyl-1-butanol (3M1B) production and the corresponding engineered Synechocystis sp. PCC 6803 strains. A Simplified pathway for IB and 3M1B biosynthesis. Endogenous pathways are written in black; heterologous pathways are written in red; targeting pathway for metabolic engineering are written in blue. Multiple enzymatic reactions are represented as dashed lines. Abbreviations of enzymes: FBA, aldolase (encoded by fbaA ); TK, transketolase (encoded by tktA ); PCK, phosphoenolpyruvate carboxykinase (encoded by pckA ); TPI, triosephosphate isomerase (encoded by tpiA ); PK, pyruvate kinase (encoded by pyk1 ). Abbreviations of intermediates: RuBP, ribulose-1,5-bisphosphate; 3PGA, 3-phosphoglycerate; G3P, glyceraldehyde-3-phosphate; DHAP, dihydroxyacetone phosphate; FBP, fructose-1,6-bisphosphate; SBP, sedoheptulose-1,7-bisphosphate; F6P, fructose-6-phosphate; S7P, sedoheptulose-7-phosphate; E4P, erythrose- 4-phosphate; Xu5P, xylulose-5-phosphate; R5P, ribose-5-phosphate; Ru5P, ribulose-5-phosphate; PEP, phosphoenolpyruvate; OAA, oxaloacetate. B Schematic diagram of plasmids used to generate Synechocystis strains in Fig. 5C and Fig. 6 A–J. P5 is a self-replicating plasmid; P6, and P14-P21 are integrative plasmids targeting various sites of Synechocystis chromosome. C Genetic background of the engineered Synechocystis strains The first two enzymes tested are aldolase (FBA) and transketolase (TK), which are involved in the Calvin–Benson–Bassham (CBB) cycle (Fig. 5 A) and the oxidative pentose phosphate (OPP) or glycolysis pathway. Overexpression of FBA and TK has positive effects on cell growth as well as ethanol production in engineered Synechocystis strains [ 36 , 37 ]. In Synechocystis , both class I and class II FBAs are present, encoded by slr0943 and sll0018 , respectively [ 24 ]. Class II FBA contributes to approximately 90% of total activity of the reversible alcohol condensation of dihydroxyacetone phosphate (DHAP) and glyceraldehyde 3-phosphate (G3P) [ 38 ]. Therefore, codon-optimized gene sequences of sll0018 and sll1070 , encoding class II FBA and TK, were synthesized, and used for building genetic constructs. An engineered strain HX74 was generated, with additional FBA and TK overexpression, when compared to the control strain HX75 (Fig. 5 B, C). All overexpressed proteins were successfully identified through Western-immunoblot, though the band of FBA protein is barely visible (Fig. 6 A). FBA expression was further verified by increasing the crude protein loading amount from 20 μg to 162 μg (Additional file 1 : Fig. S1). A distinct growth difference between the two strains was observed after day 7, the OD 750 of the control strain HX75 declined faster than strain HX74 (Fig. 6 B). There was no significant difference of IB titer and IB production per cell between strains with or without FBA and TK co-overexpression (Fig. 6 C, D). In contrast, a significant increase of 3M1B titer and 3M1B production per cell of HX74 was observed (Fig. 6 C, D). The obtained improved 3M1B production may result from the critical roles of FBA and TK in ribulose-1,5-bisphosphate (RuBP) regeneration within the CBB cycle. Fig. 6 Positive effects of rewiring central carbon metabolism on photosynthetic isobutanol (IB) and 3-methyl-1-butanol (3M1B) production. A Western-immunoblot analysis of all overexpressed enzymes. L, ladder (in kDa). Ten micrograms, 20 μg, and 20 μg of total soluble proteins were loaded for each strain to detect Strep-tagged, Flag-tagged, and His-tagged proteins, respectively. Protein size: Kivd S286T , 61 kDa; PK, 53 kDa; PCK, 60 kDa; TPI, 27 kDa; AlsS, 62 kDa; FBA, 39 kDa; IlvD, 65 kDa; IlvC, 54 kDa; Slr1192 OP , 36 kDa; TK, 72 kDa. B Growth profile of engineered Synechocystis sp. PCC 6803 strains HX74, HX75 and HX77. C Relative IB and 3M1B titers of the engineered Synechocystis strains HX74, HX75 and HX77 on day 10. D Relative IB and 3M1B production per cell of the engineered Synechocystis strains HX74, HX75 and HX77 on day 10. E Relative IB and 3M1B titers of the engineered Synechocystis strains HX81 and HX87 on day 10. F Relative IB and 3M1B production per cell of the engineered Synechocystis strains HX81and HX87 on day 10. G Growth profile of engineered Synechocystis strains HX79 and HX86. H Relative IB and 3M1B titers of the engineered Synechocystis strains HX79 and HX86 on day 10. I Relative IB and 3M1B production per cell of the engineered Synechocystis strains HX79 and HX86 on day 10. J Growth profile of engineered Synechocystis strains HX81 and HX87. Results are the mean of three biological replicates, each with three technical replicates. Error bars represent standard deviation. Asterisk represents significant difference between different strains (one-way ANOVA, * p < 0.05, ** p < 0.005) Apart from the CBB cycle as the main carbon assimilation machinery in Synechocystis with ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) as the key carbon fixation enzyme, there is a second major carbon-fixing enzyme, named phosphoenolpyruvate carboxylase (PEPc). It was reported that 25% of inorganic carbon assimilation may be through the PEPc catalyzed reaction in Synechocystis under mixotrophic or heterotrophic conditions [ 39 ]. Phosphoenolpyruvate (PEP), one of the precursors for pyruvate synthesis, is converted by PEPc to generate oxaloacetate (OAA), which further feeds into the tricarboxylic acid cycle (TCA cycle) for building block biosynthesis. Complete knock-out of PEPc in Synechocystis is challenging due to its essential role in phototrophic growth in cyanobacteria [ 40 ]. Alternatively, heterologous expression of phosphoenolpyruvate carboxykinase (PCK) from E. coli is one of the approaches to partially eliminate the carbon flow from PEP to TCA cycle and channel more carbon flow towards pyruvate for IB and 3M1B biosynthesis (Fig. 5 A). It was experimentally verified in Synechococcus elongatus PCC 7942 that PCK expression significantly improved the aldehyde production [ 41 ]. Moreover, under phototrophic conditions, the CBB cycle is the dominant pathway for carbon assimilation. G3P, an intermediate of CBB cycle, is a connection node between the CBB cycle and the glycolysis pathway. Within Synechocystis cells, assimilated carbon flows either from the CBB cycle or glycolysis pathway into G3P. Starting from G3P, pyruvate and acetyl-CoA are synthesized and involved in various complex metabolic pathways. It may be interesting to overexpress enzymes connecting two metabolites involved in both the CBB cycle and the glycolysis pathway, to cause perturbation of central carbon metabolism, which may have unexpected effects for pyruvate-derived product biosynthesis. It has been suggested that overexpression of triosephosphate isomerase ( TPI) in E. coli could effectively enhance pyruvate-derived phloroglucinol production [ 42 ] by directing the glycolysis flux into pyruvate formation. Taken together, these two above-mentioned strategies may be promising to further enhance the IB and 3M1B biosynthesis in Synechocystis . To test a combined effect of simultaneous expression of TPI and PCK, originating from E. coli , on IB and 3M1B production in Synechocystis , two engineered strains HX77 and HX87 were generated (Fig. 5 B, C). HX77 was constructed by integrating tpiA and pckA in the neutral site II (NSII) [ 43 ] of strain HX42 [ 15 ], and strain HX87 was constructed by integrating tpiA and pckA in the NSII of strain HX28 [ 15 ]. In both strains, the expression of both genes was driven by the strong synthetic P trc promoter. Protein TPI expression was successfully identified by Western-immunoblot in both strains, whereas PCK expression was only confirmed in strain HX77 (Fig. 6 A). The unsureness of PCK expression in strain HX87 is due to the similar estimated protein size of PCK and Kivd S286T , with less than 1.5 kDa difference. Based on the obtained results, two possible explanations may be made: PCK was successfully expressed, but the detected band overlapped with the band of Kivd S286T ; or PCK was not expressed in the provided cultivation condition and harvesting time point. However, in strain HX87, PCK and TPI were expressed from a single operon (Fig. 5 B) and the second gene in the operon, tpiA , was expressed successfully (Fig. 6 A), suggesting that most probably the first gene in the operon was also successfully expressed. Further attempts were invested to explore an optimized condition for the Western-immunoblot, e.g., amount of crude protein loading, SDS-PAGE running conditions. Unfortunately, it is still challenging to visualize two clearly separated bands (Additional file 1 : Fig. S1). Strains HX75 and HX81 were generated as control strains for strains HX77 and HX87, respectively. After cultivated in short-term screening condition, strain HX77 produced significantly higher 3M1B titer and 3M1B production per cell by 1.3-fold and 1.4-fold, respectively, relative to the control strain HX75 (Fig. 6 C, D). Meanwhile, the produced IB was only slightly improved in strain HX77, 1.1-fold, which is not statistically significant (Fig. 6 C, D). On the other hand, when compared to the control strain HX81, strain HX87 accumulated significantly higher IB and 3M1B titers (Fig. 6 E). Similarly, after normalized to optical density (OD 750 ), IB and 3M1B production per cell of HX87 were both improved by 1.3-fold relative to strain HX81 (Fig. 6 F). In conclusion, positive effects of a co-expression of PCK and TPI on photosynthetic IB and 3M1B production were experimentally verified using two different genetic backgrounds. Metabolic engineering was successfully performed in the carbon fixation pathway as well as the pyruvate-derived 2-keto acid pathway. However, there is still space for further improvement through speeding up carbon flow between carbon fixation and the pyruvate-derived 2-keto acid pathway. In Synechocystis , pyruvate is synthesized from G3P in five enzymatic steps catalyzed by glyceraldehyde 3-phosphate dehydrogenase (Gap1), phosphoglycerate kinase (Pgk), 2,3-bisphosphoglycerate-independent phosphoglycerate mutase (Gpm), enolase (Eno), and pyruvate kinase (PK) (Additional file 1 : Fig. S1). Singly overexpression of PK in Synechococcus resulted in significantly improved isobutyraldehyde production through the 2-keto acid pathway [ 44 ] and singly overexpression of Gpm or Eno also had positive effects on pyruvate-derived isoprene production in Synechocystis [ 45 ]. To further test if overexpression of Gpm, Eno and PK may promote IB and 3M1B production in Synechocystis and if there is any additive effect of overexpression of these enzymes, multiple plasmids were designed and generated (data not shown). However, there was an obstacle preventing further characterization, as it was impossible to acquire positive Synechocystis transformants after several attempts with traditional natural transformation methods. Developing novel genetic engineering tools and having them optimized and ready for generating engineered Synechocystis strains efficiently and precisely are in progress to overcome the encountered challenges. Strain HX86, expressing PK and PCK, was constructed by integrating pyk1 and pckA in the NSII of strain HX15 (Fig. 5 B, C) [ 15 ]. Meanwhile, a control strain, HX79, was constructed by integrating an erythromycin resistance cassette in the NSII of strain HX15 (Fig. 5 B, C). All expressed proteins were successfully identified and confirmed by Western-immunoblot (Fig. 6 A), except for PCK, which was difficult to be separated from the band of Kivd S286T due to similar expected protein size (Additional file 1 : Fig. S1). Strain HX86 grew significantly worse than control strain from three aspects: it had a longer lag phase in the beginning of cultivation; the OD 750 of HX86 declined faster after day 8; and the measured OD 750 was lower than that of control strain throughout the entire cultivation time-period (Fig. 6 G). The observed severe growth retardation may be caused by PCK expression, similar to what was observed in an engineered Synechococcus strain with PCK expressed using a metal-inducible promoter [ 41 ]. Interestingly, the severe growth inhibition phenomenon was not observed for strains HX77 and HX87, both of which had PCK expressed. The cause of the different growth phenotype is currently unknown. Further detailed analysis, e.g., proteomics and metabolomics analysis, are needed to identify and clarify the cause. As shown in Fig. 6 H, strain HX86 achieved an increased 3M1B titer by 1.2-fold, while a comparable IB titer, after expression of PK and PCK. Due to the slower growth rate of HX86 (Fig. 6 G), the IB and 3M1B production per cell was significantly enhanced by 1.2-fold and 1.4-fold, respectively, relative to strain HX79 (Fig. 6 I). Among the engineered strains, the molar ratio of IB and 3M1B observed in strains HX86 and HX77 differed significantly from its corresponding control strain (Additional file 1 : Fig. S2). The observed redistribution of end-products, IB and 3M1B, is consistent with what was claimed previously by Cheah et al. [ 41 ] that heterologous expression of PCK caused a redistribution of aldehyde production (isobutyraldehyde and isovaleraldehyde) in Synechococcus . Interestingly, both strains HX86 and HX77 had PCK additionally expressed, compared to its corresponding control strain. Therefore, PCK expression may directly or indirectly rearrange the metabolic flux of the branched-chain amino acid biosynthesis pathway, and further affect the molar ratio of IB and 3M1B. In conclusion, expression of the five selected target genes of central carbon metabolism showed positive effects on IB and 3M1B production. Co-expressing two of the selected targets successfully enhanced IB or 3M1B titer and production per cell (Fig. 6 C–F, H–I). Not only being specifically valuable for evaluating and improving IB and 3M1B production derived from the 2-keto acid pathway, the identified gene targets may potentially be applied in metabolically engineering Synechocystis to produce various pyruvate-derived compounds."
} | 11,405 |
38896623 | PMC11186498 | pmc | 2,275 | {
"abstract": "Flexible and stretchable electronic devices are subject to failure because of vulnerable circuit interconnections. We develop a low-voltage (1.5 to 4.5 V) and rapid (as low as 5 s) electric welding strategy to integrate both rigid electronic components and soft sensors in flexible circuits under ambient conditions. This is achieved through the design of conductive elastomers composed of borate ester polymers and conductive fillers, which can be self-welded and generate welding effects to various materials including metals, hydrogels, and other conductive elastomers. The welding effect is generated through the electrochemical reaction–triggered exposure of interfacial adhesive promotors or the cleavage/reformation of dynamic bonds. Our strategy can ensure both mechanical compliance and conductivity at the circuit interfaces and easily produce welding strengths in the kilopascal to megapascal range. The as-designed conductive elastomers in combination with the electric welding technique provide a robust platform for constructing standalone flexible and stretchable electronic devices that are detachable and assemblable on demand.",
"introduction": "INTRODUCTION Flexible and stretchable electronics, the cutting edge of wearable ( 1 ), on-skin ( 2 ), robotics ( 3 ), biomedical ( 4 ), and bioelectronic ( 5 ) techniques, have gained persistent developments in both materials and structural layouts ( 6 , 7 ). Nonconductive polymers ( 8 ), polymeric conductors ( 9 ) and semiconductors ( 10 ), liquid metals ( 11 ), gel ( 12 ) and elastomer ( 2 ) electrolytes, and conductive elastomers (CEs) ( 13 ) that are soft, deformable, and shape-adaptable, have been widely explored for the substrates, circuits, sensors, and electronic components of flexible and stretchable electronic devices. Alternatively, structural layouts including serpentine ( 14 ), mesh ( 15 ), microcrack ( 16 ), and longitudinal wave ( 17 ) provide possibilities for achieving device flexibility and stretchability from rigid metals and silicon-based electronic components. However, flexible and stretchable electronic devices lack facile, adaptable, and reliable interconnection techniques and have long been plagued by vulnerable circuit interfaces ( 18 , 19 ). Key issues lie in achieving stable interconnections between different electronic components [including soft sensors ( 20 – 22 ), deformable electronic components ( 23 – 25 ), commercially available rigid electronic components ( 26 – 28 ), etc.] using flexible and stretchable circuits. Among numerous conductive materials, CEs are preferred candidates for constructing flexible and stretchable circuits ( 29 – 31 ). There are intrinsic CEs typically exampled by stretchable conjugated polymers ( 32 , 33 ) and exogenous CEs adopting fillers ( 34 – 37 ) as the conductive components and polymers as the elastic substrates. Recent advances in materials science and manufacturing techniques have led to an explosion of CEs and continuous updates of their mechanical and conductive performances. However, a standalone flexible and stretchable electronic device not only requires deformable circuits but also consists of some indispensable electronic components such as signal collectors and processors, circuit controllers, signal transmitters, and energy suppliers, which can be soft and rigid and have access ports with different material compositions. It remains challenging to connect CEs with different electronic components, while simultaneously ensuring interfacial adhesion strength, mechanical compliance, and conductivity at the circuit interfaces. Metallic solder bonding techniques such as tin welding ( 38 ) and laser welding ( 39 ) have been directly applied in the circuit interconnection of CE-enabled electronic devices. However, the high temperature generated during welding can easily damage the flexible substrates and circuits ( 40 ). Conductive past strategies involve the combination of polymer adhesives and conductive metal bumps ( 41 ), utilization of anisotropic conductive films ( 42 ), and using light/heat-curable silver adhesives ( 43 ). These methods usually suffer from the problems of low interfacial adhesion strength, mechanical mismatch, or complicated postprocessing. Self-healable materials ( 44 , 45 ), liquid metals ( 11 , 46 – 48 ), and interlocking microbridges ( 49 ) facilitate the self-connection of CE circuits, while they show limited adhesiveness toward independently fabricated electronic components. There are also attempts of using nonconductive polymer adhesives to enhance the interfacial adhesion between electronic components and CE circuits ( 50 , 51 ). An ingenious example is the design of interpenetrated conductive nanostructures in a self-adhesive thermoplastic elastomer, which provides a plug-and-play assembly strategy to form flexible and stretchable devices ( 51 ). However, this strategy requires not only a premodification of the substrate surfaces with the polymer adhesive but also a careful control over the thickness of the adhesive layer to balance the interfacial adhesion and conductivity. The future practical use and the commercialization of flexible and stretchable electronic devices are determined by the development of facile and reliable circuit interconnection techniques.",
"discussion": "DISCUSSION Metal electric welding ( 60 ) and plastic fusion ( 61 ) are essentially a physical process of melting. Our concept of electric welding is a chemical process comprising both electrochemical and dynamic bond reactions. In particular, the reversible reaction of dynamic bonds is programmed by the electrochemical reaction. This method can weld CEs to different conductive materials and electronic components with both rigid and soft mechanical characteristics, thus providing an all-in-one solution for the challenging issues in flexible and stretchable electronics including interface adhesion, mechanical matching, and interface bonding stability. Compared with the previously reported interface connection methods, our electric welding technique succeeds in both material adaptability and operation simplicity (table S2). This electric welding technique may also find potential applications in the fields of bioelectronics ( 8 ), energy storage ( 62 ), and robotics ( 20 , 63 ), because the welding procedure is simple, rapid, and universal. Moreover, the concept of using electrochemical strategies to trigger or control the reversible reactions in the interior of materials is implemented for achieving new functions."
} | 1,628 |
29575478 | null | s2 | 2,276 | {
"abstract": "A major goal of nanotechnology and bioengineering is to build artificial nanomachines capable of generating specific membrane curvatures on demand. Inspired by natural membrane-deforming proteins, we designed DNA-origami curls that polymerize into nanosprings and show their efficacy in vesicle deformation. DNA-coated membrane tubules emerge from spherical vesicles when DNA-origami polymerization or high membrane-surface coverage occurs. Unlike many previous methods, the DNA self-assembly-mediated membrane tubulation eliminates the need for detergents or top-down manipulation. The DNA-origami design and deformation conditions have substantial influence on the tubulation efficiency and tube morphology, underscoring the intricate interplay between lipid bilayers and vesicle-deforming DNA structures."
} | 201 |
24921050 | null | s2 | 2,277 | {
"abstract": "The preparation, characterization, and use of a UV responsive non-woven nanofiber polymeric mesh is reported that transitions from being hydrophobic to hydrophilic. Three distinct wetting profiles are observed during the wetting process. 3D hydrophilic cavities were created within the hydrophobic bulk material by using a photo mask to control the geometry and UV exposure time to control the depth of the region."
} | 103 |
35975089 | PMC9347211 | pmc | 2,278 | {
"abstract": "Although a wide range of self-healing materials have been reported by researchers, it is still a challenge to endow exceptional mechanical properties and shape memory characteristics simultaneously in a single material. Inspired by the structure of natural silk, herein, we have adopted a simple synthetic method to prepare a kind of elastomer (HM-PUs) with stiff, healable and shape memory capabilities assisted by multiple hydrogen bonds. The self-healing elastomer exhibits a maximum tensile strength of 39 MPa, toughness of 111.65 MJ m −3 and self-healing efficiency of 96%. Moreover, the recuperative efficiency of shape memory could reach 100%. The fundamental study of HM-PUs will facilitate the development of flexible electronics and medical materials.",
"conclusion": "Conclusions In summary, inspired by silkworm fibril, we have developed one kind of elastomer with a number of excellent properties such as high transmittance (99%), high tensile strength at break (39 MPa), splendid toughness (111.65 MJ m −3 ), high self-healing efficiency, notch insensitive character, and shape memory property. The great mechanical properties are originated from the hydrogen bond arrays, and the network that soft segments contribute to the high elasticity. Due to the special structure comprising several dynamic multiple hydrogen bonds, it endows this elastomer with sustainability and durability. The elastomer we designed by a simple synthetic method is a kind of material with self-healing and shape memory properties, not only paving the way for designing and manufacturing the materials with extraordinary mechanical properties, but also for application in the fields of flexible electronics, medical apparatus, and shape memory robots.",
"introduction": "Introduction Synthetic polymers are a kind of material with entropic elasticity, which have been widely applied in the fields of rubber products, flexible electronics, aerospace and encapsulants. Moreover, various hydrogels constructed by polymer chains have been used in medical research. 1–8 Nevertheless, polymers will presumably be damaged by external force in practical application; therefore, it is important to reinforce the robustness and retrievability of traditional polymers. Up to now, two self-healing mechanisms have been proposed in healable materials, namely extrinsic and intrinsic self-healing materials. 9–17 Extrinsic self-healing materials have been limited in application on account of their limited self-healing times, and the intrinsic self-healing strategy has accordingly occupied the leading role in self-healing materials. Combining tough properties and high self-healing efficiency is a paradox. Therefore, healable materials with high self-healing efficiency and excellent mechanical properties have intrigued scientists. Unfortunately, self-healing materials based on weak dynamic bonds tend to lack good mechanical properties. To address this conundrum, scientists have devoted a good deal of attention to it. 2-Ureido-4[1 H ]-pyrimidinone (UPy) is a kind of quadruple hydrogen bond structure, and it is widely utilized in self-assembled polymers and molecular structure design. On account of the superiorities of the UPy structure, scientists have introduced UPy and its consequent benefits in the molecular structures of healable materials. For instance, Bao et al. reported a kind of supramolecular polymer material with good mechanical properties and highly stretchable performance by condensation polymerization assisted by UPy; with the aid of quadruple hydrogen bonds, the maximum stress could reach 3.74 MPa and the material exhibited high self-healing efficiency (∼88%). 18 In Lee's work, UPy groups were used to functionalize polyurethane, and an elastomer with high tensile strength (9.44 MPa) and elongation (2340%) was prepared; furthermore, the maximum self-healing efficiency showed desirable performance (∼70%, in organic solvent). 19 In both the cases above, the mechanical properties of these healable polymers functionalized by UPy groups could not meet the ideal level, and the solubility of UPy is too low in polar organic solvents to use in organic syntheses. 20,21 Hence, two problems urgently need to be addressed: (1) which kind of hydrogen bond structure should be adopted to replace the UPy with quadruple hydrogen bond structures; (2) how to balance the excellent mechanical properties and high self-healing efficiency ( Scheme 1 ). Scheme 1 Schematic diagram of the silkworm fibril. In nature, biomaterials exhibit extraordinary characteristics that man-made materials cannot compare with. 22–25 Silkworm silk is a kind of natural long fiber made from its secretion, which possesses excellent characters such as lightweight, good flexibility, high tenacity, admirable biodegradability, and biocompatibility, has been widely used in these areas of fabrics, composite materials, and biosensors. 26–31 Most importantly, the super robustness of silkworm silk originates from the hydrogen bond arrays in β-sheet nanocrystals. Hydrogen bond arrays play two roles in silk fibril: (1) assemble nanocrystals to strength the fibril; and (2) dissipate energy effectively under external force via hydrogen bond breakage. 32–34 Inspired by the structure of the silkworm fibril, multiple hydrogen bonds were introduced in polyurethane by condensation polymerization. Since the diffusion of polymer chains plays a paramount role in the self-healing process, we proposed a particular molecular design that embedded asymmetric alicyclic structure in polyurethane networks as it would affect the self-healing efficiency in cracks. 35 Also, the high toughness and shape memory property of the kind of elastomer is a result of the density and configuration of the hydrogen bond arrays, and these properties are tuned by macromolecule design. Herein, HM-PUs were synthesized from 4,4′-methylenebis(cyclohexyl isocyanate) (HMDI) as the hard segment, poly(tetramethylene ether)glycol (PTMEG) as the soft segment, and isophorone diamine (IPDA) as the chain extender and the self-healing elastomer, which not only exhibits remarkable tensile strength of 39 MPa but has a high speed of shape memory. Furthermore, this kind of elastomer even outperforms some commercial elastomers, 36–38 and the fracture energy of the elastomer is higher than some alloys. 42 As a kind of dynamic bond, the force-induced hydrogen bonds will dissipate energy to improve the toughness of the polymer networks. Under strain, ordered hydrogen bonds formed between polymer chains lead to the appearance of a metastable crystal, and accordingly, the toughness increases. More importantly, HM-PU with high transparency (∼100%), high self-healing efficiency (>90%), and high tensile strength (39 MPa) was developed; also, this material possesses remarkable shape memory characteristic (recuperative efficiency: ∼100%).",
"discussion": "Results and discussion Molecular design and general characterization We intended to synthesize a kind of molecular structure that resembled the structure of silkworm fibril as the structure of multiple hydrogen bond was a superb option to assemble linear macromolecule, and the hydrogen bonds between the carbonyl groups and hydrogen atoms can constitute the polymer matrix network with the help of hard segments as the node and soft segments as the crosslinker. Of course, this special structure we designed can adequately dissipate the energy under external force, avoiding the rupture in materials. Herein, a two-step polycondensation reaction ( Scheme 2 ) was adopted to prepare a kind of healable elastomer that constituted the crosslinked matrix with hydrogen bonds. Also, HMDI as the hard segment and IPDA as the extender were chosen as the nodes in the matrix, and PTMEG as the soft segment is the crosslinker that joins these nodes. The asymmetric alicyclic structure of IPDA will weaken the forces of hydrogen bonds between carbonyl groups, resulting in high diffusion speed of the molecular chains and fast self-healing rate. Scheme 2 Synthetic route of HM-PUs containing multiple hydrogen bonds. In the synthetic reaction, three kinds of HM-PUs with different molar ratios of hard and soft segments were prepared, showing various properties, and the information of HM-PUs is shown in Table S1. † The GPC results show that M n and M w are capable of coming up to a high order of magnitude (>10 4 ), and the small PDI shows that the molecular weights of HM-PUs are evenly distributed. As the hydrogen bond structure domains are adequately small, the visible light transmittance of HM-PUs was up to 99% in the wavelength range from 300 cm −1 to 800 cm −1 . 39 In addition, the chemical structure of HM-PU2 was analyzed by the 1 H NMR spectrum, as shown in Fig. 1c . In the FTIR spectra of HM-PUs, the –NCO peaks completely disappeared, indicating that the reaction was finished, and the –NCO groups had transformed to C \n \n\n<svg xmlns=\"http://www.w3.org/2000/svg\" version=\"1.0\" width=\"13.200000pt\" height=\"16.000000pt\" viewBox=\"0 0 13.200000 16.000000\" preserveAspectRatio=\"xMidYMid meet\"><metadata>\nCreated by potrace 1.16, written by Peter Selinger 2001-2019\n</metadata><g transform=\"translate(1.000000,15.000000) scale(0.017500,-0.017500)\" fill=\"currentColor\" stroke=\"none\"><path d=\"M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z\"/></g></svg>\n\n O groups. It is worth noting that the characteristic peaks of C O in urethane at 1718 cm −1 and C O in urea in the wavenumber range from 1628 cm −1 to 1658 cm −1 varied with the molar ratio of hard and soft segments. The thermal properties of HM-PUs were investigated by DSC analysis ( Fig. 1d ); the thermograms show the variational tendency of HM-PUs due to the change in the intermolecular forces between hard and soft segments, and the glass transition temperature ( T g ) of HM-PU1 reaches up to −56.4 °C than any others, indicating that hydrogen bonds and hard segments will significantly affect the diffusion of polymer chains. Also, we obtained the results of decomposition temperature to be approximately 273 °C from the TGA curves in Fig. S1. † In addition, the FTIR spectra of HM-PUs in the range from 1600 cm −1 to 1750 cm −1 are assigned to C O groups, and the absorptive region of C O is deconvoluted into four subpeaks; these subpeaks can be assigned to C O in urea and urethane groups (Table S2 † ). 40,41 Then, the content of four kinds of C O groups are listed in Fig. 2d such as free urethane amide, urethane amide (C O), and urea amide (C O). We can come to a conclusion from the bar diagram that the C O content of HM-PU3 (19%) in the free urethane amide region is much higher than the contents of HM-PU1 (7%) and HM-PU2 (8%), and the ordered hydrogen bonds of HM-PU1 built by the C O groups in urea amide reaches the maximum value (19.5%), indicating that the intermolecular force is stronger than that of the others (HM-PU2 and HM-PU3). As these analyses mentioned above, the results of DSC and FTIR indicate that free urethane amide will accelerate the diffused rate of polymer chains, and the metastable crystal in HM-PU is formed by ordered hydrogen bonds between the urea amide groups and hydrogen atoms. Fig. 1 General characterization of HM-PUs. (a) UV-vis spectra of HM-PUs in the visible light wavelength range. (b) FTIR spectra of HM-PUs in the wavenumber range from 400 cm −1 to 4000 cm −1 . (c) 1 H NMR spectrum of HM-PU2. (d) DSC curves of HM-PUs in the temperature range of −85 °C to 85 °C. Fig. 2 FTIR spectra of HM-PUs in the C O stretching region. Self-healing and mechanical properties of HM-PUs The polymer network of HM-PU can self-heal after rupturing due to the interesting process of breakage and rebuilding of hydrogen bonds, as illustrated in Fig. 3a . Subsequently, a bloom-shaped film was cut into two pieces, followed by placing the two sections in contact in 50 °C after 5 h, and the pieces completely connected under stretching. The three samples such as HM-PU1, HM-PU2, and HM-PU3 were put into different temperature atmospheres (100 °C, 70 °C, and 50 °C), respectively, for 30 min. Thus, the optical microscopy images show that HM-PUs have a good self-healing property. Fig. 3 Molecular network of HM-PU and optical photos. (a) Chemical structure and self-healing mechanism of HM-PU. (b) The self-healing process of the bloom-shaped film. (c) Optical microscopy photos of the scratch films showing a good appearance during the self-healing process (HM-PU1-100 °C-30 min, HM-PU2-70 °C-30 min, HM-PU3-50 °C-30 min). One attractive phenomenon we can see is that the mechanical properties are variational by adjusting the molar ratios of soft and hard segments of HM-PUs ( Fig. 4 and Table S3 † ). In Fig. 4b , HM-PU1 shows an extreme tensile strength of 39.7 MPa, and the tensile strength, toughness, and Young's modulus obviously decreased promptly with decreasing content of IPDA as the extender. The following reasons account for this: (1) As is usual, the hydrogen bonds between urea groups are stronger than the hydrogen bonds between urethane groups, and the content of the segments of IPDA in the polymer network of HM-PU1 are more than that of the two other HM-Pus; (2) The more the number of soft segments networks, the faster the diffusion speed of the polymer chains. It is interesting to note that the HM-PUs films exhibit enormous crack tolerance due to the extraordinary energy dissipation capacity. In Fig. 4a , the HM-PU1 film with a notch of 1 mm was stretched up to 400% (approximately 4 times) of its original length, and the crack was extended as a smooth notch, indicating that the elastomer with cracks can still work at high elongation. The fracture energies of HM-PUs were calculated to be in the range from 45 kJ m −2 to 60.34 kJ m −2 in the ESI (Fig. S4 † ). In addition, the fracture energy is higher than the fracture energies of most elastomers and some alloys. 42 In order to further evaluate the self-healing efficiency of HM-PUs, the specimens were tested by tensile tests of HM-PUs after different times at a specific temperature ( Fig. 4c–e ). The self-healing efficiency is defined as the ratio of the stress or elongation to that of the virgin samples, and the self-healing efficiency is characterized by the two formulas as follows. 1 2 η σ is the self-healing efficiency calculated by the stress, σ healed is the tensile strength of the healed samples, σ virgin is the tensile strength of the virgin samples, η l is the self-healing efficiency calculated by the elongation, l healed is the elongation of the healed samples, and l virgin is the elongation of the virgin samples. Fig. 4 The mechanical properties of HM-PUs. (a) A notched HM-PU1 specimen at a strain of 400%. (b) Typical stress–strain curves of the virgin and notched samples. Typical stress–strain curves of virgin and self-healed samples. (c) HM-PU3 healed at 50 °C. (d) HM-PU2 healed at 70 °C. (e) HM-PU1 healed at 100 °C. The self-healing efficiency depicted in Fig. 4 and S5 † shows its variation tendency in terms of stress and elongation as the variable factors of temperature and healing time, respectively. As we can see, the self-healing efficiency of HM-PUs improves with time or temperature. Also, HM-PU3 could heal in a moderate atmosphere (50 °C) for a short time, while the corresponding self-healing efficiency of the self-healed samples of HM-PU1 and HM-PU2 at 50 °C is so weak that the two samples fractured rapidly; hence, the results will not be shown in this article. When the healing temperature is set as 70 °C and 100 °C, respectively, the self-healing efficiency of HM-PU1 and HM-PU2 increases stepwise. After 10 h at 50 °C, the healing efficiency of HM-PU3 exhibits appreciable value (90% in terms of elongation, 80% in terms of stress). Compared with HM-PU1 and HM-PU2, the self-healing performance is not enough, showing a macroscopic decline. Certainly, the self-healing mechanism could be interpreted by the FTIR analysis in the section of general characterization. It should be thought to be due to the dynamic consecutive exchange and recombination of enhanced hydrogen bonds as the molecular chains diffuse to the fracture surface area, and the mechanical properties were recovered; most importantly, the key factor of diffusion of molecular chains is the number of soft segments. Thus, the healing efficiency of HM-PU3 is the highest than the two others. Also, benefiting from the dynamic property of hydrogen bonds, HM-PUs have a good recyclable character in the field of environmental protection (Fig. S6 † ). In summary, on account of the significant effects of hydrogen bond in the elastomer, in consequence, we can get one kind of specific material with ideal properties by adjusting the balance between the self-healing property and the mechanical properties. 43,44 The diffused properties of polymer chains In order to study the self-healing mechanism, UV-vis spectra and molecular dynamics simulation (MD) were utilized to study the properties of polymer chains. The pieces (0.4 g) of HM-PU1, HM-PU2, and HM-PU3 were added into three bottles and immersed in THF solvent (17 mL) for 8 h. The mutative process of HM-PUs fragments in THF had been recorded by a camera by taking photos at different times. In the initial state, the pieces in the bottles can be clearly observed, whereas these pieces disappeared after 8 h, exhibiting transparent liquids, indicating that a good deal of pieces had been dissolved. Subsequently, the solution in the bottles was removed, and some residues in the bottom of the bottles display three kinds of forms as the colloidal gel, colloidal gel/viscous liquid, and viscous liquid for HM-PU1, HM-PU2, and HM-PU3, respectively. Therefore, HM-PU3 is easier to dissolve in organic solvents, revealing that the diffusion coefficient of its polymer chains is higher than others on account of more soft segments and fewer hydrogen bonds. In order to represent the quantitative level of polymer chains in the procedure of diffusion, UV-vis measurements were employed. As shown in Fig. 5 , obvious characteristic peaks belonging to HM-PU1, HM-PU2, and HM-PU3 at about 241 nm, 240 nm, and 238 nm, respectively, are due to carbonyl in urea and urethane groups, wherein the types of electronic transition are n–π* and n–σ. After 2 h, the absorption peak of HM-PU1 is as high as 0.59, whereas the peak of HM-PU3 as low as 0.2. The reasons we can conclude is as follows: (1) the number of C O groups in polymer chains dissolved in THF reached the highest value compared to others; (2) The hyperchromic effect of –NHR groups in urea is stronger than that of the –OCR groups in urethane. Moreover, the characteristic peak of C O shifts to the region of long wavelength due to the hyperchromic effect of –NHR groups, meaning redshift. The most important conclusion we can obtain from Fig. 5 is that the absorption peaks of HM-PU3 grow more quickly than that of HM-PU1 and HM-PU2 with time passing by, revealing the phenomenon that the polymer chains of HM-PU3 can diffuse more quickly, indicating the HM-PU3 elastomer can heal in a shorter time. Fig. 5 (a) The mutative process of HM-PUs immersed in THF. UV-vis spectra of the solution of HM-PUs in THF after certain times. (b) The UV-vis spectra of HM-PUs after 2 h. (c) The UV-vis spectra of HM-PU1 in the time range of 2 h to 8 h. (d) The UV-vis spectra of HM-PU2 in the time range of 2 h to 8 h. (e) The UV-vis spectra of HM-PU3 in the time range of 2 h to 8 h. In order to describe the diffused behavior of the molecular chains accurately, molecular dynamics simulation (MD) was used to simulate their dynamic behaviors. In these cases, all MD simulations for the two cell systems were carried out with the commercial software Materials Studio (MS) (Accelrys Inc., San Diego, Version 2019) with COMPASS force field. The amorphous cells were built by the amorphous cell module containing ten molecular chains, which include the single molecular chain with one PTMEG segment, one IPDA extender, and that with two PTMEG segments, one IPDA extender for PU 1:1 and PU 2:1 , respectively ( Fig. 6a and b ). The density of all the cells is 0.5 g cm −3 . Then, the two cells were constructed utilizing COMPASS force field to assign the charges, followed by minimizing the structures via Forcite module until convergence. They were carried out for MD using the NPT ensemble for 10 000 ps with 1 fs time step, and one frame was recorded every 10 steps; eventually, full trajectories were saved. The MSD can be calculated using the following formula 45 3 where s ( i ) ( t ) is the position vector of i atom at the t moment, s ( i ) (0) is the position of i atom at the initial moment, and N is the total number of molecular chains. Fig. 6 Snapshots of the MD simulation of the cell structures of HM-PU with ten representative polymer chains. (a) One PTMEG segment and one IPDA extender. (b) Two PTMEG segments and one IPDA extender. (c) The energies of simulated cell systems. (d) The results of the mean-squared displacements (MSD) of HM-PU. Whereafter, the two structures were annealed via one annealing cycle of a linear heating and cooling process from 300 K to 600 K and from 600 K to 300 K, respectively. The simulation was implemented under the NPT ensemble for 25 000 steps and the time step was 1 fs. Subsequently, the cohesive energy per chain, defined as the average energy of every molecular chain diffusing from the condensed state to an infinite distance from one another, was calculated. In the amorphous cells we constructed above, the cohesive energy per molecular chain can be calculated by the following formula 46 4 where E s is the average potential energy of the single polymer chain, E t is the average potential energy of the amorphous cell containing ten polymer chains, and E ce is the cohesive energy per polymer chain. Also, the potential energies were calculated via the energy task of Forcite module (all systems were in the equilibrium state). To sum up, the MSD and cohesive energy were calculated using MD simulation; the cohesive energy of PU 1:1 (73.53 kcal mol −1 ) is much higher than that of PU 2:1 (67.97 kcal mol −1 ), indicating that PU 2:1 per chain needs more energy to be at an infinite distance, and the tendencies of PU can also be observed in the MSD curves. It can be said that the UV-vis spectra and MD simulation were used to verify the inner mechanism of the healing process, revealing a phenomenon that the more the number of hard segment polymer chains, the lower the diffused rate. The studies of residual strain and energy dissipation of HM-PUs In order to investigate the energy dissipations and residual strains of HM-PUs, they were further subjected to the mechanical tests of cyclic loading/unloading with various maximum strains; the cyclic speed was 100 mm min −1 ( Fig. 7 ), and all the samples could recover quickly. In Fig. 7b , the residual strain and energy dissipation increase with the maximum due to a great deal of hydrogen bonds, and some hydrogen bonds could fracture as the strain increases. 47 To explore the elastic properties of HM-PU, the residual strain ( ε rs ) was executed to be normalized by the maximum strain ( ε ms ), and the computational formula is ε rs / ε ms . The data of normalized strain ( ε rs / ε ms ) can be calculated from Fig. 7a ; the ε rs / ε ms of HM-PU1 increased with the maximum rapidly. Also, the slope for ε rs / ε ms of HM-PU2 and HM-PU3 is almost zero, indicating that their rupture degree and microphase separation is not enhanced at high elongation; nevertheless, HM-PU1 exhibits the reverse phenomenon. In addition, the variational tendency of energy dissipation is analogous to that of residual strain ( Fig. 7d and S2 † ). In summary, the polymer chains with more hard segments will dissipate more energy in the mechanical deformation process. Fig. 7 (a) Typical cyclic curves of HM-PUs. (b) Residual strain of HM-PUs at different maximum strains. (c) Normalized residual strain obtained from (a). (d) Energy dissipations of HM-PUs at different maximum strains. The shape memory properties of HM-PUs On account of dense multiple hydrogen bonds in polymer chains of HM-PU, they could reversibly alternate between one state of the transformed shape and another state of the original untransformed shape through the stimulus of heat. 48 In Fig. 8 , we designed three models such as “OK”-shaped film, spiral film, and box-shaped film. All the films were placed under 70 °C to deform, followed by placing them at room temperature, and the shapes were fixed; subsequently, these distorted films recovered to their initial shapes under the stimulus of 70 °C temperature. For the purpose of quantifying the rate of the shape memory, the film of HM-PU1 was deformed at 70 °C, whereafter it was fixed in room temperature atmosphere, followed by heating it at 70 °C once again, and the process of changing the angle was monitored via a camera. The angle of this film recovered from 0° to 180° in 25 s, and the interval times between the two angles we recorded to show a linear increase, indicating that the efficiency of shape recovery decreases with time. Fig. 8 The shape memory behavior of HM-PU1. (a) Reversible shape transition of different shapes (“OK”, “spiral”, and “box”) stimulated at 70 °C. (b) The changing process of the angle. Furthermore, the HM-PU1 film stretched to 300% was employed in the process of shape fixing and recovery, and underwent six shape memory cycles to test the durability of shape memory ( Fig. 9 ). As shown in Fig. 9c , the recovery property of HM-PU1 is excellent; even after undergoing six cycles, the wastage of shape memory performance is negligible. But the fixed shapes of HM-PU2 and HM-PU3 cannot be maintained in the elongation of 300% and rebound to a value lower than 300% because the energies of hydrogen bonds in the stretched polymer chains cannot completely counteract the energy of resilience of the stretched polymer network. In order to reveal the mechanism of the shape memory behavior, the XRD was used to test HM-PU1 at different elongations. In the XRD curves ( Fig. 9f ), the characteristic peaks changed with elongation, and some weak peaks appear at 40°; these peaks disappeared until the film was heated to the initial elongation, revealing that the microstructure of HM-PU1 undergoes a conversion from the uncrystallized form to the crystalline form and as well from the crystalline form to the uncrystallized form in the shape memory process (Fig. S7 † ). 49 Based on the mechanism, four strings of HM-PU1 were used to suture two pieces of papers, followed by heating it by a heat gun, and the gap between the papers was disappeared quickly; thus, it can be seen that this kind of material can be applied in the medical field such as in sutures. Fig. 9 The shape memory properties of HM-PUs. (a) Photos of HM-PU1 in the original state, the state stretched to 300% and the recovered state at 70 °C. (b) The two pieces of papers were sutured by four bands of HM-PU1 films. The HM-PU films were stretched to 300% and fixed at room temperature, then recovered to their original state for six cycles at 70 °C. (c) HM-PU1. (d) HM-PU2. (e) HM-PU3. (f) XRD of HM-PU1 in the original state and the states at 50%, 100%, 200% and 300%."
} | 6,853 |
35412890 | PMC9169844 | pmc | 2,279 | {
"abstract": "Significance Inoculation of cereals with diazotrophic (N 2 -fixing) bacteria offers a sustainable alternative to the application of nitrogen fertilizers in agriculture. While natural diazotrophs have evolved multilayered regulatory mechanisms that couple N 2 fixation with assimilation of the product NH 3 and prevent release to plants, genetic modifications can permit excess production and excretion of NH 3 . However, a lack of stringent host-specificity for root colonization by the bacteria would allow growth promotion of target and nontarget plants species alike. Here, we exploit synthetic transkingdom signaling to establish plant host-specific control of the N 2 -fixation catalyst nitrogenase in Azorhizobium caulinodans occupying barley roots. This work demonstrates how partner-specific interactions can be established to avoid potential growth promotion of nontarget plants.",
"discussion": "Discussion In this work, we have stably engineered a homozygous rhizopine biosynthesis pathway into barley and concurrently improved rhizopine signaling circuitry in Ac to establish plant host-dependent control of nifA - rpoN cassettes driving expression and activity of nitrogenase. Despite the 10 3 -fold improved sensitivity of in vitro rhizopine perception, our confocal microscopy and flow-cytometry experiments indicated that only a small fraction of the Ac Cherry (pSIR02) population occupying the RhiP barley roots responded to rhizopine signaling by expressing P mocB::GFP ( Fig. 2 ). Likewise, in situ nitrogenase activity by Ac Δ nifA (pSIN02) colonizing RhiP barley was suboptimal compared to wild-type Ac colonizing wild-type or RhiP barley ( Fig. 4 E ). It seems unlikely that plasmid loss was responsible for these results given that all the Ac Cherry (pSIR02) colonies recovered from RhiP barley tested positive for kanamycin resistance conveyed by pSIR02. Silencing may have impeded efficient in situ rhizopine signaling and nitrogenase activity, although 58 to 72% of colonies recovered from T2 RhiP barley roots retained the capacity for SI induction of P mocB::GFP on agar plates ( SI Appendix , Fig. S5 ). Finally, we found that Ac Δ nifA (pSIN02) are mildly defective in their ability to colonize barley roots, which may reduce nitrogenase activity per plant compared to the wild-type system. Our results overall suggested that increased rhizopine production by RhiP plants translated to stronger individual and population level SI perception and more effective nitrogenase expression and activity by bacteria ( Figs. 2 and 4 ). Thus, further optimization of the rhizopine sensor and tuning of rhizopine biosynthesis gene expression may be necessary to generate a synthetic interaction that is better suited for practical use. It is crucial to note that excessive overexpression of SI biosynthesis genes would likely cause defects in plant metabolism that might influence biomass accumulation and increase the potential for enrichment of specialized bacteria capable of catabolizing the signal as a source of carbon or N in the field ( 28 – 30 ), as was previously reported for transgenic opine-producing plants ( 20 – 23 ). Thus, retaining SI biosynthesis at the minimum level required for a functional bacterial response will be crucial to optimize this circuitry. Building on rhizopine control of N 2 fixation demonstrated here, our future aim is to engineer a more complete “synthetic symbiosis” where bacteria are also prompted to release fixed N to the host plant. This will require rewiring of nitrogen metabolism as Ac does not naturally release large quantities of fixed N into the growth media ( 39 ) but can be engineered to do so by deletion of the P II genes glnB and glnK ( 40 – 42 ) or potentially by other strategies ( 18 , 43 – 45 ). In addition, rhizopine signaling could be deployed by bacteria to initiate response signals to plants, establish biocontainment of bacteria in the root system, and develop relay signaling circuitry that will allow genetically incompatible bacteria to indirectly respond to rhizopine ( 9 ). Ultimately, we intend to utilize rhizopine signaling circuitry for control of bacterial traits within engineered barley nodules or “pseudo-nodules” where more fine-tuned signal and nutrient exchange would presumably exist ( 11 , 12 , 46 ). The biotechnological achievements made in this study represent a significant milestone toward the development of efficient partner-specific synthetic symbioses between cereals and bacteria."
} | 1,130 |
39723137 | PMC11668738 | pmc | 2,280 | {
"abstract": "Arbuscular mycorrhiza (AM) represents a symbiotic mutualistic association between most land plants and Glomeromycota fungi. AM fungi develops specialized intraradical and highly branched structures, called arbuscules, where bidirectional exchange of nutrients between plant and fungi partners occurs, improving plant growth and fitness. Transcriptional reprogramming and hormonal regulation are necessary for the formation of the arbuscules. SlDLK2 , a member of the third clade from the DWARF14 family of α , β -hydrolases closely related to the strigolactone receptor D14, is a negative regulator of arbuscule branching in tomato, but the underlying mechanisms are unknown. We explored the possible role of SlDLK2 on the regulation of hormonal balance. RNA-seq analysis was performed on roots from composite tomato plants overexpressing SlDLK2 and in control plants transformed with the empty vector. Analysis of transcriptomic data predicted that significantly repressed genes were enriched for genes related to hormone biosynthesis pathways, with a special relevance of carotenoid/apocarotenoid biosynthesis genes. Stable transgenic SlDLK2 overexpressing (OE) tomato lines were obtained, and hormone contents were analyzed in their roots and leaves. Interesting significant hormonal changes were found in roots of SlDLK2 OE lines with respect to the control lines, with a strong decrease on jasmonic acid and ABA. In addition, SlDLK2 OE roots showed a slight reduction in auxin contents and in one of the major strigolactones in tomato, solanacol. Overall, our results suggest that the negative regulation of AM symbiosis by SlDLK2 is associated with the repression of genes involved in the biosynthesis of AM-promoting hormones.",
"introduction": "1 Introduction Arbuscular mycorrhiza (AM) represents a symbiotic mutualistic association between most land plants and fungi from the Glomeromycota . The interaction benefits plant and fungi with the exchange of nutrients between the two partners. Plants in association with AM fungi improve their growth and fitness, and AM fungi receive plant carbohydrates and lipids essentials for their development ( Shi et al., 2023 ). Functional AM development requires fundamental reprogramming of root cells, to allow the formation of symbiotic structures required for nutrient exchange. Several stages in the establishment of AM have been identified, including the exchange of diffusible signals for mutual recognition, induction of AM-related genes in the host for cellular rearrangement that allows accommodation of the AM fungus, and creation of the arbuscule, the symbiotic structure which provides an appropriate interface for the exchange of nutrients ( Choi et al., 2018 ). The interaction is highly regulated by both partners, namely, plant and AM fungi, at the cellular, molecular, and genetic levels. Host plant cells regulate the development and functioning of the mutualistic association by a complex transcriptional reprogramming that includes, among others, hormone-related genes (mainly strigolactones and gibberellins), common symbiotic signaling pathway (CSSP) genes, transcription factors, and genes for transport, metabolism, and cellular processes required for functional AM symbiosis ( Ho-Plágaro and García-Garrido, 2022 ). Particularly important are the extensive transcriptional changes that are induced during arbuscular formation, and a precise spatiotemporal regulation of gene expression is essential for proper arbuscule development. Therefore, the identification of the mechanisms mediating these gene expression changes is crucial to understand how arbuscule formation and function are regulated ( Pimprikar and Gutjahr, 2018 ). Several studies have highlighted the potential role of apocarotenoids and related compounds in regulating the arbuscular mycorrhizal symbiosis cycle. In earlier research, Ho-Plágaro et al. (2021) recently identified a tomato gene encoding an apocarotenoid-like receptor protein, DLK2 , which plays a regulatory function in arbuscule formation. DLK2 proteins form a third clade within the DWARF14 family of α , β -hydrolases, closely related to the strigolactone receptor D14. The expression of the DLK2 gene has consistently been used as a marker for strigolactone (SL) and karrikin (KAR) signaling ( Waters et al., 2012 ; Sun et al., 2016 ). SLs, which are plant hormones derived from carotenoids, were initially identified as soil compounds that trigger the germination of the parasitic plant Striga lutea ( Cook et al., 1966 ). They were later discovered to play a crucial role in facilitating the symbiotic relationship between arbuscular mycorrhizal fungi (AMF) and plant roots ( Akiyama et al., 2005 ; Akiyama et al., 2010 ) and as important regulators of plant development ( Gomez-Roldan et al., 2008 ; Umehara et al., 2008 ). KARs, a group of butenolide compounds found in smoke, were first identified as stimulants for seed germination in fire-following adapted species. Genetic analysis of KAR signaling revealed an unexpected link to SLs. There is compelling evidence suggesting that KARs act as natural analogs of an unidentified endogenous signal known as the KAI2 ligand (KL). This KAR/KL signaling pathway regulates various plant developmental processes, including germination, photomorphogenesis in seedlings, and root and root hair growth ( Waters and Nelson, 2023 ). In addition, KAR/KL signaling has been shown to influence arbuscular mycorrhizal symbiosis ( Gutjahr et al., 2015 ). Tomato DLK2 ( SlDLK2 ) is a new component of the complex plant-mediated mechanism regulating the life cycle of arbuscules in AM symbiosis. Interestingly, SlDLK2 interacts with DELLA, a protein that regulates arbuscule formation/degradation in AM roots ( Ho-Plágaro et al., 2021 ). The DELLA-gibberellin module plays a central role in regulating arbuscule formation ( Floss et al., 2013 ; Martin-Rodriguez et al., 2016 ). In a complex with DELLA proteins, CYCLOPS regulates the expression of RAM1 ( Pimprikar et al., 2016 ), which encodes a GRAS-domain transcription factor that constitutes a master regulator for the expression of genes involved in arbuscule development and nutrient exchanges ( Rich et al., 2015 ). The previous study by Ho-Plágaro et al. (2021) showed that SlDLK2 ectopic expression downregulates AM-responsive genes, even in the absence of symbiosis, including well-known AM marker genes involved along several stages of arbuscule life cycle. In the present study, we performed an in-depth analysis of changes directed by SlDLK2 overexpression (OE) in tomato roots based on previous RNA sequencing data. We compared and evaluated in detail differentially expressed genes and the associated gene ontology (GO) terms enriched in tomato roots under two different conditions: ectopic overexpression of SlDLK2 or AM colonization. Our primary aim was to identify differentially expressed genes involved in the response of tomato roots to AM formation and mediated by SlDLK2 , and we found a clear overrepresentation of genes involved in different hormone biosynthesis pathways important for AM symbiosis. Further hormone content analyses confirmed that SlDLK2 has a relevant role in regulating hormonal balance in the roots.",
"discussion": "4 Discussion D14 and KAI2 receptors that differentiate plant responses to SLs and KARs, respectively, belong to the RsbQ-like family of a,b-hydrolases. A third clade from this family is composed by the DLK2 (DWARF 14-LIKE2) proteins, which are structurally similar to the D14/KAI2 receptors, but whose function is not so well-known. In tomato, SlDLK2 was recently shown to be involved in the complex plant-mediated signaling mechanism that regulates the life cycle of arbuscules and plays a central role in the negative regulation of arbuscule branching during AM formation ( Ho-Plágaro et al., 2021 ). Clear evidence shows that most phytohormones have an essential regulatory role from early stages in the presymbiotic signaling to later stages of AM development (revised by Pozo et al. (2015) , Bedini et al. (2018) , and Liao et al. (2018) ). We observed that many gene ontology terms associated with hormone regulation, response, and signaling were commonly overrepresented upon SlDLK2 OE and mycorrhization, suggesting that the role SlDLK2 on the regulation of mycorrhization might be mediated by a regulation of hormone balance. In this study, we show an in-depth analysis of the transcriptional changes triggered by SlDLK2 overexpression in roots from composite plants based on the data obtained by Ho-Plágaro et al. (2021) , and we focused our attention on a general repression of genes involved in hormone biosynthesis, with a special interest on the isoprenoid biosynthesis-related genes. In addition, we obtained stable-transformed SlDLK2 OE lines, and we confirmed that the content of several hormones (JA, ABA, SLs, and probably auxins) was effectively reduced upon SlDLK2 overexpression. Gene expression and hormone content analyses revealed that several genes involved in jasmonic acid (JA) biosynthesis ( OPR3 , AOC , AOS, and LOX ) were repressed in SlDLK2 OE roots and that JA was the measured hormone showing the most strongly reduced contents (approximately a 10-fold decrease) upon SlDLK2 overexpression ( Figures 1 , 3 ). Similarly, abscisic acid (ABA) and indoleacetic acid (IAA) contents were reduced in the SlDLK2 OE roots ( Figure 3 ), and this effect was accompanied by a repression of genes putatively involved in ABA ( AAO3 , ABA3 , NCED1 , and NCED2 ) and auxin ( TAR2a , AMI1 , and AAO1 ) biosynthesis ( Figure 1 ), and an induction of genes putatively involved in ABA degradation ( Table 3 ). Finally, strigolactone (SL) biosynthesis genes were also repressed ( CCD7 , D27 , CCD8, and MAX1 ), and SL contents decreased in SlDLK2 OE roots. Notably, for all these hormones (JA, ABA, SLs, and auxins) whose contents were reduced upon SlDLK2 overexpression, a positive role on mycorrhization has been described. In the case of jasmonic acid (JA), although some conflicting data exist, multiple studies support its role as a positive regulator during mycorrhization ( Herrera-Medina et al., 2008 ; Leon-Morcillo et al., 2012 ). For instance, tomato mutants deficient in JA ( spr2 ) exhibit reduced mycorrhizal colonization, whereas the overexpression of prosystemin, which displays elevated JA levels, results in the opposite effect ( Tejeda-Sartorius et al., 2008 ; Leon-Morcillo et al., 2012 ; Song et al., 2013 ; Casarrubias-Castillo et al., 2020 ). Abscisic acid (ABA), an apocarotenoid hormone, has been shown to significantly enhance mycorrhizal colonization and increase arbuscule intensity when applied exogenously, especially at low concentrations ( Charpentier et al., 2014 ; Mercy et al., 2017 ). Conversely, tomato mutants with reduced ABA levels ( sitiens ) demonstrate lower mycorrhizal colonization and fewer well-developed arbuscules ( Herrera-Medina et al., 2007 ; Martín-Rodríguez et al., 2011 ). Regarding strigolactones (SLs), another class of apocarotenoid phytohormones are critical for pre-symbiotic signaling. Under phosphate deficiency, SLs are secreted from plant roots into the rhizosphere, signaling the presence of a suitable host for colonization by arbuscular mycorrhizal (AM) fungi ( Yoneyama et al., 2007 ; Kretzschmar et al., 2012 ). This signal stimulates fungal spore germination and hyphal growth, increasing the chances of physical contact between the fungus and host roots and preparing the fungus for symbiosis establishment ( Akiyama and Hayashi, 2006 ; Besserer et al., 2006 ; Besserer et al., 2008 ; Kobae et al., 2018 ; Waters et al., 2017 ). SLs also induce fungal release of diffusible signals, such as short-chain chitin oligomers, which activate the common symbiosis signaling pathway (CSSP) in epidermal root cells, allowing initial colonization ( Akiyama et al., 2005 ; MacLean et al., 2017 ; Waters et al., 2017 ). Consistent with these findings, mycorrhizal colonization is markedly reduced in plant mutants that are deficient in SL biosynthesis and transport ( Koltai et al., 2010 ; Kretzschmar et al., 2012 ; Yoshida et al., 2012 ). Regarding auxins, several studies have also indicated that this hormone plays a role in the initiation of AM symbiosis as well as in the development and functionality of arbuscules ( Hanlon and Coenen, 2011 ; Etemadi et al., 2014 ; Li et al., 2023 ). A positive correlation between endogenous indole-3-acetic acid (IAA) levels and the extent of mycorrhization, particularly arbuscule formation, has been demonstrated, suggesting that maintaining cellular auxin homeostasis is key to regulating AM symbiosis ( Chen et al., 2022 ). Moreover, root auxin levels are associated with strigolactone exudation, and auxin may control early events in AM symbiosis by modulating SL levels ( Foo, 2013 ). Finally, ethylene is mostly described as a negative regulator of AM in tomato ( de Los Santos et al., 2011 ) and, although we found that genes related to ethylene biosynthesis were repressed upon SlDLK2 overexpression, no changes in ethylene levels were observed in a preliminary analysis of ethylene exudation in control and SlDLK2 OE plants ( Supplementary Figure 1 ). In summary, experimental evidence highly supports the symbiotic positive role of all the hormones showing a reduction in the SlDLK2 roots, suggesting that SlDLK2 OE triggers a repression of JA, ABA, and auxin biosynthesis genes, what reduces the contents of these hormones in the roots and consequently contributes to negatively regulate mycorrhizal colonization, as illustrated in Figure 5 . Figure 5 Model of SlDLK2-mediated signaling on hormone balance during AM symbiosis. SlDLK2 gene induction occurs in arbuscule-hosting cells. The encoded SlDLK2 receptor represses a number of genes involved in the biosynthesis of several AM-promoting hormones including jasmonic acid (JA), ABA, indoleacetic acid (IAA), and strigolactones (SL), what triggers a negative effect on mycorrhizal colonization and arbuscule development. Interestingly, a wide number of genes from the carotenoid/apocarotenoid pathway were repressed ( Figure 1 ). Apocarotenoids are isoprenoid molecules produced in the plastids through the MEP pathway. In plants, the precursor of all isoprenoids is prenyl diphosphate (prenyl-PP), which is synthesized by two independent pathways: the mevalonate (MVA) pathway in the cytoplasm and the 2-C-methyl-d-erythritol 4-phosphate (MEP) pathway in plastids ( Vranová et al., 2013 ). A number of studies show that the MEP pathway is induced during mycorradicin and is responsible for the production of many apocarotenoids that accumulate or are important during AM symbiosis, including not only ABA and strigolactones (SLs) but also other apocarotenoids such as C13 α -ionols, C14 mycorradicin, and zaxinone ( Herrera-Medina et al., 2007 ; Yoneyama et al., 2007 ; Walter, 2020 ; Ablazov et al., 2023 ). In our study, we observed that the MEP pathway was repressed in roots overexpressing SlDLK2 , suggesting that the lower mycorrhization upon SlDLK2 overexpression might be due to the reduced biosynthesis of these AM signaling molecules that derive from the MEP pathway. By contrast, we observed that many genes putatively involved in the mevalonate pathway were induced in SlDLK2 OE roots ( Supplementary Table 5 ), probably as an indirect plant response to provide cytosolic IPP from the MVA pathway to the plastids to compensate the reduced IPP plastid precursors in these roots, as many studies show a crosstalk between cytosolic and plastidial IPP ( Hemmerlin et al., 2003 ; Mendoza-Poudereux et al., 2015 ; Henry et al., 2018 ; Wagatsuma et al., 2018 ). Supporting the strong induction of the MEP pathway upon SlDLK2 overexpression, we observed that SlDLK2 OE repressed the AM-inducible phytoene synthase PSY3 gene ( Figure 2 ). In Medicago truncatula , the PSY3 enzyme is known to be necessary for the production of strigolactones and C13 α-ionol/C14 mycorradicin apocarotenoids, and its knockdown negatively affects mycorrhization measured with the fungal marker RiBTUB ( Stauder et al., 2018 ). Moreover, in agreement with our results, PSY3 isogenes from tomato and Medicago are shown to be co-regulated with upstream genes ( DXS2 ) and downstream carotenoid cleavage steps toward SLs ( D27, CCD7, and CCD8 ), suggesting a coordinated induction of the carotenoid and apocarotenoid pathways for the delivery and usage of precursors for apocarotenoid formation, as proposed by Stauder et al. (2018) . Moreover, we also observed that the AM-inducible gene encoding the zaxinone synthase ZAS4 was also repressed in SlDLK2 OE roots ( Figure 2 ). The homolog genes OsZAS and OsZAS2 from rice have an expression associated with arbusculated cells and are involved in the biosynthesis of AM-related apocarotenoids, being required for proper mycorrhizal colonization ( Wang et al., 2019 ; Votta et al., 2022 ; Ablazov et al., 2023 ). Although OsZAS and OsZAS2 form zaxinone in vitro , an apocarotenoid that regulates strigolactone biosynthesis ( Votta et al., 2022 ; Ablazov et al., 2023 ), contradictory results in experiments with exogenous zaxinone treatments suggest that in addition to zaxinone, these AM-related zaxinone synthases can form in planta a yet unidentified apocarotenoid for optimal mycorrhization, as proposed by Votta et al. (2022) . Finally, we investigated whether the hormonal alterations observed in roots of SlDLK2 OE plants also occurred in leaves. The DLK2 receptor is thought to have alternative potential functions in other tissues apart from its role in AM symbiosis, as SlDLK2 is highly expressed in leaves and a photomorphogenic phenotype has been observed in DLK2 mutants in the non-mycorrhizal plant Arabidopsis ( Végh et al., 2017 ). In this study, we show that alterations in hormonal contents in the leaves of SlDLK2 OE plants were not relevant ( Figure 4 ), indicating that SlDLK2 overexpression may participate in different specific signaling pathways in roots and leaves, probably by the coordinated action of DLK2 with other elements that are specific of the different tissues. Although overexpression can cause pleiotropic off-target effects by influencing multiple biological processes beyond the gene’s intended function, to date, ectopic gene expression is considered a valuable tool for gene functional characterization and for identifying candidate target genes through RNA-seq analyses. The role of SlDLK2 during mycorrhization has been previously described by Ho-Plágaro et al. (2021) , using both RNAi and overexpression in composite plants with hairy roots, validating the use of this model for characterizing SlDLK2 functionality. Moreover, our RNA-seq results point to specific effects of SlDLK2 overexpression, rather than off-target effects, as many genes from specific pathways are altered in the same direction. Reinforcing this idea, we observed that alterations in hormone content in leaves upon SlDLK2 overexpression completely differ from those occurring in roots, where SlDLK2 is biologically active. In summary, our study clearly shows that SlDLK2 overexpression in the roots triggers an overall repression of genes for the biosynthesis of different hormones, with the consequent reduction in the levels of hormones with a previous described AM-promoting role, including jasmonic acid, auxins, and the apocarotenoid compounds ABA and strigolactones. Moreover, the general repression of genes from the MEP pathway and the apocarotenoid biosynthesis pathway indicates that the reduction in other apocarotenoid compounds might be also crucial for the regulatory function of DLK2 . Altogether, we conclude that the DLK2 receptor might be an important element for the repression of the biosynthesis of important hormones and apocarotenoids to negatively regulate mycorrhization. To deepen the understanding of the underlying mechanisms, it would be highly interesting to perform protein–protein interaction assays between SlDLK2 and hormone-biosynthetic enzymes or hormone-related transcription factors to elucidate the mechanisms behind the SlDLK2-mediated regulation of hormonal biosynthesis pathways during mycorrhization. In this regard, the suppressor proteins JASMONATE ZIM DOMAIN PROTEIN (JAZ) and MYC2 are key components in the crosstalk between jasmonic acid (JA) and other plant hormones in plant growth and stress responses. Specifically, the molecular cascade involving the JAZ-MYC2-DELLA-PIF signaling module has been suggested to participate in the crosstalk between JA and GA signaling pathways ( Yang et al., 2019 ). In addition, JAZ-MYC2 participates in the crosstalk between JA and ABA signaling pathways, influencing plant growth and defense responses ( Chen et al., 2011 ). Interestingly, SlDLK2 interacts with DELLA ( Ho-Plágaro et al., 2021 ), a protein that regulates arbuscule formation and degradation in AM roots. Accordingly, we can speculate that specific SlDLK2–DELLA interactions may interfere with both DELLA’s role as an activator of the transcription of plant genes required for arbuscule formation ( Pimprikar et al., 2016 ), as well as the hormonal signaling crosstalk pathways in which DELLA is implicated. Furthermore, although we have proposed a direct suppression of JA and ABA biosynthesis by SlDLK2, the complex feedback loops intrinsic to hormone signaling pathways warrant further investigation. For instance, it is well-established that ABA can induce JA biosynthesis under certain conditions ( Wang et al., 2018 ; Ju et al., 2019 ), suggesting that suppression of ABA could be expected to decrease JA levels as well."
} | 5,453 |
31396058 | PMC6664086 | pmc | 2,281 | {
"abstract": "Cortical synapse organization supports a range of dynamic states on multiple spatial and temporal scales, from synchronous slow wave activity (SWA), characteristic of deep sleep or anesthesia, to fluctuating, asynchronous activity during wakefulness (AW). Such dynamic diversity poses a challenge for producing efficient large-scale simulations that embody realistic metaphors of short- and long-range synaptic connectivity. In fact, during SWA and AW different spatial extents of the cortical tissue are active in a given timespan and at different firing rates, which implies a wide variety of loads of local computation and communication. A balanced evaluation of simulation performance and robustness should therefore include tests of a variety of cortical dynamic states. Here, we demonstrate performance scaling of our proprietary Distributed and Plastic Spiking Neural Networks (DPSNN) simulation engine in both SWA and AW for bidimensional grids of neural populations, which reflects the modular organization of the cortex. We explored networks up to 192 × 192 modules, each composed of 1,250 integrate-and-fire neurons with spike-frequency adaptation, and exponentially decaying inter-modular synaptic connectivity with varying spatial decay constant. For the largest networks the total number of synapses was over 70 billion. The execution platform included up to 64 dual-socket nodes, each socket mounting 8 Intel Xeon Haswell processor cores @ 2.40 GHz clock rate. Network initialization time, memory usage, and execution time showed good scaling performances from 1 to 1,024 processes, implemented using the standard Message Passing Interface (MPI) protocol. We achieved simulation speeds of between 2.3 × 10 9 and 4.1 × 10 9 synaptic events per second for both cortical states in the explored range of inter-modular interconnections.",
"introduction": "1. Introduction At the large scale, the neural dynamics of the cerebral cortex result from an interplay between local excitability and the pattern of synaptic connectivity. This interplay results in the propagation of neural activity. A case in point is the spontaneous onset and slow propagation of low-frequency activity waves during the deep stages of natural sleep or deep anesthesia (Hobson and Pace-Schott, 2002 ; Destexhe and Contreras, 2011 ; Sanchez-Vives and Mattia, 2014 ; Reyes-Puerta et al., 2016 ). The brain in deep sleep expresses slow oscillations of activity at the single-neuron and local network levels which, at a macroscopic scale, appear to be synchronized in space and time as traveling waves (slow-wave activity, SWA). The “dynamic simplicity” of SWA is increasingly being recognized as an ideal test bed for refining and calibrating network models composed of spiking neurons. Understanding the dynamical and architectural determinants of SWA serves as an experimentally grounded starting point to tackle models of behaviorally relevant, awake states (Han et al., 2008 ; Curto et al., 2009 ; Luczak et al., 2009 ). A critical juncture in such a logical sequence is the description of the dynamic transition between SWA and asynchronous, irregular activity (AW, asynchronous wake state) as observed during fade-out of anesthesia, for instance, the mechanism of which is still a partially open problem (Curto et al., 2009 ; Steyn-Ross et al., 2013 ; Solovey et al., 2015 ). To help determine the mechanism of this transition, it may be of interest to identify the factors enabling the same nervous tissue to express global activity regimes as diverse as SWA and AW. Understanding this repertoire of global dynamics requires high-resolution numerical simulations of large-scale networks of neurons which, while keeping a manageable level of simplification, should be realistic with respect to both non-linear excitable local dynamics and to the spatial dependence of the synaptic connectivity (as well as the layered structure of the cortex) (Bazhenov et al., 2002 ; Hill and Tononi, 2005 ; Potjans and Diesmann, 2014 ; Krishnan et al., 2016 ). Notably, efficient brain simulation is not only a scientific tool, but also a source of requirements and architectural inspiration for future parallel/distributed computing architectures, as well as a coding challenge on existing platforms. Neural network simulation engine projects have focused on: flexibility and user friendliness, biological plausibility, speed and scalability [e.g., NEST (Gewaltig and Diesmann, 2007 ; Jordan et al., 2018 ), NEURON (Hines and Carnevale, 1997 ; Carnevale and Hines, 2006 ), GENESIS (Wilson et al., 1989 ), BRIAN (Goodman and Brette, 2009 ; Stimberg et al., 2014 )]. Their target execution platforms can be either homogeneous or heterogeneous (e.g., GPGPU-accelerated) high-performance computing (HPC) systems, (Izhikevich and Edelman, 2008 ; Nageswaran et al., 2009 ; Modha et al., 2011 ), or neuromorphic platforms, for either research or application purposes [e.g., SpiNNaker (Furber et al., 2013 ), BrainScaleS (Schmitt et al., 2017 ), TrueNorth (Merolla et al., 2014 )]. From a computational point of view, SWA and AW pose different challenges to simulation engines, and comparing the simulator performance in both situations is an important element in assessing the general value of the choices made in the code design. During SWA, different and limited portions of the network are sequentially active, with a locally high rate of exchanged spikes, while the rest of the system is almost silent. On the other hand, during AW the whole network is homogeneously involved in lower rate asynchronous activity. In a distributed and parallel simulation framework, this raises the question of whether the computational load on each core and the inter-process communication traffic are limiting factors in either cases. We also need to consider that activity propagates for long distances across the modeled cortical patch, therefore the impact of spike delivery on the execution time depends on the chosen connectivity. Achieving a fast and flexible simulator, in the face of the above issues, is the purpose of our Distributed and Plastic Spiking Neural Networks (DPSNN) engine. Early versions of the simulator (Paolucci et al., 2013 ) originated from the need for a representative benchmark developed to support the hardware/software co-design of distributed and parallel neural simulators. DPSNN was then extended to incorporate the event-driven approach of Mattia and Del Giudice ( 2000 ), implementing a mixed time-driven and event-driven strategy similar to the one introduced in Morrison et al. ( 2005 ). Here, we report the performances of DPSNN in both slow-wave (SW) and AW states, for different sizes of the network and for different connectivity ranges. Specifically, we discuss network initialization time, memory usage and execution times, and their strong and weak scaling behavior.",
"discussion": "4. Discussion We presented a parallel distributed neural simulator, with emphasis on the robustness of its performance and scaling with respect to quite different collective dynamical regimes. This mixed time- and event-driven simulation engine ( Figure 1 ) has been used to simulate large-scale networks including up to 46 million point-like spiking neurons interconnected by 70 billion instantaneous current synapses. The development of DPSNN originated from the need for a simple, yet representative benchmark (i.e., a mini-application) developed to support the hardware/software co-design of distributed and parallel neural simulators. Early versions of DPSNN (Paolucci et al., 2013 ) have been used to drive the development of the EURETILE system (Paolucci et al., 2016 ) in which a custom parallelization environment and the APEnet hardware interconnect were tested. DPSNN was then extended to incorporate the event-driven approach of Mattia and Del Giudice ( 2000 ), implementing a mixed time-driven and event-driven strategy inspired by Morrison et al. ( 2005 ). This simulator version is also currently included among the mini-applications driving the development of the interconnect system of the EXANEST ARM-based HPC architecture (Katevenis et al., 2016 ). In the framework of the Human Brain Project ( https://www.humanbrainproject.eu ), DPSNN is used to develop high-speed simulation of SW and AW states in multi-modular neural architectures. The modularity results from the organization of the network into densely connected modules mimicking the known modular structure of cortex. In this modeling framework, the inter-modular synaptic connectivity decays exponentially with the distance. In this section, we first discuss the main strengths of the simulation engine and then put our work into perspective by comparing the utility and performances of such a specialized engine with those of a widely used general-purpose neural simulator (NEST). 4.1. Speed, Scaling, and Memory Footprint at Realistic Neural and Synaptic Densities DPSNN is a high-speed simulator. The speed ranged from 3 × 10 9 to 4.1 × 10 9 equivalent synaptic events per wall-clock second, depending on the network state and the connectivity range, on commodity clusters including up to 1,024 hardware cores. As an order of magnitude, the simulation of a square cortical centimeter (~27 × 10 9 synapses) at realistic neural and synaptic densities is about 30 times slower than real time on 1024 cores ( Figure 7 ). DPSNN is memory parsimonious: the memory required for the above square cortical centimeter is 837 GB (31 byte/synapses, including all library overheads), which is in the range of commodity clusters with few nodes. The total memory consumption ranges between 25 and 32 byte per recurrent synapse for the whole set of simulated neural networks and all configurations of the execution platform (from 1 to 1,024 MPI processes, Figure 5 ). The choice of representing with 4 bytes the identities of presynaptic and post-synaptic neurons limits the total number of neurons in the network to 2 billions. However, this is not yet a serious limitation for neural networks including thousands of synapses per neuron to be simulated on execution platforms including few hundreds of multi-core nodes. The size of neural ID representation will have to be enlarged for simulation of systems at the scale of human brains. Concerning synaptic weights, as already discussed, static synapses are stored with only two bytes of precision, but injection of current and neural dynamics is performed with double precision arithmetic. When plasticity is turned on, single precision floating-point storage of LTP and LTD contributions is adopted. The engine has very low initialization times. DPSNN takes about 4 s to set up a network with ~17 × 10 9 synapses ( Figure 4 ). We note that the initialization time is relevant, especially when many relatively short simulations are needed to explore a large parameter space. Its performance is robust: good weak and strong scaling have been measured in the observed range of hardware resources and for all sizes of simulated cortical grids. The simulation speed was nearly independent from the mean firing rate ( Figure 9 ), the range of connectivity ( Figure 8 ), and from the cortical dynamic state (AW/SW) ( Figure 7 ). 4.2. Key Design Guidelines of the Simulation Engine A few design guidelines contribute to the speed and scaling of the simulation engine. We kept as driving criteria the goals of increasing the locality in memory accesses, the reduction of interprocess communication, an ordered traversal of lists and a reduction of backward searches and random accesses, the minimization of the number of layers in the calling stack, and a complete distribution of computation and storage among the cooperating processes with no need for centralized structures. We stored in the memory of a process a set of spatially neighboring neurons, incoming synapses, and outgoing axons. For the kind of spatially organized neural networks described in this paper, this reduces the size of the payload and the required number of interprocess communications. Indeed, many spiking events will need to reach target neurons (and synapses) stored in the same process of the spiking neuron itself. An explicit ordering in memory is adopted for outgoing and ingoing communication channels (representing the first branches in an axonal like arborizations). Explicit ordering in memory is also adopted for other data structures, like: the lists of incoming spiking events and recurrent synapses, the set of incoming event queues associated to different synaptic delays. Lists are implemented as arrays, without internal pointers. Also explicit ordering is used for the set of target neurons. Neurons are progressively numbered with lower bits in their identifiers associated to their local identity and higher bits conveying the id of the hosting process. The explicit ordering of neurons and synapses reduces the time spent during the sequence of memory accesses that will be followed during the simulation steps. In particular, lists are traveled only once per communication step, e.g., first searching for target synapses that are targets of spiking events and then through an ordered list of target neurons. A third design criteria has been to keep the stack of nested function calls short, following the scheme described in Figure 1 , and attempting to group at each level multiple calls to computational methods accessing the same memory structure. As an example, both the generation of external synaptic events (e.g., Poissonian stimulus) and the temporal reordering of recurrent and external synaptic events targeting a specific neuron, is deferred to the computation of the individual neural dynamics. The execution of this routine is supported by local queues storing all the synaptic currents targeting the individual neuron. This queue of events is accessed during a single call of the routine computing the dynamics of the neuron. In this case the data structure supports both locality in memory and in computation. Similar design guidelines would be problematic to follow for general purpose simulation engines that are supposed to support maximum flexibility in the description and simulation of the data structures and of the dynamics of neurons, axonal arborizations and synapses. Moreover, higher abstraction requires separating the functionality of the simulation engine into independent modules. This would dictate a higher number of layers in the calling stack, more complex interfaces and data structures that hide details like their internal memory ordering. 4.3. Comparison With a State-of-the-Art User-Friendly Simulator and Motivations for Specialized Engines There is a widely felt need for versatile, general-purpose neural simulators that offer a user-friendly interface for designing complex numerical experiments and provide the user with a wide set of models of proven scientific value. This boosted a number of initiatives (notably the NEST initiative, now central to the European Human Brain Project). However, such flexibility comes at a price. Performance-oriented engines, missing all the layers required for offering user generality and flexibility, contain the bare minimum code. In the case of DPSNN, this resulted in higher simulation speed, reduced memory footprint, and diminished initialization times (see section 3.4 and Table 3 ). In addition, optimization techniques developed for such engines on use-cases of proven scientific value can offer a template for future releases of general-purpose simulators. Indeed, this is what is happening in the current framework of cooperation with the NEST development team. Finally, engines stripped down to essential kernels constitute more easily manageable mini-application benchmarks for the hardware/software co-design of specialized simulation systems, because of easier profiling and simplified customizations on system software environments and hardware targets under development. 4.4. Future Works The present implementation of DPSNN demonstrated to be efficient for homogeneous bidimensional grids of neural columns and for their mapping of up to 1,024 processes, and this facilitates a set of interesting scientific applications. However, further optimization could improve DPSNN performance, either in the perspective of moving simulations toward million-core exascale platforms or targeting real-time simulations at smaller scale (Simula et al., 2019 ), in particular addressing sleep-induced optimizations in cognitive tasks like classification (Capone et al., 2019a ). For instance, we expect that the delivery of spiking messages will be a key element to be further optimized (e.g., using a hierarchical communication strategy). This will also be beneficial for an efficient support of white-matter long-range connectivity (brain connectomes) between multiple cortical areas. A full exploitation of the model requires parameters tuned exploiting information provided by experimental data. Addressing this goal our team is improving tools for the analysis of both electrophysiological (De Bonis et al., 2019 ) and optical (Celotto et al., 2018 ) recordings (micro-ECoG and wide-field Calcium Imaging). The main aim is the spatio-temporal characterization of SWA. We also plan to apply inference methods to obtain refined maps of connectivity and excitability for insertion in the simulated model."
} | 4,348 |
28211847 | PMC5437931 | pmc | 2,282 | {
"abstract": "Most anoxic environments are populated by small (<10 μm) heterotrophic eukaryotes that prey on different microbial community members. How predatory eukaryotes engage in beneficial interactions with other microbes has rarely been investigated so far. Here, we studied an example of such an interaction by cultivating the anerobic marine flagellate, Carpediemonas frisia sp. nov. (supergroup Excavata), with parts of its naturally associated microbiome. This microbiome consisted of so far uncultivated members of the Deltaproteobacteria, Bacteroidetes, Firmicutes, Verrucomicrobia and Nanoarchaeota. Using genome and transcriptome informed metabolic network modeling, we showed that Carpediemonas stimulated prokaryotic growth through the release of predigested biomolecules such as proteins, sugars, organic acids and hydrogen. Transcriptional gene activities suggested niche separation between biopolymer degrading Bacteroidetes, monomer utilizing Firmicutes and Nanoarchaeota and hydrogen oxidizing Deltaproteobacteria. An efficient metabolite exchange between the different community members appeared to be promoted by the formation of multispecies aggregates. Physiological experiments showed that Carpediemonas could also benefit from an association to these aggregates, as it facilitated the removal of inhibiting metabolites and increased the availability of prey bacteria. Taken together, our results provide a framework to understand how predatory microbial eukaryotes engage, across trophic levels, in beneficial interactions with specific prokaryotic populations.",
"introduction": "Introduction Small heterotrophic eukaryotes (<10 μm) perform essential functions in marine ecosystems ( Pernthaler, 2005 ; Edgcomb, 2016 ). For example, these eukaryotes capture and digest similar-sized microbes and release a mixture of nutrients, dissolved organic carbon and detritus to the environment, stimulating growth of prokaryotes. At the same time, microbial eukaryotes can become prey for the marine mesofauna and so link prokaryotic primary production to higher trophic levels ( Sherr and Sherr, 2002 ). In environments with high influxes of organic carbon, microbial activity often leads to a depletion of oxygen ( Glud, 2008 ). Most eukaryotes cope with anoxia by switching from oxygen respiration to fermentation ( Müller et al. , 2012 ; Stairs et al. , 2015 ). Just like in prokaryotes, fermentation in eukaryotes proceeds via glycolysis followed by the decarboxylation of pyruvate ( Müller et al. , 2012 ). In strictly anerobic microbial eukaryotes, pyruvate decarboxylation often takes place in mitochondria that lost their capability to respire oxygen ( Boxma et al. , 2005 ). These mitochondria recycle reducing equivalents by transferring electrons to organic metabolites or protons (H + ). Depending on the fermentation pathway used, this leads to the production of molecular hydrogen, fatty acids, alcohols and amino acids ( Müller et al. , 2012 ). In marine sediments, fermentation reactions may be syntrophically linked into a cascade of downstream metabolism, including respiration and further fermentation reactions ( Fenchel and Jørgensen, 1977 ). This mediates a direct consumption of fermentation products and keeps their steady-state concentrations low. It is well established that such syntrophic interactions provide important bioenergetic advantages to prokaryotes ( Sieber et al. , 2012 ). From the thermodynamic perspective, consumption of fermentation products enables fermentation reactions, which require low product concentrations. At low product concentrations, fermentative prokaryotes can maximize their growth efficiency and colonize environments with scarce energy resources ( McInerney e t al. 2007 ). Among eukaryotes, similar nutritional interactions are found in ciliates or flagellates that harbor hydrogen-scavenging symbionts ( van Hoek et al. , 2000 ; Boxma et al. , 2005 ; Ohkuma et al. , 2015 ). Parts of these symbionts might also be digested by their host, providing it with an additional carbon source. Yet, their main function appears to be the creation of a biochemical environment that favors fermentation reactions in the host’s cytosol ( Ohkuma et al. , 2015 ; Hamann et al., 2016 ). Even though symbiont-free eukaryotes are vital components of anerobic microbial communities ( Edgcomb et al. , 2002 ; Wylezich and Jürgens, 2011 ), it has remained largely unaddressed how their physiological activity is affected by syntrophic interactions with free-living prokaryotes and vice versa. This prompted us to enrich the marine flagellate Carpediemonas frisia sp. nov. (Fornicata, supergroup Excavata), with part of its naturally associated bacterial and archaeal community. For this enrichment culture, we applied an experimental approach that combined physiological experiments with metabolic network modeling informed by genomics and transcriptomics. Our analysis focused on two main questions: (i) how do prokaryotes take advantage from the predatory behavior and metabolic activity of C. frisia ? (ii) How does the biochemical activity of prokaryotes affect the fitness of C. frisia ?",
"discussion": "Discussion In this study, we showed that anoxic incubation of a marine sediment supplemented with prey bacteria as carbon source and sulfate as electron acceptor led to the enrichment of the predatory protist C. frisia and co-enrichment of a prokaryotic community. After five subsequent culture transfers in sediment-free medium, this community consisted of distinct populations affiliated with Bacteroidetes, Firmicutes, Deltaproteobacteria, Verrucomicrobia and Nanoarchaeota. Metabolic interactions were inferred from the analysis of metagenomic and transcriptomic sequence data. C. frisia appeared to provide two major nutritional opportunities for prokaryotes: First, the degradation of incompletely digested macromolecular organic material and second, the oxidation of sugars, organic acids and hydrogen with sulfate. Physiological experiments confirmed that the fitness of C. frisia was directly impacted by the hydrogen oxidizing activity of Deltaproteobacteria. Our genomic and transcriptomic analysis suggested that C. frisia produced hydrogen through the simultaneous (confurcating) oxidation of NAD(P)H and ferredoxin using an enzyme complex that consisted of the catalytic subunits of respiratory Complex I (NouE/ NuoF) and a Fe-hydrogenase. This reaction is very common among fermentative bacteria ( Schut and Adams, 2009 ) and has also been predicted to be used by a variety of eukaryotic species ( Stairs et al. 2015 ). However, the available free energy released during NAD(P)H-dependent hydrogen production decreases exponentially with increasing product concentration. For the synergistic production of molecular hydrogen from NAD(P)H and ferredoxin to proceed, the maximum hydrogen concentration is 0.1 m m (at NADH/NAD + =10; Stams and Plugge, 2009 ). Many fermenting eukaryotes have evolved symbiotic relationships with hydrogen-scavenging prokaryotes, which allow them to maintain intracellular hydrogen levels below this threshold. For example, ciliates and amoeba often harbor hydrogen-consuming methanogens inside their cytosol ( van Bruggen et al. , 1985 ; van Hoek et al. , 2000 ; Boxma et al. , 2005 ). Another example is the breviate Lenisia limosa , which forms epibiotic associations with hydrogen oxidizing Arcobacter ( Hamann et al. , 2016 ). In this association, the maximum hydrogen concentration has been reported to be 20 times lower than predicted for C. frisia . This can be explained by the observation that L. limosa mainly uses NAD(P)H for hydrogen production, a reaction that is more prone to product inhibition than the simultaneous oxidation of NAD(P)H and ferredoxin. This may also explain why C. frisia can sustain its metabolism without the tight symbiotic association observed for L. limosa and other hydrogen-producing eukaryotes. In any case, if the maximum hydrogen concentration is exceeded, NAD(P)H needs to be re-oxidized by reduction of organic metabolites such as pyruvate or succinate. Less energy is conserved in this metabolic bypass, leading to a reduced growth efficiency. In addition, NAD(P)H-dependent pyruvate oxidation may lead to the production of ethanol and acetaldehyde ( Müller et al. , 2012 ), which are growth inhibitors at elevated concentrations ( Maiorella et al., 1983 ). The removal of inhibitory metabolic byproducts by Deltaproteobacteria therefore provides an important fitness benefit to C. frisia . The second major metabolic function performed by the bacteria co-enriched with C. frisia appeared to be the degradation of incompletely digested, macromolecular organic material. Microscopic imaging showed that this material originated from fecal pellets excreted by C. frisia . As we did not provide any organic carbon other than prey bacteria to our enrichment, this material likely represented a major carbon source for bacterial growth. This also explains the high abundance of bacterial species in our enrichment that are adapted to use this type of material. Bacteroidetes are known as consumers of macromolecules ( Fernández-Gómez et al. , 2013 ) and were inferred to use detritus as their substrates. The presence and activity of the two populations of Firmicutes also suggested the availability of smaller organic molecules such as oligopeptides, amino acids and sugars. These might be released either directly by C. frisia , or during the hydrolysis of proteins and polysaccharides by Bacteroidetes. Similar niche partitioning between Bacteroidetes and Firmicutes is known for many other organic carbon-rich environments (for example, the digestive tract of animals ( Lozupone et al. , 2012 ) or marine oxygen minimum zones ( Wright et al. , 2012 )). The processing of organic waste products by Bacteroidetes and Firmicutes is, therefore, in good agreement with the known physiological capabilities of these species. A typical characteristic for syntrophic bacterial communities is the formation of multispecies aggregates. Aggregate formation can enhance the exchange of metabolites between different populations, thereby accelerating fermentative reactions and degradation of organic material ( Schink and Thauer, 1988 ). In our enrichment, aggregates most likely arose from bacterial colonization of detritus excreted by C. frisia . As C. frisia was often directly associated to these aggregates, it could likely benefit from the high turnover of hydrogen and other metabolites within these aggregates. In addition, C. frisia may benefit from an increased availability of prey bacteria in aggregates. This way, excretion of detritus by C. frisia and the subsequent bacterial colonization of this material, can be interpreted as a primitive form of microbial agriculture. It will be an interesting research avenue, to address how widespread similar bi-directional metabolic interactions between predatory flagellates and bacteria are in the environment. To date, there are many reported cases of specific protists-bacterial co-occurrences, which may be a result of metabolic interdependencies. For example, community fingerprinting by 16S rDNA amplicon sequencing showed that a bloom of Carpediemonas and related Diplomonads was accompanied by a specific enrichment for Bacteroidetes, Firmicutes and known sulfate-reducing Deltaproteobacteria ( Holmes et al. , 2013 ). Another study based on single-cell sequencing of flow-sorted planktonic eukaryotes showed that Firmicutes and Deltaproteobacteria are preferentially associated with heterotrophic protists ( Martinez-Garcia et al. , 2012 ). So far, a common assumption is that an enrichment of specific prokaryotic phenotypes stimulates the growth of flagellates that are adapted to prey on these species ( Pernthaler, 2005 ; Holmes et al. , 2013 ). The results presented here, however, show that for anoxic environments, the opposite provides an equally compelling explanation—predatory microbial eukaryotes can also stimulate the growth of specific prokaryotic populations, which are adapted to consume their metabolic waste products. In conclusion, we have shown that C. frisia is a newly identified anerobic flagellate, that likely engages in a hydrogen- and organic carbon-based metabolic syntrophy with sulfate-reducing Deltaproteobacteria. This syntrophy appears to benefit C. frisia by removing inhibitory metabolites and by creating a biochemical environment that promotes fermentative hydrogen production. In return, C. frisia appears to provide its associated microbiota with predigested organic macromolecules, sugars, organic acids and hydrogen. The described metabolic interdependences provide a framework to predict the ecological importance of specific co-occurrences between heterotrophic microbial eukaryotes and prokaryotic populations. In addition, the results show that the concept of metabolic syntrophy, as a form of microbial mutualism, can also apply for inter-kingdom interactions between anaerobic flagellates and free-living prokaryotes. Genus Carpediemonas Ekebom et al. 1996 Carpediemonas frisia sp. nov. Hamann et al. 2017 Description. \n Carpediemonas frisia is a flagellated marine protist with characteristics of the genus ( Simpson and Patterson, 1999 ). The cells are typically around 5 μm long, pear-shaped, slightly compressed laterally, with two flagella of unequal length. Starving cells can be as small as 3 μm. Flagella emerge from the anterior end of the ventral groove. The anterior flagellum is about 5 μm long. The posterior flagellum is two to three times longer and has a thickened, shuffle-like morphology near its origin. The anterior flagellum performs a sweeping motion, bending its tip backwards. In swimming cells, the longer posterior flagellum beats and provides additional motive force. The cells preferentially swim along surfaces, holding the anterior side in contact with the substrate. Small filamentous pseudopodia emerge from all parts of the cell. No cysts are observed. C. frisia is differentiated from Carpediemonas membranifera by the presence of mitochondria-related organelles with slight membrane folding. Habitat. This species was isolated from marine tidal flat sediment collected from a sulfidic sediment layer. The sampling site is commonly known as ‘Janssand’ and located in the German Wadden Sea south of the island Spiekeroog (53.73585 N, 7.69905 E). Etymology. The species name frisia is Latin and refers to the coastal region of the southeastern North Sea ‘Friesland’, which is the historical settlement area of the Friesians. Gene sequence data. The nearly complete SSU rRNA gene of this isolate (strain S16) is deposited in GenBank under accession no. KY031954."
} | 3,702 |
35207157 | PMC8875440 | pmc | 2,283 | {
"abstract": "Electrospinning is a unique technique that can be used to synthesize polymer and metal oxide nanofibers. In materials science, a very active field is represented by research on electrospun nanofibers. Fibrous membranes present fascinating features, such as a large surface area to volume ratio, excellent mechanical behavior, and a large surface area, which have many applications. Numerous techniques are available for the nanofiber’s synthesis, but electrospinning is presented as a simple process that allows one to obtain porous membranes containing smooth non-woven nanofibers. Titanium dioxide (TiO 2 ) is the most widely used catalyst in photocatalytic degradation processes, it has advantages such as good photocatalytic activity, excellent chemical stability, low cost and non-toxicity. Thus, titanium dioxide (TiO 2 ) is used in the synthesis of nanofibrous membranes that benefit experimental research by easy recyclability, excellent photocatalytic activity, high specific surface areas, and exhibiting stable hierarchical nanostructures. This article presents the synthesis of fiber membranes through the processes of electrospinning, coaxial electrospinning, electrospinning and electrospraying or electrospinning and precipitation. In addition to the synthesis of membranes, the recent progress of researchers emphasizing the efficiency of nanofiber photocatalytic membranes in removing pollutants from wastewater is also presented.",
"conclusion": "4. Conclusions The electrospinning technique is an advantageous method for the manufacture of nanofiber membranes generated from different polymers. The morphology of the resulting membranes is determined directly by the parameters of the solution and the operating conditions. Nanofibers produced by the electrospinning technique are an attractive choice for environmental applications. Through the electrospinning technique it is possible to produce membranes with relatively uniform pore size distribution and high pore interconnectivity, compared to conventional techniques. To increase the functionality of membranes obtained by electrospinning, research has led to the incorporation of functionalizing agents (i.e., nanoparticles) in these membranes which has improved their performance for certain applications such as removing bacteria, viruses, or metal ions from wastewater. Membranes obtained by electrospinning can be post-treated, chemically or thermally, with the aim of modifying characteristics such as hydrophobicity, pore size, mechanical integrity, or electrical conductivity. Fabricating TiO 2 fibers and thus immobilizing TiO 2 has been highly applied in the photodegradation of wastewater pollutants. In order to modify the morphology of the fibers, improve their properties and fabricate tough composites, electrospinning is combined with other methods (electrospray, coaxial electrospinning). For the degradation of pollutants, the most widely used photocatalyst is TiO 2 . It shows good performance in dye degradation and has good photochemical stability in comparison to other types of photocatalysts. The anatase/rutile performance was observed to be greater than that of anatase on its own for photodegradation, therefore, calcination has a role in inducing phase transformation. Carbonization or calcination of electrospun TiO 2 may improve crystal growth, create fiber porosity, and modify surface roughness, which affects the fibers’ hydrophilicity. ENPM membranes made of various polymeric materials are used in water treatment due to their efficiency, accessibility, durability, higher surface area-volume ratio, higher reactivity, operation without the addition of chemicals, and smaller pore size. The impressive capacity of electrospun nanofibrous photocatalytic membranes (ENPM) for degradation of harmful organic pollutants is shown in Table 1 . It is observed that the use of an ENPM to remove methylene blue from wastewater showed a 100% treatment efficiency in only 60 min contact time. In addition, by using various ENPMs, a complete degradation of wastewater from Propranolol, Cimetidine (CMT), 4-Chlorophenol, and Bisphenol A was achieved in 60, 40, 100 and 100 min, respectively. However, the development of ENPM has presented several challenges such as neglected filtration properties, low throughput, high energy consumption, low mechanical stability, or leakage of the photocatalyst from the substrate. According to this review, most research has shown that nanofiber photocatalytic membranes should be introduced into wastewater for the photocatalytic degradation of pollutants, but their use as filter membranes has been neglected. Even if huge progress has been made, further research can be done to achieve better performance of electrospun membranes depending on the solution parameters and operating conditions.",
"introduction": "1. Introduction Lately, oil spills as well as industrial wastewater discharges have led to major damage to water resources, seriously affecting human health and leading to an ecological imbalance. In this case, it is necessary to develop efficient technologies for the depollution of wastewater affected by oil spills and industrial wastewater discharges. Technologies such as filtration [ 1 ], adsorption [ 1 , 2 ], catalysis [ 3 , 4 , 5 ], centrifugation [ 6 ], electrocoalescence [ 7 ] or biological treatment [ 4 ] for wastewater treatment have been widely reported, studied, and applied. In recent years, membrane-based technologies have been of interest and have been studied for wastewater treatment due to their relatively low cost, high separation efficiency and ease of operation. Membranes form barriers between two phases, so substances are transported selectively [ 8 ]. Depending on the structure of the membranes, they can be classified as porous membranes or dense membranes [ 9 ]. The selectivity of membranes and their transport properties are strongly dependent on the structure of their pores. The transport mechanisms are different depending on the porous or non-porous membranes and are presented in Figure 1 [ 10 ]. Nanotechnology has become one of the most important areas of research. Nanofibers with high size uniformity, large specific surface area, and high porosity are expected to become efficient nanomaterials for wastewater treatment. To prepare nanofibers, several techniques have been studied such as electrospinning [ 11 ], drawing [ 12 , 13 ], phase separation [ 14 ], template synthesis [ 15 ], self-assembly [ 16 ], interfacial polymerization [ 17 ], etc. Possibly the most attractive technology used to manufacture nanofibers is electrospinning due to its low cost and ease of operation. After reviewing the synthesis methods, materials required and applications of electrospun-based fibers, the researchers proved that these fibers are potential materials for wastewater depollution applications. Due to the fact that titanium dioxide (TiO 2 ) is the most widely known and studied photocatalyst, this article provides a guideline to the fabrication of electrospun-based TiO 2 fibers, and their application in the degradation of organic pollutants. The article begins with electrospinning technology and a comparison with other composite fiber synthesis methods. This is followed by a comprehensive review of some recent electrospun-based TiO 2 fiber applications in organic pollutants degradation and TiO 2 modification. In recent decades, more and more research has been done on electrospun nanofibrous photocatalytic membranes, also called ENPM. The photocatalyst has been integrated in the nanofibrous membrane, and this composite membrane has proven to be effective in degrading wastewater pollutants [ 18 ]."
} | 1,917 |
34557731 | PMC8454563 | pmc | 2,284 | {
"abstract": "Summary The collection of physiological signals as well as the electrical stimulation to the biotissues are significant but challenging. There is a huge gap between the living systems and electronics. Biotissues are wet and soft, while electronics are dry and relatively stiff; biotissues conduct ions, while electronic materials often conduct electrons. As a result, forming a stable interface for bidirectional electrical communications between electronics and the living systems is difficult. In this perspective, we review recent landmark progresses made in this field, and propose a few future directions that scientists may further work on."
} | 161 |
35064156 | PMC8782828 | pmc | 2,285 | {
"abstract": "Research on various neuro-inspired technologies has received much attention. However, while higher-order neural functions such as recognition have been emphasized, the fundamental properties of neural circuits as advanced control systems have not been fully exploited. Here, we applied the functions of central pattern generators, biological neural circuits for motor control, to the control technology of switching circuits for extremely power-saving terminal edge devices. By simply applying a binary waveform with an arbitrary temporal pattern to the transistor gate, low-power and real-time switching control can be achieved. This binary pattern generator consists of a specially designed spiking neuron circuit that generates spikes after a pre-programmed wait time in the six-order range, but consumes negligible power, with an experimental record of 1.2 pW per neuron. This control scheme has been successfully applied to voltage conversion circuits consuming only a few nanowatts, providing an ultra-low power technology for trillions of self-powered edge systems.",
"conclusion": "Conclusion In this study, a binary pattern generator in analogy to the biological central pattern generator was first applied to controlling the switching circuits for IoT devices. We utilized spiking neuron circuits to generate an arbitrary waiting time from 100 ns to 100 ms, and constructed a binary pattern generator for a switching-circuit control with ultra-low power. The generated binary waveform with an arbitrary temporal pattern was used to drive the gates of the switching transistors, and it is shown that the DC-DC voltage conversion circuit can be controlled with only several nanowatt. This is due to an extremely low power consumption of the spiking neuron circuits, as small as 1.2 pW at minimum which is the lowest ever among experimentally demonstrated neuron circuits. The binary pattern generator allows various control functions to be implemented without worrying about the power overhead of the control circuit in self-powered devices. In particular, energy harvesting circuits often involve various types of switching operations 25 , 43 , which could be simply implemented by the binary pattern generator. Low-power sensing could also exploit the advantage of the binary pattern generator; for example, the self-heating gas sensors require fast switching operations to minimize heating power 44 . Finally, it may also be useful for low-power wireless communication especially the one based on the pulse signals, where the intermittent pulse generation is the key to achieve extreme low power on average in the order of nanowatt 45 , 46 .",
"introduction": "Introduction Due to the development of neuroscience and the success of machine learning technology, there has been increasing interest in neuro-inspired technologies that focus on the neural circuits from an engineering viewpoint. The interest is not limited to algorithmic research, but also extends to hardware research, which aims to implement neuro-inspired circuits and systems with lower power consumption and low latency 1 – 8 . However, most of the current hardware research targets the understanding of biological processes or the implementation of machine learning algorithm, and may not fully exploit the tremendous potential of biological neural circuits. In the engineering perspective, biological neural circuits are excellent control systems that control body movements with low power consumption and low latency. It compromises the trade-off between power consumption and response time by automating and decentralizing individual motor control, rather than fast centralized feedback control of body movements. There are two types of such decentralized motor control: one using reflexes that show a fixed response to a specific sensory input, and the other using a central pattern generator that drives motor organs according to a programmed time pattern 9 , 10 . While the reflex circuit is a relatively simple input–output system, the central pattern generator is a neuronal network that autonomously generates temporal patterns without input (Fig. 1 a), and is responsible for complex and rhythmic motor control such as walking, chewing, breathing, and swallowing 11 , 12 . The central pattern generator generally uses spike signals that are heterogeneous with respect to time. This is essentially different from the clock signal in digital circuits that ticks at a fixed period independent of the environment. Compared with the temporally homogeneous clock signal, the spikes at heterogeneous timings can save redundancy, and hence, more suitable for real-time operation. Figure 1 Binary pattern generator. ( a ) A central pattern generator in the human neural network which controls biological motors. ( b ) A concept of an artificial pattern generator which controls switching circuits for DC-DC voltage conversion. ( c ) A binary pattern generator, a specially designed pattern generator for purpose of controlling switching circuits. It consists of a chain of waiting time generators with a wide range of preprogrammed waiting times. ( d ) The spike or bit output signals from the binary pattern generator. The length of the bit signal corresponds to the preprogrammed waiting time of each waiting time generator. This bit signal is directly used for the gate control in the switching circuits. So far, several central pattern generators have been artificially fabricated in the form of electronic circuits for controlling robot actuators 13 – 17 , biological muscles 18 , and physiological systems 19 . However, there are few examples of their application to controlling more general electronic circuits, such as the switching circuits (Fig. 1 b). The switching circuit consists of several transistor switches that are turned on and off rapidly to provide a fine-tuned averaged function. They are essential building blocks in various fields of electronics including the Internet of Things (IoT) devices. In parallel with the artificial central pattern generators for actuators and biological systems, the pattern generators for switching circuits have the potential to provide the real-time and energy-saving control scheme in the future IoT devices. In this study, we constructed an artificial pattern generator that is optimized for switching circuits rather than for actuators or biological systems. The pattern generators in the previous studies often uses the spike rate coding, which convey analog information by spiking frequency. On the other hand, the switching circuits only needs timing information of the switching event, and therefore, it is more convenient to use temporal coding, which convey information as the timing of each spike. One of the essential components for temporal coding is the waiting time generator, which generates a spike signal after a preprogrammed waiting time from the onset of the input voltage. In addition, a wide range of waiting times are needed because the switching circuits sometimes change rapidly on the order of nanoseconds while waiting for most of the time to save energy. In this study, we experimentally demonstrated this wide range of waiting times from 100 ns to 100 ms over six orders of magnitude. In order to generate the arbitrary waiting time and an output spike signal, a technique of spiking neuron circuits with integrate-and-fire function was adopted 20 – 24 . The spiking neuron circuits were optimized solely for the purpose of waiting time generation based on the complementary metal oxide semiconductor (CMOS) technology, and any other biological function was not implemented intentionally. As a result, they achieved extremely low energy consumption in the order of 100 fJ per spike, corresponding to the average power consumption of 1.2 pW in minimum, which is the smallest ever among past experimental demonstrations 20 – 24 . For controlling the switching circuit, the input and the output of the above CMOS spiking neuron circuit were connected to a set-reset latch circuit to generate a binary wave form with a length of the preprogrammed waiting time as shown in Fig. 1 c,d. Then, this binary wave form was simply applied to the transistor gates to control the switching circuits asynchronously. This simple scheme of the “binary pattern generator” can be applied to a versatile switching circuits and can achieve real time control in an extremely low power. For demonstration, we used this binary pattern generator to control the DC-DC voltage conversion circuit, which is essential for IoT devices, and showed by simulation that it can generate a wide range of output power between 8.36 nW and 1.16 mW at approximately 90% efficiency, with negligible control power consumption around three orders smaller than the output power. Thus, controlling switching circuits with a binary pattern generator provides a powerful means to alleviate the power constraints and realize various functions in self-powered IoT terminal devices. In this paper, we first explain the detailed concept of the binary pattern generator, which is a specially designed pattern generator for controlling the switching circuits. Then, we visualize how it works in the demonstration of the switching circuit for DC-DC voltage conversion. Finally, the simulation and experiments are shown for the CMOS spiking neuron circuits that are the essential building blocks for the binary pattern generator. Binary pattern generator Artificial central pattern generators have been previously studied in the field of robotics 13 – 17 . There, the information is conveyed as the analog value of the spike rate (spike rate coding), which can be used to control the actuators of robots. The power consumption is dominated by the driving power of the actuators, and there is little need to reduce the control power consumption to the extreme. The control time scale is also determined by the motion speed of the robot, and so there is little need for a large range of control time down to nano seconds. On the other hand, if the artificial pattern generator is to be used for the switching circuits, it requires extremely low power consumption and a wide range of time scales down to nano seconds. Therefore, for controlling switching circuits, a novel artificial pattern generator is needed to meet all these demands. To address this issue, a binary pattern generator was constructed based on the control scheme of temporal coding. Previously, the similar binary waveforms were also exploited in some of the studies on robot control 13 , 14 . To construct a binary patter generator, we used CMOS spiking neuron circuits with the integrate-and-fire function, which can generate a spike after a preprogrammed waiting time from the onset of the input step voltage. The details will be discussed in the later sections. Each spiking neuron circuit is combined with a set-reset latch circuit, a simplest memory circuit, as shown in Fig. 1 c,d. When the latch circuit is turned on by an input spike signal, the latch circuit generates the output voltage of 1 V. Then, this output voltage is applied to the subsequent spiking neuron circuit, resulting in the output of a spike signal after a preprogrammed waiting time. This output spike from the spiking neuron circuit resets the initial latch circuit and simultaneously turns on the latch circuit of the next stage. By repeating this operation, spike signals are generated at arbitrary time intervals ( V A , V C , and V E in Fig. 1 c,d), and a binary pattern can be generated from the output of the latch circuits ( V B and V D ). This binary pattern can then be used to control transistor gates in the switching circuits. Here, a combination of a latch circuit and a spiking neuron circuit are referred to as a waiting time generator (dashed square in Fig. 1 c). It should be noted the spiking neuron circuit in the binary pattern generator is directly connected to the set-reset latch circuit, and therefore, should be compatible with the CMOS logic circuits. Specifically, it should operate under the same 1 V power supply as the logic circuits, and it should output spikes with sufficiently short rise times and fall times on the scale of nano seconds. All these requirements are satisfied by a special design of CMOS spiking neuron circuits as shown later. Application to voltage converter circuit In order to visualize how the binary pattern generator works in the practical system, a simulation of the switching circuit for DC-DC voltage conversion is shown in Fig. 2 . The DC-DC voltage conversion is an essential component for power supply circuits, especially for an energy harvesting IoT device 25 , where the time-varying generated power is buck-boost converted via an inductor and stored in a capacitor at a certain voltage level (Fig. 2 a). The circuit operation consists of two periods: the first period in which the switches of S1 and S3 are turned on (red arrow), and the second period in which only the switch S2 is turned on (green arrow). In the first period, the charge stored in the primary capacitor ( C 1 ) flows to ground through S1 and S3, and the electrostatic energy of C 1 is converted to magnetic flux energy of the inductor. In the second period, current flows from the ground to the secondary capacitor ( C 2 ) via S2, and the magnetic flux energy is converted into electrostatic energy of C 2 . Figure 2 Application example of the binary pattern generator. ( a ) A block diagram of the DC-DC voltage converter circuit that consists of an inductor and several switches. ( b ) A circuit diagram of the simulated DC-DC voltage converter circuit, where the switching operation is controlled by the binary pattern generator. The binary pattern generator consists of a spiking neuron circuit defining the switching period, and two waiting time generators for switching sequences. The topology of the DC-DC voltage converter circuit is a typical buck-boost converter with a 200 mH inductor. ( c ) The voltage or current wave forms of the binary pattern generator and the DC-DC voltage converter circuit. ( d ) The wave forms of the binary pattern generator and the voltages around the S 1 switch. The time span corresponds to the blue shadowed region in (c). ( e ) The simulated output power and the control power as a function of the input power when V C1 is around 3 V and V C2 is around 5 V. The input power is varied by changing the period of V SPK1 from 3 s to 50 μs. The control power is negligible, several orders smaller than the input or output power. ( f ) The simulated power conversion efficiency (the ratio of the output power to the input power) as a function of the output power, which remains a relatively high value down to an extremely low output power level. ( g ) The classification of the control power (2.23 nW) when the input power is 84 μW, which corresponds to a point with a solid circle in ( e ). The power consumption of the binary pattern generator (“Neurons”) is sufficiently small, even smaller than the logic circuit. To control this switching operation, the binary voltages from the binary pattern generator were applied to the gate electrodes of the metal–oxide–semiconductor field effect transistors (MOSFETs) as shown in Fig. 2 b. A binary pattern ( V BIT2 ) was generated from the spike signals V SPK1 and V SPK2 as shown in Fig. 2 c, and was used to control the gates of S1, S2 and S3. First, the V SPK1 is generated at a certain period, for example 300 μs, by the first-stage spiking neuron circuit, and the V SPK2 is generated 30 μs after V SPK1 by a waiting time generator. Then, the created binary pattern V BIT2 was inverted by an inverter and used for gate control of S2. At the same time, the voltage level of V BIT2 was raised from 1 to 2 V by a level shifter and used for gate control of S3. To control the gate voltage ( V S1 ) of the S1, which is connected to the high voltage side in the circuit, a binary pattern ( V BIT3 ) was generated from V SPK2 and an additional spike signal ( V SPK3 ). When V BIT3 is turned on, V OFF is pulled down, V S1 is pulled up, and S1 is turned off (Fig. 2 d). On the other hand, when V BIT2 is turned on, V S1 is pulled down and S1 is turned on. In this way, the voltage level of the binary pattern is properly converted and used for gate driving, which makes the switching circuit extremely simple and low power. When the input power of this circuit was varied from 9.28 nW to 1.33 mW by changing the period of V SPK1 from 3 s to 50 μs, the output power varied almost proportionally from 8.36 nW to 1.16 mW (Fig. 2 e), and overall, the efficiency was around 90% (Fig. 2 f). The comparison with the previous studies in Supplementary Note 1 shows the converter with a binary pattern generator maintains a higher efficiency than the previous ones without degradation down to 8.36 nW output power. The high efficiency at low output power is mainly because the power consumption of the control circuit was 2–4 orders of magnitude smaller than the output power throughout the range, for example, 17.5 pW for the output power of 8.36 nW, and 114 nW for the output power of 1.16 mW (Fig. 2 e). In actual IoT devices, various control circuits need to be implemented, but as long as they can be controlled by the binary pattern generator, there may be no need to worry about the overhead of the control power even for an extremely low output power. In order to clarify the cause of the low control power consumption, we examined the origins of the total control power of 2.23 nW when the input power was 84.4 μW (a point with a solid circle in Fig. 2 e). The results in Fig. 2 g show that the binary pattern generator operates at even lower power than the logic circuit, and this allows the control circuit as a whole to achieve very low power consumption. It should be noted the control power corresponds to the whole switching circuits but does not contains the power for the voltage detection on either side of the input or the output. This result suggests the possibility of the next-generation ultra-low power electronics, in which various functions can be implemented even with a very limited power of less than 1 μW if the control is based on a binary pattern generator. Simulation of spiking neuron circuit The most essential building blocks for the binary pattern generator are the waiting time generators which consist of spiking neuron circuits. So far, various neuron circuits have been created with different purposes in reference to biological neurons. In biological neurons, the membrane potential increases with each input, and when the threshold potential is reached, the influx of Na ions and the efflux of K ions alternate 26 , generating a spiking potential (Fig. 3 a). This is accurately described by the Hodgkin–Huxley equation 27 , but a more simplified model is used in the design of neuron circuits 28 . In previous studies, relatively accurate analogue neuron circuits have been fabricated that implement the exponential behavior of ion channels in the subthreshold region of the transistor 29 , 30 . On the other hand, a further simplification of functions has been carried out in order to achieve large scale systems that includes a number of neuron circuits. Such simplified models include a integrate-fire neuron which is the simplest version 31 , a leaky-integrate-fire neuron which ignores variation of neuron dynamics 32 , an Izhikevich neuron which treats the firing process algorithmically 33 , and a phase transition neuron which implements functions with material properties 34 , 35 . In these analogue neuron circuits, there is a trade-off between accuracy and simplicity, but by focusing on the mathematical structure of the nonlinear neuron dynamics, a circuit that balances both has been devised 36 , 37 . Recently, circuits have also been proposed that reduce energy consumption to the utmost limit by lowering the supply voltage to a few hundred mV or using an extremely small tunneling current 23 , 24 , 38 . These neuron circuits have been used for machine learning applications 3 , and also for optic, auditory, or other sensory signal processing 39 – 41 . Figure 3 CMOS spiking neuron circuit. ( a ) A schematic illustration of a biological neuron. ( b ) A CMOS spiking neuron circuit with the integrate-fire function, which is used for the waiting time generator. ( c ) A schematic illustration of V IN , V 1 , and V OUT in ( b ) as a function of time. A waiting time corresponds to the period between the onset of V IN and the output of V OUT . The details of all the spiking neuron circuits with different waiting times are presented in Supplementary Note 3 . In this study, we fabricated a spiking neuron circuit, which is specially designed for waiting time generation in ultra-low power consumption. For the purpose of waiting time generation, the spiking neuron circuit implements the integrate-fire function while all the other biological functions were excluded intentionally. It is also designed to generate a nanosecond-width square pulse wave as the output spike for seamless connection with CMOS logic circuits with a common 1 V supply. The fabricated spiking neuron circuit consists of two parts: one part generates waiting time, and the other part generates a spike (Fig. 3 b). In the former part, the input current is created by the ON current or subthreshold current of the transistor under the application of 1 V, and charges the capacitor with the approximately constant current. Then, after a waiting time that is determined by the ratio of the capacitance to the current, the capacitor potential ( V 1 ) reaches the threshold voltage of the inverter (around 0.5 V) and activates the spike generation part as shown in Fig. 3 c. The capacitance was designed to be as small as possible to suppress power consumption, and metal–oxide–semiconductor (MOS) capacitors of several tens of fF or even smaller capacitance which is parasitic to transistors and wiring were used. The spike generation part consists of CMOS circuit. When the V 1 reaches a threshold, positive feedback and delayed feedback are activated in turn to generate a spike output. Generally speaking, to reduce the power consumption of integrate–fire neuron circuits, methods such as lowering the supply voltage or using capacitive feedback have been used 23 , 24 , 42 , but here, in order to use a 1 V supply and reduce the use of capacitors as much as possible, we used only CMOS circuits. To elucidate a detailed operation of the spiking neuron circuit, simulations are shown for the spiking neuron circuit that generates a waiting time of approximately 100 ms as shown in Fig. 4 a. Transistor with 5 V withstand voltage in the TSMC 0.18 μm process was used (see Supplementary Notes 2 and 3 for more details) because the 5 V transistor under a 1 V supply leads to near-threshold computing and dramatically reduce power consumption. As shown in Fig. 4 b, when 1 V is applied to V in (input voltage in gray color), the V 1 gradually increases (red color), and the spike voltage is output as V out (blue color). Here, after the V 1 reaching the threshold potential, the positive feedback raises V 1 close to 1 V (red color), which contributes to the steep rise of the V out . Then, by resetting V 1 to 0 V with delayed feedback, the V out falls steeply and the spike waveform is completed. An enlarged view of the firing process (Fig. 4 c) shows that the rise or the fall of the waveform becomes steeper with each successive inverter, in the order of V 1 , V 2 , V 3 , V 4 , and V out . Thus, by connecting CMOS inverters in multiple stages, a steep waveform can be obtained at the output side in a digitally compatible level, no matter how long the waiting time is at the input side. The energy consumption for a series of operations is sufficiently small, only 0.16 pJ per spike operation (Fig. 4 d). This corresponds to an average power consumption of 1.7 pW, which is more than one order of magnitude lower than previous experimental demonstrations 20 – 24 . Here, to suppress energy consumption, diodes were inserted at the top and bottom of the first stage inverter (Fig. 4 a), otherwise, the through current flows for a long time as the V 1 approaches the inverter threshold and increases energy consumption. Simulation of a spiking neuron circuit with a shorter waiting time (approximately 1 μs) is also presented in Supplementary Note 4 and the simulations for all the other spiking neuron circuits are also summarized in Supplementary Note 5 . Figure 4 Simulation of the spiking neuron circuit. ( a ) The simulated spiking neuron circuit for the 100 ms waiting time. ( b ) The voltage wave forms at several different nodes in ( a ). The red shadowed region indicates the waiting time of this neuron circuit. ( c ) The magnification of ( b ) in the vicinity of the spike generation event, which is indicated by the grey shadow. ( d ) The simulated energy consumption of the spiking neuron circuit, which rapidly increases as approaching the spiking event due to the increase in the through current at the first-stage inverter. The detailed operation of the 1 μs spiking neuron circuit is also presented in Supplementary Note 4 , and the simulations of all the spiking neuron circuits with different waiting times are presented in Supplementary Note 5 . Experiments on spiking neuron circuits Based on the simulation, the spiking neuron circuits were experimentally fabricated using the TSMC 180 nm BCD process as shown in Fig. 5 a,b. By controlling the input current of the spiking neuron circuit (Supplementary Note 3 ), different lengths of waiting time can be generated from an input voltage of 1 V, which were measured based on the experimental setting as shown in Fig. 5 c. In Fig. 5 d,e, we used the ON current of the PMOS transistor to generate a waiting time in the order of 100 ns. Here, since the ON current of the PMOS can be tuned down to 10 nA with the length and width of the channel, the waiting time as short as 100 ns can be generated with a parasitic capacitance of a few fF. In Fig. 5 f,g, instead, the sub-pA OFF current of the lower-threshold PMOS transistor with the 2 V withstand voltage was used to generate a waiting time in the order of 100 ms. In this case, an additional MOS capacitor of 33 fF was also utilized to elongate the waiting time. In this way, we succeeded in experimentally generating an arbitrary waiting time spanning six orders of magnitude from 100 ns to 100 ms on the chip. The circuits and the device parameters for all the other waiting times are also summarized in Supplementary Note 3 . Figure 5 Experiments of the spiking neuron circuit. ( a ) An optical micrograph of the fabricated chip for spiking neuron circuits with various waiting times between 100 ns and 100 ms. ( b ) The magnification of the chip region for the spiking neuron circuit with the 100 ms waiting time. The red dashed square indicates the whole spiking neuron circuit while the red solid square corresponds to the spike generation part and the membrane capacitor. ( c ) The experimental setting for the measurement of the spiking neuron circuits. V IN and V OUT are the voltages inside the I/O buffers, and V IN-EX and V OUT-EX are the voltages outside the I/O buffers, which were measured by the oscilloscope. ( d ) A circuit diagram of the fabricated spiking neuron circuit for the 100 ns waiting time (see Supplementary Note 3 for details). ( e ) Experimentally measured wave forms of V IN-EX and V OUT-EX for the 100 ns spiking neuron circuit in ( d ). The interval between spikes are approximately 100 ns. ( f ) A circuit diagram of the fabricated spiking neuron circuit for the 100 ms waiting time (see Supplementary Note 3 for details). ( g ) Experimentally measured wave forms of V IN-EX and V OUT-EX for the 100 ms spiking neuron circuit in ( f ). The interval between spikes are approximately 100 ms. The experiments of all the spiking neuron circuits with different waiting times are presented in Supplementary Note 5 . The waiting time in the simulation and the one obtained in the experiment were in general agreement throughout the six-order range of the waiting time (Fig. 6 a and Supplementary Note 5 ). As shown in Fig. 6 b, the output spike width was approximately 40 ns for the waiting time up to 100 μs, and approximately 400 ns for the longer waiting times due to the insertion of diodes at the first stage inverter as mentioned previously in Fig. 4 a. This spike width can be converted to approximately 40 ns by using a spike width conversion circuit as shown in Supplementary Note 6 . The steep rise and fall of the output spike guarantee the spike is compatible with CMOS logic circuits. The experimentally measured energy consumption per spike operation was found to be between 60 and 120 fJ for all the waiting times (Fig. 6 c). It is interesting to note that the past examples of a neuron circuit with a long spike interval in the order of 100 ms is limited, and our 100 ms neuron circuit has the lowest power consumption of 1.2 pW among all the experimental demonstrations (Table 1 ). A comparison in Table 1 clearly shows the fabricated spiking neuron circuits have a unique feature of digital-circuit compatibility in the sense of the spike width and the supply voltage, and at the same time, a reasonably low energy consumption and a wide control range of the waiting time. It should be emphasized that all of these features are optimized for the waiting time generation inside a digital circuit and asynchronously controlling the switching circuits, rather than for the implementation of the biological functions or for the simple reduction of energy consumption in a single neuron circuit as in the case of previous studies 21 , 23 , 24 . Figure 6 Comparison between experiments and simulations for spiking neuron circuits. ( a ) The experimentally measured waiting times of the fabricated spiking neuron circuits are plotted as a function of the simulating waiting times (post-layout simulation). The plot indicates the experimental results are approximately consistent with the simulation. ( b ) Experimentally obtained spike width and the simulated spike width as a function of the waiting time. ( c ) Experimentally obtained energy consumption per spike and the simulated energy consumption per spike as a function of the waiting time. In the experiment, the energy consumption was obtained by measuring I SOURCE in Fig. 5 c with Keithley 6430. Table 1 Benchmark of neuron circuits. References Voltage (V) Spike width (µs) Digital circuit compatibility Frequency (Hz) Energy per spike (pJ) Power (pW) Process (nm) Area (µm 2 ) Function 20 2.25 100 No 100 17 1700 350 1187 Spike frequency adaptation, positive feedback, refractory period 21 0.6 2600 No 100 0.4 40 90 442 Integrate-fire 22 1.8 – Yes 8 2.8 22.4 180 116 Adaptive exponential I&F 23 0.2 17 No 26,000 0.004 105 65 35 Morris-Lecar 24 0.2 5000 No 16,000 0.002 30 65 31 Axon-Hillock This work 1.0 0.4 Yes 6.6 0.18 1.2 180 504 (906) Integrate-fire This work 1.0 0.037 Yes 7,800,000 0.079 6,160,000 180 234 (401) Integrate-fire The characteristics of the spiking neuron circuit in this work are compared with the state-of-the-art neuron circuits that were experimentally fabricated in the past literature. The area only includes the membrane capacitor and the subsequent circuits while the values in the brackets include the current generation MOSFETs at the input of the spiking neuron circuits."
} | 7,908 |
34443716 | PMC8400133 | pmc | 2,287 | {
"abstract": "The silver particles were grown in situ on the surface of wood by the silver mirror method and modified with stearic acid to acquire a surface with superhydrophobic and antibacterial properties. Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), and X-ray energy spectroscopy (XPS) were used to analyze the reaction mechanism of the modification process. Scanning electron microscopy (SEM) and contact angle tests were used to characterize the wettability and surface morphology. A coating with a micro rough structure was successfully constructed by the modification of stearic acid, which imparted superhydrophobicity and antibacterial activity to poplar wood. The stability tests were performed to discuss the stability of its hydrophobic performance. The results showed that it has good mechanical properties, acid and alkali resistance, and UV stability. The durability tests demonstrated that the coating has the function of water resistance and fouling resistance and can maintain the stability of its hydrophobic properties under different temperatures of heat treatment.",
"conclusion": "4. Conclusions In this experiment, a convenient and effective method for preparing superhydrophobic and antimicrobial wood was proposed by the organic–inorganic mixing method. The Ag particles were prepared on a wood surface with the disposal of a silver ammonia complex solution and then reduced with formaldehyde. The Ag particles were coated with low surface energy stearic acid and further modified to obtain a superhydrophobic surface with a WCA of 158.7°. The formation mechanism and antimicrobial mechanism of the rough structure were analyzed. The surface morphology was characterized, and the stability test was carried out. The mechanical wear resistance, UV resistance, and acid and alkali resistance of wood@Ag@SA were tested. The results showed that it could maintain the superhydrophobic property in a harsh environment. The antifouling and water absorption tests showed that the modified wood was waterproof and fouling resistant. The poplar wood coating showed excellent properties in hydrophobicity, antimicrobial property, and durability, making it suitable for self-cleaning and biomedical applications.",
"introduction": "1. Introduction As a natural material, wood is one of the most important materials for structural construction, furniture, energy, and aerospace industries [ 1 ]. However, the negative characteristics of wood, such as water absorption, UV degradation, and poor resistance to microbial action, have limited the application of wood [ 2 , 3 ]. Therefore, special treatment of the wood surface to obtain hydrophobicity is necessary to prolong the service life of wood and give greater application to the value of wood [ 4 , 5 , 6 , 7 , 8 , 9 ]. Because superhydrophobic materials have special wettability [ 10 ], it is widely used in the fields of self-cleaning [ 11 , 12 ], antifouling [ 13 ], antisepsis [ 14 , 15 ], and oil–water separation [ 16 , 17 ]. The methods of preparing superhydrophobic wood include the etching method [ 18 ], the vapor deposition method [ 19 ], and the layer-by-layer self-assembly method [ 20 ]. When water droplets contact the surface, an air pad will be formed in the groove of the micro-nano rough structure on the superhydrophobic surface. Thus, a solid–liquid-gas three-phase interface is formed, and the contact surface between water droplets and the surface is increased to prevent water droplets from infiltrating the surface [ 21 , 22 , 23 ]. Superhydrophobic surfaces are usually obtained in two steps: the first is to use nanoparticles such as ZnO [ 24 ], SiO 2 [ 19 ], TiO 2 [ 25 ], and CaCO 3 to generate rough structures on the surface of the medium [ 26 , 27 ]; the second is to modify a rough surface with a low-surface energy substance [ 28 , 29 ]. At present, it is more common to attach nanoparticles to the wood surface by polymer and treat them with the substrate of low surface energy; however, due to the low adhesion between nanoparticles and wood, superhydrophobic wood often has poor mechanical properties. To improve the bonding force between interfaces and enhance the mechanical wear resistance of the wood surface, nanoparticles can be generated in situ on the wood surface through chemical reactions [ 13 , 30 ]. At the same time, considering that wood contains cellulose, hemicellulose, lignin, starch, sugar, and other components, it is a hotbed for microbial growth [ 31 , 32 ]. When the temperature and humidity and other environmental conditions are appropriate, these microorganisms will attach to the wood for a large number of growth and reproduction, resulting in wood decay, discoloration, and moldy, which influence the strength, toughness, and permeability of wood [ 6 , 33 , 34 , 35 ]. It also poses a threat to the life and health of the users [ 36 , 37 , 38 ]. Therefore, antibacterial treatment of wood is needed to maintain the properties of wood and prevent the spread of disease [ 39 ]. At present, the preparation of antimicrobial wood methods includes siloxane antifouling coating, copolymer, polymer, and other methods to inhibit the growth of bacteria; however, these polymer-based materials are often easily removed from the wood, showing poor mechanical properties. Therefore, it is necessary to develop a wood coating with high mechanical durability and antimicrobial properties [ 40 ]. The organic–inorganic mixed coating prepared by the sol-gel method can substitute for polymer-based antimicrobial coating, which is usually prepared from metallic alkanes or functional alkanes as precursors. It has environmental benefits and ensures strong mechanical properties and antifouling performance [ 41 , 42 ]. Silver nanoparticles have a strong antibacterial effect, which is since silver ions can inhibit the synthesis of enzymes on the extra-mercaptan groups of proteins on the microbial membrane, and eventually lead to the death of cells [ 43 ]. At the same time, silver nanoparticles have a smaller particle size and larger specific surface area, so they can dissolve in solution and release silver ions, thus achieving stronger antibacterial properties [ 44 ]. In general, the physical and chemical properties of silver nanoparticles, such as size, shape, or surface characteristics, can be controlled by chemical substances such as polymer, metal–ion interaction, or various reducing agents to construct functional surfaces [ 45 ]. Duan [ 46 ] prepared the wood with superhydrophobic antibacterial properties by self-polymerization of dopamine and hydrophobic modification of copper nanoparticles on the surface of the wood with fluoro silane. The coating showed good resistance to acid or alkali corrosion and good mechanical properties. Gao [ 47 ] treated wood with sodium hydroxide and silver nitrate, then reduced nano-silver ions on the wood surface with glucose and modified with fluoro silane to prepare a kind of superhydrophobic and oil-phobic wood with electrical conductivity. These methods of preparation of superhydrophobic wood used the chemical reaction directly on the wood surface to grow nanoparticles, guarantee the superhydrophobic coating and the cohesive strength between wood, thus showed excellent mechanical properties. This method will be beneficial to expand the application forms of superhydrophobic wood, and has a high application value in the fields of self-cleaning, biomedicine, and electronic information. Here, we successfully deposited silver particles on the wood surface through the reaction of silver ammonia solution and sodium hydroxide with a simple silver mirror reaction [ 48 ]. First, the poplar wood was pretreated with NaOH solution to make the surface of it negatively charged. It absorbed Ag(NH 3 ) 2 + after being impregnated with a silver ammonia complex solution. Ag(NH 3 ) 2 + was reduced to Ag particles in situ on the wood surface by formaldehyde reduction. After that, the treated wood was modified by stearic acid with a long-chain alkyl group [ 49 ] to obtain a superhydrophobic surface with a contact angle of 158.7°. The modified surface has waterproof, fouling, and antibacterial properties, and stable superhydrophobic properties under mechanical wear cycle, acid and alkali solution, ultraviolet light, and different temperatures of heat treatment."
} | 2,081 |
16522219 | PMC1431723 | pmc | 2,288 | {
"abstract": "Horizontal gene transfer (HGT) is more important than gene duplication in bacterial evolution, as has recently been illustrated by work demonstrating the role of HGT in the emergence of bacterial metabolic networks."
} | 53 |
18846089 | null | s2 | 2,289 | {
"abstract": "Genomic data allow the large-scale manual or semi-automated assembly of metabolic network reconstructions, which provide highly curated organism-specific knowledge bases. Although several genome-scale network reconstructions describe Saccharomyces cerevisiae metabolism, they differ in scope and content, and use different terminologies to describe the same chemical entities. This makes comparisons between them difficult and underscores the desirability of a consolidated metabolic network that collects and formalizes the 'community knowledge' of yeast metabolism. We describe how we have produced a consensus metabolic network reconstruction for S. cerevisiae. In drafting it, we placed special emphasis on referencing molecules to persistent databases or using database-independent forms, such as SMILES or InChI strings, as this permits their chemical structure to be represented unambiguously and in a manner that permits automated reasoning. The reconstruction is readily available via a publicly accessible database and in the Systems Biology Markup Language (http://www.comp-sys-bio.org/yeastnet). It can be maintained as a resource that serves as a common denominator for studying the systems biology of yeast. Similar strategies should benefit communities studying genome-scale metabolic networks of other organisms."
} | 332 |
38402147 | PMC10893638 | pmc | 2,290 | {
"abstract": "Background 1,2-propanediol (1,2-PDO) is widely used in the cosmetic, food, and drug industries with a worldwide consumption of over 1.5 million metric tons per year. Although efforts have been made to engineer microbial hosts such as Corynebacterium glutamicum to produce 1,2-PDO from renewable resources, the performance of such strains is still improvable to be competitive with existing petrochemical production routes. Results In this study, we enabled 1,2-PDO production in the genome-reduced strain C. glutamicum PC2 by introducing previously described modifications. The resulting strain showed reduced product formation but secreted 50 ± 1 mM d -lactate as byproduct. C. glutamicum PC2 lacks the d -lactate dehydrogenase which pointed to a yet unknown pathway relevant for 1,2-PDO production. Further analysis indicated that in C. glutamicum methylglyoxal, the precursor for 1,2-PDO synthesis, is detoxified with the antioxidant native mycothiol (MSH) by a glyoxalase-like system to lactoylmycothiol and converted to d -lactate which is rerouted into the central carbon metabolism at the level of pyruvate. Metabolomics of cell extracts of the empty vector-carrying wildtype, a 1,2-PDO producer and its derivative with inactive d -lactate dehydrogenase identified major mass peaks characteristic for lactoylmycothiol and its precursors MSH and glucosaminyl-myo-inositol, whereas the respective mass peaks were absent in a production strain with inactivated MSH synthesis. Deletion of mshA , encoding MSH synthase, in the 1,2-PDO producing strain C. glutamicum Δ hdpA Δ ldh (pEKEx3- mgsA - yqhD - gldA ) improved the product yield by 56% to 0.53 ± 0.01 mM 1,2−PDO mM glucose −1 which is the highest value for C. glutamicum reported so far. Conclusions Genome reduced-strains are a useful basis to unravel metabolic constraints for strain engineering and disclosed in this study the pathway to detoxify methylglyoxal which represents a precursor for 1,2-PDO production. Subsequent inactivation of the competing pathway significantly improved the 1,2-PDO yield. Supplementary Information The online version contains supplementary material available at 10.1186/s12934-024-02337-w.",
"conclusion": "Conclusion Libraries with genome-reduced strains provide a valuable tool to identify novel targets for metabolic engineering. In this study, the genetic background of the genome-reduced C. glutamicum PC2 strain was the basis to disclose the pathway for detoxification of methylglyoxal with MSH to d -lactate and its relevance for 1,2-PDO production. Subsequent inactivation of this competing pathway significantly improved 1,2-PDO production in C. glutamicum . Regarding the detoxification of methylglyoxal, the low molecular weight thiol MSH in C. glutamicum , seems to play a role comparable to GSH in other organism, such as E. coli . However, the enzymatic machinery catalyzing the detoxification process with MSH has to be identified in future studies.",
"introduction": "Introduction The C3-diols 1,2- and 1,3-propanediol (PDO) are important building blocks and are widely used in the polymer, food, cosmetic and drug industry [ 1 , 2 ]. For 1,3-PDO and 1,2-PDO a global market size of around 1.4 billion and 0.4 billion US dollars is expected in the next years [ 2 , 3 ]. Both diols are mainly produced from fossil fuels, but bio-based production processes, utilizing renewable resources, are favorable to tackle the concerns of climate change and the limited availability of fossil resources. Notably, first commercial microbial processes for 1,3-PDO production are readily available [ 4 ]. Although several natural 1,2-PDO producers are known and established microbial systems have been extensively engineered for its production, a sustainable process at the industrial level is still missing [ 2 ]. Corynebacterium glutamicum is a facultative anaerobic Gram-positive soil bacterium which is generally recognized as safe (GRAS), robust, and grows with several sugars, organic acids and phenolic compounds as single or combined carbon and energy sources [ 5 – 10 ]. Ample knowledge about the physiology, metabolic and regulatory networks has been gathered and a versatile toolbox for genetic engineering is available [ 6 , 11 – 14 ]. This bacterium is known as industrial powerhouse for the production of amino acids such as l -glutamate and l -lysine at a scale of 6 million tons per year [ 5 ]. Moreover, sophisticated metabolic engineering approaches extended the product portfolio rapidly [ 15 ], also to diols such as 2,3-butanediol [ 16 – 19 ], 1,3-PDO [ 20 ] and 1,2-PDO [ 21 ]. Production of 1,2-PDO was initially achieved by heterologous expression of the Escherichia coli gene mgsA (encoding methylglyoxal synthase) and overexpression of a putative aldo-keto reductase (probably functioning as a methylglyoxal reductase) (Fig. 1 ; [ 21 ]). Biosynthesis was further improved by additional heterologous expression of the E. coli genes gldA (encoding glycerol dehydrogenase) and yqhD (encoding aldehyde reductase). Deletions of hdpA (encoding dihydroxyacetone phosphate phosphatase) and ldh (encoding l -lactate dehydrogenase) avoided the secretion of the side-products lactate and glycerol and improved 1,2-PDO production further (Fig. 1 ; [ 22 ]). Methylglyoxal is a key intermediate in the 1,2-PDO production route, but also represents a potent cytotoxic compound which reacts with arginine, lysine, and cysteine residues and might lead to protein inactivation [ 23 , 24 ]. Consequently, methylglyoxal is in many organisms detoxified to d -lactate by a GloAB-mediated glyoxalase system with the tripeptide glutathione (GSH). d -lactate is further oxidized to pyruvate which is subsequently channeled back into the central carbon metabolism (Fig. 1 ) [ 23 , 24 ]. A characteristic of Actinobacteria , such as C. glutamicum , is the absence of GSH biosynthesis. However, in these organisms the low molecular weight thiol and non-enzymatic antioxidant mycothiol (MSH) is synthesized for detoxification instead [ 25 ]. MSH is synthezised from the intermediates myo-inositol and N -acetyl-glucosamine via glucosaminyl-myo-inositol by five enzymatic steps encoded by the genes mshA , mshA2 , mshB , mshC and mshD [ 26 ]. However, no GloAB homologs have been identified in C. glutamicum and the mechanism for detoxification of methylglyoxal is elusive so far. Besides MSH as antioxidant, the transcriptional regulator OxyR plays an important role in the oxidative stress response in C. glutamicum . Under unstressed conditions this master regulator represses genes encoding enzymes such as catalase. As a result, an oxyR deletion mutant shows increased resistance to H 2 O 2 [ 27 ]. The concept of genome reduction to improve physiological characteristics and create optimized hosts for the industrial environment has been applied to many organisms [ 28 ] as well as to C. glutamicum [ 29 , 30 ]. In a systematic top-down approach all genes of the C. glutamicum genome were ranked by their relevance for growth in minimal medium with glucose as sole carbon and energy source. The proposed gene clusters were individually deleted and the genome-reduced strains (GRS) were screened for their growth phenotype [ 29 ]. In a follow-up study the deletions of the most promising genomic regions were combined in a stepwise manner, leading to the pre-chassis strains PC1 and PC2, with 8.5% and 12.6% reduced genomes, respectively. While both of these strains grew like the wildtype, the finally engineered chassis strain C1 (13.4% genome-reduction) showed a growth deficit on acetate due to an unwanted mutation in the promoter of the ramA gene, which was repaired yielding the final genome-reduced derivative C1* [ 30 ]. Recently, some intermediate strains of the aforementioned genome-reduction approach were utilized to screen for improved heterologous cutinase secretion in C. glutamicum [ 31 ]. In this study, we introduced known genetic modifications to enable 1,2-PDO production [ 22 ] into the genome-reduced strain C. glutamicum PC2 [ 30 ] and its parental strain GRS [ 29 ] (in later studies also called CR099). Only in the PC2 genetic background, we observed the formation of d -lactate as byproduct, which was attributed to the lack of the respective d -lactate dehydrogenase and pointed to a yet unknown pathway relevant for 1,2-PDO production. Further analysis indicated that in C. glutamicum methylglyoxal, the precursor for 1,2-PDO synthesis, is converted with MSH by a glyoxalase-like system to lactoylmycothiol and further to d -lactate which is channeled back into the central carbon metabolism on the level of pyruvate (Fig. 1 ). \n Fig. 1 Overview of the 1,2-PDO pathway introduced into C. glutamicum [ 22 ], including the proposed bypass from methylglyoxal to d -lactate and pyruvate via lactoylmycothiol in blue. Black arrows represent native pathways; dotted arrows indicate more than one reaction; green and disrupted arrows represent heterologously expressed proteins and deletions of gene sequences of mentioned proteins, respectively. Abbreviations: TCA, tricarboxylic acid cycle; DHA, dihydroxyacetone; DHAP, dihydroxyacetone phosphate; GAP, glyceraldehyde 3-phosphate; UDP-GlcNAc, uridine diphosphate-N-acetylglucosamine; 1- l -Ins-1-P, 1- l - myo -inositol 1-phosphate; GldA, glycerol dehydrogenase (from E. coli ); HdpA, dihydroxyacetone phosphate phosphatase; MgsA, methylglyoxal synthase (from E. coli ); Dld, quinone-dependent d -lactate dehydrogenase; LldD, quinone-dependent l -lactate dehydrogenase; Ldh, NAD-dependent l -lactate dehydrogenase; YqhD, aldehyde reductase (from E. coli ); MshA, glycosyltransferase; “GloA”, proposed lactoylglutathione lyase homolog in C. glutamicum ; “GloB”, proposed hydroxyacylglutathione hydrolase homolog in C. glutamicum",
"discussion": "Discussion Genome-reduced strains have been engineered to improve properties of microbes for industrial application, to study the physiological role of gene sets and to indentify essential genes to eventually design minimal genomes with streamlined functions [ 28 , 36 ]. This concept has also been applied to the industrially relevant bacterium C. glutamicum which yielded a strain genealogy varying in the degree of genome reduction [ 30 ]. This genome-reduced strain library was recently utilized to screen for strains with improved heterologous secretion of cutinase [ 31 ]. In this study, we harnessed the genome-reduced derivative C. glutamicum PC2 [ 30 ] and its parental strain GRS [ 29 ] to unravel metabolic bottlenecks for the production of the industrial relevant bulk chemical 1,2-PDO. The reconstruction of the previously described genetic modifications [ 22 ] to enable 1,2-PDO production in C. glutamicum PC2 led to reduced product formation accompanied by the accumulation of lactate. In C. glutamicum three enzymes are known, which are responsible for synthesis and utilization of lactate. The NADH-dependent l -lactate dehydrogenase LdhA, encoded by ldh , which is mainly responsible for l -lactate formation under excess of NADH [ 22 , 37 ]. Further, C. glutamicum harbors quinone-dependent l - and d -lactate dehydrogenases (LldD and Dld), which are specific for the respective enantiomer and essential for its utilization [ 32 , 38 ]. C. glutamicum PC2Δ hdpA Δ ldh (pEKEx3- mgsA - yqhD - gldA ) lacks LdhA and due to the genome reduction Dld, as well. The applied enzyme assay identified accumulating lactate to be the d -enantiomer. These results suggested the presence of a metabolic route relevant for 1,2-PDO synthesis and with d -lactate as intermediate which was interrupted by the inactive Dld in the PC2 background. Further, no other genes encoding proteins obviously related to the described production pathway could be identified in the PC2 background (Table S2 ). However, the PC2-derivative showed higher accumulation of lactate and less propanediol compared to the wildtype background harboring the same genetic modification in addition to dld deletion (50 ± 1 mM vs. 10 ± 2 mM lactate and 24 ± 1 mM vs. 20 ± 7 mM propanediol; Figs. 2 B and 3 C and D). Methylglyoxal is a precursor for 1,2-PDO synthesis but it also represents a cytotoxic intermediate leading to protein inactivation and oxidative stress [ 23 , 24 ]. In E. coli and other organisms methylglyoxal is detoxified with GSH, forming the intermediate S-lactoylglutathione, which is further metabolized via the glyoxalase system GloAB to d -lactate [ 23 , 24 , 39 ]. In actinobacteria , such as C. glutamicum , the low molecular weight thiol MSH is present instead of GSH [ 25 ], however, a glyoxalase-like system based on MSH has not been identified so far. Recently, in the actinobacterium Streptomyces coelicolor a Ni 2+ -activated and MSH-dependent glyoxalase I enzyme was described [ 40 ]. However, BLASTx analysis [ 41 ] did not identify a homolog in C. glutamicum (data not shown). Also the deletion of the annotated gloA and gloB homologs cg1073 as well as cg0071, cg1482 and cg1856 in C. glutamicum Δ hdpA Δ ldh (pEKEx3- mgsA - yqhD - gldA ) did not influence growth, glucose consumption and product formation. Therefore, the identification of the glyoxylase system remains elusive and has to be elucidated in future studies. It was shown that overexpression of mshA , encoding a glycosyltransferase catalyzing the first step of MSH synthesis, and deletion of the gene mstX , encoding a MSH transferase, affected the robustness of C. glutamicum towards methylglyoxal [ 33 , 42 ]. Taken together, these results indicate that, in analogy to GSH-dependent systems, methylglyoxal in C. glutamicum is detoxified with MSH via S-lactoylmycothiol by an unknown glyoxalase-like system yielding d -lactate. Notably, LC-MS-Qtof-based analysis identified major mass peaks characteristic for lactoylmycothiol, MSH and glucosaminyl-myo-inositol in cell extracts of the wildtype, C. glutamicum Δ hdpA Δ ldh (pEKEx3- mgsA - yqhD - gldA ) and its derivative with inactivated d -lactate dehydrogenase. For lactoylmycothiol, the exact mass and isotope pattern matches the expected sum formula C 20 H 34 N 2 O 14 S, and, although alternative structures cannot be excluded, the observed fragmentation pattern is in accordance with the proposed lactoylmycothiol structure. Moreover, in C. glutamicum Δ hdpA Δ ldh Δ mshA (pEKEx3- mgsA - yqhD - gldA ), with inactive MSH synthase, lactoylmycothiol and both precursors were not detectable. These findings support the presence of the MSH-dependent detoxification system converting methylglyoxal with MSH to d -lactate in C. glutamicum . The secretion of up to 24 ± 1 mM of d -lactate by C. glutamicum Δ hdpA Δ ldh Δ dld (pEKEx3- mgsA - yqhD - gldA ) indicated an increased rerouting of methylglyoxal back into the central carbon metabolism in the parental strain with active Dld. Consequently, to improve the precursor availability, we inactivated MshA which significantly improved 1,2-PDO production. C. glutamicum Δ hdpA Δ ldh Δ mshA (pEKEx3- mgsA - yqhD - gldA ) showed a Y P/S of 0.53 ± 0.01 mM 1,2−PDO mM glucose −1 which is 56% higher compared to C. glutamicum Δ hdpA Δ ldh (pEKEx3- mgsA - yqhD - gldA ) and represents the highest reported value for C. glutamicum . Also in E. coli inactivation of the methylglyoxal detoxification system by deletion of gloA proved to be beneficial for 1,2-PDO production from glucose [ 43 ]. It should be noted, that the strain lacking mshA showed reduced growth and glucose consumption (Figure S2 ) which might indicate an impaired stress tolerance. In the closely related Mycobacterium tuberculosis MSH is essential for survival whereas for M. smegmatis it is not [ 44 ]. For E. coli GSH is dispensable, however, GSH-deficiency results in an increased sensitivity towards oxidative stress [ 45 ]. Also in C. glutamicum MSH synthesis is not essential but its absence was shown to increase the sensitivity against oxidative stress and some toxic compounds indicated by reduced growth and cell viability [ 26 , 42 , 46 ]. Notably, a MSH-deficient C. glutamicum strain, lacking mshC , possessed impaired growth in bioreactor cultivations at a pO 2 of 30%, whereas at a pO 2 level of 20% this growth defect was abolished [ 46 ]. The relevance of a proper functioning oxidative stress response for 1,2-PDO production is also indicated by the result that overexpression of oxyR in C. glutamicum Δ hdpA Δ ldh (pEKEx3- mgsA - yqhD - gldA ) strongly diminished 1,2-PDO production. OxyR is the master regulator of the oxidative stress response in C. glutamicum [ 27 ] and under unstressed conditions this transcriptional regulator is acting as a repressor of its target genes (e.g. katA encoding catalase). Deletion of oxyR resulted in an increased resistance of the C. glutamicum mutant to hydrogen peroxide whereas its overexpression completely inhibited growth in the presence of H 2 O 2 [ 47 ]. Therefore, a carefully adjusted oxygen transfer into the culture broth might be crucial for scale-up of fermentation processes based on MSH-deficient 1,2-PDO production strains."
} | 4,298 |
26870079 | PMC4740370 | pmc | 2,292 | {
"abstract": "Application of hyperaccumulator-endophyte symbiotic systems is a potential approach to improve phytoremediation efficiency, since some beneficial endophytic bacteria are able to detoxify heavy metals, alter metal solubility in soil, and facilitate plant growth. The objective of this study was to isolate multi-metal resistant and plant beneficial endophytic bacteria and to evaluate their role in enhancing plant growth and metal accumulation/translocation. The metal resistant endophytic bacterial strain E6S was isolated from stems of the Zn/Cd hyperaccumulator plant Sedum plumbizincicola growing in metalliferous mine soils using Dworkin and Foster salts minimal agar medium with 1-aminocyclopropane-1-carboxylate (ACC) as the sole nitrogen source, and identified as homologous to Achromobacter piechaudii based on morphological and biochemical characteristics, partial 16S rDNA sequence and phylogenetic analysis. Strain E6S showed high level of resistance to various metals (Cd, Zn, and Pb). Besides utilizing ACC, strain E6S exhibited plant beneficial traits, such as solubilization of phosphate and production of indole-3-acetic acid. Inoculation with E6S significantly increased the bioavailability of Cd, Zn, and Pb in soil. In addition, bacterial cells bound considerable amounts of metal ions in the following order: Zn > Cd >Pb. Inoculation of E6S significantly stimulated plant biomass, uptake and bioaccumulation of Cd, Zn, and Pb. However, E6S greatly reduced the root to shoot translocation of Cd and Zn, indicating that bacterial inoculation assisted the host plant to uptake and store heavy metals in its root system. Inoculation with the endophytic bacterium E6S homologous to A. piechaudii can improve phytostabilization of metalliferous soils due to its effective ability to enhance in situ metal rhizoaccumulation in plants.",
"conclusion": "Conclusion Endophytic bacterium E6S homologous to A. piechaudii isolated from metal hyperaccumulator S. plumbizincicola stems was able to facilitate plant growth through the production of IAA and ACC deaminase and solubilization of P. The characterization studies showed that the isolate resisted high concentrations of Cd, Zn, and Pb, via extracellular biosorption in metal containing liquid media and increased water extractable metal concentration in metal contaminated soils. The inoculation of E6S significantly enhanced plant growth and rhizoaccumulation of Cd, Zn, and Pb by host S. plumbizincicola , however, the TFs of Cd and Zn remarkably declined in the presence of bacterial inoculation. Effective metal biosorption and mobilizing capacities as well as the potential to adapt to/survive multi-metal stress conditions along with various plant beneficial traits are clear indications of the advantages of employing this microorganism as a bioinoculant for ameliorating metal phytotoxicity and thus enhancing phytostabilization efficiency. Further work will address the mechanism of the selected bacterial strain contributing to reduce metal translocation from roots to shoots and its effect on the plant biomass yield and metal phytostabilization in field experiments.",
"introduction": "Introduction Mining activities produce waste tailings containing high levels of metal pollutants that have significant environmental impacts and can affect human health through the food chain ( Moreno et al., 2010 ). During mineral ore processing mine tailings are generated and mostly left without proper management, therefore leading to metal contamination of surrounding soils, with detrimental impacts on the soil microbial community and consequent reduction in ecosystem functioning ( Kossoff et al., 2014 ). Phytoremediation, application of plants to remove (phytoextraction), stabilize (phytostabilization), or volatilize (phytovolatilization) heavy metals in situ in a more attractive and cost-effective manner than the conventional physicochemical technologies, has received increasing attention over the last decades ( Raskin and Ensley, 2000 ). Particularly, phytostabilization of metal contaminated mine tailings, which uses plants that minimize metal accumulation into aboveground tissues, seems to be most promising for remediating polluted sites when phytoextraction is not a feasible option ( Mendez and Maier, 2008 ). Because mine tailings have low nutrient contents, to overcome the limitations of plant establishment, soil amendments with fresh or composted organic matter (biochemical amendments) can enhance plant colonization and reduce metal toxicity and solubility, thereafter improving phytostabilization efficiency ( Lee et al., 2011 ). However, most of those biochemical amendments, such as cyclonic ashes, steel shots. and superphosphate, are toxic to plants and their associated microbes ( Ribeiro Filho et al., 2011 ). Some beneficial bacteria have been successfully employed for environmental applications (so-called bioaugmentation) due to their ability to: (1) promote plant growth by producing beneficial metabolites [e.g., siderophores, indole-3-acetic acid (IAA) and 1-aminocyclopropane-1-carboxylate (ACC) deaminase] and solubilizing phosphate (P); and/or (2) alter soil metal mobility without disturbing soil ecological structure and function through various mechanisms such as metal biosorption/bioaccumulation, redox reaction, chelation, or complexation ( Glick, 2010 ; Ma et al., 2011a ). Amongst all the beneficial features, the production of ACC deaminase is considered as one of the major plant growth promoting traits of bacteria ( Ma et al., 2011a ). Mesa et al. (2015) reported that the inoculation of autochthonous rhizobacteria stimulated the biomass of native Spartina maritima , and enhanced metal (As, Cu, Pb, and Zn) accumulation in the root of plants (rhizoaccumulation) grown in multi-metal contaminated soils. Since bacterial endophytes have more intimate association with host plants than rhizobacteria, they could be reliable bioinoculants for improving metal phytostabilization. Although the effects of inoculating endophytic bacterial strains on growth promotion of various host and/or non-host plants have been reported ( Ma et al., 2011b ; Shin et al., 2012 ; Visioli et al., 2014 ), little is known on application of endophytes for enhanced phytoremediation of natural non-sterile polluted soils. Sedum plumbizincicola is known to hyper-accumulate or extract Cd and Zn from soils ( Jiang et al., 2010 ). Despite the potential of S. plumbizincicola to remove various metals from polluted soil, its slow growth is a limitation that needs to be overcome. The objectives of this study were to: (1) isolate and characterize metal resistant endophytic bacteria that can utilize ACC as the sole nitrogen (N) source; (2) assess plant beneficial activities, metal biosorption and mobilization capacities of a selected endophytic bacterium; (3) examine the effect of the endophytic bacterium on plant growth and metal accumulation/translocation in host S. plumbizincicola in non-sterile multi-metal contaminated soils.",
"discussion": "Results and Discussion Isolation and Identification of Metal Resistant and ACC-Utilizing Endophytic Bacterium Although the interaction between endophytic bacteria and their host plants is not completely understood, some endophytic bacteria isolated from heavy metal hyperaccumulators appear to exert beneficial effects on their hosts ( Shin et al., 2012 ), such as amelioration of metal stress, stimulation of plant establishment and growth, and biocontrol of phytopathogens ( Ma et al., 2011a ). Consequently, such beneficial endophytic bacteria could be isolated and selected for their application in assisting phytoremediation of metal contaminated soils ( Rajkumar et al., 2009 ). In this study, we isolated a metal resistant endophytic bacterial strain from stems of Zn/Cd hyperaccumulator S. plumbizincicola and assessed its effect as a bioinoculant on multi-metal phytoremediation by its host plant. The initial screening was based on the morphological differences of bacterial colonies and resulted in the isolation of 42 metal (100 mg L -1 of Cd, Zn, and Pb) resistant endophytic strains from interior tissues of S. plumbizincicola . Out of 42 strains, isolate E6S was specifically chosen based on its fast growth and capability of utilizing ACC as the sole N source. Based on morphological and biochemical characteristics ( Table 1 ), comparative analysis of 16S rDNA sequence and phylogenic analysis ( Figure 1 ), strain E6S was identified as being homologous to Achromobacter piechaudii (100% similarity). The sequence obtained (858 bp) was submitted to the NCBI databases under the accession number KC151254. Strain E6S was gram-negative, motile, rod shaped and positive for oxidase and catalase. It was able to produce H 2 S, utilize citrate and hydrolyze starch ( Table 1 ). Table 1 Morphological, physiological, and biochemical characteristics of endophytic bacterium E6S homologous to Achromobacter piechaudii . Characteristic E6S Gram staining – Cell shape Rod Oxygen requirements Aerobic Motile + Growth at 15–40°C + Growth at 6% NaCl + Oxidase + Catalase + Indole production – Voges–Proskauer test – H 2 S production + Nitrate reduction + Nitrite reduction – Utilization of Arabinose – Mannitol – Maltose – Glucose – Citrate + Lactate – Hydrolysis of Casein – Starch + Gelatin – Esculin – +, positive; –, negative. FIGURE 1 Phylogenetic tree showing the relationship of partial 16S rDNA gene sequences from endophytic bacterium E6S homologous to Achromobacter piechaudii with other related sequences from Achromobacter . \n Escherichia coli was used as the out-group. The value on each branch is the percentage of bootstrap replications supporting the branch. Biochemical Properties of Endophytic Bacterium Under various abiotic stresses (e.g., heavy metals, drought, and salinity), microorganisms face a constant battle for limited resources and try to adapt to unfavorable environmental conditions by acting as stress ameliorators ( Hibbing et al., 2010 ). Hence, strain E6S was tested for the ability to grow on both metal and antibiotic-supplemented agar media. Strain E6S was found to exhibit resistance to heavy metal (Cd, Zn, and Pb) and various antibiotics ( Table 2 ). The order of the toxicity of metals to the isolate was found to be Cd > Zn > Pb. This strain exhibited high tolerance to multiple metals, which could be attributed to the fact that it was isolated from tissues of a Zn/Cd hyperaccumulator plant, likely containing high levels of these bioavailable metal ions ( Idris et al., 2004 ). Among six antibiotics tested, strain E6S showed resistance to ampicillin and tetracycline. Table 2 Heavy metal tolerance, antibiotic resistance, and plant growth promoting traits of endophytic bacterium E6S homologous to Achromobacter piechaudii . Parameter Unit E6S Metal tolerance mg L -1 Cd 300 Zn 750 Pb 1200 Antibiotic resistance mm Ampicillin (10 μg) 3 (R) Tetracycline (30 μg) 9 (R) Streptomycin (20 μg) 25 (S) Chloramphenicol (30 μg) 15 (I) Kanamycin (30 μg) 23 (S) Plant growth promoting traits ACC deaminase μm α-KB mg -1 h -1 11 ± 0.9 P solubilization mg L -1 135 ± 16 IAA production mg L -1 22 ± 6 Siderophore (CAS) cm nd R, resistant (<10 mm); I, intermediate (10–15 mm); S, susceptible (>15 mm); ACC deaminase, 1-aminocyclopropane-1-carboxylate deaminase; α-KB, α-ketobutyrate; P, phosphate; IAA, indole-3-acetic acid; CAS, chrome azurol S; nd, not detected. Endophytic bacteria can stimulate plant growth directly by solubilizing unavailable nutrients (e.g., P, N, and potassium), sequestering iron by siderophores and phytohormones (e.g., IAA) or indirectly by inducing a systemic resistance in plants against various types of pathogens ( Ma et al., 2011b ). These features make them perfect choices for improving phytoremediation. Strain E6S was able to produce ACC deaminase, IAA and solubilize P, whereas no biosynthesis of siderophore was found ( Table 2 ). In general, bacterial IAA at low level has been implicated in promoting primary root elongation through cell division, however, a high IAA level can inhibit primary root growth but stimulate lateral root formation ( Gravel et al., 2007 ). A low level of IAA production by strain E6S (22 mg L -1 ) suggests close relationship between plant growth promotion activity ( Table 3 ) and IAA production. Strain E6S exhibited high tricalcium phosphate-solubilizing ability (135 mg L -1 ; Table 2 ), which may compensate for P deficiency-induced plant growth retardation in metal contaminated soil. ACC-utilizing bacteria have been found to prevent the inhibition of root growth by hydrolyzing the ethylene precursor ACC into ammonia and α-KB and inhibiting ACC synthase activity ( Arshad et al., 2007 ). Although strain E6S showed relatively low ACC deaminase activity (11 μm α-KB mg -1 h -1 ), it seemed to be efficient in promoting plant growth under metal stress ( Table 3 ). The results suggest that the endophytic bacteria, which can synthesize beneficial metabolites such as IAA, ACC deaminase, and solubilize P should be considered as promising biofertilizers. Table 3 Effects of endophytic bacterium E6S homologous to Achromobacter piechaudii on the growth of Brassica napus in phytagar assay and Sedum plumbizincicola in pot experiment. Type Treatment Relative elongation ratio (RER) of root a RER of shoot b Fresh weight (mg) Dry weight (mg) Ratio of root/shoot dry weight c Phytagar assay Control Blank – – 53.5 ± 3.1 b 2.6 ± 0.0 b 0.74 ± 0.08 b E6S 1.62 ± 0.18 b 1.15 ± 0.10 b 67.6 ± 5.3 a 3.2 ± 0.3 a 1.04 ± 0.08 a 10 mg L -1 Cd Blank – – 28.8 ± 4.4 c 1.9 ± 0.3 c 0.32 ± 0.04 c E6S 2.78 ± 1.00 a 1.54 ± 0.20 a 47.6 ± 4.2 b 2.8 ± 0.2 b 0.62 ± 0.03 b Pot experiment Pristine soil Blank – – 1735 ± 169 z 221 ± 16 z 0.09 ± 0.05 w E6S 1.26 ± 0.13 x 1.27 ± 0.10 w 2416 ± 157 y 275 ± 19 y 0.12 ± 0.01 w Metal polluted soil Blank – – 6086 ± 403 x 739 ± 24 x 0.02 ± 0.01 x E6S 1.89 ± 0.16 w 1.36 ± 0.10 w 8211 ± 481 w 946 ± 51 w 0.03 ± 0.01 x Average ± standard deviation from five samples. Data of columns indexed by the same letter within each treatment (with or without Cd; pristine or polluted soil) are not significantly different between bacterial treatments according to Fisher’s least significant difference (LSD) test ( p < 0.05). a RER of root = Mean root length of tested plant/Mean root length of control × 100. b RER of shoot = Mean shoot length of tested plant/Mean shoot length of control × 100. c Ratio of root/shoot dry weight = Dry weight of root/Dry weight of shoot. Metal Mobilization and Biosorption Potential of Endophytic Bacterium Endophytic bacterium E6S homologous to A. piechaudii displayed the potential for biological mobilization of Cd, Zn, and Pb in multi-metal contaminated soil ( Figure 2 ). The inoculation of E6S significantly increased ( p < 0.05) water extractable Cd, Zn, and Pb in metal contaminated soils by 3.9-, 5.8- and 6.0-fold, respectively, compared to the controls. The observation indicates that strain E6S facilitated the release of metals from non-soluble phase in the soil matrix, thus improving their bioavailability to plants. This may be attributed to organic acid production and phosphate solubilization mediated reduction in soil pH ( Ma et al., 2009 , 2015 ). FIGURE 2 Effect of inoculation with E6S on the mobilization of Cd, Zn, and Pb in soil. Bars represent standard deviations of triplicates. An asterisk ( ∗ ) denotes a value significantly greater than the corresponding control value according to Student’s t -test ( p < 0.05). Biosorption capacity of bacteria plays an important role in reducing metal phytotoxicity by limiting the entry of metal ions into plant cells and may contribute for enhanced plant growth in metal contaminated soils ( Ma et al., 2011a ). At 150 mg L -1 initial metal concentrations, strain E6S was able to remove significant amounts of Cd, Zn, or Pb within 10 h incubation ( Figure 3 ). After 8 h incubation, maximum biosorption by E6S was reached, thereafter remaining constant. This was probably due to the achievement of specific equilibrium for metal concentrations. The highest content of metal biosorption was observed with Zn (10.9 mg g -1 of cell dry weight), while the lowest was seen with Pb (3.2 mg g -1 of cell dry weight). This was probably due to ionic radius of each metal ion ( Karakagh et al., 2012 ), since Zn (0.88 Å) with smaller ionic radius may be more rapidly complexed by bacterial cell wall/membrane compared with Cd (0.97 Å) and Pb (1.2 Å). FIGURE 3 Biosorption of Cd, Zn, and Pb on E6S cells. Bars represent standard deviations of triplicates. Effects of Endophytic Bacterial Inoculation on Plant Biomass In phytagar assay, inoculation of E6S induced significant increases in fresh and dry weight, ratio of root/shoot dry weight of B. napus in both unpolluted and polluted (10 mg L -1 Cd) phytagar media ( Table 3 ). For instance, E6S induced an increase in fresh and dry weight, and ratio of root/shoot dry weight in Cd polluted media by 65, 47, and 94%, respectively, compared to non-inoculated control. Bacterial inoculation greatly enhanced the biomass production of B. napus under non-stressed and metal-stressed conditions. Plants inoculated with strain E6S presented a significantly higher root/shoot ratio than the respective controls, indicating that with the help of endophytic bacterium E6S, plants can acquire nutrients more efficiently from soils for plant biomass production, especially for roots. Moreover, the relative elongation rate (RER) of root and shoot in unpolluted media (1.62 and 1.15) were lower than that in polluted treatment (2.78 and 1.54). A possible explanation might be that the endophytic bacteria could exert more efficient functions that help plants to cope with adverse environmental stress ( Rajkumar et al., 2009 ). In pot experiment, inoculation of E6S significantly improved fresh and dry weight of S. plumbizincicola in both pristine and multi-metal polluted soils under non-sterile conditions ( Table 3 ). For example, inoculation of E6S considerably increased plant fresh and dry weight by 35 and 28%, respectively, in metal polluted soils. The beneficial effects of E6S on growth of host S. plumbizincicola were also observed on growth of non-host B. napus in phytagar assay. These effects may be attributed to beneficial metabolites produced by strain E6S, such as IAA and ACC deaminase, which alleviated metal phytotoxicity and thereafter stimulated plant development. Additionally, inoculation of E6S did not significantly influence ratio of root/shoot dry weight in neither pristine nor polluted soils. The presence of metals declined ratio of root/shoot dry weight, compared with non-polluted control. These results concur with the earlier observations of Wan et al. (2012) that elevated metal concentrations mainly impaired root growth, in spite of S. plumbizincicola being qualified as a Zn/Cd hyperaccumulator. It is therefore concluded that bioinoculant E6S could serve as biofertilizer for phytoremediation purposes. Interestingly, the fresh and dry weights of S. plumbizincicola grown in metal contaminated soil were higher than in pristine soil ( Table 3 ). A possible explanation could be that the current levels of heavy metals in soil are not toxic to the plant and can stimulate its growth. This may be attributed to both a beneficial effect of improved metal nutrition and/or the activation of stress scavenging mechanisms, which can help the plant to cope with environmental stresses ( Dalcorso et al., 2013 ). Endophytic Bacteria-Enhanced Phytostabilization The changes in metal solubility in polluted soils caused by biological amendments can contribute to facilitate metal accumulation in plants ( Ma et al., 2009 ). Therefore, the effects of metal mobilizing bacterium E6S on metal uptake and translocation by S. plumbizincicola were evaluated. In general, inoculation of strain E6S significantly improved plant uptake of Cd, Zn, and Pb ( Table 4 ). For instance, strain E6S increased Cd, Zn, and Pb concentration in S. plumbizincicola by 32, 37, and 89%, respectively, which is in accordance with significant improvements of BCF of metal (Cd, Zn, and Pb) induced by E6S. This corroborates the data shown in Figure 2 for bacterial metal mobilization, indicating that inoculation of E6S facilitated metal (Cd, Zn, and Pb) bioavailability in soils and thereby their uptake by plants. However, the bacterial inoculation significantly decreased TF of Cd and Zn ( p < 0.05; Table 4 ) and metal accumulation in shoots (data not shown). The TF was 2.7 and 1.5 (>1) for Cd and Zn in non-inoculated plants, but it decreased to 0.5 and 0.7 (<1) after inoculation with strain E6S. S. plumbizincicola has been reported to be a hyperaccumulator for Cd and Zn phytoextraction due to their TF > 1, however, our data showed that the inoculation of strain E6S inhibited the plant-self translocation of metal from roots to shoots and helped plants store metals in their roots, which is desirable for phytostabilization purposes. The present observations indicate that strain E6S effectively increased the bioavailability of metal (Cd, Zn, and Pb) in the rhizosphere soils and also promoted the growth of host S. plumbizincicola plants, consequently increasing the total plant metal uptake, while diminishing the translocation of metals from roots to shoots. Results suggest that endophytic bacterium E6S can not only protect the plants against the inhibitory effects of multiple metals, but also effectively improve rhizoaccumulation of toxic metals. Therefore, it can be used to assist phytostabilization of heavy metals in the plant root system. Previously, Srivastava et al. (2013) also reported that Staphylococcus arlettae NBRIEAG-6 enhanced As accumulation in roots of B. juncea and helped in As phytostabilization. Recently, Sura-de Jong et al. (2015) reported that selenium hyperaccumulators harbor a diverse endophytic bacterial community. It is becoming apparent that some endophytic bacteria play an important role in metal hyperaccumulation and translocation processes, contributing to phytoextraction, while others seem to be effective in improving rhizoaccumulation, contributing to phytostabilization. Future studies are, therefore, needed to access the structure and diversity of bacterial endophytes of S. plumbizincicola and their influence in phytoremediation processes. Table 4 Effects of endophytic bacterium E6S homologous to Achromobacter piechaudii on metal uptake, translocation factor and bioaccumulation factor by Sedum plumbizincicola . Treatment Cd Zn Pb Plant uptake (mg kg -1 dw) TF BCF Plant uptake (mg kg -1 dw) TF BCF Plant uptake (mg kg -1 dw) TF BCF Control 127 ± 8 2.7 ± 0.3 ∗ 21.6 ± 1.4 2262 ± 61 1.5 ± 0.1 ∗ 3.1 ± 0.1 114 ± 10 0.2 ± 0.0 0.7 ± 0.1 E6S 168 ± 25 ∗ 0.5 ± 0.1 28.4 ± 4.2 ∗ 3095 ± 380 ∗ 0.7 ± 0.1 4.9 ± 0.2 ∗ 215 ± 20 ∗ 0.2 ± 0.0 1.4 ± 0.1 ∗ Average ± standard deviation from five samples. ∗ A value significantly greater than the corresponding control according to Student’s t-test ( p < 0.05). TF, translocation factor; BCF, bioaccumulation factor; dw, dry weight. Bacterial Colonization After 75 days of inoculation onto plants, strain E6S exhibited high-density colonization of the rhizosphere [3.9 × 10 5 colony-forming units (CFU) g -1 ], and fresh roots, stems and leaves (5.4, 3.7, and 0.8 × 10 3 CFU g -1 , respectively) of S. plumbizincicola , indicating the great potential of this strain to establish, survive, and develop inside plant tissues and in the root zone. Since beneficial bacteria have great potential to contribute to sustainable plant growth promotion and metal rhizoaccumulation, the colonization and survival properties of introduced beneficial strains are crucial features to evaluate the capacity to assist their host plants in coping with metal stress and therefore phytostabilization efficiency in contaminated sites ( Ma et al., 2011a )."
} | 6,002 |
30410711 | PMC6171786 | pmc | 2,295 | {
"abstract": "ABSTRACT Since the development of DNA origami by Paul Rothemund in 2006, the field of structural DNA nanotechnology has undergone tremendous growth. Through DNA origami and related approaches, self-assembly of specified DNA sequences allows for the ‘bottom-up’ construction of diverse nanostructures. By utilizing different sets of small ‘staple’ DNA strands to direct the folding of a long scaffold strand in diverse ways, DNA origami has particularly been incorporated into a variety of prototypical applications beyond the two-dimensional (2D) smiley face. In this review, the basis of DNA nanotechnology, methods of self-assembly, and Rothemund’s DNA origami breakthrough are discussed first. Next, some of the most promising applications of structural DNA nanotechnology since 2006 are summarized. These include utilizing DNA origami as a tool for creating 3D nanostructures (including DNA bricks), as well as structural (ligand capsid binding, viral capsid binding, DNA NanoOctahedron, DNA mold, photonic devices, energy transfer units), and dynamic (DNA box-with-lid, DNA nano-robot, DNA barges, amphipathic DNA structures, DNA nanocircuits) applications of DNA origami.",
"introduction": "1. Introduction A multitude of new DNA applications have been introduced since the discovery of the double helix structure in the mid-twentieth century ( Figure 1 ). One application that is proving to be promising in a variety of fields is DNA nanotechnology. DNA nanotechnology refers to the design, study, and application of synthetically created DNA nanostructures. The physical and chemical properties of DNA, rather than its genetic properties, are particularly malleable for various applications in the field of DNA nanotechnology [ 1 ]. In other words, the field of DNA nanotechnology aims to design synthetic DNA constructs that exhibit structures and functions that are not found for DNA in nature. Ultimately, natural properties of DNA are harnessed as tools that can be manipulated and applied in a variety of settings related to the field of biotechnology. 10.1080/20022727.2018.1430976-F0001 Figure 1. 3D DNA double-helix structure comprised phosphate backbones (purple) and hydrogen bonds between nitrogenous bases (yellow). Created using ANSYS software. \n The property of self-assembly, as per Watson-Crick base pairing, has largely been exploited in the field of DNA nanotechnology. The process of self-assembly is representative of what naturally occurs in DNA. That is, two single strands of known DNA sequences will self-assemble if they have complementary nitrogenous base sequences. Self-assembly is deemed ‘bottom-up’ approach which exploits local interactions between components within a system to yield a desired structure. A ‘bottom-up’ approach allows for a more precise, controlled, and predictable DNA nanostructure [ 2 , 3 ]. This is opposed to a ‘top-down’ approach, which uses external intervention to create an object by removing or adding material in a spatially controlled manner [ 3 ]. One advantage of self-assembly is the potential for molecular precision, which is difficult to achieve with ‘top-down’ approaches such as lithography. ‘Bottom-up’ approaches ultimately allow for the faster manufacturing of smaller nanostructures [ 2 ]. However, disadvantages of the self-assembly exist; first, the technique often leads to less than 100% yield of the desired product due to mis-assembly or mis-folding. Second, everything about the structure must be specified through local interactions, which requires many unique building blocks so that sufficient information becomes available; this can often be very expensive. The self-assembly process of DNA was first showcased with the creation of 2D nucleic acid junctions and lattice shapes [ 4 ]. These junctions were developed as clusters; the clusters were linked directly to each other, or with interspersed linear DNA pieces (later coined as ‘sticky ends’) [ 4 ]. Subsequently, creation of immobile branched junctions allowed for a building framework upon to which other molecules could be attached [ 5 ]. Specifically, a closed cube-like structure containing six faces, eight vertices, and 12 double helical edges was developed [ 5 ]. Ultimately, 2D crystalline DNA forms were developed from synthetic DNA double crossover molecules [ 6 ]. ‘Sticky ends’ of DNA allowed for intermolecular interactions between each unit, leading to the formation of specific patterned DNA crystals [ 6 ]."
} | 1,112 |
25889067 | PMC4374363 | pmc | 2,296 | {
"abstract": "In ancient Chinese philosophy, Yin-Yang describes two contrary forces that are interconnected and interdependent. This concept also holds true in microbial cell factories, where Yin represents energy metabolism in the form of ATP, and Yang represents carbon metabolism. Current biotechnology can effectively edit the microbial genome or introduce novel enzymes to redirect carbon fluxes. On the other hand, microbial metabolism loses significant free energy as heat when converting sugar into ATP; while maintenance energy expenditures further aggravate ATP shortage. The limitation of cell “powerhouse” prevents hosts from achieving high carbon yields and rates. Via an Escherichia coli flux balance analysis model, we further demonstrate the penalty of ATP cost on biofuel synthesis. To ensure cell powerhouse being sufficient in microbial cell factories, we propose five principles: 1. Take advantage of native pathways for product synthesis. 2. Pursue biosynthesis relying only on pathways or genetic parts without significant ATP burden. 3. Combine microbial production with chemical conversions (semi-biosynthesis) to reduce biosynthesis steps. 4. Create “minimal cells” or use non-model microbial hosts with higher energy fitness. 5. Develop a photosynthesis chassis that can utilize light energy and cheap carbon feedstocks. Meanwhile, metabolic flux analysis can be used to quantify both carbon and energy metabolisms. The fluxomics results are essential to evaluate the industrial potential of laboratory strains, avoiding false starts and dead ends during metabolic engineering.",
"conclusion": "Conclusions We have discussed the Yin-Yang concept as the underlying regulatory mechanism in cell metabolism. Biosynthesis of diverse useful products requires sophisticated genetic pathway engineering to steer a high flux to the final product while energy fitness requires the cell metabolism to be wisely changed. Since the powerhouse in microbial cell factory is not limitless, energy shortage eventually leads to metabolic shifts and reduced cell productivity in engineered microbes. The Yin-Yang balance may caution against the assumption that the host metabolism can be modified extensively to produce any desired products. By using fluxomics, we can formulate guidelines to avoid many false starts and dead ends during metabolic engineering. In addition, industrial bioprocess always faces numerous constraints and trade-offs (mass transfer limitations in fermentation, sterilization, strain stability, contaminations, and aeration costs). Feedstock selections, downstream product separation, and waste treatment are critical issues that impact product profitability. Thus, the design-build-test-learn cycle should cover both strain development and economic analysis. Nevertheless, the Yin-Yang philosophy provides general insights into all biotechnology tradeoffs.",
"introduction": "Introduction In the past decade, molecular biology tools have been developed rapidly and now offer new opportunities for metabolic engineering of microbial hosts [ 1 - 6 ]. These tools include the selection of plasmids with different copy numbers, promoter engineering, codon optimization, synthetic scaffolds, directed evolution or rational design of enzymes, ribosome binding sites editing, and competitive pathways deletion. Advanced genome engineering (e.g., CRISPRs and TALENs) and automation of conventional genetic techniques (e.g., MAGE) provide efficient capabilities for editing genomes and evolving new functions. At the same time, systems biology (e.g., genomics, transcriptomics, and proteomics) can characterize complex cell networks, mine useful genes, discover new enzymes, reveal metabolic regulations, and screen mutant phenotypes. The advent of these powerful tools seems to lead researchers into a new epoch of bioprocess industries using GMMs (genetically modified microorganisms) in the near future. However, that is not the whole story. The golden age of industrial biotechnology dawned in the early 1940s, driven by the mass production of penicillin and enjoyed a fast growth in the 1950s ~ 1980s. Microbial bioprocess has produced diverse commodity chemicals (such as ethanol, amino acids, citric acid, and lactate) as well as recombinant proteins and antibiotics in the last century. Those commercial products mainly rely on natural strains or strains with minor genetic modifications (usually only one or few new genes). Since the recent decade, in the hope of producing chemicals at low costs and reducing greenhouse gas emissions, an enormous amount of investment has been devoted to metabolic engineering in many nations. Although modern biotechnologies can engineer microbial platforms to synthesize diverse products in laboratories, there are only a few GMM products that have become commercially promising in the past decade (e.g., artemisinic acid and 1, 4-butanediol). Novel GMMs are also used for chemical manufactures, such as short-chain alcohols and isoprene [ 7 - 9 ]. Recently, Gevo and Butamax introduce the keto-acid/Ehrlich pathway into yeasts to produce isobutanol [ 10 ]. Amyris extend the mevalonate pathway in Saccharomyces cerevisiae for branched and cyclic terpenes (e.g., farnesene) synthesis. However, these companies have not achieved strong net profit margin yet. To date, the industrial-scale biofuel is still ethanol, which is cheaply manufactured from sugar cane in Brazil. In this perspective, we address one of the hidden constraints in microbial cell factories (i.e., energy metabolism)."
} | 1,376 |
39209850 | PMC11362583 | pmc | 2,297 | {
"abstract": "The geosphere and the microbial biosphere have co-evolved for ~3.8 Ga, with many lines of evidence suggesting a hydrothermal habitat for life’s origin. However, the extent that contemporary thermophiles and their hydrothermal habitats reflect those that likely existed on early Earth remains unknown. To address this knowledge gap, 64 geochemical analytes were measured and 1022 metagenome-assembled-genomes (MAGs) were generated from 34 chemosynthetic high-temperature springs in Yellowstone National Park and analysed alongside 444 MAGs from 35 published metagenomes. We used these data to evaluate co-variation in MAG taxonomy, metabolism, and phylogeny as a function of hot spring geochemistry. We found that cohorts of MAGs and their functions are discretely distributed across pH gradients that reflect different geochemical provinces. Acidic or circumneutral/alkaline springs harbor MAGs that branched later and are enriched in sulfur- and arsenic-based O 2 -dependent metabolic pathways that are inconsistent with early Earth conditions. In contrast, moderately acidic springs sourced by volcanic gas harbor earlier-branching MAGs that are enriched in anaerobic, gas-dependent metabolisms (e.g. H 2 , CO 2 , CH 4 metabolism) that have been hypothesized to support early microbial life. Our results provide insight into the influence of redox state in the eco-evolutionary feedbacks between thermophiles and their habitats and suggest moderately acidic springs as early Earth analogs.",
"introduction": "Introduction The microbial biosphere and the geosphere (comprising the lithosphere, hydrosphere, cryosphere, and atmosphere) have changed in concert for ~3.8 Ga 1 – 3 . Subaerial hydrothermal systems (i.e. hot springs) are among the earliest known habitats to support microbial life, as evinced by ~3.5 Ga fossil and geochemical evidence of microorganisms and their activities preserved in ancient hot spring deposits 4 , 5 . Phylogenetic and inferred physiologic data also consistently place microbial thermophiles as among the earliest-branching lineages and suggest that they were supported by anaerobic, chemoautotrophic metabolisms dependent on geogenic energy substrates 6 , 7 . Concomitantly, contemporary continental hydrothermal systems host diverse and abundant microbial communities 8 – 11 , largely attributed to the extensive geochemical variation and abundant energy substrates available in these environments 8 , 9 , 12 . At the highest temperatures (>~74 o C in circumneutral springs and >~54 o C in acidic springs), photosynthetic metabolism is excluded in hot springs and microbial productivity is driven by chemoautotrophic metabolism supported by geogenic energy substrates like H 2 S, H 2 , CO 2 , and S 0 13 – 16 . Consequently, contemporary high-temperature springs and the communities they support provide an opportunity to understand how chemosynthetic populations were supported on early Earth and how they have evolved alongside their hydrothermal habitats over geologic time. The extensive variation in the geochemical composition observed within hot springs is generated by a convergence of several surface and subsurface processes. In most systems, oxic meteoric or saline waters infiltrate the crust and are heated to high temperature, forming a hydrothermal aquifer 17 , 18 . These hydrothermal aquifer fluids tend to be solute rich (e.g. of Na + and Cl − ) due to extensive high-temperature water-rock reactions and are infused with magmatic gases like CO 2 , 3 He, and SO 2 , the latter of which disproportionates at high temperature in the presence of water to form SO 4 2− and H 2 S or S 0 , depending on temperature and sulfur concentrations 19 . During circulation in the subsurface and their ascent to the surface, hydrothermal fluids become depleted in oxygen and are further enriched in crustal or meteoric gases such as CH 4 , 4 He, and H 2 17 , 18 , 20 , 21 . The development of stark chemical variation in waters principally arises from fluids undergoing decompressional boiling during their ascent along fractures and faults, resulting in separation of fluids into a lower-density vapor phase carrying volatile gases (e.g. CO 2 , H 2 S, H 2 , and CH 4 ) and a higher density liquid phase comprising non-volatilizing solutes (e.g. Na + and Cl − ) 17 , 18 , 22 . The low-density vapor phase can continue ascending to the surface and form fumaroles or otherwise condense with near-surface waters 18 , 23 . Condensation of vapor phase gases with near surface fluids equilibrated with atmospheric O 2 can result in the oxidation of H 2 S, resulting in the formation of S 0 and ultimately sulfuric acid (H 2 SO 4 ) 23 – 25 . The oxidation of H 2 S and S 0 compounds is likely driven, at least in part, by O 2 -dependent thermoacidophilic Archaea 25 – 29 that may have diverged from neutrophilic ancestors over only the past ~1 Ga 28 . The separation of fluids into a vapor phase and a liquid phase and subsequent oxidation of H 2 S to form SO 4 2− and acid leads to the bimodal distribution of hot spring pH observed globally 30 . This includes the two primary types of hot springs in contemporary geothermal fields: acid-sulfate and circumneutral-alkaline springs. Given that O 2 only began to accumulate in the atmosphere ~2.4 Ga and only reached present-day levels at ~0.8 to 0.6 Ga (as summarized in ref. 31 ), acidic hot springs and the thermoacidophiles they host are likely to be relatively recent phenomena on Earth 28 . Circumneutral to alkaline pH hot springs are the surface expression of the liquid phase and thus tend to be gas- and oxidant-poor, the latter of which is due to long residence times in the subsurface 22 that permit extensive water-rock reactions and consumption of oxidants 17 . Paradoxically, these springs are often dominated by obligately aerobic bacteria within the Aquificales order 32 – 34 and have recently been argued to only be habitable due to the infusion of atmospheric O 2 once the waters reach the surface 22 , 35 . It is consequently unclear whether acidic hot springs or circumneutral/alkaline hot springs could have been prevalent microbial habitats on early Earth when O 2 was not readily available. Springs with pH intermittent to acid-sulfate and circumneutral/alkaline types are globally more rare 30 and occur due to dilution of hydrothermal water by meteoric water or by mixing of meteoric and/or hydrothermal water/gasses 36 , 37 . Thus, while hot spring ecosystems are often invoked as suitable analogs for investigating early life 4 , 38 , it remains unclear how well their contemporary geochemistry and microbiology reflect early Earth conditions. The most widely studied hydrothermal system is in Yellowstone National Park (YNP) that is host to >10,000 geothermal features that widely vary in their geochemical composition 12 , 17 , 24 , 25 , 39 . Extensive spatial geochemical variation among hot springs selects for unique assemblages of microorganisms with diverse functionalities 8 , 10 , 11 , 34 . However, the provenance of this taxonomic and functional diversity and their distributions remain unclear, in particular, given that the geochemical compositions of hydrothermal waters are likely to have substantively changed over Earth history as the Earth became more oxidized, as described above. Here, a census of hot spring genomic and functional diversity was generated alongside detailed geochemical measurements in 34 high-temperature hot springs (>61.9 o C) with conditions that preclude photosynthesis to evaluate the adaptive evolution of thermophilic lineages and their functions in coordination with their hydrothermal habitats. A total of 64 geochemical analytes were measured and 1022 metagenome-assembled-genomes (MAGs) were generated from community metagenomes from these springs. These data were also analyzed in conjunction with 444 MAGs generated from 35 metagenomes in our other recent studies 8 , 29 , 35 , 40 – 43 .",
"discussion": "Discussion Hot springs have long served as analogs to understand the environmental conditions and metabolisms that supported the earliest life on Earth. However, hot springs vary widely in their geochemical and microbiological composition, rendering it unclear the extent to which contemporary hydrothermal environments actually reflect the types of environments that might have supported early life on Earth. Here, a comprehensive and integrated assessment of geochemical, physiological, and phylogenetic data of YNP high-temperature springs and their communities revealed that the two most abundant spring types in contemporary continental hydrothermal environments (acidic and circumneutral/alkaline springs) host relatively later-branching archaeal and bacterial lineages, while also hosting communities that are especially dependent on atmospheric O 2 for their functioning. In contrast, moderately acidic springs (pH 5–7) that are generally rare in continental hydrothermal systems today hosted relatively earlier-branching lineages comprising populations especially supported by anaerobic, gas-dependent metabolisms. Moderately acidic springs are formed by mixing of volatile (e.g. H 2 S, H 2 , CO 2 CH 4 ) rich gas with meteoric or hydrothermal waters and represent precursors to the formation of acidic springs (pH < 4). However, the transition from moderately acidic (pH 5–7) conditions to acidic springs (pH < 4) requires sufficient oxidizing power (e.g. O 2 or O 2 -derived oxidation products such as Fe(III)) to generate strong acid (e.g. sulfuric acid) and long enough residence times to allow for acidification to occur. In many instances, such moderately acidic springs form at higher elevations in YNP, likely facilitating lower water residence times 20 , 21 that enables incomplete acidification and suggests a geomorphological component to the geochemical-microbiological associations identified herein. Consequently, prior to the emergence of oxygenic photosynthesis and the generation of widely available O 2 during Earth history, moderately acidic springs were likely much more prevalent. Thus, the generally anoxic, gas-rich conditions observed in contemporary moderately acidic springs may better reflect conditions of such springs. Infusion of O 2 from the atmosphere or by input of O 2 -containing meteoric water is unavoidable in hot spring environments today. Nevertheless, moderately acidic springs in contemporary continental geothermal fields and the microbial populations they host may serve as useful platforms to understand volcanic gas-influenced, similar springs that may have been more prevalent on a mostly anoxic early Earth. The results of this study provide insight into the distribution of genomic biodiversity, functional potential, and phylogenetic patterning of thermophilic Archaea and Bacteria as a function of geochemical variation in YNP hot springs. It should nevertheless be noted that the YNP geothermal system is a singular example that has arisen primarily due to rhyolitic volcanism. Consequently, similar studies should be conducted in other geologic contexts, although the prevailing geophysical and geochemical processes that lead to geochemical variation in YNP should be similar globally and should lead to similar eco-evolutionary feedbacks within local environments. Consistently, our recent 16S rRNA gene-based analysis of hot spring community composition among globally distributed, geologically distinct settings demonstrated similar co-variation in geochemical-taxonomic composition patterns across systems 61 . These collective results suggest that complex influences from geologic settings, dispersal limitation, and localized geologic/hydrologic characteristics likely influence observed biogeographic patterns of global hot spring biodiversity. Notably, tectonic setting has recently been shown to significantly influence coordinated geochemical and microbiological variation observed among continental 61 , 119 , 120 and deep-sea 62 hydrothermal systems. Additional targeted comparisons of such systems may provide further insights into the coordinated evolution of thermophiles and their hydrothermal environments across Earth history and across global hydrothermal contexts."
} | 3,050 |
29382050 | PMC5855892 | pmc | 2,299 | {
"abstract": "A recent trend in the development of high mass consumption electron devices is towards electronic textiles (e-textiles), smart wearable devices, smart clothes, and flexible or printable electronics. Intrinsically soft, stretchable, flexible, Wearable Memories and Computing devices (WMCs) bring us closer to sci-fi scenarios, where future electronic systems are totally integrated in our everyday outfits and help us in achieving a higher comfort level, interacting for us with other digital devices such as smartphones and domotics, or with analog devices, such as our brain/peripheral nervous system. WMC will enable each of us to contribute to open and big data systems as individual nodes, providing real-time information about physical and environmental parameters (including air pollution monitoring, sound and light pollution, chemical or radioactive fallout alert, network availability, and so on). Furthermore, WMC could be directly connected to human brain and enable extremely fast operation and unprecedented interface complexity, directly mapping the continuous states available to biological systems. This review focuses on recent advances in nanotechnology and materials science and pays particular attention to any result and promising technology to enable intrinsically soft, stretchable, flexible WMC.",
"conclusion": "7. Conclusions The primary use of any embodiment of RSDs as non-volatile memory and finally as neuromorphic devices makes this field interesting and technologically challenging for the integrated circuit industry, to expand in fields such as medicine. The vision of wearable memory and computing machines, in the sense of e-textiles, describes the future electronic systems for smart personal assistance, augmentation and recovery. In the present review, we have tried to go through some important discoveries which report flexible, bendable and, some of them, even stretchable and biodegradable devices that have important future prospects to be used in wearable or even implantable devices. We do believe that epochal changes occur following a brand new route, unexpectedly. WMCs do not need to follow the same story of integrated electronics, therefore rely on solid state, opaque/rigid devices, in particular integrated circuits. Whenever the computational demand of a specific task assigned to WMCs would be outside its capabilities, the integrated wearable system could use the cloud for high performance real-time computing/storage (as happens for smartphones during complex tasks such as speech or image recognition).",
"introduction": "1. Introduction “Flexible”, “Stretchable”, “Bendable” and “Wearable” are some of the most common terminologies which may dramatically change our individual interaction scheme without impacting on the way of living in the very near future. This will occur by enabling a more natural interaction between humans and electron devices, shifting the paradigm from the actual “User”, somebody that physically interacts voluntarily with a physical object such as a smartphone or a computer, to the next evolutionary step, the “I-user”, somebody that physically interacts involuntarily by simply wearing her/his outfit (physically unbound from the user’s body, yet connected to it by means of non-invasive electrodes). With an ultimate effort, the last evolutionary step will make us “A-user”, augmented users that integrate new or augmented biological functionalities through bio-compatible integrated electron devices (physically implanted in the user’s body). Research in field of stretchable and bendable electronics is thriving day by day in the field of smart textiles or electronic textiles (e-textiles) [ 1 ]. In the present scenario, these flexible electronic devices are in high demand from the medical industry [ 2 ], for real-time patient monitoring. Nonetheless, from the most technical and specific usage to the most common and private one, smart wearables have the potentiality to fill the gap between standard clothes and electronics even for the most reluctant. We here introduce the terminology Wearable Memories and Computing devices (WMC) to create a distinction between this area, where each component is meant to be fabricated with innovative technologies and novel materials, and the Wearable Computers, known since the 1970s, integrated into clothes using the basic discrete electronics components encased and adapted to textiles. The former topic is dealt here, while the latter is not object of our treatise. Wearable Memories and Computing devices (WMC) will enable each of us to contribute to open and big data systems as individual nodes, providing extremely valuable real-time and super fine granularity information about physical and environmental parameters. Such data are of primary importance, and include air pollution monitoring, sound and light pollution, chemical or radioactive fallout alert, network availability, and so on. Furthermore, WMC could be directly connected to human brain [ 3 ] and enable extremely fast operation and unprecedented interface complexity, directly mapping the continuous states available to biological systems. We strongly believe that a real, feasible, reasonable advancement towards cybernetics, smart prosthesis and life augmentation, goes beyond the use of both standard Complementary Metal Oxide Semiconductor (CMOS) circuits and CMOS based neuromorphic emulators. Literature already reports several promising and interesting results regarding totally printed Thin Film Transistors (TFTs) (e.g., [ 4 , 5 , 6 ]), as well as other key enabling technologies, such as printable conductors to realize electrical connections and transfer energy or signal across a textile [ 7 , 8 ]. Additive/digital manufacturing appear suited for the development of really flexible, stretchable, wearable devices, while standard, solid-state electronics still will need a (weak, brittle) connection to the real macroscopic world. However, the present review delves into the development of resistive switching devices (RSD), otherwise known as memristors, as the most promising candidates towards WMC applications, including their potentiality for neuromorphic computing and direct coupling with biological brains. This conclusion is supported by several authors, to cite one, in an extremely synthetic and logic reasoning [ 9 ]: (1) a brain-inspired computing system should ideally employ some form of non-volatile memory; and (2) the dominant non-volatile technology, flash, is expected to be superseded by novel technologies, such as Phase Change Memory (PCM), spin transfer torque random access memory (STT-RAM) or resistive random access memory (ReRAM). RSDs (PCM + STT-RAM + ReRAM + other technologies not yet commercial), theoretically predicted over forty years ago [ 10 ] and experimentally studied for a decade [ 11 ], are among the few emerging technologies that catalyzed unprecedented worldwide attention due to a variety of applications, including instant turn-on computers, analog memories with continuum states for learning machines, nanoscale memristive synapses [ 12 ], and so on. Device speed, scalability and the ease of fabrication are some of their features [ 13 ]. RSDs have two or more discrete resistive states: in the former case, the high resistance state (HRS) corresponds to the Off state and the low resistance state (LRS) corresponds to the On state; in the latter, apart from the high and low resistance states, it also has intermediate resistance states (IRS) [ 14 , 15 ]. By applying the apt set voltages, the device can be moved from HRS through IRS to LRS. Multilevel RSDs are the most preferred for multilevel storage, thus enhancing the storage density without much change in the technology [ 15 ]. An RSD or a memristor can switch between these states by the application of an appropriate electric stimulus. The transition from Off to On state is known as Set and the transition from On to Off state is known as Reset. The voltage at which the device moves from Off to On is known as the set voltage and the voltage at which the device moves from Off to On state is known as the reset voltage. RSD is said to be non-volatile based on its ability to retain its past resistance state, after the electrical stress is removed; therefore, it represents the best solution for low energy permanent memories. Switching between states can be classified into unipolar and bipolar. The unipolar switching occurs when the reset takes place at a higher current and at a voltage below the set voltage ( Figure 1 a), while the bipolar switching occurs when the set and reset take place at two different polarities ( Figure 1 b) [ 16 ]. The advantages of using RSDs as storage class memories are: small feature size, lower power requirements, higher memory density, and its advantages go on [ 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 ]. The final goal in the field of RSDs is to physically implement an artificial neural network (ANN) for artificial intelligence (AI) purposes [ 22 , 24 , 28 , 29 , 30 , 31 ], through the emulation of biological synapses and biological phenomena such as spike-timing-dependent plasticity (STDP), a mechanism providing enhancement or depression of a specific communication channel based on the synchronization of pulses of the pre-synaptic and post-synaptic neurons [ 32 ]. Enhancing or depressing a response corresponds intuitively to storing or cancelling an information, respectively. We may say that “neurons that fire together wire together” [ 33 ], and add that musical instruments that do not play well in time are silenced by the director. In a similar way, the frequency of propagating pulses in a neuromorphic circuit are shifted by a variable time phase (that could be positive or negative) producing qualitatively a similar programming effect. We refer to an extremely detailed review (and references therein) for a complete screening of the device level engineering, showing how to implement such functionalities using available materials and processes [ 33 ]. In the following sections, we have classified literature in four main sections, highlighting the emergent properties fundamental for WMC applications: flexible, bio-based and biodegradable, stretchable and threaded devices. Of course, stretchable devices are also flexible. Optical transparency has not been considered, as this feature is not important for standard wearables, whose opacity guarantees everybody’s privacy—and, in the case of emergency, eventually safety and visibility. However, we should mention the importance of transparency for military applications, as appointed in the Emerging Security Technologies field [ 34 ], but this is outside the scope of the present review. Regarding the testing of flexible and stretchable of wearable devices, we refer to specific literature [ 35 , 36 ]."
} | 2,708 |
17394648 | PMC1852104 | pmc | 2,300 | {
"abstract": "Background The Archaea are highly diverse in terms of their physiology, metabolism and ecology. Presently, very few molecular characteristics are known that are uniquely shared by either all archaea or the different main groups within archaea. The evolutionary relationships among different groups within the Euryarchaeota branch are also not clearly understood. Results We have carried out comprehensive analyses on each open reading frame (ORFs) in the genomes of 11 archaea (3 Crenarchaeota – Aeropyrum pernix, Pyrobaculum aerophilum and Sulfolobus acidocaldarius ; 8 Euryarchaeota – Pyrococcus abyssi, Methanococcus maripaludis, Methanopyrus kandleri, Methanococcoides burtonii, Halobacterium sp. NCR-1, Haloquadratum walsbyi , Thermoplasma acidophilum and Picrophilus torridus ) to search for proteins that are unique to either all Archaea or for its main subgroups. These studies have identified 1448 proteins or ORFs that are distinctive characteristics of Archaea and its various subgroups and whose homologues are not found in other organisms. Six of these proteins are unique to all Archaea, 10 others are only missing in Nanoarchaeum equitans and a large number of other proteins are specific for various main groups within the Archaea (e.g. Crenarchaeota, Euryarchaeota, Sulfolobales and Desulfurococcales, Halobacteriales, Thermococci, Thermoplasmata, all methanogenic archaea or particular groups of methanogens). Of particular importance is the observation that 31 proteins are uniquely present in virtually all methanogens (including M. kandleri ) and 10 additional proteins are only found in different methanogens as well as A. fulgidus . In contrast, no protein was exclusively shared by various methanogen and any of the Halobacteriales or Thermoplasmatales. These results strongly indicate that all methanogenic archaea form a monophyletic group exclusive of other archaea and that this lineage likely evolved from Archaeoglobus . In addition, 15 proteins that are uniquely shared by M. kandleri and Methanobacteriales suggest a close evolutionary relationship between them. In contrast to the phylogenomics studies, a monophyletic grouping of archaea is not supported by phylogenetic analyses based on protein sequences. Conclusion The identified archaea-specific proteins provide novel molecular markers or signature proteins that are distinctive characteristics of Archaea and all of its major subgroups. The species distributions of these proteins provide novel insights into the evolutionary relationships among different groups within Archaea, particularly regarding the origin of methanogenesis. Most of these proteins are of unknown function and further studies should lead to discovery of novel biochemical and physiological characteristics that are unique to either all archaea or its different subgroups.",
"conclusion": "Conclusion Comparative analyses of sequenced archaeal genomes presented here have led to identification of large numbers of proteins that are distinctive characteristics of either all archaea or its different main groups. Based upon these proteins, all of the main groups within Archaea (e.g. Crenarchaeota, Euryarchaeota, Halobacteria, Thermococci, Thermoplasmata, Methanogens) and their subgroups can now be clearly distinguished in molecular terms. The species distribution of these signature proteins strongly suggests that their genes have evolved or originated at various stages in the evolution of archaea, but once evolved, they are indicated to be generally stably retained in various descendents of these lineages with minimal gene loss or LGTs. Based upon the species distributions of these proteins, the evolutionary stages where the genes for these proteins have likely evolved are shown in Fig. 4 . The evolutionary relationships among archaea have thus far been mainly inferred on the basis of their branching in phylogenetic trees based on 16S rRNA and certain protein sequences [ 2 , 7 , 13 , 23 - 25 ]. The results of our analyses although they support many inferences reached based on phylogenetic trees (viz. identification of all of the main clades in phylogenetic trees in molecular terms) (Fig. 1 ) [ 2 , 7 , 13 , 23 - 25 ], they also differ from them in important regards. In particular, our results shed important light on certain phylogenetic relationships that were very puzzling or were not resolved based on earlier studies. Some of these novel inferences are discussed below. In phylogenetic trees based on 16S rRNA and various proteins sequences, the methanogenic archaea form at least two distinct clusters (see Fig. 1 ) [ 13 , 29 , 34 , 56 , 106 ]. In addition, in many of these trees, M. kandleri branches distinctly from all other methanogenic archaea [ 13 , 34 , 48 ]. The methanogenic archaea in these trees are interspersed by other groups of non-methanogenic archaea such as Halobacteriales, Archaeoglobus, Thermoplasmatales and Thermococcales (see Fig. 1 ) [ 13 , 34 , 48 ]. This has led to important questions concerning the origin of methanogenesis i.e. whether it evolved only once and its absence in the intervening lineages [ 13 , 29 , 35 , 76 ]. To account for these results, it has been suggested that methanogenesis evolved once in a common ancestor of the above groups, i.e. different methanogenic archaea, Halobacteriales, Archaeoglobus, Thermoplasmatales and also possibly Thermococcales, comprising virtually all euryarchaeota, but that the various genes involved in this process were subsequently lost from different groups except the methanogens [ 13 , 29 , 56 ]. This scenario, in essence, proposes that the common ancestor of different physiologically and metabolically distinct groups within euryarchaeota was a methanogen and this capability was independently lost in all other lineages. In contrast to this proposal, our phylogenomics analyses have identified 31 proteins that are uniquely present in virtually all methanogens, as well as many proteins that are specifically shared by different subgroups of methanogens. Of these proteins only about 1/3 are indicated to be directly involved in methanogenesis and the cellular functions of others are presently not known. The unique presence of such large numbers of proteins by nearly all methanogens, but none of the above groups of archaea, strongly indicates that the genes for these proteins evolved in a common ancestor of various methanogens. These results strongly suggest that all methanogenic archaea form a mononphyletic lineage exclusive of all other groups of archaea (Fig. 4 ). Importantly, these studies have also identified 10 proteins that are uniquely shared by all methanogens as well as by A. fulgidus . In contrast, we have not come across any protein that various methanogenic archaea uniquely share with any of the Halobacterales or Thermoplasmatales. These observations are highly significant because they strongly suggest that Archaeoglobus and all of the methanogens shared a common ancestor exclusive of all other archaea. In other words, the ancestral lineage that led to the origin of methanogenesis very likely evolved from the Archaeoglobus lineage (Fig. 4 ). It is also significant that of the proteins that are uniquely shared by Archaeoglobus and methanogens, several form part of complexes that are important for nitrogen assimilation and methanogenesis. These results support the view that these characteristics have their origin within the Archaeoglobus lineage. The present work also provides clarification regarding the phylogenetic position of M. kandleri . In phylogenetic trees based on 16S rRNA or different protein sequences, the branching of this species is highly variable [ 13 , 34 , 47 , 48 ] and it often forms the deepest branch within the Euryarchaeota. In the present work, we have identified 31 proteins that are uniquely shared by all methanogens including M. kandleri , as well as 10 proteins that M. kandleri specifically shares with various Methanobacteriales and Methanococcales, and 15 additional proteins that are only found in M. kandleri and the two Methanobacteriales species ( M. thermoautotrophicus and M. stadtmanae ). These observations reliably place M. kandleri with other methanogenic archaea with the Methanobacteriales as its closest relatives (Fig. 4 ). Our results also suggest a closer relationship of the Thermococcales to the Archaeoglobus and methanogenic archaea, although this relationship is not as strongly supported as between Archaeoglobus and Methanogens. The observed differences in the evolutionary relationships among methanogens based upon phylogenomics analyses versus those by traditional phylogenetic methods can in principle be accounted for by three explanations. First, it is possible that the branching patterns of various clades in phylogenetic trees are misleading and they have been affected by factors such as long branch attraction effect [ 107 , 108 ]. Second, the polyphyletic branching of methanogens can also be explained (as indicated earlier) if the genes uniquely shared by all methanogens evolved in an early branching lineage such as M. kandleri , but subsequently they were either completely or partially lost from various non-methanogenic (viz. Halobacteriales, Thermoplasmatales and Archaeoglobus) groups that lie in between the two methanogenic clusters (Fig. 1 ). Third, lateral transfer of these genes from one methanogenic archaea to all others can also explain these results. Of these possibilities, we favour the first explanation, as the last two require extensive gene loss or LGT from (or into) multiple independent lineages. The present work also supports the placement of N. equitans within the Euryarchaeota lineage. N. equitans has a very small genome (only 0.49 Mb), which is at least 3 times smaller than any other archaeal genome. Due to its very small size, there are only 6 genes that N. equitans uniquely shares with all other archaea. However, our analysis indicates that whereas N. equitans shares a few genes (PAB2404 and PAB 0188) with most of the Euryarchaeota, it does not share any gene uniquely with most of the Crenarchaeota species, indicating its closer affinity for the former lineage. Although our analysis of the N. equitans genome has not revealed any strong signals indicating its specific affinity for any of the Euryarchaeota groups, the shared presence of some proteins by N. equitans and Thermococci (and in some cases also A. fulgidus and methanogens) suggest that it may be related to the Thermococci. However, because of the extensive gene losses that have occurred in this genome, we are not able to draw any reliable inference in this regard. Therefore, although we have depicted N. equitans as a deep branching lineage within Euryarchaeota (Fig. 4 ), based upon our analysis, its placement within Euryarchaeota is not resolved. The present work also suggests that Thermoplasmatales might be a deeper branching lineage within Euryarchaeota in comparison to the Thermococcales, Halobacteriales, Archaoglobous and Methanogens. This inference is suggested by the observation that a number of proteins that are uniquely present in almost all other Euryarcheota species are missing in the Thermoplasmatales. Although the absence of these proteins in the Thermoplasmatales can be explained by specific gene loss, the possibility that the genes for at least some of these proteins have evolved after the branching of Thermoplasmatales deserves serious consideration. The deeper branching of the Thermoplasmatales within the Euryarchaeota will place it closer to the Crenarchaeota. Such a placement could prove helpful in understanding why so many genes (i.e. 30) are uniquely shared by various Thermoplasmatales and the Sulfolobales. For the archaeal-specific proteins identified in the present work, sequence information at present is available from only a limited number of archaeal species. Hence, it is important to obtain information for these genes/proteins from other archaeal species to confirm whether these proteins are distinctive characteristics of the specified groups or a subgroup of such species. These proteins in addition to their utility for phylogenetic and taxonomic studies also provide valuable means for understanding archaeal biology [ 35 , 38 ]. The cellular functions of most of these proteins are not known and further studies in this regard should prove very helpful in the discovery of novel biochemical and physiological characteristics that are unique to either all or different groups of archaea [ 38 ]. Lastly, the primary sequences of many of these genes/proteins are also highly conserved and they provide novel means for identification of different groups of archaea in various environmental settings by means of PCR amplification and other molecular biological and immunological methods.",
"discussion": "Results and discussion A. Phylogenetic analyses of archaeal species Prior to undertaking comparative studies on archaeal genomes, phylogenetic analysis of sequenced archaeal species was carried out so that the results of phylogenomics analyses could be compared with those obtained by traditional phylogenetic approaches. Phylogenetic trees for the archaeal species based on 16S rRNA as well as concatenated sequences of translation and transcription-related proteins have been published by other investigators [ 7 , 28 , 32 , 44 ]. In the present work, we have constructed phylogenetic trees for 29 archaeal species (see Table 1 ) using a set of 31 universally distributed proteins that are involved in a broad range of functions [ 45 ]. The sequence of Haloquadratum walsbyi DSM 16790, which became available afterward, was not included in these studies. Phylogenetic trees based on a concatenated sequence alignment of these proteins were constructed using the neighbour-joining (NJ), maximum-likelihood (ML) and maximum-parsimony (MP) methods. The results of these analyses are presented in Fig. 1 . All three methods gave very similar tree topologies except for the branching positions of M. kandleri and Methanospirillum hungatei , which were found to be variable. Except for this, the branching pattern of the archaeal species based on our dataset is very similar to that reported by Gribaldo et al. [ 13 , 32 ] based on concatenated sequences of translation and transcription-related proteins. In the tree shown, the Crenarchaeota and Euryarchaeota , the two major phyla within Archaea were clearly distinguished from each other. The phylogenetic affinity of Nanoarchaeum , which has a long-branch length, was not resolved in this or various other trees [ 32 , 46 ]. Within Crenarchaeota , Pyrobaculum was indicated to be a deeper branch, and Aeropyrum branched in between the Pyrobaculum and Sulfolobus . Within Euryarchaeota , the clades corresponding to Halobacteria, Thermococci and Thermoplasmata were resolved with high bootstrap scores, but the methanogens were split into 2–3 clusters. One of these clusters that has low bootstrap score consisted of Methanobacteriales and Methanococcales with M. kandleri ( Methanopyrales ) branching in its vicinity [ 34 , 47 , 48 ]. The second cluster, with higher bootstrap score, showed a grouping of Methanomicrobiales and Methanosarcinales . These two clusters, which are separated by Thermoplasmata, Archaeoglobi and Halobacteria , have been referred to as Class I and Class II methanogens by Bapteste et al. [ 29 ]. B. Phylogenomic analyses of archaeal genomes To search for proteins (or ORFs), which are uniquely present in either all Archaea or various subgroups of them, blast searches were performed on each open reading frame (ORF) from a total of 11 archaeal genomes (see Table 1 ; shaded species in Fig. 1 ). These genomes included 3 Crenarchaeota (viz. Aeropyrum pernix, Pyrobaculum aerophilum and Sulfolobus acidocaldarius ) [ 49 - 51 ] and 8 divergent Euryarchaeota species covering all main functional and phylogenetic groups (see Table 1 and Fig. 1 ). The Euryarchaeota genomes analyzed included: Pyrococcus abyssi from extremely thermophilic sulfur metabolizing archaea [ 52 ], Methanococcus maripaludis [ 53 ] from Methanococcales, Halobacterium sp. NRC-1 and H. walsbyi from extreme halophiles [ 54 ], Thermoplasma acidophilum and Picrophilus torridus belonging to the cell wall-less archaea [ 19 , 55 ], Methanococcoides burtonii from Methanosarcinales and Methanopyrus kandleri from the Methanopyrales order [ 56 ]. The chosen genomes should provide information regarding all archaeal proteins that are shared at a taxonomic level higher than a genus. The analysis of the remainder of the genomes, which was expected to provide information regarding proteins that are only unique to a given species, was not carried out. Each ORF from these genomes was examined by means of blastp and PSI-blast searches against all available sequences from different organisms to identify proteins that are specific for only archaeal lineages. The methods and the criteria that we have used to identify proteins that are specific for either all or various subgroups of archaea are described in the Methods section. Generally, a protein was considered to be specific for a given archaeal lineage if all significant hits or alignments in the blastp and PSI-blast searches with the query protein were from the indicated group of archaeal species. In a few cases, where 1–2 isolated species from other groups also exhibited significant similarity, such proteins were retained as they provide interesting examples of lateral gene transfer (LGT) from archaea to other groups. Our analyses have identified 1448 proteins that are unique to different groups of Archaea and for which no homologues are generally found in any bacterial or eukaryotic species. Based on their specificity for different taxonomic groups, these proteins have been divided into a number of different groups (see Tables 2 , 3 , 4 , 5 , 6 , 7 and Additional files). A brief description of the different subsets of archaeal-specific proteins and functional information regarding them, where known, is given below. In the description of these proteins that follows, the 'APE', 'HQ', 'Mbu', 'MK', 'MMP', 'PAB', 'PAE', 'PTO', 'Saci', 'Ta', 'VNG', and 'NEQ' part of the descriptors in proteins indicate that the original query protein sequence was from the genome of A. pernix K1, H. walsbyi DSM 16790, M. burtonii DSM 6242, M. kandleri AV19, M. maripaludis S2, P. abyssi GE5, P. aerophilum str. IM2, P. torridus DSM 9790, S. acidocaldarius DSM 639, T. acidophilum DSM 1728, Halobacterium sp. NRC-1 and N. equitans , respectively. (a) Proteins that are specific for all Archaea Table 2(a) shows a group of 16 proteins that are present in nearly all archaeal species but whose homologues are not found in any Bacteria or Eukaryotes with a single exception. Of these, the first 6 proteins in the left column (Table 2a ) viz. PAB0063, PAB0252, PAB0316, PAB1633, PAB1716 and PAB2291, are present in all sequenced archaeal genomes. The observed E-values for these proteins from archaeal species are very low, close to 0, indicating that these proteins show very high degree of sequence conservation in various archaea. The unique presence of these proteins in all sequenced archaeal genomes indicates that these proteins could be regarded as distinctive characteristics or molecular signatures for the archaeal domain. The genes for these proteins likely evolved in a common ancestor of the Archaea and were then vertically acquired by other archaeal species. Makarova and Koonin [ 35 ] have also mentioned 6 proteins that are commonly shared by different archaea, but the identity of such proteins was not specified. These proteins are likely the same. The remaining 10 proteins in Table 2(a) are missing only in N. equitans , which is a tiny parasitic organism containing only 536 genes [ 57 , 58 ]. The species distribution pattern of these proteins can be accounted for by one of the following two possibilities. First, it is possible that N. equitans is the deepest branching lineage within archaea, as has been suggested [ 57 , 58 ] and the genes for these 10 proteins evolved in a common ancestor of the other archaea after its divergence (Fig. 2a ). Alternatively, similar to the first 6 proteins, the genes for these 10 proteins evolved in a common ancestor of all archaea, but they were then selectively lost in N. equitans (Fig. 2b ) [ 35 , 46 , 58 ]. Based upon our results, one cannot distinguish between these two possibilities. However, in view of the fact that the genome of N. equitans has undergone extensive genome shrinkage (only 0.49 Mb) and it is at least 3 times smaller than the next smallest archaeal genome (see Table 1 ), we favour the latter possibility (Fig. 2b ) [ 35 , 46 , 58 ]. Of the proteins that are uniquely present in all archaea, PAB0063 corresponds to tRNA nucleotidyltransferase (CCA-adding enzyme), which builds and repairs the 3' end of tRNA [ 59 ]. Functionally similar enzymes are also present in bacteria and eukaryotes (assigned as Class II), but their sequences share very little homology with the archaeal CCA-adding enzyme (Class I), which explains why no homologs were detected in any bacteria or eukaryotes in blast searches. The main mechanistic difference between class I and class II enzymes is that the tRNA substrate is required to fully define the nucleotide binding site in class I enzyme, whereas class II has a preformed nucleotide binding site that recognizes CTP and ATP in the absence of tRNA [ 60 ]. Another protein PAB0316 is assigned as archaeal type DNA primase, which also has its synonymous counterparts in bacterial and eukaryotic species, but shows very little homology to them [ 61 , 62 ]. In the same way, protein PAB1633 is annotated as a PilT family ATPase, which showed very little similarity to bacterial ATPases involved in type IV pili biogenesis [ 54 ]. Further studies of this protein could provide insights into novel aspects of the archaeal flagellar system. A number of other proteins viz. PAB1716, PAB0018a, PAB0075, PAB0475 and PAB2104, have also been assigned putative functions based on sequence analysis, but their exact roles in archaeal cells remains to be determined. Interestingly, for protein PAB0075, two gene copies with acceptable E-values are also present in the genomes of Dehalococcoides ethenogenes 195, Dehalococcoides sp. CBDB1 and Dehalococcoides sp. BAV1, which belong to Chloroflexi [ 2 ]. Because no homologue of PAB0075 is present in other bacteria, it is likely that this protein was transferred from archaea to the common ancestor of Dehalococcoides followed by a gene duplication event. Table 2(b) lists 20 additional proteins, which are specific to archaea but missing in a small number of species. Because these proteins are present in most Euryarchaeota as well as Crenarchaeota species, but not detected in Bacteria or Eukaryotes except one LGT case (PAB2342, see note in Table 2 ), we consider them also to be distinctive characteristics of most Archaea. Of these proteins, 11 proteins (viz. PAB0654, PAB0950, PAB1135, PAB1906, PAB7388, PAB0547, PAB0552, PAB0623, PAB1272, PAB1429 and PAB1721) are mainly missing in the 4 Thermoplasmata species. Thermoplasmata are thermoacidophilic archaea which lack cell envelope [ 19 , 55 , 63 ](see Table 1 ). Some studies have suggested that high temperature and very low intracellular pH exert selective pressure favouring smaller genomes [ 19 ]. Thus, it is possible that genes for these proteins were selectively lost in the Thermoplasmata lineage. Most of these proteins are of unknown function. However, 8 of them have been assigned putative functions with the title of \"archaeal type\"'. For example, PAB0301 is archaeal sugar kinase, PAB0950 is archaeal transcription factor E α-subunit, PAB1387 is archaeal flagella accessory protein, PAB7094 is archaeal chromatin protein, and PAB0552 is archaeal type Holliday junction resolvase. These proteins do not show detectable sequence similarity to their counterparts in Bacteria or Eukaryotes, and some studies indicate that they also differ in terms of their structure, function or interaction with other cell components [ 64 , 65 ]. (b) Proteins that are specific for Crenarchaeota As mentioned in the introduction, the Archaea are divided into 2 main groups, Crenarchaeota and Euryarchaeota, based on 16S rRNA trees as well many other gene trees and characteristics. The Crenarchaeota are also indicated to differ from Euryarchaeota in terms of their ribosome structure [ 30 , 31 ]. In comparison to Euryarchaeota, which contain physiologically and metabolically diverse groups of organisms, the Crenarchaeota were thought to be a pure collection of extreme thermophiles and most members metabolize sulfur. However, recent studies indicate that Crenarchaeota are much more diverse in their physiology and ecology than was previously believed [ 28 , 66 ]. Many species living in the cold ocean also belong to this group based on their branching pattern in 16S rRNA trees, although most of them have not been cultivated [ 67 ]. Currently, this phylum is comprised of one single class Thermoprotei containing three orders: Thermoproteales, Desulfurococcales and Sulfolobales. Fortunately, every order has a completely sequenced representative (see Table 1 )[ 50 , 51 , 68 , 69 ], which provide a platform to explore the characteristics that are unique to crenarchaeal species. Comparative genomic surveys have revealed some molecular features that are shared by crenarchaea but not euryarchaea, such as the lack of histones, absence of the FtsZ-MinCDE system and distinctive rRNA operon organization [ 69 ]. Lake et al. have also identified distinctive differences in ribosome structure and an insert in elongation factor EF-G and EF-Tu, which can be used to distinguish Crenarchaeota from Euryarchaeota [ 6 , 30 , 70 ]. However, these features are not unique characteristics of the Crenarchaeota. Blast searches on each ORF from the genomes of A. pernix and S. acidocaldarius DSM 639 [ 49 , 50 ] have identified 11 proteins which are shared by all five crenarchaeal species, but whose homologs are not found in other archaea, or any bacteria or eukaryotes with only 3 exceptions (see Table 3(a) ). The genes for these proteins likely evolved in a common ancestor of the Crenarchaeota and they provide potential molecular markers for species from this phylum. Additionally, 22 proteins that are listed in Table 3(b) are only found in A. pernix and three Sulfolobus genomes. These proteins suggest that Aeropyrum and Sulfolobus may have shared a common ancestor exclusive of Pyrobaculum . However, we have also come across 9 proteins that are shared by Aeropyrum and Pyrobaculum (Table 3(c) ) and 14 proteins that are exclusively present in the 3 Sulfolobus species and Pyrobaculum (see Table 3(d) ). Hence, based upon the species distributions of these proteins, the relationships among the Aeropyrum , Sulfolobales and Pyrobaculum are not entirely clear (Fig. 2a ). In phylogenetic trees Thermoproteales (i.e. Pyrobaculum ) branches consistently earlier than Desulfurococcales (i.e. Aeropyrum ) and Sulfolobales (Fig. 1 ) [ 32 , 44 ]. This observation in conjunction with the fact that Aeropyrum and Sulfolobus share larger numbers of proteins in common with each other suggests that these two groups likely shared a common ancestor exclusive of Pyrobaculum (Fig. 2b ). The proteins that are only found in Aeropyrum and Pyrobaculum , or in Sulfolobus and Pyrobaculum , most likely evolved in a common ancestor of the crenarchaea, but were subsequently lost in either the Sulfolobales or A. pernix lineages. In addition to these proteins that are uniquely present in either all sequenced Crenarchaeota genomes or different groups of Crenarchaeota species, these analyses have also identified 264 proteins that are unique for the Sulfolobales species (see Additional file 1 ). Of these, 184 proteins are present in all 3 sequenced Sulfolobus genomes, whereas the remaining 80 are present in at least two of the three Sulfolobus genomes. In this work, since blast analyses were not carried out on all three Sulfolobus genomes, it is likely that the numbers of genes or proteins that are uniquely shared by only two Sulfolobus genomes is much higher than indicated here. Chen et al. [ 50 ] have previously analyzed the genome of S. acidocaldarius DSM 639 and indicated the presence of 107 genes that were specific for Crenarchaeota and 866 genes that were specific to Sulfolobus genus. However, in the present work, relatively few genes that are uniquely shared by various Crenarchaeota species were identified. This difference could be due to more stringent criteria that we have employed for identification of proteins that are specific to different groups. The genome of Thermofilum pendens Hrk 5, which belongs to Thermoproteales, has also been partially sequenced and information for large numbers of genes/proteins from this species is available in the NCBI database. By carrying out blast searches on each ORF from P. aerophilum genome [ 51 ], we have identified 42 proteins that are only found in the above 2 Thermoproteales species (see Additional file 2 ). The numbers of proteins shared by these two species will likely increase once complete genome of T. pendens becomes available. Many of these proteins are expected to provide markers for the Thermoproteales order. (c) Proteins that are specific for Euryarchaeota The Euryarchaeota, which comprise a majority of the cultured and sequenced archaea, is a morphologically, metabolically and physiologically diverse collection of species as evidenced by the presence in this group of various methanogens, extreme halophiles, cell wall-less archaea and sulfate reducing microbes [ 2 , 13 ]. No unique biochemical or molecular characteristic that is commonly shared by all of the different lineages is known. The present study has identified 20 proteins that are only found in Euryarchaeota species with 3 exceptions (see Table 4 ). In this Table, the first 7 proteins (Table 4(a) ) are present in most euryarchaeota species. Of these proteins, PAB0082 and PAB2404 were found in all sequenced euryarchaeota species. PAB2404 was also present in N. equitans , supporting its placement within the Euryarchaeota [ 35 , 46 ]. The protein PAB0082 is annotated as archaeosine tRNA-ribosyltransferase (ArcTGT), which catalyzes the exchange of guanine with a free 7-cyano-7-deazaguanine (preQ 0 ) base, as the first step in the biosynthesis of an archaea-specific modified base, archaeosine (7-formamidino-7-deazaguanosine) [ 71 ]. It should be mentioned that there is another protein PAB0740 in the same genome, which is also annotated and experimentally confirmed as ArcTGT [ 72 ]. The latter belongs to a family of proteins that are highly conserved in all archaea species (including Crenarchaeota) and some bacteria. It seems that PAB0082 might be involved in RNA modification since it possesses a PUA domain (named after pseudouridine synthase and archaeosine transglycosylase), but its function is likely different from PAB0740. The protein PAB2404, which is annotated as DNA polymerase II large subunit, is highly conserved within Euryarchaeota, but is not found anywhere else except in Nanoarchaeum . This enzyme is the major DNA replicase in Euryarchaeota and also a distinctive molecular marker for this group [ 73 , 74 ]. The genes for the above proteins likely evolved in a common ancestor of Euryarchaeota (Fig. 2 ) and they provide molecular markers for this diverse group of organisms. Another 13 proteins listed in Table 4(b) are found in almost all euryarchaeota, but they are missing in Thermoplasmata. Their distribution suggests that either Thermoplasmata is a deep branching lineage within Euryarchaeota or that the genes for these proteins have been selectively lost from Thermoplasmata [ 55 ]. Of these proteins, PAB0188 is also present in N. equitans supporting its placement with Euryarchaeota. Five other proteins from the first two columns in Table 4 (viz. MMP0243, Ta0062, VNG1263c, MMP1287, and VNG2408c) are also not found in the 4 Thermococci species. These results can again be explained by either selective loss of these genes from these particular groups or deeper branching of these lineages within the Euryarchaeota species. On the basis of proteins listed in Table 4 , although one can infer that Thermoplasmata and Thermococci are deeper branching lineages within Euryarchaeota in comparison to methanogens, their relative branching order cannot be resolved. (d) Proteins that are specific for different main groups within Euryarchaeota Proteins specific for methanogenic archaea and their various subgroups Currently, the methanogens form the largest group within the Euryarchaeota. They are distinguished from all other prokaryotes by their ability to obtain all or most of their energy via the reduction of CO 2 to methane or by the process of methanogenesis. In the Bergey's manual [ 75 ], the methanogenes are divided into 5 distinct orders (viz. Methanobacteriales, Methanococcales, Methanomicrobiales, Methanosarcinales and Methanopyrales). Some studies have suggested that these organisms possess a set of unique enzymes which are responsible for methanogenesis, such as coenzyme M, Factor 420 and methanopterin [ 76 ]. However, no systematic study has been carried out thus far to identify proteins that are uniquely present in different methanogens. Our blast searches of proteins from different methanogens have led to identification of 31 proteins, which are uniquely found in various methanogenic archaea. Twenty of these 31 proteins are present in all sequenced methanogens, while 11 proteins are missing only in M. stadtmanae , which is a human intestinal inhabitant (see notes in Table 5 ). This archaeon generate methane by reduction of methanol with H 2 and lacks many proteins present in the genomes of other methanogens [ 77 , 78 ]. Thus, it is highly likely that the 11 proteins missing in M. stadtmanae were selectively lost from this species. Therefore, it is very likely that the genes for these 31 proteins that are commonly shared by virtually all methanogens (Table 5(a) ) evolved in a common ancestor of all methanogens. These analyses have also identified 10 proteins that are uniquely shared by various methanogens as well as A. fulgidus (see Table 5(b) ). The genes for these proteins likely evolved in a common ancestor of A. fulgidus and various methanogenic archaea and they point to a close relationship between these two groups of organisms (Fig. 3 ). Ten additional proteins are present in A. fulgidus as well as various Methanosarcinales and M. hungatei (Methanomicrobiales) (Table 5(c) ). It is likely that the genes for these proteins also evolved in a common ancestor of A. fulgidus and various methanogenic archaea, but they were selectively lost in other methanogens. Of the proteins that are commonly shared by A. fulgidus and various methanogenic archaea, MMP0607 is reported to be a novel repressor of nif and glnA genes, which are involved in nitrogen assimilation [ 79 ]. Interestingly, 2 homologs of this protein are also found in 3 Dehalococcoides species, but nowhere else, which are very likely due to LGT. Protein MMP0984 is the ε-subunit of carbon-monoxide dehydrogenase complex, which is made up of five subunits in different methanogens [ 80 ]. The epsilon subunits are required for the reversible oxidation of CO to CO 2 [ 81 ]. All of the other components could be found in a few bacterial species, while the ε-subunit is restricted to methanogenic archaea and A. fulgidus [ 82 , 83 ]. Protein MMP1499 is identified as a transcriptional regulator with a Helix-turn-helix (HTH) motif, but its exact role has not been reported. Among the genes that are uniquely shared by various methanogenic archaea (or these archaea plus A. fulgidus ), two large gene clusters responsible for methanogenesis are found. The proteins MMP1346, MMP1560–MMP1564 and MMP1566–MMP1567 (Table 5 ) are parts of an eight-component complex, coenzyme M methyltransferase (Mtr), which catalyzes an energy-conserving, sodium-ion-translocating step in methanogenesis from H 2 and CO 2 [ 84 ]. M. maripaludis contains all of the known Mtr subunits, but the gene coding for MtrF is fused into the N-terminal region of MtrA [ 53 ]. All other methanogenic archaeal genomes contain complete set of mtr genes. It is of interest to note that for the protein MMP1567 (MtrH), homologues with low E-values are also found in two Desulfitobacterium hafniense strains as well as in three Rhizobiales species ( Aminobacter lissarensis, Methylobacterium chloromethanicum , and Hyphomicrobium chloromethanicum ; α-proteobacteria) (see note in Table 5 ). These three rhizobiae species can use methyl halides as a sole source of carbon and energy, and all of them possess a set of cmu genes which are essential for methyl chloride degradation [ 85 ]. In particular, the CmuB protein which is homologous to MMP1567 transfers a methyl group to methylcobalamin:H 4 folate (H 4 F), which is analogous to the reverse of the reaction catalyzed by MtrH in archaea [ 86 ]. In view of the sequence and functional similarity between MtrH and CmuB proteins, it is likely that the mtrH gene was laterally transferred from a methanogenic archaeon to the common ancestor of the above three rhizobiae species to serve the new functional role. The function of the laterally transferred mtrH related gene in D. hafniense is not known at present. The proteins MMP1555–MMP1559 in Table 5 form another gene cluster, encoding the subunits of Methyl-coenzyme M reductase (MCR). This complex catalyzes the final reaction of the energy conserving pathway in which methylcoenzyme M and coenzyme B are converted to methane and the heterodisulfide CoM-S-S-CoB [ 87 , 88 ]. Except for these proteins, the other proteins listed in Table 5 are of putative or unknown functions. It is likely that these proteins are involved in some aspects of methanogenesis or other unknown pathways unique to methanogenic archaea. These proteins provide molecular markers for methanogens, which can be used for identification of new archaeal species capable of methane production. The blast searches of the M. maripaludis [ 53 ] and M. kandleri [ 56 ] genomes have identified 10 proteins that are uniquely shared by all of the following species belonging to the orders Methanobacteriales ( M. thermoautotrophicus ), Methanococcales ( M. jannaschii , M. maripaludis ) and Methanopyrales ( M. kandleri ) (Table 6(b) ). Of these, only 2 proteins are present in M. stadtmanae , which is also a Methanobacteriales that has lost most of its genes due to its adaptation to the human intestine [ 78 ]. The genes for these 10 proteins likely evolved in a common ancestor of the above groups of methanogens (Fig. 3 ), which corresponds to the cluster of methanogenic archaea referred to as \"Class I methanogens\" [ 13 ]. Interestingly, these studies have also identified 10 proteins that are uniquely shared by these methanogenic orders and M. hungatei (see Table 6(a) ), which branches distantly in phylogenetic trees [ 13 ]. The unique presence of these proteins in these methanogens suggests that species from these groups shared a common ancestor exclusive of other methanogenic archaea (Fig. 3 ). Fifteen additional proteins discovered in this work (Table 6(c) ) are uniquely present in M. kandleri and various Methanobacteriales indicating that these two groups are more closely related to each other than the Methanococcales (Fig. 3 ). We have also come across 7 proteins that are uniquely shared by Methanococcales and Methanobacteriales (Table 6(d) ), and 4 proteins that are only present in Methanococcales and Methanopyrales (Table 6(e) ). The most likely explanation to account for the species distributions of these latter proteins is that their genes also originated in a common ancestor of the above three groups of methanogens, but were selectively lost in either the Methanobacteriales or Methanopyrales lineages. These analyses have also identified 14 additional proteins that are uniquely present in all 5 Methanosarcinales species (Table 6(f) ), as well as 7 proteins that are only found in various Methanosarcinales and M. hungatei (Table 6(g) ). Lastly, these studies have also identified 55 proteins that are uniquely present in M. maripaludis and M. jannaschii (Methanococcales, see Additional file 3(a) ) and 68 proteins that are only present in M. burtonii and 3 Methanosarcina species, all belonging to the Methanosarcinaceae family (see Additional file 3(b) ) (Fig. 3 ) indicating that they are likely distinctive characteristics of species from these groups. Of the proteins that are uniquely found in Methanococcales, Methanobacteriales, Methanopyrales and Methanomicrobiales, 12 proteins viz. MMP1448–MMP1454, MMP1456, MMP1458–MMP1460 and MMP1467 are from a big gene cluster eha , which encodes the multisubunit membrane-bound [Ni-Fe] hydrogenase [ 89 ]. Two of these proteins, MMP1456 and MMP1458, are only found in Methanococcales (Table 6(e) ). The whole eha operon is composed of 20 ORFs in the genome of M. thermoautotrophicus and of these only these 12 proteins are restricted to these methanogens while the other subunits have counterparts in bacteria. The precise roles of these 12 proteins, which are predicted to be integral membrane proteins in the hydrogenase complex, have not been determined [ 89 ]. Among the other proteins that are specific for these groups of methanogens, MMP0127 and MMP1716 are Hmd homologs, which catalyze the reversible dehydrogenation of N 5 , N 10 -methylenetetrahydromethanopterin [ 90 ]. In the proteins that are specific for the Methanococcales (see Additional file 3(a) ), one large gene cluster (MMP0233–MMP0240) is found, but no information is available concerning its possible function. Except for these proteins, all other proteins that are specific for these methanogenic archaea are of unknown or putative function. Proteins that are specific for Thermococci Thermococci are obligately thermophilic, strictly anaerobic cocci, which are able to convert elemental sulfur to hydrogen sulfide. Thus, they are so called \"extremely thermophilic sulfur metabolizer\", which comprise one of the main functional groups within Euryarchaeota. According to the Bergey's Manual [ 75 ], the class Thermococci contains only one family, Thermococcaceae, consisting of 2 genera: Thermococcus and Pyrococcus . Currently, 4 species from this family have been completely sequenced ( Pyrococcus abyssi, P. horikoshii, P. furiosus and Thermococcus kodakarensis ; see Table 1 ) [ 52 , 91 - 93 ]. The blast searches on each protein from P. abyssi have identified 141 proteins that are shared by all 4 of these species (see Additional file 4(a) ). All of these proteins show high degree of conservation within Thermococci and they do not have homologs in any other prokaryotes or eukaryotes except one possible LGT event (PAB1493, see note in Additional file 4 ). The genes for these proteins have likely evolved in a common ancestor of the Thermococci (Fig. 3 ). Of these proteins, PAB1510 is annotated as TBP-interacting protein (TIP), which forms complex with TBP (TATA-binding protein) to regulate transcription [ 94 ]. It is known that the archaeal transcription machinery is strikingly similar to that in eukaryotes [ 23 ], but no TBP-binding component was found in archaeal species until the discovery of the TIP in T. kodakaraensis [ 95 , 96 ]. Most other Themococci-specific proteins are of unknown function, although in a few cases limited similarity to domains in known protein families have been noted. A number of proteins (viz. PAB0643–PAB0644.1n; PAB1821–PAB1826) are clustered together in the P. abyssi genome, and it is possible that they may form functional units and are involved in related functions. Cohen et al. [ 52 ] have reported a large number of proteins which are restricted to the Pyrococcus genus. However, a number of proteins from their list are also found in T. kodakarensis KOD1 [ 93 ], whose genome was not available when their work was published. Some proteins are not specific for either Pyrococcus or Thermococci according to our criteria and some of them are only found in one species – P. abyssi . Our analysis of the P. abyssi GE5 genome has also identified 43 proteins that are unique to the Pyrococcus genus (see Additional file 4(b) ). Again, almost all of these proteins are of unknown function except PAB2241, which is annotated as RNase P, but this annotation seems arbitrary as it does not show significant sequence similarity to known RNases. The proteins that are uniquely found in the 3 Pyrococcus genomes likely evolved in a common ancestor of this genus (Fig. 4 ). Proteins that are specific for Halobacteria Extreme halophiles constitute another major class within Euryarchaeota. They require 5–10 times the salinity of seawater (ca. 3–5 M NaCl) for optimal growth [ 17 , 97 ]. In order to grow in such high salinity environments, they have developed a set of physiological adaptation, such as: high internal concentration of potassium chloride, acidic proteome with low pI value, high GC content with GC bias in the wobble position, unique chloride pumps to maintain osmotic balance, etc. [ 17 , 98 , 99 ]. Among archaea, halobacteria also have the unique ability to use solar energy to generate a proton gradient to synthesize ATP. So far, the Class Halobacteria harbors one family with 15 genera and 4 species have been completely sequenced, including Halobacterium sp. NRC-1, Haloarcula marismortui , H. walsbyi and Natronomonas pharaonis [ 54 , 98 , 100 , 101 ]. By performing blast searches on each protein in the Halobacterium sp. NRC-1 genome, we have identified 127 proteins, which are only present in all 4 Halobacteria species with only 3 exceptions (see Additional file 5 ). Of the proteins listed in this Table, VNG0016H, VNG1096H, VNG2414H and VNG2563H are annotated as DNA-binding proteins or regulators because of the presence of HTH domain, but their exact functions have not been reported. VNG0667G is an ATP-binding protein of ABC transporter family. Several other proteins, such as VNG2089H and VNG2628H, have also been assigned possible functions based on weak similarity to known conserved domains in the CDD database [ 102 ], but their exact functions remain to be determined. Because of their high degree of conservation and uniqueness to halobacteria, the genes for these proteins likely evolved in a common ancestor of Halobacteria (Fig. 4 ) and they are presumably involved in unique physiological functions related to their adaptation to the hypersaline environment. Because of their specificity for Halobacteria, these proteins provide useful biomarkers for this group of species. In addition to these proteins that are specific to all sequenced halobacterial species, we have also identified a large number of proteins either uniquely shared by 3 halobacterial species or only found in 2 halobacterial species (see Additional files 6 and 7 ). Surprisingly, these proteins are present in different combinations of halobacterial species. The four-halobacterial species are from 4 different genera within the Halobacteriales order and their relationships are unclear at present. The largest numbers of these proteins (i.e. 56) are uniquely shared by the Haloarcula, Haloquadratum and Natronomonas species, followed by 49 proteins that are restricted to Haloquadratum and Haloarcula . These results suggest that of these three species, Haloquadratum and Haloarcula are more closely related to each other and that Halobacterium might be the deepest branching of the four available halobacterial species (Fig. 4 ). However, the genome size of these halobacterial species varies and some of these protein sequences are present on plasmids found in these species, which makes it difficult to reliably infer their relationships solely based on the number of shared proteins. Among the proteins that are specific for halobacteria, only few have been assigned possible functions. Protein VNG2178H is annotated as PhiH1-like repressor and VNG0584H is assigned as a Rieske Fe-S protein. Two additional proteins VNG1720H and VNG2562H have been annotated as iron-binding proteins because of their similarity with FhuD and TroA_a domains, respectively [ 102 ]. All of the other proteins are of unknown function. Proteins that are specific for Thermoplasmata The Thermoplasmata group is comprised of cell wall-less archaea, which resemble the bacterial Mycoplasma species [ 63 ]. Generally, they are thermoacidophilic, aerobic or facultative anaerobic, and are able to reduce sulfur to H 2 S under anaerobic conditions [ 19 , 55 ]. To date, this class include three families-Thermoplasmaceae, Picrophilaceae, and Ferroplasmaceae, each represented by one genus [ 103 , 104 ]. Three complete genomes from this class ( T. acidophilum, T. volcanium and P. torridus ) are available at present (see Table 1 ) [ 19 , 55 , 63 ] and Ferroplasma acidarmanus Fer1 genome is draft assembled and sequence information for this is also available in the NCBI database. Our analyses have uncovered 77 proteins that are uniquely present in all four species belonging to this class (see Additional file 8(a) ) (Fig. 4 ). Most of these proteins are present in all four available genomes, but a few are missing in one or two species, which is probably due to gene loss. Besides, we have also identified 33 proteins, which are shared only by the two Thermoplasma species (see Additional file 8(b) ) and 17 proteins unique to P. torridus and F. acidarmanus (see Additional file 8(c) ). The latter proteins indicate that species from Picrophilaceae and Ferroplasmaceae families are more closely related to each other (Fig. 4 ). All of these proteins are of unknown or predicted functions. Proteins restricted to several archaeal lineages or showing sporadic distribution In addition to the above proteins that are restricted to specific lineages of archaea, we have also identified 63 proteins, which are shared by several archaeal groups (see Table 7 ). The distribution pattern of these proteins could provide useful insights concerning the phylogenetic relationship between different groups. However, their distribution patterns could also be explained by gene losses in specific lineages or LGT between particular groups. Table 7 shows many proteins that are uniquely shared by various methanogenic archaea, Archaeoglobus and Thermococci. The first 5 proteins in Table 7(a) (PAB0076, PAB0138, PAB0965, PAB1927 and PAB1994) are present in all of the Thermococci and most of the methanogens. Four of these proteins are also present in A. fulgidus . The next 13 proteins in this Table are also uniquely found in most of the Thermococci as well as a number of methanogens and also in many cases in A. fulgidus . In addition, 6 proteins listed in Table 7(b) are only found in various Thermococci and A. fulgidus . These results suggest a closer relationship between the methanogenic archaea, A. fulgidus and Thermococci within the Euryarchaeota lineage. In conjunction with our earlier inference that A. fulgidus forms an outgroup of the methanogenic archaea, these results suggest that the above three groups are related in the following manner: Thermococci → A. fulgidus → Methanogens. Although the relationship suggested above is the most likely explanation for the observed results, we have also come across three proteins (VNG1263c, MMP11287 and VNG2408c) that are uniquely present in various Halobacteria, A. fulgidus and different methanogens. To account for their species distribution, one has to postulate that their genes have been selectively lost from the Thermococci. In addition, 9 proteins are only found in various Halobacteria as well as Methanosarcinales and Methanomicrobiales (Table 7(c) ). Their distribution requires again either selective gene losses from other lineages or LGT from Halobacteria to these methanogens. Our analyses have also uncovered 30 proteins that are uniquely shared by species from Thermoplasmata and Sulfolobus (see Table 7(d) ). Among these proteins, 7 are present in all Thermoplasmata and Sulfolobus species for which sequence information is available, while the remainder are missing in 1 or more species. It has been reported that there has been much lateral gene transfer between T. acidophilum and S. solfataricus , both of which inhabit the same environment [ 55 ]. However, the shared presence of these proteins in these two groups could also result from a unique shared ancestry of these thermo-acidophilic archaea. Another 43 Archaea-specific proteins are sporadically present in different archaeal species (see Additional file 9 ). A number of proteins in this group are present in a limited number (between 3 to 6) of archaeal species belonging to different groups. There are 2 possible explanations that can account for their sporadic distribution: First, it is possible that some of these genes are the remnants of sequences that also originated in an ancestral lineage of Archaea but they have been selectively lost in many species because they are not required for growth. Second, the sporadic presence of these genes in a number of archaeal species can also be explained if some of these genes originally evolved in a particular group or species of archaea and then transferred to other archaea by LGT [ 105 ]. However, in view of the observed specificity of these genes/proteins for archaea, the LGTs in these cases need to be selective and limited to within archaea."
} | 13,540 |
32566852 | PMC7301548 | pmc | 2,305 | {
"abstract": "Wettability is an important surface property owing to\nits useful characteristics such as self-cleaning, antifrosting, and\nanticorrosion. In particular, an oleophobic surface, which can overcome\nthe limitation of the antifouling performance of a hydrophobic surface,\nis a considerably valuable research subject. Magnesium alloys are\nwidely used in various industrial fields owing to their superior mechanical\nperformance; however, a technology that is applicable for surface\nmodification has been limited due to their chemical properties. In\nthis study, a new method to prepare a highly oleophobic magnesium\nalloy AZ31 surface is introduced; this method involves applying a\nhierarchical structure and fluorination. The hierarchical structure\nwas formed via two-step anodization and magnesium hydroxide formation,\nand a self-assembled monolayer (SAM) coating method was applied to\nfluorinate the surface. This hierarchical structure with low surface\nenergy can reduce the contact area between the surface and droplets,\nthereby decreasing the adhesive force. Contact angles were measured\nusing various test liquids to evaluate the oleophobic surface, and\nall test liquids, including rapeseed oil (35.0 mN/m), were repelled\nby the surface.",
"conclusion": "3 Conclusions In this study, we developed\na method to impart highly oleophobic properties to the surface of\nmagnesium alloy AZ31. This was achieved by two-step anodization, magnesium\nhydroxide formation, and surface fluorination. A two-scale hierarchical\nstructure was formed by creating a small-scale magnesium hydroxide\nstructure on a large-scale structure by anodization. Subsequently,\nthe contact area between the droplet and surface was minimized by\nlowering the surface energy using the SAM coating of FDTS. The oleophobic\nsurface exhibits sufficient repellency for various solutions with\nlow surface tension as well as for water. Magnesium has many restrictions\nin forming the surface structure by etching, anodization, etc. because\nof its high chemical reactivity compared to those of other metals.\nThe successful formation of a hierarchical structure on the magnesium\nsurface is expected to provide new opportunities for future magnesium\nsurface studies.",
"introduction": "1 Introduction The wetting\nproperties of\nmaterial surfaces have been widely studied because of their potential\nfor various applications. It is known that the surface wettability\nvaries depending on the surface structure and surface energy. The\nWenzel model shows that the surface wettability increases as the surface\nroughness increases. 1 Therefore, if the\nsurface roughness is sufficiently high, the surface can be fully wetted\nby water. In addition, the Cassie–Baxter model states that\na proper surface structure and proofing agent provide superhydrophobic\ncharacteristics that prevent the surface from getting wet. 2 Based on this model, many studies related to\nvarious fields such as self-cleaning, antifrosting, antifouling, and\nanticorrosion have been conducted by making use of superhydrophobic\nproperties realized through surface structure and surface energy modification. 3 − 10 However, a simple\nsuperhydrophobic surface cannot repel liquids with a relatively low\nsurface tension, such as oil. In this case, the surface may be easily\ncontaminated by other materials that are entangled with wet oil, resulting\nin the loss of superhydrophobic properties of the surface. This has\nled to the development of oleophobic surfaces with repulsion not only\nfor water but also for liquids with low surface tension. 11 In the case of an initially developed oleophobic\nsurface, an air layer was formed in the lower part of the structure\nin the form of a column having a wider top area than the lower part,\nthereby preventing the penetration of the liquid. 12 − 14 Thus, a droplet is supported by a columnar\nstructure, and the contact area between a liquid and surface is drastically\nreduced. Accordingly, the adhesion force between the surface and the\nliquid is also greatly reduced. However, to manufacture such a special\nstructure, anisotropic etching through a special system is required.\nTherefore, several studies have been conducted to reduce the adhesion\nforce as much as possible through a relatively easy method. Consequently,\nmany methods to fabricate oleophobic surfaces using anodization and\nchemical etching on the surfaces of various metals, such as aluminum,\ncopper, zinc, and titanium, have been published. 15 − 23 However, research on the oleophobic surface of magnesium has not\nprogressed, even though magnesium is widely used in aerospace, automotive,\nand electronics industries owing to its many advantages such as low\nweight, high ductility, specific rigidity, and specific strength.\nThe reason for this is the high electrochemical reactivity of magnesium.\nMoreover, anodization, chemical etching, etc., which can be applied\nto other metals to form various structures, do not form a special\nstructure on the magnesium surface that can increase the roughness\nsharply. Thus, the existing studies on superhydrophobic magnesium\nsurfaces focus on increasing the repulsive force against liquids by\nforming particles on the surface or covering the film layer; 24 − 32 however, in this way, the repulsive force against a solution having\nlow surface tension is limited. In this study, we introduce\na new fabrication method to realize a highly oleophobic magnesium\nsurface on the magnesium alloy AZ31, using two-step anodization, magnesium\nhydroxide particle formation, and silane coating process, as shown\nin Figure 1 . While\nthe large-scale wall-shape structure acts as a support for a droplet,\nthe flake-like small-scale structures dramatically reduce the contact\narea between the surface and droplet, thereby dramatically reducing\nthe adhesion. In this way, a surface that can repel various solutions\nwith low surface tension is realized. Figure 1 Preparation process on the surface of\nhighly oleophobic magnesium alloy AZ31.",
"discussion": "2 Results\nand Discussion When a surface has\na complex structure, in the general case, a liquid can permeate through\ngaps in the structure and the surface is completely wet. Under certain\ncircumstances, however, the air layer is trapped under the liquid,\nand they form a complex interface between the surface, liquid, and\nair. At this time, the liquid can be repelled by the surface and does\nnot get wet. This complex phenomenon is generally described by the\nCassie–Baxter (CB) model. According to the CB model, the apparent\ncontact angle θ CB is given by the following equation 2 1 where θ c is the intrinsic contact angle on the flat\nsurface and f s is the contact area fraction\nof the solid. According to eq 1 , the large intrinsic angle and the small contact area with\nthe surface can lead to a high apparent contact angle. It is necessary\nto form a complex structure on the surface and lower the surface energy\nto realize that. By forming a complex structure on the surface, it\nis possible to secure a gap between the structures in which an air\nlayer can exist. In addition, the decrease of the surface energy can\nlead to an increase of the repulsive force of the surface against\nthe liquids, so it is possible to allow the air layer to be formed\ninstead of the liquid permeating into the gap between the secured\nstructures. In this research, two-scale hierarchical structure\nis formed to make the gap between structures. To form the hierarchical\nstructure, two processes were applied: anodization for large-scale\nstructure and magnesium hydroxide particle formation for small-scale\nstructure. Figure 2 shows the anodized AZ31 surface by timeline. The AZ31 specimen connected\nto the anode loses electrons, forms magnesium ions, and carves the\nsurface of the specimen. Some of these magnesium ions react with the\nhydrogen carbonate and hydroxide ions dissolved in the electrolyte\nto form magnesium carbonate precipitates. In addition, magnesium ions\nreact with the hydroxide ions formed at the cathode to form magnesium\nhydroxide precipitates. A series of reactions that occur during anodization\nis as follows 2 3 4 5 In the\nfirst anodization step, the anodization started partially and spread\nout from the starting points; therefore, the anodized surface exhibits\nan irregular structure, as shown in Figure 2 a. In addition, the bottom of the surface\nis cracked, and the cracks grow; hence, the first anodized surface\ncan show superhydrophilic characteristics because any liquid can easily\npenetrate into the cracks and spread along the cracks. However, the\nsurface cannot show superhydrophobic or oleophobic properties because\nthe contact area between the droplet and surface is too large even\nif the surface is fluorinated. Therefore, it is important to reduce\nthe contact area to realize an oleophobic surface. Figure 2 Field emission\nscanning electron microscopy (FE-SEM) images\nof the surface after (a) first anodization and second anodization\nfor (b) 5 min, (c) 10 min, and (d) 20 min. For this\nreason, a second anodization was performed. The second anodic oxidation\nalso occurs according to the same reactions as the first anodic oxidation,\nbut the reaction resulting in the formation of magnesium carbonate\ndoes not occur because the sodium hydroxide solution is used as an\nelectrolyte. Therefore, anodization proceeds through the reactions\nof (2), (3), and (5). After 5 min of the second anodization step,\nthe cracks that can be seen in Figure 2 a grow wide, and a sunken lowland is formed if the\nsecond anodization time reaches 10 min, as shown in Figure 2 b,c. In particular, wall-like\nstructures are formed as a recessed area is formed, and the wall-like\nstructures reduce the contact area between the droplet and the surface.\nHowever, the contact area, especially the top side of the structure\nthat is directly in contact with the droplet, is still large. In addition,\nthe top area can be widened after magnesium hydroxide formation; therefore,\nthe anodization was conducted for 20 min to reduce the area. As shown\nin Figure 2 d, the wall\nstructure was thinned due to considerable anodization, and increasing\nthe anodization time did not produce a significant difference in the\nstructure. Therefore, the anodization time of 20 min was selected\nas the optimized time. After two-step anodization, magnesium\nhydroxide particles are formed on the surface. When the magnesium\nsulfate solution containing a magnesium specimen is heated to 180\n°C in an autoclave, the magnesium specimen reacts with water\nand forms hydrogen gas and magnesium oxide. Then, the magnesium oxide\nreacts with water to form the initial magnesium hydroxide ( eqs 6 and 7 ). Meanwhile, MgOH + crystal nuclei are formed by the reaction\nof magnesium ions and water in the aqueous solution, and the nuclei\ncombine with hydroxide group on the surface of the magnesium specimen\nto form new magnesium hydroxide ( eqs 8 and 9 ). By repeating this process,\nthe magnesium hydroxide structure grows. 6 7 8 9 Figure 3 shows the images of the AZ31 surfaces on which magnesium hydroxide\nparticles form for 30, 60, 90, and 120 min. Figure 3 a shows the result of the 30 min formation.\nIt can be seen that almost no magnesium hydroxide is formed as a whole,\nand a small magnesium hydroxide seed is formed in the low zone of\nthe surface and the bottom of the wall structure. When the reaction\ntime is increased to 60 min, magnesium hydroxide is still not widely\nformed as a whole, but the magnesium hydroxide seed grows, and the\nparticle formation proceeds to the lowland region and the wall that\nare not yet formed ( Figure 3 b). As can be seen in Figure 3 c, when the reaction occurred for approximately 90\nmin, most of the surface was covered with magnesium hydroxide particles,\nbut the particles were not completely formed on the top of some wall\nstructures. When the particle formation time increased to 120 min\n( Figure 3 d), it was\nconfirmed that magnesium hydroxide was completely formed over the\nentire surface of the specimen. It can be seen from Figure 3 e–h that the magnesium\nhydroxide particles grew in size with increasing reaction time. The\ngrown magnesium hydroxide particles can greatly reduce the contact\narea with the droplets, thereby reducing the adhesion force, helping\nthe droplets to fall off the surface easily. Figure 3 FE-SEM images\nof the AZ31 surface after magnesium hydroxide\nformation for (a, e) 30 min, (b, f) 60 min, (c, g) 90 min, and (d,\nh) 120 min. Surface fluorination\nis also indispensable for reducing the contact area between the surface\nand droplets. Without surface energy modification, the droplets penetrate\ninto the gaps of the structure, resulting in a wide contact area.\nTherefore, the surface formed by the structure has oleophilic properties,\nand when the droplet touches the surface, it can be seen that it spreads\nvery widely as if absorbed by the paper. By performing fluorination\non the surface, the repulsive force of the structure against the droplets\nis increased to form air layers instead of droplets between the structures,\nthereby minimizing the contact area between the droplets and the surface. Surface fluorination is achieved by a self-assembled monolayer\n(SAM) method. 33 , 34 When 1H,1H,2H,2H-perfluorodecyltrichlorosilane\n(FDTS) comes into contact with magnesium hydroxide, Cl ions bound\nto Si of FDTS can react with the OH ions of magnesium hydroxide to\nform the Si–O–Mg bonds ( Figure 4 a). In addition, when the FDTS in which Cl\nis bound to Si is converted to OH comes into contact with magnesium\nhydroxide on the surface, the Si–O–Mg bond is formed\nand water is produced as a byproduct ( Figure 4 b). The newly produced water again replaces\nCl bonded to Si of another FDTS with OH, which then combines with\nthe OH group of another magnesium. By repeating this process, each\nOH group on magnesium hydroxide is combined with FDTS, but the opposite\nside of FDTS is composed of −CF2 or −CF3; therefore,\nno reaction occurs. In this way, the single-molecule coating of FDTS\non the magnesium structure surface is possible, and the thickness\nof the coating is very thin, below 1.4 nm. Therefore, it is possible\nto form a large number of fluorine groups on the surface to lower\nthe surface energy without affecting the surface structure. Figure 4 Scheme of two\ntypes of fluorination mechanism by (a) original FDTS and (b) FDTS\nreplaced by the OH group. Figure 5 shows the\nX-ray photoelectron spectroscopy (XPS) analysis before and after surface\nfluorination with SAM coating. The source is Al Kα, and the\nphase shift is calibrated by the C 1s binding energy of 284.5 eV.\nIn the spectrum of the samples before and after fluorination, the\npeaks of the binding energies of approximately 1304 and 530 eV can\nbe seen in common, which are the values of Mg 1s and O 1s, respectively,\nconstituting the surface structure. In addition to the common peak,\nthe surface after fluorination can be seen to have a binding energy\nof approximately 685 eV, indicating that the content of −CF2\nand −CF3 bound to the surface by the SAM coating increased.\nBy this surface energy modification, the surface energy can be lowered,\nand the oleophobic surface can be realized with the structure formed\nabove. Figure 5 XPS spectra of the hierarchically\nstructured surface before\nand after fluorination. The oleophobic surface produced by two-step anodization,\nmagnesium hydroxide formation, and fluorination was tested for the\nrepellency of various solutions. Four kinds of liquids were used for\nthis test: DI water (72 mN/m), glycerol (62.5 mN/m), ethylene glycol\n(48.2 mN/m), and rapeseed oil (35.0 mN/m). For the untreated AZ31,\nthe surface showed hydrophilic and oleophilic properties due to the\noxide layer consisting of magnesium oxide and magnesium hydroxide\non the surface. Therefore, droplets could not maintain their hemispherical\nshape and spread widely, as shown in Figure 6 a. On the other hand, in the case of the\noleophobic surface, the droplets did not spread on the surface and\nremained spherical owing to the low surface energy and surface, as\nshown in Figure 6 b.\nTo confirm this quantitatively, the contact angle between the droplet\nand surface was measured. As shown in Figure 6 c, the contact angles were over 150°\nfor all of the droplets at 120 min of the formation of magnesium hydroxide\nthroughout the structure. When the magnesium hydroxide formation time\nwas less than 60 min, the contact angles were similar because magnesium\nhydroxide was partially formed only at the lower and side portions\nof the structure. The contact angle is increased under the 90 and\n120 min condition in which much magnesium hydroxide is formed at the\nupper end of the structure as compared to in the lower 60 min condition. Figure 6 Optical droplets images\non (a) bare surface and (b) oleophobic\nsurface. (c) Contact angles of various droplets. However, the tendency of the contact angle with water is different\nfrom the general oleophobic surface using other materials. In general,\nthe contact angle is proportional to the surface tension of a droplet\nwhen fluorination is applied to the surface. However, this oleophobic\nsurface shows that the contact angle of water is lower than that of\noil. In the case of water, a droplet maintains a high contact angle\nat the first contact with the surface, but it gradually spreads over\ntime, and the contact angle gradually decreases. However, this trend\ndoes not appear for oil, and further study is needed to explain the\ndifference between water and oil behavior."
} | 4,374 |
35208661 | PMC8879951 | pmc | 2,306 | {
"abstract": "Because microorganisms are the primary driving force behind litter decomposition, they play an important role in maintaining ecosystem material and chemical cycling. Arbuscular mycorrhizal (AM) fungi can improve host plant tolerance to various environmental stressors, making their application in mining area remediation important. In this study, litter from the dominant plant species ( Imperata cylindrica ) in a copper tailings mining area was selected as the experimental material. We conducted a greenhouse-based heavy metal stress experiment to investigate how AM fungi affect litter microbial community characteristics and key ecological factors. Results showed that AM fungi species, heavy metal treatments, and their combined interaction had significant impacts on litter pH. Additionally, enzyme activities in litter were significantly affected by interactions between AM fungi species and heavy metal contaminates. Ralstonia was significantly positively correlated to lead (Pb) content, indicating that Ralstonia had a certain tolerance to Pb pollution. Sucrase and urease activity were increased when plants were inoculated with Rhizophagus irregularis under Pb stress. Furthermore, Microbacterium , Brevundimonas , and Pseudonocardia all may play important roles in litter decomposition, while a certain tolerance was observed in Kushneria and Roseivivax to heavy metal pollution when plants were inoculated with Glomus mosseae . Results showed that AM fungi affected litter bacterial community structure and function by influencing plant litter properties. By exploring interactions between AM fungi and bacterial communities in plant litter under heavy metal stress, we will better understand associative processes that promote the cycling of soil organic matter and nutrients contaminated by non-ferrous metal tailings.",
"conclusion": "5. Conclusions Arbuscular mycorrhizal fungi play an important role in mining remediation. AM fungi had an effect on the microbial community characteristics of litters by influencing litter properties and enzyme activities, especially some key microbial groups with potential for heavy metal remediation. In the future, more attention should be paid to the dynamics of microbial community during litter decomposition, so as to deepen the understanding of the association between contaminated soil and nutrient cycling in mining areas.",
"introduction": "1. Introduction Litter is the link that connects plants and soil [ 1 ]. Moreover, its decomposition encourages the development of soil organic matter. It also plays an important role in the formation of soil carbon (C) pools and the release of mineral elements as well as various types of nutrients. Additionally, the role that litter decomposition plays is also important for regulating plant growth processes, while affecting the net productivity of terrestrial ecosystems [ 2 ]. As a type of symbiotic (mycorrhizal) microorganism, arbuscular mycorrhizas (AM) species are directly associated with soil and plant root systems and play critical roles in the host plant litter decomposition process. Many studies have been conducted on how AM fungi species influence litter decomposition, but results have been generally inconsistent. A study on woody plant litter reported that a particular AM fungi species ( Rhizophagus irregularis ) was able to significantly reduce its litter burden [ 3 ]. Another study reported that AM fungi can promote organic C decomposition under high CO 2 and N 2 concentrations [ 4 ]. Moreover, two AM fungi species ( Glomus mosseae and Glomus claroideum ) have been shown to significantly reduce decomposition coefficients in the root litter of Leymus chinensis [ 5 ]. The mechanism by which AM fungi affect litter decomposition is still unclear. Most studies have concluded that AM fungi symbiosis is obligate in nature and is subsequently unable to uptake nutrition saprophytically. However, by influencing the properties of host plant litter or microbial communities, AM fungi increased the nutrient uptake from substrate to host plant, and they may interact with other soil organism (e.g., free living bacteria) to decompose litter [ 6 , 7 ]. Isotopic labelling has also shown that AM fungi can directly decompose litter [ 7 ], and this type of mycorrhiza can significantly increase the initial nitrogen (N) and phosphorus (P) content of alfalfa ( Medicago sativa ) root litter, while also increasing the decomposition rate of root litter [ 8 ]. Additionally, interactions between AM fungi and soil microbial communities are complex. Associated mycelial secretions have been shown to contain H + , OH − , and organic acids, which all affect the microbial community structure of the rhizosphere [ 6 ]. Moreover, one AM fungi species ( Funneliformis mosseae ) has been shown to inhibit the growth of soil fungi and significantly reduce its overall biomass [ 9 ]; it can also promote the enrichment of Mycena in soil. Mycena is a typical lignin-decomposing bacterium that promotes the decomposition of litter [ 10 ]. Therefore, any changes in the microbial community induced by AM fungi may subsequently alter the production of metabolites and, consequently, impact litter decomposition. At present, the focus of most relevant studies has been on the relationships between AM fungi and soil microbial communities in natural ecosystems. Thus, associated interactions between AM fungi and litter microbial communities in contaminated ecosystems remain poorly understood. A previous study [ 11 ] reported on severe heavy metal pollution in a copper tailings dam in the southern region of Shanxi Province, China, and the role of the dominant plant species ( Imperata cylindrica ) was found to be important in the restoration of the plant community in this mining area. That study also showed that a large amount of I. cylindrica litter had accumulated at the end of plant growing season in this copper tailings area. Another study found that organic C content and changes to the litter decomposition process closely correlated to the adsorption and accumulation of metal elements in litter as well as to affiliated fixation and synergistic deposition in a heavy metal polluted environment [ 12 ]. Such findings can broaden our understanding of the influencing effects that AM fungi have on microbial litter communities as well as associative contributions to litter decomposition. To sum up, we hypothesized that AM fungi, heavy metal treatments, and their interaction would impact on the litter properties and enzyme activities, and different AM fungi would affect the microbial community structure and function of litters. This study can also improve the reparative status of degraded environments, namely, by exploring the role of AM fungi regarding nutrient cycling in copper mining areas subject to severe heavy metal pollution. For example, this study could select some available strains, which could be used for restoration in mining area copper tailings area in copper mining areas. It is of great significance to improve the ecological restoration efficiency of copper tailings areas.",
"discussion": "4. Discussion 4.1. Effects of AM Fungi on Litter Properties Litter properties are important for litter decomposition, and it is generally believed that the decomposition rate of litter is inversely proportional to the C/N ratio [ 16 ]. Previous studies have found that AM fungi had no significant effect on antecedent aboveground litter content for Leymus chinensis and M. sativa , while no significant difference was observed in litter decomposition rates [ 5 ]. These findings were consistent with results from previous study. Heavy metal content also affected litter decomposition. It was reported that Pb can inhibit cellulase and urease activities in litter and subsequently effect the decomposition of Phyllostachys pubescens litter [ 17 ]. However, cellulase and urease activities were promoted after AM inoculation, which increased litter decomposition under Pb stress. Moreover, AM fungi has been shown to promote host plant growth and increase both aboveground and belowground plant biomass [ 18 ]. This ultimately causes a biological “dilution effect” which subsequently reduces the heavy metal burden of plants [ 18 ]. In this study, we found that AM fungi inoculation can significantly influence the litter pH. Previous studies showed that there have an indirectly plant driven observed between soil pH and microbial community structure [ 19 ]. Soil pH was closely correlated with bacterial diversity and soil microbial communities [ 20 ]. Similarly, there is a close relationship between AM fungi and litter pH, which could affect the microbial community of litters. 4.2. Induced AM Fungi Alteration of Bacterial Community Composition in Litter In this study, the inoculation of AM fungi and heavy metal stress altered the bacterial community composition in host plant litter. Specifically, the AM fungal inoculation treatments increased the relative abundance of both Alphaproteobacteria and Bacteroidia. Moreover, AM fungi can cause a significant correlation between Alphaproteobacteria and ligninase activity for various cellulase types [ 21 ]. Alphaproteobacteria consists of many genes that can encode lignocellulolytic enzymes, which indicated that Alphaproteobacteria can synthesize a variety of lignin and cellulase enzymes [ 22 ]. Furthermore, AM fungi inoculation significantly enriched Bradyrhizobium, Microbacterium, Ellin6055, Rheinheimera, and Limnohabitans . Among these bacterial genera, Bradyrhizobium and Microbacterium are two typical lignin-decomposing bacteria [ 23 ]. Furthermore, our study found that Microbacterium was significantly correlated to both catalase and polyphenol oxidase activity. It has been reported that the role that Rheinheimera plays in organic C decomposition is important [ 24 ]. These results suggested that AM fungi inoculation may attract and promote microbial groups associated with litter decomposition to colonize litter, subsequently promoting litter decomposition. Additionally, Ornithinimicrobium, Nesterenkonia , and Sinobaca significantly correlated to sucrase, cellulase, and catalase activities while exhibiting a certain tolerance to heavy metals (Cd, Pb, etc.). Previous studies showed that bacteria had effects on heavy metal remediation. AM fungi indirectly affect the heavy metal speciation by changing the microbial community structure and physicochemical properties of rhizosphere soil [ 25 ]. AMF strains isolated from different contaminated sites tended to have different tolerance ranges to heavy metals [ 26 , 27 ]. Moreover, AM fungi changed the heavy metal oxidizing and bacterial community structure in rhizosphere soil, and reduced the bioavailability of heavy metals [ 28 ]. Results therefore expounded the significance of these bacteria of I. cylindrica litter in heavy metal polluted areas. 4.3. Effects of AM Fungi on the Functional Characteristics of Bacterial Communities in Litter In this study, we found that the inoculation of AM fungi would decrease the abundance of genes that encode catalase activity within litter bacterial communities, which was consistent with catalase activity trends. This indicated that AM fungi inoculation may reduce the ability of bacterial communities to decompose recalcitrant C sources, such as lignin. Moreover, AM fungi inoculation can increase the abundance of genes that encode cellulase and hemicellulase. The differences found between functional gene abundance and gene expression could have potentially resulted from activities associated with other microbial groups, except for bacteria during litter decomposition, such as fungi and protists. It was also reported that Basidiomycota members have the ability to secrete different varieties of cellulase and ligninolytic enzymes, which have a regulatory effect on litter decomposition [ 29 ]. Additionally, the role of oomycota is also important in the decomposition of organic matter [ 30 ]. Functional gene distribution can only predict the metabolic potential and the ecological function of bacterial communities; however, it is unable to predict actual metabolic conditions and the functional characteristics of bacterial communities [ 31 ]. Furthermore, a few other studies have also found differences between gene abundance and gene expression [ 32 , 33 ] Thus, it is important that we further investigate ways in which AM fungi effect functional characteristics of bacterial communities in combination with transcriptome, proteome, and other associated regulatory mechanisms."
} | 3,150 |
36923901 | PMC10009464 | pmc | 2,307 | {
"abstract": "The expression of many virulence genes in bacteria is regulated by quorum sensing (QS), and the inhibition of this mechanism has been intensely investigated. N -acetylcysteine (NAC) has good antibacterial activity and is able to interfere with biofilm-related respiratory infections, but little is known whether this compound has an effect on bacterial QS communication. This work aimed to evaluate the potential of NAC as a QS inhibitor (QSI) in Pseudomonas aeruginosa PAO1 through in silico and in vitro analyses, as well as in combination with the antibiotic tobramycin. Initially, a molecular docking analysis was performed between the QS regulatory proteins, LasR and RhlR, of P. aeruginosa with NAC, 3-oxo-C12-HSL, C4-HSL, and furanone C30. The NAC sub-inhibitory concentration was determined by growth curves. Then, we performed in vitro tests using the QS reporter strains P. aeruginosa lasB-gfp and rhlA-gfp , as well as the expression of QS-related phenotypes. Finally, the synergistic effect of NAC with the antibiotic tobramycin was calculated by fractional inhibitory concentrations index (FIC i ) and investigated against bacterial growth, pigment production, and biofilm formation. In the molecular docking study, NAC bound to LasR and RhlR proteins in a similar manner to the AHL cognate, suggesting that it may be able to bind to QS receptor proteins in vivo . In the biosensor assay, the GFP signal was turned down in the presence of NAC at 1000, 500, 250, and 125 μM for lasB-gfp and rhlA-gfp ( p < 0.05), suggesting a QS inhibitory effect. Pyocyanin and rhamnolipids decreased ( p < 0.05) up to 34 and 37%, respectively, in the presence of NAC at 125 μM. Swarming and swimming motilities were inhibited ( p < 0.05) by NAC at 250 to 10000 μM. Additionally, 2500 and 10000 μM of NAC reduced biofilm formation. NAC-tobramycin combination showed synergistic effect with FIC i of 0.8, and the best combination was 2500–1.07 μM, inhibiting biofilm formation up to 60%, besides reducing pyocyanin and pyoverdine production. Confocal microscopy images revealed a stronger, dense, and compact biofilm of P. aeruginosa PAO1 control, while the biofilm treated with NAC-tobramycin became thinner and more dispersed. Overall, NAC at low concentrations showed promising anti-QS properties against P. aeruginosa PAO1, adding to its already known effect as an antibacterial and antibiofilm agent.",
"conclusion": "4 Conclusion To our knowledge, this is the first study to show NAC as a potential QSI against P. aeruginosa PAO1. The in silico analyses suggest a possible inhibitory mechanism, since NAC interacts with LasR and RhlR proteins similarly to cognate AHLs and, sometimes, even better than the QSI furanone C30. Assays with biosensor strains and QS-related phenotypes demonstrated the in vitro inhibitory potential, besides swarming and swimming motilities were strongly inhibited at the same concentration of NAC that inhibited biofilm formation The combination of NAC-tobramycin seems to be an interesting way to reduce the drug dose and inhibit biofilm formation by P. aeruginosa PAO1, as visualized by confocal microscopy. We suggest further studies to evaluate the expression of QS target genes of P. aeruginosa PAO1 in the presence of NAC in different concentrations and in combination with tobramycin to elucidate the mechanism of action, as well as to perform in vivo assays with animal models and also test other strains of P. aeruginosa .",
"introduction": "1 Introduction Bacteria have sophisticated communication systems that enable them to send and receive chemical messages to and from other bacteria [ 1 ]. This mechanism is known as quorum sensing (QS), which is based on the production, secretion, and response to extracellular signaling molecules called autoinducers (AIs), whose concentration correlate with the populations' cell density [ [1] , [2] , [3] ]. QS is stimulated through the recognition of a threshold concentration of AIs, which allows communication between cells, leading to the expression of specific target genes [ 4 , 5 ]. Therefore, this communication enables bacteria to monitor the environment and act as a group, which facilitates population-dependent adaptive behavior, such as bioluminescence, surface motility, biofilm formation, production of toxins, sporulation, conjugative DNA transfer, among others [ [6] , [7] , [8] , [9] ]. The AIs are chemically diverse molecules and conventionally separated into three major groups. First, N -acyl-homoserine lactone (AHL, or AI-1) is a well-known QS system, used by Proteobacteria. Second, a group of molecules derived from 4,5-dihydroxy-2,3-pentanedione (DPD), commonly referred to as AI-2, is found in both Gram-positive and Gram-negative bacteria, and is thought to be a type of interspecies communication mechanism [ 10 , 11 ]. Third, the oligopeptides or auto-inducing peptides (AIPs) are usually secreted by ABC-type carrier proteins and used as signaling molecules by Gram-positive bacteria [ 3 , 12 ]. Additionally, there are several other types of molecules that can also be used as signal molecules among bacteria [ 3 ]. Pseudomonas aeruginosa is an opportunistic human pathogen, especially dangerous to cystic fibrosis patients and burn infections [ 13 ]. This bacterium expresses many virulence genes through QS regulation, and has become the main model in bacterial anti-virulence strategy studies [ 14 ]. The complexity of QS systems in this microorganism is one of the main factors responsible for its selective adaptation and environmental versatility [ 15 ]. The QS network of P. aeruginosa consists of four interconnected systems, namely las, rhl, pqs , and iqs , organized in a hierarchical manner. Two synthase/receptor pairs LasI/LasR and RhlI/RhlR, are central to the QS network. LasI resides at the top of the hierarchy, synthesizing the AHL called N -3-oxo-dodecanoyl- l -homoserine lactone (3-oxo-C12-HSL), which binds to LasR and directs the expression of various genes. Likewise, RhlI produces N -butyryl- l -homoserine lactone (C4-HSL) that binds to its RhlR cognate transcription regulator, activating a suit of other genes [ 5 , 16 , 17 ]. The P. aeruginosa pathogenicity depends on the ability of this bacterium to produce many virulence factors, such as rhamnolipids and pyocyanin, that are under the direct regulation of the QS system by RhlR-RhlI [ 14 , 18 ]. Rhamnolipids are glycolipids that have strong surfactant abilities and play a key role in microbial motility and biofilm development on host surfaces under stressful conditions [ 15 , 19 , 20 ]. Pyocyanin is a blue pigment and redox-active molecule belonging to the class of phenazines, which is crucial for P. aeruginosa virulence, as it is toxic to host cells [ 21 , 22 ]. Since the expression of many virulence factors in bacteria is regulated by QS, there is an increased interest in finding QS inhibitors (QSI). Besides, the emergence of multi-drug resistant bacteria is very worrisome and the use of antibiotics are not always effective in eradicating them and their biofilm [ 23 ]. Thus, new non-antibiotic therapies are urgently needed to treat infections and stop the spread of resistant bacteria [ 24 ]. Anti-virulence approaches have been widely studied as they can interrupt a pathogen's virulence by inhibiting the production or activity of virulence factors without interfering in bacterial growth and presenting a lower selective pressure against resistance [ 5 , 25 ]. This is one of the reasons to study antimicrobial agents at low concentrations or even to repurpose non-antibiotic drugs as it has been demonstrated with aspirin [ 26 ] and ibuprofen [ 20 ], both inhibiting QS-regulated phenotypes in P. aeruginosa . N -acetylcysteine (NAC), the N -acetyl derivative of the natural amino acid l -cysteine, is a safe and inexpensive medication commercially accessible since the 1960s. It has been used directly or in combination with other medications for treatment of various diseases like chronic bronchitis, asthma and as a mucolytic agent. Besides, NAC acts directly as a scavenger of free radicals, being considered a powerful antioxidant, as demonstrated by several in vitro and in vivo studies conducted in animals and humans [ 27 , 28 ]. In vitro studies have indicated that NAC has good antibacterial properties against Gram-positive and Gram-negative bacteria, and is able to interfere with biofilms-related respiratory infections and for eradicating mature biofilms in the environment [ 23 , 29 ]. Nonetheless, there are no studies demonstrating whether NAC has an effect on bacterial QS communication. Thus, this work aimed to evaluate the potential of NAC as a QSI in P. aeruginosa PAO1 through in silico analyses by molecular docking and in vitro analyses of QS-regulated phenotypes. In addition, we also evaluated the combined effect of NAC and the antibiotic tobramycin on growth and biofilm formation.",
"discussion": "3 Results and discussion 3.1 NAC binds in silico to LasR and RhlR proteins of P. aeruginosa PAO1 Molecular modeling was used to predict the 3D structure of RhlR protein of P. aeruginosa PAO1. This process is considered to be the most accurate of the computational structure prediction methods, and has many applications in drug discovery [ 56 ]. In the present work, the verification and validation results of the 3D generated structures RhlR named 4Y15-PAO1 and 4Y17-PAO1 revealed that they are reliable ( Figs. S1 and S2 ). The Ramachandran plot analysis showed that more than 96% of the residues of both structures were in favored regions. In addition, both structures had more than 91% of the amino acid residues with scores greater than or equal to 0.2 in the Verify3D, indicating a compatibility between an atomic model (3D) with its own amino acid sequence (1D). Finally, both structures showed a threshold higher than the cut off of 91% in ERRAT overall quality factor ( Figs. S1 and S2 ). Similar results for RhlR protein obtained by homology modeling are reported [ [57] , [58] , [59] ]. Molecular docking was performed between different structures of LasR and RhlR proteins of P. aeruginosa PAO1 with four compounds: 3-oxo-C12-HSL ( Fig. 1 A), an AHL synthesized by LasI, which binds to LasR [ 60 ]; C4-HSL ( Fig. 2 A), an AHL synthesized by RhlI, which binds to RhlR [ 14 , 61 ]; NAC ( Fig. 1 , Fig. 2 B), the compound in study, and furanone C30 ( Fig. 1 , Fig. 2 C), a known QSI [ 62 ]. This analysis is important to identify a possible QS inhibition, and a possible mechanism, by interference in LuxR-type transcriptional regulators, i.e. , LasR and RhlR proteins. The binding affinities and binding amino acid residues are shown in Table 1 , Fig. 1 , Fig. 2 . Fig. 1 Molecular docking of 2UV0 structure of LasR protein with 3-oxo-C12-HSL (A), NAC (B), and furanone C30 (C). Backbone representation with hydrogen bond between the amino acid residues and compounds (D to F), surface and backbone representation (G to I) and surface representation showing the compound completely buried in the active binding pocket (J to L). Backbone representation of LasR protein in gray; black arrow indicates the binding site; blue dashed line indicates hydrogen bond. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 1 Fig. 2 Molecular docking of 4Y17 structure of RhlR protein with C4-HSL (A), NAC (B), and furanone C30 (C). Backbone representation with hydrogen bond between the amino acid residues and compounds (D to F), surface and backbone representation (G to I) and surface representation showing the compound completely buried in the active binding pocket (J to L). Backbone representation of RhlR protein in gray; black arrow indicates the binding site; blue dashed line indicates hydrogen bond. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 2 Table 1 Results of molecular docking of different structures of LasR and RhlR proteins of P. aeruginosa PAO1 with 3-oxo-C12-HSL, C4-HSL, NAC, and furanone C30. Table 1 Protein - structure Molecule Binding residue Hydrogen bond score Steric interaction score Score total Rank LasR - 2UV0 3-oxo-C12-HSL Y56, W60, Y64, D73, S129 −10.12 −77.05 −81.77 1 NAC Y56, Y64, D73, T75, T115, S129 −13.11 −28.49 −39.91 2 Furanone C30 T75, T115, S129 −6.00 −31.95 −37.95 3 LasR - 6D6A 3-oxo-C12-HSL Y56, Y64, T75, S129 −8.00 −72.68 −76.01 1 NAC Y56, D73, T75, T115, S129 −15.10 −26.76 −39.89 2 Furanone C30 T75, T115, S129 −6.00 −32.08 −38.08 3 LasR - 6D6L 3-oxo-C12-HSL Y56 −3.37 −78.54 −75.73 1 NAC W60, D73, T75, T115, S129 −13.99 −29.64 −41.96 3 Furanone C30 T75, T115, S129 −7.34 −38.45 −45.78 2 LasR - 6D6O 3-oxo-C12-HSL Y56, W60, S129 −6.90 −74.04 −76.36 1 NAC Y56, W60, T75, T115, S129 −11.68 −33.49 −43.03 3 Furanone C30 T75, T115, S129 −7.71 −38.71 −46.42 2 LasR - 6D6P 3-oxo-C12-HSL Y56, D73 −5.08 −78.74 −77.73 1 NAC W60, D73, T75, T115 −9.19 −34.57 −42.22 3 Furanone C30 T75, T115, S129 −6.90 −38.05 −44.95 2 RhlR - 4Y15-PAO1 C4-HSL Y64 −4.00 −41.17 −43.90 1 NAC D81, Y64, Y72, S135 −8.00 −25.84 −32.36 3 Furanone C30 No binding 0.00 −33.90 −33.90 2 RhlR - 4Y17-PAO1 C4-HSL Y64, W68, D81 −6.79 −39.91 −45.54 1 NAC Y64, W68, D81 −8.93 −29.07 −36.65 2 Furanone C30 No binding 0.00 −34.81 −34.81 3 The molecular docking assay generates a score which mimics the potential energy change when the protein and the compound interact. The score is based on hydrogen bonds, metal ions and steric interactions, in which lower scores (more negative) correspond to higher binding affinities [ 39 ], as it can be observed in Table 1 . The different results obtained for each compound can be explained by the conformational changes of LasR and RhlR proteins when complexed with different compounds, as well as by the structural differences of these compounds. Nonetheless, all compounds were well buried in the active binding pocket of both proteins ( Fig. 1 , Fig. 2 ), remaining completely inside the pocket ( Fig. 1 , Fig. 2 L). For the LasR protein, 3-oxo-C12-HSL showed the highest binding affinities for all analyzed structures ( Table 1 ). Although NAC did not show binding scores as high as those of this AHL, the values were close to those obtained for furanone C30. Furthermore, NAC binding sites were similar to those of 3-oxo-C12-HSL, such as Y56, W60, Y64, D73, T75, and S129 ( Table 1 , Fig. 1 D and E). Similar results were observed for RhlR protein ( Table 1 ). The C4-HSL also shows the best score, while NAC and furanone C30 presented similar scores. In addition, the binding sites of NAC and C4-HSL were exactly the same for 4Y17 structure, such as Y64, W68, and D81 ( Fig. 2 D and E). The lack of hydrogen bond in furanone C30 indicates the importance of steric interaction in molecular docking, since the steric interaction is the result of the combination of non-polar interactions, non-polar-polar contacts, and the repulsive contacts ( Fig. 2 F). These results indicates that NAC and cognate AHLs potentially compete for binding to the same amino acid residues in both proteins and suggest that NAC may be able to bind to LasR and RhlR proteins equally or better than a known QSI molecule, in this case, furanone C30. Halogenated furanones produced by the marine red algae Delisea pulchra are a well-known group of anti-QS compounds [ 62 ]. Furanone C30 is effective at reducing QS-regulated phenotypes and consequently has been used as positive control for QS inhibition in many experiments, including inhibition of violacein production, biofilm formation and swarming motility [ 53 , 63 ]. Concerning the QS inhibition mechanism, Paczkowski et al. [ 48 ] showed that furanones function non-competitively by destabilizing LasR protein and promoting its degradation. Unfortunately, furanones exhibit toxicity to human cells and their use is not suitable in human hosts as therapeutic agents [ 48 , 64 , 65 ]. 3.2 Sub-MICs of NAC inhibit QS systems of P. aeruginosa The growth curves of P. aeruginosa PAO1 showed that 25000 μM of NAC completely inhibited the bacterial growth ( Fig. 3 A). Concentrations lower than 12500 μM, i.e ., 6200, 3100, 1500, and 800 μM did not influence the growth of this bacterium at the evaluated condition. Hence, these concentrations were used in the QS inhibition tests to ensure that NAC did not interfere in bacterial growth in the QS related bioassays. Fig. 3 Growth curves of P. aeruginosa PAO1 with concentrations ranging from 800 to 25000 μM of NAC in LB broth at 37 °C for 24 h (A). Dose-response curves of P. aeruginosa lasB-gfp (B) and rhlA-gfp (C) monitor strains treated with concentrations ranging from 125 to 1000 μM of NAC and 50 μM of furanone C30. Data shown as mean values ± standard error of three replicates and data shown as mean values of three replicates. Fig. 3 The use of reporter strains is an important asset to evaluate whether a compound potentially inhibits QS. For this assay, we used P. aeruginosa PAO1 harboring either a lasB-gfp or a rhlA-gfp fusion, which produce a modified and unstable version of GFP, as a QS reporter. So, expression of gfp gives rise to a burst of fluorescence when lasB or rhlA is induced, which are QS-regulated genes. This system is very sensitive, and the GFP signal is turned down in the presence of a QSI [ 40 ]. In the present study, the fluorescence peak of untreated control was compared with other peaks. The GFP signal was reduced ( p < 0.05) in the presence of 50 μM of furanone C30 by 76% for lasB-gfp and 60% for rhlA-gfp. NAC also significantly inhibited ( p < 0.05) both QS systems. For lasB expression, the best inhibitory concentrations were 500 μM (88%) and 1000 μM (85%). For rhlA expression, the best inhibitory concentrations were 250 μM (17%) and 125 μM (16%), compared to the fluorescence peak of the untreated control ( Fig. 3 B and C). Thus, we expected that NAC would have a greater effect in inhibiting phenotypes regulated by the las system in this assay. However, since the QS network is robust and interconnected in P. aeruginosa , an interference with LasI/LasR, which is at the top of the hierarchical network, will likely influence the expression of the other systems and their respective virulence factors. This is the case for pyocyanin and rhamnolipids, which are regulated by RhlI/RhlR and will be discussed below. Many compounds have been identified as QSI by the same biosensor strains. Jakobsen et al. [ 19 ] found that ajoene, the major bioactive compound present in garlic extract, is capable of inhibiting the expression of lasB-gfp and rhlA-gfp fusion in a dose-dependent manner, in concentrations ranging from 6.7 to 854 μM. Gökalsın et al. [ 66 ] showed that the carotenoid zeaxanthin at 12 μM inhibited las and rhl QS systems by 68 and 66%, respectively, when compared to the untreated control group. Beenker et al. [ 67 ] tested the effect of gregatins, a group of related fungal secondary metabolites, on GFP expression of lasB-gfp, rhlA-gfp and pqsA-gfp QS reporter strains, in concentrations ranging from 15.6 to 2000 μM. They observed that gregatins had the least effect or no effect on the rhlA-gfp reporter, similar to the result found in the present study with NAC. 3.3 Sub-MICs of NAC inhibited pyocyanin, rhamnolipids, and motility Pyocyanin and rhamnolipids are virulence factors that play an important role in biofilm development, in the establishment of infection and colonization of the host by P. aeruginosa [ 45 ]. Both are under control of QS; therefore, their inhibition is a good indicator of QS inhibition. Here, pyocyanin production decreased 19, 22, and 34% in the presence of NAC at 500, 250, and 125 μM, respectively ( p < 0.05) ( Fig. 4 A and Table S1 ). In these same concentrations, the rhamnolipids production was decreased by 28, 33, and 37%, respectively ( p < 0.05) ( Fig. 4 B and Table S1 ). Similar results were observed by Dai et al. [ 20 ] for ibuprofen at 100 μg/mL, which reduced the production of pyocyanin up to 24% and rhamnolipids up to 34% by P. aeruginosa . Natural compounds, such as 6-gingerol at 100 μM [ 63 ], coumarins at 2000 μM [ 68 ], and naringenin at 4000 μM [ 69 ] were also reported as inhibitors of pyocyanin and rhamnolipids production by P. aeruginosa . Fig. 4 Effect of NAC at different concentrations on the production of pyocyanin (A) and rhamnolipids (B), swarming (C) and swimming motilities (D) when compared to the control ( P. aeruginosa PAO1 without NAC). Data shown as mean values ± standard error of three replicates. Means followed by different letters differ ( p < 0.05) by Dunnett's test and by the same letters do not differ ( p > 0.05). Fig. 4 In the present study, we observed that the inhibition of these phenotypes was not concentration-dependent, since all three tested concentrations inhibited ( p < 0.05) both virulence factors statistically equally ( Fig. 4 A and B). Other studies have also reported similar results. For instance, Beenker et al. [ 67 ] observed that increasing concentrations of gregatins from 15 to 2000 μM exhibited different behaviors in biofilm formation, pyocyanin and rhamnolipids production, showing nonlinear inhibitions, with the highest concentration not always having the best inhibition. Likewise, Dai et al. [ 20 ] evaluated the levels of 3-oxo-C12-HSL and C4-HSL in P. aeruginosa with ibuprofen at different concentrations and time points and they did not observe a concentration-dependent inhibitory effect of this compound. Since rhamnolipids are involved in swarming motility due to surfactant properties, we hypothesized that NAC would inhibit this type of motility. Swarming motility is an organized microbial movement on surfaces, dependent on extensive flagellation, cell–cell contact and QS, which contributes to biofilm formation and infection [ 70 ]. The swimming motility allows bacterial movement on semisolid surfaces, which is important in absorption of nutrients and biofilm formation [ 49 , 71 ]. Both motilities are strongly associated with pathogenesis in P. aeruginosa since they enable bacteria to colonize environments, attach to different surfaces and to form biofilms [ 70 , 72 ]. So, we examined the anti-QS potential of NAC against both types of motilities, in concentrations up to 10000 μM, which were the same tested in the biofilm assay. As observed in Fig. 4 , the highest tested concentrations of NAC (500, 2500, and 10000 μM) statistically inhibited swarming ( Fig. 4 C and Table S2 ) and swimming ( Fig. 4 D and Table S2 ) motilities. Furthermore, the concentration of 10000 μM of NAC inhibited swimming motility about 80%. On the contrary, no inhibition was observed for the lowest concentration of NAC (125 μM). Other studies have also reported the inhibition of swarming and/or swimming motilities of P. aeruginosa with potential QSIs, like curcumin at 50, 75, and 100 μg/mL [ 49 ], furanones at 7.5 mM [ 72 ], methyl gallate ranging from 16 to 256 μg/mL [ 47 ], and 5-hydroxymethylfurfural at 1.25 μL/mL [ 73 ]. 3.4 NAC-tobramycin combination has effect on bacterial growth, pigment production, and biofilm formation Antimicrobial resistance is a serious concern in clinical practice, and an emergent strategy in combating resistant bacteria is the coadministration of antibiotics with other compounds [ 42 , 50 , 74 ]. The synergistic activity of two drugs is achieved if they have different modes of action and such action can bring about a higher killing effect on the bacteria [ 50 ]. Here, the idea is that NAC would work as a QSI, acting in the inhibition of cellular communication, and, consequently, on the attenuation of virulence factors, while tobramycin works through a cell killing effect. Tobramycin belongs to aminoglycoside class of antibiotics and is commonly used to treat P. aeruginosa infections as well as infections from other susceptible organisms including Serratia, Proteus, Acinetobacter, and Klebsiella. Tobramycin is one of the most commonly monitored aminoglycosides in clinical laboratories, together with amikacin and gentamicin [ 75 ]. Besides, it is the main antibiotic recommended to treat chronic bacterial infections [ 76 , 77 ]. NAC-tobramycin combination against P. aeruginosa PAO1 showed a synergistic effect with FIC i of 0.8, considering MIC of NAC alone (25000 μM), MIC of NAC combined (10000 μM), MIC of tobramycin alone (5.35 μM), and MIC of tobramycin combined (2.14 μM). The data on the effectiveness of NAC-tobramycin combinations against this bacterium are represented in a checkerboard, observed visually ( Fig. 5 A). The observation of low turbidity and lack of pigment production (pyocyanin and pyoverdine), characteristics of P. aeruginosa , indicated the inhibitory effect of drug combinations. The highest concentration of NAC (10000 μM) and the highest concentration of tobramycin (4.28 μM), alone or in combination, were effective in partially inhibiting growth and the production of pigments. The lowest concentrations that reduced pigment production were 5000, 2500, and 1250 μM of NAC combined with 1.07 μM tobramycin ( Fig. 5 A). Fig. 5 Checkerboard assay showing effect of NAC-tobramycin combinations against P. aeruginosa PAO1. Yellow color represents growth inhibition (additive/MIC); cream color indicates synergistic effect with lower growth inhibition and no production of pigments; green color shows no effect, no reduction in growth based on the visible turbidity and on pigments production (A). Result of biofilm formation by crystal violet assay (B). Synergy maps in 2D and 3D plots of biofilm formation; red color shows synergism while green color shows no reduction of biofilm formation (no-synergy) (C). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 5 For biofilm assay, the first step was to treat the formed biofilm with crystal violet, and the result is shown in Fig. 5 B and Table S3 . It is possible to observe that the highest concentrations of NAC and tobramycin alone reduced the biofilm formation of P. aeruginosa PAO1 by 81% and 89%, respectively ( Fig. 5 B). These data were analyzed in the synergism assay, and the combination of 2500–1.07 μM (NAC-tobramycin) was the most synergistic combination for inhibiting biofilm formation, as observed in Fig. 5 C. In the synergism assay, the summary synergy scores can be interpreted as the average excess response due to drug interactions. For example, a synergy score of 10 corresponds to 10% of response beyond expectation [ 54 , 78 ]. In the present study, the global score was negative (−16.368), indicating that the interaction is likely to be antagonistic. The score of most synergistic area (MSA, which represents the most synergistic 3-by-3 dose-window in a dose-response matrix, Fig. 5 C) is near zero (−2.66), indicating an additive interaction. However, the score in the red point (2500–1.07 μM of NAC-tobramycin) is 26.24, indicating a synergistic interaction at this combination. In this scenario, biofilm formation was inhibited by 63.5% ( Fig. 5 B). This is an interesting result because these drug concentrations are lower than MIC, and the use of low concentrations may prevent the development of resistance. In this scenario, we evaluate the formed biofilm in different conditions by using confocal laser scanning microscopy ( Fig. 6 ). CLSM images for P. aeruginosa PAO1 ( Fig. 6 A) revealed a stronger, dense and compact biofilm. We observed that treatment with NAC altered the biofilm architecture. Fig. 6 B and C showed a biofilm thinner and more spaced, besides it is possible to visualize the background of the slide. These changes were greater in the combinations of NAC-tobramycin ( Fig. 6 D and E), since CLSM images showed an even less dense biofilm. The most synergistic combination, 2500–1.07 μM of NAC-tobramycin ( Fig. 6 E) also showed the dispersed and thinner biofilm with spaced adhered cells, proving the efficiency of the drug combination. NAC can increase the sensitivity of P. aeruginosa to tobramycin and is a potential agent to be co-administered with antibiotics. Fig. 6 Biofilms formed on glass coverslips by P. aeruginosa PAO1 (A) in the presence of NAC at 10000 μM (B) and 2500 μM (C), as well as in the presence of NAC in combination with tobramycin (Tobr.) at 10000 μM + 2.14 μM (D) and 2500 μM + 1.07 μM (E). Fig. 6 Quorum sensing regulates the expression of many genes associated with biofilm formation, and biofilms are associated with reduced sensitivity to antibiotics [ 79 ]. Thus, QS interference can be explored as an alternative to reduce bacterial virulence and increase the efficacy of antibiotic treatment. The combination of QSIs with antibiotics was demonstrated by Mion et al. [ 79 ], combining lactonases with fluoroquinolones and efficiently decreasing the amount of antibiotics required to fight P. aeruginosa clinical isolates. The flavonoid quercetin exhibited synergism between tobramycin and amikacin, providing better anti-biofilm activity against P. aeruginosa strains compared to their MIC dose [ 50 ]. Combinations of gallic acid-ampicillin and hamamelitannin-erythromycin inhibited the growth, biofilm viability, and motilities of Escherichia coli better than individual antibiotics, being promising candidates for eradicating pathogenic E. coli in humans and animals, according to Hossain et al. [ 51 ]. Finally, it is important consider that the virulence factors, motility, and biofilm formation are controlled by many pathways and molecules, in addition to quorum sensing. Thus, it is important to identify the exact target and mechanism of inhibitory compounds to ensure the influence on QS pathways [ 80 ]. For instance, more robust analysis could be performed, such as molecular and genetic techniques, gene expression assays of target QS genes, molecular dynamic analysis, and in vivo assays with animal models. Besides, it is necessary to treat bacteria with a specific concentration range, since low concentrations may not show an anti-QS effect, whereas high concentrations may be toxic or influence bacterial growth [ 67 ]. Here, in some tests, we used concentrations well below the MIC, and higher concentrations may present even better results for inhibiting the tested phenotypes."
} | 7,623 |
23924783 | PMC3869008 | pmc | 2,308 | {
"abstract": "Coral reefs are threatened throughout the world. A major factor contributing to their decline is outbreaks and propagation of coral diseases. Due to the complexity of coral-associated microbe communities, little is understood in terms of disease agents, hosts and vectors. It is known that compromised health in corals is correlated with shifts in bacterial assemblages colonizing coral mucus and tissue. However, general disease patterns remain, to a large extent, ambiguous as comparative studies over species, regions, or diseases are scarce. Here, we compare bacterial assemblages of samples from healthy (HH) colonies and such displaying signs of White Plague Disease (WPD) of two different coral species ( Pavona duerdeni and Porites lutea ) from the same reef in Koh Tao, Thailand, using 16S rRNA gene microarrays. In line with other studies, we found an increase of bacterial diversity in diseased (DD) corals, and a higher abundance of taxa from the families that include known coral pathogens (Alteromonadaceae, Rhodobacteraceae, Vibrionaceae). In our comparative framework analysis, we found differences in microbial assemblages between coral species and coral health states. Notably, patterns of bacterial community structures from HH and DD corals were maintained over species boundaries. Moreover, microbes that differentiated the two coral species did not overlap with microbes that were indicative of HH and DD corals. This suggests that while corals harbor distinct species-specific microbial assemblages, disease-specific bacterial abundance patterns exist that are maintained over coral species boundaries.",
"introduction": "Introduction One of the most recognized features of tropical, shallow-water corals is their symbiosis with photosynthetic unicellular algae (zooxanthellae) that provide photosynthetically fixed carbon to satisfy their host's respiratory requirements ( Muscatine and Cernichiari 1969 ) and facilitate calcification ( Gattuso et al. , 1999 ). Corals also live in association with numerous other microorganisms such as bacteria, archaea, protists, endolithic algae, fungi and viruses ( Rosenberg et al. , 2007 ), the significance of which is only partially understood ( Bourne et al. , 2009 ; Kimes et al. , 2010 ). The sum of all organisms is referred to as the coral holobiont ( Rosenberg et al. , 2007 ). It is now being recognized that bacteria contribute significantly to the biology of higher-order organisms ( Ezenwa et al. , 2012 ), and accordingly, bacteria associated with corals are considered a vital component of the coral holobiont. Their potential roles include nitrogen fixation ( Lesser et al. , 2004 ), decomposition of organic materials ( DiSalvo, 1969 ), production of antibiotic compounds ( Kelman et al. , 2006 ; Ritchie, 2006 ) and occupation of space to prevent colonization by pathogens ( Ritchie and Smith, 2004 ). Coral-associated bacteria have been shown to be host species-specific, diverse and complex ( Rohwer et al. , 2001 , 2002 ; Sunagawa et al. , 2010 ), and this assemblage comprises a unique signature that differs from bacterial communities in the surrounding water column ( Rohwer et al. , 2001 ; Frias-Lopez et al. , 2002 ; Bourne and Munn, 2005 ). Several studies have been conducted that highlight the role of bacteria in coral diseases ( Denner, 2003 ; Barash et al. , 2005 ; Rosenberg et al. , 2007 ; Bourne et al. , 2009 ; Meron et al. , 2009 ; Sunagawa et al. , 2009 ; Kimes et al. , 2010 ; Cardenas et al. , 2012 ; Cróquer et al. , 2013 ). Coral diseases appear as changes in tissue color in the form of patches or bands on the coral surface, associated with subsequent tissue damage, necrosis and tissue loss ( Richardson, 1998 ). In many areas, disease outbreaks have led to massive die-offs of reef-building corals that resulted in habitat loss for reef-associated organisms, with propensity for irreversible ecosystem change ( Richardson, 1998 ; Richardson et al. , 2001 ; Pandolfi et al. , 2005 ; Weil et al. , 2006 ). To date, the exact number of coral diseases remains unknown ( Pollock et al. , 2011 ). Their characterization is mainly based on field observations of altered phenotypes. As a result, the same disease might have been defined several times and in different ways depending on the species or the region affected ( Richardson, 1998 ). For most diseases, our knowledge on causative agents, modes of transmission or disease reservoirs is missing ( Weil et al. , 2006 ). It is unknown whether the same pathogens cause similar/same disease characteristics in different coral hosts or whether the same shifts in microbial assemblages result in the same disease phenotype in different coral species ( Lesser et al. , 2007 ; Rosenberg and Kushmaro, 2011 ). Furthermore, it is not known whether diseases with a similar phenotype are caused by similar underlying mechanisms, that is, if they are associated with comparable bacterial changes or species ( Lesser et al. , 2007 ). Answers to these questions might not only enable a clearer disease nomenclature but will also result in a better understanding of the mechanisms driving coral disease outbreak and progression and will eventually lead to a better understanding of coral holobiont pathology ( Rogers, 2010 ; Pollock et al. , 2011 ). White Plague Disease (WPD) is one of the first described coral diseases ( Dustan, 1977 ). Records show that WPD was responsible for several virulent outbreaks, and it is held responsible for major reef declines worldwide, especially in the Caribbean ( Richardson et al. , 1998b ; Aronson and Precht, 2001 ; Richardson et al. , 2001 ; Navas-Camacho et al. , 2010 ; Pollock et al. , 2011 ). Corals affected by a WPD phenotype show a pronounced line of bright, white tissue that separates the colored (living) part of the coral from bare, rapidly algal-colonized skeleton ( Richardson et al. , 2001 ). Three types of WPD, I ( Dustan, 1977 ), II ( Richardson et al. , 1998b ) and III ( Richardson et al. , 2001 ), have been described that differ in the rate of progression across a coral's surface and affect different species ( Richardson et al. , 2001 ; Sutherland et al. , 2004 ). Richardson et al. (1998a) initially suggested a species of Sphingomonas as the causative pathogen but Denner (2003) proposed Aurantimonas coralicida as the final WPD-causing pathogen in corals from the Caribbean. Similarly, Thalassomonas loyana ( Thompson et al. , 2006 ) has been proposed to be the causative agent of White Plague-like disease in the Red Sea. However, neither of these bacteria could be unequivocally verified as the responsible pathogen in subsequent studies ( Pantos et al. , 2003 ; Barash et al. , 2005 ; Sunagawa et al. , 2009 ; Cardenas et al. , 2012 ). Consequently, it is debatable whether a definitive pathogen for WPD exists or whether different pathogens or bacterial consortia produce a similar disease phenotype in different coral species. Given the inherent difficulties of assigning a pathogen to WPD, and thereby proving a causal relationship, Willis et al. (2004) suggested that coral diseases from the Great Barrier Reef (and by extension the Indo-Pacific) that produce a phenotype of white bands of tissue and/or skeleton should be referenced collectively as White Syndrome, unless the underlying disease etiology is known. Here we employed an alternative approach and tested whether healthy (HH) and diseased (DD) coral colonies displaying a WPD-characteristic phenotype ( Dustan, 1977 ; Richardson et al. , 2001 ) from the Indo-Pacific share similarities in underlying microbial community patterns and are comparable to WPD-affected corals and studies from the Caribbean. Sunagawa et al. (2009) was the first study that used 16S rRNA gene microarrays (PhyloChips, Second Genome) to assess bacterial community changes in WPD in Montastraea faveolata and demonstrated the overall feasibility of the method. In this study, we used PhyloChips to profile microbial communities of HH and DD colonies of two coral species ( Porites lutea and Pavona duerdeni ) displaying signs of WPD collected from the same reef in Koh Tao, Thailand. Our aim was to examine microbial community differences within and between species and between coral health states (HH vs DD). Additionally, 16S rRNA gene clone library sequencing was conducted to compare the two different methods for assaying coral-associated bacterial community structure.",
"discussion": "Discussion Coral-associated microbes constitute an essential component in coral holobiont functioning ( Rosenberg et al. , 2007 ). In particular, bacteria seem to have important roles in coral health and disease that still need to be further defined. One approach to identify common bacterial species is to conduct microbial studies in a comparative coral species framework. By choosing two species from the same coral reef, we limited variation in environmental variables in order to focus on the difference between coral species and coral health states. Here we characterized the abundance patterns of bacterial OTUs associated with HH and DD samples of P. duerdeni and P. lutea in a standardized comparison via 16S rRNA gene microarrays. The general feasibility of the PhyloChip platform to assess microbial community patterns in coral disease has been established by Sunagawa et al. (2009) . With regard to taxonomic diversity and identification of OTUs from corals collected at Sairee Beach in Thailand, PhyloChip microarrays yielded comparable results to clone library sequencing efforts. Both methods identified all OTUs to the phylum level and half of the OTUs to the family level, whereas about 60% of all the sequences failed to be annotated to the level of OTU with either method. We found a higher number of OTUs in our study (between 2756 OTUs in P. duerdeni HH and 10 848 OTUs in P. lutea DD) in comparison to sequence-based studies that looked at bacterial diversity in corals (for example, Barott et al. (2011) : between 163 and 461 OTUs per sample; Cardenas et al. (2012) : between 256 and 378 OTUs per sample; Koren and Rosenberg (2006) : 400 OTUs; Lins-de-Barros et al. (2010) : 354 OTUs). However, our estimates are well in line with estimates from Kellogg et al. (2012) that identified between 1112 and 9240 OTUs with PhyloChips in a comparison of sampling methods for coral microbial community analysis. Our data suggest that a lower bacterial diversity and abundance is associated with HH corals, which has also been reported by Pantos et al. (2003) , Sunagawa et al. (2009) and Cróquer et al. (2013) . We identified Pseudomonadaceae and Rhodobacteraceae as prominent families promoted in colonies displaying WPD signs. Rhodobacteraceae have been proposed to be opportunistic due to uncontrolled propagation in disease by Sunagawa et al. (2009) . Furthermore, bacterial taxa of the family Vibrionaceae were more abundant in DD samples as has been shown previously ( Sunagawa et al. , 2009 ; Mouchka et al. , 2010 ; Pollock et al. , 2011 ). Cardenas et al. (2012) conducted a study with a similar experimental design and compared the microbiome of HH and WPD-affected corals from two species ( Diploria strigosa and Siderastrea siderea ) in the Caribbean via 16S rRNA gene amplicon sequencing, but the authors did not find consistent bacterial shifts over coral species. The use of pooled replicates by Cardenas et al. (2012) for the different conditions and species might have influenced the ability to statistically test for coral species or condition specificity. Alternatively, WPD-affected corals in the Caribbean might display a different pattern. We did not find A. coralicida (GenBank ID EF512716.1), the putative WPD pathogen from the Caribbean, in any of the coral samples using clone libraries or the PhyloChip microarray. Also, T. loyana (GenBank ID AY643537.2), a proposed causative agent of White Plague-like disease from the Red Sea, was neither identified during our cloning efforts, nor detected on the microarray. This is consistent with results of other WPD-investigating studies that failed to discover either of these bacteria ( Pantos et al. , 2003 ; Sunagawa et al. , 2009 ; Cardenas et al. , 2012 ), which might be due to investigating phenotypically similar but not identical diseases ( Willis et al. , 2004 ; Lesser et al. , 2007 ). It could also be argued that pathogens are subject to evolutionary change, which has been shown in other coral diseases ( Reshef et al. , 2008 ). In this regard, the loss of pathogenicity due to changes in environmental conditions ( Meron et al. , 2009 ), repression through a newly, more favorably structured holobiont microbial assemblage ( Reshef et al. , 2006 ) or control through bacteriophages ( Cohen et al. , 2013 ) could be possible explanations. When comparing HH and DD samples, there is a clear trend from bacterial communities low in diversity and abundance (HH) to mixed and variable assemblages with high numbers of unclassified bacteria (DD), many of which were also identified in the surrounding water (data not shown). Most notably, we found no overlap between OTUs differentially abundant between coral species and their health states. Our data indicate that phenotypically similar coral diseases are accompanied by a common shift in bacterial communities in the two different coral species collected from the same reef. At the same time, corals display species-specific bacterial communities that are different from disease-associated bacteria. Health and disease were as strong a discriminator between samples as species. One important consequence is that microbial community patterns (‘bacterial footprints') might exist, which classify HH and DD coral specimens over species boundaries. In this regard, our study represents an approach to compare and analyze microbial assemblages of coral disease in a standardized framework (that is, via PhyloChip profiles) that might aid in the classification and categorization of coral diseases. Future studies should incorporate measures over geographical distances in the same and different species in order to understand whether these patterns are only regionally or globally conserved."
} | 3,575 |
37152640 | PMC10158823 | pmc | 2,309 | {
"abstract": "Replacing traditional substrates in industrial bioprocesses to advance the sustainable production of chemicals is an urgent need in the context of the circular economy. However, since the limited degradability of non-conventional carbon sources often returns lower yields, effective exploitation of such substrates requires a multi-layer optimization which includes not only the provision of a suitable feedstock but the use of highly robust and metabolically versatile microbial biocatalysts. We tackled this challenge by means of systems metabolic engineering and validated Escherichia coli W as a promising cell factory for the production of the key building block chemical 2-ketoisovalerate (2-KIV) using whey as carbon source, a widely available and low-cost agro-industrial waste. First, we assessed the growth performance of Escherichia coli W on mono and disaccharides and demonstrated that using whey as carbon source enhances it significantly. Second, we searched the available literature and used metabolic modeling approaches to scrutinize the metabolic space of E. coli and explore its potential for overproduction of 2-KIV identifying as basic strategies the block of pyruvate depletion and the modulation of NAD/NADP ratio. We then used our model predictions to construct a suitable microbial chassis capable of overproducing 2-KIV with minimal genetic perturbations, i.e., deleting the pyruvate dehydrogenase and malate dehydrogenase. Finally, we used modular cloning to construct a synthetic 2-KIV pathway that was not sensitive to negative feedback, which effectively resulted in a rerouting of pyruvate towards 2-KIV. The resulting strain shows titers of up to 3.22 ± 0.07 g/L of 2-KIV and 1.40 ± 0.04 g/L of L-valine in 24 h using whey in batch cultures. Additionally, we obtained yields of up to 0.81 g 2-KIV/g substrate. The optimal microbial chassis we present here has minimal genetic modifications and is free of nutritional autotrophies to deliver high 2-KIV production rates using whey as a non-conventional substrate.",
"conclusion": "4 Conclusion In this study we implemented an iterative approach to the production of valuable building blocks such as 2-KIV using non-conventional feedstocks and cell factories, i.e., whey lactose and the non-model E. coli W strain, and we demonstrated the latter’s suitability to deal with complex mixes of mono and disaccharides. In order to reduce the operational costs of the bioprocess, we designed a microbial chassis free of autotrophies and rich media requirements for growth and we rerouted the carbon flux towards pyruvate. Finally, we applied cutting-edge synthetic biology to the construction of a recombinant E. coli W strain capable of overproducing large titers of 2-KIV at the highest Yp/s reported so far. In what is a perfect fit for the demands of the circular economy, our approach and the W1262 strain pave the way for cost-effective production of key building blocks using recalcitrant feedstocks.",
"introduction": "1 Introduction Success in the implementation of a fully circular economy demands sustainable alternatives to substrates traditionally used in fermentative processes to avoid competition with human and animal food, e.g., glucose. Fulfilling this simple and yet very ambitious goal is the challenge driving interest in the use of agricultural and agro-industrial waste as feedstock for large-scale fermentation processes. However, non-conventional carbon sources often deliver lower yields due to their limited degradability and the presence of toxic byproducts ( Ren et al., 2011 ; Lopes et al., 2019 ). Overcoming these challenges therefore requires multi-layer optimization to deliver suitable alternative feedstocks and non-conventional, highly robust and metabolically versatile microbial biocatalysts. In recent years, much research has been conducted into the use of whey as a fermentation medium to produce lipids, organic acids and alcohols ( Cao et al., 2020 ; Mano et al., 2020 ; Novak et al., 2020 ). The physicochemical composition of whey is behind this growing interest, as it typically contains around 88% lactose, 4% protein, 1.4% (w/w) lipids and trace minerals in powder product ( Carranza-Saavedra, Sánchez Henao and Zapata Montoya, 2021 ). While it is usually more expensive to grow bacteria on pure lactose than on glucose, using whey as a source of lactose significantly reduces the cost ( Silva et al., 2015 ; Amado et al., 2016 ), thus making whey a promising non-conventional feedstock for microbial biotechnology. Searching for a new non-conventional bacterial chassis able to deal with such a novel feedstock while providing additional metabolic and/or genetic advantages is also a challenge. Strains free from catabolic repression phenomena, among other features, would be the ideal candidates ( Calero and Nikel, 2019 ). The Escherichia coli W strain has been used previously to produce natural metabolites and it brings potential improvements over E. coli K-12 strains: 1) it is able to use a broader range of carbon sources, at high concentration, 2) it is the only safe E. coli strain that uses sucrose as a carbon source ( Archer et al., 2011 ; Felpeto-Santero et al., 2015 ; Erian et al., 2018 ), 3) it releases smaller amounts of acetate during fermentation while generating more biomass in batch culture ( Archer et al., 2011 ), and 4) it tolerates increased stress conditions, such as high ethanol concentrations, acidic pH, high temperature and osmotic stress ( Shiloach and Bauer, 1975 ; Gleiser and Bauer, 1981 ; Alterthum and Ingram, 1989 ; Nagata, 2001 ). On the other hand, despite the long history of success using traditional metabolic engineering strategies, successful integration of modern biotechnology into the circular economy requires the development of new holistic approaches and tackling the complexity of living organisms from a systems-level perspective. The combination of computational methods based on metabolic models and cutting-edge synthetic biology has so far delivered a high degree of success in metabolic engineering endeavors. Indeed, it has paved the way for systematic exploration of the metabolic solution spaces required to produce target metabolites ( Wu et al., 2016 ; Gudmundsson and Nogales, 2021 ). Design-Build-Test-Learn (DBTL) iterative cycles stand up as popular examples of such multidisciplinary approaches to the production of a plethora of chemical compounds ( Gurdo et al., 2022 ). DBTL cycles are iterative designs combining the advances in systems and synthetic biology to deliver rational genetic modifications and high-throughput phenotyping of strains. DBTL-driven research leads to the establishment of sound conclusions from experimental results, knowledge building and the generation of new hypotheses. Hence, it contributes to unlocking novel biotechnological processes supporting the circular economy. In this context, an interesting idea is to apply multidisciplinary approaches to optimize biotechnological platforms towards the cost-effective production of key building blocks, which in turn are the precursors of a variety of value-added compounds. 2-ketoisovalerate (2-KIV) is one of these key building blocks and it has been the focus of significant attention ( Gu et al., 2017 ). 2-KIV is an important precursor in the biosynthesis of branched chain amino acids (BCAAs) such as L-valine, cofactors such as pantothenate, coenzyme A and other biologically active compounds such as glucosinolates ( Felnagle et al., 2012 ; Lee et al., 2019 ; Zhao et al., 2022 ). In microorganisms and plants, synthesis of 2-KIV requires two pyruvate molecules via three consecutive reactions included in the BCAAs pathway which are catalyzed by acetolactate synthase (AHAS), ketol-acid reductoisomerase (KARA) and dihydroxy-acid dehydratase (DHAD) ( Figure 1 ). Production of 2-KIV has been addressed using Klebsiella pneumoniae ( Gu et al., 2017 ), Corynebacterium glutamicum ( Buchholz et al., 2013 ), Pseudomonas putida ( Batianis et al., 2022 ) and, more recently, E. coli ( Zhou et al., 2022 ). 2-KIV production with E. coli has also been addressed extensively in the context of L-valine and isobutanol production ( Blombach et al., 2007 ; 2011 ; Park et al., 2007 ; Atsumi et al., 2008 ; Holátko et al., 2009 ; Krause et al., 2010 ; Park et al., 2011a ; Bartek et al., 2011 ; Park et al., 2011b ; Li et al., 2011 ; Hou et al., 2012 ; Lee et al., 2012 ; Buchholz et al., 2013 ; Hasegawa et al., 2013 ; Chen et al., 2015 ; Gu et al., 2017 ; Liang et al., 2018 ; Schwentner et al., 2018 ; Westbrook et al., 2018 ; Noda et al., 2019 ; Hao et al., 2020 ; Jung et al., 2020 ). However, such studies used conventional feedstocks and highly engineered strains harboring multiple autotrophies, thus requiring complex and expensive production media to support the bioprocess. FIGURE 1 L-valine biosynthesis in E. coli via 2-KIV. Acetohydroxyacid synthase (AHAS), ketol-acid reductoisomerase (AHAIR), dihydroxyacid dehydratase (DHAD), valine transaminase (VALTA), valine-pyruvate aminotransferase (VPAMTr), branched-chain amino acid transport system 2 carrier protein (BRNQ). Gene names are shown in purple. In this study, we used systems and synthetic biology to produce 2-KIV in a microbial factory fueled by whey and driven by the non-conventional and promising E. coli W strain. The strain’s design featured minimal genetic intervention in order to preserve its growth performance in minimal medium without the addition of expensive nutritional supplements. 24 h batch culture assays returned a 3.22 ± 0.07 g/L titer for 2-KIV and 1.40 ± 0.04 g/L for L-valine. We also obtained yields (Yp/s) of up to 0.8 g/g for 2-KIV.",
"discussion": "3 Results and discussion 3.1 Systems-level analysis of complex sugar metabolism in E. coli W It is well-known that E. coli W has a complex sugar metabolism and is able to use several types of carbohydrate as sole carbon and energy sources ( Wang et al., 2019 ). To assess the metabolic performance of E. coli W growing on such compounds, we monitored the growth kinetics of cultures feeding on a variety of monosaccharides (glucose, fructose and galactose), disaccharides (lactose, sucrose and maltose) and a mixture of them all. Glucose and fructose were rapidly assimilated ( Figure 2 ), so they turned out to support the shortest lag phase. They were followed by lactose and sucrose, while galactose and maltose displayed the longest lag phases. Despite their longer latency periods, disaccharides provided the highest growth rates (lactose > sucrose > maltose) and final OD 600 (sucrose > lactose > maltose). Interestingly, we observed the shortest latency period, highest growth rate and greatest final biomass with the sugar mix, which strongly suggests a synergistic and complementary metabolism supporting fast nutrient consumption and growth. The absence of diauxic growth curves strongly argues in favor of a limited effect of catabolic repression in the metabolism of mixture of sugars in this strain. These results show not just that E. coli W is able to use a variety of carbohydrates as single carbon and energy sources efficiently, but more critically that the availability of a carbohydrate mix has a positive impact on its growth performance. In addition, negligible catabolic repression (CCR) is an advantage over other conventional E. coli strains like E. coli K12, which suffered from the negative impacts of CCR in experiments using glucose and non-PTS sugars as carbon source ( Luo et al., 2014 ). FIGURE 2 Growth kinetic of E. coli W using a variety of mono and disaccharides as sole carbon and energy source. 2 g/L of each single carbon source were used while for the mixture, a total of 0.33 g/L of each compound were used. Vertical bars, ± represent standard deviation and lines are a guide for the eye. Similar lowercase letters per column indicate no statistical difference between treatments ( p < 0.05). We implemented a series of growth experiments using glucose at different ratios as an additional carbon source to further assess the role of CCR in E. coli W and verify whether its growth performance improves when using carbohydrate mixes instead of single carbon sources ( Figure 3 ). Although there is no evidence of a marked diauxic behavior in our kinetics assessment, µ max analysis for each of the different substrates ( Supplementary Table S5 ) shows a slight affectation when using glucose with galactose and maltose ( Figures 3B, E ). However, there are no significant differences ( p < 0.05) in µ max between other glucose-containing mixes. Significant differences between glucose-lactose mixes and other mixes, particularly when they contained glucose, show that the former favor growth. Overall, our results suggest that sugar-mix-powered E. coli W microbial cell factories are very promising, not just because they are virtually free of CCR but because their growth performance is improved when multiple sugars are simultaneously available. FIGURE 3 Growth kinetic of E. coli W in mix sugars with glucose for studying diauxic growth. All mix and pure carbon source were used at of 2 g/L. Vertical bars represent standard deviation and lines are a guide for the eye. (A) : Kinetic with lactose, (B) : kinetic with galactose, (C) : kinetic with sucrose, (D) : kinetic with fructose, (E) : kinetic with maltose. 3.2 Legacy and model-based design of a set of 2-KIV overproducer E. coli W strains \n E. coli W’s efficient carbohydrate metabolism ensures a large carbon flux around pyruvate, a key metabolic hub in sugar catabolism and the main precursor of 2-KIV. However, cost-effective production of pyruvate-derived metabolites such as 2-KIV requires additional flux rerouting ( Gu et al., 2017 ; Noda et al., 2019 ). Nowadays, redirection of carbon flux towards pyruvate has been unlocked in a large variety of microorganisms ( Buchholz et al., 2013 ; Gu et al., 2017 ; Novak et al., 2020 ), although most studies have focused on disabling pyruvate consumption along competing pathways ( Noda et al., 2019 ; Novak et al., 2020 ). For instance, detailed analysis of studies dealing with overproduction of 2-KIV from glucose ( Supplementary Figure S4 ; Supplementary Table S6 ) revealed three main knockout strategies. The first is the preferred choice and it avoids consumption of pyruvate as a precursor of: 1) acetyl-CoA via the pyruvate dehydrogenase complex (PDH, ace EF) or pyruvate formate lyase (PFL, pfl ABCD), 2) acetate via pyruvate oxidase (PO, pox B), 3) lactate via lactate dehydrogenase (LDH, ldh A), 4) L-alanine via glutamate-pyruvate aminotransferase (GPA, yfd ZQ) and 5) L-valine via branched-chain amino acid aminotransferases ( ilv E) ( Wang et al., 2018 ). Deletion of PDH seems to be of critical importance and is thus the most widely deleted competing pathway to increase levels of pyruvate in the cell ( Supplementary Figure S4 ) ( Blombach et al., 2007 ; 2011 ; Park et al., 2007 ; Park et al., 2011a ; Bartek et al., 2011 ; Buchholz et al., 2013 ; Chen et al., 2015 ; Nitschel et al., 2020 ). Alternatively, deleting anaplerotic reactions such as phosphoenolpyruvate carboxylase (PPC, ppc ) reduces gluconeogenesis from pyruvate and therefore delivers higher levels of 2-KIV ( Buchholz et al., 2013 ; Hasegawa et al., 2013 ; Schwentner et al., 2018 ). Beyond blocking pyruvate-consuming pathways, the second strategy is aimed at unbalancing the NAD/NADP ratio via deletion of malate dehydrogenase (MDH, mdh ) ( Park et al., 2007 ; Park et al., 2011b ) or redirecting the flux towards PPP via deletion of phosphate glucose isomerase (PGI, pgi ). This has been often used to increase the carbon flux through the branched chain amino acid (BCAA) pathway ( Park et al., 2007 ; Park et al., 2011a ; Bartek et al., 2011 ; Noda et al., 2019 ). Finally, deleting metabolite-specific competing reactions in order to streamline the optimized pathway is also a recurrent strategy resulting in improved titers ( Supplementary Figure S4 ). For instance, efficient production of L-valine required blocking the biosynthesis of alternative 2-KIV derivatives, such as L-leucine and/or pantothenate ( Radmacher et al., 2002 ; Park et al., 2007 ; Holátko et al., 2009 ; Park et al., 2011b ; Westbrook et al., 2018 ). However, these deletions resulted in auxotroph strains requiring complex culture media to grow, thus increasing the operational cost of the bioprocess. It seems reasonable to minimize the number of knockouts while maintaining high production in minimal medium to deliver a cost-effective 2-KIV production bioprocess. Therefore, we assessed 2-KIV production in silico to measure the performance of a set of knockout scenarios harboring minimal deletions. These included the removal of pyruvate-consuming pathways and generation of cofactor imbalances. Specifically, we used Monte Carlo Sampling to explore the metabolic space of three in silico mutant strains, i.e., ace F- mdh , ppc - mdh and ldh A- pfl B- adh E ( Figure 4 ). Selection of this set of reactions was based on the outcomes of previous studies ( Supplementary Table S6 ; Supplementary Figure S4 ) using caution to avoid auxotrophies in the final chassis. FIGURE 4 Flux frequency and variability through central metabolism of E. coli W as predicted by Monte Carlo sampling. Pyruvate kinase reaction (PYK), acetolactate synthase reaction (ACLS), dihydroxy-acid dehydratase reaction (DHAD1), extracellular transport of 2-KIV (EX-2KIV), extracellular transport of L-valine (EX-valine), malate dehydrogenase ubiquinone-8 reaction (ME1-2), pyruvate dehydrogenase reaction (PDH), phosphoenolpyruvate carboxylase reaction (PPC), malate dehydrogenase reaction (MDH), extracellular transport of acetate (EX-acetate), extracellular transport of formate (EX-formate), isocitrate lyase reaction (ICL), isocitrate dehydrogenase—NADP reaction (ICDHyr), oxogluterato deshidrogenasa reaction (AKGDH), pyruvate formate lyase (PFL), aconitase (ACONTb), citrate synthase (CS), fumarate (FUM), malate (MAL), oxaloacetate (OAA), succinate (SUCC), succinyl-CoA (SUCCoA), 2-oxoglutarate (AKG), isocitrate (ICIT), cis-aconitate (ACON), citrate (CIT), glyoxylate (GLX), phosphoenolpyruvate (PEP), pyruvate (PYR), Acetyl-CoA (ACCoA), 2,3-Dihydroxy-3-methylbutanoate (23 dhmb). Gene names are shown in green. Quantitative analysis of flux frequency shows that, of all tested mutants, the double ace F -mdh in silico strain exhibited the highest flux through the ACLS, DHAD and Ex-2KIV reactions ( Figure 4 ). These results strongly suggest an increased flux through the BCAA pathway, which is in agreement with results from previous studies dealing with L-valine production ( Park et al., 2007 ). Similarly, the ace F- mdh double mutant has a higher frequency and metabolic flux through the ME2 reaction. This should significantly contribute not only to providing additional levels of NADPH to the KARA reaction ( Figure 1 ), but also to increasing pyruvate levels ( Noronha et al., 2000 ), thus resulting in higher levels of 2-KIV. Finally, the double ace F -mdh mutant also exhibited a significantly lower acetate secretion compared to the other mutants, which greatly enhances this strain’s potential. Overall, we found that the double ace F -mdh was the most promising. 3.3 Carbon flux rerouting towards pyruvate promotes L-valine production in E. coli W We constructed a 2-KIV biotechnological platform comprising three major stages: 1) deletion of the ace F and mdh genes to re-route the carbon flux , 2) overexpression of the 2-KIV biosynthesis pathway and 3) removal of L-valine negative feedback ( Figure 5 ). FIGURE 5 Schematic diagram of E. coli W engineering for 2-ketoisovalerate (2-KIV) overproduction. (A) : Main genetic interventions addressed to construct a 2-KIV overproducer E. coli W strain, (B) : Engineering E. coli W strains. Phosphotransferase system (PTS), Glucose-6-phosphate (Glucose-6P), acetylcoenzyme A (ACCoA), Tricarboxylic Acid Cycle (TCA cycle). Gene names are shown in italic. First we rerouted carbon flux towards pyruvate using scarless genome editing ( Kim et al., 2014 ) to sequentially delete ace F and mdh . We then monitored growth performance of the resulting E. coli W1288 strain using glucose and lactose as sole carbon sources and compared it with the parental strain. We found significant differences in the engineered strain’s growth performance irrespective of the carbon source used in shake flask assays ( Figure 6 ). For instance, E. coli W1288 exhibited lower substrate consumption and biomass production using glucose and lactose. As predicted in silico, deletion of ace F and mdh putatively resulted in lower levels of Acetyl-CoA in this phenotype and, therefore, TCA blockage thus explaining the growth performance observed. We registered significantly higher amounts of L-valine in cultures of E. coli W1288 using lactose as carbon source (up to 0.2 g/L) ( Figure 6 ). This behavior was in line with in silico predictions and it strongly suggested that the double deletion increased flux through the BCAA pathway. FIGURE 6 Kinetic of growth, substrate uptake and 2-ketoisovalerate and L-valine production of E. coli W (A) and E. coli W1288 (B) in shake flask experiments using glucose and lactose. Error bars indicate the difference between replicate cultures. 3.4 Engineering a 2-KIV E. coli W overproducer strain We assembled a synthetic pathway to express genes als S, ilv C and ilv D under heterologous expression driven by the XylS/ Pm system ( Figure 7 ). ilv C and ilv D encode the ketol-acid reductoisomerase and dihydroxy-acid dehydratase, which are responsible for synthetizing 2-KIV from acetolactate, while als S from Bacillus subtilis encodes a L-valine-insensitive acetolactate synthase broadly used for L-valine production ( Felpeto-Santero et al., 2015 ; Hao et al., 2020 ). The synthetic pathway was constructed using Golden Standard technology ( Blázquez et al., 2022 ) in conjunction with a high-copy-number host vector ( Figure 7 ). The resulting plasmid, pKIV, was further expressed in wild-type E. coli W and E. coli W1288, thus yielding the E. coli W1294 and E. coli W1262 strains, respectively ( Figure 5 ). FIGURE 7 Sequential modular assembly of pKIV plasmid from basic DNA parts by using Golden Standard (See method). te1: T1 Terminator, L1, L2, L3 and L4: linker terminals and promoters, te6: BBa B1006 Terminator, std: consensus RBS, Ap: ampicillin resistance, Km: kanamycin resistance, Gm: gentamycin resistance, fusion sites of BsaI (blue square) and BpiI (yellow square) enzymes. Despite no toxicity due 2-KIV has been previously reported ( Zhou et al., 2022 ), we found that irrespective of the carbon source, expression of pKIV placed a metabolic burden on carrier strains that appeared to be greater in the wild-type genetic background of E. coli W1294 ( Figure 8 ). Under production conditions, final biomass of W1294 was halved, while only a slight reduction was observed with the W1262 strain ( Figures 6 , 8 ). Beyond the potential metabolic burden induced by the replication and maintenance of pKIV ( Fakruddin et al., 2013 ), we noticed a significant reduction in W1294s substrate consumption compared to the wild-type strain which contributed to its poor growth performance. Production of L-valine in large amounts (around 1 g/L) is likely the reason behind W1294s poorer growth performance and it suggests the absence of feedback inhibition of AHAS by L-valine using als S from B. subtilis. \n FIGURE 8 Kinetic of Growth, substrate uptake and 2-ketoisovalerate and L-valine production of E. coli W1294 (A) and E. coli W1292 (B) in shake flask experiments using glucose and lactose. Error bars indicate the difference between replicate cultures. Regarding the W1262 strain, we observed significant levels of L-valine production with minimal changes in growth and substrate consumption ( Figure 8 ). Interestingly, the combination of carbon rerouting towards pyruvate, overexpression of the 2-KIV pathway and removal of feedback inhibition concurring in strain W1262 resulted in large titers of 2-KIV (2.2 ± 0.08 with glucose and 2.4 ± 0.00 g/L with lactose) ( Figure 8C ). In addition, we did not found side compounds such as acetate, formate and alcohols in the supernatants. The absence of byproducts suggests that 1) all the pyruvate available was funneled to 2-KIV and L-valine and 2) genetic modifications triggered no evident metabolic overflows, which is in agreement with in silico predictions. 3.5 Production of 2-KIV from non-conventional feedstock using non-conventional microbial cell factories The main aim of our work was to produce 2-KIV efficiently using whey (specifically the lactose contained in whey - see Materials and methods) as a non-conventional feedstock. To this end, we assessed the growth performance and 2-KIV production potential of strains W1288, W1294 and W1262 using whey lactose as sole carbon and energy source. Overall, we observed minor changes in terms of growth and L-valine production compared with the previous experiments performed with glucose and lactose. It was also noteworthy that the lactose in the whey was not fully consumed, especially not by strains harboring the pKIV plasmid, i.e., W1294 and W1262 ( Figure 9 ). In what seems to be the corroboration of the suitability of whey lactose as an optimal carbon source for E. coli W, we found that strain W1262 delivered the greatest titers (3.22 ± 0.07 g/L 2-KIV and 1.40 ± 0.04 g/L L-valine) with whey lactose as carbon source, it is worth mentioning that, in all cases, a pH reduction in the medium between 6-6.5 was observed, which suggests that using a buffered medium with greater ionic strength would help to maintain a stable pH. Overall, we obtained up to 4.62 g/L of products with 3.9 g/L of whey lactose ( Figure 8 ). These high yields are probably related to the presence of additional nutrients in the whey, i.e., up to 0.06% protein and 0.02% fat ( Carranza-Saavedra et al., 2021 ; Tsermoula et al., 2021 ). Altogether, these results highlight the potential of whey lactose as feedstock for value-added compounds derived from pyruvate, such as 2-KIV and L-valine. FIGURE 9 Kinetic of growth, substrate uptake and production of 2-ketoisovalerate and L-valine with E. coli W1288, E. coli W1294 and E. coli W1262 in shake flask experiments using whey’s lactose. Biomass production and carbon source uptake (A) , biosynthesis of products (B) . Error bars indicate the difference between replicate cultures. Finally, we compared our results with previously published works where authors had used alternative microbial cell factories and feedstocks ( Table 2 ). Although a direct comparison is challenging due to differing methodologies and operation modes, we found that product yields per unit of biomass using E. coli W1262 exceeds yields reported for K. pneumoniae , C. glutamicum and E. coli B0016 irrespective of the carbon source, i.e., glucose, lactose and whey lactose ( Table 2 ). Despite recording lower substrate-to-product conversion yields (Yp/s) with glucose and pure lactose than in recent studies with P. putida and E. coli B0016 ( Batianis et al., 2022 ; Zhou et al., 2022 ), in the presence of whey lactose we achieved the highest Yp/s reported so far. This strongly highlights the potential of whey, a non-conventional feedstock, as a promising carbon source with the necessary potential to reduce production costs in microbial fermentation processes. TABLE 2 Production of 2-KIV by different strains. Strain Operation mode Medium Time (h) Total substrate uptake 2-KIV (g/L) Yx/s (g/g) Yp/s (g/g) Yp/x (g/g) References \n Corynebacterium glutamicum \n Fed-batch bioreactor CGXII 56 82 g/L glucose approx.; 24 g/L potassium acetate approx 25.56 0.20 0.31 1.58 \n Krause et al. (2010) \n 20 g/L ammonium sulfate; 1 g/L yeast extract; 5 g/L urea; 0.2 mg/L biotin \n Corynebacterium glutamicum \n Fed-batch bioreactor CGXII 44 82 g/L glucose approx.; 5 g/L potassium acetate approx 35.00 0.14 0.18 1.34 \n Buchholz et al. (2013) \n 10 g/L yeast extract; 10 mM L-valine, L-isoleucine and L-leucine \n Klebsiella pneumoniae \n Batch bioreactor 5 g/L yeast Extract; 4 g/L corn steep liquor; 5 g/L (NH4)2SO4; 0.4 g/L KCl and 0.1 g/L MgSO 4 \n 26 81 g/L glucose approx 17.40 0.05 0.21 4.68 \n Gu et al. (2017) \n \n Pseudomona putida \n Batch shake flask M9 24 12 g/L glucose approx.; 0.12 g/L acetate approx 0.81 0.07 approx 0.40 0.95 approx \n Batianis et al. (2022) \n \n E. coli B0016 Fed-batch bioreactor M9 26 130 g/L glucose approx 55.8 0.15 approx 0.55 2.93 approx \n Zhou et al. (2022) \n 5 g/L yeast extract; fed with 90 g/L yeast extract and 15 g/L peptone \n E. coli W1262 Batch shake flask M9 24 10.2 g/L glucose 2.18 0.04 0.21 5.17 This study 1 g/L yeast extract \n E. coli W1262 Batch shake flask M9 24 9.0 g/L lactose 2.41 0.05 0.27 5.15 This study 1 g/L yeast extract \n E. coli W1262 Batch shake flask M9 24 3.9 g/L whey’s lactose 3.22 0.08 0.81 9.71 This study 1 g/L yeast extract"
} | 7,299 |
26366145 | PMC4553704 | pmc | 2,310 | {
"abstract": "A novel switchable adhesive, inspired by the gecko's fibrillar dry attachment system, is introduced. It consists of a patterned surface with an array of mushroom-shaped pillars having two distinct heights. The different pillar heights allow control of the pull-off force in two steps by application of a low and a high preload. For low preload, only the long pillars form contact, resulting in a low pull-off force. At higher preload, all pillars form contact, resulting in high pull-off force. Even further loading leads to buckling induced detachment of the pillars which corresponds to extremely low pull-off force. To achieve the respective samples a new fabrication method called double inking is developed, to achieve multiple-height pillar structures. The adhesion performance of the two-step switchable adhesive is analysed at varying preload and for different pillar aspect ratios and height relations. Finally, the deformation behavior of the samples is investigated by in situ monitoring.",
"conclusion": "4. Conclusions In this paper we present a bioinspired switchable adhesive where the adhesive state can be switched by application of compressive preload. The system has—in contrast to other switchable adhesives—three adhesive states, namely low adhesion at low preload, high adhesion at medium preload, and very low adhesion at high preload. To obtain such an adhesive we developed a new fabrication process using two subsequent inking and dipping steps, resulting in samples with pillars of two distinct lengths. By in situ observation we could show that the low adhesive state is caused by attachment of only LP, the high adhesive state by full contact of all pillars, and the very low adhesive state by collective buckling of the pillars. Adhesion experiments with different AR and testing velocity showed that adhesion drops with increasing AR and testing velocity, which may be explained by viscoelastic behavior of the chosen elastomeric material. We also presented a simple model which allows estimation of the different adhesive states based on single pillar measurements. For future studies, an optimized and automated fabrication method might lead to a higher reproducibility in between different samples. Also, by fabrication of samples having pillars of multiple lengths, more than three adhesive states may be accessible. Such samples could give access up to an almost continuous switching behavior. These new adhesives may find various applications, for example, in transportation, handling, and robotics.",
"introduction": "1. Introduction Bioinspired adhesives, as, for example, found on gecko toes, have fascinated mankind since a long time.[ 1 ] Especially during recent years there have been notable breakthroughs in understanding the mechanisms of bioinspired adhesives and mimicking such biological dry adhesion systems.[ 2 , 3 ] The reason for the large attention of the so-called “gecko effect” is due to the unique combination of adhesive properties: high adhesion, easy detachment, reversibility, and cleanliness. The dry adhesion system of geckos is based on a complex hierarchical structure of long tilted keratinous stems, which branch into very fine anisotropic contacts.[ 4 ] This complex geometry allows the gecko to form a large contact area even with rough surfaces, thus enhancing the adhesive interaction. Especially the anisotropy of the structures in combination with the biomechanics of their locomotion is responsible for the switchability of the adhesion system,[ 5 ] namely quick attachment and detachment, which is one of the key properties of bioinspired dry adhesives. Numerous approaches to mimic gecko-inspired adhesives have recently been published.[ 3 , 5 , 6 ] Lately, the focus was set to fabrication of switchable bioinspired adhesives. For example, Northen et al. succeeded in developing a composite consisting of an adhesive polymer and magnetic nickel cantilevers.[ 7 ] By means of a magnetic field they were able to change the orientation of the cantilevers and thus the adhesive performance. Paretkar et al. showed that adhesion of elastomeric pillars can be switched off by mechanical overloading which causes reversible buckling of the structures.[ 8 ] Del Campo and co-workers have used a liquid crystal elastomer which allowed to approach and retract an array of pillars, either allowing or preventing them to form adhesive contact.[ 9 ] By applying a low-voltage pulse, Vogel and Steen succeeded in controlling the surface tension force within liquid bridges to generate switchable adhesion.[ 10 ] Nadermann et al. used a metastable, film-terminated, structure which allowed them to switch between a collapsed and an uncollapsed state using air pressure.[ 11 ] The approach to pattern a wrinkled surface by Jeong et al. led to an adhesive which was switchable upon stretching due to a change in orientation of the adhesive structure along the surface caused by the change in waviness of the wrinkles.[ 12 ] Shape memory polymers allow a specific change of their material properties by an external stimulus such as light or temperature. Reddy et al. used a shape memory polymer where the fibrils were mechanically bent and recovered by increasing temperature,[ 13 ] thus changing the adhesive performance. Despite these promising approaches, there is still a challenge to improve the controllability and reversibility of switchable adhesion. In the present study, we investigated a novel aspect of switchability in bioinspired adhesives, namely to achieve a switchable adhesive with more than two adhesive states. While current bioinspired switchable adhesives usually have a high and a low adhesion state, we have designed a switchable adhesive, which allows switching mechanically between “low,” “high,” and “very low” adhesive state. This multistep switchable adhesion is achieved by macrofibrils with defined tip geometry and two distinct lengths. Depending on the applied compressive load, the number of contacting fibrils can be controlled. This concept may lead to the development of adhesion systems which allow precise control of the pull-off force as a function of preload.",
"discussion": "3. Discussion Earlier publications have reported on various different ways to fabricate mushroom shaped tips. With our new method we succeeded in both increasing the diameter of selected mushroom pillar tips and, at the same time, increasing the pillar length. The resulting samples were tested for their adhesive properties and exhibited the expected three different adhesive states. 3.1. Low Preload The longest pillars make contact with the probe, causing a medium pull-off force. The seven LP were identified to first form contact with the probe directly by in situ observation and indirectly by receiving seven distinct detachment peaks during pull-off, where each peak corresponds to the detachment of an individual pillar as can be seen in Figure 2 a. In principle the adhesive properties at low preload can be controlled by the number of LP chosen for the sample. With increasing number of LP, the pull-off force and work of adhesion is expected to increase, meeting a natural limit at the values achieved for samples with only one distinct length of pillars. 3.2. Medium Preload Increasing the compressive preload causes the LP to be compressed in such way that SP also form contact. This sudden increase in contact elements results in two effects. First, the effective stiffness of the sample changes and becomes stiffer. This is reflected in the suddenly increasing slope in the compressive part of the force–displacement curve in Figure 2 a. And second, the overall contact area is suddenly increased, leading to a higher pull-off force, which is reflected in the absolute value of the pull-off force, the increase in the tensile part of Figure 2 b which corresponds to the work of adhesion, and in the increasing number of detachment events as shown in Figure 2 b. While the pull-off force for low preload is expected to be only dependent on the number of LP, the pull-off force at medium preload where LP and SP are in contact depends on both the ratio of LP to SP, on the height difference between SP and LP and on their AR. 3.3. High Preload If a certain preload range is overcome, the pillars buckle as can be seen in Figure 2 c. In earlier studies, the buckling event was closely investigated and it was found that a minimum AR is necessary for buckling as well as that buckling leads to a loss in tip contact which results in almost complete loss in adhesion.[ 8 , 14 ] A similar effect is found here, where the buckling reduces adhesion significantly. 3.4. Analysis of Adhesive Properties The following equations describe the mechanical response of the adhesive as a function of geometric parameters and of adhesive and mechanical properties. Consider the preload P p to be below the threshold P p,1 where the SP form contact with the probe. The pull-off force P c is then defined simply by the number of LP, n LP , multiplied with the adhesive force generated by each pillar F LP : (1) As soon as the critical threshold P p,1 is overcome a number of SP, n SP , form contact with the probe and contribute with an additional adhesive force F SP to the overall adhesion. However, the LP need to be compressed to allow the SP to form contact. Thus, elastic energy is stored in the LP which reduces adhesion. The force generated by the stored elastic energy is identical to the force necessary to compress the LP, F compr,LP , multiplied by the number of LP, thus: (2) whereas P p,2 is the preload at which the structures buckle. Finally, at high preload beyond the buckling preload P p,2 the pull-off force drops to the buckling pull-off force P c,buck , or: (3) Thus, by measuring the adhesive force of a single SP and a single LP and extracting the adhesive force and the load necessary for a certain compression may be sufficient to describe the elastic deformation behavior and the adhesive properties of the samples with pillars of different lengths. The adhesive forces F LP and F SP are a function of tip radius r and AR,[ 15 ] while the compressive force F compr,LP mainly depends on the AR and the Young's Modulus E .[ 16 ] Note that no coupling between the pillars is assumed in this simple model. Also, the collective buckling of pillars with different lengths is difficult to determine. The results in Figure 3 indicate that both the AR and the measurement velocity influence the adhesive performance. We found that an increase in AR leads to a decrease in adhesion. While the difference between AR 4.0 and AR 4.5 are negligible, the samples with AR 5.0 show a notable lower adhesion. Earlier studies on micropatterned surfaces have reported that a higher AR results in higher adhesion,[ 17 ] if the samples are tested with a spherical probe. This can be explained by the reduced effective modulus of the high AR structures, which allows a deeper indentation of the spherical probe into the structures and, thus, resulting in a larger contact area at similar preload. A more recent study using macroscopic pillars and a flat probe revealed that, above a certain AR, no increase of adhesion as a function of AR is found.[ 15 , 18 ] Thus, it is surprising that in the present study adhesion is reduced with increasing AR. A possible explanation for our results is that LP are less stable with respect to bending compared to shorter pillars. This might influence the fabrication of double mushroom tips, especially the second to last step indicated in Figure \n 4 . A decreased bending resistance might lead to a slightly tilted mushroom tip which will certainly decrease adhesion. Figure 4 Schematic of mushroom tip preparation. A substrate is silanized and spin coated with liquid PDMS. A previously prepared pillar sample is dipped into the liquid PDMS layer forming droplets on the pillar tips. After pressing the pillars onto a heating stage, mushroom tips are formed. Subsequently, droplets of liquid pillars are applied to selected pillars and the curing process is repeated, resulting in mushroom shaped pillars of different length. The results in Figure 3 also show that adhesion depends on the measurement velocity. While for the AR 4.0 structures the pull-off forces are similar for both tested measurement velocities of 20 and 60 μm s −1 , tests on AR 4.5 and AR 5.0 samples show that the pull-off force is significantly reduced if tested with the higher velocity. This effect is surprising as it is usually found that adhesion increases with increasing testing velocity due to an increased viscoelastic loss.[ 19 ] Following the argument of less stable pillars with increasing AR and a resulting imperfection of the pillar tip alignment during sample preparation for high AR pillars as stated in the previous paragraph, this effect may be explained as follows. With increasing testing velocity, the time to form intimate contact is reduced. PDMS is known to show viscoelastic behavior, especially for loading frequencies between 0.1 and 100 Hz.[ 20 ] Thus, increasing testing velocity may deteriorate the conformation of the pillar tip to the probe. This would result in a decrease of adhesion with increasing testing velocity. Indeed, the simpler to fabricate AR4 structures show basically no velocity effect, while the more difficult to fabricate AR4.5 and AR5 pillars show a significant velocity effect. Thus, it is likely that the velocity effect may be caused by the sample preparation. One final remark has to be given with regard to the scatter of the experimental data. From the data given in Supporting Information it becomes clear that the relatively large error bars in Figure 3 do not origin from poorly defined adhesive properties but are caused by the variation between different samples and/or the tested sample rotation. The switching behavior for single tests is usually much better defined as suggested by Figure 3 and is well represented with the result shown in Figure 2 . The reason for the large variety between different samples is most probably due to the manual fabrication method of the double mushroom. Precise positioning of droplets with exactly the same size and preparation of LP with exactly the same length is a process which is almost impossible by manual fabrication. Thus, the adhesive properties at different switching levels will be directly influenced, causing the observed large overall error."
} | 3,607 |
35252703 | PMC8892856 | pmc | 2,312 | {
"abstract": "A surface with a\ngradient physical or chemical feature, such as\nroughness, hardness, wettability, and chemistry, serves as a powerful\nplatform for high-throughput investigation of cell responses to a\nbiointerface. In this work, we developed a continuous antifouling\ngradient surface using pyrogallol (PG) chemistry. A copolymer of a\nzwitterionic monomer, sulfobetaine methacrylate, and an amino monomer,\naminoethyl methacrylate, were synthesized (pSBAE) and deposited on\nglass slides via the deposition of self-polymerized PG. A gradient\nof pSBAE was fabricated on glass slides in 7 min in the presence of\nan oxidant, ammonium persulfate, by withdrawing the reaction solution.\nThe modified glass slide showed a wettability gradient, determined\nby measuring the water contact angle. Cell adhesion and protein adsorption\nwere well correlated with surface wettability. We expect that this\nsimple and faster method for the fabrication of a continuous chemical\ngradient is applicable for high-throughput screening of surface properties\nto modulate biointerfaces.",
"conclusion": "3 Conclusions In this work, we developed\na simple, fast, and economical method\nto develop continuous pSBAE gradient substrates. Oxidative polymerization\nof PG facilitates immobilization of antifouling pSBAE. The pSBAE gradient\nwas prepared using the draining method in 7 min. Cell adhesion and\nprotein adsorption were well correlated with surface wettability.\nWe expect that this simple and faster method for the fabrication of\na continuous chemical gradient is applicable for high-throughput screening\nof surface properties to modulate biointerfaces.",
"introduction": "1 Introduction Adhesion,\nspread, growth, and differentiation of cells on a surface\nare important aspects for cell responses to biomaterials. A surface\nwith a “gradient” feature, such as roughness, hardness,\nwettability, surface energy, or chemistry, acts as a powerful tool\nfor the systematic investigation of cell behavior in response to physiochemical\nparameters of various biomaterials. 1 , 2 Surfaces with\na gradient of hydrophilicity–hydrophobicity, softness-hardness,\nbiomolecule/polymer concentration, roughness, or pore size have served\nas a high-throughput platform for the investigation of cell interactions\nwith surfaces of biomedical devices. 3 − 8 Numerous techniques have been employed to fabricate gradient surfaces\nsuch as gradual immersion/retraction, 9 microstamping, 10 microfluidic lithography, 11 electrochemical, 6 , 12 plasma treatment, 13 , 14 corona irradiation, 15 , 16 UV photolithography, 17 , 18 chemical degradation, 19 diffusion, 20 and polymer curing. 8 Among these methods, gradual immersion/retraction of samples into/from\na solution of monomers, solvents, or etchants, thus creating a gradient\nbased on immersion/retraction rate, is a simple and common method\nto fabricate a continuous gradient. A gradient of cell adhesion\ncould be achieved on a surface with\na concentration gradient of an antifouling polymer. Antifouling polymers,\nincluding polysaccharides, poly(ethylene glycol) (PEG), poly(vinyl\nalcohol), poly( N -vinylpyrrolidone), and poly(zwitterions),\nare highly hydrophilic and adsorb a great amount of water to form\na barrier to reduce nonspecific cell adhesion and protein adsorption.\nSurface immobilization of antifouling polymers could resist or reduce\ncell adhesion. In recent years, zwitterionic polymers such as poly(2-methacryloyloxyethyl\nphosphorylcholine), poly(sulfobetaine methacrylate), and poly(carboxybetaine\nmethacrylate) have been increasingly used because of their excellent\nlow-fouling properties. 21 − 27 Zwitterionic monomers contain one cationic and one anionic group,\nwhich leads to strongly bound water molecules. 23 , 28 , 29 A previous study showed that grafting poly(sulfobetaine)\nto surfaces reduced nonspecific protein adsorption to less than 0.3\nng/cm 2 from single protein solutions. 30 Zwitterionic polymers have been applied in various\nbiomedical applications because of their excellent antifouling ability. 31 − 41 Antifouling polymers could be conjugated on a substrate in\na gradient\nmanner, thus resulting in a cell adhesion gradient. For example, PEG\nattached to positively charged polylysine was adsorbed on a negatively\ncharged titanium dioxide surface through electrostatic interactions\nin a time-dependent manner; thus, a PEG coverage gradient was created\non titanium dioxide surfaces to modulate cell adhesion and spreading. 42 A poly(methacryloyloxyalkyl-phosphorylcholine)\ngradient surface was created using a corona discharge treatment to\ngenerate a cell gradient. 15 Ren and coworkers\nfabricated two-component gradient of a zwitterionic sulfobetaine polymer\nand a KHIFSDDSSE peptide using the controlled reaction method by solution\ninjection, which was used to study cell adhesion and the migration\nof Schwann cells and fibroblast cells. 43 The objective of this study was to develop a simple and universal\nmethod to create a gradient surface of poly(sulfobetaine) for gradient\ncell adhesion/protein adsorption. Recently, we developed a simple\none-step method for the fabrication of low-fouling coatings of zwitterionic\npolymers via pyrogallol (PG) deposition. 44 PG, a polyphenolic molecule, could spontaneously undergo polymerization\ninto a robust coating on a wide variety of substrates under mild alkaline\nconditions. 45 Similar to coatings based\non dopamine 46 and aminomalononitrile, 47 PG coating has strong and universal interfacial\nadhesiveness to substrates and chemical reactivity toward nucleophiles,\nsuch as amines and thiols. 48 Previously,\nwe applied the mechanism to the immobilization of antifouling polymers,\nsuch as poly(ethylene glycol) and poly(sulfobetaine) on surfaces of\nseveral biomaterials using their copolymers with primary amine monomers\nand then immobilized the copolymers to several substrates via dopamine,\naminomalononitrile, and polypyrogallol coatings to effectively inhibit\ncell adhesion. 44 , 49 − 51 In this study,\nwe would like to extend the coating technology to the creation of\ngradient surfaces. In this study, a copolymer (pSBAE) of sulfobetaine\nmethacrylate\n(SBMA) and 2-aminoethyl methacrylate (AEMA) was synthesized and then\ndeposited onto glass slides via PG deposition. The AEMA moiety provides\nanchorage to polypyrogallol via Michael addition. 52 PG forms polypyrogallol via oxidation, 53 so oxidizing agents were added to expedite the process.\nThe pSBAE gradient was created on glass slides using a solution draining\nmethod. The coating was expedited via the addition of oxidizing agents.\nThe formation of a pSBAE gradient was verified using the water contact\nangle (WCA) measurement, X-ray photoelectron spectroscopy (XPS), and\natomic force microscopy (AFM). The adhesion of L929 and MG-63 cells\nwas investigated as well as the adsorption of bovine serum albumin\n(BSA) on the gradient substrates. Finally, cell adhesion and protein\nadsorption were correlated with the surface gradient.",
"discussion": "2 Results and Discussion 2.1 Synthesis and Characterization\nof pSBAE The pSBAE yield was ∼53%. The water solubility\nof the copolymer\nwas poor but well soluble in 0.1 M phosphate buffer (pH 7.4). It is\nprobably due to inter-zwitterion interactions between the sulfonate\nand quaternary ammonium groups of SBMA, which limits its solubility\nin water. 54 The interactions could be overcome\nin the presence of salts. The molecular weight of the copolymer was\ndetermined as ∼130 kDa using gel permeation chromatography.\nThe monomer composition of the copolymer was determined using 1 H NMR ( Figure S1 in the Supplementary\nMaterials). Peaks at 3.75 and 4.2 ppm are associated with SBMA and\nAEMA, respectively. 49 The ratio of the\nintegral areas of the two peaks was used as the molar ratio of SBMA\nto AEMA in pSBAE. The molar ratio of AEMA in pSBAE was estimated to\nbe 9%, close to 10% AEMA in the monomer mixture. 2.2 Formation of Poly(PG) in the Presence of Oxidizing\nAgents PG polymerization under alkaline conditions with oxidizing\nagents is a time-consuming process, usually requiring several hours.\nIt is not a reasonable duration for the fabrication of gradient surfaces.\nVarious oxidizing agents, such as ammonium persulfate [(NH 4 ) 2 S 2 O 8 ], methanol, sodium periodate\n(NaIO 4 ), and copper sulfate (CuSO 4 ), have been\nshown to facilitate the polymerization of dopamine and polyphenols. 55 − 58 Therefore, we used oxidizing agents to accelerate the PG coating. The oxidative polymerization of PG turns the originally transparent\nsolution dark brown, so the degree of PG oxidization could be visually\nevaluated based on the color of a PG solution. In this study, several\noxidizing agents were tested: ammonium persulfate (APS), sodium persulfate,\nsodium periodate, potassium iodate, copper sulfate, potassium chlorate,\nand 20% methanol. Potassium iodate, sodium periodate, and copper sulfate\nturned the PG solution dark brown immediately after the addition of\nthese oxidizing agents ( Figures S2–S4 ), showing a very strong oxidizing property even at low concentrations.\nFast reactions make it difficult to control the coating process. On\nthe other hand, potassium chlorate and 20% methanol did not significantly\nspeed up PG polymerization ( Figures S5 and S6 ). They turned the PG solution light brown after 3 h. Sodium persulfate\nhad a moderate oxidizing capacity but turned the PG solution turbid\n( Figure S7 ). We found that APS seems to\nmeet our requirements, turning the PG solution dark brown in a couple\nof minutes ( Figure 1 ). The speed of PG oxidation is dependent on the PG/APS molar ratios.\nEqual-molar PG/APS formed brown precipitates very quickly to be suitable\nfor surface coating. The PG/APS solution at a molar ratio of 2 became\ndark brown in ∼3 min without apparent precipitates. The transition\ntime increased further with increasing PG/APS ratios, indicating that\nPG oxidation could be controlled by varying APS concentrations. Thus,\nAPS was chosen as the oxidizing agent for PG/pSBAE coatings. Figure 1 Oxidative polymerization\nof PG (8 mg/mL) in the presence of APS\nfor different PG/APS molar ratios (1, 2, 3, and 4) at pH 7.4. The\nnumbers indicate the incubation times (minutes). Next, we tested the deposition of PG/pSBAE in the presence of APS\nto inhibit cell adhesion. In this experiment, the concentrations of\nPG and pSBAE were fixed at 8 and 40 mg/mL, respectively, with various\namounts of APS. From PG/APS molar ratios of 0.5∼4, cell adhesion\nto PG/pSBAE was significantly decreased compared to PG ( Figure 2 ). The lowest cell adhesion\nappeared on the substrate with a PG/APS molar ratio of 3 (weight ratio\n∼ 1.66). Thus, the ratio was chosen to create a gradient pSBAE\nsubstrate. Figure 2 L929 cell adhesion on the pristine TCPS, PG, or PG/pSBAE-modified\nTCPS. The concentrations of PG and pSBAE were 8 and 40 mg/mL, respectively,\nwith different PG/APS molar ratios. The coating time was 7 min. Value\n= mean ± standard deviation, n = 3. *, p < 0.001 vs PG. 2.3 Fabrication and Characterization of the pSBAE\nGradient Our previous study showed that one-step coating\nof PG/pSBAE could effectively inhibit cell adhesion. 44 Therefore, we tried creating a gradient of PG/pSBAE by\nthe one-step process in the presence of APS, but could not achieve\na constant gradient for cell adhesion. Therefore, two modifications\nof the process were used to improve gradient fabrication. First, the\nglass substrates were premodified in PG solution for 2 min to improve\nthe coating efficiency of the subsequent PG/pSBAE coating. Second,\nin the literature, gradient substrates were frequently created via\nthe immersion and removal method. Thus, we first tried immersing a\nglass slide into the PG/pSBAE/APS solution and gradually pulled the\nglass slide up. However, no constant gradient could be obtained. Thus,\nwe used a different approach in which PG/pSBAE/APS solution was gradually\ndrained from the bottom of the sample tube. We found that the formation\nof the gradient was more consistent and stable, so the PG/pSBAE gradient\nwas fabricated using the draining method (see Section\n4.3 ). Therefore, the glass slide was divided into three areas:\npristine glass, PG, and the gradient ( Figure 10 ). We would like to investigate the\nsurface gradient formation from 0 cm at the border of the PG area\nand the gradient area to the end of the gradient area. Therefore,\nthe surface properties of the modified glass slides were characterized\nfrom the PG/gradient interface (i.e., 0 cm) to the positions that\nwere 0.4, 0.8, 1.2, and 1.6 cm away from the border ( Figure 10 ). However, the 0 cm line,\nwhich was supposed to be PG without pSBAE, was not clear-cut because\nof the fluctuation of the liquid interface, so its surface characterization\nwas performed on the PG area. Gradient formation was first investigated\naccording to the WCAs along the glass slides. The WCA on the pristine\nglass was 65.4°, while the PG coating decreased the WCA to 54.5°\n( Figure 3 ), indicating\nthat the PG coating improves the wettability of the glass. After gradient\ndeposition of PG/pSBAE, the WCA decreased further, varying with APS\nconcentrations. In the presence of 3 mg APS/mL, the WCA decreased\nfrom 54.5° to 39.2°, 35.7°, 32.4°, and 31.1°\nat locations of 0.4, 0.8, 1.2, and 1.6 cm apart from the PG area,\nrespectively. The steepest hydrophilicity gradient appeared in the\npresence of 9 mg APS/mL, resulting in the smallest WCA of 18.21°\nat 1.6 cm. Therefore, 9 mg APS/mL was used for the subsequent gradient\ncoating to resist cell adhesion and protein adsorption. Figure 3 WCA measurement\nat different locations over the gradient glass\nslide: glass, 0 (PG), 0.4, 0.8, 1.2, and 1.6 cm. The gradient was\ncreated on a glass slide from a solution of PG (6 mg/mL)/pSBAE (40\nmg/mL) with different APS concentrations from 3 to 12 mg/mL. XPS was used to determine the elemental composition\nat different\nlocations along the gradient slides. The Si signal comes from the\nglass substrate and is expected to decrease with increasing coating.\nSulfur, which comes from the SB moiety of pSBAE, is an indicative\nelement for the pSBAE coating. It is expected that the sulfur content\nincreases with increasing pSBAE coating. The results showed that the\npristine glass and the PG-coated region do not have any sulfur ( Figure 4 ). No sulfur signal\nwas detected in the 0.4 cm region, although the measurement of the\nWCA indicates an increase in hydrophilicity compared to PG. Sulfur\nwas found at the 0.8, 1.2, and 1.6 cm spots with a value of 1.23,\n1.745, and 2.66%, respectively ( Figure 4 ). The XPS measurement also confirms the formation\nof a PG/pSBAE gradient coating over the glass slide. Figure 4 Elemental compositions\non several locations of the pSBAE gradient\nsurface. The gradient was created on a glass slide from a solution\nof PG (6 mg/mL), pSBAE (40 mg/mL), and APS (9 mg/mL). XPS analysis\nwas performed at several locations over the gradient glass slide:\nglass, 0 (PG), 0.4, 0.8, 1.2, and 1.6 cm. The surface roughness at the different locations (Glass, 0 (PG),\n0.4, 0.8, 1.2, and 1.6 cm) was determined using AFM, as shown in Figure 5 . The surface roughness\nof the PG substrate was 0.46 ± 0.04 nm, which is similar to that\nof the glass (0.40 ± 0.03 nm), indicating a smooth and uniform\ndeposition of PG on the glass surface. The deposition of pSBAE increased\nthe surface roughness with increasing deposition time. We found that\nthe deposition of pSBAE generated linear protuberances. The roughness\nwas increased to a plateau value of ∼2.27 nm at 1.6 cm. Although\nthe PG/pSBAE coating increased the surface roughness, the difference\nwas very small, which might not affect cell adhesion significantly. Figure 5 Gradient\nwas created on a glass slide from a solution of PG (6\nmg/mL), pSBAE (40 mg/mL), and APS (9 mg/mL). The surface roughness\nwas determined using AFM at the different locations: glass, 0 (PG),\n0.4, 0.8, 1.2, and 1.6 cm. n = 3. 2.4 Cell Adhesion and Protein Adsorption to Gradient\nSurfaces The adhesion of L929 and MG63 cells was investigated\non the gradient surface. L929 cells adhered and spread well on PG,\nwhile the number of cells decreased and the morphology of cells became\nround ( Figure 6 A).\nThe cell number and the cellular spreading area decreased along the\npSBAE gradient. We found that compared to the number of cells on PG,\nthe adhesion of the cells was reduced to 69.2, 51.5, 32.1, and 19.8%\nat 0.4, 0.8, 1.2, and 1.6 cm, respectively ( Figure 6 B), indicating that a cell gradient was successfully\ncreated. The difference in cell numbers between 0 and 1.6 cm was ∼5-folds.\nThe cell spreading area also decreased from 634 μm 2 at 0 (PG) to 589, 561, 528, and 482 μm 2 at 0.4,\n0.8, 1.2, and 1.6 cm, respectively ( Figure 6 C). The created pSBAE gradient generated\na cell adhesion gradient with respect to the cell number and cell\nspreading. Figure 6 Adhesion of L929 cells on the gradient substrate. The gradient\nwas created on a glass slide from a solution of PG (6 mg/mL), pSBAE\n(40 mg/mL), and APS (9 mg/mL). (A) Microscopic images of L929 cells\nat different locations. (B) Normalized cell adhesion to 0 (2.13 ×\n10 4 cells/cm 2 ) and (C) Averaged cell area of\nL929 cells. The values were determined randomly at five points at\neach location of a sample. The values from three samples were averaged.\nThe error bars represent the standard deviation of three samples.\n*** p < 0.001, ** p < 0.01,\nand * p < 0.05 vs PG. The response of MG63 cells to the gradient surface was similar\nto that of L929 cells, that is, a decrease in both the number of cells\nand the spreading area along the gradient ( Figure 7 A). The decrease in the adhesion of MG63\ncells was more abrupt along the gradient compared to that of L929\ncells. The cell number at 0.4 cm was less than half of that of 0 (PG)\n(47.8%), and only 10.7% cell adhesion appeared at 1.6 cm ( Figure 7 B). The cell spreading\narea decreased greatly from 1093 μm 2 at 0 (PG) to\n639 μm 2 at 0.4 cm ( Figure 7 C). The cell spreading area was reduced further\nto ∼500 μm 2 from 0.8 to 1.6\ncm. Figure 7 Adhesion of MG63 cells on the gradient substrate. (A) Microscopic\nimages of MG-63 cells at different locations. (B) Normalized cell\nadhesion to that on 0 (2.62 × 10 4 cells/cm 2 ) and (C) Averaged cell area of MG63 cells for the condition APS\n= 9 mg/mL. The values were determined randomly at five points at each\nlocation of a sample. The values from three samples were averaged.\nThe error bars represent the standard deviation of three samples.\n*** p < 0.001 and ** p < 0.01\nvs. PG. We successfully applied PG chemistry\nto fabricate a gradient substrate\nfor cell adhesion. The gradients created using different APS ratios\n(3, 6, and 12 mg/mL) have shown a similar trend in mediating cell\nadhesion. The results are shown in the Supplementary Materials ( Figures S8∼S13 ). Next, protein adsorption\nto the gradient surface was investigated.\nFITC-labeled BSA was used as a model protein. Protein adsorption at\ndifferent locations was determined according to surface fluorescence\nand normalized to the fluorescence intensity at PG. The fluorescence\nintensity decreased linearly from PG to the pSBAE area ( Figure 8 ). The fluorescence at 1.6\ncm was about 40% of that at PG. The result shows that the pSBAE gradient\nsurface resulted in a gradient of protein adsorption. Figure 8 Protein adsorption on\ndifferent regions of the gradient coating\nfor APS = 9 mg/mL. The values were determined randomly at five points\nat each location of a sample. The values from three samples were averaged\nand normalized to the value of PG (0 cm). For poly(sulfobetaine), a zwitterionic polymer, when conjugated\nto a substrate, provides the antifouling capacity of the surface,\nby improving its wettability to attract water molecules to form a\nbarrier to prevent nonspecific protein adsorption and cell adhesion. 44 , 49 , 59 − 61 The surface\ncoverage of poly(sulfobetaine) affects the resistance of the substrate\nto cell adhesion and protein adsorption. 62 In this study, we showed that a gradient of pSBAE could be formed\nusing the draining method. The hydrophilicity of the glass slides\nincreased with increasing\nincubation time in PG/pSBAE solution. The pSBAE gradient substrate\nresults in gradients of cell adhesion and protein adsorption. Surface\nhydrophilicity increases with an increasing amount of bound pSBAE.\nTo consider the relationship between surface wettability and biological\ninteractions, we correlated surface wettability with cell adhesion\nand protein adsorption. The adhesion for both L929 cells and MG63\ncells was correlated linearly with the WCAs ( Figure 9 ). Similarly, BSA adsorption also showed\na nice positive correlation with the WCAs in a polynomial pattern\n( Figure 9 ). Our results\nindicated that a pSBAE gradient was successfully fabricated using\nthe draining method for the modulation of biointerfaces. Figure 9 Correlation\nbetween the WCA and cell adhesion/protein adsorption."
} | 5,239 |
40248242 | PMC12004222 | pmc | 2,316 | {
"abstract": "A promising approach for developing self-healing polymer materials involves the formation of reversible dynamic crosslinked networks. However, most self-healing systems require external stimuli such as temperature or pressure, to achieve effective healing. In this study, we successfully developed self-healing epoxidized natural rubber (ENR) materials that do not depend on external stimuli by incorporating borax as a crosslinking agent. The results demonstrated that borax facilitates self-healing efficiency through the reversibility of borate–ester and hydrogen bonds. ENR with 10 phr borax exhibits remarkable self-healing performance, achieving 94% efficiency in tensile strength and 95% efficiency in elongation at break after healing at ambient temperature for 24 h. Moreover, the effect of borax loadings on the chemical structure, thermal stability, and mechanical properties of the crosslinked ENR materials were investigated. These findings highlight the crucial role of borax in imparting self-healing properties to ENR without requiring external stimuli, offering an effective approach for developing self-healing elastomers.",
"conclusion": "Conclusions The incorporation of borax in ENR induced a dynamic crosslinked network through borate–ester and hydrogen bonding. The borax incorporation significantly influenced the degree of crosslinking, mechanical properties and self-healing performance of the ENR/borax films. The optimum concentration of borax was determined to be 10 phr. The dynamic nature of borate–ester and hydrogen bonds allowed chain interdiffusion and bond network reconstruction after damage, resulting in remarkable self-healing properties of the ENR/borax films. Notably, for ENR with 10 phr of borax, their self-healing efficiencies in all mechanical properties were higher than 90% after healing at room temperature for 24 h. Furthermore, these films exhibited outstanding self-healing efficiency (>80%) after just 4 h of healing at room temperature. Extending the healing time slightly promoted the healing efficiency of the crosslinked ENR/borax materials whereas increasing the healing temperature hindered their self-healing ability. This study presents a promising strategy to fabricate a functional rubber material with the self-healing properties.",
"introduction": "Introduction The increasing production and consumption of polymer-based materials have raised severe environmental concerns due to their non-biodegradable nature and inadequate waste disposal such as landfilling and incineration. 1 Therefore, innovative solutions are required to reduce polymer waste and extend the lifespan of polymers. Moreover, the concept of a circular economy emphasizes waste minimization, reuse, and recycling of polymer products, promoting more sustainable material management. 2 Accordingly, self-healing polymeric materials have become considerably attractive due to their ability to autonomously repair damage and restore their original properties and functionality. 3 These materials offer a promising strategy for sustainability by enhancing the reliability and durability of polymer-based products, while extending their lifespan. Based on their self-healing mechanisms, self-healing polymeric materials can be categorized into two main types: extrinsic and intrinsic healing. Extrinsic healing involves the incorporation of microcapsules or microvascular structures filled with healing agents that are released when damage occurs. 4 However, this approach is limited by the finite supply of healing agents, which restricts the number of possible healing cycles. In contrast, intrinsic healing relies on the introduction of reversible dynamic bonds, such as hydrogen bonds, ionic bonds, aliphatic disulfide bonds, Diels–Alder, and trans -ester bonds. 5–11 These dynamic bonds can rearrange under ambient or specific conditions, enabling effective healing while preserving or even improving the mechanical properties of the polymers. 12 In recent years, intrinsic self-healing approaches have gained significant research interest due to their ability to autonomously repair damage without requiring external healing agents and their ability for multiple healing cycles. 13,14 This makes intrinsic self-healing a promising solution for sustainable and long-lasting material applications. The dynamic bonds can be incorporated into self-healing polymeric materials by modifying polymer backbones, introducing dynamic crosslinkers, or adding functional fillers. These strategies enhance the material's ability to form the dynamic bonds at interfaces of polymer–polymer, polymer–filler, and filler–filler, thereby improving self-healing efficiency and mechanical performance. 15,16 For example, Liu et al. modified polybutadiene with acetoacetyl groups (PBAA) and subsequently crosslinked PBAA with diamines to form vinylogous urethane dynamic bonds. The crosslinked PBAA was reprocessed under 10 MPa at 150 °C for 30 min. 17 Also, Cui et al. incorporated extracted lignin into diglycidyl ether terminated polyethylene glycol (PEG) in the melt phase. The interaction between lignin and PEG resulted in a crosslinked network of both covalent and hydrogen bonds. The hydrogen bond networks provided rapid self-healing, achieving ∼90% recovery within 1 h at ambient temperature. 18 Epoxidized natural rubber (ENR) is a derivative of natural rubber (NR), where epoxide groups are introduced randomly along the rubber chains via in situ epoxidation using formic acid and hydrogen peroxide. 19 The introduction of epoxide rings enriches ENR with enhanced properties, such as solvent and oil resistance, superior wipe grip, gas impermeability, and low rolling resistance. 19–21 Notably, under acidic conditions and high temperatures, ENR undergoes a ring-opening reaction, leading to the formation of diols, hydroxyl groups, and carbonyl groups in the ENR molecular chains. 22 These functional groups contribute to the self-healing properties of ENR through polar interactions. 23,24 Moreover, these oxygen-containing groups can further enhance the self-healing properties of ENR through interactions with other functional groups. For example, Feng et al. 7 developed dodecanoic acid-crosslinked ENR supplemented with a small amount of oligoaniline. This material exhibited self-healing properties with a healing efficiency of 80% after undergoing trans-esterification treatment at 200 °C for 30 min. Furthermore, Xu et al. 25 synthesized ENR composites with citric acid-modified bentonite, achieving restoration of tensile strength and elongation at break to 94% and 96%, respectively, after healing at 150 °C for 3 h. However, achieving sufficient healing properties often necessitates high healing temperatures, which can lead to adverse effects such as thermal degradation, excessive crosslinking, and an increase in glass transition temperature ( T g ) of rubber materials. These factors may negatively impact on the mechanical properties, healing performance, and overall service life of the materials. 26,27 Borax is well-known for its role in crosslinking diol-containing polymers during hydrogel fabrication. 28–30 Researchers have reported the efficacy of borax as a crosslinker in hydrogels, endowing them with their self-healing properties without the need for external stimuli. For example, Wang et al. utilized borax as a crosslinker to prepare polyvinyl alcohol (PVA) hydrogel through borate–ester and hydrogen bonds, imparting excellent mechanical properties. The compressive strength of the PVA hydrogels increased from 2.1 kPa with 2 wt% of borax to 5.1 kPa with 5 wt% of borax. The hydrogels also exhibited reversible sol–gel conversion characteristic, highlighting the dynamic nature of borate–ester and hydrogen bonds, as well as their self-healing properties. 31 This strategy has been widely used to develop self-healing hydrogels. Yan et al. fabricated dual networks of the PVA, sodium alginate (SA), and borax hydrogels. The synergetic reversible interactions, including borate–ester and hydrogen bonds between PVA, SA, and borax, act as the energy dissipated centre, enhancing stretchability, toughness and providing rapid self-healing properties at ambient temperature. 32 Moreover, Farajpour et al. reported that the incorporation of borax into ethylene–propylene–diene terpolymer/Kavlar/carbon fiber composites not only improved mechanical properties and abrasion resistance but also enhanced heat insulator properties. 33 Similarly, Intharapat et al. fabricated boric acid-supported NR through a simple reaction between boric acid and NR containing hydroxyl groups. These hydroxyl groups were obtained via the ring-opening reaction of oxirane rings in ENR. The study revealed that the incorporation of boric acid enhanced thermal resistance and flame retardancy of the rubber materials. 34 Our previous research has pioneered the incorporation of borax into ENR. 35 We found that pH regulations considerably influenced the effect of borax on ENR properties. When prepared under neutral conditions (pH 7), most of the borax was transformed into boric acid (B(OH) 3 ) within ENR/borax films, hampering the effective promotion of crosslinked network formation within ENR. Conversely, under alkaline conditions (pH 11), most of the borax turned into borate anions (B(OH) 4 − ), acting as a crosslinker capable of establishing borate–ester and hydrogen bonds with the diol groups of ENR. Moreover, the ENR/borax films preliminary exhibited self-healing properties owing to the dynamic nature of both bonds. However, there is limited evidence regarding the interactions between ENR and borax, including mechanical properties and self-healing performances. 35 Herein, we synthesized ENR under alkaline conditions (pH 11) using borax as a crosslinker, as at this pH, the majority of borax complex transforms into borate anions (B(OH) 4 − ), enhancing crosslinking efficiency. This study investigated the effect of varying borax loadings (0–30 parts per hundred of rubber, phr) on the crosslinked network, thermal properties, mechanical properties, and self-healing properties of ENR. Additionally, we examined the influence of healing time and healing temperature on the healing efficiency of the ENR materials to determine the optimum conditions for achieving superior self-healing properties.",
"discussion": "Results and discussion Preparation and characterization of ENR/borax films The possible bonding mechanisms within the ENR/borax films are proposed in Scheme 1 . Epoxide, diol, hydroxyl, and carbonyl groups were randomly introduced along the rubber chains through epoxidation and ring-opening side reactions. 22,37 The hydrolysis of borax generated acid–based pairs of boric acid and borate anions, which could form borate–ester bonds with the diol functional groups of polymer chains. 38 The diol complexation of borate anions occurred in two steps: first, the complexation between the diols of the polymer and adjacent hydroxyl groups of borate anions to form mono-chelate (L-B), followed by the formation of tetragonal bis chelate (L-B-L) with other diol groups of polymer chains. 39 Our previous research found that the alkaline condition (pH 11) facilitated the formation of B(OH) 4 − , crucial in the diol complexation between ENR and borax. The results showed that alkaline conditions could more effectively promote a crosslinked reaction between ENR and borax compared to neutral conditions (pH 7). 35 It should be noted that the acidic condition could facilitate epoxidation and ring opening reaction during drying process, resulting in uncontrolled of epoxide and hydroxyl content. 22 Thus, acidic condition does not suitable for synthesis ENR/borax films. Scheme 1 Proposed bonding mechanisms within the ENR/borax films. \n Fig. 1(a) shows FTIR spectra of the neat ENR and the ENR/borax films. The absorption peak located at 835 and 1660 cm −1 corresponds to C C vibration and C C stretching of cis -1,4-polyisoprene, respectively. Moreover, the absorption peaks located at 870 and 1250 cm −1 are assigned to the asymmetric and symmetric stretching vibration of epoxide, respectively, confirming successful epoxidation. 40,41 Furthermore, the peaks observed at 1064 and 1732 cm −1 and the broad peak appearing at 3395 cm −1 are attributed to C–O stretching of ester and aliphatic alcohol, C O stretching, and OH stretching, respectively. 22,42,43 These peaks were indicative of the ring-opening side reaction that may occur owing to the high temperature and acidity during the ENR synthesis. 22,37 After adding borax, new peaks appeared at 661 and 1347 cm −1 , corresponding to B–O–B bending and the asymmetric stretching of B–O–C, respectively, in the borate network. 44,45 The hydroxyl peak (3395 cm −1 ) also shifted slightly to a higher wavenumber (3425 cm −1 ), indicating the formation of hydrogen bonding and borate–ester bonding between ENR and borate anions. 32,45 The Raman spectra were further analyzed to confirm the interaction between ENR and borax, as shown in Fig. 1(b) . Several characteristic peaks of ENR were also observed in Raman spectra, including 870 (C–O–C stretching), 1000 (C–CH 2 stretching), 1130 (C–O), 1248 (C–O–C), 1359 and 1452 (CH 2 deformation), and 1658 cm −1 (C C stretching). 46–48 With the presence of borax, the peak at 1130 cm −1 shifted to 1125 cm −1 and exhibited broadening, attributed to B–O–C bonding. Additionally, the emergence of a new peak at 860 cm −1 was assigned to the B–O stretching of B(OH) 4 − . 49,50 Consequently, Boron complexation chemistry between ENR and borax was investigated by 11 B-NMR. As shown in Fig. 2 , B0 presented no signal of any boron complex due to the absence of borax. While the ENR samples with borax displayed 4 deconvoluted peaks as represented structure of boron species ( Fig. 2(b) ). The signal at 0 ppm corresponds to a mono-chelate complex (L-B) combined with free borate anion. The second peak at 5 ppm is a bis-chelate complex (L-B-L), which refers to a crosslinked form with borax. The third and fourth peaks at 10 ppm and 15 ppm are related to boron atom exchange between boric acid and borate anion and free boric acid. 51,52 The presence of L-B and L-B-L confirmed the formation of borate–ester bonds between ENR and borax. Moreover, the degree of crosslinking was roughly estimated by the ratio of the area peaks (L-B-L/L-B), as reported in Table 1 . It was found that the L-B-L/L-B ratio increased with an increase in the amount of borax until the amount of borax exceeded 10 phr. After that, the L-B-L/L-B ratio decreased. The decrease of the L-B-L/L-B ratio was attributed to an excessive amount of borax (in the form of borate anion) and an increase of L-B complex. It suggested that the suitable amount of borax as a crosslinker in this system was 10 phr. These findings confirmed not only the formation of the crosslink network between ENR and borate anions but also the existence of the borate network within the ENR. Fig. 1 (a) FTIR spectra and (b) Raman spectra of neat ENR and ENR/borax films. Fig. 2 (a) 11 B-NMR spectra and deconvoluted spectra using the Lorentz function and (b) schematic structure of boron species. Table 1 The area peak (%) of the deconvoluted B-NMR spectra and ratio of L-B-L/L-B Samples Area (%) Ratio of L-B-L/L-B 0 5 10 15 B0 — — — — — B5 27.66 43.03 27.49 1.82 1.56 B10 14.01 34.47 24.91 26.61 2.46 B20 19.15 40.28 37.98 2.59 2.10 B30 29.25 43.04 11.38 16.33 1.47 To further investigate the formation of crosslinking between ENR and borax, equilibrium swelling experiments and gel content analyses were conducted. As shown in Fig. 3(a) , the crosslink density of 10B notably increased by 154%, from 1.04 × 10 −6 to 2.65 × 10 −6 mol cm −3 . Beyond a borax loading of 10 phr, no significant change ( p > 0.05) in crosslink density was observed. The gel content followed a similar trend to that of the crosslink density, showing a 37% increase for 10B compared to the neat ENR (0B). However, when borax content exceeded 10 phr, a decrease in gel content occurred, with reductions of 16% and 23% for 20B and 30B, respectively, compared to 10B. This decrease might be attributed to the presence of voids resulting from the washing of unreacted borax (B(OH) 4 − and B(OH) 3 ) and L-B complexes, as shown in Fig. S1, † allowing solvents to permeate and dissolve rubber chains. In Fig. 3(b) , T g , as evaluated from the DSC thermograms of the prepared rubbers, gradually increased with increasing borax loadings. This suggested an enhanced restriction of chain mobility owing to the formation of a three-dimensional crosslinked network through borate–ester bonding and hydrogen bonding between ENR and borax, correlating with the increase in the crosslink density. 51 These results confirmed the behavior of borax as a crosslinker in ENR/borax films. Fig. 3 (a) Crosslink density and gel content, (b) DSC curves, (c) TGA curves, (d) DTG curves of the neat ENR and ENR/borax films and (e) relationship between the residue content of ENR/borax samples and the borax concentration (%). Superscripts (a, b, α and β) denote statistical significance ( p < 0.05) among the tested samples, while T onset and T max abbreviate the onset degradation temperature and the maximum degradation temperature, respectively. Thermal properties of ENR/borax films The TGA and derivative thermogravimetric (DTG) curves, shown in Fig. 3(c and d) , illustrate the thermal degradation behaviors of the ENR/borax materials, while Table 2 provides the onset degradation temperature ( T onset ), the maximum degradation temperature ( T max ), and residues (%) of the neat ENR and the ENR/borax films under a nitrogen atmosphere. The neat ENR (0B) undergoes a two-step thermal degradation process. Its initial thermal degradation, with a T onset of ∼152 °C, was attributed to the removal of chemical residues from the epoxidation process used to prepare ENR. Subsequently, the main degradation state, corresponding to the degradation of the ENR chains, occurred with a T onset of ∼368 °C. 34 On the contrary, the crosslinked ENR exhibited thermal degradation in three distinct stages. The incorporation of borax led to an earlier onset of thermal degradation compared to the neat ENR, as evidenced by a decrease in T onset,1 by ∼15 °C and T max,1 by ∼20 °C. This phenomenon was attributed to the removal of bonded water within borax and the formation of boron oxide. 34,53 The second degradation stage of the crosslinked ENR was similar to that of the neat ENR, indicating comparable degradation behaviour. However, a third degradation step was observed between 400 °C and 500 °C, as indicated by changes in the TGA slope and the shape of DTG curves. This degradation step could be attributed to the complexation between borax and ENR, as confirmed by FTIR, Raman and B-NMR spectra of the ENR/borax materials. Table 2 Thermal degradation temperature and residues of the neat ENR and ENR/borax films a Sample \n T \n onset,1 (°C) \n T \n max,1 (°C) \n T \n onset,2 (°C) \n T \n max,2 (°C) Residues (%) 0B 151.65 ± 0.45 180.44 ± 1.89 368.46 ± 0.18 395.77 ± 0.45 0.82 ± 0.40 5B 134.08 ± 2.67 156.15 ± 0.49 367.62 ± 0.01 392.13 ± 0.04 1.68 ± 0.39 10B 133.45 ± 0.95 158.53 ± 1.34 367.80 ± 0.01 392.00 ± 1.12 2.07 ± 0.18 20B 132.47 ± 1.18 156.0 ± 0.74 366.09 ± 1.68 390.74 ± 0.88 2.67 ± 0.04 30B 142.37 ± 0.87 162.46 ± 0.03 366.95 ± 0.12 392.67 ± 0.22 3.57 ± 0.05 a Abbreviations: T onset , onset degradation temperature; T max , maximum degradation temperature. The observed decrease in thermal stability of the crosslinked ENR, in comparison to the neat ENR, demonstrated that the presence of borax accelerated the thermal degradation of ENR. This observation indicated potential limitations in the application of the crosslinked ENR prepared in this study at temperatures exceeding 100 °C. Therefore, our future research will focus on incorporating reinforcing agents to enhance the thermal and mechanical properties of the crosslinked ENR materials for use at a higher temperature. Furthermore, the crosslinked ENR demonstrated higher char yield contents, compared to the neat ENR, with a linear relationship observed between the ENR/borax residue and borax concentration (%), as shown in Fig. 3(e) . An increase in char residue as increasing borax concentration has also been reported in other studies using borax as crosslinker in polymer such as cellulose, guar gum, and starch/PVA blend. 30,54,55 This increase in the char yield content, associated with a decrease in DTG peak height, could be ascribed to the combined effects of borax dehydration during degradation and the formation of a glassy coating layer of boron oxide on the carbonaceous residue of ENR. These mechanisms could effectively delay degradation by preventing the transfer of heat to the substrate and promoting an increase in char residues, indicating the flame-retardant properties of the ENR/borax films. 34 Mechanical properties of ENR/borax films The effect of borax acting as a crosslinker on the mechanical properties of ENR was evaluated through tensile testing at room temperature, and the stress–strain curves, along with the relevant tensile properties of the neat ENR and ENR/borax films, are shown in Fig. 4 . The one-way ANOVA and post hoc comparison using Turkey's HSD demonstrated that adding borax significantly affected the mechanical properties of the ENR/borax films. Generally, the tensile strength, modulus at 100% strain, and tensile energy of ENR significantly increased with increasing borax content until reaching the maximum at 10 phr. Compared with the neat ENR (0B), the tensile strength, modulus at 100% strain, and tensile energy of 10B increased from 0.42 ± 0.02 to 0.59 ± 0.05 MPa (with an improvement of ∼43%), 0.17 ± 0.02 to 0.26 ± 0.005 MPa (with an improvement of ∼53%), and 0.17 ± 0.02 to 0.20 ± 0.00 J (with an improvement of ∼18%), respectively. However, a decline was observed in tensile strength, modulus at 100% strain, and tensile energy when the borax loading exceeded 10 phr (20B and 30B). This decrease could be ascribed to the excess borax remaining within the ENR matrix, which might act as defects similar to those found in the formation of agglomerated particles in ENR materials and the existence of pores/voids, as shown in Fig. S1. † 56,57 Moreover, the presence of borax led to a decrease in elongation at break of the crosslinked ENR. The elongation at break of 10B slightly decreased to 2452 ± 91%, while 0B did not break even at the extension limit of the machine (2600% strain). The decrease in elongation at break was supported by the FE-SEM images of 0B and 10B, as shown in Fig. S2, † where 0B showed a ductile fracture, while 10B exhibited brittle fracture behavior. This behavior was attributed to the formation of the crosslinked network between ENR and borax, which restricted chain mobility of ENR, as evidenced by the increased crosslink density of the ENR/borax materials. Fig. 4 Stress–strain curves of (a) the original and (b) healed states of the neat ENR and ENR/borax films. (c) Tensile strength, (d) elongation at break, (e) modulus at 100%, and (f) tensile energy of the original ENR/borax films and after healing it for 24 h at room temperature (30 °C). Superscripts (a, b, α, and β) denote statistical significance obtained from one-way ANOVA and post hoc comparison ( p < 0.05) among the tested samples. The symbolic (*) indicates significant difference between the original samples and healed samples using the t -test method. The term N/B refers to a sample that did not break. Self-healing capabilities of ENR/borax films The dynamic nature of the borate–ester and hydrogen bonding has been reported to enable the healing ability of polymer materials. 28,29,39 Additionally, the presence of polar groups in ENR facilitates polar–polar interaction, serving as a driving force for chain diffusion. 23,32,58 Moreover, the dynamic crosslinking of ENR facilitates the chain interdiffusion to the surface, thereby enhancing the self-healing ability of the materials. 59 Herein, we further investigated the effect of borax contents, healing time, and healing temperature of the ENR/borax materials on their self-healing properties. To investigate the effect of borax contents, the samples underwent a 24-hour healing period at room temperature (30 °C). The healing properties of the ENR materials were subsequently evaluated through tensile and macroscopic tests, as demonstrated in Fig. 4 and 5 . Tensile tests were performed on both the original and healed ENR/borax films to quantify the healing efficiency. In the absence of borax, the mechanical properties of the healed sample were significantly low comparing to the original one. This was owing to the absence of the dynamic bonding. Specifically, the healing efficiencies based on tensile strength and tensile energy of 0B were 41.78 ± 0.06% and 1.45 ± 0.07%, respectively. On the other hand, upon the addition of borax, the mechanical properties of the healed sample did not show any significant difference compared to the original samples. This indicated the remarkable self-healing ability of the crosslinked ENR with borax, which could restore its mechanical properties. This effect was attributed to the high chain mobility resulting from the low T g and synergetic effect between borate–ester and hydrogen bonds to effectively reconstruct its dynamic crosslinked network. The healing efficiencies of all mechanical properties increased dramatically to over 90%. Notably, the healing efficiency was calculated based on their original mechanical properties. While the original mechanical properties of 20B and 30B were slightly lower than those of 10B owing to their pores/voids and excess borax, as aforementioned, the amount of bonds based on the crosslink density did not considerably differ from that of 10B. Consequently, the healing efficiencies of 20B and 30B were close to 10B. Moreover, the healing efficiencies of all mechanical properties, except for % elongation at break, of 5B were slightly lower than those of 10B. This could be attributed to the fewer dynamic bonds, as evidenced by the lower L-B-L/L-B ratio and crosslink density of 5B compared to 10B. The healing efficiencies of tensile strength, elongation at break, modulus at 100% strain, and tensile energy of 10B increased to 94.25 ± 0.12%, 95.00 ± 0.10%, 101.31 ± 0.05%, and 97.95 ± 0.17%, respectively, compared to 0B. Additionally, the healing efficiency of the elongation at break of 30B gradually decreased, which might be attributed to the restriction of chain movement resulting from the formation of a crosslinked network between ENR and borax. The macroscopic results are illustrated in Fig. 5 . The ENR/borax films displayed remarkable resilience, exhibiting no signs of fracture during bending and twisting. The 10B sample tolerated a loading of 100 g for over 120 s without showing any signs of fracture. In contrast, the uncrosslinked 0B sample clearly exhibited fractures at the joint section after healing under the same condition ( Fig. 5(a and b) ). Moreover, to delve deeper into the self-healing performance, the ENR/borax samples were coated with silver paint and connected to a light bulb circuit, as shown in Fig. 5(c and d) . When subjected to bending and twisting, the bulb connected to 0B did not light up, and upon stretching with a 50 g mass, the bulb slowly faded over time, extinguishing completely after 15 s owing to a small tear at the joint fraction of 0B. Conversely, the bulb connected to the silver-coated 10B remained illuminated even when subjected to bending or twisting and continued to shine for over 120 s under stretching with a 50 g load. Furthermore, the healing areas of 0B and 10B after a 24-hour healing time were monitored, as presented in Fig. 6 . There was no indication of healing at the healing position of 0B, whereas the two fracture surfaces of 10B exhibited well-fused characteristics, indicating the interdiffusion of ENR chains and the reconstruction of dynamic bonds at their interfaces. Considering the mechanical properties and self-healing efficiencies of 10B, the borax concentration of 10 phr represented an optimal condition for preparing crosslinked ENR materials. Notably, their observed healing efficiency exceeding 100% could be attributed to the low crosslink density and dynamic crosslink network, which allowed the rubber chains to move and entangle easily at fracture, facilitating the rearrangement of the dynamic bonds. Moreover, the effective chain mobility also induced the adjustment of the chain in the bulk into an optimal state. 60 Fig. 5 (a) Digital photographs of the healed 0B, 5B, 10B, 20B, and 30B samples under bending (left) and twisting (right). (b) Appearances of the healed 0B and 10B under tension (loading 100 g). Appearances of the healed 0B and 10B coated with silver paint connected to a circuit with a light bulb under (c) bending (left), twisting (right), and (d) under tension with a load of 50 g. Fig. 6 SEM images of the healing areas in (a) 0B and (b) 10B after healing at room temperature for 24 h. The effects of healing times and healing temperatures on the healing efficiencies of 10B are presented in Fig. 7 . The samples were healed at room temperature for various healing durations, including 1, 4, 12, and 24 h. Notably, after 1 h of healing, the healing efficiencies of tensile strength, elongation at break, modulus 100% strain, and tensile energy of 10B were 61.99 ± 0.21%, 38.02 ± 0.56%, 92.62 ± 0.14%, and 31.52 ± 0.50%, respectively. With an extended healing duration of 4 h, these healing efficiencies considerably increased to 92.02 ± 0.32%, 84.72 ± 0.36%, 123.36 ± 0.16%, and 81.46 ± 0.40%, respectively. The notable increase in the healing efficiencies of 10B with extended healing times aligned with previous findings. 8,61,62 This could be attributed to the lower T g of 10B compared to room temperature and the dynamic crosslink network between ENR and borax. These characteristics permit chain diffusion and interdiffusion of rubber molecules between cut surfaces, followed by the formation of dynamic bonds, including hydrogen and borate–ester bonds between ENR and borax, facilitating self-healing. Consequently, prolonged healing times allowed enhanced interdiffusion and provided ample time for the reconstruction dynamic bonding, leading to enhanced self-healing efficiency of 10B, evident in the healing efficiency exceeding 90% after 24 h of healing. Although extended healing durations could offer improved self-healing efficiency, the healing efficiencies at 4 hour of healing duration were higher than 80%. Which might be sufficient to yield ENR materials with mechanical properties similar to the original ones. Fig. 7 (a) Stress–strain curves and (b) healing efficiency of 10B after healing at 1, 4, 12, and 24 h at room temperature. (c) Stress–strain curves and (d) healing efficiencies of 10B after healing for 24 h at 30 °C (room temperature), 60 °C, and 80 °C. Furthermore, chain mobility can be promoted by thermal treatment. The effect of healing temperature on healing efficiency of 10B, with temperatures varied at 30 °C (room temperature), 60 °C, and 80 °C while maintaining a fixed healing time of 24 h was further investigated ( Fig. 7(b) ). Interestingly, increasing healing temperature reduced the healing efficiency of 10B. For example, the healing efficiencies based on tensile strength, elongation at break, modulus at 100% strain, and tensile energy of 10B were restored to 86.71 ± 0.13%, 63.98 ± 0.16%, 105.49 ± 0.13%, and 62.18 ± 0.11%, respectively, after healing at 60 °C. However, for 80 °C, the healing efficiencies were slightly decreased to 75.65 ± 0.21%, 58.91 ± 0.12%, 97.18 ± 0.12%, and 49.62 ± 0.13%, respectively. This reduction could be attributed to the reversible and exothermic reactions of the borate–ester bond between ENR and borax and the weakening of the hydrogen bonding at higher temperatures. 31,63,64 Although higher chain mobility was found from thermal treatment, the reconstruction of hydrogen and borate–ester bonds might decrease with increasing temperature, resulting in a reduction of the self-healing ability. However, due to the T g of the prepared ENR, which was lower than room temperature, along with the dynamic nature of borate–ester and hydrogen bonds, interdiffusion of rubber chains between the cut surfaces and the reconstruction of the dynamic bond networks occurred. This led to excellent self-healing properties without the need or external heat. \n Table 3 compares the healing efficiencies and healing conditions of the crosslinked ENR obtained from this study and those of ENR with other crosslinkers. The crosslinked ENR materials with borax prepared in this study demonstrated excellent healing efficiency (exceeding 80%) after only 4 h of healing without requiring additional thermal energy. It is noteworthy that the thickness of the sample is a critical parameter influencing self-healing efficiency. In a prior study conducted by Yoon et al. the self-healing mechanisms of star polymers crosslinked with disulfide bonds by varying both the film thickness and the width of cut were examined. Their findings indicated that, with an increase in the initial thickness of the sample, larger cut could be effectively healed because surface tension was proportional to the exposed surface area, and the surface tension served as a driving force for the viscoelastic reflow of the exposed surfaces, facilitating their recontact to recovery damaged areas. 66 The specimens of this work possessed a thinner dimension (0.4 mm) compared to those investigated in other studies (2 mm); however, the observed healing efficiency remains comparable to that of the thicker specimens. This finding suggested that borax could be a promising alternative as a crosslinker for fabricating rubber materials with superior healing properties at room temperature. Table 3 Comparison of the healing efficiencies of ENRs prepared in this study and others a Sample Conventional crosslinker Self-healing driving force External stimuli conditions Sample thickness (mm) Healing efficiency (%) Ref. ENR/CABT (20 wt%) — Trans-esterification 150 °C, 3 h 0.5 TS ≈ 100, EB ≈ 100 \n 25 \n ENR25/ZDMA (30 phr) 1 phr of DCP Ionic bonds 80 °C, 30 min 1 TS ≈ 70, EB ≈ 76 \n 6 \n ENR/CNCs (20 wt%) — Hydrogen bond supramolecular network 50 °C, 12 h 3 TS ≈ 95, EB ≈ 98 \n 65 \n ENR/ZnO-CNF (5 phr) 0.5 phr of DCP and 1.6 phr of sulfur Reversible ionic bond and hydrogen bond 80 °C, 1 h and room temperature, 3 h 0.4 TS ≈ 70, EB ≈ 92 \n 36 \n ENR/CHI/A-CNC (2 phr) — Hydrogen bond supramolecular network 80 °C, 4 h 2 TS ≈ 90 \n 61 \n ENR/borax (10 phr) — Borate–ester bond and hydrogen bond Room temperature (30 °C), 4 h 0.4 TS ≈ 92, EB ≈ 85, TE ≈ 81 This work ENR/borax (10 phr) — Borate–ester bond and hydrogen bond Room temperature (30 °C), 24 h 0.4 TS ≈ 94, EB ≈ 95, TE ≈ 98 a Abbreviations: TS, tensile strength; EB, elongation at break; TE, tensile energy; t-CNs, tunicate cellulose nanocrystals; CABT, citric acid-modified bentonite; ZDMA, zinc dimethacrylate; CNCs, chitin nanocrystals; ZnO-CNF, zinc oxide-modified cellulose nanofibers; CHI, chitosan; A-CNC, APTES-modified cellulose nanocrystals. The mechanical properties of ENR/borax materials are relatively low due to the low crosslink density and the lack of a permanently crosslinked network. Xu et al. found similar phenomena in natural rubber (NR)/zinc dimethacrylate (ZDMA). They fabricated the self-healing NR by introducing an ionic crosslink network from ZDMA. Without the permanently crosslink network from DCP, the total crosslink density (ionic crosslink density) was around 0.33 × 10 −4 mol cm −3 and tensile strength was relatively low (∼0.63 MPa). However, its superior self-healing was obtained after healing at 20 min at ambient temperature. 67 Incorporating conventional crosslinker agents such as sulfur and peroxide in combination with borax, may offer a strategy to enhance the mechanical properties of ENR/borax materials. Notably, a higher concentration of permanently crosslinked networks could reduce chain mobility, decreasing the self-healing capability of rubber or requiring external stimuli to achieve promising self-healing performance. Moreover, the incorporation of reinforcing materials can further improve mechanical properties. In the ENR/borax materials developed in this study, the presence of large polar groups, particularly the hydroxyl groups from borax and oxygen-containing groups in ENR, may improve compatibility between the matrix and polar filler such as cellulose, chitin, and chitosan. These fillers could form dynamic bonds (hydrogen and borate–ester bonds) with ENR and borax, potentially enhancing the mechanical properties of the ENR/borax materials while retaining their self-healing properties. 61,65,68 Therefore, the development of self-healing materials with superior mechanical properties and higher crosslink density would be our future focus."
} | 9,287 |
25888221 | PMC4381447 | pmc | 2,317 | {
"abstract": "Background Quorum Sensing (QS) systems influence biofilm formation, an important virulence factor related to the bacterial survival and antibiotic resistance. In Acinetobacter baumannii , biofilm formation depends on pili biosynthesis, structures assembled via the csuA/BABCDE chaperone-usher secretion system. QS signaling molecules are hypothesized to affect pili formation; however, the mechanism behind this remains unclear. This study aimed to demonstrate the possible role of QS signaling molecules in regulating pili formation and mediating the ability to form biofilms on abiotic surfaces. Results Real-time quantitative PCR analysis showed the expression of the csuA/BABCDE genes distinctly increased when co-cultured with C6-HSL ( P < 0.05). Under the same experimental conditions, expression of BfmS and BfmR was significantly higher than the control strain ( P < 0.05). A subsurface twitching assay showed a switch from a small to a large and structured clone that may result from enhanced twitching motility ( P < 0.05). Transmission electron microscopy analysis of cells lifted from a MH broth co-cultured with C6-HSL showed more abundant pili-like structures than the control strain. We then tested the idea that the addition of a QS signal, and therefore induction of chaperone-usher secretion system genes, provides a greater benefit at higher biofilm densities. An assay for the total fluorescence intensity of the biofilm using Confocal Laser Scanning Microscopy revealed an obvious increase. Conclusion Our study demonstrated that, increased transcription of the BfmS and BfmR genes, QS signaling molecules enhance the expression of the chaperone-usher secretion system, and this expression is required for twitching motility in A. baumannii . The concomitant pili expression and strain twitching allowed A. baumannii to attach easily to abiotic surfaces and form biofilms at an earlier timepoint.",
"conclusion": "Conclusion In summary, we provided data demonstrating how, increased expression of BfmS and BfmR , the QS signaling molecule C6-HSL enhanced expression of the chaperone-usher secretion system, and that bacterial pili are required for twitching motility in A. baumannii . Furthermore, the concomitant pili expression and strain twitching allowed A. baumannii to easily attach to abiotic surfaces and form biofilms at an earlier timepoint. QS signaling molecules are required for cell attachment to solid surfaces and the development of biofilms. Our study describes the biofilm formation of A. baumannii in response to a QS signaling molecule, a finding that provides a comprehensive insight into the role of bacterial pili, which play a key role in bacterial biofilm development.",
"discussion": "Results and discussion Impact of C6-HSL on chaperone-usher complex expression The capacity of A. baumannii to form biofilms is a decisive advantage for its survival in the hospital environmental. Recent studies have linked biofilm development with quorum-sensing pathways and bacterial factors, such as A. baumannii pili [ 15 , 16 ]. It is known that disruption of the csuC and csuE ORFs, which belong to the csuA/BABCDE bacterial pili structure gene cluster, results in non-piliated cells and abolishes cell attachment [ 14 ]. However, the exact mechanism of how QS pathways and csu influence biofilm formation is unclear. To directly examine all the genetic components of the csuA/BABCDE, and their regulators, the BfmS-BfmR regulating system that includes response factor ( BfmR ) and sensor kinase ( BfmS ), we provide data on the comprehensive expression of the pili structure gene cluster and the impact of C6-HSL on this chaperone-usher secretion system. Our results showed expression of bacterial pili structure genes, including csuA/B , csuA , csuB , csuC , csuD and csuE , significantly increased after addition of 100 μmol/L C6-HSL, and the transcript levels of the csuA/BABCDE chaperone-usher complex were increased >1.5-fold over the control group ( P < 0.05, Figure 1 ). Furthermore, at the same experimental conditions, expression of chaperone-usher regulators ( BfmS and BfmR ) were higher than those of the control strain, and the regulators were increased approx 1.33-fold ( P < 0.05, Figure 2 ). Figure 1 \n Transcript levels of genes within the \n csu \n operon. Quantitative RT-PCR assays of ATCC19606 cells grown in LB broth without AHLs (control) or with the addition of 100 μmol/L AHLs (C6-HSL). Transcription of each gene of the chaperone-usher complex were increased >1.5-fold. Figure 2 \n Transcript levels of the csuA/BABCDE chaperone-usher complex regulating genes \n BfmS/R. Quantitative RT-PCR assays of ATCC19606 cells grown in LB broth without AHLs (control) or with the addition of 100 μmol/L AHLs (C6-HSL). Both genes were increased approx 1.33-fold. Subsurface twitching motility and transmission electron microscopy Despite the lack of flagella, A. baumannii can spread rapidly over surfaces, probably due to twitching motility [ 17 ]. Twitching is a form of surface motility mediated by type IV pili [ 18 ]. In Pseudomonas aeruginosa , twitching has been implicated in biofilm development [ 19 ] and a correlation has been found between twitching motility activity and biofilm production [ 20 ]. To determine whether C6-HSL affects bacterial twitching motility, we performed a subsurface twitching assay at the agar/glass interface comparing the diameter of twitching motility zones between the control group and A. baumannii treated with 100 μmol/L of C6-HSL. The results showed that A. baumannii co-cultured with 100 μmol/L C6-HSL had markedly increased movement from 1.75 to 8.38 mm in 24 hours ( P < 0.05, Figure 3 ), which may result from enhanced twitching motility. Transmission Electron Microscopy (TEM) was used to confirm that pili formation in A. baumannii cells was stimulated by C6-HSL. The TEM showed there were abundant pili-like structures around the bacteria treated with C6-HSL, while structures of pili were not observed on the top of the control bacterial cells (Figure 4 ). Figure 3 \n Impact of C6-HSL on twitching motility. Four individual colonies grown on separate plates incubated at 37°C for 24 h. The twitching zones were stained and their diameters measured at least three times. Results shown represent the means ± standard deviations. Figure 4 \n TEM images of an \n A. baumannii \n bacterium grown in solution with or without C6-HSL. TEM images were captured at magnification of ×12,000 (left column) and at ×15,000 (right column). (a) 12, 000-power magnification and (b) 15,000-power magnification of bacterial cell grown on glass slips incubated in MH broth without shaking over night at 37°C. (c) 12,000-power magnification and (d) 15,000-power magnification of bacteria cell growth in MH with 100 μmol/L C6-HSL forms obvious pili-like structures. Confocal laser scanning microscopy If the increase in expression of bacterial pili seen in A. baumannii in response to C6-HSL is responsible for enhanced twitching motility, maintaining this quorum sensing stimulation in pili expression should obviously increase the capacity of the bacteria to form biofilms. With 100 μmol/L C6-HSL conditions, A. baumannii ATCC19606 was shown to form mature biofilms faster than the control group grown in MH medium, which yielded undeveloped biofilms. The results of the confocal laser scanning microscopy (CLSM) show the total fluorescence intensity of biofilms significantly increased in the C6-HSL group and the pili assembling from the surface of the cell was more abundant after C6-HSL stimulation (Figure 5 ). Figure 5 \n Impact of C6-HSL on \n A. baumannii \n ATCC19606 biofilm formation. The three-dimensional reconstruction of biofilms (a) in MH medium and (b) with 100 μmol/L C6-HSL added into MH medium were reconstructed after 4 days cultured without shaking at 37°C. The total fluorescence intensity, including (c) fluorescence volume and (d) fluorescence area was analyzed, and the results were averaged from three randomly selected positions of each sample. A recent study [ 21 ] reported that a strain of A. baumannii with a BfmS knockout displayed a reduction in biofilm formation, loss of adherence to eukaryotic cells and greater sensitivity to serum killing. Our results demonstrated the expression of BfmS and BfmR regulated their target genes, the family of csuA/BABCDE chaperone-usher secretion system genes, to produce and assemble bacterial pili. Taken together, the result that the csuA/BABCDE chaperone-usher secretion system was essential to bacterial loci encoding secretion and surface motility (required in the early steps of biofilm formation) combined with our twitching assay results led us to conclude C6-HSL may promote A. baumannii pilus biosynthesis and assembly, as well as strain twitching ability, thereby ensuing formation of biofilms. However, QS signaling molecules are chemically diverse and many bacteria possess more than one AHL synthase [ 22 ]. In A. baumannii, many other QS signaling molecules have been verified, such as 3-oxo-C12-HSL, 3-hydroxy-C12-HSL and C8-HSL [ 12 , 23 ]. It is important to keep in mind that we focused only on the impact of C6-HSL on A. baumannii , which limited our study. In the future, it would be important to test other AHLs commonly produced by A. baumannii and other bacteria."
} | 2,349 |
24497994 | PMC3908949 | pmc | 2,318 | {
"abstract": "Invasive species may owe some of their success in competing and co-existing with native species to microbial symbioses they are capable of forming. Tall fescue is a cool-season, non-native, invasive grass capable of co-existing with native warm-season grasses in North American grasslands that frequently experience fire, drought, and cold winters, conditions to which the native species should be better-adapted than tall fescue. We hypothesized that tall fescue’s ability to form a symbiosis with Neotyphodium coenophialum , an aboveground fungal endophyte, may enhance its environmental stress tolerance and persistence in these environments. We used a greenhouse experiment to examine the effects of endophyte infection (E+ vs. E−), prescribed fire (1 burn vs. 2 burn vs. unburned control), and watering regime (dry vs. wet) on tall fescue growth. We assessed treatment effects for growth rates and the following response variables: total tiller length, number of tillers recruited during the experiment, number of reproductive tillers, tiller biomass, root biomass, and total biomass. Water regime significantly affected all response variables, with less growth and lower growth rates observed under the dry water regime compared to the wet. The burn treatments significantly affected total tiller length, number of reproductive tillers, total tiller biomass, and total biomass, but treatment differences were not consistent across parameters. Overall, fire seemed to enhance growth. Endophyte status significantly affected total tiller length and tiller biomass, but the effect was opposite what we predicted (E−>E+). The results from our experiment indicated that tall fescue was relatively tolerant of fire, even when combined with dry conditions, and that the fungal endophyte symbiosis was not important in governing this ecological ability. The persistence of tall fescue in native grassland ecosystems may be linked to other endophyte-conferred abilities not measured here (e.g., herbivory release) or may not be related to this plant-microbial symbiosis.",
"conclusion": "Conclusions Our data suggest that regardless of endophyte status, tall fescue growth was stimulated after being burned. Water stress negatively affected tall fescue growth, and did so equally for E+ and E− plants in this experiment. When we did observe significant effects of endophyte on growth of fescue plants, it was opposite that expected, with E− plants having greater tiller length and biomass compared to E+ (one notable exception being belowground biomass prior to the treatments being applied). These results add to the growing body of literature that shows differences in E+ and E− tall fescue plant response to stress may depend on a number of factors (i.e., soil fertility, tall fescue and fungal endophyte genotype interactions, climatic factors, etc.) and are not universal across its range in the Eastern U.S. Our study indicated no apparent role of symbiosis with Neotyphodium in the ability of tall fescue to regrow following fire even under dry conditions such as are commonly experienced in North American grasslands. This result suggests that the persistence of tall fescue in native grassland ecosystems may be linked to other endophyte-conferred abilities not measured here (e.g., herbivory release) or not related to this plant-microbial symbiosis.",
"introduction": "Introduction Plant species may be considered invasive when they successfully spread outside their native range [1] , and may use a number of mechanisms to gain competitive advantage over native species. They may be released from their natural enemies and thus able to thrive better in the new environment [2] , they may simply be better competitors for resources in disturbed environments [3] , and/or they may use plant-soil feedbacks, including so-called “novel weapons” [4] , to negatively impact co-occurring native plants. In some cases, these effects may not be coming from the plants alone, but may be mediated by their association with microorganisms [5] . Many plant functional traits have been linked to association with bacterial and fungal microorganisms (reviewed in [6] ), with fungal endophytes of grasses (in the family Clavicipitaceae) being one of the most studied associations [7] , [8] , [9] . Association with these fungal endophytes has been linked to success of the invasive annual Italian ryegrass [10] , including conferring increased herbicide resistance [11] (but see [12] ). In addition, many of the grass functional traits affected by fungal endophytes could be considered traits that make the grasses more competitive [7] and potentially more able to successfully persist and/or invade novel habitats. Tall fescue ( Schedonorus phoenix (Scop.) Holub) is a non-native C 3 grass species, introduced in the late 1800’s which now covers 14 million hectares in the United States, with its adapted range being the entire eastern U.S. and areas within the Pacific Northwest [13] , [14] . In areas being managed for native warm-season grasslands in North America, tall fescue is considered an undesirable species, in part because it can outcompete native grassland species [15] , and in part due to negative effects on wildlife [16] . Prescribed fires are used widely in management of many grasslands today [17] \n [18] , either alone or in combination with herbicide application [19] . The persistence of tall fescue in what are largely C 4 -dominated grass systems, which often undergo frequent fire and/or water limitation (e.g. [20] ), suggests it is tolerant of these conditions. Other non-native cool-season perennial grasses have been successfully eliminated with prescribed burns [21] , but this has not been the case for tall fescue, which experienced no growth suppression following prescribed burns in the field [16] , [19] , [22] . One factor that may impact growth response of tall fescue to management practices such as prescribed fire is its frequent association with the fungal endophyte Neotyphodium coenophialum (whose presence was unknown in the studies reported in [16] , [19] , [22] ). The tall fescue- Neotyphodium symbiosis is known to increase tall fescue’s stress tolerance over that of endophyte-free (E−) individuals [8] , [23] . Endophyte presence within tall fescue populations can vary across the landscape: within a single field, some areas may have no individuals infected, whereas in other areas, all individuals present are infected. Extensive surveys of tall fescue populations in North America show that on average >50% of tall fescue tillers in an area test positive for endophyte presence [24] , [25] , [26] . Surveys of 17 tall fescue pastures being targeted for restoration across the state of Kentucky found all but one had endophyte infection frequencies (EIF) >80% [27] . ‘Kentucky-31’, the variety of tall fescue that is most common in pastures in this region, has a higher occurrence of fungal endophyte symbiosis than other varieties [25] . The physiological benefits to tall fescue of hosting N. coenophialum are thought to be most pronounced under water [28] , [29] , [30] or nutrient deficiency [31] (but see [32] ), and the fungus may actually serve as a physiological drain or sink when the plant is not under such stress [31] . Endophyte-infected (E+) fescue has been shown to have larger belowground biomass compared to E− tall fescue [33] , [34] , [35] , which could serve as a greater resource from which to recover following management activities that negatively impact aboveground growth of the plant. E+ plants have also been shown to respond to increased nutrient availability (which may be influenced by management) more than E− plants [28] , [32] . Prescribed fire is used as a management tool in many grasslands, and can affect the abiotic and biotic components of the ecosystem. In mesic grasslands effects of fire on the abiotic environment include increased light levels and decreased soil moisture at the surface, and increased nutrient availability [36] . These abiotic effects may in turn affect biotic components of the grassland systems where they occur. Burning has been shown to reduce cover of some non-native species [37] , [38] , [39] and C 3 grasses [38] , [39] , [40] while simultaneously increasing native warm-season grass tillering [41] . The behavior of fire (which determines impacts to the abiotic environment and effectiveness as a management tool) can also be impacted by the vegetation present [42] , and areas dominated by C 3 ’s, like tall fescue, often experience reduced intensity of early spring burns [43] \n [18] , as they may have already begun to grow. In some ways, the effects of prescribed fire on tall fescue might be similar to those of grazing (e.g. removal of aboveground biomass), but comparative effects of these two common grassland disturbances most likely vary depending on the severity or intensity of the events and their distribution in space and time [44] . Prior studies have shown that fungal endophyte presence within tall fescue can alter herbivory [45] and improve plant persistence and performance under grazed conditions (e.g. [46] , [47] , [48] , [49] ). However, we are aware of no studies examining prescribed fire and its interaction with N. coenophialum on tall fescue survival and regrowth. Given that fire has been shown to affect some of the same abiotic parameters also known to be important in determining whether endophyte symbiosis increases tall fescue’s competitive ability or reduces it (e.g., increases light availability, lowers soil moisture, increases nutrient availability), we wanted to explore whether endophyte infection confers greater tolerance to fire. Given previous research indicating physiological benefits of Neotyphodium being most pronounced under water stress, we also incorporated two levels of water availability in the experiment. We designed a controlled greenhouse experiment to test differences in growth following prescribed burn and water availability treatments for E+ and E− tall fescue. This experiment used established tall fescue plants (variety Kentucky-31, either with (E+) or without (E−) the common toxic strain of N. coenophialum ) to which we applied a water availability treatment, providing half the plants with adequate water supply (‘wet’), and half the plants with half as much water (‘dry’). We included an unburned control, a single burn treatment (1x), and a two burn treatment (2x). Based on prior work that suggests the fungal endophyte symbiosis is generally mutualistic, especially under stressful abiotic conditions, we hypothesized E+ plants would have higher biomass and growth compared to E− plants, and that differences would be most pronounced under the dry treatment. We also thought differences between the E+ and E− plant responses would be greatest for those individuals that received the presumably more stressful 2x burn treatment.",
"discussion": "Discussion Of the different treatments imposed during this experiment (endophyte status, water regime, burn), water regime had the most pronounced and consistent effect on tall fescue growth, with those plants under the dry water regime having less growth than those under the wet regime throughout the entire course of the experiment. This result was not surprising given that tall fescue is a C 3 species that cannot perform well during warm temperatures unless adequate water is supplied [55] . However, contrary to our hypothesis and expectations, the effects of water stress imposed by the dry regime were equally detrimental for both E+ and E− plants and across burn treatments. This was surprising, given that others have observed endophyte-related differences in growth responses, especially under dry conditions [28] , [29] , [30] , [50] , [56] , although in some cases these effects have been varied by host plant genotype [50] , [56] . It is possible that if we had controlled for plant genotype (e.g. using genetic clone pairs of E+ and E− individuals) we would have found a different result. It is also possible that our ‘dry’ treatment was not dry enough to stimulate such endophyte effects, although it should be noted it was dry enough to depress tall fescue growth (total biomass) by approximately 32% at the end of the experiment. The relatively cool temperatures of the greenhouse and the frequency of watering (dry treatment received 50% less water than the wet treatment but was applied at the same frequency) may have played a role in not seeing the expected endophyte x water interaction. Endophtye effects on biomass were opposite those expected (E−>E+), and as stated previously, there were no significant interactions with water regime or burn treatment. The only time E+ plants had higher biomass than E− was at the beginning of the experiment for initial root weight. E+ fescue has been shown in a number of cases to have greater shoot [28] , [35] , [57] , [58] , [59] , [60] and root [33] , [34] , [35] , [58] mass compared to E−. However, the magnitude of these differences observed in the previously mentioned studies varied widely (e.g., E+ plants 4.4% [60] to 70% [35] more biomass than E−), and there are a few studies in which no endophyte effect was observed. It is possible that enhanced root biomass reservoir might increase the ability of E+ tall fescue to regrow following aboveground biomass removal, through either fire or grazing. However, in our study, greater root biomass in E+ individuals at the start of the study appeared to have no effect on growth responses following disturbance. Similarly, endophyte presence did not affect leaf elongation, tiller density or dry weight per tiller in studies conducted by Elbersen and West [50] and Newman et al. [61] . It did result in earlier flowering in the Newman et al. study [61] , but in our experiment, date of flower was not significantly affected by endophyte presence either. Some might speculate that endophyte effects are better seen in field studies than in greenhouse studies, but in a climate change experiment in the field at the same research farm where the tall fescue used here originated from (and using tall fescue propagated from seed collected in the plots from which our material came), Brosi also observed relatively few endophyte effects on tall fescue tiller growth [62] . Host plant genotype [29] , [50] , [56] , [60] , [63] , [64] and fungal genotype [29] , [51] , [63] , [64] , [65] have both been shown to influence the dynamics of symbiosis within the tall fescue- N. coenophialum system. It may be that the combination used in our study simply does not exhibit the differences in growth seen in other cases, although it should be noted that our combination (variety ‘Kentucky-31’ and common toxic endophyte) was the same as in some of this previous work and is the most common pairing of tall fescue cultivar-endophyte on the landscape. Physiological benefits of symbiosis with N. coenophialum to host plants can vary depending on soil fertility [28] , [31] , [32] , but the results are not consistent. Cheplick et al. found higher biomass of E+ seedlings compared to E− at high nutrient levels and lower biomass for E+ at low nutrient levels [32] , but Arechavaleta et al. [28] and Malinowski et al. [31] saw higher biomass for E+ at lower nutrient levels and no difference [28] or reduced biomass [31] for E+ at high nutrient levels. The plants used in the current study were grown in the relatively fertile (especially for phosphorus; see [27] ) soil from which they originated. Malinowski et al. [31] and Rahman and Saiga [66] looked at tall fescue growth in response to different P levels, and our results are consistent with what both studies found in high P soils, E+ biomass was lower than E−. It may be that if we had performed this experiment in less fertile soil we would have seen a different outcome with regard to the potential endophyte effects on growth. Given the variability in growth responses in previous studies and this one, it seems there is still much to be learned about the conditions under which fungal endophyte symbiosis is strongly mutualistic for this species. The response of tall fescue to fire might be dependent on its life history (specifically life form and bud characteristics), which Pyke et al. used to characterize plant species’ fire tolerance [42] . With tall fescue being a cryptophyte (sensu [67] ), Pyke et al. predicted the growth response following fire to be neutral or positive if buds are insulated by soil, but negative if buds are closer to the surface and fire temperatures are hot enough [42] . In a review of fire effects on invasive weeds, DiTomaso et al. list cool-season perennial grasses as a category that can be controlled with burning, and while they do not specifically address tall fescue; they do cite successful reductions in Kentucky bluegrass with mid-late spring burns [21] . However, in our study, tiller length was greater for the burned pots (1x or 2x) compared to the control, but biomass (tiller and pot total) was suppressed in 2x compared to 1x or unburned control, so there was a slight reduction in material in the pots burned twice at the end of the experiment (leaf sheaths were the same lengths but apparently not as thick). The rapid growth rate following the second burn was surprising, and likely indicates that given more time prior to harvest (2x burned plants were harvested only 59 days after the second burn, but 1x burned plants were harvested 117 days after the first burn) the 2x burn pots may have regrown all, if not more than, the material lost to fire. Our study did not aim to detect whether fire could actually kill the endophyte, but when we tested for endophyte presence at the end of our experiment, we found more pots in the 2x burn treatment that differed from either 0 or 100% infection (1 such pot in control, 2 pots in 1x, and 7 pots in 2x burn). In fact, the two pots that were excluded from the study were 2x burn that had less than 100% infection. Neotyphodium coenophialum is known to be sensitive to heat, as heat treatments are regularly employed to remove the fungus from infected seed lots [68] . It is possible that prescribed fires may negatively affect the fungus, but more work exploring this topic is required. Tall fescue experiences two periods of growth during a single season with a period in the mid-summer of slow growth [69] , [70] , and it is possible that the timing of fire might interact with the seasonal growth cycle of tall fescue to alter the plant’s response. Based on growth rates prior to burns, this experiment imposed the first burn during the period of early summer growth, and tall fescue took longer to recover compared to when the second burn applied, which occurred as the plants were entering their slower growth mid-summer period. Prescribed fires are most often conducted in February or March in the eastern U.S., which coincided with the timing of our first prescribed burn (during the initial spring growth period). Based on our data, a burn applied at this time appears to allow plenty of time for plants to recover aboveground material, and they can do so in a relatively short period of time (∼3 weeks in this greenhouse experiment). A burn during the mid-summer period (which is when the second burn in this experiment occurred) resulted in rapid recovery in length (2x burn plants had the same tiller length as 1x and unburned control within 2 weeks following the second fire), although it should be noted that the greenhouse was maintained at a daytime temperature lower than ambient summer temperatures which normally produce a “summer slump” or drop in production for tall fescue and other cool-season grasses [69] , [70] . A summer prescribed burn applied to a field dominated by another C 3 grass, Texas wintergrass, resulted in 2x higher yield of that species compared to a winter (Feb/Mar) burn or no burn [40] . A burn during the autumn growing period would allow less time for recovery before the winter dormant period, and might be predicted to reduce tall fescue dominance better over the long-term than summer or spring burns, but Madison et al. found that fall burning did not reduce tall fescue cover [16] . Our results indicate that tall fescue is able to readily recover following fire, even if applied twice in a single growing season, under wet or dry conditions, and irrespective of endophyte status."
} | 5,134 |
27152937 | PMC5113850 | pmc | 2,319 | {
"abstract": "The myxobacteria are a family of soil bacteria that form biofilms of complex architecture, aligned multilayered swarms or fruiting body structures that are simple or branched aggregates containing myxospores. Here, we examined the structural role of matrix exopolysaccharide (EPS) in the organization of these surface-dwelling bacterial cells. Using time-lapse light and fluorescence microscopy, as well as transmission electron microscopy and focused ion beam/scanning electron microscopy (FIB/SEM) electron microscopy, we found that Myxococcus xanthus cell organization in biofilms is dependent on the formation of EPS microchannels. Cells are highly organized within the three-dimensional structure of EPS microchannels that are required for cell alignment and advancement on surfaces. Mutants lacking EPS showed a lack of cell orientation and poor colony migration. Purified, cell-free EPS retains a channel-like structure, and can complement EPS − mutant motility defects. In addition, EPS provides the cooperative structure for fruiting body formation in both the simple mounds of M. xanthus and the complex, tree-like structures of Chondromyces crocatus. We furthermore investigated the possibility that EPS impacts community structure as a shared resource facilitating cooperative migration among closely related isolates of M. xanthus .",
"introduction": "Introduction Multicellular life has evolved independently several times, giving rise to a variety of macroscopic organisms: land plants, animals, fungi and so on. This advance allows these organisms to coordinate multiple cellular responses cohesively ( Blackstone, 2013 ). Examining the origins of multicellularity continues to provide new perspectives on life: for example, the transition of choanoflagellates from solitary to multicellular aggregates in response to predator prey relationships ( Alegado et al. , 2012 ). Likewise, quorum sensing and biofilm formation enable groups of bacterial cells to act as multicellular organisms that can extend their survival and competitive abilities beyond those of isolated individuals ( Davey and O'toole, 2000 ; Waters and Bassler, 2005 ; Beauregard et al. , 2013 ). In natural settings, biofilms are often heterogeneous, consisting of mixtures of different species that lack predictable organization. Even within single species biofilms, the structures that arise can be highly variable ( Baum et al. , 2009 ; Habimana et al. , 2010 ). By contrast, the myxobacteria typically produce single species biofilms with species-specific architecture ( Dawid, 2000 ). The myxobacteria are common soil bacteria that obtain nutrition through the degradation of prey bacteria, plant debris or macromolecules ( Berleman and Kirby, 2009 ). As population levels rise and nutrition is reduced, individual cells aggregate to form 0.1–1-mm tall fruiting bodies. The fruiting bodies contain 10 5 –10 7 cells of a single species; some species form simple mound shaped fruiting bodies while others form elaborate three-dimensional (3D) branched structures ( Zhang et al. , 2003 ; Hyun et al. , 2008 ). Within fruiting bodies, starved vegetative cells convert to spores; the more complex branched structures of some species show spores within sporangia at branch tips. Myxobacteria colonize surfaces as organized aggregates of cells. They move with a combination of single cell gliding (A-motility) and social (S)-motility ( Pelling et al. , 2006 ; Kraemer and Velicer, 2011 ; Sun et al. , 2011 ; Nan et al. , 2013 ). S-motility is powered by the extension and retraction of Type IV pili (T4P) that are extruded from one of the cell poles ( Nudleman et al. , 2006 ). Previous studies have shown that pili can bind to matrix exopolysaccharide (EPS) found on surfaces and presumed to coat individual cells ( Li et al. , 2003 ). Cells are pulled forward by pilus retraction, which exerts a force >100 pN ( Clausen et al. , 2009 ). According to current models, EPS forms a cell envelope that serves as an anchor for pilus binding from a neighboring cell; pilus retraction would then pull cells closer to each other ( Wall and Kaiser, 1999 ; Mauriello et al. , 2010 ). However, although a polysaccharide coat is common in bacteria such as Klebsiella and Azotobacter sp. ( Sabra et al. , 2001 ; Lin et al. , 2013 ), there is no direct evidence for an EPS cell coat covering M. xanthus cells ( Palsdottir et al. , 2009 ; Remis et al. , 2013 ), nor is there a clear explanation of how EPS-coated cells could promote movement away from other cells into new territory. In this study, we used Ruthenium Red staining of M. xanthus biofilms to visualize the carbohydrate-rich EPS ( Sutherland and Thomson, 1975 ). Our findings show that most of the EPS produced by M. xanthus is deposited on surfaces and sculpted into microchannel structures that guide cell movements. Our analysis indicates that EPS microchannels are important for the multicellular life of the myxobacteria by mediating the organization of cells during surface branch migration, fruiting body formation and intra-species interaction.",
"discussion": "Discussion Biofilm formation has an important role in the life cycle and resistance to antibiotics of many bacterial species. Biofilms can form on practically any surface; however, the ability of bacteria to shape their EPS during biofilm formation is still underappreciated. In this paper, we used genetic and ultrastructural studies to reveal an unexpected level of complexity to EPS organization in M. xanthus . We found that for M. xanthus , EPS is a highly organized structural component of the bacterial biofilm, in microchannels that align cells and promote directed movement. These honeycomb-like channels surround small groups of cells, directing the passage of the bacteria towards new territory. Essentially, the EPS microchannels allow M. xanthus to act as a multicellular organism, accomplishing tasks that no single cell could manage alone. The microchannels guide cell movement and likely serve as convenient anchors for T4P binding and retraction, allowing even those cells that do not make direct contact with the substratum (for example, agar surface) a mechanism to travel forward. EPS does not appear to be required for cell movement, but rather for efficient multicellular migration as epsZ- cells were observed to move but not to maintain consistent orientation or efficient migration patterns ( Figures 4 and 5 ). This helps explain the confounding result that epsZ- colonies can initially spread out on soft agar, but the migration gradually comes to a halt. The EPS microchannels then allow maintenance of cell alignment so that cellular movements are coordinated as cells move towards new territory during swarming or towards aggregation centers during fruiting body formation. Under the conditions tested, mutants that lack the ability to synthesize EPS do not exhibit the microchannels, lack cell alignment and show severe migration defects. However, swarming migration can be rescued by adding purified EPS to the surface. As it is unlikely for the 3D architecture of microchannels to be maintained throughout the purification process, purified EPS may provide a motility substrate rather than a defined radial orientation for cells. This is similar to other polysaccharides, like methylcellulose, which allow for T4P-based migration without showing social motility branch structures. This notion is supported by the fact that eps − strains migrate as a thin layer in regions where the EPS had been provided. Our work here confirms previous studies that show that M. xanthus cells are typically aligned on surfaces, with the long axis of the cells in parallel to the orientation of the multicellular swarms ( Wall and Kaiser, 1998 ). In those reports, end-to-end cell–cell signaling was proposed to provide the mechanism for maintaining cell orientation and organization. Other reports show that cell–cell packing can occur readily on hard surfaces, where individuals move freely through A-motility ( Wei et al. , 2011 ). Gliding movement on soft surfaces has different challenges, and the biophysical constraints of the EPS microchannels provide an alternative mechanism for multicellular organization. Cell–cell signaling may still have an important role, particularly in the area of regulating the construction process rather than in keeping cells aligned. We can hypothesize that microchannel construction occurs through a succession of events including: surface detection, EPS secretion and proper shaping of EPS by the movement of cells. The specific mechanism for correct localization of EPS construction will require further experimentation, but one prediction would be that extracellular enzymes are present on the outside surface of cells allowing dehydration reactions of polysaccharides into a cohesive microchannel structure. Cell migration through microchannels would allow efficient colonization of a surface as the crowded cells in the swarm interior push towards the low cell density/higher nutrient areas present at the swarm edge. While swarm migration across surfaces allows for colonization of new territory, the synthesis of EPS microchannels may also provide a structural support for the raised structures that are prominent during fruiting body formation both in simple M. xanthus mounds as well as in the complex fruiting bodies of C. crocatus. Our data indicate EPS provides the primary structural component of the fruiting bodies as this material is not observed in fruiting bodies of an epsZ- strain or when carbohydrate-staining Ruthenium Red is not used in TEM analysis. Microchannels were observed within the stalk of C. crocatus and EPS microcompartments observed surrounding spores of both species within the fruiting bodies. The transition from unicellular to multicellular life requires extracellular structural components that bind cells together. Plants use cellulose, hemicellulose, pectin and lignin to build cell walls ( Yang et al. , 2013 ); animals use collagen and other extracellular proteins such as laminin and glycoproteins ( Chagnot et al. , 2012 ). Here, we found that bacteria use EPS to help determine multicellular organization. Our analysis of intra-species interactions suggests that although purified EPS from various sources may serve as general substrate for migration of cells, there was no implicit sharing of EPS by co-culturing of closely related strains, but these interactions may also be greatly influenced by other factors such as cell–cell signals or antibiotics. This suggests that microbial community formation could depend on whichever organism has the keystone role in providing the predominating EPS structure. Given its importance in intra-species interactions, it will be of great importance in future experiments to study the role of EPS in inter-species interactions and microbial community structure."
} | 2,734 |
35519096 | PMC9055642 | pmc | 2,320 | {
"abstract": "A novel approach for thermo-responsive wettability has been accomplished by surface roughness change induced by thermal expansion of paraffin coated on titanate nanostructures. The surface exhibits thermo-responsive and reversible wettability change in a hydrophobic regime; the surface shows superhydrophobicity with contact angles of ∼157° below 50 °C and ∼118° above 50 °C due to a decrease of surface roughness caused by thermally-expanded paraffin at higher temperatures. Reversible wettability change of ∼40° of a contact angle allows for fast and multi-directional droplet transport. The present approach affords a versatile selection of materials and wide variety of contact angles, promoting both scientific advancement and technology innovation in the field of smart surfaces.",
"conclusion": "Conclusions We demonstrated thermo-responsive wettability induced by surface roughness change on titanate nanorod brushes that were coated with paraffin (paraffin/TNR brush). The paraffin/TNR brush exhibited thermo-responsive and reversible wettability change in a hydrophobic regime: superhydrophobicity with a CA of ∼157° below 50 °C, a less hydrophobic state with a CA of ∼118° above 50 °C. The reversible and large wettability change was attributed to surface roughness change induced by thermal expansion of paraffin accompanied by phase-transition, while the surface chemical properties remained unchanged. A spatial heating of the paraffin/TNR brush at room temperature allowed for a fast and multi-directional droplet transport owing to the reversible and large wettability change. The present approach for the fabrication of a surface with stimuli-responsive wettability based on the expansion of paraffin will be applied to other substances that can change the volume by external stimuli, which will promote scientific advancement and technology innovation in the field of smart surface."
} | 470 |
27260327 | PMC4893272 | pmc | 2,323 | {
"abstract": "Background Lignocellulosic raw materials have extensively been examined for the production of bio-based fuels, chemicals, and polymers using microbial platforms. Since xylose is one of the major components of the hydrolyzed lignocelluloses, it is being considered a promising substrate in lignocelluloses based fermentation process. Ralstonia eutropha , one of the most powerful and natural producers of polyhydroxyalkanoates (PHAs), has extensively been examined for the production of bio-based chemicals, fuels, and polymers. However, to the best of our knowledge, lignocellulosic feedstock has not been employed for R. eutropha probably due to its narrow spectrum of substrate utilization. Thus, R. eutropha engineered to utilize xylose should be useful in the development of microbial process for bio-based products from lignocellulosic feedstock. Results Recombinant R. eutropha NCIMB11599 expressing the E. coli xylAB genes encoding xylose isomerase and xylulokinase respectively, was constructed and examined for the synthesis of poly(3-hydroxybutyrate) [P(3HB)] using xylose as a sole carbon source. It could produce 2.31 g/L of P(3HB) with a P(3HB) content of 30.95 wt% when it was cultured in a nitrogen limited chemically defined medium containing 20.18 g/L of xylose in a batch fermentation. Also, recombinant R. eutropha NCIMB11599 expressing the E. coli xylAB genes produced 5.71 g/L of P(3HB) with a P(3HB) content of 78.11 wt% from a mixture of 10.05 g/L of glucose and 10.91 g/L of xylose in the same culture condition. The P(3HB) concentration and content could be increased to 8.79 g/L and 88.69 wt%, respectively, when it was cultured in the medium containing 16.74 g/L of glucose and 6.15 g/L of xylose. Further examination of recombinant R. eutropha NCIMB11599 expressing the E. coli xylAB genes by fed-batch fermentation resulted in the production of 33.70 g/L of P(3HB) in 108 h with a P(3HB) content of 79.02 wt%. The concentration of xylose could be maintained as high as 6 g/L, which is similar to the initial concentration of xylose during the fed-batch fermentation suggesting that xylose consumption is not inhibited during fermentation. Finally, recombinant R. eutorpha NCIMB11599 expressing the E. coli xylAB gene was examined for the production of P(3HB) from the hydrolysate solution of sunflower stalk. The hydrolysate solution of sunflower stalk was prepared as a model lignocellulosic biomass, which contains 78.8 g/L of glucose, 26.9 g/L of xylose, and small amount of 4.8 g/L of galactose and mannose. When recombinant R. eutropha NCIMB11599 expressing the E. coli xylAB genes was cultured in a nitrogen limited chemically defined medium containing 23.1 g/L of hydrolysate solution of sunflower stalk, which corresponds to 16.8 g/L of glucose and 5.9 g/L of xylose, it completely consumed glucose and xylose in the sunflower stalk based medium resulting in the production of 7.86 g/L of P(3HB) with a P(3HB) content of 72.53 wt%. Conclusions Ralstonia eutropha was successfully engineered to utilize xylose as a sole carbon source as well as to co-utilize it in the presence of glucose for the synthesis of P(3HB). In addition, R. eutropha engineered to utilized xylose could synthesize P(3HB) from the sunflower stalk hydrolysate solution containing glucose and xylose as major sugars, which suggests that xylose utilizing R. eutropha developed in this study should be useful for development of lignocellulose based microbial processes.",
"conclusion": "Conclusions Ralstonia eutropha was successfully engineered to utilize xylose, the second abundant sugar from lignocellulosic biomass, by the introduction of a heterologous xylose-utilizing pathway from E. coli . While the wild-type R. eutropha could not utilize xylose as a sole carbon source and it in the presence of glucose, recombinant R. eutropha expressing the E. coli xylAB genes could successfully utilize xylose as a sole carbon source resulting in the production of P(3HB). In addition, this strain could co-utilize xylose even in the presence of glucose. Finally, it efficiently utilized the hydrolyzed glucose and xylose derived from sunflower stalk as a model lignocellulosic biomass resulting in the accumulation of P(3HB) with a high content. Further engineering of xylose utilizing pathway in R. eutropha should be useful for enhanced economic feasibility of lignocellulose based production of polymers, chemicals and fuels.",
"discussion": "Results and discussions Construction of recombinant R. eutropha able to utilize xylose as a carbon source Ralstonia eutropha NCIMB11599 was firstly examined for the production of P(3HB) using xylose as a sole carbon source. As shown in Fig. 1 a, R. eutropha NCIMB11599 could not grow using xylose as a sole carbon source when it was cultured in a chemically defined MR medium containing 20.39 g/L xylose as previously reported by the genome analysis of R. eutropha suggesting its narrow range of substrate usage that cannot support xylose utilization [ 26 – 29 ]. Fig. 1 Time profiles of flask-cultures of a wild-type R. eutropha NCIMB11599, b Recombinant R. eutropha (pKM212-XylAB) in sole carbon of xylose based medium for the synthesis of P(3HB). (Symbols are: filled circle , xylose concentration; open triangle - up , cell growth or cell dry weight; open square , P(3HB) concentration; open diamond , P(3HB) content) Recombinant R. eutropha NCIMB11599 (pKM212-XylAB) expressing the E. coli xylAB genes was constructed to make this strain utilize xylose as a sole carbon source. The metabolic pathway of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) is illustrated in Fig. 2 . Recombinant R. eutropha NCIMB11599 (pKM212-XylAB) cultured in chemically defined MR medium containing 19.25 g/L of xylose, could consistently consumed xylose during the flask culture and grew up to an OD 600 of 11 (Fig. 1 b). In addition, when it was cultured in nitrogen free chemically defined MR-N medium containing 19.25 g/L of xylose, it accumulated P(3HB) to a concentration of 0.45 g/L with a P(3HB) content of 72.58 wt% and yield of 30.31 wt% (Table 1 ). Fig. 2 Metabolic pathways for biosynthesis of P(3HB) in recombinant R. eutropha strain from glucose and xylose as carbon sources used in this study. The overall metabolic pathway is shown together with the introduced metabolic pathways for the production of P(3HB). Xylulose 5-phosphate is generated by E. coli xylose isomerase (xylA) and xylulokinase (xylB). 3-Hydroxybutyryl-CoA is generated by R. eutropha β-ketothiolase (phaA) and acetoacetyl-CoA reductase (phaB). In R. eutropha strain, all the genes involved in PHA biosynthesis are in the chromosomal DNA except the xylAB gene encoding E. coli xylose isomerase and xylulokinase, which is additionally expressed by the introduction of pKM212-XylAB Table 1 Flask culture of R. eutropha NCIMB11599 expressing E. coli xylAB genes in nitrogen free chemically defined MR-N medium containing xylose as a sole carbon source Strain Carbon source DCW (g/L) P(3HB) concentration (g/L) P(3HB) content (wt%) \n R. eutropha pKM212-XylAB Xylose 0.64 ± 0.01 0.45 ± 0.04 72.58 ± 2.80 Examination of a mixture of glucose and xylose as carbon sources in recombinant R. eutropha NCIMB11599 (pKM212-XylAB) Recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was examined for its growth characteristics in MR medium containing both glucose and xylose as carbon sources. Although several microbial strains such as Lactobacillus brevis , and Lactobacillus plantarum , which are not able to utilize xylose as a sole carbon source, have been reported to be able to utilize xylose in the presence of glucose [ 30 ], it was found out that R. eutropha NCIMB11599 could not still consume xylose even in the presence of glucose. When R. eutropha NCIMB11599 was cultured in MR medium containing 9.78 g/L of glucose and 10.43 g/L of xylose, it completely consumed glucose in 72 h to grow up to an OD 600 of 21.12 (data not shown). However, the concentration of xylose was maintained as high as the initial concentration throughout the culture. On the other hand, recombinant R. eutropha NCIMB11599 (pKM212-XylAB) could utilize both glucose and xylose when it was cultured in MR medium containing these two substrates. As shown in Fig. 3 a, R. eutropha NCIMB11599 (pKM212-XylAB) consumed 5.97 g/L of glucose and 6.51 g/L of xylose, growing up to an OD 600 of 21.14, from 9.16 g/L of glucose and 11.43 g/L of xylose, respectively. The rates of glucose and xylose utilization were 0.0622 g/L/h and 0.0678 g/L/h, respectively. By the way, when recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was cultured in the medium initially containing 13.68 g/L of glucose and 5.76 g/L of xylose, it consumed 8.29 g/L of glucose and 3.32 g/L of xylose, with the rates of glucose and xylose utilization of 0.0864 g/L/h and 0.0346 g/L/h, respectively, to grow up to an OD 600 of 19.01 (Fig. 3 b). The rate of xylose utilization by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) tended to decrease as the initial ratio of glucose to xylose in the medium increased. In contrast, the increased substrate ratio could speed up the glucose consumption rate of the host strain. Fig. 3 Time profiles of flask cultures of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR medium containing mixed sugars of glucose and xylose as carbon sources. a 9.25 g/L of glucose and 10.97 g/L of xylose were used as carbon sources b 13.76 g/L of glucose and 5.78 g/L of xylose were used as carbon sources. (Symbols are: filled square , glucose concentration; filled circle , xylose concentration; open triangle - up , cell growth) When recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was cultured in a nitrogen free chemically defined MR-N medium containing both glucose and xylose as carbon sources, it exhibited higher P(3HB) production capacities than that obtained by the flask culture using xylose as a sole carbon source. In addition, production of P(3HB) was enhanced as the initial ratio of glucose to xylose in MR-N medium increased. The concentration of P(3HB) and P(3HB) content obtained by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) cultured in MR-N medium containing 9.85 g/L of glucose and 10.01 g/L of xylose were 1.49 g/L and 86.58 wt%, respectively. The consumed glucose and xylose were 2.21 g/L and 3.27 g/L of xylose, respectively. On the other hand, 1.65 g/L of P(3HB) and 93.33 wt% of P(3HB) content were obtained by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) cultured in MR-N medium containing 14.36 g/L of glucose and 5.63 g/L of xylose by consuming 2.23 g/L of glucose and 1.09 g/L of xylose (Table 2 ). Also, the P(3HB) yields of 49.96 wt% obtained in MR-N medium containing 14.36 g/L of glucose and 5.63 g/L of xylose was higher than that of 23.54 wt% obtained in MR-N medium containing 9.85 g/L of glucose and 10.01 g/L of xylose. Taken together, it seems to be that R. eutropha NCIMB11599 (pKM212-XylAB) still favors glucose to xylose in terms of substrate utilization and P(3HB) production capacity even though it has the functional xylose-utilizing pathway. Table 2 Results of analysis of bacterial growth, P(3HB) concentration, and P(3HB) contents by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) using mixed carbon sources during flask cultivations using MR-N medium containing glucose and xylose as carbon sources Strain Glucose (g/L) Xylose (g/L) DCW (g/L) P(3HB) concentration (g/L) P(3HB) content (wt%) \n R. eutropha (pKM212-XylAB) 9.25 10.97 1.49 ± 0.15 1.29 ± 0.10 86.58 ± 0.52 13.76 5.78 1.65 ± 0.03 1.54 ± 0.11 93.33 ± 0.65 The mechanism of xylose uptake by R. eutropha has not yet been clearly understood. The sequence similarity search by the protein BLAST indicates that R. eutropha seems not to possess a possible xylose transporter like E. coli . There are several reports suggesting bacterial xylose uptake through different mechanisms. A facilitated diffusion mechanism has been reported to be functional in Lactobacillus pentosus for xylose transportation [ 31 ]. On the other hand, Corynebacterium glutamicum has been suggested to possess several different mechanisms for xylose transportation [ 32 ]. Taken together, further studies are required in order to understand the xylose uptake mechanisms in R. eutropha . Examination of P(3HB) production in recombinant R. eutropha NCIMB11599 (pKM212-XylAB) from a mixture of glucose and xylose as carbon sources in fermentation The growth and P(3HB) production profiles of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was further examined in batch fermentations in a 2.5 L jar fermentor (CNS Co. Ltd., Korea) using xylose or mixtures of xylose and glucose as carbon sources, respectively. When recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was cultured in MR-B medium containing 20.18 g/L of xylose, it completely consumed the substrate in 120 h (Fig. 4 a). Considering that recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was able to utilize glucose within 12 h when it was cultured in MR-B medium containing 20 g/L of glucose as a sole carbon source (data not shown), it showed relatively a long lag phase of 48 h in the xylose fermentation. In addition, it took three times longer to consume 20.18 g/L of xylose than to consume the same amount of glucose (data not shown). After the lag phase, R. eutropha NCIMB11599 (pKM212-XylAB) gradually consumed xylose and then grew up to a DCW (dry cell weight) of 7.46 g/L in 120 h (Fig. 4 a). The highest P(3HB) concentration P(3HB) content, and P(3HB) yield achieved in 120 h were 2.31 g/L, 30.95 wt%, and 11 wt% (g polymer/g xylose), respectively. The xylose consumption rate was 0.164 g/L/h. Fig. 4 Time profiles of batch fermentations of a recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-B medium containing 20.18 g/L of xylose as a sole carbon source, b recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-B medium containing 10.05 g/L of glucose and 10.91 g/L of xylose as carbon sources and c recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-B medium containing 16.94 g/L of glucose and 6.15 g/L of xylose as carbon sources. (Symbols are: filled square , glucose concentration; filled circle , xylose concentration; open triangle - up , dry cell weight; open square , P(3HB) concentration; open diamond , P(3HB) content) When 10.91 g/L of xylose and 10.05 g/L of glucose were used as carbon sources, recombinant R. eutropha NCIMB11599 (pKM212-XylAB) completely consumed glucose in 40 h and xylose in 72 h, respectively, resulting in the highest DCW of 9.08 g/L in 72 h (Fig. 4 b). While R. eutropha NCIMB11599 (pKM212-XylAB) was able to utilize glucose in 25 h, it began to utilize xylose in 30 h after inoculation. The highest concentration of P(3HB), P(3HB) content, P(3HB) yield, P(3HB) productivity of 5.71 g/L, 78.11 wt%, 40.0 wt% (g polymer/g glucose + xylose), and 0.12 g/L/h, respectively, were obtained by consuming 10.91 g/L of glucose and 4.19 g/L of xylose in 48 h. The rates of glucose and xylose consumption were 0.25 g/L/h and 0.15 g/L/h, respectively. Although the mixed-substrates fermentation was completed within a much shorter time than the xylose fermentation, the rate of xylose consumption decreased in the presence of glucose. In addition, fermentation of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-B medium containing 6.15 g/L of xylose and 16.94 g/L of glucose resulted in different profiles of cell growth, substrate consumption rates, and P(3HB) production from the previous fermentations (Fig. 4 a, b, and c). Recombinant R. eutropha NCIMB11599 (pKM212-XylAB) began to utilize glucose in 12 h after inoculation and then completely consumed the substrate in 51 h. On the other hand, it was able to consume xylose in 24 h after inoculation and then completely consumed the substrate in 60 h, growing to the highest DCW of 10.01 g/L. The highest concentration of P(3HB) of 8.79 g/L was obtained in 56 h by consuming 16.94 g/L of glucose and 3.97 g/L of xylose with in a P(3HB) content of 88.69 wt%, a P(3HB) yield of 42.0 wt%, and a productivity of 0.15 g/L/h, respectively. The rates of glucose and xylose consumption were 0.33 g/L/h and 0.09 g/L/h, respectively. While the consumption rate of glucose increased by 31 % than that obtained with 10.05 g/L of glucose and 10.91 g/L xylose, the consumption rate of xylose decreased by 40 %. The growth and P(3HB) production profiles of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) were further investigated in fed-batch fermentations in a 2.5 L jar fermentor (Fig. 5 ). The initial concentrations of glucose and xylose were 16.01 and 7.48 g/L, respectively. While the DCW and P(3HB) concentration gradually increased, the P(3HB) content reached to the highest point (80.51 wt%) at 84 h and was not significantly changed during fermentation. The P(3HB) yield for the batch phase and that for the feeding phase were 22.63 and 35.89 wt%, respectively, resulting in the overall yield of 34.87 wt%. Considering that the ratio of xylose to glucose in the feeding solution was 1:3 and the consumption rate of xylose (0.09 g/L/h) was less than one-third of that of glucose (0.33 g/L/h) in the previous batch fermentation, it was interesting that the xylose concentration was maintained at about 6 g/L, instead of being accumulated during the fed-batch fermentation. It is interesting to note that this result was mainly due to the increased consumption rate of xylose during the feeding phase (0.27 g/L/h). Fig. 5 Time profiles of fed-batch fermentations of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-B medium containing 16.01 g/L of glucose and 7.48 g/L of xylose as carbon sources. (Symbols are: filled square , glucose concentration; filled circle , xylose concentration; open triangle - up , dry cell weight; open square , P(3HB) concentration; open diamond , P(3HB) content) Since the highest cell growth, P(3HB) concentration, content, weight yield, and productivity achieved by batch fermentations of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) increased as the initial ratio of glucose to xylose in the culture medium increased, it could be concluded that glucose is still a preferred substrate to xylose for its cell growth and P(3HB) production, even though the engineered R. eutropha NCIMB11599 (pKM212-XylAB) successfully utilized xylose as a sole carbon source as well as co-utilized it in the presence of glucose. This phenomenon is quite common in microbial species including E. coli , Saccharomyces cerevisiae , Bacillus subtilis , Pseudomonas putida , Lactococcus lactis , Lactobacillus casei , and C. glutamicum [ 33 – 36 ]. Among them, C. glutamicum , which is not a native xylose utilizing organism, has also been investigated for xylose utilization by introducing the E. coli xylAB genes used in this study [ 33 ]. It was reported that expression of the xylAB genes is enough for the construction of recombinant C. glutamicum to consume xylose as a carbon source without further engineering to improve xylose consumption such as introduction of heterologous transporters for xylose and overexpression of native transporters. However, engineered strains of C. glutamicum to utilize xylose still preferred glucose when it was cultured in a medium containing both sugars, glucose and xylose. Recombinant C. glutamicum that has been engineered to efficiently use xylose as a carbon source still showed twice higher consumption rate of glucose than that of xylose when it was cultured in a medium containing 20 g/L of glucose and 10 g/L of xylose [ 33 ]. On the other hand, previous studies have focused on using native xylose-utilizing hosts such as E. coli [ 37 ] and Burkholderia cepacia [ 38 ] for the production of PHAs from xylose. We successfully developed an engineered R. eutropha to utilize xylose as a sole carbon source as well as co-utilize xylose in the presence of glucose. However, it was observed that, like other well-known microbial hosts mentioned above, recombinant R. eutropha engineered in this study is still carbon catabolite repression (CCR)-sensitive in the presence of glucose. To overcome this problem, it would be necessary to apply the engineering approaches examined in other strains into R. eutropha for improving xylose catabolism along with the reduction of CCR [ 36 , 39 – 41 ]. For example, enhanced expression of the transketolase operon in the pentose phosphate pathway by replacing its native promoter with the strong sod promoter in C. glutamicum could lead to the increase of the product yield from xylose [ 42 ]. In addition, an engineered S. cerevisiae overexpressing the transaldolase gene exhibited a higher xylose consumption rate than its parental strain when it was cultured in the medium containing both glucose and xylose [ 43 ]. Introduction of heterologous xylose transporters may also be a good strategy to increase xylose consumption rate along with reduced CCR [ 33 , 34 ]. In addition, production of different types of PHA copolymers from xylose can be achieved by employing PHA synthase with a promiscuous substrate specificity and by introducing additional metabolic pathways to supply various hydroxyacyl-CoA (HA-CoA) with different carbon numbers and R-pending groups to PHA synthase [ 44 ]. Since xylulose 5-phosphate, synthesized from xylose by the enzymatic reaction of XylA and XylB, may enter into the pentose phosphate pathway yielding glyceraldehyde 3-phosphate, which is an intermediate of glycolysis, most of engineering strategies for PHA copolymers based on glycolysis would be applicable for the production of PHA copolymers from xylose as a carbon source. Biosynthesis of P(3HB) in recombinant R. eutropha NCIMB11599 (pKM212-XylAB) from sunflower stalk hydrolysate solution To examine the possible application of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in a lignocellulose-based PHA production process, sunflower stalk has been chosen as a raw material for biosugar. The hydrolysate solution of sunflower stalk was prepared by hydrothermal pretreatment and subsequent hydrolysis of the raw material. The sunflower stalk hydrolysate solution contained 8.7 wt% glucose, 2.4 wt% hemicellulosic sugars, 0.006 wt% furfural, and ash. Hemicellulosic sugars contained 85.1 wt% of xylose, 0.2 wt% of galactose, and 14.7 wt% of mannose, respectively. The hydrolysate solution of sunflower stalk may be successfully employed for microbial fermentation process since it contains quite low concentrations of possible inhibitors such as acetic acid, furfural, and 5-hydroxy-methyl-2-furaldehyde (HMF), all of which are known to be generated from hydrothermal pretreatment of lignicellulosic biomass. Growth of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) from sunflower stalk hydrolysate solution was firstly examined in flask cultures, in which MR medium containing 1:5 diluted hydrolysate solution of sunflower stalk was used. Recombinant R. eutropha NCIMB11599 (pKM212-XylAB) could grow to an OD 600 of 27.67 by consuming 13.49 g/L of glucose and 4.37 g/L of xylose in 96 h (Fig. 6 ). Also, galactose and mannose existing in sunflower stalk hydrolysate solution was completely consumed by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in 96 h as previous reported [ 45 , 46 ]. When this strain was cultured in MR-N medium containing 1:5 diluted hydrolysate solution of sunflower stalk, it produced 2.42 g/L of P(3HB) with a P(3HB) content of 75.42 wt% by consuming 5.61 g/L of glucose, 0.431 g/L of xylose, 0.00101 g/L of galactose, and 0.0744 g/L of mannose. The P(3HB) yield was 39.50 wt% (g polymer/g glucose + xylose + galactose + mannose) which is higher than that synthesized by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-N medium containing 10.05 g/L of glucose and 10.91 g/L of xylose, but lower than that in MR-N medium containing containing 16.94 g/L of glucose and 6.15 g/L of xylose (Table 3 ). Fig. 6 Time profiles of flask cultures of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR medium containing sunflower stalk hydrolysate solution as a carbon source. The initial XGM is composed of 85.13 % xylose, 0.23 % galactose, and 14.64 % mannose. (Symbols are: filled square , glucose concentration; filled circle , xylose, galactose, and mannose (XGM) concentration; open triangle - up , cell growth) Table 3 Results of analysis of bacterial growth, P(3HB) concentration, and P(3HB) contents by recombinant R. eutropha NCIMB11599 (pKM212-XylAB) during flask cultivation in MR-N medium containing sunflower stalk hydrolysate solution as a carbon source Strain Carbon source DCW (g/L) P(3HB) concentration (g/L) P(3HB) content (wt%) \n R. eutropha (pKM212-XylAB) Sunflower stalk a \n 2.46 ± 0.04 2.09 ± 0.01 84.96 ± 0.89 \n a Initial concentrations of carbon sources in the medium were 15.55 g/L of glucose, 5.41 g/L of xylose, 0.01 g/L of galactose, and 0.92 g/L of mannose Batch fermentation of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) was carried out in a 2.5 L jar fermentor with MR-B medium using the sunflower stalk hydrolysate solution as a carbon source. The time profiles of cell growth and P(3HB) production are shown in Fig. 7 . Recombinant R. eutropha NCIMB11599 (pKM212-XylAB) grew to a DCW of 10.97 g/L and completely consumed all fermentable sugars in 60 h. The highest P(3HB) concentration of 7.86 g/L was achieved with a P(3HB) content of 72.53 wt% and a P(3HB) yield of 34.52 wt% in 54 h. Fig. 7 Time profiles of batch fermentation of recombinant R. eutropha NCIMB11599 (pKM212-XylAB) in MR-B medium containing sunflower stalk hydrolysate. The initial XGM is composed of 85.13 % xylose, 0.23 % galactose, and 14.64 % mannose. (Symbols are: filled square , glucose concentration; filled circle , xylose, galactose, and mannose (XGM) concentration; open triangle - up , dry cell weight; open square , P(3HB) concentration; open diamond , P(3HB) content) As the initial composition of fermentable sugars in the sunflower stalk based medium was similar to that in the mixed sugar based medium (about 16.94 g/L of glucose and 6.15 g/L of xylose), recombinant R. eutropha showed similar profiles of cell growth and substrate consumption in both media. However, the highest concentration of P(3HB) synthesized by recombinant R. eutropha (pKM212-XylAB) in the sunflower stalk based medium was lower than that in the mixed sugar based medium even though it grew slightly better on the former medium than on the latter medium in terms of the highest DCW. In other words, the sunflower stalk hydrolysate solution can be a better fermentation feedstock for R. eutropha (pKM212-XylAB) than the mixture of glucose and xylose in terms of cell growth, but not of P(3HB) production. As previously reported by the studies on the examination of lignocellulosic biomass as raw materials for fermentation, unknown components present in the hydrolysate solution seemed to negatively affect the host cell to synthesize target products [ 17 , 47 , 48 ]. R. eutropha has also been examined for PHA production using lignocellulosic raw materials [ 49 ]. In this study, the alkali pretreated hydrolysates of rice paddy straw, sunflower husk, soybean straw, and wood straw have been examined for the cell growth and P(3HB) production by R. eutropha , in which the hydrolysate of rice paddy straw resulted in the highest P(3HB) concentration, yield, and weight. It was found out that both the cell growth and P(3HB) synthesis were hindered by the paddy straw hydrolysate, whereas sunflower hydrolysate solution used in present study did not significantly affect cell growth and P(3HB) synthesis efficiency. This might be from the different characteristics of sunflower stock and hydrothermal pretreatment process employed in this study. Since only a few lignocelluloses have been examined for PHA production in recombinant and natural R. eutropha strains, it is difficult to evaluate economic feasibility of PHA production process due to the limited data on the effect of lignocellulosic feedstock cost on the entire production cost. Thus, investigation of more raw materials from different lignocelluloses for PHA production are needed to evaluate their economic feasibility, from which fermentable sugars are obtained by employing appropriate pretreatment and hydrolysis methods. In addition, availability of raw materials in the area of production process should be considered because transportation cost may exceed production cost if a process is constructed far away from the farmland along with the digestibility and efficiency of lignocelluloses in fermentation process after hydrolysis process. Finally, all of process parameters including efficiency of microbial producers, fermentation types, purification techniques, process design, capital costs, and operation costs should be taken into consideration for the development of economically feasible process for the production of PHA based on lignocellulosic materials."
} | 7,295 |
40212327 | PMC11935989 | pmc | 2,324 | {
"abstract": "Electronic skins have been developed for sensing vertical pressure and tensile strain individually. Here, a stretchable, multifunctional, and 3D electronic skin (SMTE) is fabricated to imitate the squid's natural structure to achieve strain visualization and pressure sensing at the same time. The SMTE consists of an anisotropic and elastic light shutter to tune the natural color or fluorescence, which converts the in‐plane strain to the visual changes in color and intensity. The shutter also works as an elastic dielectric layer for vertical pressure sensing based on the principle of a single‐electrode triboelectric nanogenerator. The performance of the SMTE is demonstrated by integrating it on a model hand as a tactile sensor for touch, bending, and gesture interpretation, which has potential applications in human–machine interaction, soft robots, and artificial intelligence.",
"conclusion": "3 Conclusion In this work, a stretchable, multifunctional, and 3D E‐skin is fabricated to achieve in‐plane strain visualization and vertical pressure sensing. The visual in‐plane strain sensing part consists of fluorescent elastomer films and a light shutter with many parallel microcracks. Induced by the external force, the size of microcracks controls the transmittance of the bottom fluorescence to display the local strain distribution. The microcracks open up in the transverse direction while closing up in the longitudinal direction due to the Poisson's ratio. Therefore, the light shutter can distinguish the transversely and longitudinally tensile strains. Compared with the rigid film with polyvinyl alcohol (PVA) and other materials, the microcracks in the SMTE can be controlled precisely by the microengineering process. By developing optimum cracks, SMTE can perform transversely tensile sensing in the strain range of 0–120% with an intensity sensitivity up to 20.8. The results show that the color coordinates in the CIE diagram followed an approximately linear variation from the blue region to the orange region. SMTE realizes the longitudinal tensile sensing in the strain range of 0–100%, with the fluorescence intensity decreasing over 60%. Moreover, the fabricated SMTE can realize sensitive and stable visual in‐plane strain sensing. The vertical pressure part of the SMTE is based on the single‐electrode mode TENG. A stretchable conductive network of AgNWs is embedded in the medium layer, and the elastomer shutter works as the triboelectric and dielectric layer. The open‐circuit voltage can reach 28.3 V, and the transfer charges can reach 6 nC. The SMTE achieves a high sensitivity up to 40.22 V kPa −1 or 0.61 nC kPa −1 for vertical pressure sensing. The fabricated triboelectric part can work stably for long cyclings. Finally, the finger SMTEs are attached to the hand model to demonstrate its 3D sensing of external stimuli. The finger SMTEs are appropriate to the finger shape and can respond to the model gestures with corresponding visual and electrical signal outputs. Moreover, the multidimensional properties of the finger SMTE are demonstrated by the anisotropic letter shapes. In conclusion, the SMTE has potential applications in human–machine interaction, soft robotics, and artificial intelligence.",
"introduction": "1 Introduction Skin is one of the most sensitive and complicated sensory organs for humans and living bodies. [ \n \n 1 \n \n ] It converts environmental stimuli into physiological signals, which nervous systems can further transport or interpret. [ \n \n 2 \n \n ] Some species possess the extraordinary ability to change colors to gain advantages against prey or hide from potential predators. [ \n \n 3 \n \n ] The bioinspired design has been an attractive strategy for electronic skins (E‐skins) to imitate the natural structure.[ \n 3 \n , \n 4 \n ] Recently, E‐skins have been developed for sensing vertical pressure and tensile strains. [ \n \n 5 \n \n ] However, the lack of 2D anisotropic tensile sensing properties has greatly limited their applications. Triboelectric nanogenerators (TENGs) can generate electrical signals based on contact electrification and Maxwell displacement current. [ \n \n 6 \n \n ] When TENGs work as self‐powered sensors, the output signals are correlated to the intensity of external stimuli. Furthermore, a wide range of flexible or stretchable materials can be adopted to manufacture TENGs and E‐skins. [ \n \n 7 \n \n ] Lin et al. demonstrated a TENG sensor array for self‐powered static and dynamic pressure detection. Yang et al. developed a single‐electrode TENG for self‐powered sensing of human touch. Moreover, E‐skins can have the bioinspired function to change the color or luminance with stress or strain‐controlled microstructure for communication and camouflage. [ \n \n 8 \n \n ] It is meaningful for E‐skins to develop other properties, such as the optical properties of tunable reflected color and fluorescence intensity.[ \n 5 \n , \n 8 \n ] In this work, a stretchable, multifunctional, and 3D electronic skin (SMTE) was fabricated to achieve strain visualization and vertical pressure sensing. [ \n \n 9 \n \n ] The SMTE consisted of an elastic and stretchable shutter for the light transmission. The transverse stretching could open up the shutter to let light through and change the SMTE color. [ \n \n 10 \n \n ] In contrast, the longitudinal stretching closed the shutter due to the Poisson's ratio and only changed the intensity of the original color. [ \n \n 11 \n \n ] Another vertical pressure sensing was achieved by following the principle of single‐electrode TENG.[ \n 9 \n , \n 12 \n ] The shutter also worked as an elastic dielectric layer for vertical pressure sensing. The electrical and optical performance of the SMTE was demonstrated by integrating it on a hand model. Such a combination of visual strain and electrical sensing is aimed to enhance the sensing performance as tactile sensors and effectively increases the information flux in a short time. The SMTE could work as a tactile sensor for touch, bending, and gesture interpretation, which has potential applications in human–machine interaction, soft robots, and artificial intelligence.[ \n 8 \n , \n 13 \n ]",
"discussion": "2 Results and Discussion 2.1 Concept and Working Principle of SMTE Squids utilize the pigments to change the skin color for instant camouflage. Figure \n 1 a shows the squid and its epidermal cells. The pigment cells are in the capsule and only have an exposure area of 100 μm, which is almost invisible to the naked eye. Muscles are attached around them. When the muscle expands, the pigment cells rapidly expand to the millimeter range, displaying color patterns instantly. Inspired by this natural structure, the SMTE is designed to respond to tensile strains by anisotropic optical signals. The SMTE changes its color under transverse stretching and light intensity under longitudinal stretching. Figure 1 The design of the SMTE. a) Schematic diagram of the squid and its color‐change structure. b) Hierarchically schematic illustration of the SMTE on the enlarged functional area. c) The schematic illustration of the fabrication process of the SMTE. d) The SEM images of the phosphors and AgNWs. e,f) PL spectra of blue and orange phosphors. © 2021 WILEY‐VCH GmbH The hierarchical structure of the SMTE on the enlarged functional area is shown in Figure 1b . The strontium aluminate phosphor doped with Eu and Dy is used as the visual material. The Ecoflex films are mixed with these phosphor powder to attain the corresponding color in the natural light. In addition, the phosphors also have fluorescence after UV irradiation. In the SMTE design, the orange layer is like the special pigment cells of squids, whereas the blue one is like the muscle to change the overall visualization. Between the orange and blue layers, another Ecoflex film is mixed with carbon black as a shading layer to prevent the light leak from the orange bottom layer. The shading layer is synchronously stretchable with the blue layer. Therefore, the orange layer bears the main stress under tensile strains, whereas there is a slight change in the blue and shading layers. During the expansion, the blue and shading layers work as the light shutter, and the orange layer finally changes the SMTE color. A silver nanowires (AgNWs) layer is fabricated under the blue layer as the electrode for the single‐electrode TENG. The fabrication process is shown in Figure 1c . First of all, the Ecoflex mixed with orange phosphors was spin coated on the precleaned acrylic plate to cure. Then AgNWs solution was sprayed on the orange layer, with conducting wires fixed for the following step. Another Ecoflex precursor mixed with carbon black was spin coated on the orange and AgNWs layers as a shading layer. Next, an Ecoflex layer mixed with blue phosphors was coated on the shading layer. After complete cure, the blue and carbon layers were cut by laser cutting to form regular and parallel cracks. The orange bottom layer was intact and exposed to external environments. Finally, the whole film was peeled off from the acrylic plate. The detailed fabrication process is in Experimental Section. Figure 1d shows the scanning electron microscopy (SEM) images of the phosphor powder and AgNWs. The particle size of the orange phosphor was around 13 μm, while the blue phosphor was about 1.5 μm. Figure S1, Supporting Information, shows the X‐ray diffraction (XRD) data of the orange and blue phosphors, and the main component is strontium aluminate. The corresponding photoluminescence (PL) spectra of orange and blue phosphors were measured by an Edinburgh spectrometer, with the results in Figure 1e,f . The main PL peak of the blue phosphor was around 484 nm, and the main PL peak of the orange phosphor was around 584 nm. The photos of phosphors under the natural light and dark environment were in the insets, which the fluorescence in the dark was excited by a broadband UV lamp. 2.2 Visual Strain Sensing of the SMTE in the Transverse Direction SMTE was designed to achieve a highly sensitive and visual transverse stretching sensor, as shown in Figure \n 2 a . The SMTE had many parallel and periodic microcracks which were along the longitudinal direction. The cracks were precisely cut by the laser cutting, penetrating the upper layers to the orange bottom layer to attain a light shutter to regulate the size of the color area and the light leak from the orange bottom layer. When the SMTE was stretched transversely, the orange layer bore the main tensile stress, whereas there was little stress in the shutter. Therefore, the exposed area of the orange bottom layer increased gradually to emit corresponding fluorescence under environmental light excitation, realizing visual strain sensing. Figure 2b shows the photos of stretching SMTE under different tensile strains. The SMTE displayed its visual changes under natural and UV light. In the case of no stretching, the SMTE mainly emitted the blue glow from the upper layer. However, the width of microcracks increased gradually with the increasing tensile strain. The color changes could display the local stress distribution. The stretched Ecoflex film was wide on both ends and narrow in the middle due to the Poisson's ratio. The light shutter on the microscale was also observed under the optical microscope, as shown in the bottom part of Figure 2b . The microcracks changed from about 100 to 800 μm with increasing tensile strain, whereas the hills attained slight changes. Considering that the microcracks were uniformly fabricated in the transverse direction, the difference in the width resulted from the local stress distribution. Figure 2 Visual strain sensing of the SMTE in the transverse direction. a) The working mechanism of the SMTE in the transverse direction to achieve visual color changes. b) The photos of the stretching SMTE under different tensile strains with the excitation from natural (top) and UV (middle) light. The microcracks were broadened under an optical microscope (bottom). c) The corresponding PL spectra of the SMTE under different tensile strains in the transverse direction. d) The CIE coordinates move from the blue to the orange region with increasing tensile strains. e) The 584 nm peak intensity of the PL spectra increased with the transversely tensile strains. f) The real‐time 584 nm intensity of the PL spectra in five stretching cycles under different transverse strains. g) Cycling tests of the SMTE to record the 584 nm peak intensity under 100% tensile strain (excited by the UV lamp). © 2021 WILEY‐VCH GmbH Figure 2c shows the fluorescent spectra of SMTE at different stretching strains of 0–120%. The detailed spectra of stretching strains with stepwise 10% increases were shown in Figure S2, Supporting Information. The corresponding spectra were attributed to the relative changes in the 484 and 584 nm peaks, which were adjusted by the light shutter. The peak intensity of 484 nm, mainly from the blue layer, decreased by 88%. On the contrary, the peak intensity of 584 nm, mainly from the orange layer, increased by 1866%. However, the orange layer also made contributions to the 484 nm peak. The 484 nm peak intensity no longer decreased when the orange peak reached saturation at the 120% tensile strain. In this situation, the width of the microcracks was too large to maintain the function of the light shutter to adjust the orange fluorescence. The color changes could identify the transverse stretching strain of the SMTE. The coordinates in the commission internationale de L'eclairage (CIE) color space could describe the color changes under different transverse tensile strains, as shown in Figure 2d . In the measurement, the SMTE was excited by a stable 365 nm light to emit fluorescence. With the strain increasing, the coordinate gradually moved from the blue region to the orange region. The star markers in Figure 2d show the stepwise color changes from 0% to 120% stretching states. The results show that the color coordinates in the CIE diagram followed an approximately linear variation. Moreover, the changes of 584 nm intensity to strains were defined as the gauge factor for the transversely stretching SMTE, as shown in Figure 2e . There were sensitive visual responses to the continuously increasing strain. The gauge factor reached 20.8 at the 80% strain. Furthermore, the real‐time curves of the fluorescence intensity were measured during repetitive cycles of stretching. A portable spectrometer equipped with an optical fiber tested the fluorescence intensity of SMTE under tensile strains. The different strains could be distinguished by the intensity of the 584 nm peak, as shown in Figure 2f . The dynamic visual changes are shown in Video S1, Supporting Information. The SMTE had excellent sensing stability and reversibility in the transverse stretching process. The cycling test was carried out for the SMTE to record the real‐time curve of the 584 nm intensity. The tensile strain was set as 100%, and the period was 2 s in the experimental setup. The result is shown in Figure 2g . The light intensity varied between 15 counts for the release state and 203 counts for the 100% strain during the cycles, achieving the changes over 12 times. After more than 5000 cycles, the SMTE maintained the stable and sensitive sensing properties for tensile strain. The stable visual sensing was attributed to the elastic properties of the elastomer and the light shutter. The stretching strain below 200% was in the range of the elastic deformation of the SMTE. The measurement system to test the real‐time PL intensity of SMTE is shown in Figure S3, Supporting Information. 2.3 Visual Strain Sensing of the SMTE in the Longitudinal Direction To further discuss the anisotropic properties of in‐plane visual strain sensing, the SMTE performance was measured by stretching longitudinally. The schematic illustration is shown in Figure \n 3 a . The SMTE was stretched longitudinally, parallel to the microcrack direction, and measured its optical properties simultaneously. Under longitudinal tensile strain, the microcrack size decreased to close up. Therefore, the orange bottom layer could be hidden under the elastic shutter. Figure 3b shows the optical photos of different stretching states of SMTE under natural and UV light excitation. The SMTE remained to glow blue color even though the longitudinal strains increased from 0% to 100%. Simultaneously, the fluorescent intensity gradually decreased because the areal density of blue phosphor particles decreased. The bottom photos in Figure 3b show the microcracks under longitudinal strain tended to be closed due to the large Poisson's ratio of the elastic Ecoflex. Therefore, the orange bottom layer was hidden under the shutter. Figure 3 Visual strain sensing of the SMTE in the longitudinal direction. a) The working mechanism of the SMTE in the longitudinal direction to achieve visual strains. b) The photos of stretching the SMTE under different tensile strains with the excitation from natural (top) and UV (middle) light. The microcracks were closed under an optical microscope (bottom). c) The corresponding PL spectra of the SMTE under different longitudinal tensile strains. d) The CIE coordinates stayed in the blue region with increasing tensile strains labeled as the red markers. Only the intensity of the fluorescence changed. e) The 484 nm peak intensity of the PL spectra decreased with the tensile strains in the longitudinal direction. f) The real‐time 484 nm intensity in 5 stretching cycles under different longitudinal strains. g) Cycling tests of SMTE to record the 484 nm peak intensity under 100% tensile strain (excited by UV lamp). © 2021 WILEY‐VCH GmbH The fluorescent spectra resulting from longitudinal tensile strains are shown in Figure 3c . The detailed spectra of the stretching states with a stepwise 10% strain increase were shown in Figure S4, Supporting Information. The main peak of the spectra was around 484 nm, indicating that the emitting fluorescence was mainly contributed to the blue top layer under longitudinal strains. The 484 nm peak intensity decreased by about 72% from 0% to 110% tensile strains because the tensile strains reduced the areal density of fluorescent particles. However, there were much smaller changes around the 584 nm peak. The results show that the longitudinal strain had little influence on the orange bottom layer, screened by the light shutter. The shading layer with carbon black prevented the excitation light from transmitting and screened the bottom fluorescence. Therefore, the CIE coordinates remained stable under longitudinal tensile strains, as shown in Figure 3d . The black stars are the coordinates of the transverse strains, and the red dots are the coordinate of the longitudinal stretching states. There was a distinguished difference between the two tensile conditions, which provided enough information to realize the in‐plane visual strain sensing. Angle tests for a square SMTE were carried out as stretching directions of 0°, 45°, 90°, 135°, and 180° to further discuss the in‐plane visual strain sensing. The visual changes displayed the strain distribution for this soft and stretchable device, as shown in Figure S5, Supporting Information. The 484 nm peak intensity of the PL spectra decreased with the longitudinal tensile strains. Figure 3e shows the approximate linear curve of 484 nm peak intensity to the longitudinal stretching strains. The slope value of the curve was −0.61, as the longitudinal gauge factor of SMTE. Unlike the transverse sensing results, the fluorescent sensitivity of SMTE to longitudinal strain was related to the concentration of phosphor powder in the Ecoflex composite film or the areal density of phosphor particles. Moreover, the real‐time 484 nm peak intensity at different longitudinally tensile strains was measured. The detailed intensity curves of the tensile strains are shown in Figure 3f , displaying that SMTE had good reversibility and stability under longitudinal stretching. The SMTE was also tested for more than 5000 cycles to record the cycling changes in the fluorescence intensity, as shown in Figure 3g . The experiment was carried out under 100% strain and a 2 s period. During the stretching process, the 484 nm intensity was 380 counts at the release state and 221 counts at the 100% stretching strain. The light intensity waveform did not change significantly, indicating that SMTE had stable and long‐lasting sensing properties for longitudinal tensile strain. 2.4 Triboelectric Pressure Sensing of the SMTE in the Vertical Direction The microengineering structure usually works as the dielectric layer for capacitive and triboelectric sensors. [ \n \n 14 \n \n ] The light shutter based on the microcracks also has pressure sensitivity in the vertical direction. In contrast, AgNWs could work as the flexible and stretchable electrode for the sensor because AgNWs conducting network has excellent conductivity and stability in the tensile state. [ \n \n 15 \n \n ] The AgNWs were sprayed on the cured Ecoflex film and randomly oriented and evenly distributed to form a conductive network. The resistance of the AgNWs network was about 85 Ω before the laser cutting with a distance of 5 cm. Based on the microstructured Ecoflex and AgNWs electrode, the fabricated single‐electrode TENG could work as a self‐powered pressure sensor in the SMTE to transform the vertical pressure into electrical signals. Figure \n 4 a shows the working mechanism of the single‐electrode TENG in the SMTE. In the measurement setup, the light shutter contacted another Al film face to face. There was a charge transfer between the Al film and SMTE at the several initial contacts. The electrons in the Al film transferred onto the surface of Ecoflex due to the triboelectric series. When the Al film moved away from the Ecoflex, the electrostatic induction from the Al film was weakened. AgNWs gained electrons from the ground to maintain the overall electrostatic balance and became positively charged. Until the Al layer was farthest away from Ecoflex, the positive charges in AgNWs eventually reached an electrostatic balance with Ecoflex. There were several millimeters away to reach the highest open‐circuit voltage. When the Al film started to approach Ecoflex again, the positive charges in AgNWs decreased until the Al film contacted Ecoflex again. Figure 4 Electrical performance of the SMTE for vertical pressure sensing. a) The single‐electrode working principle of the SMTE as the vertical pressure sensing. The triboelectric device was shown in the inset. b,c) The open‐circuit voltage of the SMTE. b) The real‐time signals from 0.05 to 0.55 kPa; c) the curve of voltage to pressure in the range below 15 kPa. d,e) The output transfer charges of the SMTE. d) The real‐time signals from 2.5 to 8.75 kPa; e) the curve of transfer charges to pressure in the range below 40 kPa. f) The relationship between the short‐circuit current of the SMTE and vertical pressure. g) The cycling current test for the SMTE under 6.67 kPa pressure and 0.5 Hz frequency. © 2021 WILEY‐VCH GmbH The experiment tested the electrical performance of the SMTE for vertical pressure sensing. In this part, a measurement system was set up to characterize the electrical performance of the SMTE for vertical pressure sensing, as shown in Figure S6, Supporting Information. In the experimental setup, the SMTE was fixed on a flat plate and faced a dynamometer. The periodically vertical pressure is applied through a linear motor, whereas the electrical signals were recorded in real‐time by the data acquisition device. The size of the SMTE was 1.5 × 1.5 cm 2 in the experiment. The relationship between the measured open‐circuit voltage ( V \n OC ) and vertical pressure is shown in Figure 4b,c . Figure 4b shows the real‐time pressure signals from 0.05 to 0.55 kPa. As the applied pressure increased, the V \n OC increased. The generated voltage signals from the triboelectric pressure sensing were recorded by a real‐time monitor, as shown in Video S2, Supporting Information. Moreover, the electrostatic potential was maintained as the pressure load was on the SMTE. The voltage to pressure curve is shown in Figure 4c , by concluding the V \n OC values in the pressure ranging from 0.05 to 13.75 kPa. Under 13.75 kPa, the V \n OC increased to 28.3 V finally. In addition, the curve was mainly divided into three regions according to the different sensitivities in the large pressure region. In the first low‐pressure region (0.005–0.68 kPa), the SMTE had a higher‐pressure response sensitivity of 40.22 V kPa −1 , and in the second medium‐pressure region (0.68–3.75 kPa), the sensitivity decreased to 2.04 V kPa −1 . Finally, the sensitivity was as low as 0.25 V kPa −1 in the third high‐pressure region (3.75–13.75 kPa). The difference in sensitivity was mainly due to the states of contacting surface and the dielectric shutter between Al and Ecoflex. The Ecoflex and its grating structure had a small deformation at the low‐pressure region but apparent contact electrification and electrostatic induction, which achieved the highest sensitivity. While the pressure was in the medium region, there was an increase in the contact area and pressing strain. However, the electrical output of the SMTE was limited by the relatively small change in the contact area, which reduced the sensitivity to a certain extent. The dielectric layer was then more difficult to compress at the high‐pressure region, so the sensitivity became small. The contact area and compressive modulus were two competitive factors for the voltage signals. Compared with the microengineering SMTE, the flat‐surface device was tested to study the voltage to vertical pressure curve, as shown in Figure S7, Supporting Information. The highest V \n OC reached 35 V due to the larger contact area. However, the sensitivity decreased in the low‐pressure region because of the larger compressive modulus. The medium and high pressure imposed the larger contact area to improve the sensitivity. Similarly, the changes of transfer charges ( Q ) to the vertical pressure were measured. Figure 4d shows the corresponding Q in the pressure range of 2.5–8.75 kPa. The vertical pressure enhanced the output Q to over 6 nC at the pressure of 8.75 kPa. The various Q values were concluded in the pressure range from 2.5 to 35 kPa to draw a Q to pressure curve in Figure 4e . The curve was mainly divided into three pressure regions according to the different sensitivity. In the first low‐pressure region (lower than 4.9 kPa), the SMTE had a pressure sensitivity of 0.61 nC kPa −1 . The sensitivity decreased to 0.14 nC kPa −1 in the medium‐pressure region (4.9–15.2 kPa), further decreasing to 0.03 nC kPa −1 in the third high‐pressure region (15.2–35.1 kPa). Moreover, the output short‐circuit current ( I ) was related to the vertical pressure. There was a current increase under larger vertical pressure, as shown in Figure 4f . Different from the voltage, the current followed a dynamic mechanical response. The cycling test for the current characterized the electrical stability and durability of the SMTE. The device was set under 6.67 kPa pressure and tested for over 6000 cycles, as shown in Figure 4g , displaying excellent electrical stability. 2.5 Demonstration as Finger SMTEs The SMTE worked based on the in‐plane visual strain sensing and vertical pressure sensing. The SMTEs were attached on the fingers of a hand model. The hand model was utilized to grasp a shuttlecock and a stamp, with different gestures shown in Figure \n 5 a,b . The bending finger joints drove the SMTEs, of which the deformation could be observed in the natural light, UV, and UV(dark) conditions. Video S3, Supporting Information, shows the bending finger SMTE that was excited by UV light. As the strain increased, the orange bottom layer was exposed to the environment. The SMTE color was related to the bending degree of the joint, which was described in Figure S8, Supporting Information. The spectra of finger SMTEs were close measured by a portable spectrometer with an extended optical fiber. There were three joints at each finger of the hand model except for the thumb with two joints. The SMTE could reflect the local strain that was induced by the bending joints. The visual strain changes could be observed in the Video S3, Supporting Information. Figure 5c,d shows the visual strain distribution of two different gestures, which was described by the peak intensity ratio of 584 to 484 nm at the joints. The gestures of grasping shuttlecock and stamp were translated into different optical signals. The results in Figure 5a–d show that the SMTE could quantitatively display the local strains in visual changes, which was easy to recognize by human eyes or cameras instead of lots of wiring and power units. Furthermore, SMTE was based on a highly reversible and stable structure to achieve visual changes. These visual changes were controllable by the microcracks and composite, which could display its characteristics under natural and UV light. An overview of the color distribution or the detailed postprocess could easily recognize the gesture. Furthermore, the SMTE could be utilized for display optics. The experiment utilized a “BINN, GXU” patterned SMTE to describe the anisotropic properties. The “BINN” were composed of parallel microcracks in the longitudinal direction, whereas the microcracks of “GXU” were in the transverse direction. When stretching from the original shape, different orientations induced the color changes of the letters, as shown in Figure 5e . As the manufacture methods of microcracks and color composites were controllable, the SMTE could be fabricated into complex outlines and shapes, which expanded its usage scenarios. Figure 5 The electrical and visual performance of the finger SMTEs. a,b) Photos of the finger SMTEs to display the strain distribution based on the color change. Two gestures were tested to grasp the a) shuttlecock and b) stamp under the natural light, UV light, and UV (full dark) environments. The insets at the left bottom corners were the corresponding objects. c,d) The ratio of the PL intensity of the finger SMTEs at 584 and 484 nm when grasping the c) shuttlecock and d) stamp. The light signals were collected from three joints of the fingers except for the thumb with two joints. e) Photos of “BINN, GXU” on the SMTE in transverse and longitudinal tensile directions. f) Open‐circuit voltage of a finger SMTE in the conditions of different pressure, different locations, and large effective working regions. g) Open‐circuit voltage from the finger SMTEs on five fingers. © 2021 WILEY‐VCH GmbH The finger SMTE could also respond to external pressure while maintaining the visual strain sensing. Figure 5f shows the open‐circuit voltage of a finger SMTE in the conditions of different pressure, different locations, and large effective working regions. The pressure was applied by finger touch. Under relatively light pressure, the output voltage of the SMTE was about 5.2 V, and it increased to about 9.8 V under a larger pressure. Similarly, the output voltage at different parts of the bending finger SMTE was measured. The simultaneous pressure at several locations could also enhance the voltage output. However, the variation of voltage was usually related to the contact area. For example, there was a difference in voltage signals between the joint and phalanx, which should be standardized to precisely sense in different conditions. Moreover, the SMTE could work as an electrically tactile array, as shown in Figure 5g . Each finger SMTE could work independently to sense the pressure. The experiment tested the voltage signals at joint 2 when the fingers were bending. The bending degrees in different fingers influenced the output voltage. The voltage that could be read out at the thumb was about 11.8 V and was about 2.51 V at the little finger. In this way, SMTE could sense small touches, demonstrating the coexist of self‐powered electrical and visual sensing and supplying more tactile information. Another Supporting Information, Video S4, displayed that the external pressure had little influence on the visual strain of the SMTE supported by a hard substrate. In a word, the SMTE had potential applications for smart skins, human–machine interaction, and prostheses."
} | 8,131 |
39935465 | PMC11811698 | pmc | 2,325 | {
"abstract": "The increasing demand for rare earth elements (REEs) necessitates the development of more efficient and environmentally friendly leaching methods. This study investigates the use of biological pretreatment to improve metal recovery from REE ore obtained from the Mushgia Khudag deposit. Characterization of the ore revealed a total REE content of 6.99%, with X-ray diffraction analysis identifying calcite and apatite as the dominant minerals, while REEs were primarily found in the forms of monazite and parisite. Experimental results demonstrated that ore pre-treated with a mixed thiobacteria culture (Tmix) achieved a 1.40-fold increase in metal recovery compared to direct acid leaching. Additionally, Bacillus sp. ( B. sp. ) bacteria improved recovery by 1.07-fold. Monitoring changes in pH, oxidation–reduction potential (ORP), and zeta potential during the pre-treatment process indicated that the bacteria did not directly dissolve the REEs but rather modified the surface charge and mineral structure of the ore, facilitating more efficient acid leaching. The use of Tmix bacteria for pretreatment significantly improved leaching efficiency, reduced acid consumption, and minimized environmental impact.",
"conclusion": "4. Conclusion This study highlights the significant potential of bio-pretreatment in enhancing the extraction of REEs from monazite ore using Tmix and B. sp. bacteria. The results demonstrate that this biological pretreatment approach improved the metal recovery rates by 1.4 and 1.07 times, respectively, compared to acid or biological leaching under ambient conditions. The successful application of microorganisms underscores their capacity to facilitate the leaching of REE-bearing minerals, offering an environmentally friendly alternative to traditional extraction approaches. The bioleaching process effectively facilitated the dissolution of apatite, the primary mineral in the ore, promoting the release of associated REE minerals. This was confirmed through FTIR and XRD analyses, which also revealed distinct patterns in the pH, ORP, and surface charge. Specifically, the monazite ore exhibited a negative surface charge, while the Tmix bacteria carried a positive charge. These contrasting charges enabled both direct and indirect interactions that enhanced mineral leaching. In comparison, the B. sp. bacteria were limited to indirect interactions, suggesting their suitability for pretreatment purposes. While these findings are promising, further optimization of the bio-preparation process is necessary to maximize the recovery efficiency. Ongoing research into the underlying mechanisms of bioleaching will provide further valuable insights, and should facilitate the development of sustainable REE extraction methods through the integration of biological and chemical approaches.",
"introduction": "1. Introduction In recent years, rare earth elements (REEs) have gained significant attention owing to their critical role in renewable energy technologies, electrical devices, and nuclear applications. Naturally, REEs occur in various forms, with monazite (20%), bastnasite (70%), ion-exchange clays (7%), xenotime (2%), and other forms (1%) by weight. 1 Among these, REE ores derived from monazite are particularly rich in light rare earth elements such as La, Ce, Pr, and Nd. 2–4 Monazite, a phosphate mineral with the composition (Ce, La, Nd, Th)PO 4 , has a crystalline structure that imparts high thermal stability, making it challenging to leach. Consequently, leaching and extracting these elements typically require strong acids or alkalis under controlled conditions. 5–7 Commonly employed acids in acid leaching systems for REEs containing monazite include hydrochloric acid, 8,9 phosphoric acid, 10 nitric acid, 11,12 and sulfuric acid. 7,13,14 Under ambient conditions, the metal recovery of monazite-containing ores through acid leaching usually ranges from 30% to 70%. 15,16 However, this recovery can significantly increase to as much as 98% at elevated temperatures. 2,7 For example, Harry Watts et al. achieved a total rare earth element (TREE) recovery of 95.2% by leaching monazite sands with concentrated (85%) phosphoric acid at high temperatures of 260 °C. 10 Similarly, using 98% sulfuric acid, Helaly et al. studied monazite leaching in a temperature range of 160–300 °C and achieved a maximum metal recovery of 93.7% at 220 °C. 13 Additionally, Panda et al. conducted a two-step leaching process with diluted HCl, followed by 6 N HCl on Korean monazite, achieving a metal recovery increase of 3.6 times. 8 Kuzmin et al. also demonstrated a TREE recovery of 97.4% by leaching Chuktukon ore containing monazite with 6 M nitric acid at 200 °C for 2 hours. 11 Despite the high efficiency of REE leaching through acid methods, these processes are energy-intensive and generate toxic gases and liquid phases that require further neutralization. To address these challenges, implementing pretreatment processes that facilitate leaching under milder conditions presents a potential solution. 17 However, research into the pretreatment of monazite ores remains limited. For instance, Mei Li et al. roasted a mixed rare earth concentrate (53.59% bastnasite-rare earth oxide and 8.75% monazite-rare earth oxide) with sodium hydroxide at 550 °C for one hour, followed by leaching with 6 M hydrochloric acid at 90 °C, achieving a metal recovery of 92.6%. 18 Although activation methods show promise, they still require high temperatures, extensive chemical use, significant energy consumption, and may have negative environmental impacts. Therefore, finding ways to enhance the efficiency under mild or ambient conditions remains a considerable challenge. Recently, there has been growing interest in bioleaching, an environmentally friendly, low-cost, and sustainable technology, as a viable method for leaching REEs under ambient conditions. 3,19,20 Several studies have explored this approach. For instance, Fathollahzadeh et al. used Enterobacter ( E. ) aerogenes and Acidithiobacillus ( A. ) ferrooxidans bacteria to leach REE-rich monazite and florencite minerals at 30 °C and 120 rpm over 12 days. They observed a continuous increase in REE leaching over time, with significant enhancements in the presence of both bacteria due to synergistic effects. It was suggested that E. aerogenes , which is known to be tolerant of acidic conditions and capable of producing organic acids, positively influenced the leaching environment created by A . ferrooxidans . 21 Additionally, Corbett et al. investigated the impact of carbon sources on bioleaching using Klebsiella ( K. ) aerogenes , Burkholderia T48, Pseudomonas ( P. ) putida , and Gluconobacter ( G. ) oxydans on high-grade monazite ore. Their findings revealed that fructose significantly enhanced the leaching efficiency when used with K. aerogenes and Burkholderia T48, while P. putida and G . oxydans were more effective with galactose. 20 Furthermore, Fathollahzadeh et al. examined the indirect mechanisms of bioleaching, focusing on bacterial interactions. Their study showed that organic acids produced as by-products of bacterial metabolism played a crucial role in dissolving REEs. 21 Overall, the bioleaching of rare earth elements is an environmentally friendly method that operates under milder conditions compared to traditional acid leaching. While it generally requires a longer duration and results in relatively lower recovery, incorporating bacteria in the pretreatment phase can significantly enhance the leaching efficiency. This biological approach can aid in the effective release of rare earth elements from ores, making the process more sustainable and efficient. Building upon these findings, this research aimed to isolate bacterial cultures from natural ores and apply bio-pretreatment under standard conditions to enhance the REE recovery from monazite-containing ores. In this study, REE ore from the Mushgia Khudag deposit was treated with bacterial cultures at room temperature for 7 days. Following bacterial treatment, the ore was filtered, dried, and subjected to acid leaching, which resulted in a total metal recovery increase of 1.4 times compared to the untreated ore.",
"discussion": "3. Results and discussion 3.1 Chemical and phase analysis of the raw material Geological investigations have confirmed that the economically significant ore minerals at the Mushgia Khudag deposit include parisite, calcite, apatite, phosphate, and monazite. These minerals are associated with ore bodies composed of phosphate–carbonate, sulfate–silicate–carbonate, carbonate, and quartz–fluorite–carbonate veins and lenses. 29,30 A comprehensive understanding of the leaching behaviour of REE ores requires detailed knowledge of the chemical composition and mineral phases present in the initial sample. The chemical composition of the Mushgia Khudag ore analysed in this study is provided in Table 1 . Chemical composition of the Mushgia Khudag ore Content, wt% SiO 2 Al 2 O 3 CaO MgO TFe 2 O 3 TiO 2 K 2 O Na 2 O MnO P 2 O 5 21.28 1.74 26.21 0.46 10.98 0.22 0.66 0.45 0.25 14.61 LOI* CO 2 SO 3 Sr La Ce Pr Nd Sm Y 8.96 6.23 6.39 2.53 2.32 2.84 0.67 0.87 0.09 0.08 Content, mg kg −1 As Ba Cd Co Cr Cu Ga Gd Ni 130.34 306.34 1.65 14.00 18.15 76.97 107.14 250.91 32.75 Be Bi V Zn Zr Yb Lu Se Ta 9.50 63.80 153.88 425.53 116.23 43.31 7.32 71.80 120.27 Tb Mo Nb Er Dy Eu Li Pb Te 43.77 68.38 75.56 132.68 542.60 129.54 51.34 274.21 161.65 The initial sample was characterized by high concentrations of SiO 2 , CaO, P 2 O 5 , and TFe 2 O 3 , each exceeding 10%. In contrast, Sr, CO 2 , and SO 3 were present in moderate amounts, ranging from 2.5% to 6.4%. The ore's economic significance stems from a notable concentration of light REEs, especially lanthanum (La) and cerium (Ce), accompanied by significant amounts of praseodymium (Pr), neodymium (Nd), yttrium (Y), and samarium (Sm). These elements are critical to various high-tech industries, including light emitting diodes, lasers, electric vehicles, and NiMH batteries, making efficient REE extraction from this ore strategically and economically valuable. The total REE content was 6.99%, classifying the Mushgia Khudag ore as light REE-rich, consistent with previous studies. 31 To further correlate the chemical composition with the corresponding mineral phases, X-ray diffraction analysis was performed. The results of the analysis are presented in Fig. 2 . Fig. 2 X-ray diffraction pattern of the ore sample from the Mushgia Khudag deposit. Consistent with the chemical analysis ( Table 1 ), the Mushgia Khudag ore contained minerals such as quartz (SiO 2 ), hydroxyapatite (Ca 5 (PO 4 ) 3 OH), calcite (CaCO 3 ), gypsum (CaSO 4 ·2H 2 O), fluorite (CaF 2 ), goethite (FeOOH), and celestine (SrSO 4 ). XRD analysis further revealed peaks corresponding to REE-bearing minerals, including monazite ((La, Ce, Nd)PO 4 ) and parisite (Ca(Ce, La) 2 (CO 3 ) 3 F 2 ). These findings align with the observations of Jargalan, 30 who reported that monazite, the primary REE-bearing mineral, occurs in association with apatite, often filling the spaces between apatite grains or along fractures within the apatite structure. 3.2 Combined acid and bacterial leaching The bioleaching was conducted over 7 days using B. sp. , isolated from the Mushgia Khudag ore, and Tmix bacteria, derived from high-grade chalcopyrite copper ore. After the bioleaching step, sequential acid leaching was performed to compare the efficiency of both leaching processes ( Fig. 3 ). This study showed how bacterial pretreatment improved the subsequent acid leaching. Bioleaching significantly enhanced REE recovery from the Mushgia Khudag ore. The bioleaching of REE ore using the B. sp. and Tmix bacteria resulted in relatively low REE concentrations in the leachate, with recovery rates of merely 0.003% (1.12 mg L −1 ) and 2.4% (838.68 mg L −1 ), respectively. These findings indicate that the bacteria were not highly effective in directly leaching REEs into the solution. Certain types of bacteria, including Thiobacteria , 32 Pseudomonas , 33,34 Mesorhizobium , 33 Acetobacter , 33 Enterobacter , 32,34 Burkholderia , 3,20 Klebsiella , 20 and Pantoea , 34 have been investigated for the bioleaching of rare earth elements (REEs) from monazite ore or concentrate. Among these studies, the highest total REE (TREE) recovery recorded was 25.5 mg L −1 , 34 while the lowest recovery was 4 mg L −1 . 32 In comparison to our study, although these values exceed the leaching efficiency obtained with B. sp. , it was noteworthy that the TREE recovery associated with Tmix surpassed that of all the aforementioned studies. This finding suggests that while B. sp. may yield lower TREE recovery, the efficiency of Tmix resulted in a higher concentration of TREEs, underscoring their potential applicability in bioleaching processes. Fig. 3 TREE recovery for the different leaching systems. However, acid leaching following bacterial pretreatment showed a significant improvement in REE recovery. The total REE recovery after pretreatment with Tmix and B. sp. bacteria increased from 3173.44 mg L −1 to 4539.50 mg L −1 and to 3387.97 mg L −1 , respectively, compared to REE recovery from direct acid leaching. These results suggest that, although the bacteria may not directly dissolve REEs, Tmix bacteria, in particular, play a crucial role in pre-treating the ore. This pretreatment likely alters the surface area and surface charge of associated minerals, thereby enhancing the effectiveness of the subsequent acid leaching. 35 \n Fig. 4 presents the individual metal recovery for each REE, comparing the results of acid leaching with and without bacterial pretreatment. The data highlight the significant impact of bacterial pretreatment on enhancing metal recovery, with improved leaching efficiency demonstrated for most REEs following microbial treatment. Fig. 4 Metal recovery of REEs with different leaching systems. In all cases, the use of Tmix bacteria prior to acid leaching led to the most significant improvement in metal recovery. Notably, Ce and Sm demonstrated the greatest enhancements, with recovery rates 1.44 times higher than those obtained from acid leaching alone. The smallest increase with Tmix bacteria, 1.15 times, was observed for Er. In contrast, with the B. sp. bacteria, Ce and Sm showed the highest recovery increases of 1.07 times, while Er had the smallest increase at 1.04 times. 36 Several studies have explored various approaches to enhance REE recovery through thermal pretreatment followed by acid leaching. For example, Archana reported that roasting monazite (containing 13.5% lanthanum) with KOH at 250 °C, followed by leaching with 1.0 M HCl and 10% H 2 O 2 at 80 °C for 120 min, achieved an 80% TREE leaching efficiency. 37 However, while this approach demonstrated high efficiency, it required the use of highly concentrated acids and strong oxidizing agents, along with substantial energy consumption and harsh processing conditions, posing environmental and operational challenges. In contrast, bioleaching and bacterial pretreatment approaches using Tmix and B. sp. bacteria offer a more sustainable alternative with the potential to reduce energy consumption and minimize environmental impact. However, further optimization of these biological pretreatment approaches is necessary to improve the leaching efficiency and make them viable for large-scale REE recovery. 3.3 Potential application of bacteria for the pretreatment of acid leaching The ability of Tmix bacteria to enhance the leaching of REE ores indicates its potential as an effective pretreatment approach prior to acid leaching. To evaluate the feasibility of this approach, it was crucial to monitor several key parameters throughout the leaching process. These included pH fluctuations, changes in ORP and surface charge, and alterations in the mineral composition of the solid phase ( Fig. 5 ). Fig. 5 (a) pH and (b) ORP values for the different pretreatment and acid leaching approaches. When the ore was treated with Tmix and B. sp. bacteria, the pH of the leachate increased by 4.1 and 0.6, respectively. After subsequent acid leaching, the pH rose further by 0.91 and 0.94. In comparison, when untreated ore was directly subjected to acid leaching, the pH increase was only 0.69. As the REE ore originates from a carbonatite deposit and contains a high concentration of calcite, the interaction between the carbonate rock and hydroxide ions from the Tmix bacterial solution resulted in a sharp increase in pH. This increase was attributed to the leaching of carbonate minerals under the influence of hydroxide ions, creating favourable conditions for efficient acid leaching. Acid leaching of the ore pretreated with Tmix bacteria resulted in a significantly higher metal recovery compared to that obtained from direct acid leaching ( Fig. 3 ). An interesting trend was observed regarding the ORP ( Fig. 5b ). When the ore was treated with bacteria, the ORP decreased initially but then increased during subsequent acid leaching. Typically, the ORP increases when oxidized species dominate the system and decreases in the presence of reduced species. The initial ORP values of the bacterial solutions were 618 mV for Tmix and 441 mV for B. sp. The ORP is influenced by several factors, including the solution's pH, temperature, and chemical species, such as dissolved oxygen. Since the leaching experiments were conducted at a constant temperature, the observed increase in ORP during acid treatment of the bacteria-pretreated ore could be attributed to the influence of oxygen in the system, aligning with the following half-reaction: O 2 + 4H + + 4e − → 2H 2 O As no aerobic bacteria were present in the system, this reaction directly reflects the role of oxygen during the acid leaching process. Conversely, when the ore was treated with bacteria, the pH of the solution increased while the ORP decreased. This reduction in ORP was due to the consumption of dissolved oxygen during bacterial respiration. As the bacteria metabolize, the oxygen levels decrease, resulting in a lower ORP value, which, combined with the increase in pH, leads to enhanced leaching conditions. To further evaluate the effect of bacterial treatment on the mineral phases of the REE-containing rock, XRD analysis was conducted on the solid residue following pretreatment of the Mushgia Khudag ore with Tmix and B. sp. bacteria, followed by acid leaching. The results of this analysis are presented in Fig. 6 . Fig. 6 XRD patterns after bio-pretreatment and acid leaching. The analysis showed that in both cases where the rare earth element oxide was pretreated with bacteria, the intensity of the apatite peak decreased, while the gypsum peak intensity increased following treatment with Tmix bacteria. The increase in the gypsum peak was attributed to the high concentration of sulfate ions (15.2 g L −1 ) present in the 9 K nutrient medium used for the Tmix bacteria. Gypsum formation was also linked to the release of calcium ions resulting from the leaching of apatite and calcite minerals, as described by eqn (2) and (3) . 2 CaCO 3 + 2H + + SO 4 2− + H 2 O → CaSO 4 ·2H 2 O + CO 2 3 Ca 5 (PO 4 ) 3 OH + 10H + + 5SO 4 2− + 9H 2 O → 3H 3 PO 4 + 5CaSO 4 ·2H 2 O After acid leaching, peaks corresponding to minerals such as apatite, calcite, and fluorite, which are known to dissolve in acidic conditions, could no longer be detected in the X-ray diffractogram. In contrast, the peak intensities of quartz and goethite, which are insoluble in acid, remained unchanged, while the peak intensity of celestine decreased ( Fig. 5b ). The reduction in celestine's peak intensity was likely not due to its leaching but rather the overlapping of its peaks with those of the newly formed minerals. Meanwhile, the peak intensity of gypsum increased, and the presence of bassanite, a new mineral phase, could be identified, which likely formed during the acid leaching of the REE ore through specific interactions, as described by eqn (4) . 4 2CaCO 3 + 4H + + 2SO 4 2− → 2CaSO 4 ·1/2H 2 O + 2CO 2 + H 2 O 5 2Ca 5 (PO 4 ) 3 OH + 20H + + 10SO 4 2− + 3H 2 O → 6H 3 PO 4 + 10CaSO 4 ·1/2H 2 O The bacterial leachate was acidic, with pH values of 1.6 for Tmix and 4.0 for B. sp. However, the system's pH sharply increased due to the leaching of ores rich in calcite, as shown in eqn (2)–(5) ( Fig. 5a ). Additionally, the decrease in peak intensity of the REE minerals monazite and parisite suggested that the leaching of associated host rocks, such as calcite and apatite, led to a release of REE minerals, thereby enhancing the overall leaching efficiency. To confirm whether the chemical bonds and compositions changed after bacterial and acid treatment, FTIR spectral analysis was performed, with the results presented in Fig. 7 . Fig. 7 FTIR spectra after bio-pretreatment and acid leaching. The initial analysis of the infrared spectrum revealed the characteristic peaks of calcite 38 at 875.68, 1423.47, 1782.23, and 2515.18 cm −1 ; however, these peaks disappeared following bacterial pretreatment. This observation further corroborated the dissolution of calcite mediated by bacterial activity ( eqn (2)–(5) ). In all cases, except for the initial ore, new peaks corresponding to P–O stretching vibrations at 460 cm −1 and P \n \n\n<svg xmlns=\"http://www.w3.org/2000/svg\" version=\"1.0\" width=\"13.200000pt\" height=\"16.000000pt\" viewBox=\"0 0 13.200000 16.000000\" preserveAspectRatio=\"xMidYMid meet\"><metadata>\nCreated by potrace 1.16, written by Peter Selinger 2001-2019\n</metadata><g transform=\"translate(1.000000,15.000000) scale(0.017500,-0.017500)\" fill=\"currentColor\" stroke=\"none\"><path d=\"M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z\"/></g></svg>\n\n O stretching vibrations at 601.79 cm −1 were identified. 39 This change was indicative of the dissolution of apatite in the ore during both the pretreatment and acid leaching processes ( eqn (5) ), resulting in phosphate ions being adsorbed onto the surfaces of the solid residual particles. Following the treatment with Tmix and subsequent acid leaching, two weak signals were observed at approximately 601.79 and 667.37 cm −1 , as well as at 1118 and 1143 cm −1 , which were attributed to the asymmetric stretching vibrations of sulfate. 40,41 Moreover, the peaks observed around 1622.13–1685.79 and 3408.22–3549.02 cm −1 corresponded to the bending and stretching vibrations of water molecules present in gypsum. 42,43 This observation further validated the formation of gypsum ( eqn (2)–(5) ), as supported by the XRD analysis results, illustrating the influence of the sulfate ions derived from the Tmix cultural medium and sulfuric acid. The leaching of associated minerals, such as apatite and calcite, through bacterial involvement altered the surface charge of the ore. Therefore, to evaluate the impact of bacterial treatment on the leaching process, the surface charge was measured, with the results presented in Fig. 8 . Fig. 8 Changes in surface charge: (a) bacterial culture, (b) after bacterial pretreatment, (c) after acid leaching. The particles of the REE ore were negatively charged, 28 resulting in a repulsion of similarly charged species or the attraction of oppositely charged species. In parallel, the B. sp. bacteria also exhibited a negative charge, while the Tmix bacteria were positively charged ( Fig. 8a ). As the negatively charged Mushgia Khudag ore interacted with ions and B. sp. bacteria, the zeta potential value decreased in both cases ( Fig. 8b ). This decrease indicated that surface interactions occurred, as corroborated by the XRD and FTIR analysis results. The attraction of positively charged ions to the negatively charged ore facilitated surface interactions, leading to the formation of smaller, negatively charged particles. 44 Furthermore, the reduction in particle size was correlated with an increase in the number of negatively charged ore particles, 45 resulting in a significant decrease in zeta potential. The increase in the number of smaller particles suggested a higher leaching rate. After acid treatment, the surface charge increased in all cases ( Fig. 8c ), indicating that the acid interacted with the ore particles, further modifying their surface properties. 44 The increase in surface potential during acid leaching of the ore treated with Tmix bacteria was greater than that observed with the B. sp. bacteria. The interaction between the strongly positively charged Tmix bacteria and protons led to a significant rise in surface potential. In this case, the amount of dissolved oxygen in the initial leaching solution played a primary role, indicating that the interaction of O 2 and protons predominated over the bacterial count. This finding supports the observation that metal recovery from acid leaching was higher for the ore pretreated with Tmix bacteria compared to ore pretreated with B. sp. bacteria ( Fig. 3 ). This suggests that B. sp. bacteria may have primarily facilitated bioactivation through indirect interactions with the Mushgia Khudag ore. In contrast, the Tmix bacteria, due to their positive charge, attracted negatively charged particles, facilitating interactions with the REEs through a cooperative mechanism. 46 Additionally, hydronium ions produced by Thiobacillus species , along with organic acids released as metabolic by-products from Bacillus species , 47–49 played a crucial role in this bioleaching process. 46 As illustrated in Fig. 5 , the bacterial activity throughout the ore treatment was monitored by measuring the pH and ORP. The free protons or hydronium ions derived from Thiobacillus , and the organic acids produced by Bacillus species interacted with REE ore particles, 46 leading to the formation of a solution containing anions such as CO 3 2− and PO 4 3− , as well as cations, including Ca 2+ and REE 3+ . These cations tend to form complexes with trace amounts of organic acid residues, 46,47 while the anions adsorb onto the surfaces of the particles, 47 resulting in the generation of negatively charged particles. The increased intensity of the PO 4 3− bands observed in the infrared spectra aligned with the adsorption of PO 4 3− onto the particles, which contributed to their negative charges. As noted in the results regarding the changes in surface charge, the degree of negative charge on the particles treated with Tmix was significantly greater than that observed with the particles treated with Bacillus species . This discrepancy could be attributed to the more effective cooperative mechanism employed by Tmix compared to the indirect mechanism utilized by B. sp. , thereby enhancing the overall leaching efficiency."
} | 6,732 |
37328514 | PMC10275999 | pmc | 2,326 | {
"abstract": "Reservoir computing (RC) offers efficient temporal information processing with low training cost. All-ferroelectric implementation of RC is appealing because it can fully exploit the merits of ferroelectric memristors (e.g., good controllability); however, this has been undemonstrated due to the challenge of developing ferroelectric memristors with distinctly different switching characteristics specific to the reservoir and readout network. Here, we experimentally demonstrate an all-ferroelectric RC system whose reservoir and readout network are implemented with volatile and nonvolatile ferroelectric diodes (FDs), respectively. The volatile and nonvolatile FDs are derived from the same Pt/BiFeO 3 /SrRuO 3 structure via the manipulation of an imprint field ( E imp ). It is shown that the volatile FD with E imp exhibits short-term memory and nonlinearity while the nonvolatile FD with negligible E imp displays long-term potentiation/depression, fulfilling the functional requirements of the reservoir and readout network, respectively. Hence, the all-ferroelectric RC system is competent for handling various temporal tasks. In particular, it achieves an ultralow normalized root mean square error of 0.017 in the Hénon map time-series prediction. Besides, both the volatile and nonvolatile FDs demonstrate long-term stability in ambient air, high endurance, and low power consumption, promising the all-ferroelectric RC system as a reliable and low-power neuromorphic hardware for temporal information processing.",
"introduction": "Introduction Deep learning is progressing rapidly and plays an increasing role in industry and daily life. It mainly relies on two types of neural network algorithms: feedforward neural network (FNN) and recurrent neural network (RNN), which are adept at handling static spatial and dynamic temporal tasks, respectively. Reservoir computing (RC) is a simple yet efficient type of RNN well suited for processing temporal information 1 – 3 . An RC system typically consists of a reservoir that nonlinearly maps the time-varying inputs into a high-dimensional feature space, and a readout network that performs further processing through a linearly weighted summation of the reservoir outputs (see Fig. 1a ) 3 . During training, only the readout network needs to be trained while the reservoir does not. The training cost can thus be significantly reduced, which represents the most outstanding advantage of RC over other RNNs. Fig. 1 Concept of all-ferroelectric RC system. a Schematic of an RC system consisting of a reservoir with internal dynamics and a readout network. The inputs are projected into a high-dimensional feature space through the reservoir and then analyzed by the readout network. Only the weights in the readout network, i.e., W out , need to be trained. b Schematic of an all-ferroelectric RC system, where the inputs are encoded as pulse trains while the reservoir and readout network are implemented with the volatile and nonvolatile FDs, respectively. Schematics of c volatile FD with E imp and d nonvolatile FD without E imp . The E imp can cause polarization back-switching and consequent conductance decay. “TE”, “FE” and “BE” denote the top electrode, ferroelectric, and bottom electrode, respectively. Recently, emerging hardware-based RC systems have attracted great attention, not only because they have achieved prediction performance comparable to that of the software-based counterparts in many tasks (e.g., pattern classification 4 , 5 , speech recognition 6 – 8 , chaotic system forecasting 6 , 7 , 9 , and others 10 – 12 ), but also because of their boosted energy efficiency 6 , 13 . For the hardware implementation of an RC system, the constituent reservoir and readout network need to be implemented on memory devices with distinctly different switching characteristics, i.e., volatile and nonvolatile switching characteristics, respectively. Most previous studies have focused on the hardware implementation of the reservoir by using (volatile) diffusive memristors 6 , 7 , 10 , 14 – 22 , nanomagnetic systems 23 (including spintronic oscillators 24 , magnetic nanorings 25 , spin ices 26 , and magnonic systems 27 ), self-organized nano-networks 9 , 28 , electrochemical transistors 8 , 12 , 29 , and so on. Among these devices, the diffusive memristors stand out because they possess intrinsic nonlinearity and short-term memory which are the two essential properties required by the reservoir 3 , as well as high speed and excellent scalability. On the other hand, despite being less studied, the hardware implementation of the readout network has been demonstrated with (nonvolatile) drift memristors 9 , 13 , 21 , 30 , whose nonvolatile conductances are utilized to map the weights in the readout network. Notably, both the diffusive and drift memristors were mainly based on a filamentary mechanism, which can, however, lead to relatively large variations and low endurance due to the stochasticity of filament formation/rupture processes. This limits the prediction accuracy and reliability of the filamentary memristor-based RC system. Compared with filamentary switching, ferroelectric polarization switching is a more deterministic switching mechanism 31 . Ferroelectric memristors, which use polarization switching to tune the resistance 31 , 32 , can thus exhibit highly reproducible memristive responses and potentially unlimited endurance 33 – 37 . Besides, they also show high switching speed and low-power consumption 38 – 40 . Using ferroelectric memristors as building blocks may therefore facilitate the development of highly reliable, accurate, fast, and energy-efficient ferroelectric-based RC systems. However, the use of ferroelectric memristors in RC systems is currently scarce and mainly restricted to the reservoir 11 , 30 , 41 – 45 , as summarized in Supplementary Table S1 . All-ferroelectric implementation of a whole RC system still remains undemonstrated. The reason for this is probably because the ferroelectric memristors used hitherto in RC systems—ferroelectric tunnel junction (FTJ) 30 and ferroelectric field-effect transistor (FeFET) 11 , 43 , 45 —possess inherently large depolarization fields ( E dp s) arising from ultra-small ferroelectric film thickness 46 , 47 and poor screening at ferroelectric/semiconductor interface 48 , respectively. This makes them voluntary to exhibit volatile characteristics while difficult to be engineered into nonvolatile memristors to implement the readout network. To construct an all-ferroelectric RC system, alternative ferroelectric memristors capable of being engineered into both volatile and nonvolatile memristors (for the reservoir and readout network, respectively) are demanded. A promising candidate is a ferroelectric diode (FD) which operates by using polarization to modulate the interfacial Schottky barrier 49 – 51 . FD is inherently subjected to a much smaller E dp compared with FTJ and FeFET, because it comprises a relatively thick ferroelectric film (several tens to hundreds of nanometers) sandwiched between two metal electrodes with good screening ability. Consequently, FD can readily function as a nonvolatile memristor 32 , 50 , 51 . In addition, FD can also be engineered to be volatile by judiciously introducing certain mechanisms for polarization back-switching 52 – 54 . Therefore, it is quite promising to use appropriately engineered FDs to implement both the reservoir and readout network, thus realizing an all-ferroelectric RC system in hardware (Fig. 1b–d ). In this work, we experimentally demonstrate an all-ferroelectric RC system consisting of a volatile FD-based reservoir and a nonvolatile FD-based readout network. The FDs with distinctly different volatile and nonvolatile switching characteristics are derived from the same capacitor-like structure of Pt/BiFeO 3 (BFO)/SrRuO 3 (SRO), which has not been realized yet for other types of ferroelectric memristors. The key to realizing this is purposely introducing an imprint field ( E imp ) into the volatile FD while avoiding it in the nonvolatile FD. Owing to the E imp , the volatile FD exhibits spontaneous polarization back-switching and consequent conductance decay, based on which short-term memory and nonlinearity are further demonstrated. On the other hand, the nonvolatile FD with negligible E imp exhibits good polarization stability and consequent nonvolatile memristive switching, as well as long-term potentiation/depression (LTP/LTD). With these distinctly different device characteristics, the volatile and nonvolatile FDs are thus suitable building blocks of the reservoir and readout network, respectively. We then experimentally integrate them to build an all-ferroelectric RC system. Various tasks including curvature discrimination, digit recognition, waveform classification, and Hénon map prediction, are successfully implemented with the all-ferroelectric RC system, demonstrating competitive performance compared with the existing RC hardware systems. In particular, an ultralow normalized root mean square error (NRMSE) of 0.017 is achieved in the Hénon map prediction. Besides, long-term stability in ambient air, high endurance, and low-power consumption are proven in both the volatile and nonvolatile FDs, which can endow the all-ferroelectric RC system with high reliability and power efficiency. Our study showcases the development of application-specific neuromorphic devices based on ferroelectrics by manipulating the polarization dynamics and also highlights the great potential of ferroelectrics for use in hardware-based neuromorphic computing.",
"discussion": "Discussion In summary, we experimentally demonstrate an all-ferroelectric RC system in which the reservoir and readout network are implemented with the volatile and nonvolatile FDs, respectively. Both the volatile and nonvolatile FDs have the same structure of Pt/BFO/SRO, but the difference is that E imp is purposely introduced into the former while it is absent in the latter. Under the effect of E imp , the volatile FD displays spontaneous polarization back-switching and consequent conductance decay. Short-term memory and nonlinearity are further demonstrated in the volatile FD. These properties enable the volatile FD to produce well-separable responses to different temporal inputs, making it a suitable building block of the reservoir. On the other hand, the nonvolatile FD with negligible E imp exhibits good polarization stability and consequent nonvolatile memristive switching. Moreover, the LTP and LTD functions are implemented in the nonvolatile FD, qualifying it as the synapse in the readout network. Then, an all-ferroelectric RC system consisting of the volatile FD-based reservoir and the nonvolatile FD-based readout network is constructed. The all-ferroelectric RC system is used to solve various tasks including curvature discrimination, digit recognition, waveform classification, and Hénon map prediction, and it achieves competitive performance compared with the existing RC hardware systems. In particular, an ultralow NRMSE of 0.017 is achieved in the Hénon map prediction. Besides, both volatile and nonvolatile FDs demonstrate long-term stability in ambient air, high endurance, and low power consumption, making the all-ferroelectric RC system a reliable and low-power hardware platform for temporal information processing. We expect that our encouraging results will stimulate further research on the ferroelectric implementation of various emerging neuromorphic computing algorithms, e.g., e-prop 64 ."
} | 2,897 |
19384423 | PMC2668766 | pmc | 2,327 | {
"abstract": "Background Coral reefs around the world are experiencing large-scale degradation, largely due to global climate change, overfishing, diseases and eutrophication. Climate change models suggest increasing frequency and severity of warming-induced coral bleaching events, with consequent increases in coral mortality and algal overgrowth. Critically, the recovery of damaged reefs will depend on the reversibility of seaweed blooms, generally considered to depend on grazing of the seaweed, and replenishment of corals by larvae that successfully recruit to damaged reefs. These processes usually take years to decades to bring a reef back to coral dominance. Methodology/Principal Findings In 2006, mass bleaching of corals on inshore reefs of the Great Barrier Reef caused high coral mortality. Here we show that this coral mortality was followed by an unprecedented bloom of a single species of unpalatable seaweed ( Lobophora variegata ), colonizing dead coral skeletons, but that corals on these reefs recovered dramatically, in less than a year. Unexpectedly, this rapid reversal did not involve reestablishment of corals by recruitment of coral larvae, as often assumed, but depended on several ecological mechanisms previously underestimated. Conclusions/Significance These mechanisms of ecological recovery included rapid regeneration rates of remnant coral tissue, very high competitive ability of the corals allowing them to out-compete the seaweed, a natural seasonal decline in the particular species of dominant seaweed, and an effective marine protected area system. Our study provides a key example of the doom and boom of a highly resilient reef, and new insights into the variability and mechanisms of reef resilience under rapid climate change.",
"introduction": "Introduction Coral reefs are among the most biologically diverse and economically important ecosystems. However, reefs are rapidly degrading at a global scale, due to a combination of pressures, including climate change, overexploitation, coral diseases, and declining water quality [1] – [4] . Rising ocean temperatures have triggered mass coral bleaching events that have devastated many coral reefs around the world [5] and caused ecological phase or state shifts, from coral-dominance to dominance by seaweeds (fleshy algae) [6] – [8] . Current climate change models suggest increasing frequency and severity of mass coral bleaching events [5] , so that phase shifts to algal dominated states are expected to occur more frequently and last longer [9] – [11] . Critically, the recovery of degraded reefs depends on the reversibility of seaweed dominance [12] , [13] . However, all previously documented cases have found dominance by seaweeds difficult to reverse, because the algae prevent settlement of new corals, and because the algae persist, usually due to overfishing or mass mortality of key herbivorous species and to relative unpalatability of algae to herbivores [14] , [15] . Examples of natural reversals from algal dominance to coral dominated states are extremely rare (but see [16] , [17] ) and take years to decades to occur (e. g. Kaneohe Bay, Hawaii [18] ; Dairy Bull Jamaica [19] ). Rapid reversals from algal dominated states to dominance by corals and small algae have only been demonstrated at a very small scale after experimentally induced herbivore exclusion [20] . In that experiment, artificially enhanced algal biomass was rapidly consumed by grazers upon removal of exclusion cages, and reef recovery was dependent on recovery of herbivory, a process extrinsic to the corals and algae. Inshore, high latitude coral reefs of the largest reef system in the world, the Great Barrier Reef (GBR), Australia, suffered severe mass bleaching of coral in early 2006. Reefs in the area exhibit low coral species diversity and are widely dominated by Acropora corals, with branching Acropora accounting for more than 90% of the coral species [21] . Sea surface temperatures in the inshore reefs of the Keppel Islands (23°10′S, 151°00′E) in the southern GBR rose rapidly in late 2005, with some locations reaching temperatures in December that are not normally found until February. The onset of high sea temperatures early in the season triggered coral bleaching by mid January 2006 [22] . Overall, bleaching damage was severe, affecting 77–95% of coral colonies [22] , [23] . The purpose of this paper was to document some novel mechanisms for coral reef resilience based on changes in coral and seaweed abundance following the 2006 mass coral bleaching event that affected reefs of the Keppel Islands.",
"discussion": "Results and Discussion Abundance of corals and seaweeds showed strong dynamics in response to the warming-induced mass coral bleaching event ( Figs. 1 , 2 ). Cover of bleached but living coral (mainly branching Acropora spp.) on the reef slopes of Middle Island, Halfway Island, and Barren Island was high (77%–89%) during the bleaching event in January/February 2006. Five months after the onset of bleaching, coral cover was severely reduced, to values around 20–30% by July–August 2006. The coral mortality was followed by an extraordinary bloom of the brown seaweed Lobophora variegata , apparently unprecedented in magnitude on the GBR (GDP and LM personal observations, Fig. 2 ). This alga commonly grows between the branches of most Acropora colonies in the area, but under normal (i.e. undisturbed) conditions it is not able to grow beyond the base of the branches, probably due to competitive inhibition by the corals. Previous work on L. variegata growing amongst branching Porites cylindrica corals showed that the interaction is competitive, with both coral and alga inhibiting growth of the other [24] , [25] . However, seaweeds and algal turfs were apparently released from space competition with the corals due to the bleaching mortality [9] and dramatically increased in cover (200–300% increase on Middle Island and Halfway Island) by August 2006. Importantly, coral bleaching preceded L. variegata overgrowth, and overgrowth only took place on bleached or dead corals at a range of spatial scales (from cm to 10 s of kilometers; careful inspection showed negligible overgrowth of healthy coral). Nonetheless, the seaweed apparently exacerbated coral mortality by overgrowing stressed coral tissue [24] – [26] ( Figure S1D ). Algal competitiveness may have been enhanced by uptake of nutrients and carbon generated by the coral mortality [27] . There are no previous observations of such an extensive bloom of L. variegata , or indeed any single species of fleshy alga, on the GBR, although large-scale blooms of filamentous algal turfs have occurred following coral mortality [9] , [28] , [29] , and a small-scale bloom of a red seaweed was recorded in response to a ship-grounding [30] . Blooms of L. variegata are common in the Caribbean, particularly after the die-off of the sea urchin Diadema \n [14] , [31] and following coral mortality [32] , [33] (also personal observations in Islas del Rosario, Colombia and Flower Garden Banks, Gulf of Mexico, GDP and LM). 10.1371/journal.pone.0005239.g001 Figure 1 Coral bleaching, algal overgrowth of corals and coral recovery. A) Bleached corals in the Keppel Islands, Great Barrier Reef, during the mass bleaching event in January 2006. The fleshy brown seaweed Lobophora variegata grows at the base of the branches of Acropora spp. corals. B) L. variegata is released from space competition by coral mortality and overgrows coral skeletons as well as some coral tissue, causing an unprecedented algal bloom. C) Seaweed bloom on North Keppel Island after coral bleaching. The reef has lost its structural complexity and has experienced little coral recovery. D) Recovered reef on Barren Island, showing high coral cover and low cover of seaweeds. 10.1371/journal.pone.0005239.g002 Figure 2 Coral – algal dynamics in response to the 2006 warming-induced coral bleaching event. Data from the reef slopes of four islands in the Keppel Islands, southern Great Barrier Reef. % cover data are means (n = 10) ±SE, except for Feb 2006 (n = 25–26). CCA: Crustose calcareous algae. Surprisingly however, the cover of branching Acropora corals at most sites showed an extremely rapid recovery after the seaweed bloom, reaching pre-bleaching levels by December 2006–April 2007 (ca 12–14 months after the onset of bleaching, Fig. 2 , Table 1 ). This represents a 100 to 200% increase in cover of Acropora in approximately 6 months, thereby returning the system to coral dominance (P = 0.004, 0.001 and 0.006 for Tukey's comparisons of August 2006 c.f. February/March 2007 for Middle, Halfway and Barren Islands respectively). 10.1371/journal.pone.0005239.t001 Table 1 Two-way analyses of variance for the effects of sampling date and site on % cover of corals, brown seaweed Lobophora variegata , algal turfs and crustose calcareous algae (CCA). Source of variation df Mean-Square F-ratio p \n Coral cover \n Date (D) 5 1.265 20.238 <0.001 Site (S) 3 5.257 84.100 <0.001 D×S 15 0.204 3.262 <0.001 Error 214 0.063 \n Lobophora \n cover \n Date (D) 5 0.541 14.450 <0.001 Site (S) 3 5.249 140.224 <0.001 D×S 15 0.121 3.244 <0.001 Error 214 0.037 \n Algal turf cover \n Date (D) 5 0.161 5.207 <0.001 Site (S) 3 0.096 3.094 0.028 D×S 15 0.184 5.929 <0.001 Error 214 0.031 \n CCA cover \n Date (D) 5 0.203 15.964 <0.001 Site (S) 3 0.424 33.336 <0.001 D×S 15 0.018 1.378 0.160 Error 214 0.013 Data were Arc-sin transformed. Interactions between date and site were significant; therefore, data were analysed for site effects within dates and date effects within sites, using a one-way ANOVAs and Tukey's post-hoc comparisons (results not shown). Unexpectedly, the rapid reversal and increase in coral cover did not involve settlement and recruitment of coral larvae. Coral recruitment was generally very low throughout the course of the study at all sites [recruit densities for Middle, Halfway, Barren and North Keppel Islands were 0, <1, <1 and 4 recruits m −2 respectively; Kruskal-Wallis Test indicated no increases in recruit densities through time after the bleaching event, Table 2 ]. Instead, coral recovery involved a rapid regeneration and regrowth of remnant coral tissue after bleaching mortality, with branches of Acropora emerging from the algal mat to reestablish high cover much faster than could occur from growth of new recruits ( Figs. 2 , 3 ). Growth rates of branching Acropora from the Keppel Islands appear unusually high, with rates of calcification nearly 100% faster than those of corals from offshore the GBR ( Fig. 4 ). Linear extension rates of branching Acropora from other Pacific inshore reefs are also extraordinarily high, with mean values of 333 (±42 SD) mm/year [34] . This rapid, vegetative regeneration allowed the corals to out-compete and overgrow the algae settled on dead skeletons. 10.1371/journal.pone.0005239.g003 Figure 3 Coral recovery following algal overgrowth. Branches of Acropora corals died after bleaching and were subsequently colonized by a variety of benthic algae. Remnant coral tissue at the base of the coral colonies regrew upward and deposited new skeleton along the old dead coral branch, overgrowing A) algal turfs (arrows), B) fleshy seaweed Lobophora variegata , and C) crustose coralline algae. D) Coral tissue has all but completely overgrown the colonizing algae. E) Thin section of coral showing benthic algae sandwiched between old coral skeleton and a thin layer of new skeleton. Examination using a compound microscope showed that coral tissue overgrew a range of algal types. 10.1371/journal.pone.0005239.g004 Figure 4 Coral growth (calcification). Calcification rates of Acropora millepora at North Keppel Island and Davies Reef (an offshore reef). Data are means±SE (n = 12 for North Keppel Island and 8 for Davies Reef), and show unusually high growth rates in the Keppel Islands. 10.1371/journal.pone.0005239.t002 Table 2 Kruskal-Wallis one-way analyses of variance for the effects of sampling date on density of coral recruits of four islands. Source of variation df H p \n Middle I \n Date 5 0.000 1.000 \n Halfway I \n Date 5 5.000 0.416 \n North Keppel I \n Date 4 4.387 0.356 \n Barren I \n Date 5 11.308 0.046 * \n Data were log transformed. H: Statistic of the Kruskal-Wallis test; df: degrees of freedom. Dates included in the analyses are: Aug 06, Dec 06, Feb 07, Jun 07, Aug 07, Jan 08. Data missing for Dec 06 in North Keppel Island. * Recruit density was slightly higher in February 2007 but declined afterwards. We propose that this unusually rapid and successful regrowth stems from several key factors: i. the strong competitive ability of the corals; ii. the corals' ability to regrow from relatively small amounts of live tissue; iii. and a seasonal dieback in the single species of dominant seaweed. Although overgrowth by seaweeds probably inhibited coral growth, a natural seasonal decline in L. variegata , between December 2006 and March/April 2007 ( Fig. 2 ), markedly reduced the apparent effects of this competitive inhibition. Cover of L. variegata decreased significantly from 50% to <20% in Middle Island and from 75% to 45% in North Keppel Island during that period of time ( Table 1 ; P<0.005 for Tukey's comparisons of August 2006 and March/April 2007 for both islands). Removal of the seaweed L. variegata in this study appears to have been largely due to inherent seasonal dieback. Large amounts of loose L. variegata were observed at the time of the dieback, and similar seasonal changes in L. variegata have been previously observed in the GBR ( Fig. 5 ) and nearby areas [35] , apparently related to elevated seawater temperature during the austral spring and summer (GDP unpublished data). Herbivorous fishes, although largely unfished, are not generally abundant in the Keppel Islands, being generally about an order of magnitude less than on mid and outer shelf reefs [36] . Careful observations did not indicate grazing damage to the L. variegata , despite the extent of the bloom and decline, and patterns of herbivore abundance among the study reefs were not consistent with the growth and decline in L. variegata at these sites ( Fig. 6 ). The site with lowest herbivore densities had lowest L. variegata abundance (Barren Island). The site with most abundant scarids had most abundant L. variegata (North Keppel Island), while siganids were most abundant on Halfway Island, which had intermediate abundance of L. variegata . Large invertebrate herbivores, such as sea urchins, were virtually absent across all sites. Thus, whilst herbivory could have contributed to some degree, and is likely important to algal abundance on these reefs generally, the extent of decline in L. variegata in this study appears largely due to seasonality. 10.1371/journal.pone.0005239.g005 Figure 5 Seasonality in Lobophora variegata on Goold Island, inshore central GBR. Abundance of L. variegata consistently shows strong declines during the austral summer. Data are means±SE of 5 replicate quadrats. 10.1371/journal.pone.0005239.g006 Figure 6 Herbivore abundance. Herbivore density data from the study sites; data are square root transformed, means+/−SE of 5 transects per site. However, the increase in coral cover was apparently also due to strong growth rates and consequent competitive ability of the coral, and not dependent on the seasonal decline in the algal competitor, L. variegata . This is suggested by results for Middle Island ( Fig. 2 ; August–December 2006) where strong coral recovery preceded decline in L. variegata , and from Barren Island, where coral recovery involved overgrowth of non-seasonal algal turfs and crustose calcareous algae. Tissue growth may have been enhanced by heterotrophic feeding [37] , as shown elsewhere on GBR reefs [38] . Regeneration of the coral tissue apparently derived from tissue reservoirs, or areas of live coral tissue that persisted at the very base of the coral branches, underneath the seaweed canopy ( Figure S1A, B ; the “phoenix effect” in which apparently dead coral branches regenerate live tissue [39] – [41] ). Removal of the dominant seaweed mat showed that coral tissue mortality was extensive under the seaweed at all sites. However, there did remain small fragments of live coral tissue. The remnant surviving coral tissue rapidly expanded upwards along the dead coral branches ( Fig. 3 ) and actively overgrew L. variegata , as well as a range of other algal types, including filamentous algal turfs, fleshy seaweeds and crustose coralline algae ( Fig. 3A–D ). Thin sections of Acropora corals show overgrowth of several algae by new coral material, and show that overgrowth involved direct horizontal contact as well as overtopping, resulting in a “seaweed sandwich [42] ”, with algae engulfed between new and old layers of skeleton ( Fig. 3E ). Regeneration over existing coral skeletons offers an energetically efficient and rapid mechanism for recovery, by limiting the calcification required for regrowth. Whilst regeneration of corals has been observed elsewhere [42] , [43] , our findings are significant because they demonstrate the potential importance of this process for large-scale, rapid recovery even after severe climate-related mass bleaching. The rate and scale of recovery is increasingly critical as climate change causes more frequent mass bleaching events. Coral recovery and algal dynamics were not uniform in this study. Although most reefs showed rapid recovery, coral cover on North Keppel Island declined from 46% to <10% after bleaching and had recovered relatively little after two years ( Fig. 2 ), despite a marked seasonal decline in L. variegata . Coral cover on North Keppel Island prior to the bleaching event was low compared to the other reefs in the area (46% vs. 75–90% respectively) and cover of L. variegata higher. These differences may reflect differences in disturbance history, conditions less conducive to coral growth, or differences in the extent of coral mortality due to floods from the Fitzroy River (the largest river catchment along the GBR) [44] . Recovery of the reef on North Keppel Island may also have been limited by the loss of three-dimensional structure of the reef framework (most branching Acropora corals have been broken into rubble due to bioerosion, Fig. 1C ; habitat complexity has been shown to be critical for the rapid recovery of damaged reefs [19] , [45] ). At the other extreme, coral recovery at Barren Island was very strong, and abundance of L. variegata remained much lower than other sites, even after coral mortality ( Fig. 2 ). However, abundance of L. variegata was still highest following coral mortality (18%), and declined as the coral recovered (although not significantly: P = 0.131 for Tukey's comparison of August 2006 and February 2007). Barren Island is further offshore and in deeper water than the other sites, and dead coral tissue was colonized predominantly by algal turfs more typical of offshore reefs. Detailed analyses of the species composition of the algal turfs in this locality (data not shown) revealed a very different species composition of turfing algae, mainly dominated by calcareous turfing species (e.g. Jania and Amphiroa ). Recent events in the Keppel Islands provide an exceptional, but important example of the doom and boom of highly resilient reefs, and thereby provide new insight into the potential variability in mechanisms of reef resilience. Most degraded reefs globally have either failed to recover from events such as coral bleaching and other human induced disturbances [3] , or have taken several years to decades to return to pre-disturbance condition [14] , [15] , [18] , [19] , [29] , [46] , [47] . In contrast, the Keppel Islands have shown rapid recovery of coral dominance, despite repeated coral bleaching events (1998, 2002, and 2006 [48] ), severe flood plumes (e.g. 1991, 2008 [44] ), and dense algal overgrowth. If they allow recovery of coral populations within one year, instead of ten, such exceptional processes may be disproportionately important to larger-scale reef resilience. Resilience of reef coral populations is typically considered in terms of removal of algal blooms by herbivores, combined with replenishment by coral larvae. Whilst these factors are no doubt vital for reef persistence [7] , [13] , [49] , both abundance of herbivorous fishes and coral recruitment were apparently limited on the reef slopes studied here during these events. There is considerable evidence that algal abundance on coral reefs is generally related to herbivory [6] , [7] , [50] – [54] , and herbivory can be important to interactions between L. variegata and corals on the GBR [25] . However, in this instance, removal of the seaweed L. variegata appears to have been largely due to inherent seasonal dieback, more than consumption by herbivores, although experimental studies would be required to be conclusive. Importantly, this dieback is apparently species specific ( [35] , GDP unpublished data), so that its ecological significance presumably depends on the nature of the seaweed bloom as a single species. In more typical multi-species seaweed blooms, it is unlikely that all species would have similar seasonality, and competitive effects on coral regrowth would probably be stronger. In this sense, given the apparent limited abundance of herbivores, the reduction in seaweed during our study may be a fortunate coincidence of monospecific bloom and seasonal dieback in that one species. Further, had the decline in L. variegata not coincided with rapid coral growth, it is likely that a range of other algae would have colonized, potentially stabilizing the phase shift. Thus, the seasonal decline was clearly important to the resilience of these reefs in these circumstances, but should not be seen as diminishing the general importance of herbivory to reef resilience. Our results stand in contrast with many previous studies, especially studies of coral and algal dynamics on Caribbean reefs in the early 80 s, where a combination of coral mortality and hurricane damage followed by mortality of sea urchins, caused massive algal blooms (including L. variegata ) that still continue today [14] , [15] , [55] . Although L. variegata was involved in both circumstances, there are several fundamental differences that probably contribute to the different outcomes. First, the Keppel Islands are dominated by rapidly growing, branching Acropora , apparently better suited to competing with a mat-like algal growth than the massive and plate-like corals that were dominant on Caribbean reefs [26] , [55] , [56] . Coral-algal interactions will depend considerably on the particular species involved. Second, the monospecific algal bloom in the Keppel Islands was exceptionally vulnerable; most macroalgal blooms are much more diverse, imbuing the algal-dominated state with greater resilience. Studies of Caribbean reefs typically note 5–10 genera of benthic macroalgae (e.g. [31] , [57] , [58] ); after long-term herbivore exclusion on the GBR, at least 10 algal genera were abundant in algal dominated plots [7] . Third, coral recovery may be strongly influenced by the nature of the disturbance regime. Reefs subject to acute disturbances, such as the rapid bleaching in the Keppel Islands, may often recover more effectively than those subject to chronic disturbances such as in the Caribbean [46] , [59] . Similarly, the spatial scale of disturbance in our study was much smaller than that in the Caribbean. Numerous other factors can contribute to the resilience or vulnerability of a reef (e.g. [3] , [60] ). In summary, unusually rapid coral recovery in the Keppel Islands apparently stemmed from synergistic effects of factors not previously recognized as important to resilience. These factors included robust tissue regeneration, high competitive ability of the corals and a seasonal dieback in the monospecific seaweed bloom, all against a backdrop of an effective marine protected area system and moderate water quality. Understanding the variability in mechanisms underlying resilience is critical for reef management under climate change. Settlement and recruitment of new corals requires years to decades to re-establish abundant corals, whereas recovery in the Keppel Islands took less than one year. Frequent, large-scale damage may mean that reefs able to rapidly recover abundant corals may serve as key refugia, or sources of larvae for reef recovery at broader scales. Diversity in processes may well be critical to the overall resilience and persistence of coral reef ecosystems globally."
} | 6,204 |
29468126 | PMC5779732 | pmc | 2,328 | {
"abstract": "Chemical and fuel production by photosynthetic cyanobacteria is a promising technology but to date has not reached competitive rates and titers. Genome-scale metabolic modeling can reveal limitations in cyanobacteria metabolism and guide genetic engineering strategies to increase chemical production. Here, we used constraint-based modeling and optimization algorithms on a genome-scale model of Synechocystis PCC6803 to find ways to improve productivity of fermentative, fatty-acid, and terpene-derived fuels. OptGene and MOMA were used to find heuristics for knockout strategies that could increase biofuel productivity. OptKnock was used to find a set of knockouts that led to coupling between biofuel and growth. Our results show that high productivity of fermentation or reversed beta-oxidation derived alcohols such as 1-butanol requires elimination of NADH sinks, while terpenes and fatty-acid based fuels require creating imbalances in intracellular ATP and NADPH production and consumption. The FBA-predicted productivities of these fuels are at least 10-fold higher than those reported so far in the literature. We also discuss the physiological and practical feasibility of implementing these knockouts. This work gives insight into how cyanobacteria could be engineered to reach competitive biofuel productivities.",
"conclusion": "5 Conclusions We used available genome-scale models and algorithms and found metabolic engineering strategies for creating growth-coupled cyanobacteria biofuel strains. The relative dearth of NADH-utilizing reactions in Synechocystis allowed for coupling of fermentative butanol with relatively few gene knockouts. Lowering the ATP/NADPH ratio in the cell is a general approach for coupling fatty-acid derived products such as alcohols and alkanes and terpenes such as limonene. Advances in genome engineering techniques will allow testing of these genetic interventions and integration of systems-biology data will refine the models to be more accurate.",
"introduction": "1 Introduction The engineering of microbes for the production of chemicals and fuels is a pillar in a future bio-based economy ( Choi et al., 2015 , Peralta-Yahya et al., 2012 ). Autotrophic hosts such as cyanobacteria are particularly attractive cell factories for large-volume, low-value products like biofuel, as handling of plant-based feedstock can negatively affect the cost and energy balances of heterotroph-based processes ( Jiang et al., 2014 ). Many cyanobacteria strains have been developed which produce small amounts of chemicals and biofuels directly from CO 2 ( Oliver and Atsumi, 2014 ). However, systems-level metabolic engineering is needed to achieve industrially-relevant chemical productivities in cyanobacteria ( Gudmundsson and Nogales, 2014 ). Cell factory design is based on a genome-scale metabolic model (GEM), where cellular metabolic reactions are tabulated and connected into a network topology. The GEM can be subjected to flux balance analysis (FBA), which uses external nutrient uptake rates and optimization principles to estimate steady-state intracellular and extracellular reaction fluxes, including cell growth rate ( O’Brien et al., 2015 ). A suite of algorithms have been developed which can use the GEM to calculate how cellular metabolism should be changed to achieve high productivities or yields of a given product ( Machado and Herrgård, 2015 ). These algorithms report in silico modifications that could be manifest experimentally as genetic knockouts, knockdowns, or knockups. One powerful algorithm is OptKnock, which seeks to maximize flux to product while simultaneously maximizing growth rate. The result is a list of knockouts, that when executed in silico , result in a strain where product synthesis occurs at maximum growth ( Burgard et al., 2003 ). This is beneficial as there is evidence that bacterial metabolism will evolve to maximize growth ( Fong and Palsson, 2004 ). Therefore, product-growth “coupled” strains would ensure high productivity over time. A first application of OptKnock was to predict and execute the reaction knockouts necessary to link lactate production to cell growth in E. coli ( Fong et al., 2005 ). In the OptGene algorithm, reaction knockouts are implemented randomly, creating a mutant population ( Patil et al., 2005 ). The reaction fluxes of each mutant are predicted using minimization of metabolic adjustment (MOMA), which assumes a minimal deviation from wild-type fluxes ( Segrè et al., 2002 ). Mutants having higher fluxes toward product (or some other fitness metric) are selected for subsequent mutation or crossover with other mutants. OptGene is computationally efficient because it searches for local optima; i.e. mutants are subjected to an evolutionary trajectory, and each mutant is compared only to others in the population. While global optima are not found, the mutants reveal heuristics about the most effective mutations to improve productivity. OptGene was recently used to identify knockouts in yeast that improved succinate titers 30-fold when the strategy was executed in vivo ( Otero et al., 2013 ). Synechocystis GEMs have been previously used to suggest engineering strategies for increasing production of ethanol, isobutanol, fumarate, and hydrogen ( Sengupta et al., 2013 ; Erdrich et al., 2014 , Nogales et al., 2013 ). However, strategies for product-growth coupling have been elusive in cyanobacteria, as photosynthesis metabolism is robust, with several electron “valves” such as cyclic and alternative electron flows. Here we use OptGene and OptKnock on stoichiometric Synechocystis GEMs to identify in silico gene knockouts that improve production of fermentation, fatty-acid, and terpene-derived biofuels. OptGene revealed heuristics for improving productivity, while OptKnock identified reaction knockouts that couple photoautotrophic growth and biofuel production. The underlying logic of these strategies is revealed.",
"discussion": "4 Discussion We present genetic engineering strategies to obtain cyanobacteria strains where biofuel productivity is increased or coupled to growth. An advantage of growth-coupled production strains is that adaptive evolution can be used to improve both growth and production rate, as the strain evolves to the FBA-predicted flux distribution. To visualize an evolutionary progression for the M1 and M3 strains, we used MOMA to estimate the flux distributions after these knockouts. The MOMA-predicted flux of M1 was quite favorable, resulting in high butanol productivity. However, MOMA predicted no growth for strains M3 and M4 ( Fig. 7 ). This is likely due to the large number of gene knockouts needed to couple growth to biofuel, which results in a massive perturbation from the reference flux. It is likely that not all knockouts are needed to improve biofuel productivity. The M1, M3, and M4 reaction knockouts could be used to limit the reactions “pool” considered by OptGene, which could then be used find the effective combinations. When we recomputed OptGene with only M1, M3, and M4 knockouts considered (Subset C), the BPCY of mutants for 1-butanol, 1-octanol, and limonene were generally higher than for when the larger subsets were considered ( Fig. 7 ). A combination of methods for predicting intracellular fluxes is useful since the true objective function may not be biomass formation, but rather ATP generation, a minimization of total intracellular flux, or a combination of these ( Bordbar et al., 2014 ). Cyanobacteria in particular pose a challenge to the biomass assumption since their flexible metabolism, which provides robustness necessary for survival in a fluctuating marine environment, can be considered suboptimal for growth. Recent works have explored cyanobacteria metabolism under diurnal cycles and assumed that the cellular objective function changes depending on external conditions ( Knoop et al., 2013 , Rügen et al., 2015 ; Saha et al., 2016 ). With this caveat in mind, the general themes of the two coupling strategies appear to be robust and are consistent with strategies found using other methods. The M1 knockout strategy applies to other alcohols produced via reverse beta-oxidation and our work shows that modulating ATP and NADPH production can be applied to a range of products with different energy requirements. Our strategies for fatty-acid derived 1-octanol (M3) and limonene (M4) also have several overlapping knockouts with a recent work that used Elementary Mode Analysis to find knockouts which coupled ethanol production to growth in cyanobacteria ( Erdrich et al., 2014 ). The EMA approach has the benefit of visualizing many possible flux distributions (elementary modes) within the allowed flux space, including those that are not maximal growth. A desired degree of product-biomass coupling can be defined and algorithms such as CASOP can be used to compute which reaction deletions are needed to remove all possible elementary modes outside of this cutoff ( Klamt and Mahadevan, 2015 ). The knockout strategies reported here were found to be somewhat model-dependent, as transfer to a second GEM did not result in coupling unless extra reactions were knocked out ( Supplemental Note 2 ). The discrepancy among GEMs is a reflection of the incomplete knowledge of cyanobacteria metabolism, which influences the number of needed knockouts. For example, strategies that rely on co-factor imbalance require knowledge of the cofactor usage of an enzyme ( King and Feist, 2013 ). Here the models tend to be conservative and if the co-factor preference of an enzyme is not known, then both are included. In the M1 reaction target list the NDH-1 reaction using NADH is a necessary knockout while the NDH-1 reaction using NADPH is not. The preferred co-factor for NDH-1 is presumed to be NADPH ( Ma et al., 2006 ), though there is some experimental evidence supporting NADH activity ( Ooyabu et al., 2008 ). The required knockout of GlcD2 glycolate oxidase (O 2 requiring) in M1 is another illustrative example. The model iJN678 contains two glycolate oxidation reactions. The glycolate dehydrogenase GlcD1 ( sll0404 ) catalyzes NAD+-dependent glycolate oxidation and does not have detectable O 2 -dependent oxidase activity ( Eisenhut et al., 2006 ). A second glycolate dehydrogenase/oxidase GlcD2 ( slr0806 ) was identified in Synechocystis, but its activity was not characterized ( Eisenhut et al., 2008 ), so it is not known if GlcD2 has oxygen-dependent glycolate oxidase activity. The reaction may have been included in iJN678 based on the prevalence of oxygen-dependent glycolate oxidases in plants. Therefore, the knockouts of NADH-dependent NDH-1 and GlcD2 may not be necessary in practice to achieve coupling in M1. Of course, it is possible that as new reactions are experimentally verified, they will add to the reaction knockout list. The flux capacity of reactions can also be constrained based on experimental evidence, such as enzyme kinetics and expression data ( Reed, 2012 ) and could also obviate some knockouts. Integration of transcriptomics and proteomics data will improve the accuracy of the cyanobacteria GEM and could reduce the number of required knockouts. For example, three type-2 NADH dehydrogenases are required knockouts in the M1 strain, since they can oxidize NADH at high fluxes in silico . However, RNA-Seq showed that these type-2 NADH dehydrogenases are each expressed at less than 10% of the NDH-1 ( Anfelt et al., 2013 ) and experimental evidence for their activity was not found ( Howitt et al., 1999 ). A similar case can be made against pyruvate: ferredoxin oxidoreductase, which is another target in M1. This enzyme is known to be inhibited by O 2 and is likely not active in photoautotrophic conditions ( McNeely et al., 2011 ). Overall, these cofactor and flux constraints could reduce the required gene knockouts to achieve the M1 mutant butanol to 3 (FPK slr0453 , GLUSx sll1502 , MDH sll0891 ). Even considering cofactor and flux constraints, the number of gene knockouts for the M3 (fatty alcohols) and M4 (limonene) strains is at least 6. A central question is whether these strains can be realized in practice. While there are several selection and counter-selection methods available for cyanobacteria ( Begemann et al., 2013 , Cheah et al., 2012 , Viola et al., 2014 ), sequential gene knockout and antibiotic cassette curing is time-consuming and to date no strain with more than 4 knockouts has been reported. The type-II CRISPR/Cas tool is a powerful way to realize gene knockouts and has been applied recently in cyanobacteria for single knockouts ( Wendt et al., 2016 ). A CRISPRi knockdown tool could repress four genes simultaneously at 50–90% in Synechocystis ( Yao et al., 2015 ). One limitation of CRISPRi is that it is not possible to knockdown just one gene in an operon. Therefore, more advanced genetic engineering techniques must be developed in cyanobacteria in order to perform systems-level engineering."
} | 3,247 |
27093046 | PMC5030697 | pmc | 2,329 | {
"abstract": "Arbuscular mycorrhizal fungi are asexual, obligately symbiotic fungi with unique morphology and genomic structure, which occupy a dual niche, that is, the soil and the host root. Consequently, the direct adoption of models for community assembly developed for other organism groups is not evident. In this paper we adapted modern coexistence and assembly theory to arbuscular mycorrhizal fungi. We review research on the elements of community assembly and coexistence of arbuscular mycorrhizal fungi, highlighting recent studies using molecular methods. By addressing several points from the individual to the community level where the application of modern community ecology terms runs into problems when arbuscular mycorrhizal fungi are concerned, we aim to account for these special circumstances from a mycocentric point of view. We suggest that hierarchical spatial structure of arbuscular mycorrhizal fungal communities should be explicitly taken into account in future studies. The conceptual framework we develop here for arbuscular mycorrhizal fungi is also adaptable for other host-associated microbial communities.",
"conclusion": "Conclusion: community ecology from the viewpoint of a microbial symbiont We presented a conceptual framework of community assembly and coexistence adapted to a microbial symbiont group with a unique combination of characteristics. The importance of factors influencing obligate symbionts differs from those affecting free-living organisms, or even facultative symbionts ( Linardi and Krasnov, 2013 ). The host–AM fungal relationship, similarly to parasites, exhibits a hierarchical spatial structure, which should be explicitly incorporated into future studies, to enable the study of the scale dependency of the relative importance of elements of community assembly. Adapting a symbiont-centered point of view in addition to considering how the host community is affected would help to fill the knowledge gaps of coexistence research, especially in the field of non-host interactions.",
"introduction": "Introduction: applying models of community assembly and contemporary coexistence theory to communities of arbuscular mycorrhizal fungi: knowledge gaps and difficulties How communities assemble and which species can coexist in the same locale has been a central question of ecology. The recent advancement of high-throughput molecular barcoding methods has enabled researchers to obtain information on the composition and structure of natural microbial communities more easily than ever. This is especially important for organisms that are difficult to culture, such as arbuscular mycorrhizal (AM) fungi, which are obligate symbionts of plants where they form multispecies symbiont communities in the same root. Given the growing number of molecular studies in the field of AM fungal community research, it is timely to review the progress on understanding the processes that underlie AM fungal community assembly, and highlight the knowledge gaps. The theoretical framework of community assembly and coexistence ( HilleRisLambers et al. , 2012 ) combines a classic filter model of community assembly with modern coexistence theory ( Chesson, 2000 ). The filter model describes a regional pool of species from which the members of local communities are selected by passing through environmental and biotic filters. Modern coexistence theory ( Chesson, 2000 ) addresses interactions on the local scale, which arise from niche differences and fitness similarities. The combined approach therefore acknowledges factors on a wide spatiotemporal scale. Regional and local processes are connected by the neutral process of dispersal. The filter model is further nuanced by taking into account feedbacks, when organisms are not only influenced by but also have an impact on their environmental and biotic filters ( Figure 1 ). AM fungi are a monophyletic group (phylum Glomeromycota) of asexual, obligately symbiotic fungi with a unique combination of traits regarding morphology, genomic structure and ecology. Within their coenocytic mycelia and spores, multiple, potentially genetically divergent nuclei coexist, making it difficult to delineate an individual even with molecular methods ( Table 1 ). Through intraradical and extraradical mycelia, they occupy a dual niche, the plant root and soil. Both the soil environment and plant root can be described according to a simple filter model as species filters preventing certain AM fungal species from entering the local community. However, the simple filter analogy ends when taking into account that host plants and soil do not remain unchanged during community assembly: AM fungi interact with their hosts through hormonal crosstalk and actively shape soil as ecosystem engineers (See section: ' Feedbacks: AM fungi as ecosystem engineers '). Taking into account these developmental, genetic and ecological angles, the direct adoption of models for community assembly developed for other organism groups is not evident. There are several points from the individual to the community level where the application of modern community ecology terms runs into problems when AM fungi are concerned ( Table 1 ). Especially in the area of coexistence, even for the definitions of such fundamental concepts as ‘fitness' further research and discussion are needed ( Table 2 ). Here we introduce the elements of a community assembly and coexistence model by highlighting recent research on AM fungal communities. As examples for each element, we included studies that used DNA-based methods (preferentially, high-throughput sequencing approaches) to investigate AM fungal communities."
} | 1,407 |
32165719 | PMC7067859 | pmc | 2,330 | {
"abstract": "The three-dimensional hierarchical morphology of surfaces greatly affects the wettability, absorption and microfabrication properties of their hybrid materials, however few scalable methods exist that controls simultaneously complex geometric shape and spatial scattered location and their physical properties tuned. Consequently, this report describes a synthetic strategy that enables the position of well-ordered biomorph nano-microstructures on hydrophobic surfaces to be precisely controlled. The hierarchical architecture can be accurately positioned on polydimethylsiloxane (PDMS) surfaces in an unprecedented level by leveraging a solid/liquid/gas triphase dynamic reaction diffusion system strategy. The effect of salt concentrations, pH, CO 2 levels, temperature and substrate patterning on this self-assembly process has been investigated, enabling protocols to be devised that enables the hydrophobic properties of the hierarchically assembled multiscale microstructures to be tuned as required. This combined top-down/bottom-up approach can be used to produce composites with outstanding hydrophobicity properties, affording superhydrophobic materials that are capable of retaining water droplets on their surfaces, even when the material is inverted by 180°, with a wide range of potential applications in oil/water separation technology and for selective cell recognition in biological systems.",
"conclusion": "Conclusion A triphasic interface-mediated strategy has been used to attach hierarchical silica biomorph microstructural arrays to the surfaces of vertical hydrophobic micropillars PDMS substrates. The factors that affect solid/liquid/gas interface coprecipitation of the local chemical field have been evaluated to identify what parameters are responsible for controlling the self-assembly reaction at the growth front, which has enabled conditions to be identified that enable the morphology, size and direction of the resultant self-assembled structures to be controlled. Triphasic interface growth behavior was found to occur via liquid-gas and solid-liquid growth processes, with overall growth rates controlled by the ambient CO 2 concentration at the site of reaction. The artificial hierarchical structure shows remarkable super-hydrophobic property and petal effect. The super-hydrophobic property and tunable adhesion are related to the number of arms and the gap distance between each branch. SiO 2 -SrCO 3 array exhibited ultrahigh durable superhydrophobic properties under UV irradiation, humidity and pH condition. It opens intriguing perspectives for designing artificial surfaces on the basis of complex hierarchical structure to form a super-hydrophobic surface. This one-step, mild fabrication method can be used to produce various different types of silica biomorph microstructure, thus providing a versatile method for fabricating materials with superhydrophobic surfaces that have potential applications for self-healing/self-regulation 31 , 32 , oil/water separation 33 , 34 and cell culture/diagnostics/recognition 35 , 36 .",
"introduction": "Introduction Self-assembly processes present in nature have inspired scientists to develop the artificial supramolecular self-assembly systems for the synthesis of complex and hierarchical structures that allow functional materials that exhibit useful physical/chemical/structural properties to be produced at scales ranging from nanometers to millimeters 1 . In particular, the well-ordered patterns and aligned hierarchical morphologies produced by synthetic inorganic materials (e.g., silica-based systems) can impart unique anisotropies, large surface areas and enhanced charge transport properties to materials for applications in opto-electronic devices 2 – 4 , catalysis 5 , mechanics 6 , 7 , superwettable materials 8 , 9 and molecular diagnostics 10 . However, the development of efficient self-assembly techniques that enable directed mesoscale complex construction using low-cost scalable protocols for the production of a wide range of materials is still urgently required. Self-assembly processes that occur during biomineralization at ambient temperature afford highly functional materials with hierarchical architectures and stunning beauty that have prompted the development of rationally designed silicon-based materials. For example, silica-carbonate biomorphs comprised of encapsulated crystalline carbonate and amorphous silica have been synthesized by diffusing CO 2 into alkaline solutions of sodium silicate (affording CO 3 2− and H + ions) that results in the coprecipitation of SrCO 3 and SiO 2 11 – 17 . This simple approach involves mixing of cheap ingredients (Na 2 SiO 3 and Sr/BaCl 2 ) to create silicon composites with unprecedented complex morphologies (vases, corals, flowers, worms and plates) that are generated through randomly dispersed growth processes 12 , 18 . The key factor in these bioinspired mineralization systems is the presence of front reaction-diffusion processes that create the dynamic environment required for production of these precisely sculpted arbitrary architectures 19 – 24 . These processes serve to create an artificial microenvironment that creates exactly the right chemical and thermodynamic conditions required for spatial control of the microstructures that are produced 25 . Initial reports into generating silica biomorph systems in dilute silica sols or silica gels involved immersion of a substrate into a solution containing reactive silicon precursors which resulted in the random generation of microstructures with no positional control 12 , 13 , 16 , 17 . More recently, a bubble mediated substrate water/air interface strategy has been developed to induce crystallization at solid-liquid-gas triphasic contact lines that can be used to control the spatial arrangement and morphology of the self-assembled surfaces on polydimethylsiloxane (PDMS) surfaces 26 . This approach relies on the hydrophobic micropillar-based PDMS surfaces containing air gaps that enable reagents in solution to be effectively transported to their edges where they can then react to form defined microstructures. In these processes, the micropillars of the PDMS act as a solid support that simultaneously interacts with air (containing CO 2 gas) and the solution (containing reactive silicon reagents) to produce a solid-liquid-gas triphasic reaction that deposits microstructures on the surface of the micropillars. The incorporation of soft lithography techniques into this hydrophobic micropillar fabrication process, can potentially control the location of the self-assembly growth process to afford selectively patterned materials. Three-dimensional microscopic computed tomography (3D-CT) technology has been used to determine the amount of air in contact with the liquid/solid phases, thus enabling the structure of the gas transfer networks present in these systems to be determined. These studies have shown that this gas transfer network is essential, because it sustains the integrity of the liquid phase and facilitates solution based gas diffusion-reaction processes that are required for self-assembly to occur at room temperature. Aebisher and co-workers 27 have reported other types of reactive solid-liquid droplet-based systems, employing silicon phthalocyanine sensitizing particles to study the role of hydrophobic photosensitizers that generate oxygen gas in droplets for the functionalization of PDMS micropillars. Furthermore, Wang 28 and Wu 29 have reported diffusion studies on H 2 and H 2 S at triphasic interfaces under nonequilibrium conditions that enabled gold and metal sulfide dish-like microstructural arrays to be grown on the top surfaces and edges of PDMS pillars. Therefore, the ability of well-organized hydrophobic micropillar substrates to act as structural templates to precisely locate gas-liquid-solid triphasic interfacial self-assembly reactions has been firmly established. In this study, we report how self-assembly processes generated through bottom up biomineralization and top-down lithography processes that occur at hydrophobic gas-liquid-solid interfaces can be used to generate multiscale patterned architectures on pillar templates. Optimization of reaction times, reagent concentrations, pillar geometries and temperatures of this triphasic self-assembly process has enabled reproducible tunable protocols to be developed that enable the scalable interfacial growth of uniform SrCO 3 –SiO 2 microstructural arrays on hydrophobic micropillar substrates. These protocols generate stable CO 2 concentrations at the reactive triphasic interface of these hydrophobic self-assembly processes, thus enabling the location and hierarchical structures of the microstructures formed on the pillars to be studied and controlled. This new triphasic interface approach represents a novel platform for investigating crystal growth mechanisms for the formation of carbonate-silica biomorphs and is likely to be of use for the production of silicon-based arrays of use in functional devices.",
"discussion": "Results and Discussion PDMS micropillar structure was fabricated using conventional photolithographic processes to produce patterned surfaces with typical dimensions of 100 μm width, 50 μm gap, and 100μm height (Fig. 1F ), with all substrates being dried thoroughly in a vacuum before being used in self-assembly experiments. The rough surfaces of the exposed PDMS surfaces exhibited hydrophobic properties with a water contact angle of 135 ± 2° (Fig. 1G–H ) and were found to contact SrCl 2 –Na 2 SiO 3 solution (via air pockets) in a CO 2 atmosphere at room temperature (Fig. 1E ). Fabrication of the SrCO 3 –SiO 2 architectural arrays was achieved by vertically immersing the PDMS substrate in a beaker containing an aqueous solution of SrCl 2 (19.1 mM) and Na 2 SiO 3 (8.2 mM) at pH of 11.8. The beaker was then fitted with a loose cover that allowed gas diffusion from the atmosphere to the surface of the pillars (Fig. 1A,D ) and then placed in an incubator whose atmosphere and temperature could be regulated. The incubator was then pressurized with CO 2 and left for several hours to allow the hierarchical nano-microstructures to grow on the surfaces of the pillars of the PDMS substrate (Fig. 1C ). Alcohol was then added to the solution to decrease its surface tension, which enabled the fragile PDMS chip microstructures to be isolated without damage, which was then freeze-dried under vacuum. The analysis revealed that triphasic reactions had resulted in regular SrCO 3 -SiO 2 hierarchical structures being deposited at the top and edge regions of the PDMS pillars, with no deposition having occurred on their bottom regions. Protruding branches were present inside the pillar gaps located at the liquid-gas interface, as observed previously when bulk solution air-water interface condition was employed. Analysis by scanning electron microscopy (SEM) revealed that the carbonate-silica biomorphs with microflower and micrograss morphologies were present on the surface of the PDMS pillars (see Fig. 1C ). Energy dispersive spectrometric (EDS) analysis confirmed that the microstructures were comprised of silica (Fig. S8 ). Magnified TEM images confirmed the presence of protruding arms with relatively large surface areas of 20 nm width and 200-400 nm length (insert in Fig. 1I ). The amorphous silica deposits affording a single bright diffraction spot, whilst X-ray diffraction (XRD) patterns clearly revealed CP2 diffraction peaks expected for the body-centered cubic structure of SrCO 3 (JCPDS no. 14941-1-40-3) (Fig. S2B insert). Therefore, these analytical results confirm that the interface-induced microflower-shaped constructs produced in these self-assembly processes are composed of polycrystalline SrCO 3 and amorphous silica building blocks. Figure 1 Schematic and SEM/TEM images of fabricated silica biomorphs: ( A ) Micropillars vertically submerged in solution with an initial pH of 11.8. ( B , D , E ) Schematic illustration of hierarchical architectural formation occurring at the interface between the micropillar arrays and the air pocket. ( C , F ) SEM images of SrCO 3 -SiO 2 structures formed at the top of the micropillar edges. ( G ) Photographic image of a 1.5μm water droplet on a surface with a contact angle (CA) of 135 ± 2°. ( H ) 3D-CT view of a 400μm diameter droplet on a hydrophobic pillar-structured substrate. ( I , J ) TEM/HRTEM image of the SrCO 3 nanostructure. Nucleation and crystallization processes at triphasic interfaces controls the location of self-assembly in these systems, with air-water interfaces unable to support crystal nucleus growth. Nucleation processes generally occur at surface edges with small radii of curvature, which means the self-assembly process only occur at the triphasic solid-liquid-gas interface. The rate of microstructural growth at the top of the pillar (bulk liquid) is slower than that at the edges because the concentration of CO 2 required for the reaction is less at this more exposed interface (Fig. 2C,G,K,O ). The air pockets in the gaps in the pillars provides a link between the triphasic interface and the CO 2 atmosphere, with the bulk solution containing the reactive silicon precursors contacting each vertical PDMS pillar that is exposed to a differential CO 2 gradient (Fig. 1D ). Therefore, this triphasic architecture clearly has the potential for the production of multilayer surfaces with different shapes and hydrophobicities for the production of tunable silicon based microarrays for applications in chip design (Fig. 2 ). Figure 2 SEM images of the growth morphology of different micropillar patterned silica biomorphs on (i) cylindrical micropillars ( A , E , I , M ); (ii) elliptical micropillars ( B , F , J , N ); (iii) rounded nose micropillars ( C , G , K , O ), (iv) square micropillars in solution ( D , H , L , P ). Biomorphs formed using a solution of BaCl 2 (19.1 mM) and Na 2 SiO 3 (8.2 mM) at pH 11.8. Attempts to carry out triphasic interface coprecipitation reactions using patterned pillars with 100 μm gaps proved unsuccessful, because the larger liquid surface area between the two pillars generated a large amount of sediment which resulted in the architecture being disrupted and liquid flowing directly into the gaps to displace the air voids. Similarly, when the micropillars were inclined at an angle (other than 90°) in the solution, then the uncontrolled aggregation of SiO 2 and SrCO 3 resulted in the liquid entering the gaps to disrupt the self-assembly process. Consequently, all self-assembly reaction was carried out using vertically aligned PDMS templates to avoid the air-liquid interface from being disrupted and ensure free CO 2 exchange between the atmosphere and the triphasic reaction sites (Fig. 2 and Movie S1 S2 ). The structural growth behaviors of root, stem and petal segments of interface-mediated silica biomorphs were investigated over time (Fig. S2A ). Root segment growth proceeded quite slowly, indicating that a relatively low concentration of SrCO 3 and SiO 2 were present at the reactive triphasic sites in these systems. Initial nucleation events occur at the pillar top edges that provide high surface energies and tensions to establish effective triphasic interfaces for functionalization of the micropillar surface. After an hour, self-assembles stems were observed with a statistical average length of 5 μm, whose formation occurs because of continuous diffusion of CO 2 gas through the ambient-connected gas network, which facilitates the solid crystallization process. Over the next five hours, the width of the stem remains constant at 2 μm and its length rising to 40–45 μm as the number of crystals present reached saturation level. Between seven and nine hours, the rate of the self-assembly reaction process decelerated significantly, which is probably due to the pH value at the interface being reduced and the surface wetting properties of the system change. A concomitant reduction in ion concentration levels results in a change in growth behavior to a radial mode that produces agglomerates with plate vase like shapes. To demonstrate that lower ion concentration was responsible for this change in structural morphology, we repeated pillar functionalization experiments using a solution containing half the concentration of the reactive silicon precursors, which resulted in the formation of vase shape assemblies. Notably, once the tops of pillars were uniformly coated by the hierarchical architecture of a definite size, then the wetting properties of the exposed surfaces became hydrophilic, with the pillar gaps still being exposed to CO 2 gas. Interestingly, PDMS pillars whose surfaces were pretreated with oxygen plasma to produce hydrophilic surfaces produced microstructural assemblies wit rambling structures, similar to those produced in standard solution based self-assembly systems. The effect of different temperatures on the growth of silica biomorph architectures was also investigated, with increasing temperatures resulting in increased reaction rates due to air bubbles destroying the gas-liquid interface, which results in the liquid fully permeating into the pillar gaps to afford less morphologically defined microstructures. Conversely, when the temperature declines, the growth rate decreases, with no surface based self-assembly processes found to occur at 0 °C. As a result, it is recommended that a temperature range of between 25–45 °C (Fig. S2B ) is used in these self-assembly reactions, with temperature variation a potentially useful variable that can be used to modify the size and morphology of the resultant structures. The chemical reaction (see Fig. 3V ) underpinning this self-assembly reaction takes place under alkaline conditions (pH 9–13) and relies on changes in speciation of each reactive component being induced by precipitation of another component. As barium carbonate nucleates and crystallizes at the front of a developing aggregate, then the localized decrease in concentration carbonate ion induces dissociation of nearby bicarbonate ions to re-establish the perturbed HCO 3 − /CO 3 2− speciation equilibrium. This results in the concomitant release of protons that react near crystallites to dissolve silicate species, such as Si(OH) 3 O − . These localized protonation events raise the concentration of silicic acid, resulting in supersaturation of amorphous silica which precipitates out and polymerizes around the carbonate crystallites, thus preventing further growth at that site. Continuous oligomerization reactions remove protons from the precipitation site, which in turn affects local carbonate speciation by increasing the amount of CO 3 2− ions present 11 – 16 , 19 , 21 . This causes an increase in supersaturation of barium carbonate, resulting in the nucleation sequence being repeated and further growth occurring. After an initial stage involving silica-induced fractal branching of a carbonate seed crystal, this dynamic interplay results in emerging nanocrystalline assemblies initially growing as cardioid shaped quasi-2D laminar sheets that are a few microns thick. These sheets can grow along the surfaces of vessel walls, or assemble on top of each other, curling, bending and twisting in the process to produce delicate helical or worm-like morphologies. Based on this knowledge, we investigated whether the morphologies of the silica biomorphs could be controlled by modifying salt concentration and pH levels used in these self-assembly reactions. Inspired by phase diagram theory, there were four factors affecting the chemical reaction (see Fig. 3W ); with a green triangle indicating that the concentration of CO 2 and pH was unchanged for variable amounts of NaSiO 3 and SrCl 2 ; and a pink triangle with a black line indicating that the NaSiO 3 /SrCl 2 ratio and CO 2 concentration levels were unchanged for pH various from X to Y (Table S1 , S2 ). These optimized screening reactions identified conditions could be used to produce different types of microstructures on the surfaces of the PDMS pillars like flower, branch, vessel, coral, worm, like and bowl microstructures (Fig. 3A–M ) (Table S3 ). Flower-like microstructures (Fig. 3H ) with branch-like architectures were generated in five of the protocols, with the extent of their proliferation through a genetic growth mode controlled by varying growth times. Three types of nanostructure were found to influence the morphology of these systems: columnar 100–200 nm SrCO 3 nanocrystals; 20–50 nm SrCO 3 -SiO 2 nanowires; and 5–10 nm SrCO 3 -SiO 2 nanoslices (Fig. 3N–U ). When the amount of glassy SiO 2 increased, the amount of nanocrystal SrCO 3 present for growth becomes limited, resulting in different crystalline forms of SrCO 3 being produced that are then encapsulated in nanoglass SiO 2 . The various branch-like microstructures present were mainly produced on big columnar SrCO 3 nanocrystals, whilst the leaflike and wormlike shapes ware mostly present on SrCO 3 -SiO 2 nanoslices, with the rest of the shapes containing all three types of nanostructure. These experiment results fit well with Noorduin’s hypothesis 12 of structural evolution in this type of self-assembly systems occurring via three distinct growth regimes. Figure 3 Regulated synthesis of hierarchical complex structures at different solution concentrations and pH. The false-colored SEMs ( A – M ) were produced using solution concentrations and pH values described in Table S3. The TEMs ( N – U ) represent three basic nanostrutures that appear as coral-like/branch-like, flowerlike and vessellike/wormlike hierarchical assembled microstructures, respectively. These structues include 100–200 nm columnar SrCO 3 nanocrystals, 20–50 nm width SrCO 3 -SiO 2 nanowires and 5–10 nm width SrCO 3 -SiO 2 nanoslices. The schematic chemical reaction ( V ) shows the reactivity relationships ( W ) between the four reactive components (Na 2 SiO 3 , SrCl 2 , CO 2 and H + ). The mechanism of formation of the microcrystalline deposits produced in this system is dependent on the microenvironment (pH, ion concentration, temperature) at the front the oscillatory coprecipitation reaction. Three types of micro-environment interfaces where CO 2 could potentially be captured may be considered: (1) at a gas-liquid biphasic interface (Fig. S6A ); (2) at a solid-liquid biphasic interface (Fig. S6B ); and (3) at a gas-liquid-solid triphasic interface (Fig. S6C,D ). CO 2 directly contacted with the solution at a gas-liquid interface can potentially generate an explosive dendritic morphology, with surface tension effects having little effect on their morphology. By contrast, when CO 2 is consumed more gradually at increased depths in the pillar gaps, then the rough vertical sides of the pillar substrate can induce a variable CO 2 gradient distribution that negatively affect the uniformity of the self-assembly process. Therefore, accurate adjustment of parameters to account for the effects of this substrate induced CO 2 concentration gradient are difficult to achieve, and so the presence of the air pocket in the voids in the pillars is required to closely regulate the gas concentration available to the triphasic reaction centers that produce these silica biomorphs system. The surface of the template substrate in conventional processes does not influence on the nucleation and growth of crystal formation. However our methodology generates silica biomorphs at a triphasic interface that enables SrCO 3 -SiO 2 nanostructures with different morphologies to be generated depending on the reaction conditions employed (Fig. 4 ). Interestingly, the surfaces of most of the complex hierarchical structures produced in this study were shown to exhibit superhydrophobicity (i.e. lotus effect) with a maximum water contact angle of 155.7° being achieved (Fig. 5F–I,P–T ). The superhydrophobicity of a solid’s surface is dependent on its hierarchical roughness which in turn determines its surface energy, with the wettability properties of the silicon assemblies grown on the tops of all the different patterned PDMS pillars investigated shown to display superhydrophobic properties (Fig. 5P–Y ). The capillary-force-based hierarchical structures of these superhydrophobic silicon coated materials enabled them to retain water droplets when the functionalized material was rotated through different orientations (flat, tilted at 25°, 75° and 90° (Fig. 5K–N , Movie S3 ). Increasing the size and density of the silica biomorph (Fig. 5A–E ) resulted in the superhydrophobicity of the resultant synthetic hierarchical structure being amplified. Indeed, the superhydrophobic surfaces of hierarchical SrCO 3 -SiO 2 materials were even capable of retaining water droplets when they were inverted through 180° (Fig. 5K–O , Movie S3 ). As reported, there are four different kinds of contact regime for the eggbeater structure in contact with water W-W, CB-CB, W M -CB m , CB M -W m (W represent Wenzel; CB represents Cassie–Baxter; subscript “M” is for “Macro”, “m” is for “micro”) 30 . As for the mixed CB M -Wm state (lower/macro-Cassie–Baxter and upper/micro-Wenzel) with the water droplet pinned on the top of arms and an air layer is trapped by nano-microstructure arms (Fig. 5F–I,P–T ). It is noticed that superhydrophobicity is increased with the increment of higher nano-microstructure intensity or the surface roughness, however the roughness too large would make the droplet can’t stand on their surface (Movie S4 ). Together with the optical microscope images, we believe the presence of a CB M -W m regime in the middle and CB-CB regime in the edges when the SiO 2 -SrCO 3 surface is in contact with water droplet. This CB regime in the edges provides enough liquid air contact area to support the super-hydrophobic effect, and the micro-Wenzel (Wm) state in the middle makes the high CA hysteresis and petal effect (Fig. 5K–N , Movie S3 ). Figure 4 Diversity of structures grown on different geometric micropillars (triphasic interface) under different solution conditions. Images of cylindrical micropillars ( A , F , K , L , P , Q , U , V ), triangular micropillars ( B , G ), square micropillars ( C , H , M , R , W ), elliptical micropillars ( D , I , N , S , X ) and rounded nose micropillars ( E , J , O , T , Y ) microstructures that were obtained through adjusting solution composition, temperature, growth time and pillar shape. Figure 5 Superhydrophobicity of hierarchical silica biomorph arrays. (i) Time-dependent formation of microflowers affects surface wettability by increasing branch lengths and space distribution densities ( A–E ) to afford hydrophobic and superhydrophobic surfaces ( F–J ). (ii) Images of water droplets adhered to a SrCO 3 -SiO 2 hierarchical structure rotated through 0°−180° ( K–O ). (iii) Superhydrophobicities of biomorph arrays present on different pillar types ( P–T ) characterized by their different microflower shapes ( U–Y ). The long-term stability of the superhydrophobicity of the obtained surface after humidity, outdoor exposure and UV for various time intervals was also investigated. The relative air humidity was controlled by using the saturated salt solution method according to EN ISO 483. The samples were put into a closed container with different kinds of saturated salt solutions by which the relative humidity in the container could be kept constant. Hydrophobicity transfer characteristics of microstructure arrays are shown in Fig. S7A . It can be seen that the hydrophobicity transfer speed and the asymptotic value of the contact angle are both influenced by the relative humidity. The transfer speed and the asymptotic value of the contact angle decrease significantly, especially when the relative humidity is high. Figure S7B shows the CAs of microstructure array as a function of UV irradiation time. It was found that the CA declined a little from 168.3° to 163.2° after the first 12 h UV exposure, however showed no great change for further irradiation. Also it should be noted that the sample surface has still remained superhydrophobic with a CA of 158.9° before UV exposure (Fig. S7B ) after 384 h, i.e., 16 days, of UV exposure. This indicates that the superhydrophobicity of the microstructure array possesses excellent resistance to UV light. Figure S7C shows the relationship between the water CA and SA of asprepared superhydrophobic surface and the exposure time. The CA on the surface only slightly decreased to 158.5 ± 0.9° from 163.3 ± 0.7° after exposure more than 24 days, as well as the SA is less than 10°, exhibiting the long-term stability of the resulting surface. This demonstrates that the long-term stability of the obtained superhydrophobic surface is of great importance to the practical application of superhydrophobic surface. In order to analyze the chemical stability, the influence of pH of an aqueous solution and saline solution on the superhydrophobicity was studied for the as-prepared surface. A dry sample was placed inside a closed experimental cell, with water vapors saturated with respect to the test solutions. Figure S7D shows the evolution of the CA formed by drops of aqueous solutions of H 2 SO 4 with pH = 1, NaOH with pH = 13 and neutral 3.5 wt.% NaCl solution for the superhydrophobic surface. From Fig. S7D , we can find that the CA is higher than 150° after 24 h of contact with H 2 SO 4 and NaCl solutions, suggesting a good stability. However, the CA is lower than 150° after 3 h of contact with NaOH solution and finally decreased to 124° after 24 h of contact with NaOH solution. This is due to the occurrence of the chemical reaction between stearic acid and NaOH. It was investigated the pH and surface wetting properties changing before and after the reaction. Through the rate of the self-assembly reaction process carry on, the pH value decelerated and the CA increased first then decreased in Fig. S7E, F . In this case, pH change in solution reaction could affect the microstructure and hydrophobicity. Therefore, we propose that our triphasic approach is a potentially versatile approach to directly position hierarchical combinatorial architectures on to the surfaces of pillar shaped PDMS substrates, thus producing self-assembled superhydrophobic materials without the need for multiple operational processes and complicated instrumentation."
} | 7,635 |
35845416 | PMC9284946 | pmc | 2,331 | {
"abstract": "Cyanobacteria are a promising photosynthetic chassis to produce biofuels, biochemicals, and pharmaceuticals at the expense of CO 2 and light energy. Glycogen accumulation represents a universal carbon sink mechanism among cyanobacteria, storing excess carbon and energy from photosynthesis and may compete with product synthesis. Therefore, the glycogen synthesis pathway is often targeted to increase cyanobacterial production of desired carbon-based products. However, these manipulations caused severe physiological and metabolic impairments and often failed to optimize the overall performance of photosynthetic production. Here, in this work, we explored to mobilize the glycogen storage by strengthening glycogen degradation activities. In Synechococcus elongatus PCC 7942, we manipulated the abundances of glycogen phosphorylase (GlgP) with a theophylline dose-responsive riboswitch approach, which holds control over the cyanobacterial glycogen degradation process and successfully regulated the glycogen contents in the recombinant strain. Taking sucrose synthesis as a model, we explored the effects of enhanced glycogen degradation on sucrose production and glycogen storage. It is confirmed that under non-hypersaline conditions, the overexpressed glgP facilitated the effective mobilization of glycogen storage and resulted in increased secretory sucrose production. The findings in this work provided fresh insights into the area of cyanobacteria glycogen metabolism engineering and would inspire the development of novel metabolic engineering approaches for efficient photosynthetic biosynthesis.",
"introduction": "Introduction Cyanobacteria are important prokaryotic microorganisms performing oxygenic photosynthesis and playing essential roles in global carbon and nitrogen cycles on Earth ( Waterbury et al., 1979 ; Flombaum et al., 2013 ; Rousseaux and Gregg, 2014 ). Being abundant and widely distributed among diverse types of habitats, cyanobacteria provide 20–30% of the primary organic carbon in the biosphere, with their efficient oxygenic photosynthesis systems converting solar energy and carbon dioxide into organic compounds ( Flombaum et al., 2013 ; Rousseaux and Gregg, 2014 ). In cyanobacteria cells, the surplus carbon flow beyond the requirements of cellular growth and maintenance would be stored as a carbon sink, supporting materials and energy for the cells to survive in dark and stressful conditions ( Ball and Morell, 2003 ; Nakamura et al., 2005 ; Damrow et al., 2016 ). Glycogen is the most essential carbon sink compound in cyanobacteria, accounting for up to 50% of the total cellular biomass in specific species or environments ( Aikawa et al., 2014 ; Song et al., 2016 ). As promising photosynthetic microbial platforms for biotechnological and industrial applications ( Angermayr et al., 2009 ; Lu, 2010 ; Oliver and Atsumi, 2014 ; Zhou et al., 2016 ; Luan and Lu, 2018 ), diverse strategies and tools have been developed to manipulate carbon flow in cyanobacteria, and glycogen metabolism has been generally recognized as a promising target ( Carrieri et al., 2012 ; Melis, 2012 ; Zhou et al., 2016 ; Luan et al., 2019 ). Metabolic pathways for glycogen synthesis and degradation have been clearly deciphered in cyanobacteria ( Figure 1A ) and inhibition of the key enzymes for glycogen synthesis, and GlgC (ADP-glucose pyrophosphorylase, catalyzing ADP-glucose formation with glucose-1-phosphate) and GlgA (glycogen synthase, incorporating glucose monomers into the growing 1-4 α-linked glucose polymer) have successfully facilitated effective regulation of glycogen storages ( Carrieri et al., 2012 ; Melis, 2012 ; Zhou et al., 2016 ) and intracellular carbon distribution ( Xu et al., 2013 ; Hendry et al., 2017 ). However, impaired glycogen synthesis usually causes severe disturbance to cell physiology, including reduced photosynthesis, growth, respiration, and robustness toward environmental stresses ( Miao et al., 2003 ; Suzuki et al., 2010 ; Grundel et al., 2012 ; Guerra et al., 2013 ; Hickman et al., 2013 ). In many cases, blocking glycogen synthesis decreased rather than increasing the productivity of heterologous pathways in engineered cyanobacterial strains, which might be resulted from the combined effect of physiological damage and metabolic rebalance ( Davies et al., 2014 ; Jacobsen and Frigaard, 2014 ; Li et al., 2014 ; van der Woude et al., 2014 ; Work et al., 2015 ). FIGURE 1 Construction of a theophylline-responsive glgP expression system in Synechococcus elongatus PCC 7942. (A) Glycogen metabolism pathways in cyanobacteria. GlgP, glycogen phosphorylase; GlgA, glycogen synthase; GlgC, glucose-1-phosphate adenylyltransferase or ADP-glucose pyrophosphorylase; CBB cycle, Calvin–Benson–Bassham cycle. (B) Strategy for constructing the theophylline-responsive riboswitch control system for glgP expression in PCC 7942. A kanamycin resistance gene (KmR) and theophylline dose-regulated expression cassette ( Ptrc-ENYC4 ) were inserted between the native promoter of glgP ( PglgP ) and the glgP CDS, and the obtained mutant was termed as PCC 7942 XC8. The red arrows represented the primer pairs used for checking the genotypes of PCC 7942 wildtype (PCC7942-WT) and XC8 mutant (PCC7942-XC8) strains. (C) Identification of the genotypes of PCC 7942 wildtype (WT) and XC8 mutant by PCR with the primers (shown as red arrows in B). Since elimination of the glycogen synthesis flux failed to optimize the photosynthetic biosynthesis performances of the cyanobacterial cell factories, strengthening the glycogen degradation pathways to mobilize the carbon sink into central metabolism could undoubtedly be an alternative strategy worth trying ( Luan et al., 2019 ). However, approaches and effects to manipulate glycogen degradation metabolism in cyanobacteria have been relatively less reported. As shown in Figure 1A , glycogen phosphorylase (encoded by glgP ), catalyzing the glycogen digestion into glucose-1-phosphate (G-1-P), is mainly responsible for glycogen degradation in cyanobacteria ( Fu and Xu, 2006 ). In 2012, Ducat et al. overexpressed glgP in an engineered strain of Synechococcus elongatus PCC 7942 for secretory sucrose production and reported that the additional cassette for glgP expression caused a 10% reduction in sucrose productivity ( Ducat et al., 2012 ). In general, the possibilities to engineer and accelerate glycogen degradation and the subsequent effects on cellular physiology and metabolism in cyanobacteria are yet to be explored. In the present study, we aimed to manipulate glycogen metabolism by regulating the expression of glgP and the activities of glycogen degradation. Adopting a theophylline dose-responsive riboswitch, GlgP abundances, glycogen phosphorylase activities, and glycogen storage in a model cyanobacterium Synechococcus elongatus PCC 7942 (hereafter PCC 7942 for short) were successfully regulated, and the relative effects on cellular physiology and metabolism were explored. Sucrose is another important type of carbohydrate in PCC 7942 to be accumulated as osmoprotectants in response to extracellular hypersaline stress ( Hagemann, 2011 ). In the past years, interactions between sucrose synthesis and glycogen metabolism have been explored in different cyanobacteria species ( Suzuki et al., 2010 ; Guerra et al., 2013 ; Xu et al., 2013 ), and several glycogen metabolism engineering strategies have been adopted to improve the production of sucrose and other metabolic products in engineered cyanobacterial cell factories ( Ducat et al., 2012 ; Qiao et al., 2018 ; Luan et al., 2019 ). Thus, sucrose synthesis was also taken as a model to explore whether the strategy of accelerating glycogen degradation could enhance the performance of photosynthetic biosynthesis. Our findings provided fresh insights into the area of cyanobacteria glycogen metabolism engineering and would inspire the development of novel metabolic engineering approaches for efficient photosynthetic biosynthesis.",
"discussion": "Results and Discussion Construction of a Theophylline-Responsive glgP Expression System in PCC 7942 As shown in Figure 1A , glycogen phosphorylase (GlgP) plays an essential role in catalyzing the glycogen degradation process by breaking the α-1,4-glycosidic bond on the glycogen chain and removing glucose monomers in the form of glucose-1-phosphate (G-1-P). It has been confirmed that the knockout of two glgP genes in Synechocystis sp. PCC 6803 would increase the glycogen contents by about 7% with continuous illumination and 250%–450% in day–night cycles ( Shimakawa et al., 2014 ). However, so far, there have not been reports about quantitatively evaluating the effects of enhanced glgP expression on glycogen storage in cyanobacteria. We designed to regulate the expression of glgP and glycogen degradation activities adopting a theophylline-dose responsive riboswitch system ENYC4 in PCC 7942. Previously the Ptrc-ENYC4 system ( Ptrc promoter combined with the ENYC4 riboswitch sequence) had been successfully utilized to regulate the expression of glgC and facilitated dynamic regulation of GlgC abundances and glycogen contents ( Qiao et al., 2018 ; Chi et al., 2019 ). In this work, we adopted a similar strategy by replacing the native glgP promoter sequence with a Ptrc-ENYC4 sequence ( Figures 1B,C ). As designed, the expression of glgP on the chromosome of PCC 7942 would be regulated by the dose of theophylline. To evaluate the effects of the theophylline-responsive riboswitch on regulating glgP expression, concentrations of theophylline (0,100, and 1,000 μM) were supplemented into the culture broth of XC8, and the abundances of GlgP were evaluated by immunoblotting. As shown in Figure 2A ; Supplementary Figure S1 , supplementation of theophylline with all of the three concentrations into the culture medium caused minor influence on growths of the XC8 strain, showing similar growth patterns with the wildtype control. During the process, the GlgP abundances were significantly regulated by the addition of theophylline ( Figure 2B ). Samples were collected from three time points of the cultivation process, and Western blot with GlgP-specific antibodies was performed. When no theophylline was added, the GlgP abundances were significantly reduced to an undetectable level compared with those of the control. The addition of theophylline (100 and 1,000 μM) as an inducer would elevate the expressions of GlgP in XC8. In the medium containing 100 μM theophylline, GlgP abundances would significantly exceed those of the wildtype, and increasing the theophylline concentration to 1,000 μM would further improve the GlgP abundances. The results indicated that in PCC 7942 recombinant strain XC8 carrying the Ptrc-ENYC4-glgP cassette, the abundances of GlgP could be effectively regulated with theophylline dose added in the culture broth. In addition, the change in GlgP abundances also brought in consistent changes in glycogen phosphorylase activities in the XC8 strain with increasing theophylline inductions ( Figure 2C ). With 100 μM theophylline induction, glycogen phosphorylase activities were increased to nearly 4-fold higher than those of the wildtype control. FIGURE 2 Theophylline-regulated glgP expression and glycogen storage in PCC 7942. (A) Cell growths, (B) GlgP abundances, (C) glycogen phosphorylase activities, and (D) glycogen contents of XC8 strain induced with different concentrations of theophylline during the 10 days of cultivation. To confirm the effects of theophylline induction on glgP expression regulation, different concentrations of theophylline (0 μM, T0; 100 μM, T100; 1,000 μM, T1000) were supplemented into culture broth of XC8 strain in flasks. The cells would be sampled on Day 3, Day 6, and Day 9 for GlgP Western blot assays, glycogen phosphorylase activity calculations, and glycogen content determination. Glycogen Storage is Negatively Correlated With the Theophylline-Regulated GlgP Abundances in PCC 7942 It has been confirmed that the GlgP abundances and activities could be artificially regulated through theophylline dose in the XC8 strain; thus, we further determined the influence of artificially regulated glycogen phosphorylase activities on glycogen storage in Synechococcus cells. Intracellular glycogen contents of XC8 and wildtype control were calculated on Day 3, Day 6, and Day 9 of the flask cultivation process ( Figure 2D ). As expected, intracellular glycogen contents of the recombinant strain showed a significantly negative relationship with the theophylline dose and GlgP abundances. Glycogen contents in the wildtype PCC 7942 cells were maintained in the range between 10 mg/gDCW and 30 mg/gDCW. While for the XC8 recombinant strain, when no theophylline was added (meaning that the translation of glgP transcripts was still inhibited), the intracellular contents were increased by 2–3 folds compared with those of the wildtype control. When theophylline was supplemented in the culture broth, glycogen storage in XC8 cells would be correspondingly reduced. However, it is also noteworthy that although the GlgP abundances and activities could be increased by theophylline supplementation to levels much higher than those of the wildtype control, the glycogen contents would still be maintained on a normal physiological level similar to the wildtype control. With relatively weak illumination (50 μmol photons/m 2 /s) and limited carbon supply for the cells cultivated in flasks, the glycogen synthesis and storage in PCC 7942 cells were maintained at a relatively low level; thus, the effects of enhanced glgP expression and glycogen degradation activities might not be significant enough. We supposed that under conditions facilitating rapid glycogen accumulation, the effects of enhanced glycogen degradation activities might be more obvious; thus, we further evaluated the performance of the XC8 cells in column photobioreactors with bubbled 3% carbon-air gas. The illumination strengths were also elevated to 150 μmol photons/m 2 /s. As shown in Figure 3A , with enhanced illumination and carbon supply, both the wildtype and the XC8 strains grow significantly faster, with or without theophylline inductions, while the theophylline induction significantly reduced the final cell densities of XC8. During the cultivation process, intracellular glycogen contents of the wildtype were gradually increased, reaching from about 40 mg/gDCW on Day 3 to about 120 mg/gDCW on Day 6 and then were maintained on a stable level in the stationary phase; the addition of theophylline caused no influence on glycogen contents as well as cell growths ( Figure 3B ). As for XC8, when no theophylline was added, the glycogen contents would be maintained on a much higher level than those of the wildtype control, although the difference was declining with the prolonged cultivation process (about 5-fold on Day 3 and 1.75-fold on Day 6). When 500 μM theophylline was added, the glycogen contents would be reduced by about 70% to 60 mg/gDCW after 6 days of induction and cultivation, which is about 50% of the wildtype level on Day 6. The growth of the XC8 was reduced by the theophylline-induced glgP overexpression ( Figure 3A ). The final cell density was about 30% lower than that of the controls, indicating that the elevated glycogen turn-over rates might cause a non-significant cycle between glucose-1-phosphate and glycogen, leading to waste or inefficient utilization of the photosynthesis-derived energy and material flow. FIGURE 3 Effects of theophylline-regulated glgP expression on PCC 7942 cell growths and glycogen contents in column photobioreactors. (A) Cell growth and (B) glycogen contents of the wildtype and XC8 strains when cultivated in column bioreactors. The cultivation would be bubbled with CO 2 (3% in air) and supplied with elevated illuminations (150 μmol photons/m 2 /s). 0 μM (T0) and 500 μM (T500) theophylline were added to the culture broth. Enhanced glgP Expression Failed to Improve Intracellular Sucrose Accumulation of PCC 7942 Facing Salt Stress As the most important carbon sink mechanism in cyanobacteria, glycogen metabolism regulates and buffers the intracellular carbon flow distribution derived from photosynthesis. When treated with salt stress, carbon distribution in cyanobacteria cells would be rewired and the osmolyte synthesis would take up a significant portion of intracellular carbon, indicating potential interactions and competition with the glycogen metabolism network. Sucrose is the primary osmolyte synthesized in PCC 7942 to resist extracellular hypersaline conditions, and previously it has been reported that the blocking of the glycogen synthesis pathway would reduce the capacity to accumulate sucrose in PCC 7942 under hypersaline conditions and inhibit the salt tolerances ( Miao et al., 2003 ; Suzuki et al., 2010 ). To explore the effects of enhanced glgP expression on sucrose accumulation of PCC 7942 facing salt stress, we imposed salt stress on PCC 7942 wildtype and XC8 strains supplemented with concentrations of theophylline. As shown in Figure 4A , when cultivating in hypersaline conditions, growths of XC8 were not influenced by the addition of theophylline to regulate GlgP abundances and glycogen storage. Although glgP expression in XC8 was significantly inhibited when no theophylline was added and the glycogen would be accumulated to a 2-fold higher level than that of the wildtype control, glycogen storage would still be dramatically reduced with the accumulation of intracellular sucrose under hypersaline conditions ( Figure 4B ). During the cultivation process with salt stress, glycogen contents in XC8 were maintained at the same level as that of wildtype control, with or without theophylline induction. However, the theophylline induction (1,000 μM) of glgP expression failed to elevate sucrose accumulation in XC8 cells. After 12 days of salt stressed cultivation, although intracellular sucrose concentration of XC8 cells was slightly higher than that of the wildtype control, no significant difference could be detected between the samples with or without theophylline induction ( Figure 4C ). Under hypersaline conditions, sucrose would be synthesized and accumulated in Synechococcus cells to resist the hyperosmotic stress and would be maintained at a favorable concentration rather than unlimited accumulation, which might be a possible explanation for the phenomenon that the reduction of glycogen contents failed to bring in elevated sucrose yield. FIGURE 4 Effects of theophylline-regulated glgP expression on salt stress tolerance and intracellular sucrose storage in PCC 7942. (A) Growths of PCC 7942 wildtype and XC8 mutant strains facing 150 mM salt stress imposed on Day 3 with different doses of theophylline. (B) Glycogen storage of PCC 7942 wildtype and XC8 mutant strains facing 150 mM salt stress imposed on Day 3 with different doses of theophylline. (C) Intracellular sucrose concentrations of PCC 7942 wildtype and XC8 mutant strains facing 150 mM salt stress imposed on Day 3 with different doses of theophylline. Theophylline-Induced glgP Overexpression Failed to Enhance Salt Stress–Induced Secretory Sucrose Synthesis of PCC 7942 Cell Factories Previously, it has been confirmed that the introduction of the E. coli– sourced cscB sucrose permease could facilitate the secretory synthesis of sucrose, which removes the potential intracellular over-accumulation effects and significantly improved sucrose production titers in cyanobacteria ( Ducat et al., 2012 ; Song et al., 2016 ). Inspired by this, we explored whether the GlgP overexpression strategy could be used for enhancing secretory sucrose production in engineered cell factories. With a previously developed engineered strain FL130 carrying a heterologous cscB gene and an overexpressed native sps gene (encoding the sucrose-6-phosphate synthase, the rate-limiting enzyme catalyzing sucrose synthesis in PCC 7942), we utilized the Ptrc-ENYC4 system to replace the native glgP promoter sequence and obtained the recombinant strain XC14 ( Figure 5A ). As shown in Figures 5B,C , when 1,000 μM theophylline was added to induce the overexpression of glgP , growth of XC14 would be inhibited, and the glycogen contents were reduced to a similar level to those of the FL130 control. After 150 mM NaCl was added to activate Sps and sucrose synthesis, glycogen storage in all the sets would be continuously decreased, indicating that a significantly rewired intracellular carbon distribution leads to enhanced sucrose production. After 3 days of cultivation under salt stress, glycogen contents in the XC14 strain without theophylline induction were still about 2-folds higher than those of the FL130 control and the XC14 cells with induction of 1,000 μM theophylline ( Figure 5C ). However, the difference in glycogen storage and glgP expressions caused minor effects on specific productivities of secreted sucrose ( Figure 5D ) on the per cell levels, and the final sucrose titer of the theophylline-induced XC14 strain was lower than that of the control due to the reduced cell growth. Comparing the wildtype and XC8 strains cultivated in flasks and column photobioreactors, the theophylline-induced glgP overexpression failed to improve the actual performances of the sucrose synthesis cell factories stressed by 150 mM NaCl. A possible explanation would be the low contribution ratio of glycogen storage in the salt-induced sucrose synthesis process. Due to the introduction of sucrose permease, the limitation of intracellular sucrose accumulation was removed, and a much larger portion of the carbon source would be rewired into sucrose synthesis, which would be then secreted to the extracellular environments. The carbon source from mobilized glycogen storage would take a much lower ratio for sucrose synthesis than that in the non-secretory mode. Thus, enhanced glycogen degradation could only bring in negligible contribution to carbon flow distribution, while the burden on protein overexpression and theophylline further caused the weakened performance of cell growth and sucrose titers, which is in accordance with the previously reported phenomenon by Ducat et al. (2012) . FIGURE 5 Effects of theophylline-induced glgP overexpression on salt-induced sucrose synthesis and secretion in PCC 7942 cell factories. (A) Design and working mode of the theophylline-regulated glgP expression system in a previously constructed sucrose-synthesizing strain FL130. The expression of the cscB gene would be induced by the addition of IPTG, facilitating sucrose secretion out of the cells. An additional copy of the native sucrose-6-phosphate synthase gene would be expressed by a constitutive strong promoter Ptrc , and when salt stress was imposed, the intracellular Sps would be activated by increased concentrations of NaCl, leading to enhanced sucrose synthesis with the carbon source from CBB cycle and glycogen storage. The promoter sequence of the native glgP gene would be replaced with the Ptrc-ENYC4 system, and the addition of theophylline would lead to overexpression of glgP and enhanced glycogen degradation. The strain derived from FL130 was termed XC14. (B) Cell growths, (C) glycogen storage, and (D) extracellular sucrose production of FL130 and XC14 (with or without theophylline induction) strains before and after salt stress on Day 3. T0 and T1000 represented theophylline concentrations of 0 and 1,000 μM, respectively. GlgP Overexpression Improved Sucrose Synthesis in PCC 7942–Derived Cell Factories Under Non-Hypersaline Conditions Based on the effects of theophylline-induced glgP expression on glycogen storage and sucrose synthesis in XC14 facing salt stress, we supposed that this strategy might more possibly be valuable in non-hypersaline conditions to link the glycogen degradation metabolism and synthesis of desired metabolites. As for PCC 7942, sucrose synthesis would be activated through salt ion–induced activation of sucrose phosphate synthase and inhibition of invertase ( Liang et al., 2020 ), while it has been confirmed that Sps from PCC 6803 (hereafter termed as Sps 6803 for short) was not a salt-activated enzyme, and introduction of the Sps 6803 in the another Synechococcus strain UTEX 2973 successfully resulted in a secretory synthesis of sucrose without the need for salt stress induction ( Lin et al., 2020 ). We took the same strategy to construct a PCC 7942–derived cell factory producing sucrose under non-hypersaline conditions by overexpressing Sps 6803 in the FL92 strain which was previously developed and carried the same cscB -expression cassette as FL130 ( Figure 6A ), and the strain was termed as JS28 ( Ptrc-cscB ; PcpcB-sps \n \n 6803 \n ). As shown in Figure 6B , growth patterns of the FL130, JS28, and the two respective glgP-overexpressing strains (JS33 and JS34) were quite similar under non-hypersaline conditions, while the glgP-overexpression caused a slight decrease of cell densities. As shown in Figure 6D , JS28 could synthesize about 500 mg/L sucrose, about 2.5 folds higher than that of the FL130 ( Ptrc-cscB ; Ptrc-sps ) under the same non-hypersaline conditions. Glycogen storage in JS28 was also 50% lower than that of FL130, indicating that the secretory sucrose synthesis deprived a large portion of carbon flow from glycogen storage under non-hypersaline conditions ( Figure 6C ). FIGURE 6 Effects of glgP overexpression on sucrose synthesis of PCC 7942 cell factories under non-hypersaline conditions. (A) To facilitate sucrose synthesis under non-hypersaline conditions, the PCC 6803 sourced sps gene sps \n \n 6803 \n was put under the control of the PcpcB promoter to replace the Ptrc expressed native sps gene in FL130, and the obtained strain was termed as JS28. To enhance glgP expression, a PcpcB promoter–driven glgP expression cassette was introduced into FL130 and JS28, generating JS33 and JS34, respectively. Without salt stress induction, the Sps in FL130 and JS33 would be inactivated, and the sucrose synthesis would be inhibited. In JS28 and JS34, the NaCl-activation–independent Sps 6803 could normally catalyze the sucrose synthesis process. Cultivated in column photobioreactors in standard conditions without salt stress, (B) cell growth, (C) glycogen contents, and (D,E) sucrose production of FL130, JS33, JS28, and JS34 cells were calculated in 9 days. To explore the effects of enhanced glycogen degradation, we expressed glgP with a strong constitutive promoter PcpcB in both JS28 and FL130 ( Figure 6A ) and obtained the strains JS34 ( Ptrc-cscB ; PcpcB-sps \n \n 6803 \n ; PcpcB-glgP ) and JS33 ( Ptrc-cscB ; Ptrc-sps ; PcpcB-glgP ) respectively. As expected, glycogen storage in both of the two strains was reduced. In JS34, sucrose synthesis was correspondingly increased by about 2.4-folds compared with that of JS28 (with no glgP overexpression), reaching about 1,200 mg/L (accounting for up to about 40% of the fixed carbon, Figure 6E ), while in JS33 (FL130 with PcpcB expressed glgP ), the sucrose productivities and sucrose titers were not significantly increased compared with those of FL130, indicating that as for FL130 cells under non-hypersaline conditions, carbon flux was not the rate-limiting factor restricting sucrose synthesis. Previously Ducat et al. have reported that combined overexpression of glgP and cscB could not improve sucrose productivities while causing a 5%–10% decrease ( Ducat et al., 2012 ). In this work, we also evaluated the performances of similar strains JS33 ( Ptrc-cscB ; Ptrc-sps ; PcpcB-glgP ) cultivated under hypersaline conditions and got similar results when compared with the FL130 control ( Ptrc-cscB ; Ptrc-sps ). As shown in Supplementary Figure S2 , when cultivated under hypersaline conditions, expression of glgP reduced the sucrose titer from 2 g/L to 1.5 g/L after 9 days of cultivation. As mentioned above, the sharply reduced glycogen storage of salt-stress–induced sucrose-synthesizing cells provides a reasonable explanation for that phenomenon. As for FL130 strain facing continuous salt stress, a large portion of the carbon source (including the glycogen storage) has been motivated for sucrose synthesis; thus, the enhanced glgP expression and glycogen degradation activities could only bring in minor contributions but just consume more energy and resources on meaningless protein (GlgP) synthesis, which in turn causes impairment on the final sucrose yields. The salt stress–responsive sucrose synthesis serves as a special case for cyanobacteria-based photosynthetic production, and the naturally evolved condition–induced metabolism shift spontaneously plays a role in regulating glycogen metabolism, and thus the artificially implemented glycogen motivation strategy could not bring in necessary effects. However, as for a majority of cyanobacteria metabolic engineering cases, the carbon sink in glycogen storage ( Davies et al., 2014 ; van der Woude et al., 2014 ; David et al., 2018 ) and how to rewire the natural carbon sink into artificially assembled heterologous pathway was still an important issue to be solved. With the salt stress–independent sucrose synthesizing cell factories as a model, we have confirmed the feasibility of regulating the glycogen degradation pathway to engineer the glycogen metabolism and production capacities of the desired product."
} | 7,414 |
30027501 | PMC6208612 | pmc | 2,332 | {
"abstract": "Several studies have described that cyanobacteria use blue light less efficiently for photosynthesis than most eukaryotic phototrophs, but comprehensive studies of this phenomenon are lacking. Here, we study the effect of blue (450 nm), orange (625 nm), and red (660 nm) light on growth of the model cyanobacterium Synechocystis sp. PCC 6803, the green alga Chlorella sorokiniana and other cyanobacteria containing phycocyanin or phycoerythrin. Our results demonstrate that specific growth rates of the cyanobacteria were similar in orange and red light, but much lower in blue light. Conversely, specific growth rates of the green alga C. sorokiniana were similar in blue and red light, but lower in orange light. Oxygen production rates of Synechocystis sp. PCC 6803 were five-fold lower in blue than in orange and red light at low light intensities but approached the same saturation level in all three colors at high light intensities. Measurements of 77 K fluorescence emission demonstrated a lower ratio of photosystem I to photosystem II (PSI:PSII ratio) and relatively more phycobilisomes associated with PSII (state 1) in blue light than in orange and red light. These results support the hypothesis that blue light, which is not absorbed by phycobilisomes, creates an imbalance between the two photosystems of cyanobacteria with an energy excess at PSI and a deficiency at the PSII-side of the photosynthetic electron transfer chain. Our results help to explain why phycobilisome-containing cyanobacteria use blue light less efficiently than species with chlorophyll-based light-harvesting antennae such as Prochlorococcus , green algae and terrestrial plants. Electronic supplementary material The online version of this article (10.1007/s11120-018-0561-5) contains supplementary material, which is available to authorized users.",
"introduction": "Introduction Almost 140 years ago, professor Theodor Engelmann showed that light color plays an important role in photosynthesis (Engelmann 1882 ). In his classic experiment, Engelmann placed a filamentous green alga from the genus Cladophora on a microscopic slide, which he illuminated through a prism glass, thus dividing sunlight into separate wavelengths across the filament. By introducing aerotactic bacteria and observing in which regions of visible light these bacteria aggregated, he established that photosynthetic oxygen (O 2 ) production occurred in red and blue light, thereby creating the first “living” action spectrum of chlorophyll. In the following years, Engelmann continued his studies with cyanobacteria from the genus Oscillatoria , demonstrating that in these cyanobacteria, not only red and blue light but also orange light resulted in high O 2 production rates (Engelmann 1883 , 1884 ). Engelmann’s findings were criticized for many years, but 60 years later, his results were confirmed by Emerson and Lewis, who showed that the phycobiliproteins of cyanobacteria and red algae play a key role in light-harvesting for photosynthesis (Emerson and Lewis 1942 ). We now know that these phycobiliproteins make up specialized light-harvesting antennae, called phycobilisomes (PBSs), consisting of an allophycocyanin core and stacked rods of phycocyanin often in combination with phycoerythrin. These phycobiliproteins consist of an apo-protein and one or more chromophores, also known as bilins, including phycocyanobilin absorbing orange light (620 nm), phycoerythrobilin absorbing green light (545 nm), and phycourobilin absorbing blue-green light (495 nm) (Grossman et al. 1993 ; Tandeau de Marsac 2003 ; Six et al. 2007 ). Recent reviews on the structure and function of PBSs are provided by Tamary et al. ( 2012 ), Watanabe and Ikeuchi ( 2013 ), and Stadnichuk and Tropin ( 2017 ). Light energy absorbed by PBSs is effectively transferred via allophycocyanin to the chlorophyll a (Chl a ) pigments in the photosystems (Arnold and Oppenheimer 1950 ; Duysens 1951 ; Lemasson et al. 1973 ). It has long been assumed that most PBSs transfer their energy to photosystem II (PSII). It is now well established, however, that cyanobacteria are able to re-balance excitation energy by moving PBSs between photosystem I (PSI) and PSII in a process called state transitions (van Thor et al. 1998 ; Mullineaux 2008 ). As a consequence of these state transitions, which occur at time scales of seconds to minutes, the PBSs associate with PSII (state 1) or PSI (state 2) and transfer the absorbed light energy to the reaction center of the photosystem they are associated with (Kirilovsky 2015 ). At longer time scales, cyanobacteria can also adjust their PSI:PSII ratio to optimize their photosynthetic activity under different environmental conditions (Fujita 1997 ). In cyanobacteria, the PSI:PSII ratio generally ranges between 5:1 and 2:1 depending on light quality and intensity, which is higher than the approximately 1:1 ratio often found in eukaryotic phototrophs (Shen et al. 1993 ; Murakami et al. 1997 ; Singh et al. 2009 ; Allahverdiyeva et al. 2014 ; Kirilovsky 2015 ). Since blue and red light are both strongly absorbed by Chl a , and the intermediate wavelengths by the different phycobiliproteins, one would expect that these light colors are all used for photochemistry at approximately equal efficiency. However, several studies have described that blue light yields lower O 2 production rates than red light in cyanobacteria (Lemasson et al. 1973 ; Pulich and van Baalen 1974 ; Jørgensen et al. 1987 ; Tyystjärvi et al. 2002 ), in cyanolichens (Solhaug et al. 2014 ), and also in PBS-containing red algae (Ley and Butler 1980 ; Figueroa et al. 1995 ). Furthermore, other studies noted that blue light resulted in lower growth rates in a variety of cyanobacteria (Wyman and Fay 1986 ), including Synechocystis sp. PCC 6803 (Wilde et al. 1997 ; Singh et al. 2009 ; Bland and Angenent 2016 ), Synechococcus sp. (Choi et al. 2013 ), and Spirulina platensis (Wang et al. 2007 ; Chen et al. 2010 ). A possible explanation for their poor performance in blue light might be that most chlorophyll of cyanobacteria is located in PSI (Myers et al. 1980 ; Fujita 1997 ; Solhaug et al. 2014 ; Kirilovsky 2015 ), and hence, blue light induces high PSI but low PSII activity. This phenomenon is also known from fluorescence studies, where the use of blue measuring light complicates interpretation of the fluorescence signal of cyanobacteria (Campbell et al. 1998 ; Ogawa et al. 2017 ). However, although several of the above-cited studies measured growth rates and/or pigment composition in different light colors, they did not report on, e.g., O 2 production, PSI:PSII ratios, or state transitions. Conversely, other studies measured O 2 production rates or PSI:PSII ratios but did not measure growth rates or other relevant parameters. To our knowledge, more comprehensive studies of the photophysiological response of cyanobacteria to blue light are largely lacking, and no clear consensus has yet been reached on the question why their photosynthetic activity might be hampered by blue light. In this study, we compare the effect of blue light with that of orange and red light on the model cyanobacterium Synechocystis sp. PCC 6803. This cyanobacterium uses Chl a in its photosystems and phycocyanin but not phycoerythrin in its phycobilisomes, and hence effectively absorbs blue, orange, and red light. We investigate its photosynthetic performance, the composition of its photosynthetic machinery, and its growth rate at different light intensities for all three colors, and compare these results with growth of the green alga Chlorella sorokiniana and of other cyanobacteria containing phycocyanin and phycoerythrin. The experiments will make use of blue, orange, and red LED light with narrow-band wavelengths, which allows more precise investigation of the photosynthetic response to different light colors than the broad-band light filters used in the older literature. Our results demonstrate that blue light has a major impact on the photophysiology of cyanobacteria.",
"discussion": "Discussion Our results show that the cyanobacterium Synechocystis sp. PCC 6803 absorbs blue light to at least a similar extent as orange and red light (Figs. 2 , 6 ), but uses the absorbed blue light much less effectively for oxygenic photosynthesis and growth (Figs. 3 , 4 , 5 ). The poor photosynthetic performance of cyanobacteria in blue light has also been reported by several earlier studies (e.g., Lemasson et al. 1973 ; Wyman and Fay 1986 ; Wilde et al. 1997 ; Tyystjärvi et al. 2002 ; Wang et al. 2007 ; Singh et al. 2009 ; Chen et al. 2010 ; Choi et al. 2013 ; Solhaug et al. 2014 ; Bland and Angenent 2016 ), but in-depth investigations were still lacking. Our results support the hypothesis that blue light creates an imbalance between the two photosystems, with an excess of energy at the PSI-side and a deficiency at the PSII-side of the photosynthetic electron transport chain of cyanobacteria (e.g., Solhaug et al. 2014 ; Kirilovsky 2015 ). This hypothesis is explained below. In orange and red light, cyanobacteria usually have 2–5 times more PSI than PSII (Fig. 7 a–c; see also Singh et al. 2009 ; Allahverdiyeva et al. 2014 ; Kirilovsky 2015 ). Furthermore, cyanobacterial PSI contains around 100 molecules of Chl a (Jordan et al. 2001 ; Kennis et al. 2001 ), while a single PSII contains only about 35 Chl a molecules (Guskov et al. 2009 ; Umena et al. 2011 ). Hence, cyanobacteria invest much more of their chlorophyll in PSI than in PSII, and, therefore, PSI will absorb more photons than PSII (e.g., Myers et al. 1980 ; Fujita 1997 ). This dissimilarity between the two photosystems is compensated for by the light-harvesting PBSs, which tend to be associated mostly with PSII and hence transfer most of their absorbed photons to PSII (van Thor et al. 1998 ; Joshua et al. 2005 ; Mullineaux 2008 ). In this way, cyanobacteria can maintain linear electron flow by balancing excitation energy between the two photosystems, which will enable the production of both ATP and NADPH required for growth (Allen 2003 ; Nogales et al. 2012 ; Mullineaux 2014 ). In blue light, the PBSs do not absorb photons very effectively, because the short wavelength of 450 nm does not match the absorption spectrum of phycocyanin (Fig. 2 ; see also, e.g., Tandeau de Marsac 2003 ; Six et al. 2007 ). Hence, in blue light, the PBSs hardly transfer any light energy to PSII. In contrast, the chlorophylls of the more abundant PSI still absorb blue light effectively. Moreover, in cyanobacteria, β-carotenes absorbing blue and green wavelengths are also more abundant in PSI than PSII and further contribute to photosynthetic light harvesting by PSI (Ritz et al. 2000 ; Takaichi 2011 ; Stamatakis et al. 2014 ). Hence, most blue photons are absorbed at PSI, while in comparison PSII has a severe shortage of photons, which will suppress linear electron transport. This is supported by the low O 2 production rates that we measured when the cyanobacteria were exposed to limiting levels of blue light. Interestingly, once the intensity of blue light was high enough to saturate PSII, the O 2 production rate in blue light approached the rate found in orange and red light (Fig. 5 ). The important role of PBS in the distribution of excitation energy over the two photosystems is illustrated by a study of Campbell ( 1996 ) on complementary chromatic adaptation of the cyanobacterium Calothrix sp. strain PCC 7601. When acclimated to green light the PBSs of this species contain mostly phycoerythrin, while they contain mostly phycocyanin when adapted to red light. Rates of photosynthesis were similar in both light colors, but 77 K fluorescence emission revealed major differences in the distribution of light by the PBSs. Similar to our experimental results, red light (600–750 nm) applied by Campbell ( 1996 ) was absorbed by both Chl a and (allo)phycocyanin, and the PBSs of red-light acclimated cells were mostly associated with PSII while Chl a harvested light for PSI. In contrast, green light (500–575 nm) was only absorbed by phycoerythrin, and the phycoerythrin-containing PBSs of green-light acclimated cells transferred light energy to both PSI and PSII to maintain the balance between these two photosystems. Our results indicate that blue-light acclimated cells attempt to restore the balance between the two photosystems in a similar way. Fluorescence emission by the two photosystems after excitation of Chl a at 440 nm showed that blue-light acclimated cells decrease their PSI:PSII ratio (Fig. 7 a–c), which will increase light absorption by PSII relative to PSI. Furthermore, fluorescence emission after excitation of phycocyanin at 590 nm revealed that blue-light acclimated cells associate even more of their PBSs with PSII than cells acclimated to orange and red light (Fig. 7 d–f). All these results indicate that cells exposed to blue light try to increase the transfer of light energy to PSII, to restore the balance. Yet, the low O 2 production rates indicate that cells in blue light are still unable to sustain a high rate of linear electron flow in comparison to those in orange and red light, unless saturating amounts of blue light are provided. Previous research has shown that in high CO 2 conditions, cyanobacteria decrease their PSI:PSII ratio and increase phycocyanin relative to Chl a (MacKenzie et al. 2004 ; Eisenhut et al. 2007 ). These changes are similar to the changes we observed in blue-light acclimated cultures, indicative of an increased transfer of light energy to PSII at high CO 2 conditions. Since our experiments were also performed at high CO 2 conditions, the results of MacKenzie et al. ( 2004 ) and Eisenhut et al. ( 2007 ) suggest that the photosynthetic efficiency of cyanobacteria in blue light might be even lower under carbon-limiting conditions. Since red light is also absorbed by Chl a , why did red light not give the same results as blue light? The answer is probably related to the wavelength of 660 nm that we used as the red-light source in our experiments. This wavelength is already longer than the 620–650 nm-red light used in many previous studies (Wyman and Fay 1986 ; Figueroa et al. 1995 ; Wang et al. 2007 ; Singh et al. 2009 ; Chen et al. 2010 ; Choi et al. 2013 ; Solhaug et al. 2014 ; Zavřel et al. 2017 ). This range of wavelengths around 660 nm is absorbed not only by chlorophyll, but also very effectively by allophycocyanin in the core of the PBS (Lemasson et al. 1973 ; Glazer and Bryant 1975 ; MacColl 2004 ). Hence, the PBSs can still redistribute photons of 660 nm over both photosystems, as indicated by the 77 K fluorescence data of the red-light acclimated cells after excitation of the PBSs (Fig. 7 f). One would expect that cells grown with red light at a longer wavelength of 680 nm, which is much less absorbed by allophycocyanin and hence specifically targets Chl a , would display similar results as our blue-light exposed cells. Indeed, other studies provide support for this hypothesis. For instance, Murakami ( 1997 ) showed that Synechocystis sp. PCC 6714 had a substantially lower PSI:PSII ratio when grown at 680-nm red light than at 650-nm red light, resembling the low PSI:PSII ratio that we found for Synechocystis sp. PCC 6803 in blue light. Bland and Angenent ( 2016 ) found that Synechocystis sp. PCC 6803 had a lower specific growth rate at 680 nm than at 660 nm, although the growth rate was even lower in blue light (440 and 460 nm). Blue-light absorbing β-carotenes are more abundant in PSI than in PSII, and in cyanobacteria they contribute to light harvesting only for PSI (Stamatakis et al. 2014 ). Hence, we speculate that blue light causes an even stronger imbalance between PSI and PSII than red light of 680 nm, which may explain their observation (Bland and Angenent 2016 ) that the growth rate was even lower in blue light than in red light of 680 nm. Our findings show that the poor performance in blue light is not specific for the freshwater cyanobacterium Synechocystis sp. PCC 6803, but also applies to marine cyanobacteria including both phycocyanin-rich strains such as Synechococcus sp. CCY 9201 and phycoerythrin-rich strains such as Synechococcus sp. CCY 9202 (Fig. 8 ). The low rates of growth and photosynthesis in blue light even extend to red algae which also possess PBSs composed of phycocyanin and phycoerythrin (Ley and Butler 1980 ; Figueroa et al. 1995 ). Hence, although the chromophores phycoerythrobilin and phycourobilin absorb green light (absorption peak at 545 nm) and blue-green light (495 nm), respectively, these organisms are not able to absorb and redistribute deep blue light (< 450 nm) very effectively between the two photosystems. In green algae, growth rates in blue and red light are comparable (Fig. 3 b; see also Teo et al. 2014 ; Yan and Zheng 2014 ; Zhao et al. 2015 ; de Mooij et al. 2016 ), thus blue light does not seem to result in an excitation imbalance between their PSII and PSI. Green algae maintain a lower PSI:PSII ratio than cyanobacteria (Murakami et al. 1997 ; Kirilovsky 2015 ) and utilize light-harvesting antennae composed of Chl a and b (Kühlbrandt et al. 1994 ). Additionally, green algae are able to use blue light more efficiently due to a wider variety of light-harvesting carotenoids, in their photosystems and light-harvesting antennae, which also absorb photons in the blue-green part of the visible light spectrum (Takaichi 2011 ). Contrary to cyanobacteria, these carotenoids harvest light energy not only for PSI, but also for PSII (Goedheer 1969 ). Hence, in contrast to the cyanobacterial PBSs, the light-harvesting antennae of green algae absorb both blue and red light, and transfer the absorbed light energy to both PSI and PSII in a balanced way. Our results help to explain why Prochlorococcus species dominate over Synechococcus species in the oligotrophic oceans (Partensky et al. 1999 ; Biller et al. 2015 ). Because red light is strongly absorbed by water molecules, blue light (400–500 nm) prevails in the deeper water layers of the open ocean (Stomp et al. 2007 ). Instead of PBSs, Prochlorococcus utilizes light-harvesting antennae composed of divinyl Chl a and b (Chisholm et al. 1992 ; Ting et al. 2002 ). Consequently, like green algae, they can balance the amply available blue light between both photosystems. In contrast, Synechococcus species utilize blue light much less effectively. The PBSs of marine Synechococcus strains of the oligotrophic ocean often contain high contents of phycourobilin (Palenik 2001 ; Everroad et al. 2006 ; Six et al. 2007 ; Grébert et al. 2018 ), with which they do absorb blue-green light (495 nm). However, none of the phycobiliproteins described so far extends its absorption to the deep-blue wavelengths (< 450 nm) that form a major part of the underwater light spectrum characteristic of the oligotrophic ocean. Hence, not only does blue light offer a suitable habitat for the chlorophyll-based light-harvesting antennae of Prochlorococcus , as has been described by many previous studies (Scanlan and West 2002 ; Ting et al. 2002 ; Rocap et al. 2003 ; Stomp et al. 2007 ), but blue light (< 450 nm) is also less suitable for phycobilisome-containing cyanobacteria such as Synechococcus ."
} | 4,870 |
35484107 | PMC9051161 | pmc | 2,333 | {
"abstract": "A self-organizing map (SOM) is a powerful unsupervised learning neural network for analyzing high-dimensional data in various applications. However, hardware implementation of SOM is challenging because of the complexity in calculating the similarities and determining neighborhoods. We experimentally demonstrated a memristor-based SOM based on Ta/TaO x /Pt 1T1R chips for the first time, which has advantages in computing speed, throughput, and energy efficiency compared with the CMOS digital counterpart, by utilizing the topological structure of the array and physical laws for computing without complicated circuits. We employed additional rows in the crossbar arrays and identified the best matching units by directly calculating the similarities between the input vectors and the weight matrix in the hardware. Using the memristor-based SOM, we demonstrated data clustering, image processing and solved the traveling salesman problem with much-improved energy efficiency and computing throughput. The physical implementation of SOM in memristor crossbar arrays extends the capability of memristor-based neuromorphic computing systems in machine learning and artificial intelligence.",
"introduction": "Introduction Neuromorphic computing systems built with memristors could have much-improved power efficiency and computing throughput than traditional hardware 1 – 6 . Recent implementations with memristors are, however, primarily artificial neural networks for supervised learning algorithms 7 – 13 . Unsupervised learning networks inspired by biological systems can learn through data sets without labels and hence are more energy- and cost-efficient 14 – 16 . Compared with the supervised network, the unsupervised learning are more similar with human brain and has more extensive application, considering that most data and information are unlabeled in real world. Besides, unsupervised approach can cluster or pre-process the unlabeled complex data to smaller subspaces for subsequent classification through another supervised network. A self-organizing map (SOM), also called a ‘Kohonen network’, is a frequently used unsupervised algorithm inspired by the topological maps in the sensory-processing areas of the brain, where neurons responding to similar inputs are spatially located very close 17 , 18 . As a result, SOMs can identify relationships of input data and are well suited for clustering and optimization problems such as language recognition and text mining, financial predictions, and medical diagnosis 19 – 24 . Furthermore, SOM is a nonlinear dimension-reduction tool that automatically maps high-dimensional data to a lower dimension (usually two- or one-dimensional), more effectively in nonlinear distributions than classical linear algorithms such as multi-dimensional scaling 25 or principal components analysis 26 . However, implementing SOM in conventional CMOS-based hardware is limited by the complexity in calculating the similarities and determining neighborhoods, which imposes an enormous increase in computing time and power consumption as the number of neurons and features increase 27 . It is therefore imperative to seek emerging energy-efficient hardware with parallel computing capacity for SOM networks. Memristor, a two-terminal resistance switch with multiple conductance states as synaptic weights, has been organized into large-scale crossbar arrays to implement parallel and energy-efficient in-memory computing using physical laws 28 – 32 . On the other hand, experimental demonstrations of SOM using memristors are yet to be achieved due to two main challenges: finding the shortest Euclidean distance with the unnormalized vectors and implementing complex topology of SOM output layer without extra hardware cost 33 , 34 . Herein we report our experimental implementation of SOMs in a 128 × 64 1-transistor 1-memristor (1T1R) crossbar array and its applications in data mining and optimization. The similarities between inputs and weight are directly calculated through Euclidean distance in the hardware. The neighborhood function of SOM is directly realized by the topological structure of the memristor array without extra circuits. Memristor-based 1D-SOM and 2D-SOM are successfully employed to solve color clustering and traveling salesman problems. Compared with traditional hardware, the memristor-based SOM system has better power efficiency and higher parallelism in computing, extending the application range of memristor-based neuromorphic computing systems in artificial intelligence.",
"discussion": "Discussion We have experimentally demonstrated in-situ SOM in memristor crossbar arrays. The Euclidean distance is directly calculated in the hardware by adopting additional rows of 1T1R cells. The similarities between input vectors and weight vectors are computed in one readout step without normalized weights. We have further employed the memristor-based SOM in clustering and solving the traveling salesman problems. Taking advantage of the intrinsic physical properties of memristors and the massive parallelism of the crossbar architecture, the novel memristor-based SOM hardware has advantages in computing speed, throughput, and energy efficiency, compared with current state-of-the-art SOM implementation, as shown in Supplementary Table 2 . Unlike a digital counterpart, the entire read operation is performed in a single time step, so the latency does not scale with the size of the array. The energy consumptions of the memristor device are extremely low in both inference (~40 fJ) and update (~2.42 pJ) processes. And even for compute-intensive tasks such as image segment or optimization problems (e.g., solving TSP), the memristor-based SOM can achieve a high energy efficiency system due to the small energy required in a memristor-based weight array over an all-digital system. Besides, SOM hybrid systems such as SOM-MLP 52 , 53 , SOM-RNN 54 , SOM-LSTM 55 , 56 , show better performance and are more robust in pattern classification and prediction than simple artificial neural network systems. Our results encourage advances in the hardware implementation of unsupervised neural networks using emerging devices and provide a promising path towards machine learning or neuromorphic computing based on memristors."
} | 1,565 |
33536938 | PMC7848188 | pmc | 2,334 | {
"abstract": "The symbiosis between cnidarian hosts and photosynthetic dinoflagellates of the family Symbiodiniaceae (i.e., zooxanthellae) provides the energy foundation of coral reef ecosystems in oligotrophic waters. The structure of symbiont biota and the dominant species of algal symbiont partly shape the environmental adaptability of coral symbiotes. In this study, the algal symbiont cells were isolated from the tentacles of Galaxea fascicularis , a hermatypic coral with obvious differentiation in heat resistance, and were cultured in vitro with an improved L1 medium. An algal monoclonal cell line was established using separated algal culture drops and soft agar plating method, and named by GF19C1 as it was identified as Cladocopium sp. C1 (Symbiodiniaceae) based on its ITS1, ITS2, and the non-coding region of the plastid psbA minicircle ( psbA ncr ) sequences. Most GF19C1 cells were at the coccoid stage of the gymnodinioid, their markedly thickened (ca. two times) cell wall suggests that they developed into vegetative cysts and have sexual and asexual reproductive potential. The average diameter of GF19C1 cells decreased significantly, probably due to the increasing mitotic rate. The chloroplasts volume density of GF19C1 was significantly lower than that of their symbiotic congeners, while the surface area density of thylakoids relative to volumes of chloroplasts was not significantly changed. The volume fraction of vacuoles increased by nearly fivefold, but there was no significant change in mitochondria and accumulation bodies. Light-temperature orthogonal experiments showed that, GF19C1 growth preferred the temperature 25 ± 1°C (at which it is maintained post-isolation) rather than 28 ± 1°C under the light intensity of 42 ± 2 or 62 ± 2 μmol photons m –2 s –1 , indicating an inertia for temperature adaptation. The optimum salinity for GF19C1 growth ranged between 28–32 ppt. The monoclonal culture techniques established in this study were critical to clarify the physiological and ecological characteristics of various algal symbiont species, and will be instrumental to further reveal the roles of algal symbionts in the adaptive differentiation of coral-zooxanthellae holobionts in future studies.",
"introduction": "Introduction The establishment of specialized intracellular symbiotic relationship between reef-building corals and photosynthetic dinoflagellates (Symbiodiniaceae, also known as zooxanthellae) is the primary energy source for reef ecosystems to flourish in oligotrophic tropical shallow waters. Photosynthesis of symbiotic algae can supply more than 95% of the nutritional needs of the corals and contribute to the calcification of reef corals to form the carbonate framework of coral reefs ( Mallela, 2013 ). However, due to the essential differences in metabolic rates and nutritional requirements between corals and algal symbionts, the precise homeostasis necessary to maintain the symbiotic relationship is highly sensitive to environmental stresses ( Obura, 2009 ), especially to the synergistic stress of light intensity and temperature variations ( Lesser and Farrell, 2004 ; Ferrier-Pages et al., 2007 ; Hawkins et al., 2015 ). A thermal perturbation as little as 1°C above the average summer maxima could cause the breakdown of this symbiosis and lead to coral bleaching ( Hume et al., 2015 ). Since 1980s, worldwide coral bleaching events caused by global warming have become more and more frequent, with too short intervals allowing for a full recovery of mature assemblages. As a result, mass coral mortality and the severe degradation of the coral reefs structure and ecological functions occurred ( Hoegh-Guldberg et al., 2007 ; Hughes et al., 2017 ). Even more worrying, as global warming in progression, local extreme weather conditions are more frequently seen, and the coral reef ecosystems are likely to further decline ( Hughes et al., 2018 ). Therefore, analyzing the adaptation and resilience of reef-building coral holobionts has become the focus of coral reef protection and resilience. This is bound to start with the two symbiotic parties, respectively, to clarify the physiological and ecological characteristics and environmental adaptation potentials of reef-building corals ( Shinzato et al., 2011 ; Ying et al., 2018 ) and symbiotic algae ( Lin et al., 2015 ; Aranda et al., 2016 ). And then, bring it to the level of holobiont as a unique biological entity of evolutionary selection for integrated research ( Rosenberg, 2013 ). The algal symbiont used to belong to Symbiodinium , a genus with obscure morphological and taxonomic characteristics with complex phylogenetic lineages ( LaJeunesse et al., 2018 ). The species of this genus formed a complex symbiotic relationship with numerous categories of marine invertebrates ( Trench, 1993 ). In the past 30 years, DNA sequences and molecular biology techniques have been used to establish a comprehensive phylogenetic relationship for algal symbionts derived from various marine invertebrates. Based on nuclear 18S-rDNA and restriction fragment length polymorphisms (RFLPs) ( Rowan and Powers, 1991a , b ), chloroplast 23S-rDNA gene sequence ( Santos et al., 2002 ; Pochon et al., 2006 ; Pochon and Gates, 2010 ), Symbiodinium was classified into nine (A–I) genetically distinctive clades. And each clade is further divided into multiple subclades based on the nuclear internal transcribed spacer (ITS) regions ( LaJeunesse, 2001 ; Van Oppen et al., 2005 ). Recently, LaJeunesse et al. (2018) systematically revised the evolutionarily divergent Symbiodinium Clade A–G to seven genera in the family Symbiodiniaceae, and some subclades or genetic strains are described as species within those genera. With increasing phylogenetic, ecological, and biogeographic evidences available, more novel genera and species will be likely uncovered and classified in the family Symbiodiniaceae ( LaJeunesse et al., 2018 ). The host species associated with algal symbionts are highly diverse, even when spoken of reef-building coral hosts, they are also miscellaneous. Since the existence of symbiont-host specificity at the species level ( Trench, 1993 ), the lineages of genetic differentiation of symbiotic algae, combining with that of coral hosts, have indicated much greater genetic and functional diversities in the algal symbionts of reef-building corals ( Barshis et al., 2014 ). Certainly, the establishment of in vitro monoclonal culture of host-associated algal symbiont strains would be necessary to elucidate the species identification and characterization through collecting and analyzing their physiological and ecological data. In addition, such efforts could facilitate to reveal the roles of algal symbionts in building, maintaining, breaking down and reconstruction of the symbiosis. Nevertheless, due to the numerous difficulties in the establishment of in vitro culture strains ( Schoenberg and Trench, 1980a ), the studies on the host-associated monoclonal algal symbiont cultures are still limited ( Chakravarti and Van Oppen, 2018 ). Galaxea fascicularis , a massive coral with large polyps, is mainly distributed in the tropical and subtropical coral reef areas of the Indian- Pacific Ocean ( Veron, 2000 ). It is also the dominant species on the fringing reefs of Hainan Island ( Chen et al., 2013 ; Wang et al., 2013 ). G. fascicularis species includes two morphologically and genetically differentiated lineages characterized by the microbasic p-mastigophores (MpM) types of tentacular nematocyst ( Hidaka, 1992 ) and mitochondrial genotypes ( Watanabe et al., 2005 ) around Hainan Island ( Wu, 2018 ; Wepfer et al., 2020 ). These two lineages demonstrate significant differences in heat resistance ( Xu, 2019 ), indicating the potential of G. fascicularis as an ideal model for exploring the genetic basis of heat resistance differentiation of corals. Thus, the establishment of in vitro monoclonal cultures of G. fascicularis associated algal symbiont species, followed by characterization of their morphological, physiological and ecological traits, would be necessary and helpful to reveal the symbiosis flexibility behind the differentiations in environmental adaptability between the two lineages, as well as the interaction between corals and symbionts within the holobiont. In this study, we conceived and developed the techniques for successful isolation and cultivation of the monoclonal symbiotic algal strain of G. fascicularis . and then characterized the first monoclonal algal strain GF19C1 ( Supplementary Figure 1 ).",
"discussion": "Discussion The symbiosis between algal symbionts and reef-building corals were thought to emerge in the mid-Triassic period ( Stanley, 1981 ). In the ensuing 230 million years of intense differentiation and speciation of the hermatypic corals ( Shepard, 1964 ), the two mutualistic sides, corals and their endosymbiotic algae, have undergone precise coordination or a series of synergistic mutations and formed obligate interdependence ( Antonelli et al., 2016 ). The photosynthetic symbionts have established relatively stable and complex communities in specific coral species and geographic regions ( Lewis et al., 2019 ), and the symbiont biota also changed with the persistence and periodicity of environmental stress ( Lee et al., 2016 ; Lewis et al., 2019 ). On the other hand, adapted to the life in coral cells, symbionts and their host have reached a delicate metabolic balance ( Lin et al., 2019 ) and physiological compromises (such as abandoning sexual reproduction and being compatible with the coral’s immune system) ( Antonelli et al., 2016 ). Therefore, when we attempt to isolate the symbionts from coral cells and establish genetically and physiologically consistent strains, the algae should experience drastic morphological and physiological changes so as to adapt to the new artificial culture environment. Isolation and in vitro Culture of GF19C1 The isolation of algal symbionts from Galaxea fascicularis and the establishment of monoclonal strain GF19C1 have undergone a tedious process for more than 1 year, while many efforts have been taken for optimization, including culture medium components, light intensity managements, and contamination control etc. We found that the common medium (e.g., f/2 and L1 medium) could be applied to GF19C1 in vitro culture upon minor modification. One of the greatest challenges was to eliminate the contamination of diatoms, protozoa, and fungi in the culture system. Although the use of triple antibiotics (penicillin, streptomycin and kanamycin, final concentration of 200, 100, and 100 μg/mL, respectively, Lin et al., 2015 ) can effectively inhibit the bacteria growth in the cultures, however, contaminants like fungi, diatoms and protozoa that originated from the coral exoskeleton or internal polyps during isolation, were difficult to eliminate. Chakravarti and Van Oppen (2018) used an antifungal cocktail (consisting of nystatin, amphotericin, and GeO 2 , final concentration 100 μg/mL, 2.5 μg/mL and 50 μM, respectively) to inhibit the contamination of fungi and other organisms during the centrifugal collection of symbiotic algae from plenty of broken tissues. As reported here, we used a different strategy to reduce the sources of contamination. We selected the tentacles of Galaxea fascicularis , which were easily sampled and cleaned for symbiont isolation. Meanwhile, we also gently teared down the tissue to avoid the damage of symbiont cells. The algal colonies were then cultured by the soft agar plates containing triple antibiotics. We eventually obtained the stable and passaged algal -strain GF19C1 cell line. The protocol of the established GF19C1 strain mainly followed the monoclonal culture procedure from Schoenberg and Trench (1980a) , which is a rather rigorous and tedious approach with the process involving the rejection of the dominant alga to achieve culture purification during the algae culture ( Chakravarti and Van Oppen, 2018 ). The symbiotic phylotypes associated with G. fascicularis were mainly dominated by Symbiodiniaceae ITS2-C1, D1, and C21a, along with numerous of other background clade C phylotypes ( Xu, 2019 ; Wepfer et al., 2020 ). Even identified as clade C phylotype by n18S-rDNA and Taq I-generated RFLPs ( Santos et al., 2002 ), the symbiotic biota in G. fascicularis could also possess multiple composition of both dominant and background symbionts when probed with ITS2 sequence tagging. Thus, a more rigorous strategy of isolation and purification process is necessary to achieve the pure cell line. Ultrastructural Changes of GF19C1 Our study revealed that the cell wall of in vitro cultured GF19C1 could be significantly thicker (about twofold) than that of their symbiotic congeners (SC). This phenomenon is rather common ( Lesser and Shick, 1990 ), but there are several aspects to account for its causes and results. One probable explanation is that, the thinner cell wall of symbionts in hospite is an adaptation to live inside coral cells, which may facilitate the transport of nutrients between coral cell and symbiont ( Schoenberg and Trench, 1980b ). Freudenthal (1962) noted that, the algal symbiont in coral cell is haplontic and autotrophic vegetative cell. When cultured in vitro , they become vegetative cyst by thickening the cell wall, which could restore both sexual and asexual reproductive potential through producing autospores, aplanospores, or motile gymnodinioid zoospores, or possible gametes, so that they can adapt to the outside environment ( Freudenthal, 1962 ). Palincsar et al. (1988) also observed the phenomenon of cell wall thickening of algae during growth outside the host sea anemone Aiptasia pallida . This was thought to be related to an increase in mitotic rate after isolation, because during the division the entire new cell walls were synthesized resulting thickened newly produced cells ( Palincsar et al., 1988 ). In addition, the mitotic rate of GF19C1 also increased by 8% higher than that of its SC, which was similar to that (10%) of in vitro cultured algal symbiont isolated from A. pallida ( Palincsar et al., 1988 ). The volume fraction (VF) of chloroplast of GF19C1 cells was significantly lower than that of their SC, whereas the surface density of thylakoid lamellae (SDTL) showed no obvious change. Unlike their SC, in vitro cultured algal symbionts don’t have to provide photosynthetic nutrients for coral cells in exchange for their protection ( Davies, 1993 ). Therefore, their photosynthetic burden is reduced and they don’t need to possess so many photosynthetic apparatuses. Especially when cultured in vitro , the light intensity is usually greater than that in host cells. Interestingly, for the algal strain isolated from sea anemone Aiptasia pallida , its VF of chloroplast didn’t change but SDTL remarkably decreased, when compared to their SC ( Lesser and Shick, 1990 ). Perhaps, that is another photo-adaptive strategy to a greater outside ( in vitro ) light intensity. With the notable reduction in chloroplast VF, the vacuoles of GF19C1 cells could join together and merge into one or two larger irregular vacuoles. And the volume density of vacuoles increases significantly, which may favor frequent mitosis in GF19C1. Freudenthal (1962) also found in vitro cultured symbiont cells have 1–2 large vacuoles, meanwhile, contain intact chloroplasts and accumulation bodies. Such cells not only do not mean aging, but extensively appear in the culture medium containing actively-reproducing algae. The appearance of these big vacuoles could imply some unknown factors on regulating cell physiology. We speculate that it is likely to be associated with the increased demand for processing metabolic wastes and more flexibility in space as cell division accelerating. Suitable Light, Temperature and Salinity Conditions of GF19C1 Growth Light and temperature are critical environmental factors for algae growth. We found that the initially isolated algal cells show a high sensitivity to light intensity, and might bleach to death in 3 days under 50 μmol photons m –2 s –1 , with 12:12 light/dark cycle. If the light intensity was reduced to 40 μmol photons m –2 s –1 , the survival time of the algal cells could be significantly prolonged. Even within coral cells, a sudden increase in light intensity could also reduce the pigment content rather than decrease the density of symbionts, and result in coral bleaching ( Hoegh-Guldberg and Smith, 1989 ). Generally, light intensity has a greater effect on algal growth than temperature ( Sakami, 2000 ). The appropriate light intensity range for GF19C1 increased from 40 to 64 μmol photons m –2 s –1 in 8 months post-isolation, when the cell growth rate at 25 ± 1°C was significantly higher ( P < 0.05) than that at 28 ± 1°C. This may be due to the fact that GF19C1 have been kept consistently at 25 ± 1°C since isolation, which result in an inertia of temperature preference for its growth and division so that they could not adapt to the sudden increase of temperature. The growth rate K on 8 months post-isolation was rather low. This may be because, on the one hand, GF19C1 hasn’t adapted to the extracellular environment yet, or on the other hand, the restrained influence of host cytokines is still present ( Cunning et al., 2015 ). When GF19C1 was cultured for 12 months in vitro , the growth rate K increased to 0.108, and showed more rapid evolution than the coral host ( Chakravarti and Van Oppen, 2018 ). Although this K value was still low compared with other free-living dinoflagellates ( Liang et al., 2011 ), it is comparable to that of other algal symbionts cultured in vitro ( Chakravarti and Van Oppen, 2018 ). Given that when GF19C1 were transferred from coral cells to the external environment, they may be more sensitive to salinity changes. Therefore, a dense gradient (ten gradients. from 10 to 40 ppt) was set to determine the optimum salinity in vitro . The effect of salinity on aquatic organisms is mainly manifested in the regulation of osmotic pressure in cells ( Sakami, 2000 ). At day 7, the density of GF19C1 increased in all salinity groups (10–40 ppt, Supplementary Table 3 ). As the experiment went on, comparing with the high salinity groups, the growth of low salinity groups (10–24 ppt) decreased more significantly ( Figure 4 and Supplementary Table 3 ) because of the accumulation of osmotic pressure ( Maboloc et al., 2015 ), at day 21, we can observe that the cytochrome gradually faded in the low salinity groups. Compared with the algal symbionts on mantle of juvenile giant clam Tridacna gigas , which display acclimation response to salinity of 25 ppt ( Maboloc et al., 2015 ), GF19C1 was more sensitive to low salinity stress. In addition, we also observed that, the cell density declined in all salinity groups at 28 days post-experiment, that is because the batch culture model was applied during the test and at that time the algae was at the Death/Lysis phase ( Farag and Price, 2013 ). GF19C1 cells can grow normally in a salinity range of 26–40 ppt, and its optimum salinity is 28–32 ppt, comparable to the suitable salinity (32–40 ppt) of most corals ( Veron, 1986 ). In suitable ranges, the effects of salinity on algal growth are not as obvious as that of light and temperature ( Hoegh-Guldberg and Smith, 1989 ). To sum up, GF19C1 is the first in vitro cultivated monoclonal strain isolated from the endosymbiotic biota of the scleractinian coral Galaxea fascicularis . When cultured in vitro , GF19C1 cells show morphological changes, preparation for the recovery of sexual reproduction, rapid adaptation to light intensity, and rapid evolution of growth rate. Symbiodiniaceae C1 are the most widespread algal symbiont in reef-building corals ( Stat et al., 2008 ), their distribution ranged from high latitude region e.g., Korea-Jeju Island and Japan, to tropical region e.g., South China Sea ( Reimer et al., 2006 ; Ng and Ang, 2016 ; Chen et al., 2019 , 2020 ; Wepfer et al., 2020 ), thus they have experienced large sea surface temperature variations and high turbidity ( Ng and Ang, 2016 ; Wepfer et al., 2020 ). Therefore, in future study it is necessary to further explore the adaptive potential of GF10C1 to multiple environmental factors and high gradient changes. The establishment of monoclonal culture technology will make it possible for the isolation and in vitro culture of G. fascicularis and other hermatypic corals-associated algal symbiont species, which is important to elucidate the physiological and ecological characteristics of various species of symbiont, and their cooperative mechanisms within coral hosts, and will help to further clarify their role and function in the environment adaptation of scleractinian corals."
} | 5,226 |
16332160 | PMC1310649 | pmc | 2,335 | {
"abstract": "Pollination is exclusively or mainly animal mediated for 70% to 90% of angiosperm species. Thus, pollinators provide an essential ecosystem service to humankind. However, the impact of human-induced biodiversity loss on the functioning of plant–pollinator interactions has not been tested experimentally. To understand how plant communities respond to diversity changes in their pollinating fauna, we manipulated the functional diversity of both plants and pollinators under natural conditions. Increasing the functional diversity of both plants and pollinators led to the recruitment of more diverse plant communities. After two years the plant communities pollinated by the most functionally diverse pollinator assemblage contained about 50% more plant species than did plant communities pollinated by less-diverse pollinator assemblages. Moreover, the positive effect of functional diversity was explained by a complementarity between functional groups of pollinators and plants. Thus, the functional diversity of pollination networks may be critical to ecosystem sustainability.",
"introduction": "Introduction Understanding the consequences of biodiversity loss for ecosystem functioning and services is currently a major aim of ecology [ 1 , 2 ]. Animal-mediated pollination is one of the essential ecosystem services provided to humankind [ 3 , 4 ]. The negative impact of pollinator decline on the reproductive success of flowering plants has been documented at the species level [ 5 – 7 ], but little information is available at the community level [ 8 ]. Increasing the scale of study to the community level is essential to account for potential competitive or facilitative effects among species that belong to the plant–pollinator network. Such effects, which are often linked to diversity [ 9 , 10 ], are known to have large influences on ecological processes such as community productivity and stability [ 11 , 12 ]. Experimental evidence for diversity effects on the functioning of terrestrial ecosystems is mainly available for plants. As primary producers, plants play a central role in the flow of energy within ecosystems [ 13 , 14 ]. Animal-pollinated angiosperms represent up to 70% of plant species in numerous communities and ecosystems [ 15 ]. Mutualistic interactions between animals and plants form several intricate interaction webs [ 16 ]. Recent analysis of plant–pollinator and plant–frugivore interaction webs demonstrates that these contain a continuum from fully specialist to fully generalist species [ 17 , 18 ]. However, these networks are structured in a nested way [ 19 , 20 ], with specialists mainly interacting with generalists. Such a pattern might have important consequences for ecosystem functioning, because it might confer resilience to perturbations such as the extinction of species [ 21 ] if, for example, generalist pollinators buffer the loss of specialist pollinators [ 18 , 22 – 24 ]. Furthermore, this hypothesis does not take into account the dynamical properties of these networks. In a plant–pollinator community, variations in species diversity at different trophic levels may lead to an adaptation of interaction strengths [ 25 ], which may in turn affect the total effectiveness of pollination. We conclude that more information is urgently needed concerning the impacts of biodiversity loss on multispecies and multitrophic interactions. To experimentally test the effect of functional diversity on the functioning and persistence of plant–pollinator communities, we defined functional groups of plants and pollinators based on morphological traits. For plants, two functional groups with three species each were defined according to accessibility of floral rewards (pollen and nectar; see Figure 1 ). The first group (group 1) included Matricaria officinalis, Erodium cicutarium, and Raphanus raphanistrum, which have easily accessible floral rewards and will be called “open flowers.” The second group (group 2), called “tubular flowers,” included Mimulus guttatus, Medicago sativa, and Lotus corniculatus, all of which present floral rewards hidden at the bottom of a tubular corolla. For pollinators, two functional groups were defined according to mouthparts length ( Figure 1 ). The first group (group A) included three species of syrphid flies (Diptera) with short mouthparts: Saephoria sp., Episyrphus balteatus, and Eristalis tenax. The second group (group B) included three species of bumble bees with longer mouthparts: Bombus terrestris, B, pascuorum, and B, lapidarius . Note that in this case a functional trait (long mouthparts) and a phylogenetic group are confounded. Preliminary observations showed that these six insect species contribute up to 70% of all pollinating visits to flowers in our study area in France. Constructing a plant–pollinator network with these four functional groups leads to a nested structure with specialists interacting with generalists ( Figure 1 , third column). In principle, syrphid flies cannot efficiently pollinate tubular flowers because their mouthparts are too short. Figure 1 Experimental Pollination Web Summary of the characteristics upon which functional groups of pollinators (left) and plants (right) were based. In the middle, the arrows linking insect heads to flower types show the theoretical pollination network when all functional groups are present. At the beginning of spring 2003, we set up 36 4-m 2 caged experimental plant communities. There were three plant treatments following a “substitutive” design [ 26 ]. Two of them contained one of the two plant functional groups alone (group 1 or 2), whereas the third contained both plant functional groups in combination (group 3). We applied three different pollination treatments to each plant treatment, by introducing each pollinator functional group alone (group A or B), or both groups together (group C). This full factorial design led to nine experimental treatments, which were replicated four times each, making a total of 36 experimental units. The pollination treatments were applied in two consecutive years (June–July 2003 and 2004). We controlled for the total number of pollinator visits received by each plot during the two pollination seasons (1,000 visits in 2003 and 1,200 visits in 2004) to allow an unbiased comparison of pollination efficiency among the various experimental treatments. In August and September 2003, we counted the number of fruits on each plant in every plot. We also counted the number of seeds per fruit on five collected fruits per plant. Lastly, in April 2004 and 2005, we measured both the number of plant species present at the seedling stage (recruitment richness) and the total number of seedlings (recruitment density) to determine the effects of the experimental treatments on the natural recruitment of the next plant generation.",
"discussion": "Discussion Previous studies on the diversity of plant–pollinator interaction webs were either descriptive [ 16 ], carried out on a single plant species [ 6 , 7 , 28 – 30 ], or based on simulation [ 21 ] and theoretical approaches [ 22 , 31 ]. To our knowledge, this is the first experimental evidence that the persistence of a plant community can be affected by a loss of diversity of its pollinating fauna. Of course, our experimental communities differed from natural ones in several respects. Among other things, the interaction networks we studied were much simpler than those occurring in nature; in particular, they contained fewer species in each trophic level. But such simplifications from natural situations are often necessary to carry out controlled experiments. In plant communities that contained only open flowers, plants produced fewer seeds per fruit in the bumblebee treatment than in the syrphid treatment ( Figure 2 C), but this was compensated by a sufficiently high fruit production, leading to a richness and density of natural recruitment that was similar to the other pollination treatments ( Figure 3 A and 3 B left). Thus, in these communities, all pollination treatments were equally effective in the long term. In plant communities that contained only tubular flowers, syrphids were inefficient pollinators; fruit production was very low ( Figure 2 A) and insufficient to allow a good natural recruitment. Bumble bees were the most effective pollinators ( Figure 3 A and 3 C, centre). Note that in the bumble bee treatment, the very high value of average recruitment density was due to three measurements in two replicates, in which only M. guttatus seedlings were recorded at a very high density (more than 150 seedlings per quadrat). To test the effect of these outliers, we removed them and repeated our analysis. The same significant effects were observed, except for the effect of pollination treatment, which became marginally significant ( p = 0.0645). The new mean number of seedlings per quadrat for this experimental treatment was 32.17 ± 4.55 (SE), which is still slightly above the value for the pollination treatment with both pollinator groups. For plant communities that contained only tubular flowers, recruitment richness in the two pollination treatments that contained bumblebees was similar. These results are in agreement with our theoretical pollination network presented in Figure 1 . In our experimental system, syrphids can be considered as specialist pollinators since they efficiently pollinate only open flowers. Bumble bees were potentially generalists as they induced an important fruit production of the two plant types and a good recruitment in the open- and tubular-flower plant treatments. Our results on the reproductive success and recruitment of single-guild plant treatments indicate that there are strong functional group identity effects since our plant functional groups responded differently to our pollinator functional groups. However, the functional diversity of both the plant and pollination treatments was also important. Plant reproductive success tended to increase with pollinator functional diversity when the number of seeds per fruit was considered, and with both plant and pollinator functional diversity when the number of fruits per plant was considered ( Figures 2 B and 2 D). Although recruitment in single-guild plant treatments was mainly affected by the identity of functional groups, the effect of functional diversity was dramatic in the mixed plant treatment. Natural recruitment of plant communities visited by mixed pollinator guilds was largely above that in other pollination treatments. Pollination by syrphids alone allowed the reproduction of open flowers but not tubular flowers, as expected from the specialisation of syrphids. More surprisingly, however, bumble bees failed to be efficient generalist pollinators. Most of their visits occurred on tubular flowers ( Figure 4 ), resulting in a relatively poor recruitment of open flowers. The only pollination treatment that achieved a high recruitment of both open and tubular flowers when they were mixed, was the one containing the two insect functional groups ( Figure 3 , right). When syrphids and bumble bees simultaneously pollinated mixed plant communities, they each focused on their target plant functional group, leading to more efficient visits and a better distribution of visits among plant functional groups ( Figure 4 ). Ultimately, it was the pollination treatment with both pollinator functional groups that produced the highest richness and density of natural recruitment. Consequently, since most natural plant communities contain both open and tubular flowers, pollinator functional diversity should strongly enhance the persistence of these communities. Although our experimental system differed from natural communities, and information about the reciprocal effects of the functional diversity of plant communities on the diversity of pollinator communities would be useful, our study indicates that the functional diversity of plant–pollinator interaction webs may be critical for the persistence and functioning of ecosystems and should be carefully monitored and protected. The loss of pollinator functional diversity is likely to trigger plant population decline or extinctions [ 4 ], which in turn are likely to affect the structure and composition of natural plant communities and the productivity of many agroecosytems that rely on insect pollination [ 8 ]. Ultimately, higher trophic levels may be affected since the diversity and biomass of consumers depend on primary production. Our results strongly suggest that the functional diversity of complex interaction webs plays a crucial role in the sustainability of ecosystems."
} | 3,168 |
35407265 | PMC9000849 | pmc | 2,336 | {
"abstract": "Eco-friendly energy harvesting from the surrounding environment has been triggered extensive researching enthusiasm due to the threat of global energy crisis and environmental pollutions. By the conversion of mechanical energy that is omnipresent in our environment into electrical energy, triboelectric nanogenerator (TENG) can potentially power up small electronic devices, serves as a self-powered detectors and predominantly, it can minimize the energy crisis by credibly saving the traditional non-renewable energy. In this study, we present a novel bio-based TENG comprising PDMS/α-Fe 2 O 3 nanocomposite film and a processed human hair-based film, that harvests the vibrating energy and solar energy simultaneously by the integration of triboelectric technology and photoelectric conversion techniques. Upon illumination, the output voltage and current signals rapidly increased by 1.4 times approximately, compared to the dark state. Experimental results reveal that the photo-induced enhancement appears due to the effective charge separation depending on the photosensitivity of the hematite nanoparticles (α-Fe 2 O 3 nanoparticles) over the near ultraviolet (UV), visible and near infrared (IR) regions. Our work provides a new approach towards the self-powered photo-detection, while developing a propitious green energy resource for the circular bio-economy.",
"conclusion": "4. Conclusions To sum up, the PDMS/α-Fe 2 O 3 nanocomposite and processed hair film-based TENG with both superlative photoelectric and triboelectric properties has been exhibited, which illustrates an extraordinary performance without any dependence on an external power source. This well-made bio-TENG based on the coupling of triboelectric and photoelectric conversion mechanisms achieved superior photo-induced enhancement with the output voltage and current were increased by 46.73% and 37.1%, respectively. What is more, this TENG exhibits a high photo responsivity (1.42 V mW −1 ) and rapid response (<150 ms), on both white light and monochromatic light covering from the near UV to near IR regions. By scrutinizing the mechanism, we discovered that the major reason behind this TENG output boosting upon light illumination is the generation of a large number of photoelectrons trapped on the surface of negative electrification layer due to the combined effect of superior absorption characteristics and surpassing charge trapping property of the α-Fe 2 O 3 nanoparticles, ensuing in an amplification in surface charge density. Additionally, the use of the processed human hair waste-based film as the positive tribo-layer, mitigates the challenge to waste management systems by extending the human hair waste utilization while securing the social and environmental welfare. This work manifests that the flexible, bio-based TENG is a very probable choice for the applications in self-powered and ultra-broadband light detection while generating significant socioeconomic benefits for people, by developing a promising approach towards the complete utilization systems for human hair waste.",
"introduction": "1. Introduction The power generation by harvesting mechanical [ 1 ], light [ 2 ], thermal [ 3 ] and magnetic [ 4 ] energy from ambient sources has attracted significant attention because of the advanced development of self-powered techniques, small portable electronics, and wireless sensing networks. Developed on the coupling of contact electrification and electrostatic induction, a triboelectric nanogenerator (TENG) can satisfactorily convert mechanical energy, in particular the energy in vibrations, wind, waves, and frictions produced by regular human motions, into electricity. In principle, any two dissimilar materials with evidently opposite triboelectric polarity are potentially implemented as the tribo-materials in TENG [ 5 ]. Because of the superior output power density, superior energy conversion efficiency, lower cost, simple device fabrication, and mechanically robustness numerous potential applications like, environmental surveillance, self-powered sensors (i.e., motion, pressure, and acoustic sensors), power sources for portable and wearable electronic devices have been successfully exhibited [ 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 ]. Although the basic mechanism behind the triboelectric charging phenomena has been considered to be well understood [ 15 ], the expeditious development of nanotechnologies along with the vast potential applications has notably increased the requirement to upgrade our apprehension of the mechanism under various realistic conditions [ 16 ]. Photo detection is of utmost importance because of various scientific applications like fiber optical communication systems, environmental monitoring, safety and for defense-related issues. On the other hand, the menace of global environmental pollution caused by the traditional fossil fuels and the continuous thrive of global demand for energy, the demand for the eco-friendly and self-powered photo detection over the conventional modules of photo detection is attracting world-wide attention [ 17 ]. In this aspect, the integration of triboelectric technology and photoelectric conversion mechanisms can be considered as an effective approach towards the execution of self-powered photo detectors with a simple, cost effective and maintenance free self-subsisting device structure. More importantly, it has been illustrated that both the photoconductivity and surface charge density of some semiconductor materials can be altered because of the superior photoelectric properties of those semiconductor materials by absorbing the solar energy [ 18 ]. Polymers are widely preferable materials as the primary dielectric for the fabrication of the self-powered and flexible wearable TENG devices, among them polydimethylsiloxane (PDMS) is appraised to be one of the best choices, owing to its high electronegativity, transparency, inherent elasticity, mixing ability with various nanostructured materials to prepare different composites, its capability of coating other surfaces, its hydrophobic surface and most importantly excellent biocompatibility [ 19 , 20 ]. Hematite nanoparticles (α-Fe 2 O 3 nanoparticles) are acknowledged as one of the widely applied photocatalysts due to their high activity, light stability, ready availability, low cost, environmental friendliness and corrosion resistance [ 20 , 21 , 22 ]. After the illumination of UV-visible light, free electrons are excited to the conduction band (CB), thereby leaving holes in the valence band (VB). Electron excitation by photon absorption will influence the possibility of the electron transfer in contact state, thus, the output performance of TENGs. This conception can be appraised for the self-powered photo-detection. On the other hand, several designs of triboelectric energy harvesters relying on recycling natural materials which are abundant in nature have been developed in recent years as eco-friendly solutions to the energy crisis [ 20 , 23 , 24 ]. Utilizing various materials easily available around us, recognized as waste material in most parts of the world, we can build eco-friendly, biocompatible, and highly efficient devices. In this study, we describe the hybridization between triboelectricity and photoelectric properties for simultaneous harvesting of the mechanical energy and solar energy based on a flexible and transparent TENG consist of PDMS/α-Fe 2 O 3 nanocomposite film (as the negative tribo-layer) and processed hair film (as the positive tribo-layer). Human hair, which is well known to be a powerful triboelectric material, are regarded as a slowly degrading waste material and its aggregation in waste streams causes environmental pollution. Thus the use of human hair waste eases a defiance to waste management systems and promotes the development of a propitious green energy source for the circular bio-economy. The as-fabricated bio-TENG exhibits excellent responsivity, rapid response time, appreciable repeatability, and broadband detection ability that extends over the near ultraviolet (UV), visible and near infrared (IR) regions. This work demonstrates a new way to developing high performance self-powered photodetectors, while developing waste hair utilization, that will play a significant part in circumventing climate emergencies.",
"discussion": "3. Results and Discussion 3.1. Structural and Compositional Analysis Figure 2 a exhibits the XRD patterns of α-Fe 2 O 3 nanoparticles, PDMS and the PDMS/α-Fe 2 O 3 nanocomposite film to analyze the crystal structures of the as-prepared samples. The XRD patterns of α-Fe 2 O 3 nanoparticles was consistent with the typical diffraction patterns (JCPDS card no. 33-0664) [ 25 ], and the non-existence of any further peak proved the emergence of the pure phase of hematite nanoparticles. As shown in the XRD pattern of PDMS/α-Fe 2 O 3 nanocomposite film, the amorphous halo at 2θ = 12° was the indication of the unaffected crystalline domain of the PDMS films [ 26 ], while the crystallinity was also preserved due to the insertion of α-Fe 2 O 3 nanoparticles. The XRD patterns affirmed the successful addition of pure α-Fe 2 O 3 nanoparticles into the pure PDMS to prepare the PDMS/α-Fe 2 O 3 nanocomposite with no alteration in the chemical composition. The FESEM image ( Figure 2 b) demonstrated the nanorod-like morphology of the as-synthesized α-Fe 2 O 3 nanoparticles, with 50 nm widths and 150–200 nm lengths. The role of morphological changes of the electrification layer in determining the TENG output performance was examined using FE-SEM and AFM images of pure PDMS and PDMS/α-Fe 2 O 3 nanocomposite surface ( Figure S2 and Figure 2 c,d). In contrast with the pure PDMS, the PDMS/α-Fe 2 O 3 nanocomposite film features special surface morphology owing to the insertion of α-Fe 2 O 3 nanoparticles into PDMS film. In the FESEM image of the PDMS/α-Fe 2 O 3 nanocomposite film ( Figure S2b ), one can see the presence of well-defined α-Fe 2 O 3 nanoparticles, evenly-distributed and properly embedded in PDMS matrix, therby improving the mechanical strength and surface roughness of the film [ 27 ]. The AFM analysis result demonstrates the root mean square (rms) roughness of the pure PDMS film to be 0.565 nm. Upon α-Fe 2 O 3 nanoparticles insertion, the roughness was elevated up to 0.826 nm ( Figure 2 c,d) and this higher surface roughness beneficially enlarged the surface area. The infrared transmittance spectra of functional groups were acquired from the pure PDMS, PDMS/α-Fe 2 O 3 nanocomposite and α-Fe 2 O 3 nanoparticles samples by using an FT-IR spectrometer accompanied with ATR microscopy accessory. As shown in Figure 2 e the main FTIR spectrum of PDMS between the pristine PDMS and PDMS composite sample do not show considerable difference except the presence of characteristic peaks at about 540 cm −1 and 432 cm −1 matched with the bonds in α-Fe 2 O 3 [ 28 ], that again indicates the successful formtion of PDMS/α-Fe 2 O 3 nanocomposite. The typical peaks at wavenumbers of 2965 cm −1 (C-H methyl stretch), 1260 cm −1 (CH 3 symmetric bending in silicon–methyl bond), identified the different existing functional groups in PDMS [ 29 ]. The wide polymer backbone absorption band found betwixt 1130 and 1000 cm −1 also characterises the PDMS [ 30 ]. The absorption band with lower intensity at 1410 cm −1 (for vinyl) and the absence of absorption band at 2140 cm −1 (for SiH), reveals the existence of unreacted vinyl groups in very small amount and no extra hydrosilane (SiH) groups [ 30 ]. Thus, FTIR analysis indicates that the insertion of α-Fe 2 O 3 nanoparticles into pure PDMS to prepare the PDMS/α-Fe 2 O 3 nanocomposite does not make any considerable transformation in the chemical bonds of PDMS essentially. Figure S1 shows the energy-dispersive X-ray spectroscopy (EDS) analysis results for the pure PDMS, pure α-Fe 2 O 3 nanoparticles and PDMS/α-Fe 2 O 3 nanocomposite. Both Fe and O are found in almost expected amount due to the integration of the α-Fe 2 O 3 nanoparticles into the pure PDMS, to form the PDMS/α-Fe 2 O 3 nanocomposite-based negative tribo-layer. IR spectroscopy for the processed hair film had been used to characterize the compounds and to recognize the purity of a material, where infrared radiation with a clear span of 4000 to 400 cm −1 was used to scan the sample. As represented in Figure 3 a,b a broad peak arised at around 3080–3557 cm −1 , possibly arose from the asymmetric and symmetric stretching modes of H−O−H and the stretching of N−H group [ 31 ]. The respective expansion of this peak for processed hair film in comparison with that of the untreated hair could be caused by an additional modification of its chemical structure due to the heat treatment at 60 °C [ 32 ]. The small peak at 2965 cm −1 originated from the asymmetric stretching mode of CH 3 . The three important amide bands attributed to the peptide bonds in untreated hair deveopes at 1700−1580 cm −1 (amide I); 1580−1500 cm −1 (amide II); and 1320−1210 cm −1 (amide III) [ 33 ]. Although the weak and the very strong absorption peaks detected at 1573 cm −1 and 1637 cm −1 , respectively, are not generated from the pure amide modes, yet preferably from amides that hydrolyzed into carboxylic acid groups after the reaction with strong bases, while the peak at around 1637 cm −1 probably resulted from the overlapping of N−H bending and C=C stretching modes [ 31 ]. The very weak peaks emerged at 1573 cm −1 and 1396 cm −1 possibly originated from the symmetric and antisymmetric stretching modes of −COO−, developed due to ionization of the carboxylic acid groups in the ethanolic NaOH. The weak peak appeared at 1433 cm −1 arose due to the vibration of the CH 2 group in scissoring mode. The strong peak found at 1331 cm −1 possibly originated from the S=O groups emerged by the oxidation of cysteine disulfide cross-links, exists in the untreated human hair [ 31 ]. The peak at approximately 1269 cm −1 corresponds to the N-H bending mode, while 1166 cm −1 represented the ester C–O asymmetric stretching mode [ 34 ]. The intense absorption at 1047 cm −1 was may be because of the appearance of the C−O stretching in the alcohol residue still existing in the processed hair film also because of the cysteic acid [ 31 ]. The peak at 877 cm −1 arose from the distortion of the hydrocarbons present in the keratin proteins, and the sharp peak at 670 cm −1 was designated to the C−OH out-of-plane bending mode [ 31 ]. Thus the FTIR study devulged some deformation in the chemical structure of natural untreated human hair due to the NaOH treatment at 60 °C. As shown in Figure S3a , the EDS analysis revealed the weight and atomic percentages of vital chemical elements present in processed hair film were similar to the regular amounts traced in natural untreated human hair [ 31 ]. The FE-SEM and AFM image in Figure S3b and Figure 3 c represents the surface morphology of the processed hair film. The as-prepared film evinces a nanoisland like surface morphology without any cracks or trace of undissolved hair with uniform coverage over the ITO electrode. The AFM analysis offered more insight into the surface roughness, and three-dimensional features along with the surface morphology of the as-prepared processed hair film, revealing that it is a compact film with the root mean square (rms) surface roughness of 4.44 nm ( Figure 3 c). 3.2. Electrical Performance As shown in Figure 4 , the performance of the as fabricated TENGs, operated in vertical contact-separation mode was signalized by measuring its output voltage and current under various light induced conditions. In dark state, the voltage and current signals are solely from the mechanical energy provided by the linear stepper motor during the periodic contact and separation cycle of the TENG device. As the α-Fe 2 O 3 nanoparticles were introduced into the PDMS to form the PDMS/α-Fe 2 O 3 nanocomposite, the magnitude of mean peak-to-peak output voltage was intensified from 37 to 367 V with the corresponding output current showing similar way of variation from 1.74 μA to 6.41 μA. This enhancement in output signals may be due to the combined effect of special surface morphology of the film with elevated surface roughness, effective contact area and the enriched charge trapping occurring from the insertion of high dielectric constant α-Fe 2 O 3 nanoparticles into the pure PDMS film, which assists to enhance charge transfer process during the triboelectrification mechanism, as explained thoroughly in our previous study, Reference [ 20 ]. Figure 4 a,b describes the change in voltage and current outputs of the TENGs during the switching on and off the white light (intensity—13.39 mW cm −2 ) repeatedly. The voltage and current amplitude (peak-to-peak value) was immediately intensified up to approximately 1.4 times, after the PDMS/α-Fe 2 O 3 nanocomposite based-TENG was exposed to the illumination and once the illumination was cut off, the voltage and current amplitudes recover quickly to the original value, although the bare PDMS based-TENG did not exhibit any further change in the output signals. These cycles of the periodic illumination indicate outstanding repeatability of the photo-enhancement. Since the electric signal produced by the PDMS/α-Fe 2 O 3 nanocomposite-based TENG is dynamic, it could be puzzling to calculate an accurate value of the response time. Nevertheless, it can be fairly evaluated. As shown in Figure 4 a, the voltage amplitude has a considerable step increment in the first cycle just after turning on the light and step reduction for turning off the light, indicating an immediate response to the illumination. This first step up and step down variation taking place within 150 ms. Consequently, the response time of this photo-induced TENG is less than 150 ms, which is supercilious to numerous recently reported self-powered photodetectors [ 18 ]. It was observed that the amplitude of the output voltage and current signals from the PDMS/α-Fe 2 O 3 nanocomposite-based TENG under illumination intensifies as the applied light intensity rises from 0.11 to 13.39 mW cm −2 ( Figure 5 ). However, it should be mentioned that the output signals suffered from a very tiny downswing when the intensity of light is over 13.39 mW cm −2 that is 17.79 mW cm −2 , this is likely attributed to a decay of PDMS/α-Fe 2 O 3 nanocomposite film owing to the high intensity light and a longtime measurement. This light intensity dependency of the output signals qualifies the TENG to be a realizable perspective for photo-detection. Moreover, the responsivity ( R ) can be interpreted using the following equation,\n (1) R = Δ V A Δ P \nwhere Δ V is the change in the amplitude of output voltage arising from the increase in light intensity, A is the sample area under illumination, and Δ P is the intensity of light [ 35 ]. The responsivity can be calculated by Equation (1), which originates a value of 1.42 V mW −1 , presenting a remarkable improvement over the recent TENG-based self-powered photodetectors. Further, the response of output signal from the as fabricated PDMS/α-Fe 2 O 3 nanocomposite-based TENG, on different wavelengths of lights was analyzed, as the ultra broad-band detection capability is one another major quality for an acceptable photodetector. As depicted in Figure 6 , the output signals manifest signification changes in their magnitudes under illumination (intensity-13.39 mW cm −2 ) over the near UV, visible and near IR regions, which specify an ultra broad-band detection capability. The pattern of variation of the output signal under illumination of different wavelengths ( Figure 6 c,d) conforms to the UV–Vis absorption of α-Fe 2 O 3 nanoparticles [ 36 ]. 3.3. Mechanism Discussion To interpret the triboelectric charge-transfer mechanism for the photo induced enhancement of the PDMS/α-Fe 2 O 3 nanocomposite-based TENG, the contact-separation process of TENG is split into two parts for detailed discussion. As represented in Figure S4a–e , without illumination, the TENG can simply convert vibrating energy into electrical energy under repeated external force, applied through the mover of linear stepper motor. In the initial stage, no charge occurs on the friction surface of the PDMS/α-Fe 2 O 3 nanocomposite film and processed hair film. Following the two friction layers come into physical contact, negative tribo-charges are assembled on the surface of PDMS/α-Fe 2 O 3 nanocomposite film, developing identical amount of tribo-positive charges on the processed hair film surface because of the difference in electron affinity. After that, as the TENG device started getting separated under the external applied force, the tribo-charges will build a time-varying electric field in space, as well will induce the positive and negative charges on the corresponding ITO electrodes respectively. In the course of this process, these inductive electrons will flow across the external circuit to balance the generated potential difference and attain electrostatic equilibrium eventually. Conversely, when these two frictional layers come back to the initial state from the top displaced position, the electrons will move at reversed direction till the complete contact, in this way generated an oscillating electric output signal. Under illumination condition, this TENG executes a double conversion mechanism as shown in Figure 7 a–f. Other than the benefaction from the triboelectricity, photoelectric conversion is one other source for this improvement in electric output. Under illumination condition also, following the conventional mechanism of triboelectric conversion by the coupling of contact electrification and electrostatic induction, once the electrification layers come into contact with each other, due to huge differences in the electron affinities between the PDMS/α-Fe 2 O 3 nanocomposite and human hair, surface charge transfer occurs, resulting in the development of opposite charges over the surface of those friction layers ( Figure 7 b). At this stage, there is no electric potential difference generated between the two back electrodes as with opposite charges coexist at the same plane [ 37 ]. Upon releasing the two parts of the device, a potential difference is then developed betwixt the two ITO electrodes as the charge accumulated on the contact induces opposite charges on the back electrodes (ITO electrodes), generating current flowing through the external circuit from the upper part to the lower part of the TENG device ( Figure 7 c). Immediately after the application of illumination, a lot of electron-hole pairs are produced within the PDMS/α-Fe 2 O 3 nanocomposite layer, resulted from photon absorption, owing to the superior absorption characteristics of the α-Fe 2 O 3 nanoparticles. As shown in the Figure 7 g, after the insertion of α-Fe 2 O 3 nanoparticles into the PDMS matrix, distinct peaks appeared due to the α-Fe 2 O 3 nanoparticles absorption characteristics over the near UV, visible and near IR regions, while the bare PDMS exhibited no considerable absorption peaks except an absorption peak at approximately 220 nm in the middle-UV region [ 38 ]. As demonstrated in Figure 7 f, these photo generated electrons get quickly captured by the α-Fe 2 O 3 nanoparticles, embedded in the negative tribo-layer, that is recognized for its surpassing charge trapping property [ 39 ], confirmed by the appreciable enhancement in the dielectric properties of the PDMS based-composite owing to the insertion of α-Fe 2 O 3 nanoparticles into the pure PDMS ( Figure 7 h), while the generated holes then tend to get neutralized by the recombination with adsorbing negatively charged particles or ions from air [ 40 ]. Considering comparatively lower dielectric constant of the as prepared PDMS film ( Figure 7 h), as the general distinction of dielectric property is noticed because of the changes of space charge/interfacial polarization developing from differences in the surface to volume ratio of the presented structure morphology [ 20 , 41 , 42 , 43 ]. Hence, these trapped electrons successively, are amplified to the previously-available negative tribo-charges, resulting in an enhancement in the surface charge density of PDMS/α-Fe 2 O 3 nanocomposite film, thereby amplifying the output performance of this TENGs upon light illumination. Afterwards, the two tribo-layers are fully released ( Figure 7 d), the output signals fall to zero. When the TENG is pressed again ( Figure 7 e), electrons are pushed back from the top to the bottom current collector, since the electrostatic potential difference is recompensed for by the opposite charges on both tribo-layers, till the induced charges get neutralized. Therefore, it has been observed that the as-fabricated TENG exhibits fast response on both white light and monochromatic light with appreciable differences in the output signal responses under different intensities and wavelengths of light covering from the near UV to near IR regions, hence providing a novel methodology towards self-powered photodetection. The processed hair film based positive tribo-layer manifests the TENG output performance considerably alike to that of the untreated natural hair waste, as explained minutely in our previous study, Reference [ 20 ]. Therefore, it is anticipated that the as prepared PDMS/α-Fe 2 O 3 nanocomposite film will be capable of harvesting human hair movement reliably, in consequence of spontaneous gesture, in that way meeting users’ different functional necessity and executing the aforesaid broadband photodetection during any activities under the sun with this PDMS/α-Fe 2 O 3 nanocomposite film fixed inside some transparent hair accessories, for example a hair clip, hairband, hair cap and head cap, even also helmet.\n\n3.3. Mechanism Discussion To interpret the triboelectric charge-transfer mechanism for the photo induced enhancement of the PDMS/α-Fe 2 O 3 nanocomposite-based TENG, the contact-separation process of TENG is split into two parts for detailed discussion. As represented in Figure S4a–e , without illumination, the TENG can simply convert vibrating energy into electrical energy under repeated external force, applied through the mover of linear stepper motor. In the initial stage, no charge occurs on the friction surface of the PDMS/α-Fe 2 O 3 nanocomposite film and processed hair film. Following the two friction layers come into physical contact, negative tribo-charges are assembled on the surface of PDMS/α-Fe 2 O 3 nanocomposite film, developing identical amount of tribo-positive charges on the processed hair film surface because of the difference in electron affinity. After that, as the TENG device started getting separated under the external applied force, the tribo-charges will build a time-varying electric field in space, as well will induce the positive and negative charges on the corresponding ITO electrodes respectively. In the course of this process, these inductive electrons will flow across the external circuit to balance the generated potential difference and attain electrostatic equilibrium eventually. Conversely, when these two frictional layers come back to the initial state from the top displaced position, the electrons will move at reversed direction till the complete contact, in this way generated an oscillating electric output signal. Under illumination condition, this TENG executes a double conversion mechanism as shown in Figure 7 a–f. Other than the benefaction from the triboelectricity, photoelectric conversion is one other source for this improvement in electric output. Under illumination condition also, following the conventional mechanism of triboelectric conversion by the coupling of contact electrification and electrostatic induction, once the electrification layers come into contact with each other, due to huge differences in the electron affinities between the PDMS/α-Fe 2 O 3 nanocomposite and human hair, surface charge transfer occurs, resulting in the development of opposite charges over the surface of those friction layers ( Figure 7 b). At this stage, there is no electric potential difference generated between the two back electrodes as with opposite charges coexist at the same plane [ 37 ]. Upon releasing the two parts of the device, a potential difference is then developed betwixt the two ITO electrodes as the charge accumulated on the contact induces opposite charges on the back electrodes (ITO electrodes), generating current flowing through the external circuit from the upper part to the lower part of the TENG device ( Figure 7 c). Immediately after the application of illumination, a lot of electron-hole pairs are produced within the PDMS/α-Fe 2 O 3 nanocomposite layer, resulted from photon absorption, owing to the superior absorption characteristics of the α-Fe 2 O 3 nanoparticles. As shown in the Figure 7 g, after the insertion of α-Fe 2 O 3 nanoparticles into the PDMS matrix, distinct peaks appeared due to the α-Fe 2 O 3 nanoparticles absorption characteristics over the near UV, visible and near IR regions, while the bare PDMS exhibited no considerable absorption peaks except an absorption peak at approximately 220 nm in the middle-UV region [ 38 ]. As demonstrated in Figure 7 f, these photo generated electrons get quickly captured by the α-Fe 2 O 3 nanoparticles, embedded in the negative tribo-layer, that is recognized for its surpassing charge trapping property [ 39 ], confirmed by the appreciable enhancement in the dielectric properties of the PDMS based-composite owing to the insertion of α-Fe 2 O 3 nanoparticles into the pure PDMS ( Figure 7 h), while the generated holes then tend to get neutralized by the recombination with adsorbing negatively charged particles or ions from air [ 40 ]. Considering comparatively lower dielectric constant of the as prepared PDMS film ( Figure 7 h), as the general distinction of dielectric property is noticed because of the changes of space charge/interfacial polarization developing from differences in the surface to volume ratio of the presented structure morphology [ 20 , 41 , 42 , 43 ]. Hence, these trapped electrons successively, are amplified to the previously-available negative tribo-charges, resulting in an enhancement in the surface charge density of PDMS/α-Fe 2 O 3 nanocomposite film, thereby amplifying the output performance of this TENGs upon light illumination. Afterwards, the two tribo-layers are fully released ( Figure 7 d), the output signals fall to zero. When the TENG is pressed again ( Figure 7 e), electrons are pushed back from the top to the bottom current collector, since the electrostatic potential difference is recompensed for by the opposite charges on both tribo-layers, till the induced charges get neutralized. Therefore, it has been observed that the as-fabricated TENG exhibits fast response on both white light and monochromatic light with appreciable differences in the output signal responses under different intensities and wavelengths of light covering from the near UV to near IR regions, hence providing a novel methodology towards self-powered photodetection. The processed hair film based positive tribo-layer manifests the TENG output performance considerably alike to that of the untreated natural hair waste, as explained minutely in our previous study, Reference [ 20 ]. Therefore, it is anticipated that the as prepared PDMS/α-Fe 2 O 3 nanocomposite film will be capable of harvesting human hair movement reliably, in consequence of spontaneous gesture, in that way meeting users’ different functional necessity and executing the aforesaid broadband photodetection during any activities under the sun with this PDMS/α-Fe 2 O 3 nanocomposite film fixed inside some transparent hair accessories, for example a hair clip, hairband, hair cap and head cap, even also helmet."
} | 8,007 |
38551595 | PMC11008094 | pmc | 2,338 | {
"abstract": "Methane emissions present a significant environmental\nchallenge\nin both natural and engineered aquatic environments. Denitrifying\nanaerobic methane oxidation (N-DAMO) has the potential for application\nin wastewater treatment plants. However, our understanding of the\nN-DAMO process is primarily based on studies conducted on environmental\nsamples or enrichment cultures using metagenomic approaches. To gain\ndeeper insights into N-DAMO, we used antimicrobial compounds to study\nthe function and physiology of ‘ Candidatus Methanoperedens nitroreducens’ and ‘ Candidatus Methylomirabilis oxyfera’ in N-DAMO\nenrichment cultures. We explored the effects of inhibitors and antibiotics\nand investigated the potential application of N-DAMO in wastewater\ncontaminated with ammonium and heavy metals. Our results showed that\n‘ Ca. M. nitroreducens’\nwas susceptible to puromycin and 2-bromoethanesulfonate, while the\nnovel methanogen inhibitor 3-nitrooxypropanol had no effect on N-DAMO.\nFurthermore, ‘ Ca. M. oxyfera’\nwas shown to be susceptible to the particulate methane monooxygenase\ninhibitor 1,7-octadiyne and a bacteria-suppressing antibiotic cocktail.\nThe N-DAMO activity was not affected by ammonium concentrations below\n10 mM. Finally, the N-DAMO community appeared to be remarkably resistant\nto lead (Pb) but susceptible to nickel (Ni) and cadmium (Cd). This\nstudy provides insights into microbial functions in N-DAMO communities,\nfacilitating further investigation of their application in methanogenic,\nnitrogen-polluted water systems.",
"introduction": "Introduction Methane, the second most prevalent greenhouse\ngas globally after\nCO 2 , poses a serious environmental challenge as a driver\nof climate warming. 1 Anaerobic oxidation\nof methane (AOM) has been found to be an important sink of methane\nby preventing its emission into the atmosphere. 2 AOM is mediated by a diverse group of microorganisms, including\nbacteria from the NC10 phylum (Methylomirabilota) and a polyphyletic\ngroup of archaea known as ANME. 3 These\nanaerobic methanotrophs are present in natural and engineered water\nsystems, where they couple AOM to various electron acceptors such\nas nitrate, nitrite, sulfate (in syntrophy with sulfate-reducing bacteria),\nmetal oxides, and humic substances. 4 − 10 Despite the increasing number of studies on AOM in recent years,\nprimarily focusing on omics data, detailed physiological studies of\nanaerobic methanotrophs remain scarce due to the absence of pure cultures. In freshwater systems, anaerobic methane-oxidizing microorganisms\nof the genus ‘ Candidatus Methanoperedens’\nand ‘ Candidatus Methylomirabilis’\nare frequently observed. 3 ‘ Ca. Methanoperedens nitroreducens’ is an\nANME2d archaeon that utilizes a reverse methanogenesis pathway for\nmethane oxidation. 5 Methane is activated\nby the enzyme methyl-coenzyme M reductase (MCR) and further converted\nto CO 2 while reducing nitrate to nitrite. 11 On the other hand, ‘ Ca. Methylomirabilis oxyfera’ employs an intra-aerobic pathway\nreducing the nitrite that is toxic to ‘ Ca. M. nitroreducens’ to nitric oxide, followed by dismutation\nof nitric oxide to nitrogen gas and oxygen. 10 , 12 , 13 Subsequently, this oxygen is directly used\nby its particulate methane monooxygenase (pMMO) to activate methane. 10 , 12 , 13 The cooperative activities of\n‘ Ca. M. nitroreducens’\nand ‘ Ca. M. oxyfera’\nestablish an efficient denitrifying anaerobic methane oxidation (N-DAMO)\nsystem. 14 Because of the efficient\nmethane and nitrogen removal, N-DAMO holds\npotential for application as a bioremediation strategy for methanogenic\nnitrate-polluted water. 2 , 14 − 16 Consequently,\nintroducing N-DAMO into wastewater treatment plants (WWTPs) has been\nproposed as a means of methane and nitrogen removal. 2 , 14 , 15 However, in many freshwater ecosystems\nand WWTPs, concentrations of heavy metals such as lead, nickel, and\ncadmium are high ( Table S1 ). Yet, the effect\nof these toxic compounds on N-DAMO efficiency is still unknown. Therefore, in this study, we used various antimicrobial compounds\n(antibiotics and inhibitors) to study anaerobic methane oxidation\nin a complex N-DAMO community. These antimicrobial compounds may be\nused in the future to ultimately enrich N-DAMO microorganisms by removing\nflanking nonmethanotrophic community members. At the same time, we\nevaluated the biotechnological application of N-DAMO by studying the\neffect of organic solvents, ammonium, and various heavy metals commonly\npresent in wastewaters on the N-DAMO efficiency. To address these\nknowledge gaps, the effect of different antimicrobial compounds on\nmethane oxidation and nitrate reduction was tested in batch incubations\nusing an N-DAMO enrichment culture. It appeared that ‘ Ca. M. nitroreducens’ was inhibited by 2-bromoethanesulfonate\n(2-BES) and puromycin, while ammonium below 10 mM and lead (Pb) had\nlittle effect.",
"discussion": "Results and Discussion N-DAMO Community Is Sensitive toward Ethanol and DMSO In the first part of this study, the complex interactions between\nanaerobic methanotrophs ‘ Ca. M. nitroreducens’ (19% abundance in metagenome) and ‘ Ca. M. oxyfera’ (28% abundance in metagenome)\nas well as the side community ( Figure S3 ) were investigated using a series of batch experiments with various\nantibiotics and inhibitors ( Table 1 ). Testing inhibitors and antibiotics can be challenging\nin the case of moderately or poorly soluble compounds. To administer\nthese compounds, a stock solution in organic solvents such as dimethyl\nsulfoxide (DMSO), ethanol, or methanol is often prepared. Therefore,\nwe first examined the effect of these organic solvents on the AOM\nactivity. These experiments revealed that the addition of 1% (v/v)\nDMSO and 0.1% (v/v) ethanol (equivalent to 17 mM) had a significant\nimpact on AOM, resulting in >70% inhibition of AOM ( Figure 1 ). This was a surprising finding\nconsidering that DSMO is generally accepted as nontoxic below 10%\n(v/v) and biological effects of applied concentrations below 2% (v/v)\nare often unreported as assumed negligible. 18 For comparison, a study on soil bacterial communities found only\nminor effects of DMSO in concentrations up to 5% (v/v) with a reduction\nin the growth rate of up to 20%. 19 Ethanol,\neven at low concentrations (0.63 mM), showed an almost complete inhibition\nof AOM (96 ± 1%) that did not recover within 96 h, which should\nbe sufficient time for the ethanol to be consumed by the side community.\nIncubations with aerobic methanotrophs have reported similar inhibitory\neffects after addition of 0.5 mM ethanol, although the exact mechanism\nremains unclear. 20 In contrast, methanol\nat a concentration of 0.63 mM did not exert a significant influence\non the AOM rate ( Figure 1 ). The nitrate reduction rate was also significantly decreased by\nDMSO (52 ± 1% inhibition), while in the case of methanol and\nethanol, there was no significant decrease. Nitrate reduction under\nthese conditions can probably be attributed to other nonmethanotrophic\ncommunity members oxidizing these alcohols ( Table S2 ). All in all, these findings highlight the specific sensitivity\nof the AOM community to DMSO and ethanol, for which the mechanisms\nremain to be investigated. For this reason, ethanol and DMSO were\navoided as solvents in antibiotic assays in this study and only water-soluble\ncompounds were tested. Figure 1 Impact of antimicrobial compounds and organic solvents\non the AOM\nrate. The relative activity of AOM was calculated by dividing the\nAOM rate (μmol CH4 day –1 g dw –1 ) of incubations supplemented with antimicrobial\ncompounds or solvents by the positive control (with biomass, methane,\nand nitrate), accounting for the methane loss during sampling in the\nnegative control (with only medium and methane). The methane concentration\nwas measured after 4 h of incubation and measured twice daily over\nthe course of 96 h. Results are obtained from biological triplicates\n(error bars represent standard deviation). Asterisks indicate a significant\ndifference ( p < 0.05) compared to the positive\ncontrol (two-tailed, heteroscedastic t test). Identification of Effective Antibiotics against Methane-Oxidizing\nMicroorganisms To assess the contribution of ‘ Ca. M. nitroreducens’ to methane oxidation\nand nitrate reduction, antibiotics known to be effective against archaea\nsuch as neomycin, bacitracin, and puromycin were tested. We chose\nantibiotics known to affect methanogens, because of the close phylogenetic\nrelationship and similarities in metabolism between ANME archaea and\nmethanogens. 21 Neomycin and bacitracin\nin standard concentrations of 50 μg mL –1 did\nnot show a significant effect on methane oxidation and nitrate reduction\ncompared to the positive control ( Figure 1 and Table S2 ).\nThis is in contrast to the methanogenic archaeon Methanosarcina\nmazei , which has been shown to be susceptible to 20\nμg mL –1 neomycin, 22 and Methanobrevibacter smithii , which\nis inhibited by 4 μg mL –1 bacitracin. 23 These findings suggest that the N-DAMO community\nin this study is resistant to neomycin and bacitracin at the concentrations\ntested. Puromycin, another antibiotic that is known to affect\nmethanogenic archaea, however, showed significant inhibition at a\nconcentration of 10 μg mL –1 with an AOM rate\nof 42 ± 0.3% (mean ± standard deviation) lower compared\nto the positive control ( Figure 1 ). Higher concentrations of puromycin (50 and 75 μg\nmL –1 ) led to further inhibition to 73 ± 3 and\n65 ± 4%, respectively. Together, this indicates that the maximum\ninhibitory effect of puromycin on AOM is ∼70%, which is already\nreached at a concentration of 50 μg mL –1 .\nSince puromycin is effective against archaea and Gram-positive bacteria, 24 this indicates that the remaining AOM activity\ncan likely be attributed to Gram-negative ‘ Ca. M. oxyfera’. Nitrate reduction was inhibited to a lesser\nextent (up to 42 ± 12%, Table S2 ),\nwhich might be facilitated by ‘ Ca. M. oxyfera’ using storage molecules such as glycogen or by\nother nonmethanotrophic community members consuming dead biomass.\nWe estimated the concentration of puromycin required to inhibit 50%\nof the AOM activity (IC 50 ) to be <10 μg mL –1 , taking into account that the maximum effect of puromycin\ndoes not completely abolish AOM ( Table 2 ). Ultimately, puromycin sensitivity seems comparable\nto that of the methanogen M. mazei ,\nwhich is completely inhibited at 5 μg mL –1 . 22 Table 2 IC 50 Values for AOM-Affecting\nCompounds Puromycin, Ammonium, and Heavy Metals IC 50 puromycin <10 μg mL –1 ammonium 52 mM lead >1 mM nickel 0.23 mM cadmium <10 μM To explore the microbial interactions between ‘ Ca. M. nitroreducens’ and the flanking bacterial\ncommunity, a bacteria-suppressing antibiotic cocktail containing streptomycin,\nvancomycin, ampicillin, and kanamycin was used. This cocktail targets\ncell wall synthesis and translation in both Gram-positive and Gram-negative\nbacteria potentially inhibiting most of the bacteria in the community. 25 Batch incubations conducted with this bacteria-suppressing\nantibiotic cocktail demonstrated a decrease in AOM and nitrate reduction\nrates of 69 ± 3 and 42 ± 6%, respectively ( Figure 1 and Table S2 ). Since the AOM rate was not completely abolished, this\nindicates that ‘ Ca. M. nitroreducens’\nis resilient against this bacteria-suppressing antibiotic cocktail\nand is responsible for the remaining AOM activity. Interestingly,\nnitrite did not accumulate, which hints toward the use of a possible\nnitrite detoxification system such as dissimilatory nitrite reduction\nto ammonium by ‘ Ca. M. nitroreducens’. 26 Therefore, the use of this bacteria-suppressing\nantibiotic cocktail could be an effective strategy to further enrich\nand isolate ‘ Ca. Methanoperedenaceae’\nin the future. Methanogen Inhibitors 2-BES and 3-BPS Seem to Inhibit ‘ Ca. M. nitroreducens’, while 3-NOP Is Modified\nor Degraded A commonly employed strategy to inhibit methanogens\nis the addition of MCR inhibitors to the growth medium, of which 2-BES\nis the most frequently used compound. In this study, we aimed to assess\nthe effectiveness of methanogen inhibitors on MCR-containing ‘ Ca. M. nitroreducens’. Activity assays revealed\nthat 2-BES and its structural analogue 3-bromopropanesulfonate (3-BPS)\nat 20 mM concentrations similarly inhibited the AOM rate by 75 ±\n2 and 68 ± 4%, respectively ( Figure 1 ). Similar to the antibiotic conditions,\nthe nitrate reduction rate was affected less with 41 ± 3 and\n14 ± 0% inhibition, respectively ( Table S2 ). The remaining AOM activity and nitrate reduction can likely be\nattributed to ‘ Ca. M. oxyfera’\nand the side community. In the literature, conflicting observations\nregarding the effectiveness of 2-BES on ANMEs have been reported.\nA recent study of ‘ Ca. M. nitroreducens’\nin bioelectrochemical systems revealed a similar inhibition by 20\nmM 2-BES as described in this study. 27 Furthermore,\nmethane oxidation in a culture containing ANME1 and ANME2 was inhibited\nby 1 mM 2-BES, 28 while others report that\nconcentrations as high as 50 mM did not show measurable effects on\nAOM rates, including cultures where ‘ Ca. M. nitroreducens’ was the dominant species. 5 , 29 , 30 This indicates that the susceptibility\nof ANME to 2-BES highly depends on the microbial community and/or\nstrain. A third MCR inhibitor tested was 3-NOP, which has recently\nemerged as a compound capable of inhibiting MCR at concentrations\nmuch lower than 2-BES and 3-BPS. 31 Notably,\n3-NOP has been formulated by Royal DSM under the name Bovaer as a\nfeed additive for ruminants, demonstrating a 30% reduction in biogenic\nmethane emissions by ruminants. 32 Interestingly,\nour findings indicate that 3-NOP does not significantly affect AOM\nat concentrations 20 times higher (200 μM) than required for\nmethanogens (10 μM). 31 These results\nsuggest that while 3-NOP demonstrates promising efficacy in mitigating\nmethane emissions in ruminants, it is not able to inhibit anaerobic\nmethanotrophic archaea at the concentrations tested in the N-DAMO\ncommunity here. Whether this is an intrinsic feature of ‘ Ca. M. nitroreducens’ or a resistance inferred\nby the flanking community remains unclear. A possible resistance\nmechanism to 3-NOP includes the degradation\nor modification of 3-NOP. To explore this possibility, we conducted\nexperiments where we added the culture supernatant from a batch incubation\nwith the N-DAMO culture and 200 μM 3-NOP to the medium of the\nmethanogen M. formicicum . This way,\nif 3-NOP was degraded by the N-DAMO community, the supernatant would\nnot affect the growth of M. formicicum . In a control with the direct addition of 3-NOP to M. formicicum , methane production was completely\ninhibited ( Figure 2 ). Similarly, the supernatant from an abiotic control with dead N-DAMO\nbiomass and 3-NOP completely inhibited M. formicicum , indicating that 3-NOP remained chemically stable throughout the\nduration of the experiment. Furthermore, the addition of the supernatant\nfrom an N-DAMO culture without 3-NOP did not impact methane production\nby M. formicicum . Finally, the addition\nof the supernatant from an N-DAMO culture with 3-NOP did not affect\nmethane production by M. formicicum , implying that the N-DAMO community is capable of degrading or modifying\n3-NOP, rendering MCR inhibition ineffective. Further studies would\nbe required to identify degradation or modification mechanisms of\n3-NOP. Figure 2 Assessment of degradation or modification of 3-NOP by addition\nof the N-DAMO community supernatant to M. formicicum . 3-NOP was directly added to M. formicicum as well as the supernatant obtained from N-DAMO batch cultures which\nincluded a control N-DAMO incubation, an abiotic N-DAMO incubation\nwith fixed biomass, and an N-DAMO incubation with 3-NOP. Subsequently,\ncumulative methane production by M. formicicum was monitored for 96 h. Results are averaged from biological triplicates\n(error bars represent the standard deviation). 1,7-Octadiyne Seems to Inhibit Methane Oxidation by ‘ Ca. M. oxyfera’ While MCR inhibitors\ntarget ‘ Ca. M. nitroreducens’,\nthe first step in methane oxidation by ‘ Ca. M. oxyfera’ relies on the activity of a pMMO. To specifically\ninvestigate the role of ‘ Ca. M. oxyfera’ in the N-DAMO community, we examined the effect\nof the pMMO inhibitor 1,7-octadiyne (1,7-OD). Batch incubations supplemented\nwith 100 μM 1,7-OD resulted in a significant decrease in the\nAOM rate of 64 ± 23% accompanied by a decrease in the nitrate\nreduction rate of 45 ± 7% ( Figure 1 and Table S2 ). This effect\nwas comparable to the incubation with streptomycin, vancomycin, ampicillin,\nand kanamycin, and also in this case, nitrite did not accumulate.\nOur findings indicate that 1,7-OD effectively inhibits ‘ Ca. M. oxyfera’, with results comparable to\nexperiments done with aerobic methanotrophs. 33 Syntrophic Relationships Boost the Methane Oxidation Rate Tested inhibitors exhibited partial effects on methane oxidation,\nwithout completely abolishing AOM showing that both ‘ Ca. M. nitroreducens’ and ‘ Ca. M. oxyfera’ are important contributors\nto AOM in the tested cultures. Compounds targeting archaea, such as\npuromycin, 2-BES, and 3-BPS indeed seemed to inhibit ‘ Ca. M. nitroreducens’ and reduced the AOM\nrate by approximately 70%. Similarly, the bacteria-suppressing antibiotic\ncocktail containing streptomycin, vancomycin, ampicillin, and kanamycin,\nas well as aerobic methanotroph inhibitor 1,7-OD resulted in a comparable\nreduction of the AOM rate by approximately 70%. To further investigate\narchaeal and bacterial contributions to AOM, a combination of 2-BES\nand the bacteria-suppressing antibiotic cocktail was added to a batch\nculture to inhibit both methanotrophic bacteria and archaea. The addition\nof 2-BES-targeting ‘ Ca. M. nitroreducens’,\nwhile the bacteria-suppressing antibiotic cocktail targeted ‘ Ca. M. oxyfera’, resulted in the complete\ncessation of methane oxidation ( Figure 1 ). It seems that individually ‘ Ca. M. nitroreducens’ and ‘ Ca. M. oxyfera’ can reach an AOM rate of about\n30% of the total rate that can be achieved in a syntrophic relationship.\nThese findings demonstrate that unlike sulfate-dependent AOM where\nANME form an interdependent syntrophic relationship with sulfate-reducing\nbacteria, 34 ‘ Ca. M. nitroreducens’ and ‘ Ca. M. oxyfera’ can oxidize methane independently, but rates\nare greatly enhanced through a syntrophic relationship. Response of N-DAMO Community to Various Concentrations of Ammonium\nand Heavy Metal - Implications for Wastewater Treatment To\nexplore potential applications of N-DAMO in (waste)water treatment,\nwe investigated the tolerance of the N-DAMO community toward pollutants\nsuch as ammonium and heavy metals. Of particular interest were lead\n(Pb), nickel (Ni), and cadmium (Cd), as these toxic metals are commonly\nfound in wastewaters, often at high concentrations ( Table S1 ). Among the tested heavy metals, the N-DAMO\ncommunity demonstrated the highest tolerance to Pb ( Figure 3 ). Lead is generally considered\ntoxic to bacteria, as Pb ions can interfere with essential cellular\nprocesses and disrupt enzymatic activity. However, some bacteria have\ndeveloped mechanisms to tolerate or adapt to Pb exposure. Certain\nbacterial species may possess resistance mechanisms, such as efflux\npumps or enzymatic detoxification systems, which allow them to survive\nin the presence of Pb. The N-DAMO community appeared to be resistant\nand even benefited from a low concentration of Pb (10 μM), which\nsignificantly increased the AOM rate by 38 ± 7%. Concentrations\nof 100 and 500 μM Pb did not exert a significant impact on the\nAOM rate, while toxicity effects became apparent at 1000 μM\nPb. These observations indicate that the IC 50 value for\nPb exceeds 1000 μM, suggesting a high tolerance to Pb by the\nN-DAMO community ( Table 2 ). Moreover, the nitrate reduction rates did not decrease at any\nof the Pb concentrations ( Table S2 ). The\nsignificant increase of the AOM rate with 10 μM Pb remains enigmatic\nbut might be caused by introducing a selective advantage for ‘ Ca. M. nitroreducens’ and ‘ Ca. M. oxyfera’ over other nonmethanotrophic\ncommunity members that are likely less resistant to Pb. These data\nsuggest that N-DAMO could be applied in some of the most contaminated\nenvironments containing Pb with concentrations up to 336 μM\n( Table S1 ). Figure 3 Impact of ammonium, Pb,\nNi, and Cd on anaerobic methane oxidation.\nThe relative activity of AOM was calculated by dividing the AOM rate\n(μmol CH4 day –1 g dw –1 ) of incubation supplemented with ammonium or heavy\nmetals by the positive control (with biomass, methane, and nitrate)\naccounting for the methane loss during sampling in the negative control\n(with only medium and methane). The methane concentration was measured\nafter 4 h of incubation and measured twice daily over the course of\n96 h. Results are obtained from biological triplicates (error bars\nrepresent standard deviation). Asterisks indicate a significant difference\n( p < 0.05) compared to the positive control (two-tailed,\nheteroscedastic t test). The effects of Ni and Cd on the N-DAMO community\nwere much more\npronounced. In the case of Ni, a concentration of 10 μM did\nnot have a significant effect on the AOM rate. However, at increasing\nconcentrations of 100, 500, and 1000 μM, clear signs of toxicity\nwere observed, resulting in an IC 50 value of 0.23 mM ( Figure 3 and Table 2 ). Although clear signs of toxicity\nwere observed based on the AOM rate, nitrate reduction was affected\nmuch less, which could be attributed to more resistant nonmethanotrophic\ncommunity members ( Table S2 ). Nickel is\nused as a cofactor by several microbial enzymes such as urease, hydrogenase,\nand MCR. 35 Therefore, it is not surprising\nthat low concentrations of Ni do not have an inhibitory effect on\nN-DAMO, as this metal is in fact required by some community members,\nincluding methanotrophic archaea. However, at high concentrations,\nNi becomes toxic to microorganisms as it was shown to be directly\ninvolved in the inhibition of enzymes, 35 which was also the case in our experiment. In contrast, Cd\nexhibited pronounced toxicity already at a concentration\nof 10 μM, completely abolishing methane oxidation at 500 μM.\nThese findings indicate an IC 50 value for Cd below 10 μM\n( Figure 3 and Table 2 ). Similar to Ni,\nnitrate reduction was affected to a lesser extent ( Table S2 ). Cd is a toxic heavy metal with no known biological\nfunction. Besides being an enzyme inhibitor, it can have deleterious\neffects on the membrane structure and function by binding to ligands\nsuch as phosphate and the cysteinyl and histidyl groups of proteins. 36 , 37 With concentrations of Ni and Cd being as high as 424 and 8 μM\nin some wastewaters, respectively ( Table S1 ), the application and efficiency of the N-DAMO process might be\nhindered. Similarly, ammonium is a common water contaminant,\nparticularly\nin agricultural areas. Ammonium was introduced to the batch incubation\nin concentrations ranging from 1 to 100 mM ( Figure 3 ). The addition of 1 mM ammonium resulted\nin a marginal increase in the AOM rate, which was not statistically\nsignificant. However, a significant decrease in the AOM activity was\nobserved at concentrations of 20 and 100 mM, while the nitrate reduction\nrate was only affected at 100 mM ammonium ( Table S2 ). It was calculated that the AOM IC 50 value for\nammonium was 52 mM ( Table 2 ). The N-DAMO community appears to be much more sensitive\nto ammonium compared to methanogenic archaea such as Methanospirillum hungatei and Methanosarcina\nbarkeri that were shown to resist up to 200 mM NH 4 + 38 and anaerobic digester\ncommunities that can tolerate up to ∼150 mM NH 4 + . 39 However, ammonium concentrations\nin wastewater are typically around 2 mM, 40 indicating that the presence of ammonium should not be a problem\nfor application of N-DAMO in wastewater treatment. Using combinations\nof the antimicrobial compounds, puromycin, a\ncocktail of bacteria-suppressing antibiotics, 2-BES, 3-BPS, and 1,7-OD\nallowed us to study and understand the function of N-DAMO community\nmembers regarding methane oxidation and nitrate reduction. These antimicrobial\ncompounds can be used as diagnostic tools to study the AOM in environmental\nsamples. Ultimately, the prolonged use of such compounds might lead\nto the first pure culture isolation of ‘ Ca. M. nitroreducens’ or ‘ Ca. M. oxyfera’. Many studies have suggested the application\nof N-DAMO in WWTPs.\nFor this, ammonium tolerance is relevant, especially in setups where\nN-DAMO is combined with anammox. Our study shows that N-DAMO is not\naffected by low concentrations of ammonium, therefore supporting the\npossibility of combining N-DAMO with anammox in WWTPs. Moreover, wastewater\nis often enriched in heavy metals, making it essential to know their\neffect on N-DAMO. Here, we showed that the N-DAMO community is resistant\nto Pb up to 1000 μM, but less toward Ni and Cd, although lower\nconcentration of Cd (≤8 μM, Table S1 ) should be tested to pinpoint Cd thresholds for N-DAMO in\nwastewater treatment. Ultimately, these results are a step forward\ntoward understanding the application potential of N-DAMO in wastewater\ntreatment."
} | 6,333 |
32982673 | PMC7487417 | pmc | 2,339 | {
"abstract": "Simulation of large scale biologically plausible spiking neural networks, e.g., Bayesian Confidence Propagation Neural Network (BCPNN), usually requires high-performance supercomputers with dedicated accelerators, such as GPUs, FPGAs, or even Application-Specific Integrated Circuits (ASICs). Almost all of these computers are based on the von Neumann architecture that separates storage and computation. In all these solutions, memory access is the dominant cost even for highly customized computation and memory architecture, such as ASICs. In this paper, we propose an optimization technique that can make the BCPNN simulation memory access friendly by avoiding a dual-access pattern. The BCPNN synaptic traces and weights are organized as matrices accessed both row-wise and column-wise. Accessing data stored in DRAM with a dual-access pattern is extremely expensive. A post-synaptic history buffer and an approximation function thus are introduced to eliminate the troublesome column update. The error analysis combining theoretical analysis and experiments suggests that the probability of introducing intolerable errors by such optimization can be bounded to a very small number, which makes it almost negligible. Derivation and validation of such a bound is the core contribution of this paper. Experiments on a GPU platform shows that compared to the previously reported baseline simulation strategy, the proposed optimization technique reduces the storage requirement by 33%, the global memory access demand by more than 27% and DRAM access rate by more than 5%; the latency of updating synaptic traces decreases by roughly 50%. Compared with the other similar optimization technique reported in the literature, our method clearly shows considerably better results. Although the BCPNN is used as the targeted neural network model, the proposed optimization method can be applied to other artificial neural network models based on a Hebbian learning rule.",
"introduction": "1. Introduction Bayesian Confidence Propagation Neural Networks (BCPNNs), proposed by Lansner and Ekeberg ( 1989 ) and Lansner and Holst ( 1996 ), are biologically plausible brain cortex models that have been proven useful for understanding brain functions. Tully et al. ( 2016 ) implemented a BCPNN on SpiNNaker and analyzed the neural structure and dynamics inside a hypercolumn and demonstrated temporal sequence learning. Meli and Lansner ( 2013 ) studied the neural interconnection scheme from a BCPNN model. Fiebig et al. ( 2020 ) demonstrated how BCPNN could emulate the cortical working memory function. Recently, unsupervised hidden representation learning using BCPNN was benchmarked on MNIST. The BCPNN achieved 97.5% accuracy on the unseen test set (Ravichandran et al., 2020 ). Currently, the simulation of large scale BCPNNs heavily relies on high-performance computing centers equipped with supercomputers and accelerators, such as GPUs and ASICs (Farahini et al., 2014 ; Stathis et al., 2020 ), or dedicated spiking neural network simulation platform, such as SpiNNaker (Knight et al., 2016 ). We identify three categories of optimization methods: (1) reducing the amount of computation, (2) reducing the amount of memory access demand, and (3) increasing the memory access efficiency. Current studies of the BCPNN optimization are mainly focused on reducing the computation and memory access demand (Vogginger et al., 2015 ). The memory access efficiency aspect is seldom exploited. With technology scaling, memory access becomes the dominant cost for most applications (Mutlu, 2013 ). Memory optimization in terms of both reducing memory access demand and increasing the efficiency of the memory access has been done for many years both for non-Hebbian artificial neural networks and Hebbian spiking neural networks. For conventional non-spiking deep neural networks, research works like Li et al. ( 2016 ) and Yang et al. ( 2017 ) optimized both memory access demand and efficiency in deep convolutional neural networks. For Hebbian spiking neural networks, most research works target the spike-timing-dependent plasticity (STDP) learning rule (Markram et al., 2012 ). For example, Bichler et al. ( 2012 ) simplified the STDP learning rule and reduced the demand for computation and memory access. Yousefzadeh et al. ( 2017 ) further improved the method proposed by Bichler et al. by replacing the full connection to a weight-sharing connection. Thus, it further reduced the computation and memory access demand. Davies et al. ( 2012 ) changed part of the STDP learning rule and approximated the membrane potential in LTP. It reduced the computation and memory access demand and improved memory access efficiency. Jin et al. ( 2010 ) and Davies et al. ( 2018 ) delayed the update of weights and reduced the memory access demand. Pedroni et al. ( 2019 ) analyzed different synaptic matrix memory mapping strategies and proposed a variation of STDP learning rule to perform the causal and acausal update process. It reduced memory storage demand for the sparsely connected network by pointer-based compressed sparse rows and improved the efficiency for reversed access of such pointer-based data structure. Knight and Furber ( 2016 ) proposed a spike buffer and a mechanism called “flushing event” to deal with the inefficient column update in STDP and increase the memory efficiency. Morrison et al. ( 2007 ) used a dynamic spike buffer to remove column update process in STDP and increased memory efficiency. Sheik et al. ( 2016 ) also pointed out the memory problem caused by bi-directional spike-triggered learning rule and proposed a learning rule in which update is only triggered by presynaptic spikes to improve the memory access efficiency. However, almost all of these studies that target spiking neural networks focus on the STDP learning rule. Thus, it cannot be directly applied to BCPNN since its learning rule is different and more complex than STDP. By abandoning a conventional von Neumann architecture, custom neural network simulation platforms could potentially avoid the root of the memory access problems. Serrano-Gotarredona et al. ( 2013 ) and Prezioso et al. ( 2018 ) used memristors to merge computation and storage, thus eliminating the need for memory access. Though such new architectures are efficient and attractive, they are not off-the-shelf and easily accessible, and none of them supports the BCPNN learning rule. In this paper, we tackle the memory access problem introduced by the BCPNN optimization method presented in Vogginger et al. ( 2015 ). By replacing a time-driven simulation method with an event-driven one, plenty of computational requirements have been eliminated. The event-driven simulation method is called “lazy evaluation method” because it delays the evaluation computation as much as possible. The lazy evaluation simulation method requires access to the synaptic matrix stored in main memory, both row-wise and column-wise. A presynaptic spike triggers an update of a single row in the synaptic matrix, while a post-synaptic spike triggers an update of a single column. Today's memory architecture, such as DRAM, cannot handle two orthogonal directional access patterns of the same block of continuous data without sacrificing efficiency. Such access patterns will also affect the efficiency of the cache system. Therefore, we propose to remove the column update procedure and to merge the column and row update. In this way, we avoid entirely the dual memory access pattern that degrades the efficiency of the BCPNN simulation. Furthermore, by carefully designing our strategy, we can also reduce the demand in storage requirement and memory access, while increasing the overall performance. We remind readers that even though the BCPNN is the optimization target, our method is not restricted to this learning rule. Any Hebbian based neural network learning rule could potentially be optimized with a slightly modified version of our strategy. The rest of the paper is organized as follows: section 2 explains the original BCPNN learning rule, points out the memory access problem, proposes the alternative method that resolves the problem, and performs an error analysis for the proposed method. Section 3 demonstrates the benefits of the proposed method in terms of both memory storage requirements and performance. Finally, section 4 summarizes the paper and addresses the potential of the proposed method.",
"discussion": "4. Discussion In this paper, we have discussed the BCPNN memory access problem introduced by the lazy evaluation method. An algorithmic optimization has been proposed to tackle this issue. The proposed Column Update Elimination (CUE) method eliminates the column update and merges it with the row update, with the help of spike history buffer and approximation function. Using the CUE method, we gain not only memory access efficiency but also other improvements, such as the reduction of memory storage, memory access demand, etc. We also show that our algorithmic modification only introduces negligible errors and does not compromise the functionality of the BCPNN. In this section, we further examine the potential of the proposed method. We focus on the new possibilities after the column update has been eliminated, and other learning rules that CUE method can fit in. Finally, based on what we have achieved, we describe an outlook that could improve BCPNN simulation even further. 4.1. Exploiting the Temporal Locality of Spike Train Memory access efficiency can be further improved by architectural optimization. We have analyzed the pattern of spike train in section 2.4.1. Spikes are generated in burst mode regulated by stimulus pattern. Usually, the change of these stimulus patterns is infrequent. Thus, when the neural network is in the middle of a stable stimulus pattern, the firing patterns of spikes are also stable. We have analyzed the percentage of winning MCUs in both active and silent HCUs. Even in completely active H × M BCPNN (α = 1), the fraction of winning MCUs which fire frequently is just 1 M . Therefore, only a fraction of 1 M connections will be active at each simulation step, assuming uniform interconnections. Therefore, when the firing patterns are stable, the part of synaptic traces that need to be frequently updated is also stable, and the global memory access patterns of synaptic traces are stable as well. In this paper, we have eliminated the non-coalesced column-wise memory access pattern. With stable global memory access patterns, we can cache the frequently updated fraction of synaptic traces by designing a large enough cache between the memory and computation unit. The single global memory access pattern guarantees that the cache system will not be interrupted by other memory access patterns. Unfortunately, the estimated size of such a cache usually is much bigger than any off-the-shelf commercial computer architecture. The insufficient size of the cache will lead to frequent swapping data between the cache and DRAM, thus significantly compromises the memory access efficiency. The customization of the cache system becomes necessary and can only be done in custom hardware architecture, such as ASICs. 4.2. CUE Method for STDP The STDP learning rule changes the synaptic weights based on the correlation of pre- and post-synaptic spikes. The amount of change depends on the time difference between the pre- and post-synaptic spikes. A pre- and post-synaptic spike pair for a synaptic connection 〈 s i , s j 〉 occurs at time 〈 t i , t j 〉. If t i < t j , they are correlated. Otherwise, they are anti-related. One commonly used method to calculate such causal strength is described in Equation (13). (13) Δ w i j = { + A + · e - t j - t i τ , if t i < t j - A - · e - t i - t j τ , if t i > t j We can see that the update of weights in the STDP learning rule is triggered by pre- and post-synaptic spikes, just like the BCPNN learning rule. It is natural to organize the synaptic weights of STDP learning as a matrix stored in the memory. Because of its triggering mechanism and data structure, the STDP learning rule will suffer from the same dual memory access pattern issue as a lazy evaluation method. By proposing the same solution to STDP learning, we can use a post-synaptic buffer and an approximation function to delay the update of the post-synaptic triggered column update. The buffer size L , and the type of approximation function H will be different for STDP learning. Further analysis and simulation should be done to investigate the error bound. For example, the probability density function (PDF) of the firing rate in STDP might be different. It might be enough for STDP by just using the spike history buffer. However, the principle of optimizing memory access efficiency remains the same for both STDP and BCPNN learning rules. 4.3. Outlook In the future, we could also explore the option of hardware architecture to improve the BCPNN simulation further. For example, we could dimension a big enough cache that can hold the complete stationary synaptic traces to avoid DRAM memory access. We could also use non-von Neuman architecture, such as memristor, which could potentially avoid the memory access problem. Other algorithmic modifications combined with approximate computing, such as delayed stochastic row update, could also improve the overall BCPNN simulation."
} | 3,356 |
19735279 | null | s2 | 2,340 | {
"abstract": "Quorum sensing (QS) cell-cell communication systems are utilized by bacteria to coordinate their behaviour according to cell density. Several different types of QS signal molecules have been identified, among which acyl-homoserine lactones (AHLs) produced by Proteobacteria have been studied to the greatest extent. Although QS has been studied extensively in cultured microorganisms, little is known about the QS systems of uncultured microorganisms and the roles of these systems in microbial communities. To extend our knowledge of QS systems and to better understand the signalling that takes place in the natural environment, metagenomic libraries constructed using DNA from activated sludge and soil were screened, using an Agrobacterium biosensor strain, for novel QS synthase genes. Three cosmids (QS6-1, QS10-1 and QS10-2) that encode the production of QS signals were identified and DNA sequence analysis revealed that all three clones encode a novel luxI family AHL synthase and a luxR family transcriptional regulator. Thin layer chromatography revealed that these LuxI homologue proteins are able to synthesize multiple AHL signals. Tandem mass spectrometry analysis revealed that LuxI(QS6-1) directs the synthesis of at least three AHLs, 3-O-C14:1 HSL, 3-O-C16:1 HSL and 3-O-C14 HSL; LuxI(QS10-1) directs the synthesis of at least 3-O-C12 HSL and 3-O-C14 HSL; while LuxI(QS10-2) directs the synthesis of at least C8 HSL and C10 HSL. Two possible new AHLs, C14:3 HSL and (?)-hydroxymethyl-3-O-C14 HSL, were also found to be synthesized by LuxI(QS6-1)."
} | 391 |
23656326 | null | s2 | 2,342 | {
"abstract": "Microbially produced alkanes are a new class of biofuels that closely match the chemical composition of petroleum-based fuels. Alkanes can be generated from the fatty acid biosynthetic pathway by the reduction of acyl-ACPs followed by decarbonylation of the resulting aldehydes. A current limitation of this pathway is the restricted product profile, which consists of n-alkanes of 13, 15, and 17 carbons in length. To expand the product profile, we incorporated a new part, FabH2 from Bacillus subtilis , an enzyme known to have a broader specificity profile for fatty acid initiation than the native FabH of Escherichia coli . When provided with the appropriate substrate, the addition of FabH2 resulted in an altered alkane product profile in which significant levels of n-alkanes of 14 and 16 carbons in length are produced. The production of even chain length alkanes represents initial steps toward the expansion of this recently discovered microbial alkane production pathway to synthesize complex fuels. This work was conceived and performed as part of the 2011 University of Washington international Genetically Engineered Machines (iGEM) project."
} | 289 |
38004879 | PMC10673317 | pmc | 2,343 | {
"abstract": "Nanofibers have gained much attention because of the large surface area they can provide. Thus, many fabrication methods that produce nanofiber materials have been proposed. Electrospinning is a spinning technique that can use an electric field to continuously and uniformly generate polymer and composite nanofibers. The structure of the electrospinning system can be modified, thus making changes to the structure, and also the alignment of nanofibers. Moreover, the nanofibers can also be treated, modifying the nanofiber structure. This paper thoroughly reviews the efforts to change the configuration of the electrospinning system and the effects of these configurations on the nanofibers. Excellent works in different fields of application that use electrospun nanofibers are also introduced. The studied materials functioned effectively in their application, thereby proving the potential for the future development of electrospinning nanofiber materials.",
"conclusion": "4. Conclusions Electrospinning is a method of manufacturing nanofibers. Though the materials for the electrospinning processes are usually polymer solutions, the treatments on the nanofibers and extensive choices of materials allow many kinds of materials to be produced. To date, the materials that can be provided by electrospinning include polymer, polymer composite, metal, and metal oxide nanofibers. On the other hand, modifications to the electrospinning system configuration can lead to various changes in the collected nanofiber membrane. Firstly, not only does the use of special nozzles and collectors alter the morphology of the electrospun nanofibers, but the combination of different solvents and materials in one single electrospun solution can also lead to the same results. Furthermore, the after-treatment carried out on the nanofibers can convert them into different materials. Especially carbon nanofibers, which have a significantly high surface area created from the removal of sacrificing materials, can be harvested from the carbonization of electrospun nanofibers. Secondly, the addition of extra parts, which are used to control the pathway of the electrical field, can align the nanofibers following a designed pattern. In short, there is no limitation in the morphology and alignment of nanofibers that can be produced by electrospinning, opening up the potential to be further developed in the future. However, the disadvantages of the electrospinning process are that it is highly dependent on the environment humidity and stability of the power source, restricting the use of the technique in specific conditions. The applications of these nanofibers were also introduced in the review. It is clear that the electrospun nanofibers have been used in many applications, from simple applications, such as sound absorption and air purification, to highly advanced applications, like sensors, catalysts, and energy storage. The electrospinning method has great potential to provide materials with excellent properties, while it also offers many opportunities for further improvement.",
"introduction": "1. Introduction Nanofiber materials have attracted much attention because of their extraordinary advantages that can bring many benefits to applications [ 1 ]. Since the interaction of the materials with the working environment happens on their surface, improvement of the surface area can greatly enhance their working performance. Many studies have proposed that nanofiber materials possess a surface area much greater than bulk or 2D (sheet) materials [ 2 , 3 ]. Meanwhile, they are still imbued with the physical properties of the material from which they are built. These properties can help the materials to be applied in some applications that require a certain physical strength [ 4 , 5 , 6 , 7 , 8 ]. Many fabrication methods for nanofibers have been proposed in the literature, including the uses of hard, soft, and free templates. While hard and soft templates release a large amount of chemicals into the environment, the free template seems to be more environment-friendly, in which case, electrospinning is a free template technique that uses an electrical field to fabricate the polymer and polymer composite nanofibers [ 9 , 10 , 11 , 12 , 13 ]. This technique requires a very limited amount of solvent, which is used to dissolve the polymer into a high-viscosity solution. After the shape of fibers is applied to the solution, the solvent is released and leaves the material in the shape of the nanofibers [ 14 ]. The polymer used in the process must be dissolved in the solvent, which limits the selection of polymers that can be spun. However, some of the available commercial polymers are dissolvable in a low boiling point solvent, so such limitation is almost negligible. Though the mechanism of electrospinning is simple, many improvements can be carried out on the components of the system [ 15 , 16 , 17 ]. Thereby, the morphology of the produced nanofibers can be greatly altered. There are reports that described further treatment processes that have been carried out on the electrospun nanofibers [ 18 ]. As collected, the nanofibers usually stack over each other, layer after layer, and form a membrane. They can be used as a membrane sheet and built up to a 3D structure [ 19 , 20 , 21 ]. On the other hand, carbonization can transform most of the existing polymer nanofibers into carbon nanofibers (CNFs). These CNFs showed great potential in electrochemical application, which has gained increased interest in recent years [ 22 ]. Herein, the manuscript proposes a critical review of the electrospinning system. First, the modifications made to components (electrospun material, nozzle, and collector) of the electrospinning system are introduced, together with the effects on the shape of nanofibers. Later, the recent applications of nanofibers that are fabricated by electrospinning are suggested. Many modifications of nanofibers, including freeze-drying, deposition, and thermal treatment, can be applied to achieve suitable properties for many advanced applications."
} | 1,518 |
19077236 | PMC2621215 | pmc | 2,345 | {
"abstract": "Background Acidithiobacillus ferrooxidans is a major participant in consortia of microorganisms used for the industrial recovery of copper (bioleaching or biomining). It is a chemolithoautrophic, γ-proteobacterium using energy from the oxidation of iron- and sulfur-containing minerals for growth. It thrives at extremely low pH (pH 1–2) and fixes both carbon and nitrogen from the atmosphere. It solubilizes copper and other metals from rocks and plays an important role in nutrient and metal biogeochemical cycling in acid environments. The lack of a well-developed system for genetic manipulation has prevented thorough exploration of its physiology. Also, confusion has been caused by prior metabolic models constructed based upon the examination of multiple, and sometimes distantly related, strains of the microorganism. Results The genome of the type strain A. ferrooxidans ATCC 23270 was sequenced and annotated to identify general features and provide a framework for in silico metabolic reconstruction. Earlier models of iron and sulfur oxidation, biofilm formation, quorum sensing, inorganic ion uptake, and amino acid metabolism are confirmed and extended. Initial models are presented for central carbon metabolism, anaerobic metabolism (including sulfur reduction, hydrogen metabolism and nitrogen fixation), stress responses, DNA repair, and metal and toxic compound fluxes. Conclusion Bioinformatics analysis provides a valuable platform for gene discovery and functional prediction that helps explain the activity of A. ferrooxidans in industrial bioleaching and its role as a primary producer in acidic environments. An analysis of the genome of the type strain provides a coherent view of its gene content and metabolic potential.",
"conclusion": "Conclusion • Bioinformatics analysis of the complete genome of the type strain of A. ferrooxidans (ATCC 23207) provides a valuable platform for gene discovery and functional prediction that is especially important given the difficulties in carrying out standard genetic research in this microorganism. The models presented herein should facilitate the design and interpretation of future experiments and enable the experimental investigator to focus on important issues. • An analysis of the genome of the type strain provides a coherent view of the gene content and metabolic potential of this species (Figure 7 ). Figure 7 Whole-cell model for A. ferrooxidans ATCC 23270 . Genome-based model of the cellular metabolism of A. ferrooxidans including predicted transport systems; chemolithoautotrophic components; carbon, nitrogen and sulfur metabolism; and biogeochemical cycling. • Metabolic models support the key capabilities of A. ferrooxidans that pertain to its use in industrial bioleaching, including its ability to oxidize both sulfur and iron, to resist low pH, and to live in environments with potentially toxic organic and inorganic chemicals. They also suggest that it has the ability to precipitate metals in anaerobic environments, which would be deleterious to copper bioleaching activity. • Our analysis also prompts several unexpected predictions, some of which could potentially be useful in biomining such as the proposed connection between biofilm formation and central carbon metabolism and the presence of several predicted quorum sensing mechanisms. Indications of past phage infection and conjugation events suggest potentially fruitful approaches for the development of efficient methods for genetic manipulation of this microorganism. • Metabolic models also indicate how this microorganism could play an important role as a producer of fixed carbon and nitrogen and as a recycler of metals in bioleaching operations as well as in natural environments.",
"discussion": "Results and discussion 1. Genomic properties The genome of A. ferrooxidans ATCC 23270 (type strain) consists of a single circular chromosome of 2,982,397 bp with a G+C content of 58.77%. No plasmids were detected in the type strain, although they occur in several other strains of [ 24 ]. A total of 3217 protein-coding genes (CDSs) were predicted, of which 2070 (64.3%) were assigned a putative function (Table 1 and Figure 2 ). The genome encodes two ribosomal operons and 78 tRNA genes. A putative origin of replication (Figure 2 ) has been identified from marginal GC skew variations in the genome and by the localization of the dnaN and dnaA genes (AFE0001 and AFE3309). Figure 2 Circular representation of the A. ferrooxidans ATCC 23270 genome sequence . The two outer circles represent predicted protein encoding-genes on the forward and reverse strands, respectively. Functional categories are indicated by color, as follows: energy metabolism (green), DNA metabolism (red), protein synthesis (magenta), transcription (yellow), amino acid metabolism (orange), central intermediary metabolism (dark blue), cellular processes (light blue), nucleotide metabolism (turquoise), hypothetical and conserved hypothetical proteins (grey), mobile and extra-chromosomal elements (black), and general functions (brown). The third and fourth circles (forward and reverse strands) indicate major transposases and mobile elements (orange), plasmid-related genes (red), and phage elements (blue). The fifth and sixth circles (forward and reverse strands) indicate tRNA genes (gray). The seventh and eighth circles (forward and reverse strands) show genes predicted to be involved in sulfur (purple), iron (red), and hydrogen (orange) oxidation. The ninth and tenth circles show genomic GC bias and GC skew, respectively. Table 1 General features of the A. ferrooxidans ATCC 23270 genome. Characteristic Value Complete genome size, bp 2,982,397 G+C percent (%) 58.77 Total number of CDSs 3,217 Coding density (%) 97.45 No. of rRNA operons (16S-23S-5S) 2 No. of tRNA genes 78 Proteins with known function 2,070 Conserved hypothetical proteins 388 Hypothetical proteins 759 Most represented functional categories (%) Cell envelope 7.8 Transport and binding proteins 7.61 Energy metabolism 6.52 Best BLASTP comparisons against complete proteomes Number of best blast hits γ-proteobacteria 899 β-proteobacteria 791 α-proteobacteria 271 δ-proteobacteria 103 Cyanobacteria 73 Archaea 41 2. Chemolithoautotrophy A. ferrooxidans has a complete repertoire of genes required for a free-living, chemolithoautotrophic lifestyle, including those for CO 2 fixation and nucleotide and cofactor biosynthesis (Additional file 1 ). Analysis of an earlier draft genome had predicted genes for the pathways for synthesis of most amino acids, although ten genes were missing [ 23 ]. Seven of these missing assignments have now been detected: a potential 6-phosphofructokinase in the glycolysis pathway (EC 2.7.1.11; AFE1807), pyruvate dehydrogenase (EC 1.2.4.1; AFE3068-70); shikimate kinase in the chorismate synthesis pathway and required for tryptophan, phenylalanine and tyrosine biosynthesis (EC 2.7.1.71; AFE0734); homeserine kinase in the threonine biosynthesis pathway (EC 2.7.1.39; AFE3097); N-acetyl-gamma-glutamil-1-phosphate reductase in the ornithine biosynthesis pathway and required for proline biosynthesis (EC 1.2.1.38; AFE3073); pirroline-5-carboxilate reductase involved in proline biosynthesis (EC 1.5.1.2; AFE0262); and asparagine synthase (EC 6.3.5.4: AFE1353). The three genes identified in E. coli which have not been found in A. ferrooxidans encode ornithine cyclodeaminase (EC 4.3.1.12) involved in proline biosynthesis, aromatic-amino-acid transaminase (EC 2.6.1.57), and arogenate dehydrogenase involved in tyrosine biosynthesis (EC 1.3.1.43). A. ferrooxidans has two glutamyl-tRNA synthetases: a more discriminating one (D-GluRS, AFE0422) that charges only Glu-tRNA(Glu) and a less discriminating one (ND-GluRS, AFE2222) that charges Glu-tRNA(Glu) and Glu-tRNA(Gln). The latter one is a required intermediate in protein synthesis in many organisms [ 25 ]. An indirect regulation of glutamyl-tRNA synthetase by heme status suggests a potential metabolic connection between heme requirements, nitrogen, and central carbon metabolism [ 26 ]. Bioinformatic analysis supports prior experimental evidence that A. ferrooxidans has a versatile aerobic metabolism, capable of providing energy and reducing power requirements from inorganic compounds by the oxidation of Fe(II), reduced sulfur compounds, formate, and hydrogen. In addition, gene function predictions suggest that the microorganism is capable of anaerobic or micro-aerophilic growth using Fe(III) or elemental sulfur as alternative electron acceptors [ 27 ]. Many of the predictions were experimentally validated in a piece-meal fashion in a number of diverse strains of A. ferrooxidans , some of which may not belong to the same species [ 28 ]. Herein, we describe a coherent view of the metabolic potential of the type strain that will now allow a systematic appraisal of the diversity of the metabolic capacity of the A. ferrooxidans pangenome. 2.1 CO 2 fixation A. ferrooxidans fixes CO 2 via the Calvin-Benson-Bassham reductive pentose phosphate cycle (Calvin cycle) using energy and reducing power derived from the oxidation of iron or sulfur [ 29 ]. Early studies showed a relationship between the rate of iron and sulfur oxidation and the rate of CO 2 fixation in A. ferrooxidans (no strain designated) [ 30 ]. Several enzymes of the Calvin cycle have been described in A. ferrooxidans , including the key D-ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) [ 29 ]. Two structurally distinct forms of RuBisCO (I and II), with different catalytic properties, are typically present in autotrophs [ 31 ]. Genes encoding Form I (AFE3051-2) have been cloned and characterized from A. ferrooxidans [ 32 , 33 ]. Gene clusters potentially encoding a second copy of Form I (AFE1690-1) and a copy of Form II (AFE2155) were predicted and shown to be differentially expressed depending on whether A. ferrooxidans was grown on iron- or sulfur-containing medium. [ 34 ]. A gene predicted to encode a novel Rubisco-like protein known as Form IV [ 35 ] was recently identified in the genome (AFE0435) and is suggested to be involved in stress response (Esparza-Mantilla, personal communication) (Additional file ). The genomic organization of the three gene clusters encoding the Rubisco type I and II enzymes in A. ferrooxidans is similar to that found in Hydrogenovibrio marinus strain MH-110, an obligate chemolithoautotrophic, hydrogen-oxidizing, marine bacterium. In H. marinus , these three-gene clusters are regulated in response to CO 2 concentration, suggesting the ability to adapt to environmental conditions with different levels of CO 2 [ 36 ]. 2.2 Energy metabolism 2.2.1 Aerobic Iron oxidation Since ferrous iron [Fe(II)] is rapidly oxidized by atmospheric oxygen at neutral pH, iron exists primarily in the oxidized form [Fe(III)] in aerobic environments. Therefore, ferrous iron is available for microbial oxidation principally in acidic environments where chemical oxidation is slow and Fe(II) is soluble, in anoxic conditions such as in marine sediments and at the interface between aerobic and anaerobic atmospheres [ 37 ]. In anoxic conditions, phototrophic bacteria can use light energy to couple the oxidation of Fe(II) to reductive CO 2 fixation. Although little is known about the mechanisms involved, this process has been postulated to be an ancient form of metabolism and to represent a transition step in the evolution of oxygenic photosynthesis [ 38 , 39 ]. The bioinformatics analysis of the genome sequence of A. ferrooxidans has permitted the identification of the main components of the electron transport chain involved in iron and sulfur oxidation (Figure 3 ). Genes encoding iron oxidation functions are organized in two transcriptional units, the petI and rus operons. The petI operon ( petC-1, petB-1, petA-1, sdrA-1 , and cycA-1; AFE3107-11) encode the three subunits of the bc 1 complex (PetCAB), a predicted short chain dehydrogenase (Sdr) of unknown function, and a cytochrome c 4 that has been suggested to receive electrons from rusticyanin and pass them to the bc 1 complex [ 5 ]. The petI operon has been analyzed experimentally in A. ferrooxidans strain ATCC 19859 [ 5 ] and recently in strain ATCC33020 [ 8 ]. Figure 3 Genome-based models for the oxidation of ferrous iron and reduced inorganic sulfur compounds (RISCs) . Schematic representation of enzymes and electron transfer proteins involved in the oxidation of (A) ferrous iron and (B) reduced inorganic sulfur compounds (RISCs). Proteins and protein complexes are described in the text. The rus operon ( cyc2, cyc1, hyp, coxB, coxA coxC, coxD , and rus ; AFE3146-53) encodes two c-type cytochromes (Cyc1 and Cyc2), components of the aa3-type cytochrome oxidase (CoxBACD), and rusticyanin, respectively [ 40 ]. Cyc2 has been shown to accept electrons directly from Fe(II) and, given its location in the outer membrane, may carry out the first step in Fe(II) oxidation [ 41 ]. These proteins are thought to form a \"respiratory supercomplex\" that spans the outer and the inner membranes and transfers electrons from iron (or pyrite) to oxygen [ 40 , 42 , 43 ]. Based on transcriptional, biochemical, and genetic studies [ 28 ], it was proposed that electrons from ferrous iron oxidation flow through Cyc2 to rusticyanin. From there, some of the electrons feed the \"downhill electron pathway\" through c -cytochrome Cyc1 to aa 3 cytochrome oxidase, some the \"uphill electron pathway\" that regenerates the universal electron donor NADH by the reverse electron flow through c -cytochrome CycA1--> bc 1 complex-->ubiquinone pool-->NADH dehydrogenase (Figure 3a ). Genome analysis suggests a solution to a long-standing controversy. A HiPIP (high potential iron-sulfur protein) encoded by iro has been postulated to be the first electron acceptor from Fe(II) [ 44 , 45 ]. However, transcriptional studies of iro in A. ferrooxidans ATCC33020 suggested that it may be involved in sulfur oxidation. In our analysis of the type strain, iro (AFE2732) was found to be associated with the petII gene cluster thought to be involved in sulfur oxidation [ 46 , 47 ], thus making it unlikely that Iro is the key iron-oxidizing enzyme. 2.2.2 Aerobic oxidation of reduced inorganic sulfur compounds (RISCs) Genes encoding enzymes and electron transfer proteins predicted to be involved in the oxidation of reduced inorganic sulfur compounds (RISCs) were detected in the genome (Figure 3b ). The oxidative and electron transfer pathways for RISCs are more complicated than those for Fe(II) oxidation, making their prediction and elucidation more difficult [ 48 ]. To add further complication, some steps occur spontaneously, without enzymatic catalysis. Previous experimental studies in various strains of A. ferrooxidans detected several enzymatic activities involved in the oxidation of RISCs [ 1 , 28 ], but some of these activities had not been linked to specific genes. Based on genome analysis, some of these missing assignments are predicted and also some novel genes involved in the oxidation of thiosulfate, sulfide, and tetrathionate are suggested. Experimentally validated components of RISC metabolism include: the pet-II operon (AFE2727-31) and alternative quinol oxidases of the bd (AFE0954-5) and bo 3 families (AFE0631-4) [ 7 , 8 ]; a sulfide/quinone oxidoreductase encoded by sqr (AFE0267) suggested to be involved in the oxidation of sulfide to sulfur [ 49 , 50 ]; and a tetrathionate hydrolase encoded by tetH (AFE0029) thought to be involved in the oxidation of tetrathionate [ 51 ]. The two homologs of doxDA (AFE0044; AFE0048) present in the genome are predicted to encode a thiosulfate/quinone oxidoreductase. Both appear to be a fusion of the separate doxD and doxA genes that are found in other organisms such as A. ambivalens [ 52 , 53 ]. Both are located in a major gene cluster composed of two divergent gene clusters. The first region (AFE0050-47) encodes a protein with TAT-signal peptide (IPR006311, TIGR01409), a periplasmic solute-binding protein, the first doxDA gene, and a conserved hypothetical protein. The second region (AFE0046-42) encodes a conserved hypothetical protein, a rhodanese enzyme that splits thiosulfate into sulfur and sulfite [ 54 ], the second copy of doxDA , a periplasmic solute-binding protein, a second copy of a gene encoding a protein with TAT-signal peptide (IPR006311, TIGR01409), and a gene encoding a putative carboxylate transporter. We have detected a similar organization in the Gluconobacter oxydans genome. Five genes, predicted to encode thiosulfate sulfur transferase (rhodanese) proteins (AFE2558, AFE2364, AFE1502, AFE0529 and AFE0151) are dispersed in the genome [ 55 ] but their roles in sulfur oxidation remain to be firmly established. Notably, some of these predictions are based on the presence of the rhodanese PFAM00581 motif associated with phosphatases and ubiquitin C-terminal hydrolases, in addition to sulfur oxidation. Genes were not detected for several enzymatic functions that have been experimentally demonstrated in other strains of A. ferrooxidans including the sulfur dioxygenase that oxidizes persulfide-sulfur to sulfite in A. ferrooxidans strain R1 [ 1 , 56 ] and the sulfite oxidase that oxidizes sulfite to sulfate in Ferrobacillus ferrooxidans [ 1 , 57 ]. 2.2.3 Hydrogen and formate utilization Hydrogen utilization has been demonstrated experimentally in A. ferrooxidans ATCC 23270 [ 9 ] and a group 2 hydrogenase from A. ferrooxidans ATCC 19859 has been characterized [ 58 ], but there were no previous reports describing the hydrogenase genes and their genetic organization or their potential diversity. The A. ferrooxidans genome encodes four different types of hydrogenases based on the 2001 classification by Vignais et al. [ 59 ] (Figure 4 , Additional file 3 ). Group 1 [NiFe]-hydrogenases are membrane-bound respiratory enzymes that enable the cell to use molecular hydrogen as an energy source. A. ferrooxidans has both the predicted structural (AFE3283-86) and the maturation-related genes (AFE3281-2; AFE3287-90) required for production of a functional respiratory hydrogenase of this type. In addition, the small subunit of this predicted complex has the characteristic TAT-signal peptide used to target the full heterodimer to the periplasmic space [ 60 ]. The genomic arrangement of the structural genes ( hynS-isp1-isp2-hynL ) is identical to that found in a thermoacidophilic archaeon ( Acidianus ambivalens ), a hyperthermophilic bacterium ( Aquifex aeolicus ), a denitrifying bacterium ( Thiobacillus denitrificans ), and two phototrophic sulfur bacteria ( Thiocapsa roseopersicina and Allochromatium vinosum ). Like A. ferrooxidans , all of these bacteria are chemoautotrophs that live in extreme environments, use inorganic energy sources, and have an active sulfur metabolism that oxidizes and reduces inorganic sulfur compounds [ 61 ]. Figure 4 Diversity and genomic organization of predicted hydrogenases . A) Schematic representation of the four predicted types of hydrogenase. B) Organization of the predicted operons encoding the four types of hydrogenase. C) Schematic representation of similarity between the group 4 hydrogenase genes in M. barkeri with the A. ferrooxidans group 4 hydrogenase (above) and NADH dehydrogenase subunits (below). A. ferrooxidans also encodes a group 2 cytoplasmic uptake [NiFe]-hydrogenase (AFE0701-2). Group 2 hydrogenases are induced during nitrogen fixation to utilize the molecular hydrogen generated [ 62 ]. The cyanobacterial-like hydrogenase in A. ferrooxidans exhibits the characteristic features of uptake hydrogenases as determined by EPR and FTIR [ 63 ]. Divergently oriented from the group 2 hydrogenase gene cluster is a predicted σ 54 -dependent hydrogenase transcriptional regulator ( hupR ) (AFE0700). HupR together with a histidine kinase forms part of a two-component regulatory system in R. eutropha [ 64 ], but the histidine kinase appears to be absent from the A. ferrooxidans genome. Despite that, HupR is able to activate transcription in the non-phosphorylated form [ 65 - 67 ], indicating that HupR is still able to regulate transcription of the group 2 hydrogenase system in A. ferrooxidans . Adjacent to the group 2 hydrogenase gene cluster and transcribed in the same direction is a predicted cysteine regulon transcriptional activator cysB (AFE0699). This is followed by a cluster of genes potentially involved in fermentation, including a predicted σ 54 -dependent transcriptional regulator and a group of isc -like genes (AFE0672-78). The latter gene group is thought to be involved in assembling the iron-sulfur cluster of the nitrogenase used in nitrogen fixation, thus suggesting a connection between hydrogen production by the group 2 hydrogenase and nitrogen fixation [ 68 ]. The close proximity of the fermentation gene cluster suggests an additional metabolic coupling with fermentative metabolism, perhaps as part of a σ 54 regulatory cascade operating in anaerobic or microaerophilic conditions. The third predicted hydrogenase encodes a sulfhydrogenase, a group 3b cytoplasmic, bidirectional, heterotetrameric hydrogenase. This hydrogenase, in association with other proteins, binds soluble cofactors such as NAD, cofactor 420, and NADP [ 59 ]. Domain analysis predicts an F420 binding site in the α subunit (large hydrogenase subunit; AFE0937) and NAD- and FAD-binding sites in the γ subunit (AFE0939). The predicted NAD-binding site suggests that A. ferrooxidans can use NADPH as an electron donor, as has been shown for Pyrococcus furiosus [ 69 ]. A possible role for this hydrogenase could be the recycling of redox cofactors using protons or water as redox counterparts, as has been suggested for Alcaligenes eutrophus , thus serving as an electron sink under high reducing conditions [ 66 ]. The gene organization and amino acid sequence of a six-gene cluster (AFE2149-54) (Figure 4c ) shows significant similarity to the group 4 H 2 -evolving hydrogenase complex found in several organisms (e.g., Methanococcus barkeri [ 70 ]). In M. barkeri , this cluster encodes a six-subunit complex that catalyzes the energetically unfavorable reduction of ferrodoxin by H 2 , possibly driven by reverse electron transport. The reduced ferrodoxin produced then serves as a low-potential electron donor for the synthesis of pyruvate in an anabolic pathway [ 71 ]. Reverse electron flow for the production of NADH via the oxidation of Fe(II) in A. ferrooxidans has been shown to be driven by the proton motif force (PMF) across its membrane that results from the acidity of its environment [ 72 ]. The predicted activity of the group 4 hydrogenase complex may exemplify another where A. ferrooxidans exploits the natural PMF to generate reducing power and couple it to redox reactions. Another possible role for the group 4 hydrogenase complex involves the oxidation of formate. Two clusters of three genes (AFE1652-4 and AFE0690-2) potentially encode a formate dehydrogenase complex consisting of a formate dehydrogenase accessory protein FdhD-1, a hypothetical protein, and a molybdopterin formate dehydrogenase. The second cluster is divergently oriented from a gene encoding a predicted σ 54 -dependent transcriptional regulator. It has been reported that this complex associates with a hydrogenase group 4 complex in E. coli to create a formate hydrogenase supercomplex [ 73 ]. We propose a similar model for A. ferrooxidans , thus offering a biochemical basis for its ability to oxidize formate [ 10 ]. 2.2.4 Anaerobic metabolism Several strains of A. ferrooxidans have been reported to use electron acceptors other than O 2 , including the use of ferric iron for the oxidation of sulfur and hydrogen and the use of sulfur for the oxidation of hydrogen by A. ferrooxidans JCM 7811 [ 74 ]. In that strain, the reduction of ferric iron was accompanied by the increased expression of a 28 kDa c-type cytochrome that was suggested to be responsible for this activity [ 74 ]. The reduction of ferric iron during sulfur oxidation was also shown for the type strain ATCC 23270 [ 75 ]. However, a gene potentially encoding this cytochrome could not be identified in A. ferrooxidans ATCC 23270 [ 76 ]. A candidate iron reduction complex has been investigated in A. ferrooxidans AP19-3 by electrophoretic purification and enzymatic assays [ 77 , 78 ]. However, potential genes encoding this complex could not be detected in our genome analysis. The use of sulfur as an electron acceptor was investigated in A. ferrooxidans NASF-1 where aerobically grown cells were found to produce hydrogen sulfide from elemental sulfur using NADH as electron donor via a proposed sulfur reductase [ 79 ]. However, the observed molecular weights of the subunits of this sulfur reductase do not correspond to those predicted from an analysis of the group 3b hydrogenase genes in the type strain genome, with the caveat that post-translational modifications could explain the differences in molecular weights. However, a gene cluster (AFE2177-81) was detected in the type strain that is predicted to encode a sulfur reductase enzyme with significant similarity of amino acid sequence and gene order to the cluster suggested to be responsible for sulfur reduction in Acidianus ambivalens [ 80 ]. We hypothesize that this enzyme could associate with the predicted group 1 hydrogenase to form a supercomplex, facilitating the use of hydrogen as an electron and energy source with sulfur serving as the final electron acceptor. 2.3 Nitrogen metabolism A. ferrooxidans can meet its nitrogen needs by either nitrogen fixation or ammonia assimilation. Diazotrophic growth of A. ferrooxidans was first demonstrated in early studies of acetylene reduction and 15 N 2 assimilation [ 15 ] and the structural genes for the nitrogenase complex were later sequenced [ 81 - 83 ]. 2.3.1 Ammonia uptake and utilization The A. ferrooxidans genome contains genes predicted to be involved in ammonia uptake ( amt1 , amt2 and amtB ; AFE2916, AFE2911, and AFE1922). Amt1 and amt2 are located in a gene cluster that includes a gene potentially encoding a class-I glutamine amidotransferase (AFE2917) that has been shown in other organisms to transfer ammonia derived from the hydrolysis of glutamine to other substrates. GlnK-1 (AFE2915) is also present in this cluster and is predicted to encode a P-II regulatory protein involved in the regulation of nitrogen metabolism in response to carbon and glutamine availability [ 84 ]. A glnA homolog (AFE0466) is predicted to encode a type I glutamine synthase that would permit the incorporation of ammonia directly into glutamine, completing the inventory of genes necessary for ammonia uptake and utilization. 2.3.2 Nitrogen Fixation A putative nitrogenase gene cluster ( nifH-D-K-fer1-fer2-E-N-X ; AFE1522-AFE1515) (Additional file 4 ) was previously reported in the type strain [ 68 ]. These genes potentially encode the nitrogenase complex and proteins involved in the synthesis of the nitrogenase MoCo cofactor. In other organisms, nitrogenase has been shown to be oxygen sensitive and its expression and activity are regulated at both the transcriptional and post-translational levels [ 84 ]. Divergently oriented from the nif operon is a cluster of genes involved in the regulation of nitrogenase activity. The first gene of this cluster is a putative σ 54 response regulator (AFE1523). This is followed by the draT and draG genes (AFE1524, AFE1525) that encode a dinitrogen-reductase ADP-D-ribosyltransferase and a ADP-ribosyl-[dinitrogen reductase] hydrolase, respectively. These two are involved in the post-translational modulation of nitrogenase activity in response to ammonium and oxygen concentrations [ 84 ]. NifA (AFE1527) is also present in the same gene cluster. NifA potentially encodes an enhancer binding protein that, together with σ 54 , is involved in the transcriptional activation of the nif operon in response to the redox, carbon, and nitrogen status. This ensures that nitrogen fixation occurs only under physiological conditions that are appropriate for nitrogenase activity [ 85 ]. Using this genomic information, a gene network for the regulation of nitrogen fixation and ammonia uptake can be suggested for A. ferrooxidans that is consistent with similar models derived from other organisms (Figure 5 ) [ 84 ]. In this model, NifA (AFE1527) is the transcriptional activator of the nitrogenase operon and its expression is regulated by a two-component regulatory system encoded by ntrB and ntrC (AFE2902, AFE2901) that measure oxygen and nitrogen levels. These signals are integrated by the P-II proteins ( glnK-1 , AFE2915; glnB-1 , AFE2462; glnK-2 , AFE2240; and glnB-2 , AFE0429) with additional metabolic signals, such as fixed carbon and energetic status [ 86 ]. Two additional copies of ntrC and ntrB , termed ntrY and ntrX (AFE0024, AFE0023) have been detected in the genome that could allow cross talk between the sensor/regulator pairs NtrY/X and NtrB/C, as described in Azospirillum brasilense [ 87 ]. The redundancy of the regulatory genes responsible for nitrogen fixation and assimilation suggests the presence of a flexible mechanism that is responsive to environmental changes. Figure 5 Predicted regulatory models for inorganic ion uptake and assimilation . A) Phosphate and phosphonate. B) Nitrogen and ammonia. C) Ferric and ferrous iron. D) Sulfate. Proteins and protein complexes are described in the text. 3. Nutrient uptake and assimilation systems A. ferrooxidans has 72 genes (2.23%) predicted to be involved in nutrient uptake (Additional file 3 ) whereas most heterotrophic γ-proteobacteria typically dedicate about 14% of their genome information to transport functions [ 88 ]. The potential substrates incorporated include phosphate, sulfate, iron, ammonia, organic acids, amino acids, and sugars. This repertoire, especially the low representation of predicted carbohydrate uptake systems, is a signature of obligate autotrophic bacteria [ 88 ]. 3.1 Inorganic ion assimilation 3.1.1 Sulfate A gene for a predicted sulfate permease (AFE0286) of the SulP family is present in the genome adjacent to a potential carbonic anhydrase gene (AFE0287). This linkage has been observed in many bacteria [ 89 ], suggesting that the gene pair forms a sulfate/carbonate antiporter system. Sulfate taken up from the environment is thought to be reduced to sulfide for cysteine biosynthesis by a group of genes belonging to the cys regulon [ 68 ]. 3.1.2 Phosphate Previous investigations of phosphate metabolism in A. ferrooxidans provided evidence for a phosphate starvation response [ 90 ] and for a relationship between polyphosphate degradation and heavy metal resistance and efflux [ 91 ]. However, we still lacked a comprehensive understanding of all the potential components involved in phosphate metabolism, as well as their integration and regulation. Our genome analysis identified a complete repertoire of the genes necessary for phosphate uptake by the high affinity Pst-transport system. These predicted genes are arranged in two similar clusters. The first cluster (AFE1939-41) includes two genes ( pstC-1 and pstC-2 ) that encode a phosphate permease and a third gene ( pstS-1 ) that encodes a periplasmic phosphate binding protein. The second gene cluster (AFE1441-1434) includes a gene encoding a exopolyphosphatase ( ppx , AFE1441) previously described to be involved in heavy metal resistance and efflux [ 91 ], phoU (AFE1440) predicted to encode a phosphate transport regulatory protein, pstB encoding an ATP binding protein, and pstA coding for the permease component. In addition, there are genes encoding a third homolog of the phosphate permease PstC-3, a second homolog of the periplasmic phosphate binding protein PstS-2, and the two-component response regulator PhoR/PhoB. The predicted phosphonate utilization gene cluster (AFE2278-86) contains genes for C-P cleavage and an ATP-binding protein for the ABC phosphonate transport system. In spite of the experimental evidence reported about the utilization of phosphonate in this bacterium [ 92 ], the typical permease subunit that is required to complete phosphonate uptake was not found in the genome. The genome does contain a gene (AFE1876) for a predicted polyphosphate kinase (Ppk) involved in polyphosphate storage. It has been suggested to be part of a pho regulon whose expression is activated during phosphate starvation [ 89 ] and in response to heavy metal toxicity [ 91 ]. 3.1.3 Iron Genomic evidence indicates that A. ferrooxidans relies on diverse standard iron uptake mechanisms to obtain both Fe(II) and Fe(III) (93). The type strain has candidate genes (AFE2523-AFE2525) potentially encoding the FeoABC Fe(II) inner-membrane transport system and an NRAMP dual Mn(II)/Fe(II) MntH-like transporter (AFE0105). Previously reported gene context analysis indicated that feoA, feoB , and feoC form part of an iron-regulated operon, along with an ORF (AFE2522) encoding a putative porin (designated feoP ) that could facilitate entrance of Fe(II) into the periplasm [ 93 ]. A. ferrooxidans is typically confronted with an exceptionally high concentration of soluble iron in its acidic environment, as high as 10 -1 M compared to 10 -16 M in typical neutrophilic environments. This raises questions as to the mechanisms it uses for iron assimilation and homeostatic control of internal iron concentrations. Given the abundance of both Fe(II) and Fe(III) in its environment, A. ferrooxidans has a surprisingly large number of iron uptake systems, including eleven distinct putative genes encoding TonB-dependent outer membrane receptors ( tdr ) for high affinity uptake of siderophore-chelated Fe(III) ( tdrA , AFE2935; tdrC , AFE1483; tdrD , AFE1492; tdrE , AFE2040; tdrF , AFE2998; tdrG , AFE2302; tdrH , AFE2298; tdrI , AFE2292; tdrJ , AFE2288; tdrK , AFE0763; tdrL , AFE3229). Also, it has a number of copies of all the accessory genes needed to transport iron into the cytoplasm, including seven different copies of the energy transduction genes tonB (AFE3002, AFE2304, AFE2301, AFE2275, AFE2268, AFE1487, AFE0770) and exbB (AFE3003, AFE2299, AFE2273, AFE2270, AFE1485, AFE0768, AFE0485) and six copies of exbD (AFE3004, AFE2300, AFE2269, AFE1486, AFE0769, AFE0486), as well as the genes encoding two different ABC iron transporters (AFE1489-AFE1491, AFE1493-AFE1495). No genes were detected that might be involved in standard mechanisms of siderophore production. However, its multiple siderophore uptake systems suggest that it is nonetheless capable of living in environments where iron is scarce (perhaps at higher pH values) and in which other organisms capable of producing siderophores are present. For Fe(III) uptake, all the genes involved are organized in seven gene clusters, some of which include additional gene functions [ 22 ]. One cluster encodes a complete suite of proteins necessary for Fe(III) uptake (AFE1482-AFE1495) that includes not only two outer membrane receptors (OMRs) of different predicted siderophore specificities, but also three different ABC solute-binding proteins with affinity for iron and molybdenum and may be a dedicated iron-molybdenum transport system that is present in a genomic island [ 94 ]. This predicted operon also includes a putative gene ( gloA , AFE1482) predicted to encode a globin-like protein that has been suggested to be an oxygen sensor regulating the expression of Fe-Mo uptake [ 94 ]. GloA is also associated with an upstream Fur box, indicating possible regulation via the master iron regulator Fur [ 21 ]. 3.2 Carbon compound uptake 3.2.1 Amino acids Among the predicted nutrient transport genes in the A. ferrooxidans genome are five amino acid permeases of unknown specificity (AFE2659, AFE2457, AFE1782, AFE0719, and AFE0439) (the same number as found in the chemolithoautotroph T. crunogena ) and one complete ABC system for dipeptide uptake (AFE2987-92). The addition of leucine to solid media has been reported to improve the yield of A. ferrooxidans ATCC 33020 during the first ten days of growth, whereas the addition of cysteine or methionine inhibits growth [ 95 ]. More recently, the addition of minimal concentrations of glutamate to liquid media was found to accelerate the growth rate of A. ferrooxidans ATCC 23270 (Omar Orellana, personal communication). 3.2.2 Carbohydrate uptake The suite of genes for carbohydrate transport appears to be limited, as has been found in most obligate autotrophs (e.g., T. denitrificans [ 96 ], T. crunogena [ 88 ], M. capsulatus [ 97 ], N. europea [ 98 ], and N. oceanii [ 99 ]). This suite of predicted genes includes two outer membrane carbohydrate selective porins of the OPRB family (AFE2522, AFE2250), one carbohydrate transporter of unknown specificity (AFE2312) that is related to the major facilitator superfamily (PF00083, PS50850) and very similar to xylose and galactose proton symporters [ 100 ], and an MFS transporter (AFE1971) with marginal similarity to sugar/nucleoside symporters. It also includes genes for an incomplete PTS system for carbohydrate uptake (AFE3018-23) potentially encoding EII-A, a kinase/phosphatase HprK, an ATPase, an IIA component, a phosphocarrier protein Hpr, and a phosphoenolpyruvate phosphotransferase. However, we could not identify a gene encoding the IIC sugar permease component, thus making it unlikely that A. ferrooxidans has a functional sugar-transporting PTS system. Instead, we suggest that this PTS system could be involved in molecular signaling as part of a regulatory cascade involving RpoN, as described in other proteobacteria [ 101 ]. In this model, a decrease of fixed carbon leads to low levels of phosphoenolpyruvate and cyclic-AMP that in turn maintain most PTS proteins in the dephosphorylated form. This promotes the utilization of glycogen as a carbon source to replenish the phosphoenolpyruvate levels, thus restoring the levels of phosphorylated PTS proteins [ 102 ]. 4. Central carbon metabolism It has been shown in many organisms that the 3-phosphoglyceraldehyde generated by CO 2 fixation via the Calvin cycle enters the Embden-Meyerhof-Parnass pathway, thus providing fixed carbon that can be channeled in either of two directions: for glycogen biosynthesis and storage, or to provide carbon backbones for anabolic reactions. The genes predicted for these two pathways in A. ferrooxidans , together with their reactions and potential interconnections with other biosynthetic pathways, are shown in Figure 6 . Figure 6 Predicted pathways (pentose phosphate pathway, glycolysis, glycogen and interrupted TCA cycle) for central carbon metabolism. 4.1 Carbon storage and utilization The genome also contains genes predicted to encode the five enzymes required for glycogen biosynthesis from fructose-6P. As has been shown in other organisms, glucose-1P-adenylyltransferase ( glgC , AFE2838) is predicted to synthesize ADP-glucose. A specific glycogen synthase ( glgA , AFE2678) would then transfer the glucosyl moiety of ADP-glucose to a glycogen primer to form a new 1,4-glucosidic bond. Subsequently, a branching enzyme ( glgB , AFE2836) is predicted to catalyze the formation of branched 1,6-glucosidic linkages. The carbon stored in glycogen is thought to be released by glucan phosphorylase ( glgP1 , AFE1799; glgP2 , AFE0527), thus regenerating glucose-1P from the non-reducing terminus of the 1,4 chain. The pathway for the conversion of glucose-6P to 2-dehydro-3-deoxy-gluconate is also predicted to be present, except for the last step that replenishes the levels of pyruvate and 3P-glycerate. In addition, a gene encoding phosphoribulokinase was not detected, thus suggesting that either alternate genes encode the missing functions or else that A. ferrooxidans regenerates pyruvate and 3P-glycerate from stored glycogen by alternate pathways. Expression data obtained from A. ferrooxidans growing with sulfur and iron as energy sources have shown that genes involved in glycogen synthesis and utilization are differentially expressed [ 103 ]. Specifically, growth in sulfur-containing media preferentially activates genes involved in glycogen biosynthesis, whereas growth on iron-supplemented media upregulates genes involved in glycogen breakdown. This suggests that A. ferrooxidans channels fixed carbon to glycogen when sulfur is available as an energy source and uses glycogen as a reserve carbon donor when iron is the energy source. 4.2 Carbon backbone formation The genome contains three genes (AFE1802, AFE1676 and AFE3248) that are predicted to encode fructose biphosphate aldolase (EC. 4.1.2.13), the enzyme that catalyzes the formation of fructose-1,6-bP. The interconversion of fructose-1,6-bP to fructose-6P in most heterotrophic bacteria is carried out by fructose biphosphatase and phosphofructokinase enzymes. In A. ferrooxidans , a gene encoding a fructose biphosphatase enzyme was found (AFE0189) that we suggest allows a direct flux of fixed carbon to glycogen storage. A potential phosphofructokinase candidate gene (AFE1807) was also found, a member of the PfkB family of sugar kinases (cd01164). It is located near putative genes involved in glycolysis/glyconeogenesis (e.g., phosphoglycerate mutase and phosphoenolpyruvate synthase), thus generating a bidirectional metabolic path for the utilization/generation of glycogen. Putative genes for all the enzymes involved in the conversion of glyceraldehyde-3-P to pyruvate and acetyl-coA, as well as for the citric acid (TCA) cycle, were detected with the exception of genes encoding the E1-3 subunits of α-ketoglutarate dehydrogenase. Thus, the TCA cycle is incomplete, as has been described in a number of obligate autotrophic bacteria and archaea – a likely hallmark of this lifestyle [ 104 ]. 5. Heavy metal resistance Bioleaching microorganisms, such as A. ferrooxidans , typically live in environments that have high concentrations of soluble heavy metals (e.g., arsenic, mercury, and silver), as well as unusually high concentrations of potentially toxic metals (e.g., copper and iron). This has prompted numerous studies of the mechanisms employed by A. ferrooxidans for metal resistance [ 105 ]. In contrast to the genomic perspective presented herein, those investigations were conducted on multiple strains and thus do not provide a coherent view of the repertoire of heavy metal resistance genes present within one strain. Our genome analysis confirmed the presence of a divergent gene cluster (AFE2857-60) previously identified as involved in arsenic resistance. The cluster includes genes encoding an arsenate reductase (ArsC), the arsenate repressor (ArsR), the divergently-oriented arsenate efflux pump (ArsB), and a hypothetical protein (ArsH). The arsCRB gene cluster was shown to confer resistance to arsenate, arsenite, and antimonium in E. coli , but the function of arsH is unknown [ 106 , 107 ]. Mercury resistance has been investigated in several strains of A. ferrooxidans [ 108 - 110 ]. Genome analysis of the type strain identified three genes potentially encoding the well-described Mer components, i.e., the repressor accessory protein (MerD, AFE2483), the mercury reductase (MerA, AFE2481), and the mercuric ion transporter (MerC, AFE2480). Four candidate genes potentially encoding members of the family of MerR-like transcriptional regulators were also found (AFE2607, AFE2509, AFE1431, and AFE0373). The A. ferrooxidans genome also contains several genes (Additional file 3 ) predicted to be part of heavy metal tolerance systems [ 111 ], including genes for the copCD copper extrusion system, ten clusters of genes predicted to belong to the resistance-nodulation-cell division (RND) family of transporters, three genes encoding cation diffusion facilitator (CDF) proteins, three genes encoding copper translocating P-type ATPases, and two genes encoding other P-type ATPases of unknown specificity. These genome-based predictions offer new opportunities for experimental validation of heavy metal resistance in A. ferrooxidans and also provide new markers for detecting similar genes in other microorganisms 6. Extrusion of toxic organic compounds The ability to extrude toxic organic compounds is widespread, and our inspection of the A. ferrooxidans genome suggests that this bacterium is well equipped to deal with toxic organic molecules. Its genome contains a gene predicted to encode the toluene tolerance protein TtgD (AFE1830) as well as a cluster of proteins often associated with toluene resistance that includes a Tol-Pal-associated acyl-CoA thioesterase (AFE0063) and TolBARQ (AFE0064-67). The genome also includes a predicted complete ABC gene cluster (AFE0158-63) involved in drug extrusion that has significant similarity to the toluene ABC resistance proteins reported in other organisms. Resistance to toluene/xylene and related aromatic hydrocarbons and organic solvents may be needed by A. ferrooxidans when growing in runoff from coal wastes where it might encounter aromatic hydrocarbons [ 112 ] or in bioleaching operation heaps that are irrigated with recycled water containing carboxylic acids and other organic compounds from solvent extraction operations [ 113 ]. An alternative hypothesis for the role of drug-related compound extrusion mechanisms present in microbes associated with biogeochemical cycles has been proposed [ 114 ]. A homolog of TolC has been shown in Shewanella oneidensis MR-1 to excrete anthraquinone-2,6-disulfonate (AQDS) that is used as an extracellular electron shuttle. It has been proposed that AQDS may be particularly important to transfer electrons from cells embedded in the interior of biofilms to reduce Fe(III) present in the solid substrate to which the biofilm is attached. It is possible that a similar mechanism may be used by A. ferrooxidans in the reverse process, namely, to convey electrons from the oxidation of Fe(II) present in solid minerals to cells not in contact with the substrate. Two additional ABC systems potentially involved in drug extrusion are also predicted in the genome, each associated with a HlyD secretion protein family (AFE2861-64, AFE1603-7). The first is directly downstream from the ars genes; the second cluster may have originated through lateral gene transfer since it is flanked by truncated transposases and hypothetical genes and it also exhibits anomalous G+C content. A. ferrooxidans may also be resistant to some antibiotics due to the presence of a two-gene cluster (AFE1977-78) potentially encoding a fosfomidocyn resistance protein and a TonB-family protein, respectively, and also a gene potentially encoding an AmpG permease protein (AFE1961). 7. Stress responses For aerobically growing bacteria, the autooxidation of oxidases in the respiratory chain is the main source of endogenous reactive oxygen species (ROS). Increased levels of ROS can also result from exposure to redox active metals, including iron. Aerobic biomining microorganisms such as A. ferrooxidans that thrive in iron-rich environments are thus expected to be well equipped to deal with disturbances in oxidant-antiooxidant balance. Surprisingly, an unexpectedly low number of genes encoding known ROS detoxification functions were identified in the genome. These genes include a Mn-superoxide dismutase encoded by sodA (AFE1898), two non-identical copies of ahpC -like (AFE1468, AFE0985) and ahpD -like (AFE02014, AFE1814) members of the alkylhydroperoxidase family, and nox , (AFE1803) potentially encoding a NADH oxidase (FAD-dependent pyridine nucleotide-disulfide oxidoreductase family protein). This latter is considered to be important in oxygen scavenging in anaerobes because of its potential to reduce oxygen to water [ 115 ]. No genes coding for known catalases were detected. On the other hand, A. ferrooxidans is predicted to have a complete set of components needed for non-enzymatic neutralization of ROS. This mechanism maintains high levels of low molecular weight thiols in the cytoplasm that, in combination with specific disulfide reductases, provide a reducing intracellular environment and maintain the thiol/disulfide balance of other molecules (unpublished results). Seven distinct thioredoxins ( txr ; AFE2867, AFE2848, AFE2590, AFE2362, AFE1979, AFE0657, AFE0047) and one thioredoxin disulfide reductase ( trxB , AFE0375) are present. Also present are the genes of the glutathione system necessary for glutathione-tripeptide synthesis from the amino acids L-cysteine, L-glutamate, and glycine ( gshA , AFE03064; gshB , AFE03063), four distinct glutaredoxins ( gxr ; AFE3038, AFE2449, AFE2263, AFE0367), and the glutathione reductase gorA (AFE0366). In some bacteria, when basic protection is not sufficient, e.g., when sudden large increases in ROS occur, rapid global responses are induced to cope with the oxidative stress [ 116 ]. Often survival during the period of stress is aided by the simultaneous employment of multiple strategies. The strategies predicted to be available to A. ferrooxidans include repair of oxidative damage (e.g., nfo ), bypassing of damaged functions (e.g., resistant isozymes acnA, fumC ), and the exclusion of oxidative stress agents (e.g., acrAB multidrug efflux pump). Typically, many of these functions are coordinately regulated in response to superoxide by the SoxRS two-component regulator, and in response to peroxide by OxyR in Gram-negative bacteria or by PerR in Gram-positive bacteria. A. ferrooxidans lacks oxyR , soxR , and soxS orthologs, but has a Fur family regulator similar to PerR (AFE1467). The role of PerR in the control of A. ferrooxidans inducible stress response has not been investigated, but could include regulation of the divergently-transcribed AhpC family peroxidase (AFE1468). This arrangement is conserved in other microorganisms [ 117 ]. Other antioxidant defenses that are not controlled by the major oxidative stress regulators include the DNA repair enzyme endonuclease III ( nth , AFE2682), glycoylases ( mutM , AFE2758; mutY , AFE3015), DNA polymerase I ( polA , AFE3094), recombinase protein A ( recA , AFE0932), and other defenses including a peptide methionine sulfoxide reductase ( msrAB , AFE2946-45) and a molecular chaperone ( hlsO , AFE1408). 8. Flagella formation and chemotaxis Conserved fla or fla -related genes that could encode flagella were not identified in the genome, nor were che genes that encode classic chemotaxis functions. These observations conflict with Ohmura et al. (1996) [ 118 ] who proposed that the formation of flagella was a major factor mediating the adhesion of A. ferrooxidans ATCC 23270 to solid sulfur surfaces. This discrepancy could be explained if the fla genes have been lost in the particular culture used for sequencing. Since flagella genes are encoded in a multigene operon in many bacteria, their complete loss might require only one or a small number of excision events. In contrast, the multiple che genes are usually widely dispersed in bacterial genomes and their collective loss in A. ferrooxidans ATCC 23270 would presumably require multiple excision events. Alternative hypotheses to explain this discrepancy include (i) contamination of the A. ferrooxidans ATCC 23270 culture used by Ohmura et al. (1996) by a flagella-bearing microorganism, and (ii) significant differences between the culture used by Ohmura et al. (1996) and that used for our genome sequencing despite their identical designation (ATCC 23270). 9. Adhesion and biofilm formation For mineral-associated bacteria, adhesion and biofilm formation are critical steps for colonization and subsequent mineral solubilization [ 119 ]. Cell surface structures such as pili have been shown to play a critical role in auto-aggregation of microbial cells involved in biogeochemical processes [ 120 ]. A. ferrooxidans contains several gene clusters potentially involved in the formation of a type IV pilus (AFE0967-73, AFE0735-39, AFE0416, AFE0183-6, and AFE0006-7). Some of the relevant genes identified include those for the σ 54 -dependent transcriptional regulator pilR (AFE0185) and for the signal transduction histidine kinase pilS (AFE0184). In addition, candidate tad (tight adherence) genes (AFE2699-AFE2708) were also detected (Additional file 5 ). These genes are responsible for the secretion and assembly of bundled pili. In A. actinomycetemcomitans , they are essential for tight adherence, autoaggregation, and pili formation during colonization of dental surfaces [ 121 ]. They are also present in Thiomicrospora crunogena [ 88 ], a RISCs-oxidizing, chemoautotrophic bacterium found in thermal vents. The multiple copies of genes for pili biosynthesis and adhesion in A. ferrooxidans could enable attachment and colonization on various mineral surfaces, such as pyrite, chalcopyrite, and solid sulfur. The redundancy of related regulatory genes could allow A. ferrooxidans to respond successfully to environmental changes. Genes involved in quorum sensing that were previously identified and characterized include those that encode the classical autoinducer-binding transcriptional regulator LuxR (AFE1997) and the autoinducer synthesis protein LuxI (AFE1999) [ 122 ]. In addition, a second route for the production of homoserine lactones using the act system was predicted based on the presence of a gene encoding an acyltranferase ( act , AFE2584) that was shown to be involved in the production of homoserine lactones of C14 length [ 123 ]. A five gene operon, containing luxA - galE - galK - pgm - galM , was assigned gene numbers AFE1341-45, respectively. This operon has been proposed to be involved in the formation of extra-cellular polysaccharide (EPS) precursors via the Leloir pathway [ 100 ]. GalU (AFE0445) and galT -like (AFE1237) have also been predicted to form part of the Leloir pathway and genes rfbA , B , C and D (AFE3295, AFE0441, AFE3294 and AFE0442, respectively) have been proposed to be involved in the biosynthesis of the EPS precursor dTDP-rhamnose. These groups of genes have been postulated to be involved in biofilm formation in A. ferrooxidans and their patterns of transcription were characterized in growth media with and without organic carbon supplementation [ 100 ]. 10. Genetic transfer A region of the genome (AFE1013-AFE1387, Figure 2 ) is enriched (84% versus 54% for the rest of the genome) in putative genes encoding hypothetical proteins, genes for DNA metabolism and sequences related to mobile elements such as transposases, plasmids, and bacteriophage (phage), and pseudogenes. The presence of site-specific recombinases and phage integrases in this region, as well as in other regions such as AFE2397-99, AFE0833-35 and AFE0507-9, indicates that A. ferrooxidans has been the target of phage infection. Although no phages are currently known to infect A. ferrooxidans , this finding suggests that further searching might be fruitful. Such phage could facilitate study of the mechanisms of viral infection in extremely acidic conditions, as well as serve as useful transducing agents for the genetic manipulation of A. ferrooxidans , as have been shown for the acidophilic archaeon, Sulfolobus spp. [ 124 , 125 ]. The genome contains clusters of genes whose sequence and gene order show significant similarity to both the Trb system of the Ri plasmid from Rhizobium rhizogenes and the Ti plasmid from Agrobacterium tumefaciens [ 126 ]. Most of these predicted genes potentially encode structural proteins of the type IV secretion system involved in conjugative DNA transfer. However, missing from the genome are the trbC , trbH , and trbK genes that encode an inner membrane lipoprotein, a pili structural protein, and a protein involved in plasmid immunity, respectively. Notably, one of two copies of a trbG -like gene is located within a highly conserved cluster in a position usually occupied by trbH , and it may assume the role of this missing gene. The absence of critical components of the conjugation system suggests that A. ferrooxidans ATCC 23270 has lost the capacity to carry out conjugation via the Trb mechanism. The question arises as to the origin of the Ti plasmid-like sequences in A. ferrooxidans . One possibility is that it was acquired from an Agrobacterium -related microorganism or an ancestor of such through conjugation. A. ferrooxidans and a free-living or plant root-associated Agrobacterium might share the same environment at the interface of acidic drainages and anaerobic soils/water. Ten proteins predicted to be involved in plasmid stability and maintenance are present in the genome. This discovery, coupled with the detection of an extensive suite of predicted conjugation-related genes, provides additional evidence that A. ferrooxidans was capable of undergoing conjugation. Even though no natural conjugation partners are known, conjugation between E. coli and A. ferrooxidans has been achieved in the laboratory. The frequency of detectable marker transfer has been very low [ 95 , 127 , 128 ], and must be increased before this technique can be used for widespread genetic manipulation of A. ferrooxidans . Our finding of conjugation-related genes could stimulate further attempts. Forty-one IS elements were identified, of which thirty-three could be classified as members of nine families according to the scheme of Mahillon and Chandler [ 129 ] (Additional file 3 ). The largest groups, designated here as ISafe3 (8 copies) and ISafe4 (3 copies), belong to the IS110 and IS3 families, respectively [ 129 ]. ISafe1, which is associated with phenotypic switching in A. ferrooxidans ATCC 19859 [ 130 ], was not detected in the genome of the type strain. Two non-identical copies of a Tn5468 transposon (family Tn7-like) were detected, each containing tnsABCDorf5 (AFE1201-AFE1205, AFE3199-95). The first copy is embedded in a suite of genes encoding hypothetical proteins; the second is associated with the atp operon and the glmSU as described for A. ferrooxidans ATCC 33020 [ 131 ]. 11. Predicted osmotic balance and potential pH tolerance mechanisms Acidophiles exhibit functional and structural properties that allow them to survive and proliferate in extremely acidic environments (pH 3 or below) [reviewed in 132]. These include: a) impermeable cell membranes (mostly in archaea); b) selective outer membrane porins; c) the generation of positive internal potential (Δψ) to create a chemosmotic barrier inhibiting proton influx; and d) the removal of excess internal protons by active proton pumping. A putative gene ( omp40 , AFE2741) was identified that had significant similarity to an outer membrane porin found in A. ferrooxidans strain ATCC 19859 (133, 134). A large external, positively-charged loop has been predicted in Omp40 that may control pore size and ion selectivity at the porin entrance and may constitute a potential proton barrier [ 133 , 134 ]. In addition, the following related functions were predicted (Additional file 3 ): several potassium transporters including one K+ channel, one K+ uptake protein and one K+ efflux transporter; two copies of an ABC potassium import system that could be involved in the generation of a positive internal potential inhibiting proton influx; four Na/H+ antiporters and two proton P-type ATPases that could extrude excess internal protons. These predictions suggest specific areas for future experimental validation."
} | 14,983 |
36658394 | PMC10030642 | pmc | 2,346 | {
"abstract": "Closely interacting microbial species pairs (e.g., predator and prey) can become coadapted via reciprocal natural selection. A fundamental challenge in evolutionary ecology is to untangle how coevolution in small species groups affects and is affected by biotic interactions in diverse communities. We conducted an experiment with a synthetic 30-species bacterial community where we experimentally manipulated the coevolutionary history of a ciliate predator and one bacterial prey species from the community. Altering the coevolutionary history of the focal prey species had little effect on community structure or carrying capacity in the presence or absence of the coevolved predator. However, community metabolic potential (represented by per-cell ATP concentration) was significantly higher in the presence of both the coevolved focal predator and prey. This ecosystem-level response was mirrored by community-wide transcriptional shifts that resulted in the differential regulation of nutrient acquisition and surface colonization pathways across multiple bacterial species. Our findings show that the disruption of localized coevolution between species pairs can reverberate through community-wide transcriptional networks even while community composition remains largely unchanged. We propose that these altered expression patterns may signal forthcoming evolutionary and ecological change.",
"introduction": "Introduction Coevolution is the reciprocal selection imposed by pairwise or multi-way ecological interactions between species [ 1 ]. The coevolutionary process is a major force producing phenotypic and genetic diversity both at the microevolutionary scale [ 2 , 3 ] and across the wider tree of life [ 4 , 5 ]. Much about coevolution has been learned from studying microbes whose large population sizes, fast generation times, and relatively high mutation rates allow scientists to observe evolution in real-time as it unfolds [ 6 ]. Many microbial phenotypes (e.g., antimicrobial production or parasite resistance) have been shaped by antagonistic coevolution whereby hosts/prey evolve resistance to their parasites/predators [ 7 ]. Antagonistic coevolution between microbial predators and prey likely affects many ecological and evolutionary processes, including population dynamics [ 8 ], the maintenance of local genetic diversity [ 9 , 10 ], and the enrichment of biodiversity across spatially heterogeneous landscapes [ 11 ]. Coevolution between microbial grazers and prey promotes phenotypic and genotypic diversity in marine bacterial populations by selecting for altered cell size, shape, lifestyle, or physiochemical properties of the cell surface, which can increase bacterial survival [ 9 ]. For example, phagotrophic protistan grazers have a critical role in controlling the standing stock of bacterial populations [ 12 , 13 ] and are a significant link in the transfer of dissolved organic carbon from heterotrophic bacteria to higher trophic levels in many microbial ecosystems [ 14 ]. Because antagonistic coevolution can drive reciprocal phenotypic diversification of prey/hosts and predators/parasites, this may be a critical process generating intraspecific diversity and driving community composition [ 15 – 17 ]. Intraspecific diversity in a focal species has been shown to alter microbial community composition to a comparable extent to removing a predator [ 18 ] or the focal species itself [ 19 ]. Phenotypic traits shaped through coevolution may also have pleiotropic consequences for ecological functions unrelated to predator/parasite sensitivity. For example, coevolution between a marine flavobacterium species and two viruses altered the suite of carbon compounds used by the bacterium [ 20 ], while rapid resistance evolution in a marine cyanobacterial host has been shown to reduce the effect of viral lysis on dissolved nutrient recycling community stoichiometry [ 21 ]. These studies demonstrate that localized coevolution between species pairs can influence microbial community function and composition. However, key questions about the consequences of localized coevolution for ecosystem function and composition remain. Here ask how the coevolution of trophic antagonism between a microbial predator and prey species affects a larger web of interacting microbial species. We experimentally manipulated the coevolutionary history of a focal bacterial prey species and ciliate predator in the presence of a bacterial community while measuring cell densities, metabolic potential (per-cell ATP concentration), community composition, and community gene expression (Fig. 1 ). Based on past studies showing a significant effect of local adaptation on community composition [ 18 , 19 ], we predicted that altering the coevolutionary history of the predator/prey focal pair would drive overall community composition with largely independent and strain-specific transcriptional responses. Instead, we found that the coevolutionary mismatch between the focal predator and prey species had little impact on community composition but a substantial effect on community-wide transcriptional networks and ecosystem metabolic potential. We conclude by discussing how changes in transcriptional networks may be relevant for eco-evolutionary processes in general. Fig. 1 Experimental microcosm overview. A Coevolution of Pseudomonas fluorescens SBW25 and Tetrahymena thermophila in long-term selection lines (mean ± s.e.m. normalized to 0–1, data are from a prior publication [ 25 ]). Coevolved ciliate and bacteria were isolated after ~740 generations for use in the main experiment. B Treatment scheme of the main experiment. Each box represents the 30 clonal species bacterial community where colors are treatments: bacterial community + ancestor SBW25 without predation (light blue), bacterial community + coevolved SBW25 without predation (dark blue), bacterial community + ancestor SBW25 with coevolved Tetrahymena (light green), bacterial community + coevolved SBW25 with coevolved Tetrahymena (dark green). C Sampling protocol. Each treatment from B was performed in triplicate in 100 ml growth medium. On day 41, fresh media was added to the microcosm. DNA and RNA samples were collected on days 4 and 45. The experiment was terminated after 55 days.",
"discussion": "Discussion Our results demonstrate how the effects of antagonistic coevolution between a focal species pair may cascade through microbial communities by altering multi-species transcriptional networks. We expected that decoupling the coevolutionary history of Pseudomonas fluorescens SBW25 from its ciliate predator would perturb bacterial species composition as has been observed before in other microbial communities [ 18 , 19 , 64 ], while RNA sequencing would reveal the functional mechanisms underlying those ecological changes. Instead, we found that the coevolved focal prey species, by itself, had a negligible effect on community structure, which was only significantly altered by the coevolved predator. However, the coevolved prey species had a significant effect on the expression profiles of other bacteria contingent upon the presence of the coevolved predator. This coevolution-dependent effect was also evident in significantly higher community metabolic potential in the microcosms with both coevolved focal species. Later in the experiment, when predator densities dropped considerably, each treatment followed its own ecological and transcriptional trajectory. Antagonistic coevolution between bacterial hosts and viruses can generate pleiotropic effects like metabolic aberrations, reduced growth, and altered susceptibility to other hosts/viruses when the environmental context changes [ 65 ]. Coevolution between bacterivorous protists and bacterial prey likely also generates pleiotropic effects analogous to those between hosts and viruses [ 9 ]. However, the impacts of local coevolution and any potential pleiotropic effects on wider community structure and function are not understood. We observed that many mutations from coevolved SBW25 were in genes related to defense traits (e.g., motility, biofilm formation, and small molecule biosynthesis), so we expected these genes to be differentially expressed by coevolved SBW25 in the community experiments. However, none of the SBW25 genes with parallel mutations from the long-term coevolution lines (Fig. 2 ) were differentially expressed in response to experimental treatments. Indeed, very few genes from any metabolic pathway related to motility or defense phenotypes were differentially expressed in coevolved SBW25. Instead, most differentially regulated genes were involved in nutrient assimilation, sulfur metabolism, and ATP biosynthesis/oxidative phosphorylation. The differentially expressed pathways in the bacterial community were related to sulfur and carbon acquisition with the exception of type 3 fimbriae in Citrobacter koseri which contribute to cell adhesion and biofilm formation. Taken together, these patterns show how antagonistic coevolution between one species pair can generate transcriptional diversity across a community of interacting species. This finding represents a link between intraspecific diversity and interspecific phenotypic diversity in microbial ecosystems. We did not expect organic sulfur metabolism genes (e.g., aliphatic sulfonates) to be differentially regulated in response to SBW25 and Tetrahymena coevolution. Inorganic sulfate concentrations in our minimal medium were low (40 μmol l −1 ) relative to commonly used bacteria media (LB; 150 μmol l −1 with excess cysteine) and overlapped with sulfate concentrations in natural soils [ 66 ]. We speculate that this growth medium was nearly sulfur deficient for the bacterial community and that changes in the coevolved focal pair altered the bioavailability of sulfur in the growth medium, causing downstream effects related to sulfur scavenging [ 59 ]. One possibility is that the combined metabolisms of the coevolved ciliate and SBW25 depleted inorganic sulfur concentrations to the extent that it triggered a community-wide sulfur starvation response. Alternatively, the bacterial community may have been responding to aliphatic organosulfonates compounds derived from the Tetrahymena predator. For example, Taurolipids, which contain an aliphatic organosulfonate head group, are characteristic lipids of Tetrahymena species [ 67 ]. This experiment was designed to test how localized coevolution between a focal bacterial species and its ciliate predator affected wider community dynamics and function in a multi-species ecosystem. However, coevolution often occurs within networks of multiple interacting species that vary over time and space [ 11 ]. The spatial and temporal heterogeneity of species interactions is an added layer of complexity that needs to be evaluated in subsequent work. However, our findings may be relevant for understanding species’ colonization of new environments. The order and timing of colonization events can affect community assembly and function [ 68 , 69 ], and our findings add to our understanding of this process by highlighting the potential importance of the coevolutionary history of immigrants. More broadly, our study shows that localized coevolution between a single species pair over ecological time scales can drive functional changes in multi-species transcriptional networks hidden beneath a relatively static community structure. These shifts in community gene expression may be important precursors to the emergence of future adaptive alleles and potentially subsequent species evolution [ 70 ]. Thus, our findings also point to gene regulation as a potentially important component of eco-evolutionary change."
} | 2,936 |
39905014 | PMC11794851 | pmc | 2,347 | {
"abstract": "The seamless integration of rigid/flexible electronic components into stretchable substrates is imperative for the realization of reliable stretchable electronics. However, the transition from flexible-to-stretchable substrates presents inherent challenges in interfacial behavior, predominantly arising from disparities in elastic moduli, thereby rendering their integration arduous for practical deployment. Here, we introduce a bioinspired interface-engineered flexible island (BIEFI), which effectively facilitates the creation of highly stretchable electronics by optimizing the interface with flexible mechanical interlocking mechanisms, resilient to physical deformations. Various electronic components, such as light-emitting diodes (LEDs) and solar cells, are affixed onto the flexible island, showcasing its versatility as a robust platform for rigid components while ensuring the entire substrate maintains high stretchability. Additionally, a smart workout monitoring system is demonstrated by integrating a resistance band with a flexible-to-stretchable platform. This approach seamlessly integrates a wide range of rigid, flexible, and stretchable components, ensuring durability under diverse physical deformations.",
"introduction": "Introduction Flexible and stretchable electronics have been a core focus of emerging research fields, which show high potential for shaping the future electronics world 1 – 4 . Flexible electronics, which enable circuits and electronic components, show their functionality under limited physical deformations such as bending 5 , 6 . Stretchable electronics enable safe operation under diverse physical deformation modes such as stretching, twisting, and so on 7 . The potential of stretchable electronics opens a new road for electronics, such as wearable devices 8 – 10 , displays 11 , 12 , and energy harvesters 13 – 15 . Although numerous progresses have been made in developing stretchable electronics, integrating rigid, flexible substrates into the stretchable substrate remains challenge. The mismatch in mechanical properties, particularly the difference in elastic moduli between rigid and stretchable substrates, often leads to performance failures when subjected to stretching. Addressing this issue requires innovative designs that can effectively isolate and distribute strain, minimizing the concentration of mechanical stress at the interface between rigid and stretchable materials. There have been several efforts with different approaches to enhance stretchability by modifying the geometry (e.g., serpentine, kirigami, and rigid island) 16 , 17 . The serpentine and kirigami approaches have used flexible substrates such as polyimide (PI) and polyethylene terephthalate (PET) to build stretchable electronics 18 , 19 . In these examples, the flexible substrate is strategically designed and fabricated to achieve stretchability. However, these solutions are limited in their ability to fully mitigate the strain mismatch at the rigid-stretchable interface, leading to failure over extended use. On the other hand, the rigid island approach is used to minimize the mismatch of the elastic modulus of the rigid island and stretchable substrate with several strategies, such as material-based and interfacial engineering-based methods. Material-based approaches use material’s stiffness gradients to lower the discrepancy between high and low elastic moduli to increase the stretchability of the device 20 – 23 . The challenges in previous studies include issues with scalability, as complex multi-step fabrication processes limit practicality for large-scale applications. Additionally, maintaining long-term durability and stability under repeated strain has been a significant obstacle, leading to performance degradation over time in some designs. Mechanical mismatches between rigid and soft materials, as well as difficulties in embedding rigid components into flexible substrates, further hinder overall flexibility and application potential in many approaches. A recent study of an interfacial engineering-based approach utilized the mechanical interlocking method 24 . However, the proposed rigid island made from polyactic acid (PLA) demonstrated low flexibility and significant thickness, restricting its use in various flexible and stretchable electronics applications (Supplementary Table 1 ). In nature, plant roots anchor themselves deeply into the soil to enhance the interface (roots-soil interlocking), thereby reducing the mismatch between the rigid plant body and the soft soil structure 25 . Moreover, the root structure contributes to standing firmly on the ground and enduring harsh conditions: weather and natural disasters. Roots-soil interlocking postulates high adhesion and secure interfacial behaviors and is also shown in mimicking applications such as stretchable electrodes 26 and structural batteries 27 . Herein, we develop bioinspired interfacial engineered flexible islands (BIEFI) demonstrating high stretchability and establishing excellent physical deformational performances for flexible-to-stretchable electronics (Fig. 1a ). Inspired by roots-soil structure, the flexible-to-stretchable interface is designed using a flexible substrate with root structures embedded within a stretchable substrate (Fig. 1b ). By conducting a parametric study in the BIEFI design, optimized to facilitate a flexible interlocking mechanism, significant improvement in stretchability up to 700% strain can be achieved. (Fig. 1c ). The flexible mechanical interlocking mechanism extends the fatigue life for various types of physical deformations (e.g., strain, poking, and twisting). The optimized BIEFI design demonstrates applications of flexible-to-stretchable platform (e.g., stretchable LED array and solar cell array) under diverse physical deformations (Fig. 1d ). A smart resistance band (SRB) was developed to show highly stretchable applications by integrating the optimized BIEFI into a resistance band that monitors workout with a strain sensor and accelerometer (Fig. 1e ). The proposed strategy facilitates a new perspective on the integration of rigid, flexible, and stretchable components with optimized BIEFI design for high stretchability and endurance for flexible-to-stretchable platform. It also paves the way for exploring the commercial feasibility of stretchable electronics. Fig. 1 Bioinspired Interface-Engineered Flexible Island (BIEFI) within stretchable substrate for highly stretchable flexible-to-stretchable platforms. a Schematic illustration of flexible-to-stretchable substrate interface with bioinspired interfacial engineering. b Roots-soil inspired flexible-to-stretchable interface design. c Stress versus strain comparison for the Rootless Island (RLI) (gray line) and BIEFI (red line) embedded in Ecoflex under stretching. d Schematic of BIEFI applications in the flexible-to-stretchable platform for stretchable electronics applications with diverse electronic devices. e Schematic of application of BIEFI for highly stretchable electronics: Smart resistance band for workout monitoring. Source data are provided as a Source Data file.",
"discussion": "Discussion In this study, we proposed a BIEFI design for a flexible-to-stretchable platform to achieve high stretchability and superior usage life. We systematically investigated the effect of parameters on roots-soil-inspired design, such as primary roots, secondary roots, and width of the roots. Notably, the effect of stress distribution (primary roots) and flexible mechanical interlocking (secondary roots) delay and suppress the interfacial failure in the flexible-to-stretchable substrate interface. Importantly, the optimized BIEFI design realizes superior extended fatigue life under diverse physical deformations. LEDs and solar cells were utilized to demonstrate the diverse applicability of the BIEFI on flexible-to-stretchable electronics. Furthermore, a smart resistance band was developed to quantify workouts by integrating electronics into the BIEFI within the resistance band. Our research provides a promising strategy for developing and commercializing stretchable electronics in the coming years. Despite the accomplishments mentioned earlier, there are still opportunities to enhance the proposed flexible-to-stretchable platform through various approaches. First, a unidirectional parameter study is conducted and demonstrated in the applications, exhibiting the feasibility of omnidirectional design. This suggests an opportunity to design an omnidirectional interface with relevant parameter studies. The omnidirectional interface design will expand more potential application opportunities to the flexible-to-stretchable platform. Second, in this study, the effect of the adhesion parameter on stretchability was not explored. Nevertheless, the adhesion force between PI and Ecoflex substrate can be investigated through various methodologies. The increased adhesion of PI to elastomer will enhance the gripping effect of the roots, leading to improved strain at failure value. Third, the proposed root designs are relatively simplistic, consisting of only a straight configuration of primary and secondary roots. In reality, plant roots show a high degree of complexity, which allows them to have a strong gripping effect on the soil. Mimicking this complex design could also provide a higher degree of strain at failure value and stability performance for the flexible-to-stretchable platform. The developed highly stretchable and durable flexible-to-stretchable platform with BIEFI provides a promising strategy for integrating rigid, flexible, and stretchable components; thereby, the platform can be used in many various applications, such as stretchable displays, energy storage devices, medical devices, and so on. Our strategy promises a future in which the flexible-to-stretchable platform effortlessly elevates and redefines the electronics world."
} | 2,485 |
24808856 | PMC4010740 | pmc | 2,348 | {
"abstract": "In 1943 McCulloch and Pitts suggested that the brain is composed of reliable logic-gates similar to the logic at the core of today's computers. This framework had a limited impact on neuroscience, since neurons exhibit far richer dynamics. Here we propose a new experimentally corroborated paradigm in which the truth tables of the brain's logic-gates are time dependent, i.e., dynamic logic-gates (DLGs). The truth tables of the DLGs depend on the history of their activity and the stimulation frequencies of their input neurons. Our experimental results are based on a procedure where conditioned stimulations were enforced on circuits of neurons embedded within a large-scale network of cortical cells in-vitro . We demonstrate that the underlying biological mechanism is the unavoidable increase of neuronal response latencies to ongoing stimulations, which imposes a non-uniform gradual stretching of network delays. The limited experimental results are confirmed and extended by simulations and theoretical arguments based on identical neurons with a fixed increase of the neuronal response latency per evoked spike. We anticipate our results to lead to better understanding of the suitability of this computational paradigm to account for the brain's functionalities and will require the development of new systematic mathematical methods beyond the methods developed for traditional Boolean algebra.",
"conclusion": "Conclusion We proposed a new computational paradigm in which the brain consists of dynamic logic gates (DLGs) which are governed by time-dependent logic modes. The relevance of our work to the brain's functionalities has to be evaluated using many aspects including: (a) Do DLGs exist in the dynamics of a network of interconnected neurons? (b) Is the concept of DLGs robust to population dynamics and specifically to recurrent networks? (c) Is DLGs a mechanism which the brain could plausibly use to any extent and especially when it is critically rely on precise relative timing of neural activities? (d) Can one find a realistic learning mechanism, e.g., Hebb's rules, to implement DLGs? The brain is composed of large neural networks, where neurons are interconnected via excitatory and inhibitory synapses as well as sub-threshold and above-threshold synapses. In the events of weak synapses, spatial and temporal summations of excitations are required to generate an evoked spike. Hence, the examined gate architectures have to be locally embedded in such large interconnected networks. The existence of weak synapses with high probability indicates that complex DLGs, where several input chains exist, are also expected to be a common building block of such networks. We verified that the phenomenon of DLGs is robust to population dynamics and hence it is expected to be less sensitive to unexpected fluctuations in the response timings of a single neuron. However, there are many unavoidable effects of brain activity which are not assumed to carry any significant information, e.g., synaptic noise. Is the DLGs one of these unavoidable effects? The answer is not yet clear, however, we showed that the increase in the neuronal response latency to ongoing stimulations cannot be ignored, as it may double its value and therefore affect the time dependent connectivity of a recurrent network. As for the implication of such DLGs to cognitive activities, we demonstrated some preliminary tasks such as edge detections, which obviously can be generalized to more complex tasks. Nevertheless, our work is a call for advanced in-vivo experiments and theoretical studies, which can pinpoint the existence and the importance of the suggested DLGs in various functionalities of the brain. Moreover, the proposed mechanism of DLGs opens a manifold of theoretical questions regarding advanced paradigm for the brain activity including the search for efficient local learning rules for the DLGs. It is evident that the variety of possible DLGs is much larger than the abovementioned examples. For recurrent networks, the complexity is expected to be enhanced in comparison to feedforward networks. As opposed to feedforward networks with given simultaneous external stimulations, in recurrent networks the timings of the input stimulations are a function of the large scale activity of the entire network. One of the open theoretical questions is the number of realizable logic operations among P N , where each one of the N gates has P operating modes. On mathematical grounds, the key question is whether recurrent networks consisting of DLGs might go beyond the computation paradigm of the universal Turing machine (Turing, 1938 ; Maini et al., 2006 ; Dayan, 2009 ; Hodges, 2012 ). This challenge requires a careful mathematical definition and in particular, a definition of whether the stretching of the neuronal response latency has to be taken as continuous or discrete in comparison to the delays. Such networks represent a class of heterogeneous time-delayed networks composed of excitable units, where the delays are a function of the activity of the network itself. Practically, the question is whether a circuit composed of such new elements can be analyzed using the traditional systematic methods and tools developed for Boolean circuits. In the event that the presented dynamics is within traditional computational complexity, i.e., can be implemented using conventional computers, an interesting question is its advantages with respect to the implementation of the brain's functionalities. 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 This year we are celebrating the 70th anniversary of the publication of the seminal work by Warren S. McCulloch, a neuroscientist, and Walter Pitts, a logician, entitled “A logical calculus of the ideas immanent in nervous activity” (Mcculloch and Pitts, 1943 ). They attempted to understand how the brain could produce highly complex patterns by using many interconnected building blocks of the brain, the neurons. In their model, the brain is composed of Boolean entities functioning as threshold units. Such simplified units constitute pure and reliable logic-gates (e.g., AND, XOR), similar to the logic at the core of today's computers. The generalization of this simplified Boolean framework to include unreliable elements has emerged in 1956 by the innovative work of John von Neumann (Von Neumann, 1956 ). These concepts as well as the earlier pioneering work of Claude Shannon to simplify Boolean circuits (Shannon, 1938 ) are at the cornerstone of today's computational paradigm (Turing, 1938 ). The computational framework of McCulloch and Pitts had a tremendous impact on the development of artificial neural networks (Hopfield, 1982 ; Krogh, 2008 ; Qian et al., 2011 ; Gerstner et al., 2012 ; Gilja et al., 2012 ) and machine learning theory (Sutton and Barto, 1998 ; Hunt et al., 2012 ). Their concept triggered the next major development in theoretical neural networks when in 1958 Frank Rosenblatt introduced the concept of the perceptron (Rosenblatt, 1958 ), the prototypical linear classifier, which ever since has been theoretically investigated and generalized to more structured multi-layer and recurrent architectures (Litwin-Kumar and Doiron, 2012 ; Stoianov and Zorzi, 2012 ). Nevertheless, it is fair to conclude that the concept of simplified Boolean neurons had a limited impact on neuroscience, which exhibit much richer temporal dynamics (Izhikevich, 2006 ; Izhikevich and Hoppensteadt, 2009 ; Gal et al., 2010 ; Vardi et al., 2012a ). Moreover, it appears that the brain is the most ineffective environment to implement such a Boolean logical operating system, comprised of static logic-gates (SLGs). Seven decades after the proposed neuronal paradigm by McCulloch and Pitts, the fundamental concept of the computational abilities of the nervous system remains unclear (Hodges, 2012 ). On the one hand, one might conclude that the search for a comprehensive computational logic framework is irrelevant, as specialization in specific behavioral and perceptual tasks requires different “operating systems.” On the other hand, it is evident that the “hardware” implementations of all complex brain tasks are composed of similar basic interconnected building blocks (neurons) having many features in common, which are enhanced and possibly dominant when operating as an ensemble (Abeles, 1991 ). In the present study, we extend the recently demonstrated new experimentally corroborated paradigm in which the logical operations of the brain differ from the logic of computers (Vardi et al., 2013b ). Unlike a burned logic-gate on a designed chip that consistently follows the same truth-table, here the functionality of the brain's logic-gates depend on the history of their activity, the stimulation frequencies of their input neurons, as well as on the activity of their interconnections. Our results are based on an experimental procedure where conditioned stimulations were enforced on circuits of neurons embedded within a large-scale network of cortical cells in-vitro (Marom and Shahaf, 2002 ; Morin et al., 2005 ; Wagenaar et al., 2006 ; Vardi et al., 2012b ). We demonstrate that the underlying biological mechanism is the unavoidable increase of neuronal response latencies to ongoing stimulations (Aston-Jones et al., 1980 ; De Col et al., 2008 ; Ballo and Bucher, 2009 ; Gal et al., 2010 ; Soudry and Meir, 2012 ), which imposes a non-uniform gradual stretching of delays associated with the neuronal circuit (Kanter et al., 2011 ; Vardi et al., 2012a , 2013a , c ). To further support and expand the limited experimental results, we present a straightforward theoretical model based on the assumption of identical neurons with a constant increase in their neuronal response latency per evoked spike. This model, corroborated with simulations, allows us to explore the behavior of more complex structured neuronal DLGs in addition to SLG (Vogels and Abbott, 2005 ). We anticipate our results to be a starting point for larger scale in-vitro experiments and structured recurrent neuronal circuits, which will lead to a better understanding of the suitability of this computational paradigm to account for the brain's functionalities. In addition, this paradigm will require the development of new systematic methods and practical tools beyond the methods developed for traditional Boolean algebra (Chavesa et al., 2005 ; Nahin, 2012 )."
} | 2,661 |
32183145 | PMC7183087 | pmc | 2,349 | {
"abstract": "The combination of the triboelectric effect and static electricity as a triboelectric nanogenerator (TENG) has been extensively studied. TENGs using nanofibers have advantages such as high surface roughness, porous structure, and ease of production by electrospinning; however, their shortcomings include high-cost, limited yield, and poor mechanical properties. Microfibers are produced on mass scale at low cost; they are solvent-free, their thickness can be easily controlled, and they have relatively better mechanical properties than nanofiber webs. Herein, a nano- and micro-fiber-based TENG (NMF-TENG) was fabricated using a nylon 6 nanofiber mat and melt blown nonwoven polypropylene (PP) as triboelectric layers. Hence, the advantages of nanofibers and microfibers are maintained and mutually complemented. The NMF-TENG was manufactured by electrospinning nylon 6 on the nonwoven PP, and then attaching Ni coated fabric electrodes on the top and bottom of the triboelectric layers. The morphology, porosity, pore size distribution, and fiber diameters of the triboelectric layers were investigated. The triboelectric output performances were confirmed by controlling the pressure area and basis weight of the nonwoven PP. This study proposes a low-cost fabrication process of NMF-TENGs with high air-permeability, durability, and productivity, which makes them applicable to a variety of wearable electronics.",
"conclusion": "4. Conclusions In summary, we have developed a nano- and micro-fiber-based TENG (NMF-TENG) by using a nylon 6 nanofiber mat and melt blown nonwoven PP as the triboelectric layers to enhance the electrical performance. The NMF-TENG was fabricated by electrospinning nylon 6 on the nonwoven PP with Ni coated fabric electrodes. Owing to the porous structure of the triboelectric layers containing a large volume of air with a difference in electronegativity, our NMF-TENG could generate triboelectric effects even with very thin layers and without a spacer. The morphology, porosity, pore size distribution, and fiber diameters of the triboelectric layers were characterized. Moreover, the electrical output performance of NMF-TENGs was investigated. When PP15, PP30, and PP50 were applied to the triboelectric layer with a nylon 6 nanofiber mat, the electrical output performance of the NMF-TENG was improved with the increase in PP basis weight from 15 to 50 gsm. Furthermore, the electrical charge potential realized by hand tapping with different pressure areas was 33.93 V for palm, 23.10 V for blade, and 12.52 V for finger tapping. It means that the charge potential of the inner friction area is increased through the increase in pressure area. To demonstrate the capability of the NMF-TENG for energy harvesting, it was used as a power source to turn on LEDs. The triboelectric output performance of NMF-TENGs with high permeability provides the opportunity to apply them to smart and wearable device power systems and self-generated electronic systems. The large area of the NMF-TENG and its application to clothing will be studied in the near future.",
"introduction": "1. Introduction Fiber-based electronic devices are attracting extraordinary attention due to their flexibility, lightness, comfortableness, and applicability to a variety of industries and products, including physical or chemical sensors [ 1 , 2 ], biomedical monitoring [ 3 , 4 ], soft robotics [ 5 , 6 ], and wearable devices [ 7 , 8 ]. Furthermore, various portable smart devices are playing major roles in daily life, and these devices require lighter, smaller, or larger capacity power sources. Related research topics such as sensors [ 9 , 10 ], energy harvesting [ 11 , 12 , 13 , 14 ], piezoelectric nanogenerators (PENGs) [ 15 , 16 , 17 ], and triboelectric nanogenerators (TENGs) have received significant attention [ 18 , 19 , 20 , 21 , 22 , 23 ] because a large amount of available energy is generated by the human movement and clothing friction. The triboelectric effect is caused by the contact between different materials, and it can induce strong electrostatic charges. As static electricity can lead to ignition, dust explosions, and electrical shocks, the unintended static electricity is generally considered to have a negative impact in the industry. However, the combination of the triboelectric effect and static electricity has been extensively studied as a TENG [ 24 , 25 , 26 , 27 , 28 ] because it produces a sufficiently large amount of electrical energy to be used as a generator. An all-nanofiber-based stretchable TENG (S-TENG) with polyvinylidene fluoride (PVDF) and thermoplastic polyurethane (TPU) nanofiber membranes was reported by Zhao et al. [ 29 ] for energy harvesting. This S-TENG, which according to the analysis had a full separation of surface-to-surface, had an excellent triboelectric output performance. Zhu et al. [ 30 ] introduced the microfiber-based TENG in 2018. This TENG consisted of ZnO-coated polypropylene (PP) microfibers with a spacer, and it exhibited high transfer charge and output voltage. The properties of a TENG depend on various elements, such as the surface morphology, dielectric constant, spacer, and triboelectric potential difference between the triboelectric materials [ 31 , 32 , 33 , 34 ]. Since the fiber-based TENG was introduced by Zhong et al. in 2014 [ 35 ], related studies have been actively reported [ 15 , 16 , 36 ]. Notably, the TENG using nanofibers has been widely investigated because of its advantages, such as ease of production by electrospinning and high surface roughness [ 37 , 38 , 39 , 40 , 41 ]. Owing to the porous structure of nanofibers, which can contain a large volume of air with high dielectric constant, they have a large contact area, which can enhance the triboelectric effects and produce a high-output generator [ 42 , 43 , 44 ]. However, TENGs are generally designed with structures that include spacers to enhance the electrical performance, and this leads to increased fabrication process steps, cost, and total volume of devices. In addition, nanofibers can be produced only by electrospinning with high-cost and limited yield, which also results in relatively low mechanical properties due to the difficulty in improving strength through the orientation of polymers [ 45 ]. The melt blowing process is a well-known method of producing nonwoven fabrics and can be applied to various thermoplastics, including polyethylene terephthalate (PET) [ 46 ], polyolefin [ 47 ], and polylactic acid (PLA) [ 48 ]. Melt blown nonwoven fabrics, which generally consist of microfibers, can be produced with low cost on a mass scale in large areas, are solvent-free, their thickness is easily controlled, and have relatively better mechanical properties than nanofiber webs [ 49 , 50 ]. In this study, in order to maintain and mutually complement the advantages of nanofibers and microfibers and avoid their shortcomings, a nano- and micro-fiber-based TENG (NMF-TENG) was fabricated using a polyamide (PA), especially nylon 6, nanofiber mat and melt blown nonwoven polypropylene (PP) as triboelectric layers. The NMF-TENG is composed of the nylon 6 solution electrospun on the nonwoven PP and Ni-coated fabric electrodes. The nanofiber mat and nonwoven PP contain a large volume of air in the porous structure, and thus, the NMF-TENG does not require a spacer. The morphology, porosity, pore size distribution, and fiber diameters of the triboelectric layers were characterized. Moreover, the electrical output performances of NMF-TENGs were investigated. This study proposes a low-cost fabrication process of NMF-TENGs with high triboelectric output performance, air-permeability, durability, and productivity. Therefore, their application to a variety of wearable electronics is expected.",
"discussion": "3. Results and Discussion The manufacturing process of the NMF-TENG is shown in Figure 1 a. In the first step, nonwoven PP was manufactured in the pilot scale equipment. After that, the nanofiber mats were electrospun on the nonwoven PP fabric, and an integrated NMF-TENG without a spacer was fabricated. As shown in Figure 1 b, the nonwoven PP was randomly distributed. In the case of the nanofibers, the structure of the fabricated mats was more dense and compact than that of the nonwoven PP. Figure 1 c shows the chemical structure both of PP and nylon 6, and it also presents the cross-sectional FE-SEM image of the NMF-TENG specimen. Although it is combined without a spacer between the two triboelectric materials, there is no full contact between the rough fibers owing to the porous structure of the nonwoven fabric. The structural characteristics of the nonwoven PP and nylon 6 nanofiber mat are summarized in Figure 2 and Table 1 . In the case of the nonwoven PP, its average diameter was in the range of 2.4–2.7 μm as the basis weight of the nonwoven PP increased. The thickness of the nonwoven PP increased up to approximately two times with the increasing weight. The air permeability decreased with increasing weight, and it is assumed that the air pathway increases in the thickness direction. In Figure 2 a, the pore size distribution of the PP membrane with the various basis weights was exhibited. The pore size showed a relatively broad range, which was not affected by the increase in basis weight of the nonwoven PP. The average and maximum pore sizes were 16 and 33 μm. The porosity of the nonwoven PP was slightly increased, up to 80%, as the basis weight of the nonwoven PP increased. It was confirmed that the air permeability and porosity of the nonwoven PP is higher than that of other film-based materials commonly used in TENGs. In addition, the pore area of the PP50 basis weight related to the air volume was significantly increased, and the area of fiber contact between the materials could also be increased with an applied force. For the nylon 6 nanofiber mat, the fiber diameter was 283 nm, which was ten times smaller than that of the nonwoven PP. The pore size was sharply distributed around 320 nm, and the maximum pore size was 800 nm. From the previous FE-SEM image, the structure of the nylon 6 nanofiber mat was relatively denser than that of the nonwoven PP. The porosity of the nanofiber mat was approximately 87%, a high pore volume. From those results, we confirmed that both the nonwoven PP and nylon 6 nanofiber mats have a highly porous and open pore structure through the thickness direction. A schematic of the NMF-TENG under the vertical contact mode is shown in Figure 3 a. The nylon 6 nanofiber mat and nonwoven PP are chosen as the negative and positive triboelectric layers, respectively. The nylon 6 nanofiber mat is successfully deposited by electrospinning onto the nonwoven PP surface, and the Ni fabric electrodes are attached to the top and bottom of the triboelectric layers. Figure 3 b shows a schematic of the NMF-TENG to which an external force is applied and released. The triboelectric layers of the NMF-TENG contain numerous pores with a volume of air and critical contact points between the fibers. As the external force with vertical components is applied, the triboelectric layers are deformed, which results in a decrease in the air volume and an increase in critical contact points between the frictional materials. Therefore, by applying a mechanical force such as pressing and releasing, surface triboelectric charges are generated due to changes in the contact area between frictional materials having different polarities. Figure 3 c shows a schematic of the electricity generation mechanism in the vertical contact mode. In the original state, there is no generation of frictional charges or potential differences between the two electrodes [ 52 ]. When an external force is applied to the top surface of the NMF-TENG, the triboelectric layer is deformed as described above, and then triboelectric charges are generated by the change in contact area between the nylon 6 nanofibers and PP microfibers. When the external force is released, the opposite charges are separated, and electrons flow through the external circuit, inducing the potential difference. As a result of this sequence, the open-circuit voltage and short-circuit current shown in Figure 3 d are generated during one cycle by the external force applied to the NMF-TENG surface. The effect of the PP basis weight on the electrical performance of the NMF-TENGs was investigated by applying an external force to NMF-TENGs with different PP basis weights. The fabricated NMF-TENGs were periodically pressured and released under the constant force of 5 N at a frequency of 8 Hz. The NMF-TENGs generated excellent output voltage and current, as shown in Figure 4 , despite the absence of a spacer in the structure. The electrical output performance of PP15 shows V oc of 1.55 ± 0.12 V and I sc of 161.01 ± 10.12 nA. The electrical performance of PP30 shows V oc of 2.79 ± 0.13 V and I sc of 243.88 ± 12.42 nA, which is an improved output performance over that of PP15. Furthermore, when the PP basis weight is increased to 50 gsm, V oc and I sc are further improved to 3.54 ± 0.13 V and 374.39 ± 22.28 nA, respectively. Based on these results, the electrical output performance of the NMF-TENG was improved with an increase in the PP basis weight from 15 to 50 gsm. These results arise because the high basis weights of nonwoven PP contain a high porosity and large volume of air with a high dielectric constant, thereby enhancing the triboelectric effect. In this study, the NMF-TENG fabricated by the use of nonwoven PP50 and a nylon 6 nanofiber mat is selected to examine the triboelectric properties. The electrical output performances of the fabricated NMF-TENG were investigated by applying various external forces. The NMF-TENG was pressured and released for three cycles with increasing force to 0.5, 2, and 5 N in a mechanical pressure machine (SnM Tech, Korea), as shown in Figure 5 a. Figure 5 b illustrates the electrical response of the NMF-TENG in terms of voltage as the force increases. As shown in the results, it can be observed that the output voltage of the NMF-TENG increases to 1.26 ± 0.01, 2.38 ± 0.24, and 3.47 ± 0.06 V as the pressure varies between 0.5, 2, and 5 N, respectively. Figure 5 c shows the electrical response of the NMF-TENG in terms of current. The results indicate that the output current increases to 107.76 ± 7.07, 150.23 ± 9.71, and 247.58 ± 10.71 nA as the pressure gradually increases. As mentioned above, the increase in electrical performance of the NMF-TENG in response to an increasing pressure is related to a change in the interfacial contact area between the friction materials. Moreover, the electrical charge potential of the NMF-TENG was realized by finger ( Figure 5 d), blade ( Figure 5 e), and palm tapping ( Figure 5 f) with the human’s hand, and the generated output voltages were 12.52 ± 0.87, 23.10 ± 0.60, and 33.93 ± 2.09 V, respectively, despite the application of a similar pressure, from approximately 14 to 16 N. The results indicate that the area of pressure applied to the surface of the NMF-TENG is related to the output performance. It means that the charge potential of the inner friction area is increased through the increase in pressure area. To demonstrate the capability of the NMF-TENG for energy harvesting, it was used as a power source to charge capacitors through a bridge rectifier under a periodic external force of 5 N at 8 Hz, as shown in Figure 5 g. As plotted in Figure 5 h, the capacitor with 0.1 μF is rapidly and steeply charged by the NMF-TENG. However, as the capacitance becomes larger, a characteristic linear charging behavior occurred. Furthermore, we replaced the capacitor with LEDs in the bridge rectifier, and only the power source generated by the NMF-TENG was used to turn on the LEDs, as shown in Figure 5 i. The full fabric-type NMF-TENG can be considered as a power source for wearable devices."
} | 3,988 |
36684523 | PMC9843297 | pmc | 2,350 | {
"abstract": "With over a decade of tremendous effort and exciting developments, we have yet to see the successful commercialization of wearable soft electronics due to the lack of reliable energy solutions. This perspective summarizes the theoretical limits and the practical limitations for wearable energy devices and proposes strategies to address such limitations."
} | 88 |
39762374 | PMC11704319 | pmc | 2,352 | {
"abstract": "Over the past decades, human impacts have changed the structure of tropical benthic reef communities towards coral depletion and macroalgal proliferation. However, how these changes have modified chemical and microbial waterscapes is poorly known. Here, we assessed how the experimental removal of macroalgal assemblages influences the chemical and microbial composition of two reef boundary layers, the benthic and the momentum. Chemical and microbial waterscapes were spatially structured, both horizontally and vertically, according to macroalgal dominance and boundary layers. Microbes associated with reef degradation were enriched in the boundary layers surrounding macroalgal-dominated substrata. Dominant macroalgae were surrounded by a distinct chemical pool of diverse lipid classes (e.g., diterpenoids and glycerolipids) and labile organic matter (e.g., organooxygen compounds), which diffused from algal tissues to boundary layers according to their polarity. Finally, our results highlighted strong co-variations between specific algal-derived metabolites and planktonic microbes, giving insight into their roles in coral reef functioning and resilience.",
"introduction": "Introduction Coral reefs are among the most productive and diverse ecosystems on the planet. Their high productivity in oligotrophic tropical seas depends on a tight benthic-pelagic coupling and key microbial processes 1 – 3 . Benthic organisms, like stony corals, macroalgae, or turf, release an extensive amount of dissolved and particulate organic matter underpinning reef community metabolism. For example, photosynthates and coral mucus constitute labile organic matter consumed by microbial assemblages tunneling essential nutrients through the coral reef food web 1 , 4 , 5 . Part of this chemical diversity can act as powerful cues involved in communication or defense structuring reef communities 6 , 7 . As such, the chemical pool in which sessile organisms bathe mediates both positive (e.g., settlement cues, metabolites exchange) and negative (e.g., allelopathy, competition) biotic interactions between reef members 6 , 8 . The combination of molecular exchanges and microbial processes within reef waters results in complex chemical and microbial waterscapes which are just starting to be deciphered 9 – 11 . Water masses of variable thickness emerge from drag forces as water flows over the reef geomorphology, resulting in a physical stratification of reef waterscapes 12 . Three boundary layers have been described: the benthic boundary layer (BBL—m to cm scale) influenced by the overall shape of the reef and main currents; the momentum boundary layer (MBL—cm to mm scale) receiving organic matter from the benthos through advection; and the diffusive boundary layer (DBL—mm to μm scale) essentially formed by the diffusion and accumulation of benthic products 12 . These water masses are dynamic, influenced by flow velocity and benthos complexity, causing variation in the transfer of organic matter within and between boundary layers 13 . Therefore, each boundary layer may have distinct biological and chemical characteristics, reflecting both benthic member identities and physical processes. Accumulating anthropogenic pressures, such as ocean warming, pollution and overuse, have drastically altered the sessile community structure of the reef benthos, inducing phase shifts from coral to macroalgal dominance with cascading effects down to microbial scale 14 , 15 . Large-scale studies, across reefs and ocean basins, have demonstrated that the pelagic microbiome reflects the underlying benthos of shifted reefs. Specifically, reefs that have transitioned towards macroalgal dominance tend to harbor higher microbial density and abundance of copiotrophic, potentially pathogenic, microbial taxa than coral-dominated reefs 14 , 16 , 17 . Reef taxa actually exert a strong organismal influence on microbial assemblages, although limited to their immediate vicinity. Benthic primary producers, such as corals, macroalgae or turf algae, exhibit highly distinct microbial assemblages in their MBLs, which also differ from those of the upper water layer 10 , 18 . These microbial shifts are likely driven by concurrent changes in benthic-derived organic matter, yet its composition and small-scale spatial variation across boundary layers and benthic organisms remain inadequately understood. In the last two decades, several studies have provided essential groundwork for unveiling the differential influence of benthic communities on chemical diversity and the link with microbial processes 4 , 19 – 22 . Coral and macroalgal exudates have specific chemical signatures and select for taxonomically and functionally distinct microbial communities. Specifically, carbon-rich algal exudates select for copiotrophic bacteria with more potential virulence factors compared to coral exudates 4 . Yet, it is only recently, with the advancement in marine untargeted metabolomics and chemoinformatics 23 – 26 , that these metabolite pools have been shown to differ in their elemental stoichiometry and macronutrient content 21 . However, in-situ investigations of the chemical composition of reef waters, as well as their sources, remain scarce 9 , 11 , 27 . For example, molecular gradients from coral surface to overlying boundary layers have been described, comprising infochemicals, such as quorum sensing and antibacterial compounds, that may structure microbial communities surrounding the coral holobiont 27 . Conversely, in two recent studies, untargeted chemical composition of reef exometabolomes did not vary across depths 9 and did not reflect benthic composition across habitats 11 . Linking changes in metabolite pools and microbiome structure induced by macroalgal dominance is paramount to understand the ecology of transitioning reefs. Dominance of macroalgae may alter reef biogeochemical cycles through changes in microbe-metabolite interactions 1 , 14 . Additionally, water-mediated effects involving waterborne allelochemicals and microbially-mediated processes are thought to be involved in coral-algal competition 6 , 28 , 29 . If algal-associated microbial and chemical waterscapes drive the demise of coral reefs, they could create a feedback loop that reinforces macroalgal dominance 28 . In the light of accelerating coral reef degradation, a better comprehension of the effects of macroalgal assemblages on chemical and microbial signatures across reef boundary layers, as well as the origin and diffusion of algal-derived metabolites, is needed. Here we describe an in-situ manipulative experiment in the lagoon of Mo’orea, French Polynesia, designed to assess how macroalgal assemblages influence the chemical and microbial composition of two reef boundary layers: the MBL and BBL. We sampled reef waters surrounding algal-dominated and algal-removed coral bommies and investigated algal-derived metabolites and microbes using untargeted metabolomics tandem mass spectrometry (LC-MS/MS) and 16S rDNA metabarcoding. Our integrative analysis reveals a spatial structuring of the chemical and microbial waterscapes according to macroalgal dominance and boundary layers, with an enrichment of opportunist copiotrophic bacteria in algal-associated waters. By investigating compound class identities in two invasive macroalgae and their associated boundary layers, we also show diffusion gradients of distinct compound classes and highlight co-variations between algal-derived metabolites and planktonic microbes.",
"discussion": "Discussion On coral reefs, the tight coupling between microorganisms and metabolites drives reef biogeochemical processes and mediates ecological interactions 1 , 3 . In the wake of the widespread shift from coral to macroalgal dominance, deciphering the unseen diversity of microbes and chemicals and their spatial distribution is paramount to understand the consequences of benthic community changes on coral reef ecosystem function and resilience. Here, we demonstrate that macroalgal assemblages modify chemical and microbial waterscapes surrounding coral bommies. Our analyses identified specific classes of algal-derived metabolites, including diverse lipids classes (e.g., prenol lipids and glycerolipids) and labile organic matter (e.g., organooxygen compounds and carboxylic acids). Several of these classes featured a diffusion profile with decreasing relative abundance from the whole tissue to the BBL, which is consistent with the hypothesis that hydrophobic compounds remain close to emitters, while polar compounds move across boundary layers. In addition, microbiome-metabolome data integration highlighted strong co-variations between algal-associated metabolites and microbes, overall suggesting that some of these metabolites could structure microbial communities in coral reef waters. Our study shows a remarkable spatial heterogeneity in microbial and chemical waterscapes according to dominant benthic taxa and boundary layers under real flow condition. Horizontally, the description of distinct pools of bacteria and metabolites in each boundary layer between algal-dominated and algal-removed bommies is congruent with previous studies at the organismal 10 , 18 and reef 14 , 16 , 17 scales. Consistent with earlier research on microbial communities 10 , 18 , we also observed a vertical separation between the MBL and BBL surrounding both algal-dominated and algal-removed bommies. This spatial structuring is likely driven by interacting biological and physical processes. The hydrodynamic conditions, such as flow velocity and turbulences, within each boundary layer regulate the advection of benthic organic matter and microorganisms 12 , 13 . In addition, the rates at which metabolites and microbes are released into the water column, along with their interactions, further determine the occurrence of chemical and microbial concentration gradients 27 , 36 . However, some studies did not detect differences between surface and benthic waters in microbial assemblages 37 and chemistry 9 . Interestingly, the multi-omic signatures of BBLs surrounding algal-dominated and algal-removed bommies were clearly distinct. Although reef currents are expected to homogenize upper water layers, the presence of canopy-forming macroalgae, such as Turbinaria and Sargassum , likely reduces flow velocity and increases turbulence, which, in turn, may increase the retention of molecules and microbes within the BBL 38 . By exploring metabolite pools from the algal tissue to the algal-dominated BBL, this work provides a fine-scale spatial characterization of algal-derived metabolites within the reef ecosystem. For both algal species, the surface metabolome lied between the endo-metabolome and the two boundary layers (i.e., the MBL and BBL), highlighting holobiont surfaces as key interaction zones between the host, its epibiont community and its surrounding environment 7 , 27 . In addition, a substantial proportion of surface- and endo-metabolites were detected across the boundary layers, supporting the existence of chemical gradients in the water column 27 . By relating the diversity and relative abundance of metabolites to their chromatographic retention times from the algal tissue to the algal-dominated BBL (Fig. 5c ), our results suggest that algal-derived metabolites diffuse into the water column according to their polarity. This finding supports that hydrophobic compounds will act on contact or diffuse at a very short distance, while medium polar or hydrophilic compounds will have a wider range of detection and potential activities. Using spectral library matches and in silico annotations, our study demonstrates that the diffusivity of metabolites across boundary layers differs according to their molecular classes. In a recent study, metabolites from diverse lipid classes, including prenol lipids, fatty acyls, glycerolipids, and glycerophospholipids, were abundantly recovered from tropical green and red macroalgae which is congruent with our results 22 . We found that metabolites associated to sesquiterpenoids, carotenoids, and sulfolipids were abundant in the algal tissue and surface but scarce in algal waters. This limited diffusion is consistent with their surface-bound anti-fouling and cytotoxic properties 39 – 41 . In contrast, diterpenoids, conspicuous in Dictyota tissue, were also abundant in the Dictyota MBL and the algal-dominated BBL (Fig. 6c ) . While diterpenoids display a range of contact-dependent bioactivities, ranging from herbivore deterrence to allelopathy 42 – 44 , this finding indicates a water-borne action and/or a high release rate by macroalgae. In contrast, benzenoids, organooxygen compounds, carboxylic acids, and lysolipids were abundantly detected in the algal MBLs and algal-dominated BBL (Fig. 6c ; Supplementary Fig. 9 ). Their high abundance in the boundary layers suggests a non-specific origin of these compounds. For instance, benzenoids are released by diverse organisms, such as phytoplankton 45 , turf, and corals 21 . Organooxygen compounds also constitute important growth and energy substrates for microbial life 4 , 46 . While variations in physiochemical properties (e.g., size, polarity 36 ) might account for the different abundance profiles observed among the sub-categories of organooxygen compounds (Supplementary Figs. 9 and 10 ), these differences might also imply distinct lability and, therefore, biotransformation rate by microbes. Finally, the structural property of lysolipids, consisting of a single fatty acyl chain and a polar head, might make them more hydrophilic than other lipids 47 . While we highlight polarity as a potential factor influencing metabolite diffusion, future investigations into the other aforementioned factors will be paramount to understand the determinants and extent of these chemical gradients in reef waters. Our integrative approach reveals a strong co-variation between microbiome and metabolome datasets, supporting a tight microbe-metabolite coupling in the coral reef waterscape. The observed enrichment of Gammaproteobacteria and Bacteroidia in algal waters echoes numerous studies demonstrating the selective influence of energy- and carbon-rich algal exudates on these microbial classes 1 , 4 , 20 . The algal MBLs and algal-dominated BBL harbored several genera from these classes known for their copiotrophic lifestyle and capacity to degrade complex OM, including Alcanivorax , Eilatimonas, Idiomarina , Leeuwenhoekiella and Oleibacter 46 , 48 – 50 . In addition, the Alphaproteobacteria Croceicoccus and Hoeflea , abundant in the algal waters, can metabolize mannitol as their carbon source 51 , 52 . In fact, organooxygen compounds constituted a significant proportion of the algal MBLs, particularly in negative ionization mode (Fig. 6c ; Supplementary Figs. 9 and 10 ). Besides carbohydrates, carboxylic acids can also constitute strong chemo-attractants and foraging cues for coral reef bacteria 53 , 54 , including from the Rhodobacteraceae, which is consistent with the observed positive correlations in microbe-metabolite network (Fig. 7 ; Supplementary Fig. 12 ). Although the analysis of data in positive ionization mode prevails in coral and reef metabolomics and yields a greater number of putative annotations, our results encourage the exploration of negative ionization metabolomics to gain a better understanding of microbe-metabolite interactions. The enrichment of copiotrophic lineages has often been described as indicative of high macroalgal cover and signs of reef microbialization 1 , 14 , 17 . In addition, these copiotrophs can constitute virulent microbial populations capable of shifting towards a pathogenetic lifestyle when labile carbon is abundant 4 , 14 , 46 . For example, Alcanivorax , enriched in the algal MBLs (Fig. 4 ), has previously shown such metabolic versatility 46 . Copiotrophic microbes can invade stressed and diseased corals 55 , 56 , and several taxa characteristic of the algal MBLs are suspected to be coral 57 , 58 and algal 59 , 60 pathogens, among them Halomonas, Maricaulis and Nautella (Fig. 4 ). However, some taxa (e.g., Halomonas ) can also be beneficial bacteria 61 , calling for cautious interpretation of their putative roles beyond their taxonomic affiliations. Comparatively, Roseitalea and Mameliella , two probiotic bacteria of corals and Symbiodiniaceae 62 , 63 , were depleted in algal waters. Allelopathy could also represent a threat to coral holobionts. In particular, several studies on hard corals and sponges demonstrated the cytotoxicity of diterpenoids and SQDGs 41 , 42 . In this context, we propose a conceptual model for the spatial structuring of chemicals and microbes around algal-dominated bommies (Fig. 8 ). In this model, algal-derived metabolites act as structuring elements of planktonic microbial communities. We posit that boundary layers surrounding algal-dominated bommies harbor metabolites that could foster reef degradation through allelopathy and the virulence of copiotrophs. However, the extent to which these water-borne elements are effectively detrimental to underlying benthic corals, as well as their pelagic stages, urgently requires further investigation. This model is consistent with the DDAM (Disease, Dissolved organic carbon, Algae, and Microbe) and reef microbialization hypothesis, which supports a link between elevated DOC concentrations and the emergence of potential coral pathogens 14 , 28 , 46 . Fig. 8 Conceptual diagram of chemical and microbial waterscapes above the algal-dominated bommies. Algal-derived metabolites vary in their diffusivities across boundary layers (i.e., BBL and MBL), with some metabolites displaying diffusion-like profile (e.g., diterpenoids, glycerolipids). The emergence of chemical gradients structure microbial assemblages according to the molecular classes and the ecological roles of these metabolites. Metabolites can constitute trophic resources (e.g., organooxygen compounds) or allelopathic agents (e.g., diterpenoids, sulfolipids). Within the algal waterscapes, two mechanisms could contribute to coral demise and reinforce phase-shifts: water-mediated allelopathy and the virulence of copiotroph microbes fed by algal-derived OM. Allelochemicals (e.g., diterpenoids) and potential virulent bacteria (e.g., Alcanivorax) could negatively impact coral health during the pelagic larval and sedentary benthic phases, which urgently require further experimental investigations. BBL Benthic Boundary Layer. MBL Momentum Boundary Layer. The algal MBLs harbored DMSP-degrading bacteria (e.g., Idiomarina and Marinobacter ) 64 . DMSP is a significant substrate and chemical mediator for bacterioplankton 65 . It is produced by coral- and algal-associated bacteria 64 , as well as free-living taxa, including Pelagibaca and Pseudooceanicolla 66 , two Rhodobacteraceae which were enriched in the algal MBLs. Nitrogen transformation on coral reefs is complex and largely mediated by microbial activity 2 . Some of the detected taxa, such as Acuticoccus and Nitratireductor , may be involved in these processes 67 . Strong microbial interactions with sulfur- and nitrogen-containing compounds were revealed in our analysis of bipartite networks (Fig. 7 ; Supplementary Fig. 12 ), further supporting their structuring role in planktonic assemblages. Conversely, negative microbe-metabolite associations occurred with diterpenoids, SQDGs, phenols, and Lyso-PC (Fig. 7 ; Supplementary Fig. 12 ), which can inhibit bacterial settlement and growth 35 , 68 , 69 . Surprisingly, diterpenoids were also abundant in Turbinaria MBL, likely due to the presence of epiphytic Dictyota on Turbinaria thalli, thereby suggesting a chemically-mediated associational benefit between both algae 70 . In tropical macroalgae, lysophosphatidylcholines may be determinant in the establishment of host-microbe symbiosis, mediated by host immunity pathways 22 . Positive associations between these lipids and microbes from the Saprospiraceae, Flavobacteriaceae, and Rhodobacteraceae families have been reported in algal tissues 22 , while our observations revealed negative associations with two genera from the Rikenellaceae family in algal waters. Additionally, metabolites may be of bacterial origin, such as the cyclopeptide (Fig. 7a ; Supplementary Table 5a ) related to halolitoralin, an antifungal cyclopeptide produced by marine bacteria 71 . Although our study focused on algal waters, it contributes to deciphering the microbe-metabolite pairings associated with macroalgal holobionts. Promising future research should investigate the consistency of these patterns across algal tissues, surfaces, and the surrounding environment to further our understanding of ecological interactions within coral reef systems. Despite a broad classification system, our work demonstrates that coral reef macroalgae release a variety of compounds potentially involved in defensive and competitive interactions, as well as in microbial energetics 21 , 35 , 42 . However, minor modifications in the structure of compounds can change their bioactivity, making the prediction of their effects solely based on their class identities ambitious 72 . We also cannot rule out that the mentioned microbe-metabolite co-variations are not causal. Moreover, it is important to note that the relative abundance of metabolites does not necessarily reflect their absolute amounts. Quantification using appropriate standards is essential to confirm the actual concentrations of these metabolites in the samples. Without an accurate quantification of metabolite concentrations, a better annotation of unknown metabolites, and improved knowledge of the ecology of reef microbes, determining the nature of microbe-metabolite interactions will remain highly challenging. Future research should isolate specific metabolites and conduct bioassays to elucidate their role in microbial processes and interspecific interactions. Importantly, only a fraction of the chemical pool can be successfully retrieved as the successive steps of extraction, ionization, fragmentation, and annotation narrow down the pool of studied compounds, leaving a vast unknown “dark” fraction 26 . While the recent advent of in silico tools have revolutionized compound dereplication, achieving confident annotations, even at a broad level, remains highly challenging, especially for the marine environment with limited spectral libraries 26 . Overcoming these limitations will be crucial for advancing marine metabolomics and multi-omics. This study provides key insights into the influence of macroalgal assemblages on chemical and microbial waterscapes. The taxonomic composition of the microbial communities revealed an enrichment of copiotrophic bacteria, characteristic of altered reef states and compromised coral holobionts, in two boundary layers overlying macroalgal-dominated bommies. By characterizing the broad molecular classification of algal-derived metabolites, this work participates in the description of the chemodiversity on coral reefs and improves our understanding of water-mediated transport of chemical compounds and their roles as structuring and functioning elements. The data presented herein contribute to further unveil the identity, distribution and co-variations of metabolites and microbes within reef waterscapes and constitute a starting point to further investigate their complex roles in coral reef functioning and resilience. It also provides leads for more targeted research to explore the water-mediated mechanisms by which coral reef macroalgae reinforce the persistence of coral-algal phase-shifts."
} | 5,936 |
35394830 | PMC8993117 | pmc | 2,353 | {
"abstract": "Convolutional neural networks (CNNs) have gained much attention because they can provide superior complex image recognition through convolution operations. Convolution processes require repeated multiplication and accumulation operations, which are difficult tasks for conventional computing systems. Compute-in-memory (CIM) that uses parallel data processing is an ideal device structure for convolution operations. CIM based on two-terminal synaptic devices with a crossbar structure has been developed, but unwanted leakage current paths and the high-power consumption remain as the challenges. Here, we demonstrate integrated ferroelectric thin-film transistor (FeTFT) synaptic arrays that can provide efficient parallel programming and data processing for CNNs by the selective and accurate control of polarization in the ferroelectric layer. In addition, three-terminal FeTFTs can act as both nonvolatile memory and access device, which tackle issues from two-terminal devices. An integrated FeTFT synaptic array with parallel programming capabilities can perform convolution operations to extract image features with a high-recognition accuracy.",
"introduction": "INTRODUCTION Convolutional neural networks (CNNs) have been used in various applications, such as speech recognition and image classification ( 1 , 2 ). CNNs can achieve superior performance in complex image recognition processes through convolution operations that use kernels as filters to extract features from images. However, convolution operations are time-consuming because they involve repeated multiplication and accumulation operations between the input data and the kernel weight. When CNNs are implemented in hardware based on the von Neumann architecture, drawbacks such as increased power consumption and limited data processing speeds arise because large amounts of data need to be transferred between the processing and memory units ( 3 , 4 ). To overcome these limitations, compute-in-memory (CIM) has been suggested as alternative hardware for CNNs because it enables parallel data processing ( 5 – 9 ). In CIM, parallel data processing can be performed in a crossbar array structure using vector-matrix multiplication (VMM) based on Ohm’s law (for multiplication) and Kirchhoff’s law (for accumulation) ( 10 ). For accurate VMM operation, synaptic devices are required, which can precisely adjust the conductance states via analog conductance modulation ( 5 , 11 , 12 ). In previous studies, several CIM devices with a crossbar structure have been demonstrated using two-terminal devices, such as phase-change and resistive-switching memories, as synaptic devices ( 13 – 18 ). However, when CIM is implemented using crossbar arrays based on two-terminal devices, there are issues such as cross-talk, sneak path current, and nonlinear current-voltage characteristics ( 19 – 21 ). The analog conductance modulation characteristics of two-terminal devices are usually achieved by controlling the current of devices during the programming process ( 22 – 24 ). However, in arrays consisting of two-terminal devices, it is difficult to control the current of selected devices during the programming process precisely because of current flowing through the unselected devices ( 25 , 26 ). In addition, the current flowing through the unselected devices can affect the accuracy of read operation. Therefore, this issue can cause inaccurate conductance modulation in crossbar arrays based on two-terminal devices ( 26 ). Three-terminal devices have the potential to perform accurate conductance modulation in array structure because terminals for program and read operations are separated ( 19 , 27 ). In an array composed of three-terminal devices, the program and read operations are not affected by the states of the unselected devices. Therefore, three-terminal devices, such as electrochemical transistors and charge-trapping transistors, have been investigated as synaptic devices ( 10 , 17 , 19 , 20 , 28 – 31 ). Electrochemical transistors exhibit linear weight updates and low-voltage operation ( 19 , 31 ). However, the use of active ions (e.g., Li + ) may be incompatible with complementary metal-oxide semiconductor (CMOS) device fabrication and integration ( 17 , 32 ). Although charge-trapping transistors are based on mature technology for array integration, they require a high operation voltage and exhibit nonlinear weight update characteristics ( 16 , 20 ). Among the available three-terminal synaptic devices, ferroelectric transistors based on zirconium-doped hafnium oxide (HfZrO x ) are advantageous because they have CMOS compatibility, fast operation speed, low operation voltages, and high scalability ( 33 – 36 ). In particular, the conductance of ferroelectric transistors can be precisely modulated by controlling the states of polarization in the ferroelectric layer ( 33 , 37 , 38 ). These properties can be used to implement kernel weights for convolution operations in ferroelectric transistor arrays. Thus, ferroelectric transistor arrays could be used as synaptic arrays for CIM. In this study, we demonstrate integrated synaptic transistor arrays based on ferroelectric thin-film transistors (FeTFTs) composed of indium zinc oxide (IZO) and HfZrO x . In these arrays, FeTFTs are used as synaptic devices. The conductance of the FeTFTs, which represents kernel weight, is linearly controlled by adjusting the states of polarization in the ferroelectric layer. In addition, column- and row-wise parallel programming operations are demonstrated in ferroelectric synaptic transistor arrays by the selective control of polarization switching with program-inhibit operations. Using these characteristics, ferroelectric synaptic transistor arrays are trained to implement kernel weights and perform convolution operations, which can be used to extract the features of input images. The four different kernels, which are realized in ferroelectric synaptic transistor arrays, can extract the features of an image with 64 × 64 pixels. In addition, neural network simulations based on the weight update characteristics of the ferroelectric synaptic transistor array exhibit an image recognition accuracy of 90.3%. The results of this study provide important information that can contribute to the development of CIM based on ferroelectric synaptic transistor arrays.",
"discussion": "DISCUSSION In summary, we integrated FeTFTs based on IZO oxide semiconductors and ferroelectric HfZrO x to investigate the potential of the ferroelectric synaptic transistor array for use in CIM applications. To implement the weight update characteristics, we controlled the conductance of the FeTFTs by adjusting the polarization of the ferroelectric layer. In ferroelectric synaptic transistor arrays, column- and row-wise parallel programming methods were experimentally demonstrated by the selective control of the polarization switching using program-inhibit operations. Kernel weights for convolution operations were realized in the ferroelectric synaptic transistor arrays using parallel weight update processes. The FeTFTs exhibited linear current-voltage behavior and weight update characteristics, which were essential for synaptic devices used for convolution operations. Accurate convolution operations were demonstrated in the ferroelectric synaptic transistor arrays using the kernel weights in the array. These convolution operations based on ferroelectric synaptic transistor arrays enabled the extraction of the features in an input image with 64 × 64 pixels. To further investigate the potential of the ferroelectric synaptic transistor arrays as CNNs for CIM, we simulated the CNNs using the characteristics of the ferroelectric synaptic transistor arrays. In simulations based on the weight update characteristics of the ferroelectric synaptic transistor arrays, the CNNs achieved an image recognition accuracy of 90.3% for a CIFAR-10 dataset. These results suggest that ferroelectric synaptic transistor arrays based on IZO oxide semiconductors and ferroelectric HfZrO x have the potential to be used as neuromorphic hardware for CNNs."
} | 2,031 |
28346361 | PMC5412291 | pmc | 2,354 | {
"abstract": "Most species in the Leguminosae (legume family) can fix atmospheric nitrogen (N 2 ) via symbiotic bacteria (rhizobia) in root nodules. Here, the literature on legume-rhizobia symbioses in field soils was reviewed and genotypically characterised rhizobia related to the taxonomy of the legumes from which they were isolated. The Leguminosae was divided into three sub-families, the Caesalpinioideae, Mimosoideae and Papilionoideae. Bradyrhizobium spp. were the exclusive rhizobial symbionts of species in the Caesalpinioideae, but data are limited. Generally, a range of rhizobia genera nodulated legume species across the two Mimosoideae tribes Ingeae and Mimoseae, but Mimosa spp. show specificity towards Burkholderia in central and southern Brazil, Rhizobium / Ensifer in central Mexico and Cupriavidus in southern Uruguay. These specific symbioses are likely to be at least in part related to the relative occurrence of the potential symbionts in soils of the different regions. Generally, Papilionoideae species were promiscuous in relation to rhizobial symbionts, but specificity for rhizobial genus appears to hold at the tribe level for the Fabeae ( Rhizobium ), the genus level for Cytisus ( Bradyrhizobium ), Lupinus ( Bradyrhizobium ) and the New Zealand native Sophora spp. ( Mesorhizobium ) and species level for Cicer arietinum ( Mesorhizobium ), Listia bainesii ( Methylobacterium ) and Listia angolensis ( Microvirga ). Specificity for rhizobial species/symbiovar appears to hold for Galega officinalis ( Neorhizobium galegeae sv. officinalis ) , Galega orientalis ( Neorhizobium galegeae sv. orientalis ), Hedysarum coronarium ( Rhizobium sullae ), Medicago laciniata ( Ensifer meliloti sv. medicaginis ), Medicago rigiduloides ( Ensifer meliloti sv. rigiduloides ) and Trifolium ambiguum ( Rhizobium leguminosarum sv. trifolii ). Lateral gene transfer of specific symbiosis genes within rhizobial genera is an important mechanism allowing legumes to form symbioses with rhizobia adapted to particular soils. Strain-specific legume rhizobia symbioses can develop in particular habitats.",
"conclusion": "7. Conclusions Overall, the data indicate that lateral gene transfer of specific symbiosis genes within rhizobial genera is an important mechanism allowing legumes to form symbioses with rhizobia adapted to particular soils. It also maintains specificity between legume species and rhizobia species with specific symbiosis genes. Strain-specific legume rhizobia symbioses can develop in particular habitats.",
"introduction": "1. Introduction The Leguminosae (Fabaceae, the legume family) is comprised of ca. 19,300 species within 750 genera that occur as herbs, shrubs, vines or trees in mainly terrestrial habitats and are components of most of the world’s vegetation types [ 1 , 2 , 3 ]. Currently, the legume family is divided into three sub-families, the Caesalpinioideae, Mimosoideae and Papilionoideae [ 3 , 4 ]. Members of the Caesalpinioideae are grouped into four tribes, the Caesalpinieae, Cassieae, Cercideae and Detarieae comprising ca. 170 genera and 2250 species. The Mimosoideae are grouped into two tribes, the Ingeae and Mimoseae with ca. 80 genera and 3270 species, while the Papilionoideae consists of 28 tribes with ca. 480 genera and 13,800 species. However, a new classification of the legumes has been proposed with six sub-families based on the plastid matK gene sequences from ca. 20% of all legume species across ca. 90% of all currently recognized genera [ 5 ]. The six sub-families proposed are a re-circumscribed Caesalpinioideae, Cercidoideae, Detarioideae, Dialioideae, Duparquetioideae and Papilionoideae. In this system, the currently recognized Mimosoideae is a distinct clade nested within the re-circumscribed Caesalpinioideae. Species within the Cercidoideae, Detarioideae, Dialioideae and Duparquetioideae do not nodulate [ 5 , 6 ]. Most legume species can fix atmospheric nitrogen (N 2 ) via symbiotic bacteria (general term “rhizobia”) in root nodules, and this can give them an advantage under low soil nitrogen (N) conditions if other factors are favourable for growth [ 7 , 8 ]. Furthermore, N 2 fixation by legumes can be a major input of N into natural and agricultural ecosystems [ 9 , 10 , 11 , 12 ]. Generally, legume nodules can be classified as indeterminate or determinate in growth [ 13 ]. Indeterminate nodules maintain meristematic tissue, while determinate nodules have a transient meristem. Nodule type is dependent on host plant, and legume species that can produce both determinate and indeterminate nodules are rare [ 14 , 15 ]. All genera examined in the Caesalpinioideae and Mimosoideae had indeterminate nodules [ 13 ]. Within the Papilionoideae, most tribes had indeterminate nodules, but the Desmodieae, Phaseoleae, Psoraleae and some members of the Loteae show “desmodioid” determinate nodules and the Dalbergieae “aeschynomenoid” determinate nodules [ 13 ]. Desmodioid nodules have lenticels, and rhizobia “infected” tissue within them also contains uninfected cells. Aeschynomenoid nodules do not have lenticels, have uniform infected tissue and are always associated with lateral or adventitious roots. Where tested, species within the Desmodieae, Phaseoleae and Psoraleae had ureides as the main N-containing compound transported from nodules, but species in the Dalbergieae and Loteae transported amides/amino acids [ 13 ]. Indeterminate nodules have a single or branched apical meristem, and a few genera, such as Lupinus (Genisteae) and Listia (Crotalaria), have “lupinoid” nodules with two or more lateral meristems, which in some cases completely surround the subtending root [ 16 , 17 ]. Generally, indeterminate nodules have a mixture of infected and uninfected cells in the central nodule tissue, but lupinoid nodules, as for aeschynomenoid nodules (Dalbergiae), have uniformly-infected cells. The nodulation process for almost all legumes studied is initiated by the legume production of a mix of compounds, mainly flavonoids, which induce the synthesis of NodD protein in rhizobia [ 18 , 19 ]. Different legumes produce different types/mixes of compounds. The NodD protein activates the transcription of other genes involved in the nodulation process, including those required to produce Nod factors, the signal molecules produced by the rhizobia and detected by the plant, which induce nodule organogenesis [ 20 ]. The nodABC genes encode for the proteins required to make the core Nod factor structure [ 18 ]. Nod factors from different rhizobia have a similar structure of a chitin-like N-acetyl glucosamine oligosaccharide backbone with a fatty acyl chain at the non-reducing end, but differ in their length of N-acetyl glucosamine oligosaccharide backbone and the length and saturation of the fatty acid chain. The Nod-factor core is modified by species-specific proteins, which results in various substitutions, including acetylation, glycosylation, methylation and sulfation. Perception of the Nod-factor signal in legumes is mediated by Nod factor receptors, which are plasma membrane localized serine/threonine receptor kinases in the case of the model legumes Lotus japonicus and Medicago truncatula [ 18 , 19 ]. The available data indicate that rhizobia enter the roots of most legume species via root hair infection [ 13 ]. Here, rhizobia enter root hairs, and host cell wall material grows around the developing “infection”, forming an infection thread, which grows through the cortex of the root, branching repeatedly. Generally, rhizobia are released from the tips of these infection threads into membrane-bound structures within host cells called symbiosomes where they differentiate into their N 2 -fixing form known as bacteroids. However, all species examined in the Caesalpinioideae, except herbaceous Chamaecrista spp. and a few species in the Papilionoideae, retain their rhizobia within infection threads [ 2 , 13 ]. Bacteroids vary greatly in their level of differentiation and viability depending on the legume host [ 13 , 21 ]. The process of root hair infection can lead to the formation of either indeterminate or desmodioid determinate nodules. A second mode of rhizobial infection occurs with species in the Dalbergiae (aeschynemonoid nodules) where rhizobia enter roots at the sites of lateral root emergence (“crack” entry), and infection threads are not involved in the infection process [ 22 , 23 ]. Thirdly, for at least some members of the Genisteae (e.g., Lupinus spp.) and Crotalariae (e.g., Listia spp.), rhizobia enter the roots directly through the root epidermis at the junction between epidermal cells, and again, infection threads are not involved in the infection process [ 17 , 21 , 24 ]. Over the past twenty-five years, DNA-based methods have become increasingly used to characterize rhizobia. In particular, phylogenetic analyses of sequences of the 16S ribosomal RNA (rRNA) gene, a range of “housekeeping” genes and genes involved in the symbiosis have been developed as a “standard approach” [ 15 , 25 , 26 ]. The 16S rRNA gene sequence on its own can delineate rhizobia at the genus level [ 27 ]. The main symbiosis genes studied are the “ nif ” genes, which encode the subunits of nitrogenase, the rhizobial enzyme that fixes N 2 , and the “ nod ” genes, which encode Nod factors that induce various symbiotic responses on legume roots. Specific nod genes have been shown to be major determinants of legume host specificity [ 28 , 29 ]. The nif and nod genes are often carried on plasmids or symbiotic islands, and these genes can be transferred (lateral transfer) between different bacterial species within a genus and more rarely across genera [ 30 , 31 , 32 ]. Almost all rhizobia tested had nod genes. However, a few Bradyrhizobium strains, which do not possess nodABC genes, can form N 2 fixing nodules on particular Aeschynomene spp. [ 33 , 34 ]. Bacterial species from a range of genera in the α-proteobacteria (most commonly Bradyrhizobium , Ensifer ( Sinorhizobium ), Mesorhizobium and Rhizobium ) and two genera in the β-proteobacteria ( Burkholderia ( Paraburkholderia ) and Cupriavidus ) can form functional (N 2 fixing) nodules on specific legumes ( Table 1 , Table 2 , Table 3 and Table 4 ). Reports that Achromobacter and Herbaspirillum (β-proteobacteria) produce N 2 fixing nodules on Prosopis juliflora and Aspalathus linearis , respectively [ 35 , 36 ] and Pseudomonas (Gammaproteobacteria) produces N 2 fixing nodules on Robinia pseudoacacia [ 37 ] and Acacia confusa [ 38 ] have not been confirmed. Furthermore, for Lotus corniculatus , Geobacillus (phylum Firmicutes), Paenibacillus (Firmicutes) and Rhodococcus (Actinobacteria) were for the first time reported as rhizobial symbionts [ 39 ]. These bacterial species had similar nodA gene sequences to Mesorhizobium isolated from the same plants, and it was concluded that the lateral gene transfer of these genes had occurred from the Mesorhizobium . However, lateral gene transfer of symbiosis genes is much less common between than within genera, and this work needs to be independently verified. Legume species differ greatly in their specificity for rhizobial symbionts. Galega officinalis (tribe Galegeae) and Hedysarum coronarium (tribe Hedysareae) have been highlighted as being highly specific with respect to their rhizobial symbionts [ 120 , 145 , 296 , 297 ]. Both of these species are in the inverted repeat lacking clade (IRLC). The IRLC is marked by the loss of one copy of the inverted region of the plastid genome [ 298 , 299 ]. Almost all genera in the IRLC are temperate; all have indeterminate nodules, and where examined, their bacteroids were terminally differentiated and could not return to their bacterial form [ 13 ]. The IRLC contains several important temperate grain (e.g., Pisum sativum and Vicia faba ) and forage (e.g., Trifolium spp. and Medicago spp.) legumes. There is evidence that at least some of these crop legumes have a high degree of rhizobial specificity. For example, an analysis of core and symbiotic genes of rhizobia nodulating Vicia faba and Vicia sativa from different continents showed that they belong to a phylogenetically-compact group indicating that these species are restrictive hosts [ 117 ]. In contrast, Macroptilium purpureum and the grain legumes Phaseolus vulgaris and Vigna unguiculata in the tribe Phaseoleae are nodulated by rhizobia from different genera across the α- and β-proteobacteria [ 264 , 283 , 300 ]. The Phaseoleae are of tropical/subtropical origin, have desmodioid determinate nodules with bacteroids, which are not terminally differentiated [ 2 , 13 ]. Here, the literature on legume-rhizobia symbioses in field soils was reviewed and genotypically characterised rhizobia related to the taxonomy of the legumes from which they were isolated. The objectives of the work were to collate data on legume rhizobia symbioses and then assess to what extent legume specificity for rhizobial symbionts is related to legume taxonomy."
} | 3,280 |
36611085 | PMC9825370 | pmc | 2,355 | {
"abstract": "An increasing number of frequently applied portable electronics has raised the significance of self-powered systems. In this regard, triboelectric nanogenerators (TENGs) have drawn considerable attention due to their diversity of design and high power output. As a widely used material in TENG electrodes, polydimethylsiloxane (PDMS) shows attractive characteristics, such as electron affinity, flexibility, and facile fabrication. To achieve active TENG-based humidity sensing, we proposed a straightforward method to enhance the hydrophilicity of PDMS by two parallel approaches: 1. Porosity induction, 2. Carbon nanotube (CNT) compositing. Both of the mentioned processes have been performed by water addition during the synthesis procedure, which is not only totally safe (in contrast with the similar foaming/compositing routes), but also applicable for a wide range of nanomaterials. Applying the modified electrode as a single-electrode TENG-based humidity sensor, demonstrated an impressive enhancement of sensing response from 56% up to 108%, compared to the bare electrodes. Moreover, the detecting range of ambient humidity was broadened to higher values of 80% in a linear behavior. The fabricated humidity sensor based on a CNT-PDMS foam not only provides superior sensing characteristics but also is satisfactory for portable applications, due to being lightweight and desirably self-powered.",
"conclusion": "Conclusions In summary, our proposed method for facile fabrication of PDMS-based foams was investigated by the addition of CNT. Sensing features of the obtained electrode were totally outstanding, as follows: Response values of 108% (improved by more than 100%); RH range broadens to higher values of 80%; the response/recovery times remained at about 2 s. These mentioned enhancements were achieved by applying a straightforward method of simultaneous addition of porosity and hydrophilic fillers. The proposed self-powered humidity sensor based on PDMS is extremely promising for portable electronics, as well as industrial applications.",
"introduction": "Introduction Providing the required energy for launching a large number of sensors depends on supporting the sensing network via numerous power supplies, resulting in frequent battery charging and displacement. A novel approach considered in the past decade is applying the self-powered sensors 1 – 4 . Since 2012, an increasing number of self-powered systems based on triboelectric nanogenerators (TENG) have been introduced, utilizing mechanical triggers to generate electricity 5 – 14 . In order to launch a TENG-based sensor, two approaches of passive and active TENG circuits are generally adopted 15 . In passive mode, the resistance of the sensing electrode, which is located outside of the TENG structure, varies by detecting the desired species. By integrating the sensing material as one of the TENG electrodes, the change of the surface charge acts as the detecting mechanism, which is called active TENG-based sensing 16 . Among the widespread materials applied as TENG electrodes, PDMS (polydimethylsiloxane) is well known for its high electron affinity, as well as straightforward fabrication 17 . Moreover, good mechanical properties, such as high flexibility and tensile strength, provide PDMS as an appropriate candidate for highly repetitive tapping or rubbing cycles in TENGs 18 – 20 . There is also a wide range of proposed applications for PDMS, due to its inherent biocompatibility 21 and non-toxicity 22 , which is significant for applying body motion to trigger TENGs. Facile surface patterning and functionalization of PDMS provide the opportunity of controlling surface characterization of the TENG electrode 23 – 25 , which is a crucial feature to achieve active TENG-based sensors. Among the environmental properties of the domestic and industrial atmosphere, detecting the amount of ambient humidity is significant 8 . Generally, the sensing material should be inclined to adsorb water molecules, due to its porosity and/or surface adsorbing sites, such as defects and chemical groups 26 . Since conventional polymers applied as TENG electrodes, like PDMS, are hydrophobic, they are not appropriate candidates for humidity sensing. On the other hand, the adsorption of water molecules on a nonporous surface is restricted up to the medium range of relative humidity (RH) 27 . However, these limitations can be modified by surface treatments, such as surface patterning, foaming, and compositing with hydrophilic additives 28 . Producing porosity generally can be performed via using dissolvable or evaporating materials to generate hollow spaces in the bulk polymer 29 – 31 . Among the nanomaterials, carbon nanotube (CNT) has been widely used as an additive, to obtain a nanocomposite foam with desirable sensing behavior and power generation properties 32 – 35 . Here, we proposed a straightforward method to synthesize PDMS foam with water addition to produce macro- and micro-porosity. To reduce the hydrophobicity of PDMS, CNT as an aqueous solution was added to the polymer. During this method, the addition of nanomaterials in aqueous solutions is feasible, leading to a biocompatible surface, due to the fact that nanomaterials will be surrounded by PDMS. Applying the composite of PDMS with CNT as a TENG electrode resulted in enhancing humidity sensing response, as well as widening the sensing range of detecting RH. In addition to the novelty of foaming and compositing PDMS without toxic solvents, an uncomplicated humidity sensor was introduced through an active TENG-based setup, promising for the future of self-powered sensing networks.",
"discussion": "Results and discussion Characterizing the produced electrodes Figure 1 represents the porosity of PDMS and CNT-PDMS foams in various magnifications of the cross-section. The optical images of the polymer shown in Fig. 1 a,e demonstrate the increase in the number of macroscopic pores by the addition of CNT. Since CNT is hydrophilic, it helps more efficient dispersion of the water drops within the polymer, resulting in higher value of porosity after the evaporation of trapped water. Figure 1 b and f show the macro-pore distribution in the foam, where larger pores can be observed near the surface. Meanwhile, the trapped water is moving toward the surface, they are inclined to join together, due to the hydrophobicity of PDMS. In the presence of CNT, it is expected that the hydrophilicity of the matrix improves, and consequently, a smaller pore size is achieved. Figure 1 c,d,g, and h demonstrate micro-pores within the macro-size pores, especially in CNT-PDMS nanocomposite. Figure 1 Optical ( a , e ) and SEM ( b – d , f – h ) images of the cross-section of pure porous PDMS (top row) and CNT-PDMS (bottom row) electrodes. Figure 2 demonstrates the mechanism of water entrapment inside the polymer. By stirring the polymer precursors and water, the colonies of water produced inside the polymeric network (Fig. 2 a). Then, heating up the whole system resulted in both polymer curing and water evaporation, leading to obtaining a porous structure. By addition of CNTs (Fig. 2 b), the functional groups on the CNT (confirmed by FTIR analysis (Figure S1 )) form several hydrogen bonding with water molecules, facilitating water entrapment inside the polymer network during stirring. Therefore, smaller amounts of water would be trapped inside the polymeric network in CNT-PDMS (Fig. 2 b) compared to PDMS (Fig. 2 a), resulting in finer pores observed in the final sample after the evaporation of water. Figure 2 The formation of pores containing water molecules inside the PDMS network ( a ) and the CNT-PDMS network ( b ), where the presence of carbon nanotubes as a separating agent of water molecules causes the creation of smaller pores. The presence of CNT leads to an obviously darker color of the obtained polymer (Figure S2 ). Since the fabricated foams are opaque (in contrast to dense PDMS), their reflection behavior was investigated via UV–Vis spectroscopy (Fig. 3 a). According to the absence of transmittance, the lower reflection of CNT-PDMS confirms higher absorption by CNTs. DRS spectroscopy was also performed, due to the highly porous surface of the electrodes, which reveals analogous result for CNT-PDMS compared to pure PDMS (Fig. 3 b). On the other hand, the amount of carbon atoms is higher in specific points of the elemental analysis (Figure S3 ), confirming the presence of CNTs. EDX map also shows the distribution of pores within the whole thickness of the foam (Figure S4 ). The reduction of hydrophobicity of the PDMS after the addition of CNT was investigated via contact angle test, as shown in Fig. 3 c. Accordingly, more inclination to water molecules is expected, resulting in finer porosity as mentioned above. Figure 3 Characterization of fabricated foams. UV–Vis ( a ), DRS ( b ), and contact angle ( c ) tests. TENG studies To investigate the TENG performance of the fabricated electrodes, the output voltage and current of three single-electrode TENGs were studied. According to the TENG structure shown in Fig. 4 , the mechanism of power generation via finger tapping can be observed. When the finger is in contact with the PDMS electrode, negative triboelectric charges are produced on the surface of the polymer (Fig. 4 a). During the releasing step, negative charges move from aluminum toward the ground, in order to satisfy the equilibrium condition (Fig. 4 b). In the released step, the whole system is stable, resulting in no charge transfer (Fig. 4 c). In the pressing step, the electrons move in the reverse direction (Fig. 4 d), which leads to an inverted signal peak. Figure 4 TENG mechanism. The generated charge in contact ( a ), releasing ( b ), released ( c ), and pressing ( d ) steps. Figure 5 demonstrates the open-circuit voltage and short-circuit current of the TENGs at room temperature and humidity. The mean values of the output voltage of TENGs based on dense, porous and CNT-added PDMS were 35 ± 19 V, 30 ± 13 V, and 27 ± 15 V, respectively (Fig. 5 a–c). Similarly, the average amount of output current is higher for dense PDMS (4.1 µA), compared to the porous electrodes generating peaks with maximum values of about 3 µA (Fig. 5 d–f). Since the variations of maximum peaks of current were slighter than the voltage diagram, humidity sensing was performed by recording the output current. Figure 5 Generated voltage and current for TENGs based on dense PDMS ( a and d ), porous PDMS ( b and e ), and CNT-PDMS ( c and f ). The reduction of produced electricity via porous electrodes was resulted from a lower dielectric constant. According to Eq. 1 , the induced charges on the back-contact, demonstrating the amount of short-circuit current, depending on the surface charge density (σ) and dielectric constant of the polymeric electrode (ɛ): 36 1 \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$ {\\text{Q}}_{{{\\text{SC}}}} = {\\text{S}}_{{\\upsigma {\\text{x}}}} ({\\text{t}})/\\left( {{\\text{d}}/\\upvarepsilon + {\\text{x}}\\left( {\\text{t}} \\right)} \\right) $$\\end{document} Q SC = S σ x ( t ) / d / ε + x t where S and d are the effective surface and the thickness of the electrode and x(t) represents the distance between the bottom and up electrodes, which is varying during every tapping. Since CNTs do not exist on the surface of CNT-PDMS (Figure S5 ), σ is similar for all electrodes. Regarding the fact that other parameters had similar values for all TENGs, the only effective factors should be S and ɛ. Surface porosity led to decreasing S, while the main reason for ɛ reduction was resulted from pores produced inside the electrode. Since ɛ of the polymer is higher than air, the porosity led to decreasing the amount of ɛ for both foam electrodes. Therefore, the triboelectric charge produced on the surface of the polymer reduced after the foaming process. The additional finer pores, which had been observed in SEM images, caused a little more reduction of ɛ for CNT-PDMS TENG. Humidity tests Figure 6 demonstrates the variation of the output current of the fabricated TENGs under different values of RH. By increasing the ambient humidity, the surface charge density of electrodes diminished gradually. For dense and porous pure PDMS (Fig. 6 a,b), the reduction of current does not proceed up to higher values of RH, while for CNT-PDMS, decreasing the output signal is observable within the whole range of RH (Fig. 6 c). To demonstrate the variations of output current, the I OC -time diagrams were sketched as shown in Fig. 7 a,b,c, which show the sensing behavior of the self-powered sensors as discussed above. Figure 6 Variation of short-circuit current for TENGs based on dense PDMS ( a ), porous PDMS ( b ), and porous CNT-PDMS ( c ) electrodes. Figure 7 Sensing behavior of fabricated TENGs. Short-circuit current variations of dense PDMS ( a ), porous PDMS ( b ), and CNT-PDMS ( c ) based TENGs against RH values. Comparison of current response vs. RH for TENG-based humidity sensors ( d ). As shown in Fig. 7 a–c, the reduction of output current shows diverse behaviors for the fabricated TENGs. By increasing the RH around the dense PDMS electrode, the generated current decreased gradually to reach a saturation value of 2 µA at RH > 60% (Fig. 7 a). Therefore, no effective sensing is reported at higher amounts of RH. For porous PDMS, the saturation was observed at lower amounts of RH by the mean value of 2.2 µA (Fig. 7 b). In contrast to pure polymer electrodes, CNT-PDMS was able to sense RH variation within the whole range of 30–80% (Fig. 7 c). The output current gradually decreased from 2.9 to 1.6 µA by increasing RH from 30% up to 80%. Moreover, the linear behavior of humidity sensing facilitates the process of data analysis 37 . On the other hand, the precision of the sensor, which is defined as the dispersion of obtained data at the desired RH, differs apparently for each electrode. According to Fig. 7 a–c, the error bars for gathered data are obviously lower for CNT-PDMS TENG, compared to the pure PDMS sensors. According to Eq. 1 , discussed in TENG Studies Section, the amount of charge induced on the back-contact electrode via finger tapping is proportional to the amount of surface charge density (σ). Since the amount of σ decreases by elevating the ambient humidity, smaller values of charge generate at higher RH 38 . Consequently, the transferred charges in the circuit declined for every tapping cycle, leading to lower amounts of short-circuit peaks in the diagrams shown in Fig. 6 . This reduction either continues within the whole range of RH values or might be restricted up to a certain amount of RH, depending on the nature of the sensing electrode 27 . To investigate the obtained data more thoroughly, the response values of current against the relative humidity were calculated employing the following formula: (I 0 − I)/I. Here, I 0 and I represent the quantities of current at RH = 30% and the desired RH, respectively. According to Fig. 7 d, showing the curves of humidity response of the electrodes, the CNT-PDMS electrode represents the highest response to humidity variations, up to 108% for RH = 80%. This is about two times of the highest values for dense and porous PDMS, which were 56% and 50%, respectively. Therefore, the addition of CNT improved humidity response by more than 100%. For both dense and porous PDMS, a saturation of humidity sensing was observed at higher amounts of RH. In contrast, for CNT-PDMS, the current decreased continuously up to high ambient humidity circumstances, due to the wide-range of pore distribution. For the primer electrodes, physisorption of water molecules reached a maximum value at a specific RH, and consequently, no more molecules had the opportunity of diffusing toward the surface of the electrode. For the CNT-PDMS electrode, alongside the saturation of adsorption in larger pores, the diffusion proceeds inside the smaller ones, which requires higher pressure of water molecules to initiate. This phenomenon occurred due to the hydrophilicity of CNTs, especially where presented on the edge of the pores, providing water diffusion into smaller pores. Therefore, the current is reduced by the physisorption of water molecules into a wide-range-size of pores. However, the above-mentioned mechanism led to delaying response and recovery sensing behavior for CNT-PDMS TENG compared to the porous pure polymer (Fig. 8 ). The response and recovery times are defined as the recorded time to reach 90% of the considered value. To measure the amounts of response and recovery times, we applied a setup consisting of two chambers with RH = 30% (the ambient humidity of the laboratory) and RH = 90% (provided by a humidifier) (Figure S6 ). The two chambers are connected via an aperture, by means of which the humidity can be augmented abruptly. For pure porous PDMS, the recovery time is 1.3 s, while it is 2.1 s for CNT-PDMS. Analogously, the recovery time after the addition of CNT increased from 1.8 s to 2.5 s. Therefore, adding CNT to PDMS results in the improvement of sensitivity, as well as precision, while simultaneously leading to a trivial decrease in the response and recovery rates, due to the finer porosity generated in the foaming process. Figure 8 Dynamic response of porous PDMS ( a ) and CNT-PDMS ( b ). Recovery and response times of humidity sensing are indicated. The sensing characteristics of the fabricated CNT-PDMS sensor are shown in Table 1 compared to the previous studies. The first outstanding feature of current work is not only being self-powered, due to applying TENGs but also the integration of the whole setup, resulting from the active circuit design. Moreover, the fabricated foams are highly light and flexible, beneficial for wearable/portable applications. Compared to previous investigations based on PDMS, CNT, or TENG-based polymeric sensors, the response and recovery times of our humidity sensor are appreciable. The proposed foaming procedure in this research opens a new avenue of facile nanocomposite production, appropriate for enhancement of gas sensing. Table 1 Comparison of humidity sensing characteristics of similar previous investigations with our fabricated CNT-PDMS sensors. Type Sensing material Flexible/Rigid Response time (s) Recovery time (s) Ref Capacitive nanofibrillated cellulose/graphene oxide/PDMS Flexible 57 2 39 Impedance Armalcolite/PDMS Flexible 10 15 40 Resistive CNT Rigid 40 60 41 Self-powered PFSA* Flexible 30 – 42 Passive TENG PTFE Rigid – – 43 Passive TENG rGO/PVP Rigid 2.8 3.5 44 Active TENG CNT-PDMS Flexible 1.8 2.5 This work *perfluorosulfonic acid ionomer."
} | 4,731 |
28462927 | PMC5418572 | pmc | 2,358 | {
"abstract": "Biofilms are social entities where bacteria live in tightly packed agglomerations, surrounded by self-secreted exopolymers. Since production of exopolymers is costly and potentially exploitable by non-producers, mechanisms that prevent invasion of non-producing mutants are hypothesized. Here we study long-term dynamics and evolution in Bacillus subtilis biofilm populations consisting of wild-type (WT) matrix producers and mutant non-producers. We show that non-producers initially fail to incorporate into biofilms formed by the WT cells, resulting in 100-fold lower final frequency compared to the WT. However, this is modulated in a long-term scenario, as non-producers evolve the ability to better incorporate into biofilms, thereby slightly decreasing the productivity of the whole population. Detailed molecular analysis reveals that the unexpected shift in the initially stable biofilm is coupled with newly evolved phage-mediated interference competition. Our work therefore demonstrates how collective behaviour can be disrupted as a result of rapid adaptation through mobile genetic elements.",
"discussion": "Discussion Stability of cooperative interactions can determine the performance of microbes in most medically and biotechnologically relevant situations 27 28 29 30 31 32 . In recent years, understanding of microbial group behaviours and the mechanisms that prevent spreading of non-cooperative mutants has become one of the key aims of sociomicrobiology. Long timescale evolutionary experiments have already demonstrated the evolutionary plasticity of social interactions in various bacterial models 18 33 34 . Here we describe a scenario where a biofilm matrix non-producer that is initially eliminated from the population increases its performance over longer timescales and evolves the ability to better incorporate into the biofilm. The evolution of improved invasion of biofilms by non-producers was previously observed by Zhang et al . 15 . They excluded the possibility of general adaptation being responsible for the changed social dynamics in biofilms since the evolved producer did not increase its performance towards the ancestor producer. In the present work, an increased selection coefficient and improved productivity of the evolved WT could be observed in monocultures; the same, unfortunately, could not be tested for the evolved mutants because of their inability to form pellicles in monocultures. We therefore hypothesize that the evolved increased-biofilm-incorporation-ability of the mutant was a side effect of extremely fast general adaptation of both producer and non-producers driven by mobile genetic elements. Interestingly, the improved incorporation of the non-producers into biofilms was not reproduced when both WT and non-producer strains had identical evolved genetic backgrounds (that of the evolved WT strains). This means that although the general adaptation pattern in the entire population was very similar, the non-producers are evolutionarily ahead of the producers and carry certain specific changes that allow their improved performance in incorporation into biofilms formed by the evolved WT strains. We believe those specific differences are hidden within prophage elements of the evolved strains, and they could be revealed by de novo sequencing in the future. In the ancestral population, the matrix non-producers (Δ eps –Δ tasA , which do not secrete two key matrix components Eps and TasA) can hardly incorporate into pellicle biofilms formed by the WT. This result was rather unexpected for two reasons: first, previous work demonstrated that both Eps and TasA are shared with non-producing strains 20 , and, second, the production of at least one of those compounds (Eps) was proven to be costly and exploitable 13 . Although we did not study the competition mechanism in detail, a positive correlation between fitness, initial Δ eps –Δ tasA frequency and resource availability suggests that in the ancestral population the growth of Δ eps –Δ tasA is not only limited by the lack of oxygen, but also by carbon resources. We speculate that this is caused by a delay in surface co-colonization of Δ eps –Δ tasA , because the producer can partially privatize the matrix components. Since the WT is released from oxygen-limitation first, it can quickly deplete the remaining carbon resources, preventing further growth of the mutant. This model, however, awaits further studies. The pellicle incorporation mechanism of evolved Δ eps –Δ tasA does not depend on resource concentration or on the initial frequency of the mutant in the co-culture. It is likely that new antagonistic interactions involving infection and lysis of the ancestor WT by the evolved mutant delay surface colonization by the WT, giving the mutant a prolonged window of opportunity for co-colonization. A similar mechanism could play a role in the competition between the evolved mutant and the evolved WT, since the evolved WT strains spontaneously release phages into the medium and show a delay in pellicle formation. How did the new lytic properties evolve? We believe that multiple rearrangements in the genomes of the evolved strains, combined with series of SNPs in regions that were rearranged, resulted in new lytic properties of the normally inactive domesticated SPβ prophage. Since this scenario was more likely to occur on sporulation treatment (that is, treatment method B), we suspect that the multiple heat-treatments involved in this treatment might have promoted phage activation 35 or even rearrangements of phage elements in the genome 36 . The accumulation of multiple SNPs and rearrangements resembles the previously reported evolutionary response of the Streptococcus thermophilus phage to the host's CRISPR system 37 , however, no CRISPR/Cas has yet been identified in B. subtilis . Alternatively, rapid diversification within prophage regions combined with lytic induction may be a universal adaptive pattern of bacteria to a biofilm lifestyle, as it was previously also observed during experimental evolution of Pseudomonas aeruginosa biofilms 38 . Our work also demonstrates how such newly evolved phage warfare shifts social dynamics in the bacterial population in favour of biofilm non-producers. The dynamics of host-phage interactions is long studied in various experimental systems 39 . It was previously observed that lytic phages can shift the balance in competitive interactions by reducing the frequency of a winning partner 40 , or impair biofilm formation ability as a trade-off for phage immunity 41 . We hypothesize that in the case of the B. subtilis pellicles, the disadvantage of matrix producers could originate from the degeneration of toxin/secretion-related genes in the evolved wild-types that in turn became less efficient competitors than the evolved mutants. The improved fitness of the evolved WT strain in monoculture could be a direct result of the evolved spontaneous phage release. Normally, the excision of the SPβ prophage from the B. subtilis chromosome takes place before sporulation and allows reconstitution of the spsM gene involved in spore polysaccharide biosynthesis 42 . Sanchez-Vizuete et al . 43 demonstrated that removal of SPβ from the chromosome permanently restores spsM , resulting in increased biofilm thickness. We presume that frequent spontaneous excision of SPβ, or even pseudolysogeny (as demonstrated in ref. 24 ) in the evolved WT strains, could positively contribute to the biofilm productivity through spsM reconstitution. Excision of prophage from the host chromosome was recently linked to improved biofilm formation by Shewanella oneidensis facing cold stress 44 . Such a phage excision benefited the host through gene inactivation rather than reconstitution (as observed in ref. 43 ). Similar genetic switches triggered by prophage excision were also described in several other species (reviewed in ref. 45 ). The B. subtilis SPβ prophage carries a bacteriocin-immunity system 46 , several putative toxin–antitoxin systems 47 and cell wall hydrolases 48 . Several SPβ segments of >250 nucleotides exhibit >90% identity with B. subtilis chromosomal regions 25 promoting recombination events, especially in naturally competent strains. Not surprisingly, recent reports strongly indicate a key role of phage elements in rapid evolution of kin recognition mechanisms and antagonistic interactions between closely related, sympatric B. subtilis strains 49 50 . Accumulation of SNPs in the SPβ region was also observed in the evolution experiments of Overkamp et al . 51 , where B. subtilis was kept in zero-growth conditions for 42 days. Among hundreds of SNPs discovered by Overkamp et al . 51 , 80% overlapped with the SNPs reported in this study. In addition, most of the SNPs detected were synonymous and evolutionarily conserved, suggesting selection against loss of function. Recent reports show that even non-synonymous mutations can positively contribute to fitness 52 53 . This again suggests that mutations or rearrangements within phage elements can be a very important evolutionary force in B. subtilis , with a major impact on social interactions. Recently, the profound impact of prophages on the evolution of a pathogenic bacterium was experimentally demonstrated in P. aeruginosa biofilms 54 , where the presence of phages resulted in strong selection against phage recognition elements (type IV pilus), at the same time enhancing parallel evolution 54 . Similar selective pressure could emerge after fast evolution of active SPβ variants in B. subtilis biofilms, resulting in striking parallelism in evolved populations of both WT and Δ eps –Δ tasA bacteria. Our work demonstrates how social dynamics in an initially very robust biofilm can be shifted by unexpected evolutionary events. We show that an adaptive genotype that is quickly tailored by mobile genetic elements can easily spread through horizontal gene transfer. The same adaptive path, although beneficial for the producer, became maladaptive in a mixed population where producers coexisted with non-producers."
} | 2,535 |
22924473 | null | s2 | 2,359 | {
"abstract": "Designing nanoscale objects with the potential to perform externally controlled motion in biological environments is one of the most sought-after objectives in nanotechnology. Different types of chemically and physically powered motors have been prepared at the macro- and microscale. However, the preparation of nanoscale objects with a complex morphology, and the potential for light-driven motion has remained elusive to date. Here, we go a step forward by designing a nanoscale hybrid with a propeller-resembling shape, which can be controlled by focused light under biological conditions. Our hybrid, hereafter \"Au@DNA-origami\", consists of a spherical gold nanoparticle with self-assembled, biocompatible, two-dimensional (2D) DNA sheets on its surface. As a first step toward the potential utilization of these nanoscale objects as light-driven assemblies in biological environments, we show that they can be optically trapped, and hence translated and deposited on-demand, and that under realistic trapping conditions the thermally induced dehybridization of the DNA sheets can be avoided."
} | 274 |
17888177 | PMC2151768 | pmc | 2,360 | {
"abstract": "Background Biological systems are often modular: they can be decomposed into nearly-independent structural units that perform specific functions. The evolutionary origin of modularity is a subject of much current interest. Recent theory suggests that modularity can be enhanced when the environment changes over time. However, this theory has not yet been tested using biological data. Results To address this, we studied the relation between environmental variability and modularity in a natural and well-studied system, the metabolic networks of bacteria. We classified 117 bacterial species according to the degree of variability in their natural habitat. We find that metabolic networks of organisms in variable environments are significantly more modular than networks of organisms that evolved under more constant conditions. Conclusion This study supports the view that variability in the natural habitat of an organism promotes modularity in its metabolic network and perhaps in other biological systems.",
"conclusion": "Conclusion This study indicates that the modularity of metabolic networks correlates with the variability of the environment. Such a correlation supports the view that variability in the natural habitat promotes modularity. It would be important to test this more fully as data on metabolic and regulatory networks of diverse species becomes more complete. We currently know more about the structure of metabolic networks than about the ecology of the organisms. It is a challenge to see how far one can go in what might be termed 'reverse ecology' [ 36 ]: inferring from the structure of biological system information about the environment in which it evolved.",
"discussion": "Discussion This study indicates that variability in the environment correlates with enhanced modular organization of metabolic networks, while constant environment correlate with a less modular structure. One interpretation of these findings can be made in the context of previous simulation studies of evolution in modularly varying environments [ 18 ]. The metabolic goal that a bacterium faces can be considered as a combination of sub-goals. An example of a sub-goal is the biosynthesis of an amino acid such as histidine. If histidine is missing in the environment, the bacterium must synthesize it. If histidine is present, the bacterium can down-regulate the biosynthesis pathway and instead import this metabolite. When the environment changes over time, it introduces a different combination of such metabolic sub-goals. Simulations suggested that varying the sub-goals leads to the evolution of networks with a modular structure, where each module corresponds to one of the sub-goals [ 18 ]. Modular structure evolves despite the fact that it is less optimal than non-modular solutions [ 46 ]. In contrast, evolution under a goal that is constant over time leads to non-modular networks, in which many nodes participate in several functions [ 18 ]. The present findings may be interpreted within this context: Bacteria that live under varying environments typically evolve a functional module for each of the varying sub-goals. Bacteria under constant conditions tend to evolve towards a less modular design. It is interesting to note that some metabolic goals are held relatively constant even when the environment changes. An example is energy metabolism, which is needed for growth by all of the bacterial species studied, in all environments. Analysis of the metabolic networks shows that the part of the network responsible for energy metabolism (central metabolism) is less modular than other parts of the network (such as biosynthesis of amino acids, nucleotides, vitamins etc) [ 44 , 47 - 49 ]. More generally, the fraction of the metabolic network devoted to constant goals (such as central metabolism) seems to increases as the environment becomes more constant (Fig S8c in additional file 2 ). An additional observation in the computer simulations [ 18 ] is that initially modular networks rapidly degrade into non-modular but more optimal structures when the goal becomes constant over time. Examples of such a degeneration of modularity can be seen by comparing the closely related species E. coli and Buchnera . E. coli lives in a variable environment, moving between its mammalian host and the external world. Buchnera lives in a more constant environment, as an endosymbiont of aphids. Buchnera is found to fuse together pathways that are separate in E. coli , and thus to achieve its metabolic goals with a smaller set of enzymes [ 50 ]. One example occurs in the histidine and purine modules (Fig 5 ). Both modules convert the metabolic substrates PRPP to AICAR in two distinct parhways. E. coli seems to maintain these alternative pathways because under different environments (histidine/purine rich environments) only one of the pathways is utilized. For Buchnera , on the other hand, as an endosymbiont that supplies amino-acids to its host, histidine biosynthesis is a fixed goal and under no regulation. Here, the purine module can count on the histidine module for AICAR production. The two pathways were thus combined into a single module, in which many of the genes are used for both functions [ 38 , 51 ]. It would be interesting to uncover other mechanisms that degenerate or enhance modularity by comparing networks of closely related species with different environments. Figure 5 Illustration of a mechanism that reduces modularity . The connection between purine and histidine pathways is presented for a . E. coli and b . Buchnera sp. APS . Whereas in E. coli the pathways are separated, in Buchnera the pathways are partially combined [51]. One limitation of the present study is the limited knowledge of metabolic networks for diverse species. The reconstructed networks, based on genomic data, were used to generate information about putative non-directed metabolic interactions. The present network representation ignore: i) directionality of reactions ii) reaction stochiometry iii) that only a fraction of the reactions are active under given environmental conditions (hence at best it offers only a static view on modularity). The above mentioned problems can be handled by more sophisticated network analyses [ 52 ]. Such studies employ correlated reaction sets as mathematically defined modules in biochemical reaction networks. They constitute groups of reactions in a network that always appear together in functional states of that network and therefore represent a functional module of the reaction network. Previous work has shown that these sets can include non-obvious groups of reactions and differ from groupings of reactions based on structural analyses of network topology [ 53 , 54 ]. One drawback of these latter approaches is that they require carefully annotated, genome-scale metabolic network, of which is only available for a handful number of species."
} | 1,729 |
34035172 | PMC8179172 | pmc | 2,361 | {
"abstract": "Significance Army ants form some of the largest insect societies on the planet and famously forage in mass raids, in which many thousands of ants stream out of the nest in search of live prey. Here we show that this complex collective behavior has evolved from group raiding, which is practiced by relatives of army ants with smaller colonies. Through laboratory experiments, we discovered that group raids and mass raids follow similar organizational principles and that mass raids emerge from group raids when colony size is artificially increased. This suggests that ancient expansions in colony size, rather than changes in individual behavioral rules, led to the evolution of mass raids in the first army ants.",
"conclusion": "Conclusion Typically, the mechanism for behavioral evolution is thought to be the modification of neural circuits for that behavior. For instance, courtship decisions in fruit flies evolve through modifications to the internal physiology and/or synaptic strength of courtship neurons ( 27 , 28 ). Vocal learning circuits in birds are thought to have evolved via the duplication and modification of ancestral motor circuits ( 29 ). Animals typically acquire the ability to perceive new stimuli by duplicating or modifying old receptor genes or evolving new ones and/or enhancing their sensitivity to old stimuli through expansions in existing sensory processing systems ( 30 , 31 ). Similar modifications have been observed or proposed to explain a variety of evolutionary changes in motor systems ( 32 ). In all these cases, the circuits immediately involved in that behavior are modified, altering the computations they perform on the timescale of that behavior. Our data suggest a slightly less direct mechanism for the evolution of foraging behavior in army ants. We propose that all doryline ants share similar neural circuits for raiding behavior and that instead the evolution of mass raiding from group raiding depends on a change elsewhere—likely in the circuits that ultimately regulate colony size. Changes in group size are known to induce qualitative changes in collective behavior in other cases as well. For instance, golden shiners form polarized swarms or milling schools depending on their group size ( 33 ). For the last few decades, it has been understood that changes in the size of an ant colony can also have dramatic effects on the organization of its foraging behavior. Theoretical work proposes that ant colonies may transition from individual to collective foraging with increasing colony size ( 34 ) and suggests that this may be driven by positive feedback within the nest ( 35 ). Empirical work demonstrates that Pharaoh’s ants ( Monomorium pharaonis ) undergo a phase transition from disordered to ordered foraging as their colony size increases ( 36 ). Moreover, colony size has been generally associated with foraging mode across the ant phylogeny. Species with small colonies typically rely on individual foragers, species with moderately sized colonies tend to use a mixture of individual and collective foraging strategies, and species with large colonies tend to rely primarily on cooperative, collective foraging ( 37 ). Here we provide a concrete example in which such changes in colony size have been instrumental in shaping the evolution of collective behavior over tens of millions of years.",
"discussion": "Results and Discussion Group Raids Have Stereotyped Structure. To furnish a detailed understanding of group raiding we systematically studied foraging behavior in the clonal raider ant, Ooceraea biroi , the only non-army ant doryline that has been propagated in the laboratory. In our efforts to establish this species as an experimental model we have developed high-throughput, automated tracking approaches to monitor individual and collective behavior ( 15 , 16 ), allowing us to study doryline foraging behavior quantitatively and under controlled laboratory conditions. In a first experiment, we set up nine colonies each of 25 individually tagged ants and filmed and tracked their foraging behavior while offering them a single small fire ant pupa once every 12 h (for experimental details see Methods ). Overall, we analyzed tracking data for 31 raids ( Methods ) ( 17 ). We found that O. biroi , like other non-army ant dorylines ( 11 , 18 ), forages in scout-initiated group raids ( Movies S1–S6 ; for ant foraging terminology see SI Appendix , Table S1 ). We decompose group raids into six distinct phases ( Fig. 1 A and B and SI Appendix , Fig. S1 ). First, in the “search” phase, one or a few scouts explore the arena. Once a scout has discovered food, she examines it briefly before becoming highly excited. In the “recruitment” phase, she runs homeward, and as she enters the nest the ants inside become active. In the “response” phase, a large proportion of ants inside the nest run toward the scout, exit the nest in single file, and move toward the food, retracing the scout’s homeward trajectory ( Fig. 1 A – C ). Most ants then stay on or near the food for a few minutes, while some run back and forth between the food and the nest, which we call the “preretrieval” phase. Variation in the length of this phase explains most variation in raid length, but its function is currently unknown ( Fig. 1 D and SI Appendix , Fig. S2 ). Next, during the “retrieval” phase, one to three ants begin to independently drag or carry the food back home, with no apparent help from their nestmates ( Fig. 1 and Movie S2 ). Finally, in the “postretrieval” phase, the last ants outside gradually return to the nest. To visualize the temporal structure of these raids we aligned and rescaled each phase of each raid and quantified three informative features: the number of ants outside the nest, the mean distance from the nest, and the sum of the speeds of all ants ( Fig. 1 E – G ). Our analyses show that group raids are highly stereotyped and mostly vary in the duration of the phases. Moreover, these raids are remarkably similar to published descriptions of group raids in other non-army ant dorylines ( 11 , 18 ), suggesting that group raids are stereotyped not just within O. biroi but across the Dorylinae generally. Fig. 1. The anatomy of a group raid. ( A ) Trajectories of ants at each phase of a representative group raid ( Movies S1 and S2 ), separated into six sequential phases (see Methods ). The orange track in the recruitment phase depicts the path taken by the recruiting ant, whereas tracks in all other phases depict the paths of all ants in the colony. ( B ) Overlay of trajectories from all six phases. ( Inset ) Snapshot of the colony at the peak of the response phase. A short tunnel separates the nest (small circle) from the foraging arena (large circle), and the food (blue spot) is at the top left. ( C ) Heat map showing the number of ants outside the nest over time. Thirty-one raids are sorted vertically by their duration and are aligned to the start of recruitment. ( D ) Representing each phase of each raid by the same color code as in B shows that variation in raid length is primarily determined by the length of the preretrieval phase ( SI Appendix , Fig. S2 ). We do not show the postretrieval phase here, because it has constant length by definition (see Methods ). ( E – G ) Aligning and rescaling each phase of each raid (see Methods ) and plotting the time course of the mean number of ants outside the nest ( E ), their mean distance from the nest ( F ), and the sum of the speeds of all ants (a measure of collective activity) ( G ) shows that the temporal structure of group raids is highly stereotyped. The error bands in E – G represent the 95% CI of the mean. Army Ant Mass Raids Evolved from Group Raids. Next, to infer the evolutionary relationship between group raiding and mass raiding, we combined our data on O. biroi with published descriptions of doryline biology and mapped relevant life history traits (taken from the literature; see SI Appendix , Table S2 ) onto a consensus genus-level phylogeny of the Dorylinae ( 10 ). Given the cryptic lifestyle of non-army ant dorylines, information on their biology is unfortunately sparse, which resulted in gaps in our character matrix. However, as our goal here is a coarse-grained inference of the lifestyle of the most recent common ancestor of the Dorylinae, these gaps do not constitute a prohibitive constraint. First, we reconstructed the colony size of the most recent common ancestor of extant doryline ants. We classified each extant doryline genus as having either small or large colonies, separated by a threshold of 10 4 workers (see SI Appendix , Table S2 for detailed sizes). Our reconstructions (see Methods ) found that ancestral dorylines lived in small colonies and that the two origins of army ants were each associated with massive expansions in colony size ( Fig. 2 and SI Appendix , Fig. S3 and Table S3 ). Next, we reconstructed the diet of the ancestral dorylines by classifying the food spectra of extant genera ( SI Appendix , Table S2 ). We found that ancestral dorylines were specialist predators of ants ( Fig. 2 and SI Appendix , Fig. S4 and Table S3 ). Such predators are a priori unlikely to be solitary foragers, because ant colonies are well-defended and able to kill solitary intruders. In addition, there are no recorded observations of doryline ants foraging (i.e., retrieving food) solitarily, or through any form of collective behavior other than group and mass raiding. Together, this lends support to the notion that the ancestral doryline ants foraged collectively through a form of raiding behavior. Indeed, as noted earlier, all observed doryline foraging can be classified as either group or mass raiding, according to the criteria we set out earlier ( SI Appendix , Tables S1 and S2 and Supplementary Note 1 ; see Methods for the classification method). Our formal reconstructions found that ancestral dorylines indeed foraged in group raids. A recent phylogeny of the Dorylinae estimated that their most recent common ancestor lived roughly 74 Ma ( 10 ). Thus, our analyses suggest that group raiding is an ancient form of foraging and that non-army ant dorylines have employed this strategy for tens of millions of years. Moreover, our reconstructions show that the origins of army ants were associated with transitions from group raiding to mass raiding, likely independently in the New World and Old World army ants ( Fig. 2 and SI Appendix , Fig. S5 and Table S3 ) ( 4 , 10 , 13 , 14 ). In other words, these analyses show that mass raids evolved from ancestral group raids and, by extension, that studying O. biroi might provide mechanistic insight into how a group raid might be transformed into a mass raid. Fig. 2. Phylogeny of the Dorylinae, showing all extant genera, along with maximum colony size, type of raiding behavior, and prey spectrum, where known. Ancestral reconstructions on a consensus cladogram ( 10 ) are shown at the base of the tree ( SI Appendix , Figs. S3–S5 and Methods ). Photographs from top to bottom show workers of the army ants Eciton burchellii and Dorylus molestus as well as the clonal raider ant O. biroi (highlighted by a red box). E. burchellii and D. molestus images credit: D.K. O. biroi image credit: Alexander Wild (photographer). Recruitment and Response Are Homologous across Group and Mass Raids. To understand this transformation, we must be able to compare group and mass raids explicitly. This requires the use of a common vocabulary to describe the elements of group and mass raids. Our phylogenetic analyses demonstrate that group raids are homologous to mass raids, which suggests we may apply our quantitative description of group raids to mass raids as well. Specifically, we may consider mass raids to entail the same set of phases in the same sequence, beginning with search and ending after prey retrieval. This would allow us to identify homologous phases, and ask in which phase(s) the evolutionary modifications occurred. Intuitively, one might expect the response phase of a group raid to be homologous to the onset of a mass raid, because these are superficially similar: They both represent columns of ants streaming out of the nest. However, homology is better established by identifying the behavioral rules involved in each case. Based on our own observations, as well as previous work on army ants and two distantly related non-army ant dorylines ( 3 , 9 , 11 , 18 ), we hypothesized that at least two distinct, scout-derived signals determine the spatial and temporal structure of group raids. First, we asked how the scout activates nestmates during recruitment. We conducted an experiment in a modified arena that had a porous wall in the middle of the nest chamber and separate foraging arenas connected to each nest half ( Fig. 3 A ). In each trial, food was placed in one foraging arena, and when a scout with access to that arena located the food she recruited the ants in her nest half, which formed a column that traveled to the food. Shortly after the scout entered the nest, the ants in the other nest half moved toward the wall separating the two halves ( Fig. 3 A and B , SI Appendix , Fig. S6 , and Movie S7 ) ( 17 ). This suggests that the scout releases an attractive recruitment pheromone as she enters the nest ( SI Appendix , Supplementary Note 2 ). Second, we asked whether the scout lays a pheromone trail back to the nest during recruitment and whether that trail is sufficient to guide the responding ants. Scout-initiated raiding has evolved independently on a few occasions in distantly related ant subfamilies, and in several cases the scout must lead the raiding party to the target. In other words, in these other species information about target location resides primarily in the scout, rather than in a pheromone trail (e.g., refs. 19 – 23 ). In contrast, we found that in O. biroi the scout usually (in 30/31 raids) does not lead the raiding column ( Fig. 3 C ). However, the trajectories of the responding ants closely recapitulate the homebound trajectory of the scout, suggesting that the scout indeed deposits trail pheromone on her way to the nest ( Fig. 3 D ). Information about prey location therefore resides exclusively in the scout’s trail. This use of pheromones is highly reminiscent of recruitment at the raid front in army ant mass raids, where scouts lay pheromone trail from prey to the raid front where they elicit recruitment ( 9 ). Together, this suggests that group- and mass-raiding dorylines use chemical information in the same way and that the recruitment and response phases of a group raid are homologous to recruitment and response at the raid front in mass raids. Fig. 3. A trail and a recruitment pheromone determine the spatial and temporal structure of group raids, respectively. ( A ) The recruitment pheromone is attractive and acts at a distance. The image shows a modified nest with a porous barrier down the middle. On the left side, a scout releases recruitment pheromone, causing the ants to leave the nest. The ants on the right side, meanwhile, run toward the barrier instead of leaving the nest. ( B ) The distance between the barrier and the center of mass of ants on the side opposite to that of the scout as a function of time since recruitment. The center of mass travels toward the barrier after recruitment, which shows that the recruitment pheromone is attractive ( n = 31 raids, error band shows 95% CI of the mean). ( C ) A histogram of the scouts’ position in the raiding column shows that scouts do not typically lead raids. ( D ) The outbound trajectories of responding ants are significantly closer to their scout’s inbound trajectory than they are to trajectories of scouts in other group raids (see Methods ), showing that the responding ants indeed follow their scout’s trail to the food ( n = 31 raids, two-sided Welch’s t test t = 22.77, P < 7 × 10 −29 ). Behavioral Rules for Search Are Conserved across Group and Mass Raids. Considering a mass raid to have the same sequence of phases as a group raid, it follows that, as with the other phases, the search phases of group and mass raids are also homologous ( Fig. 4 A ). In other words, despite their apparent differences, the onset of a mass raid is actually homologous to the search phase of a group raid ( Fig. 4 A ). Mass raids begin when workers spontaneously and synchronously leave the nest in “pushing parties” ( 3 , 24 , 25 ). At first, small groups of workers hesitantly leave the nest to explore its immediate vicinity. They lay trail pheromone as they walk, returning after only a few steps out. Ants continue to leave the nest, walking further and further out, confidently following their predecessors’ trail. When they reach untrodden ground, they also hesitate and turn, spreading outward along the raid front. Over time, this leads to a dynamic fan of ants traveling outward, leaving a strong, elongating trail back to the nest in its wake ( 3 , 24 , 25 ). In the species with the largest colonies the ants at the raid front can be so numerous that the raid advances as a swarm ( 3 ). In our initial O. biroi experiments, however, the search phase seemed less coordinated ( Fig. 1 E – G ). Fig. 4. Group raids turn into mass raids with increasing colony size. ( A ) The onset of a mass raid is homologous to the search phase of a group raid, despite the superficial resemblance to its response phase. Arrows indicate when a column of ants leaves the nest in each type of raid. ( B ) On average, early excursions (low excursion index) terminate closer to the nest than later excursions (high excursion index) ( n = 127 excursions, colony size 25, linear regression r = 0.65, P = 2 × 10 −16 ). ( C ) Four example sequences of nest exit times, sorted by colony size. ( D ) An example distribution of interexit intervals in a colony of size 20. This distribution (in amber) deviates significantly from a simulated exponential distribution (in gray) (Anderson–Darling k-sample test P = 0.001). ( Inset ) Difference between this distribution and 1,000 simulated exponential distributions, as a function of the interval, showing an increased coefficient of variation (red line is mean difference, with 95% CI). Nest exits in close succession (i.e., short intervals) are overrepresented in the empirical distribution compared to simulated distributions. ( E ) The coordination index (see Methods ) of real interexit intervals (red data points) increases as a function of colony size ( n = 131 exit sequences, linear regression r = 0.48, P = 4.9 × 10 −9 ), but the coordination index of shuffled interval sequences (gray data points) does not ( n = 131 exit sequences, linear regression r = 0.79, P = 0.89). ( F ) Schematic of a mass raid of the army ant Aenictus laeviceps , reformatted with modifications from ref. 42 . ( G ) Snapshot (background-subtracted and contrast-enhanced; see Methods ) of an O. biroi raid in a colony with ca . 5,000 workers. The raid shows all the major features of the army ant mass raid depicted in F . Error bands in B and E depict the 95% CI of the regression line. We thus asked whether O. biroi scouts follow the same basic behavioral rules for search that translate into pushing parties in mass-raiding army ants. First, we analyzed our tracking data from colonies of 25 workers to see whether ants incrementally increase their foraging distance by extending previously traveled paths. We found that O. biroi often (in 21/31 raids) search an arena that is initially void of trail pheromone in a series of excursions (see Methods ). Further analysis of these excursions revealed that, on average, early excursions terminate close to the nest, while later excursions terminate farther away ( Fig. 4 B ). Additionally, ants walk faster ( SI Appendix , Fig. S7 A ) and spend longer outside ( SI Appendix , Fig. S7 B ) in later excursions and are more likely to follow trail at the beginning, rather than the end, of the outbound leg of each excursion ( SI Appendix , Fig. S7 C ) ( 17 ). Thus, it appears that, as in army ants, O. biroi workers lay pheromone trail as they leave the nest in the search phase, and indeed, appear to follow the same rules during search behavior. Taken together, our results suggest that the basic behavioral rules underlying search behavior are conserved between army ants and their non-army ant relatives. Despite the similarities in individual behavior, the emergent collective search patterns in a mass raid and a group raid are strikingly different. Understanding how mass raids evolved thus requires us to understand how the search phase of the ancestral group raid was modified without changes to individual worker behavior. In other words, we must explain how pushing parties emerge during search in mass raids. Unlike in army ants, where workers leave the nest en masse to go on a raid, O. biroi workers typically leave the nest during the search phase in a seemingly sporadic manner ( Fig. 1 E ). To study the temporal structure of search in O. biroi and to quantify the synchronicity in the search phase we conducted an experiment with four colonies of size 20. To control for the possibility that ants behave differently when food is in the arena we specifically selected periods when the arena was empty (i.e., the ca . 20 h after each raid each day, resulting in a total of 43 search events). We then recorded each time an ant exited the nest in each event. We analyzed the resulting sequences of interexit intervals by comparing them to the uncorrelated, exponentially distributed expectation from a random Poisson process ( Fig. 4 C and D and Methods ). We found that nearly all distributions deviated significantly from the random expectation ( SI Appendix , Fig. S8 A ), exhibiting increased coefficients of variation ( SI Appendix , Fig. S8 B ), overrepresentation of short intervals ( SI Appendix , Fig. S8 C ), and positive correlations between consecutive intervals ( SI Appendix , Fig. S9 A ), implying that workers leave the nest in quick succession more often than expected by chance ( Fig. 4 D and SI Appendix , Fig. S8 C ). This suggests that, while the apparent synchronicity is weak, a significant positive feedback—which characterizes the onset of army ant raids—also underlies the search activity of O. biroi . Increasing Colony Size Transforms Stereotyped Group Raids into Mass Raids. Army ants live in much larger colonies than non-army ant dorylines, and expansions in colony size within the Dorylinae align perfectly with the evolutionary transition to mass-raiding behavior ( Fig. 2 ). We hypothesized that this increase in colony size could explain the origins of mass raiding. To understand the effect of colony size on the emergent search and raiding behavior in dorylines we established O. biroi colonies with 10, 50, or 100 workers, alongside the colonies of 20 workers described above. Although these colony sizes do not approach those of army ants, this experiment is nonetheless informative regarding the general scaling effects of colony size. Moreover, as the nest chambers of our experimental arenas are much larger than even the largest colonies, this experiment manipulates colony size but not their effective density. Across all colony sizes, colonies mostly exhibited the same stereotypical raid dynamics ( Movie S8 ), with the number of ants participating in the raids increasing proportionally to colony size ( SI Appendix , Fig. S9 B ). Analyzing their interexit interval distributions during search ( Fig. 4 C ), we observed the same increase in their coefficient of variation compared to the random expectation as before ( SI Appendix , Fig. S8 B ). Moreover, the correlation between consecutive intervals, as measured by the autocorrelation function of the sequences, markedly increased with colony size ( SI Appendix , Fig. S9 A ), as did a “coordination index” that we computed from the autocorrelation function ( Fig. 4 E and Methods ) ( 17 ). Thus, as colony size increases, search behavior in O. biroi begins to resemble the onset of highly bursty, coordinated army ant mass raids. However, unlike in a full-blown mass raid, these bursts typically attenuate quickly. Nevertheless, we observed multiple events in colonies of ≥50 ants in which positive feedback among the ants spontaneously produced a column that traveled away from the nest, headed by an obvious pushing party that formed without recruitment, in what resembled the onset of a mass raid ( Movie S9 ). To test whether these scaling effects persist at colony sizes that approach those of army ants, we established two O. biroi colonies of roughly 5,000 workers each, an order of magnitude larger than naturally occurring colonies ( 26 ), and filmed their raids in large arenas (see Methods ). The resulting raids involved thousands of ants and displayed trail bifurcations, simultaneously targeting multiple food sources ( Fig. 4 F and G , Movie S10 , and SI Appendix , Table S4 ). Most initial recruitment events now occurred outside the nest and usually at the raid front (43 out of 47). Thus, increasing colony size eventually transforms stereotyped group raids into raids that display all the defining features of army ant mass raids ( SI Appendix , Table S1 ). Together, our results suggest that all doryline ants share fundamental rules of search and recruitment behavior. At small colony sizes, these rules manifest as scout-initiated group raids. However, as colony size increases, either experimentally within species or naturally between species across evolutionary time, these rules gradually give rise to spontaneously initiated mass raids in which many ants leave the nest in quick succession, advance in pushing parties, and recruit at the raid front rather than at the nest. The difference between search behavior in group raiders and mass raiders may thus be largely driven by the effects of increasing colony size. In other words, expansions in colony size in the ancestors of army ants are sufficient to have caused the transition from group raiding to mass raiding behavior."
} | 6,562 |
34384219 | PMC8397237 | pmc | 2,362 | {
"abstract": "A novel\nsuperhydrophobic/superoleophilic surface has been developed\nby direct surface condensation of dichloroxylene that results in a\ncontrolled coating of hyper-cross-linked polymers. Specifically, the\ncoating was successfully applied to a melamine formaldehyde sponge\nand optimized by fine-tuning the reaction variables. The resulting\nhierarchical porous sorbents stabilized by polydimethylsiloxane exhibited\nan increased surface area, good physiochemical stability, high selectivity,\nand adsorption capacities for a variety of oils and solvents. The\ncomposite can separate oil in water emulsions with ultrahigh separation\nefficiency >99% over 10 cycles in liter-scale experiments, wherein\nthe highest separation efficiency was as low as 2 ppm even with a\nshort period of filtration, suggesting strong potential for oil/water\nseparation and recovery.",
"conclusion": "4 Conclusions By simply varying the reaction\nconcentration and time of a hyper-cross-linked\nsystem, a novel HCP coating has been developed. This method introduces\nhighly porous HCPs on a 3D porous matrix, which is reported for the\nfirst time. The existing relevant works involve column packing of\nporous polymers as solid phase extraction (SPE), fiber-based solid\nphase microextraction (SPME), where preformed porous polymers are\nfixed using silicone adhesives, and preparation of magnetic porous\norganic polymers, which are not versatile strategies for other applications. 36 We anticipate this method is potentially applicable\nfor numerous desired surfaces with tunable loading, surface area,\nporosity, pore diameter, and surface chemistry for specific use. There\nare also alternative approaches using nonhalogenated reagents, which\nare possible without DCX used in this study. 20 The facile fabrication of the sorbent material requires inexpensive\nstarting materials and mild operating conditions favorable for scale-up\nand therefore practical application in oil spill cleanup and wastewater\ntreatment. The recyclability of MF-HCP indicated that the lifespan\nof such porous coating is limited. Therefore, the introduction of\nPDMS binders for robustness is necessary. This strategy guaranteed\nthe long-lasting durability and optimized surface wettability without\ncompromising the surface area of HCPs, resulting in ultrahigh efficiency\nfor O/W emulsion separation and recovery. To further remove the oil\ndroplets, for instance, below 100 nm, is also theoretically achievable\nby tuning the pore size of the substrate. We believe the HCP-decorated\nsponge has great potential for water treatment and meets the stringent\ndischarge standards of effluents containing petroleum hydrocarbons\ninto the marine environment.",
"introduction": "1 Introduction Oil\nspill incidents have an adverse impact to the marine ecosystem,\neconomy, and society all over the world. 1 − 3 To minimize these consequences,\nan appropriate material for oil spill cleanup and recovery is required.\nThis presents a number of technical challenges requiring the surface\nto be optimized in terms of chemistry, surface area, charge, roughness,\nand environmental profile. Ideal materials possess superhydrophobicity,\nhigh porosity, surface area, and selectivity to hydrocarbons. In addition,\nthe pore size of any material must be tailored to allow rapid uptake\nwhile avoiding clogging and maintaining high uptake particularly under\nrepresentative conditions that would be encountered during a spill.\nAny material must also be recyclable and easy to operate under conditions\nencountered during an oil spill response. In addition, the materials\nmust be inexpensive and scalable. All these factors mean that progress\nin this field has been challenging. One approach that has been explored\nis using nanotechnology for improved separation and recovery during\nthe oil spill response, which has been reported to have encouraging\noutcomes. 3 This includes meticulously engineered\nporous fabrics/fibers, 4 , 5 sponges/foams, 6 − 8 metal meshes, 9 − 11 membranes, 12 , 13 ceramics, 14 aerogels, 15 , 16 free-standing porous materials 17 , 18 with superwetting properties, which are typically superhydrophobic/superoleophilic\nand superhydrophilic/superoleophobic (in air or under water), and\nresponsive moieties. However, many are limited to a laboratory scale\nwith disadvantages such as expensive starting materials, complicated\nsynthesis procedures, lack of durability, low flux, fouling, challenging\nregeneration, and long-term operation. This study presents a robust\nhierarchical porous superhydrophobic sponge achieved by nanocoating\nof hyper-cross-linked polymers (HCPs) that have the potential to overcome\nsome of these issues. HCPs are an important class of permanently\nmicroporous materials\nthat have promise for a broad range of applications such as gas storage,\ncarbon dioxide capture, water treatment, drug delivery, separation,\ncatalysis, and sensing. 19 − 21 They have been extensively studied\nin the past few decades owing to many distinct advantages such as\ndiverse synthetic routes, low-cost reagents, mild reaction conditions,\nease of functionalization, and high surface area. Nevertheless, existing\nstudies are more focused on fundamental examination from a gas sorption\nperspective and synthetic routes rather than applications of HCPs. 19 While some commercial HCP sorbents have already\nbeen used as highly effective alternatives to activated carbon for\norganic and metal ion removal in wastewater treatment, 22 there is limited research on the application\nof HCPs for oil spill cleanup. We envisioned that HCPs are also capable\nof separating hydrocarbons from water in the field of oil spill response\nas demonstrated in our earlier study via hyper-cross-linking of a\npolymerized high internal phase emulsion (PHIPE). In particular, we\ninvestigated a three-dimensional (3D) hierarchically porous HCP monolith\ncomprising abundant macropores and nanopores of a high surface area\nof 196–595 m 2 g –1 with an oil\nadsorption capacity of 800–1900% of its own weight, suggesting\nthe potential application of HCPs for an oil spill response. 18 However, the fabrication of such materials involves\npolymerizing HIPEs, which presents challenges in terms of large-scale\napplications. To the best of our knowledge, we report for the first\ntime an HCP-decorated hierarchically porous sorbent where porous HCPs\nwere introduced to the macroporous substrate for oil/water separation.\nNotably, this strategy is potentially a universal method for all desired\nsurfaces, and the facile fabrication is inexpensive, affordable, and\nsimple to scale-up. Melamine formaldehyde (MF) sponges were\nselected as the ideal substrate\nin this study because the 3D interconnected framework gives rise to\nhigh permeability for oil adsorption and its intrinsic flexible skeleton\nmakes the material mechanically robust. 23 , 24 There have\nbeen some studies reporting microporous material-coated sponges, 25 − 29 but these do not include HCPs. Jiang et al. developed a hydrophobic\nmetal organic framework (MOF)-coated sponge incorporated with −CF 3 fluorine groups, overcoming the vulnerability that most MOF\nmaterials have in the presence of water/moisture. The coated sponge\ntakes up 1200–4300% of organic pollutants including oils and\ncan be used for continuous oil spill cleanup assisted with a self-priming\npump. 27 Xu et al. made a hydrophobic MOF-coated\nsponge with an excellent oil adsorption capacity ranging from 6600\nto 13,000%. 29 Sun et al. successfully imparted\nsuperhydrophobicity on the sponge via the coating of covalent organic\nframeworks (COFs) postmodified with 1 H ,1 H ,2 H ,2 H -perfluorodecanethiol. The\ncomposite exhibited remarkable sorption capacities from 6700 to 14,200%. 28 MOFs and COFs are another class of microporous\nmaterial similar to HCPs, but they are crystalline unlike the amorphous\nHCPs. They offer great promise across a number of fields, but challenges\nlie in the complex synthesis conditions and scale-up of inexpensive\nstable structures. 25 , 26 Compared with these emerging\nmicroporous materials, HCPs have advantages in terms of inherent stability,\nefficient fabrication, and comparable functions through surface coating.\nNevertheless, none of the aforementioned microporous material-coated\nsponges were reported to demonstrate the feasibility of oil/water\nemulsion separation, which is a much more serious challenge to any\nnew material. In this work, we demonstrate the controlled coating\nof nanosized\nHCPs on a sponge to transition the wettability from hydrophilicity\nto superhydrophobicity. The material was fabricated using a direct\nsurface coating of low-cost and mild operating conditions, which offers\nadvantages in terms of easier scale-up. The novel sorbent in conjunction\nwith a silicone elastomer formed a robust porous surface, 30 − 32 which has an addition of specific surface area, good mechanical,\nchemical, and thermal stability, high selectivity, and adsorption\ncapacities over various oils and organic solvents. The oil in water\nemulsion (O/W) separation performance was evaluated through large-scale\napparatus for carrying out liter-scale tests, which confirmed the\nultrahigh efficiency of HCP-coated sponges as a sorbent filter for\npractical application such as tertiary treatment.",
"discussion": "3 Results and Discussion 3.1 Synthesis of the Superhydrophobic HCP-Decorated\nSponge The major advantages of HCPs for an oil spill response\nare the molecular structure, rigid network, permanent porosity, and\nhigh surface area. However, the poor solution processability of HCPs\nhas limited their versatility for many applications. Unlike polymers\nof intrinsic microporosity (PIMs), HCPs are insoluble in solvents\nand are difficult to process into devices for many industrial uses\nwhile handling in particle form is not desirable. 33 , 34 It is also challenging to use preformed HCPs from bulk synthesis\nas a coating material due to the nature of the Friedel–Crafts\nreaction that proceeds rapidly and randomly even under ambient condition\nand often generates polydisperse particles in the micron size range\n( Figure S1 ). The appropriate particles\nfor coating upon a desired surface, however, should have a controlled\nsize, which would lead to homogenous distribution and stable attachment. In this study, we have coated nanosized HCPs upon a substrate via\ndirect surface condensation of dichloroxylene (DCX). This was achieved\nby manipulating all possible variables including building blocks (starting\nmonomers), external cross-linkers, as well as concentration, time,\npressure, and temperature of the reaction, which could determine the\nfinal coating characteristics. Ambient conditions were chosen for\nthis reaction out of practical scale-up purpose. Concentration and\ntime of the reaction were fine-tuned to control the film density (loading)\nand degree of cross-linking of the HCPs on the commercial melamine\nsponge. Therein, the degree of cross-linking is crucial to the resultant\nparticle size and surface area of the polymers. A series of\nconcentrations by mass, that is, 0.1, 0.25, 0.5, 1,\nand 2% of DCX dissolved in DCE, has been used for surface coating,\nrespectively ( Figure S2a ). It was found\nthat the higher the concentration, the faster the reaction proceeded,\nresulting in a higher degree of cross-linking as well as the amount\nof particle loading. The 1 and 2% reaction mixtures began the surface\ncondensation reaction at the mixing of DCX and FeCl 3 solutions\nin the presence of the sponge and caused too many HCP particles to\nquickly form, while the reaction rate of 0.1% was drastically retarded\ndue to the lowest concentration, and the coating could not be visually\nobserved on the sponge. Correspondingly, the appearance of the decorated\nsurface varied from dark brown to pale yellow reflecting the different\ncross-linking degrees and loadings of HCPs. Further scale-up experiments\nat 0.25% with a prolonged reaction time of 20 h was found to provide\nthe most preferable coating, in terms of particles distribution on\nthe surface ( Figure S2b ). In comparison,\nthe coating resulting from 0.5% solution was incomplete with the center\nof the sponge visually showing no coating, which was possibly due\nto the detachment of larger particles from the surface during the\nreaction ( Figure S2c ). The increased reaction\ntime contributed to a higher cross-linking degree of HCPs for coating\napplication at a given concentration. The uniformly coated sponge\nMF-HCP 0.25% had mass increase about 69%, which indicated sufficient\nloading for applications. However, it was found that the resulting\ncomposite was susceptive to mechanical loss of the HCPs. Therefore,\nPDMS was utilized to fix the HCP because it is known for its durable\nand elastic properties and has been widely used to bridge the nanoparticles\nto the surface. Hence, thermally curable PDMS of different concentrations\nwas further applied to improve the stability of the HCP coatings.\nAs seen in Figure 1 a HCP coating without PDMS can shed off completely in an ultrasonic\nbath within 15 min. By comparison, sponges coated with PDMS at concentrations\nover 1% can survive the same harsh test as well as other external\nforces such as extrusion and shredding and no obvious particle loss\ncan be observed. Moreover, the use of PDMS resulted in a considerable\nincrease in the water contact angle varying with the concentrations,\nas shown in Figure 1 b, which is favorable for oil/water separation application. The scheme\nof the fabrication of the robust superhydrophobic HCP sponge is shown\nin Figure 1 c. Figure 1 (a) Coating\nstability against mechanical forces examined by ultrasonicating\nthe samples in toluene for 15 min. The MF-HCP as the control and MF-HCP-PDMS\nsponges generated at different PDMS concentrations 0.5, 1, 2, and\n3% were labeled, respectively. (b) Contact angle of sponges as a\nfunction of PDMS concentrations. (c) Fabrication of MF-HCP-PDMS (the\ndeionized water dyed with methylene blue was used to show the wettability\ntransition of the pristine sponge from hydrophilicity to superhydrophobicity\nas a result of the HCP coating, and toluene dyed with Sudan red was\nused to demonstrate the superoleophilicity of the composite). 3.2 Surface Morphology To confirm that\na homogeneous coating was generated, SEM was utilized to observe the\nmicrostructural and surface morphology of the sponge without coating\nand compared to the HCP-decorated sponge. As shown in Figure 2 a, the pristine sponge has\nan open, interconnected network of a smooth skeleton with pore sizes\naround 150 μm, showing large macropores and volume. As shown\nin Figure 2 b, the sponge\nretains its inherent structure, pore size, and porosity after coating\nwith HCPs. The nanoparticles deposited on the skeleton homogeneously,\nresulting in an obvious change of surface morphology. A thin base\nfilm of polymers formed throughout the sponge, altering the original\nappearance and roughness of the sponge. Moreover, emerging small bulges\nand large clumps further contributed to the establishment of the hierarchical\narchitecture ( Figure S3 ). Larger clumps\nmay develop from small bulges as a result of smaller nanosized HCPs\nthat assembled and cross-linked. This is also confirmed by Fourier\ntransform infrared (FTIR) spectroscopy results ( Figure S4 ) with the emerging band around 2921 cm –1 indicating the presence of HCP containing C–H bonds. The\nmorphology looked interestingly like a cluster where the network of\nthe submicron spheres with a diameter 300–500 nm was supported\nby numerous “trunks” ( Figure 2 c). Figure 2 d–f shows a significant morphology change after\nthe incorporation of PDMS top layers; the macroporosity of the sponge\nremained the same with a slightly reduced pore diameter about 140\nμm, while the overall roughness was further enhanced, and the\nmajority of HCP particles were sparingly covered with PDMS, forming\na new intertwined porous networks; this was in good agreement with Figure S4 where the strong bands of Si–O–Si\nand Si–CH 3 appeared at 789, 1010, and 1258 cm –1 , respectively. In general, SEM imaging of this complicated\nmicro-nano hierarchical structure consisting of a base film, small\nbulges, large clumps, and a PDMS top layer has showed that this novel\nporous coating has been successfully generated upon the sponge. Figure 2 SEM images\nof (a) pristine melamine foam; (b, c) MF-HCP at a magnification\nof 500× and 1000×; (d–f) MF-HCP-PDMS-1 at a magnification\nof 100×, 500×, and 1000×. 3.3 Surface Area HCPs typically have\nhigh surface area when synthesized in the bulk phase due to the high\ndegree of cross-linking in the network in the solvent swollen state,\nwhich prevents collapse of the structure on drying. We envision that\nthe presence of HCP nanoparticles on the sponge should provide an\nextra surface area despite the limited cross-linking degree due to\nthe low reaction concentration for optimal coating. To verify the\nhypothesis, the total surface areas of MF-HCP, HCP surface , and HCP batch have been investigated as part of coating\ncharacteristics. The BET analysis shown in Table S1 confirms this. It is worth noting that the BET surface area\nof the pristine sponge is not applicable because the macroporosity\ngoes beyond the multilayer gas sorption theory. In contrast, the surface\narea of the coated sponge was 41.1 m 2 g –1 , indicating the successful incorporation of nanoporous materials.\nThe surface area of an HCP synthesized in bulk is typically >1000\nm 2 g –1 , but the overall amount of HCP\non the foam is limited, so the increase in surface area from essentially\nzero for the pristine foam to 41.1 m 2 g –1 is a strong indication of the porous structure of the HCP. HCP surface has a similar average pore size to MF-HCP, while the\nsurface area and pore volume are twofold more than that of MF-HCP,\nsuggesting that the surface area and pore volume of MF-HCP are inversely\nproportional to the weight of the composite on the grounds that MF-HCP\nconsisted of 41% nanoparticles and 59% sponge calculated by weighing\nbefore and after. The surface area of HCP surface , however,\nis not comparable with that of HCP batch , which is in good\nagreement with the previous discussion that the reaction concentration\nhas a major influence on the degree of cross-linking and thus the\nresultant particle size and surface area of the HCPs ( Figure S5 ). To further investigate the influence\nof PDMS outermost coating on MF-HCP, the surface areas of MF-HCP-PDMS-1,\nMF-HCP-PDMS-2, and MF-HCP-PDMS-3 were measured. The BET results revealed\nthat MF-HCP with 1% PDMS almost maintained its original surface area\nwhereas 2 and 3% PDMS greatly reduced the surface area because the\nexcess amount of PDMS began to block the pores of HCPs. In all cases,\nPDMS coating reduced the pore volume and pore size. Also, a slight\ndecrease in the water contact angle with 3% PDMS ( Figure 1 b) can be seen, suggesting\na counteraction effect upon the surface roughness. In summary, though\nat the expense of compromised surface area, the regulated nanosized\nparticles are advantageous for coating application over HCPs from\nbulk synthesis and the loading amount is appropriate for the desired\nsurface. Furthermore, MF-HCP-PDMS-1 was determined to be the preferable\ncandidate in terms of surface area, robustness, and superhydrophobicity,\nwhich were essential for oil/water emulsion separation. 3.4 Superwettability, Adsorption Capacity, and\nRecyclability The surface morphology and chemistry are known\nto determine the surface wettability that are closely related to oil\nadsorption. The hydrophobicity enhances the adsorption of oils simultaneously\npreventing the uptake of water. Therefore, superhydrophobic sorbents\nwith high selectivity are one of the most promising designing strategies.\nThe nanocoating of HCPs reported in this study can transform the hydrophilic\nmelamine sponge to superhydrophobic surface with a water contact angle\nof about 150° in one step ( Figure 1 c and Figure 2 b). This is attributed to the inherent physicochemical properties\nof HCPs, including the hydrophobicity from low surface energy of repeating\nbenzene units and hierarchical roughness by the hyper-cross-linked\nporous framework disturbed on the surface altered the wetting properties\ndrastically. In addition to the superhydrophobicity, the superoleophilic\nfeature of the HCP-decorated surface was found outperform the hydrophobic\nsurface coated with PDMS only as a control. Figure S6 shows that both MF-HCP-PDMS and MF-PDMS have an oil contact\nangle of 0°, but the time-lapse images reveal the exceeding adsorption\nrate of MF-HCP-PDMS compared to MF-PDMS. The modified sponge repels\nwater even by pressing down into the water and shows a “silver\nmirror” effect ( Movie S1 ). This\nphenomenon is caused by the reflection from the thin layer of air\ntrapped in the hierarchical architecture, which is a signature of\nthe Cassie state. 35 The composite bounces\noff the water around itself and floats on the top immediately when\nthe pressure is released, leaving the surface unwetted. When in the\nscenario of floating oil on the water surface or heavy oil under water,\nthis vast volume composite with superoleophilic character can readily\ntransport the hydrocarbons from water into the hierarchical pores\nat the contact of the oils and solvents though capillary effects ( Movies S2 and S3 ). To further investigate the superwettability of HCP-decorated sorbents,\na range of organic solvents and oils were used for adsorption experiment\n( Figure 3 ). Depending\non the different polarities, viscosities, and densities of sorbates,\nMF-HCP exhibits superior adsorption capacities ranging from 4800 to\n10,300%, while MF-HCP-PDMS-1 showed lower capacities 3400–8100%.\nThis can be owed to the mass increase of PDMS coating, which in turn\ndecreases the adsorption capacities calculated by adsorption of sorbates\ndividing mass in total. This trend can be confirmed by further increasing\nthe concentration of PDMS coating ( Figure S7 ). The regeneration of MF-HCP and MF-HCP-PDMS-1 was examined by five\nadsorption/desorption cycles using a model oil by centrifugation ( Figure 4 ). The use of centrifugation\ninstead of mechanical compression allows for the efficient removal\nof the sorbates from the adsorbents. More importantly, this experiment\nindicates that the adsorption of oils not only exists among the macropores\nof the sponge but also the micropores of HCPs. This can be interpreted\nfrom the adsorption capacities and residual oils of MF-HCP and MF-HCP-PDMS-1\nshown in Figure 4 .\nWithin expectations, there was a decrease in terms of overall adsorption\ncapacity and barely any residual oil was observed by the last cycle\nin the case of MF-HCP due to the significant particle loss, which\ncan be observed during the test. In contrast, MF-HCP-PDMS-1 showed\na steady adsorption capacity and residual oil, indicating no particle\nloss during the adsorption/desorption cycles. These findings were\nin good agreement with the discussions before. Figure 3 Adsorption capacities\nof MF-HCP and MF-HCP-PDMS-1 toward a variety\nof solvents and oils. Figure 4 Recyclability of (a)\nMF-HCP and (b) MF-HCP-PDMS-1 over Shell Tellus\noil. 3.5 O/W Emulsion\nSeparation To evaluate\nthe separation performance, Shell Tellus hydraulic fluids were used\nfor generating O/W emulsions. The model oil composed of refined mineral\noils with hydrocarbons from C10 to C34 and additives can easily form\nstable emulsions without the use of surfactants, which is desirable\nfor the study. MF-HCP-PDMS (2 g) was loaded as the filter in a stainless-steel\ncylinder where 2 L of 1000 ppm O/W emulsion in a reservoir tank was\npumped through for each experiment, as shown in Scheme 1 . The HCP sorbent was repeatedly regenerated\nby the solvent washing method and reused after blow-drying. Hexane\nwas employed as an inexpensive, relatively safe, and fast evaporation\nsolvent that is pumped through a custom-built separator for a specified\ntime to extract the adsorbed oils followed by decanting the mixture\nand drying the sorbent with an air compressor connected to the system.\nThe constant flow rate was set at 50 mL/min due to the upper pressure\nlimit of the system. As can be seen in Figure 5 a, the separation efficiency as a function\nof separation time indicates that the efficiency reaches 98.8% (12\nppm left) after 10 min and the maximum efficiency of 99.8% (2 ppm\nleft) within 30 min separation in the liter-scale experiment. Since\nthe size of the sponge after loading into the cylinder is compressed\nto be one third of the original, which in turn makes the average pore\nsize about 45 μm. Despite such a large pore size compared to\nmicro-nano emulsions, the sorbent filter can still effectively separate\nthe oil droplets above the average diameter of 350 nm ( Figure 5 c). This can be attributed\nto the adsorption of hierarchical pores existing in this unique superhydrophobic\nsponge ( Scheme 1 ).\nThe addition of HCPs accelerates breaking of the emulsions and diffusion\nof hydrocarbons into the sorbent. Given the particular discharge standard\nestablished by local regulators during the oil spill response, the\nrecovered water on a temporary storage device could be decanted if\nthe total petroleum hydrocarbons is below a certain amount, for instance,\n15 ppm, according to international law (MARPOL 73/78); therefore,\nthe ultrahigh efficiency of MF-HCP-PDMS is far superior for decanting\napplications. Figure 5 b shows the long-term robustness of MF-HCP-PDMS, and the performance\nis highly consistent over 10 cycles and remains the same after a total\namount of 40 L of emulsions was tested. Figure 5 (a) O/W separation efficiency\nof MF-HCP-PDMS as a function of time.\n(b) Reuse of MF-HCP-PDMS. (c) Illustration of one cycle of separation\nbefore and after, as well as the size distribution of emulsions in\nthe feed (average 3.6 μm) and filtrate (average 0.35 μm). Scheme 1 Configuration of the O/W Emulsion Separator and Separation\nMechanism\nof the HCP Sorbent Filter"
} | 6,495 |
39881373 | PMC11776243 | pmc | 2,363 | {
"abstract": "Background Fungal pretreatment for partial separation of lignocellulosic components may reduce lignocellulose recalcitrance during the production of biofuels and biochemicals. Quantitative and qualitative modification of plant lignin through genetic engineering or traditional breeding may also reduce the recalcitrance. This study was conducted to examine the effects of combining these two approaches using three white rot fungi and mulberry wood with an altered lignin structure. Results Mulberry wood prepared from homozygotes or heterozygotes with a loss-of-function in the cinnamyl alcohol dehydrogenase gene ( CAD ) was pretreated with three fungal species. Both heterozygous ( CAD / cad ) and homozygous ( cad / cad , null mutant) mulberry plants were derived from the same parents via backcrossing between Sekizaisou ( cad / cad , seed parent), a natural lignin mutant, and its F1 progeny ( CAD / cad , pollen parent). Homozygote wood and the isolated lignin exhibited an abnormal color. Lignin in homozygotes without fungal pretreatment exhibited a lower syringyl/guaiacyl ratio, molar mass, and thioacidolysis product yield than those in heterozygotes. Pretreatment with Phanerochaete chrysosporium achieved the highest delignification efficiency with a significant reduction in the cellulose content in both mulberry genotypes. In contrast, Ceriporiopsis subvermispora selectively removed lignin, with a weaker reduction in the cellulose content. The degree of delignification by C. subvermispora was significantly higher in homozygotes than in heterozygotes. Trametes versicolor tended to have a lower delignification capacity and smaller effect of subsequent enzymatic sugar release toward the wood from both genotypes than the other two fungi, making it less suitable for fungal pretreatment. Thioacidolysis assays indicated that cinnamaldehyde β-O-4, a typical subunit in the homozygote lignin, did not contribute to the high degradability of the lignin. The saccharification efficiency tended to be higher in homozygote wood than in heterozygote wood under all fungal pretreatment conditions. Conclusions Although further optimization of various system conditions is required, our findings suggest that CAD deficiency promotes delignification and subsequent enzymatic saccharification and may improve the biorefining efficiency of wood when combined with fungal pretreatment. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-025-02611-y.",
"conclusion": "Conclusions Our in-depth analysis of lignin modification and fungal degradation in two genotypically different mulberries showed that genetic modifications at the CAD1 locus had a significant effect on the lignin content, chemical composition, CEL molar mass, and fungal degradation of the wood. We found that wood deficient in CAD showed improved enzymatic saccharification when combined with fungal pretreatment. Saccharification efficiency is influenced by a variety of factors, including the species of fungi and pretreatment conditions including temperature, moisture, nutrients, and lignocellulose size. Thus, optimization of these factors may further improve the biorefining efficiency of lignin-modified wood.",
"discussion": "Results and discussion Wood appearance of Sekizaisou progeny Prior to fungal pretreatment and subsequent enzymatic saccharification, we characterized the wood of backcrossed progenies with different genotypes ( cad/cad and CAD/cad ). At harvest, the wood (xylem) under the bark had varying colors. After pulverization and sequential solvent extraction, the colors still differed, with cad/cad (homozygote) wood exhibiting a pale orange color, whereas CAD/cad (heterozygote) wood exhibited a light yellow (normal) color (Fig. 1 ). The colors of fresh and extractive-free homozygote wood and its extractive-free powder were similar to those of Sekizaisou, which also has a cad/cad genotype at the CAD1 locus and an altered lignin structure [ 40 ]. The red wood phenotype is a typical characteristic and good indicator of CAD deficiency in mutant and transgenic plants [ 25 , 41 – 43 ]. These findings confirm that the homozygous cad -null allele in the Sekizaisou progeny alters the wood color, a recessive trait, as reported in other plant species. Fig. 1 Appearance of wood powders prepared from heterozygote ( a , CAD/cad ), homozygote ( b , cad/cad ), and cellulolytic enzyme lignins ( c , d ) Chemical characterization of lignin in homozygote and heterozygote woods The composition of the progeny wood was characterized using a combination of gravimetric and chemical analyses (Table 1 ). The Klason lignin (KL) content tended to be slightly lower in homozygotes than in heterozygotes ( P = 0.11). The acid-soluble lignin (ASL) content was 17% lower in the former than in the latter ( P < 0.001). These data are partially consistent with the low lignin content in Sekizaisou, as measured using the acetyl bromide method, compared with that in other mulberry varieties [ 40 ]. The KL content in 1-year-old stems of field-grown transgenic poplars with strong CAD suppression was also reported to be 10% lower than that in wild-type poplars [ 44 ]. Table 1 Chemical composition of heterozygous ( CAD/cad ) and homozygous ( cad/cad ) samples Progeny genotype Chemical composition (dry weight %) Klason lignin Acid-soluble lignin Cellulose Hemicellulose CAD/cad 24.9 ± 0.8 1.7 ± 0.0 51.7 ± 4.1 14.6 ± 1.0 cad/cad 22.2 ± 0.3** 1.4 ± 0.0** 47.4 ± 0.7 17.1 ± 0.7 ** P < 0.01, compared with the heterozygote sample Samples from three independent progenies of the same genotype were analyzed The polysaccharide content of homozygotes and heterozygotes also differed. The cellulose content estimated from glucose released after acid hydrolysis of the wood powder was 12% lower than that of heterozygotes ( P < 0.001). In contrast, the hemicellulose content derived from the sum of four monosaccharides (mannose, galactose, xylose, and arabinose) tended to be higher in homozygotes than in heterozygotes ( P = 0.054). Sekizaisou, the seed parent of the progenies, also showed a lower cellulose (as glucose equivalent) content than did other mulberry varieties [ 38 ]. In a 1-year-old transgenic poplar with CAD suppression, the decrease in the lignin level was accompanied by a minor reduction in cellulose (average 3%) and small increase in hemicellulose levels (average 5%) compared with that in the wild-type plant [ 44 ]. Our results and those of previous studies indicate that a significant decrease in CAD activity leads to a slight decrease in the lignin content and change in the sugar composition in young woody plants. Lignin monomeric composition The monomeric composition of lignin was characterized using thioacidolysis (Table 2 ) [ 45 , 46 ]. Calculations based on detection of conventional thioethylated C 6 –C 3 monomers showed that the ratio of S to G units (S/G) in homozygous lignin (0.91) was significantly lower than that in heterozygous lignin (1.39). Similar trends were observed in other angiosperm species with CAD suppression [ 32 , 47 ] as well as in a study comparing CAD-deficient Sekizaisou with other CAD-functional mulberry varieties [ 40 ]. Although the difference was not significant, the total amount of thioacidolysis monomers was 14% lower in the homozygous lignin than that in the heterozygous lignin. Table 2 Thioacidolysis monomer yields from heterozygotes ( CAD/cad ) and homozygotes ( cad/cad ) Progeny genotype Thioacidolytic monomer yields (μmol g −1 of lignin) S/G ratio a G monomer S monomer Total CAD/cad 739.1 ± 59.4 1023.4 ± 70.9 1762.5 ± 129.8 1.39 ± 0.03 cad/cad 786.8 ± 11.1 718.4 ± 52.2* 1505.2 ± 56.0 0.91 ± 0.06** a S:G monomer ratio * P < 0.05, ** P < 0.01, compared to the heterozygote sample Samples from three independent progenies of the same genotype were analyzed CAD deficiency is commonly associated with the release of indene derivatives from the β-O-4 substructure coupled with cinnamaldehyde components in lignin under thioacidolysis [ 26 , 29 ]. To confirm the changes in the lignin structure of the homozygotes, the thioacidolytic products were analyzed using gas chromatography/mass spectrometry (GC/MS). Four candidates were detected, three of which (1 G1 , 1 S1 , and 1 S2 ; Figure S1) were identified as indenes based on their mass spectra [ 40 ]. These compounds were also detected in thioacidolysis products prepared from Sekizaisou wood [ 40 ]. No indenes were detected in the heterozygote wood. The indenes could not be quantified in the present study because authentic compounds were not available; however, S-type indenes (1 S1 , 1 S2 , and with an asterisk in Figure S1) were clearly more abundant than was G-type indene (1 G1 ) (Figure S1). This result suggests that lignin in the homozygotes contains more β-O-4 substructure with sinapaldehyde unit than that with coniferaldehyde unit as reported previously in CAD-deficient Sekizaisou [ 40 ] and CAD -suppressed poplars [ 35 ]. If re-calculation is conducted to take this finding into consideration, the S/G ratio of homozygotes lignin is expected be even larger than the value (0.91) presented in Table 2 . In addition to the colored wood phenotype, these findings confirm successful establishment of CAD-deficient progeny. Total weight loss and sugar degradation under fungal pretreatment For both genotypes of mulberry progeny, weight loss was greater in the order of pretreatment with PC, TV, and CS. Weight loss did not differ under PC, but homozygotes lost significantly more weight under TV and CS pretreatment than did heterozygotes (Table S1). These findings suggest that the change in lignin structure of homozygotes had a positive effect on degradation of the woods by these fungi under our experimental conditions. In heterozygotes, cellulose degradation occurred in the order of PC > TV > CS, with significant differences between each sample. In homozygotes, the reduction in cellulose was also significant under PC, but its degradation was slower under TV and CS pretreatments, with no significant difference between them (Table S1). In both genotypes, cellulose degradation was the strongest contributor to weight loss under PC pretreatment (Tables 3 and S1). In contrast, weight loss in TV- and CS-pretreated wood was more affected by a reduction in the KL content than by polysaccharide degradation. Table 3 Chemical composition of wood powder from heterozygote ( CAD/cad ) and homozygote ( cad/cad ) before (WF) and after fungal pretreatment Progeny genotype Fungal strain Chemical composition (wt.% of non-co-cultivated wood powder) b Klason lignin Acid-soluble lignin Cellulose Hemicellulose CAD/cad Without fungi a 24.9 ± 0.8 1.7 ± 0.0 51.7 ± 4.1 14.6 ± 1.0 PC 17.8 ± 1.1 x 1.5 ± 0.0 x 40.7 ± 1.1 x 10.1 ± 1.2 TV 19.9 ± 0.0 y,z 1.8 ± 0.0 y 47.8 ± 0.8 y 12.9 ± 0.2 CS 19.1 ± 0.1 x,z 2.4 ± 0.1 z 50.8 ± 0.8 z 13.3 ± 1.0 cad/cad Without fungi a 22.2 ± 0.3 1.4 ± 0.0 47.4 ± 0.7 17.1 ± 0.7 PC 15.2 ± 1.4 1.4 ± 0.0 x 37.2 ± 2.0 x 12.6 ± 0.9 x TV 17.7 ± 1.3 1.6 ± 0.0 y 43.1 ± 0.5 14.4 ± 0.2 y,z CS 16.4 ± 0.5 2.4 ± 0.1 z 42.8 ± 2.8 13.9 ± 1.0 x,z Samples from three or four independent progeny of the same genotype were analyzed a Values are the same as those shown in Table 1 b Different letters indicate significant differences under pretreatment with different strains within the same progeny genotype ( P < 0.05) Delignification under fungal treatment The KL levels in each wood sample after fungal pretreatment were more variable under PC and TV than under CS pretreatment, although cause of the variation was unclear. Regardless of the progeny genotype, the reduction in the wood fraction producing KL was likely larger under PC than that under TV and CS pretreatment (Tables 3 and S1), but the difference was not significant. Among the fungal pretreatments, only CS-treated homozygotes (27.6%) and heterozygotes (23.3%) differed in their KL degradation efficiencies ( P = 0.003), suggesting that the altered lignin structure of the homozygote wood was more readily degraded by CS. Although KL levels were generally lower in fungus-pretreated than in non-pretreated wood powders, the levels of most ASLs detected in both genotypes, except for the PC-treated homozygote wood, increased after fungal pretreatment. The increase in ASL levels was particularly pronounced in CS-treated homozygous wood ( P < 0.001; Table 3 ). A substantial decrease in KL and increase in ASL levels were also observed in poplar wood disks [ 19 ] and aspen wood chips under CS degradation [ 48 ]. In addition, the increased rate of ASL was higher in homozygous wood after pretreatment with PC and TV than in heterozygous wood, although the difference was only significant following PC treatment ( P < 0.001, Table S1). In non-pretreated wood, ASLs were partially composed of low-molecular weight lignin degradation products released by sulfuric acid hydrolysis [ 49 ]. The mechanism by which the acid solubility of lignin was increased under fungal degradation is unclear but was likely due to changes in the cleavage of lignin–polysaccharide and internal lignin linkages, a reduction in lignin molecular weight, and the introduction of more hydrophilic groups into lignin. Although ASL levels were significantly lower than those of KL, the remarkable increase in ASL released from CS- and PC-pretreated homozygous wood may reflect a partially unique manner of lignin decomposition. Selective degradation of lignin The selectivity index (SI, KL level reduction per unit of cellulose; Table S1) is an indicator of selective fungal degradation of lignin in lignocellulosic materials [ 50 ]. In heterozygous wood, the SI value was higher under CS (3.7) than under the other treatments ( P < 0.01 for PC vs. CS and TV vs. CS), with no significant difference detected between PC (1.2) and TV (1.7). Similarly, the SI values for homozygous wood were in the decreasing order of PC (1.3), TV (1.2), and CS pretreatments (6.5), though no significant difference was detected among the genotypes because of the large variation in the SI value for CS pretreatment. These results suggest that, as previously reported [ 50 ], lignin-selective degradation by CS occurred in both heterozygous and homozygous wood. Characterization of lignin degradation by thioacidolysis We further characterized lignin degradation under fungal pretreatment based on the release of monomeric products from fungus-degraded wood powders by thioacidolysis (Table 4 ). The levels of thioacidolytic monomers decreased in both wood genotypes after fungal pretreatment as the level of lignin decreased. Pretreatment of the heterozygote with PC and TV slightly reduced G and S monomer levels, but no significant differences were detected compared to those in non-pretreated samples. In contrast, the levels of each monomer and the total decreased significantly after CS pretreatment in the heterozygote. The pretreated homozygous wood showed a similar trend in monomer reduction as the heterozygous wood; however, there was no significant difference between the different fungal strains in terms of S monomer reduction. The lowest monomer level was observed in CS-pretreated homozygous wood, suggesting a more condensed structure with few β-O-4 linkages. Table 4 Thioacidolysis monomer yields from fungus-pretreated heterozygote ( CAD/cad ) and homozygote ( cad/cad ) wood Progeny genotype Fungal strain Thioacidolytic monomer yields (μmol g −1 of lignin) S/G ratio G monomer S monomer Total CAD/cad Without fungi a 739 ± 59 x 1023 ± 71 x 1762 ± 130 x 1.39 ± 0.03 x PC 666 ± 40 x 842 ± 41 x 1508 ± 81 x 1.26 ± 0.01 y TV 657 ± 24 x 896 ± 44 x 1552 ± 68 x 1.36 ± 0.02 x CS 440 ± 6 y 612 ± 8 y 1052 ± 14 y 1.39 ± 0.00 x cad/cad Without fungi a 787 ± 11 x 718 ± 52 x 1505 ± 56 x 0.91 ± 0.06 x PC 629 ± 53 x 556 ± 76 x,y 1186 ± 129 x,y 0.88 ± 0.05 x TV 658 ± 40 x 606 ± 65 x,y 1264 ± 105 x 0.92 ± 0.05 x CS 400 ± 39 y 387 ± 44 y,z 787 ± 82 y 0.97 ± 0.04 x a Values are the same as those shown in Table 1 b Different letters indicate significant differences under pretreatment with different strains within the same progeny genotype ( P < 0.05) A slight difference in the S/G ratio was observed among the genotypes under different fungal pretreatments, and most values were comparable before and after fungal degradation. Although preferential degradation of S units was reported in the lignin of wheat straw and oak wood under CS decomposition [ 16 , 51 ] it was not observed in the present study. The thioacidolytic dimer products after desulfurization were analyzed using GC/MS [ 52 , 53 ]. We identified 18 different dimers based on their mass spectra, of which 5, 4, 3, 3, and 3 compounds had β-5, β-1, β-β, 4-O-5, and 5–5 substructures, respectively. Seven compounds consisted of two G units (GG), four had two S units (SS), and seven had both S and G units (SG). The total level of dimers detected in non-pretreated homozygous wood was 11% lower than that in heterozygous wood (Fig. 2 ). The low yield of total thioacidolytic monomers and dimers in the homozygote suggest that the homozygote lignin has a more “condensed (rich in carbon–carbon bond)” structure than the heterozygote lignin, as was also reported in CAD-deficient transgenic poplar [ 54 ]. The relative proportion of β-β-type dimers, in which the structure with two S units predominates, was markedly lower in the non-pretreated homozygote than in the heterozygote. The relative compositions of the GG, SS, and SG dimers were 55.8%, 14.6%, and 29.6% in homozygotes, whereas they were 45.6%, 21.8%, and 32.6% in heterozygotes, respectively. The lower proportion of dimers containing S units in the homozygote was consistent with the lower S/G ratio calculated from the lignin monomers. Fig. 2 Thioacidolysis dimer yields of heterozygote (left) and homozygote (right) woods before (WF, without fungi) and after fungal pretreatment. PC, P. chrysosporium . TV, T. versicolor . CS, C. subvermispora After fungal treatment, the total amount of dimers was reduced to 5–43% in heterozygotes and 23–54% in homozygotes, suggesting that homozygote lignin is more easily degraded than is heterozygote lignin. The most abundant dimers detected after fungal treatment had the G(5–5)G substructure in both genotypes, regardless of the fungal species, indicating that the lignin substructure with the 5–5 linkage is difficult for fungi to degrade. The largest decrease in dimer levels after fungal degradation was observed in homozygous wood under CS pretreatment (28% decrease compared with that in heterozygous wood). Effect of β-O-4 interunit coupling with cinnamaldehyde units on fungal degradation of lignin As previously reported [ 40 ], CAD deficiency in Sekizaisou generated abnormal β-O-4 subunits with cinnamaldehyde residue in its lignin. These structures may contribute to delignification in CAD-deficient woods under alkaline conditions [ 31 , 38 , 44 ]. To investigate their contribution to the fungal degradation of homozygous wood, the amounts of indene derivatives released by thioacidolysis of the degraded wood were compared based on their peak area of the mass chromatograms (Figure S1) [ 26 , 29 ]. The levels of the three indenes increased slightly after fungal degradation but were not correlated with the reduction in total weight or lignin content. These results indicate that the abnormal β-O-4 substructure does not affect the degradability of homozygous wood and its lignin. Lignin molecular distribution before and after fungal pretreatment Size exclusion chromatography was used to compare the molar mass distribution of cellulolytic enzyme lignins (CELs) prepared from heterozygous and homozygous wood before and after fungal pretreatment. CEL prepared from non-pretreated homozygous wood showed smaller weight-average molar mass (Mw) and number-average molar mass (Mn) values than that from non-pretreated heterozygous wood (Fig. 3 and Table S2). These results suggest that CAD deficiency, the associated changes in the composition of lignin monomers, and resultant substructures inhibit elongation of the lignin chain in the cell wall, thereby at least partially contributing to the accelerated delignification of homozygous wood during fungal pretreatment. A decrease in the Mw of lignin was also reported for CAD-deficient transgenic poplar [ 54 ], Arabidopsis , and loblolly pine mutants [ 55 , 56 ], in which delignification of lignocellulosic materials was improved under deep eutectic solvent extraction or alkaline pulping [ 31 , 36 , 55 ]. Fig. 3 Molar mass distribution of cellulolytic enzyme lignins (CELs) from non-pretreated homozygotes ( cad/cad ) and heterozygotes ( CAD/cad ) The Mw was further decreased after fungal pretreatment, with the reduction rate roughly correlated with the delignification efficiency of each fungus, regardless of the genotype (Figure S2 and Table S2). The largest decrease in the CEL Mw was observed in both genotypes under PC, which is consistent with the largest reduction in KL levels under the same pretreatment. The reduction in CEL Mw after PC pretreatment was greater in homozygotes (35%) than in heterozygotes (27%) (Table S2). This may be partially due to the slightly higher delignification efficiency of the former (Table S2). Mw reduction was also detected in CS-degraded homozygous and heterozygous CELs; however, unlike under PC, the reduction rate was higher in heterozygotes (37%) than in homozygotes (24%). In contrast to PC and CS, there was no apparent difference in the Mw reduction rate between the two genotypes after TV pretreatment. The smaller CEL Mw in the non-pretreated homozygous wood should benefit the fungal degradation of lignin compared to that in the heterozygous wood. Combined effect of lignin alteration and fungal pretreatment on saccharification efficiency Saccharification efficiency was calculated as the amount of glucose released by enzymatic hydrolysis of wood powder with and without fungal pretreatment. The saccharification efficiency of the CS-treated heterozygous wood, based on the glucose units in the pretreated wood estimated as cellulose (78%), was higher than that under PC (60%, P = 0.07) and TV pretreatment (53%, P = 0.017) and without fungal pretreatment (62%, P = 0.11). In homozygous wood, the saccharification efficiency was greatest after CS pretreatment ( P < 0.001, compared to TV and non-fungal pretreatment), whereas the difference under PC pretreatment was less pronounced ( P = 0.055). Collectively, CS pretreatment contributed to improved glucose release from both mulberry genotypes, despite the reduction in cellulose during fungal pretreatment. In contrast to the differences in fungal pretreatments, glucose yield did not significantly differ between the different genotypes, except for in PC-pretreated woods (Fig. 4 ). However, after pretreatment with TV and PC, the yield tended to be higher in homozygotes than in heterozygotes. As backcross progeny plants ( CAD/cad and cad/cad ) were generated by artificial crossing of Sekizaisou ( cad/cad ) and its F1 progeny ( CAD/cad ), their genotypes may vary greatly at some gene loci, even if individual plants had the same genotype at the CAD locus. Furthermore, as pretreatment was performed under semi-solid culture with wood powder, the fungal mycelia likely grew nonuniformly. Therefore, the genetic diversity of the wood samples and non-uniform fungal culture conditions in the present study may have limited the detection of differences between the mulberry genotypes under TV and CS pretreatments. Fig. 4 Glucose yields from heterozygotes ( CAD/cad ) and homozygotes ( cad/cad ) with and without (WF) fungal pretreatment, PC, P. chrysosporium . TV, T. versicolor . CS, C. subvermispora . Data were analyzed using Student’s t -test ( n = 3)"
} | 5,972 |
38433268 | PMC10910722 | pmc | 2,365 | {
"abstract": "Background The severity and frequency of drought are expected to increase substantially in the coming century and dramatically reduce crop yields. Manipulation of rhizosphere microbiomes is an emerging strategy for mitigating drought stress in agroecosystems. However, little is known about the mechanisms underlying how drought-resistant plant recruitment of specific rhizosphere fungi enhances drought adaptation of drought-sensitive wheats. Here, we investigated microbial community assembly features and functional profiles of rhizosphere microbiomes related to drought-resistant and drought-sensitive wheats by amplicon and shotgun metagenome sequencing techniques. We then established evident linkages between root morphology traits and putative keystone taxa based on microbial inoculation experiments. Furthermore, root RNA sequencing and RT-qPCR were employed to explore the mechanisms how rhizosphere microbes modify plant response traits to drought stresses. Results Our results indicated that host plant signature, plant niche compartment, and planting site jointly contribute to the variation of soil microbiome assembly and functional adaptation, with a relatively greater effect of host plant signature observed for the rhizosphere fungi community. Importantly, drought-resistant wheat (Yunhan 618) possessed more diverse bacterial and fungal taxa than that of the drought-sensitive wheat (Chinese Spring), particularly for specific fungal species. In terms of microbial interkingdom association networks, the drought-resistant variety possessed more complex microbial networks. Metagenomics analyses further suggested that the enriched rhizosphere microbiomes belonging to the drought-resistant cultivar had a higher investment in energy metabolism, particularly in carbon cycling, that shaped their distinctive drought tolerance via the mediation of drought-induced feedback functional pathways. Furthermore, we observed that host plant signature drives the differentiation in the ecological role of the cultivable fungal species Mortierella alpine ( M . alpina ) and Epicoccum nigrum ( E. nigrum ). The successful colonization of M . alpina on the root surface enhanced the resistance of wheats in response to drought stresses via activation of drought-responsive genes (e.g., CIPK9 and PP2C30 ). Notably, we found that lateral roots and root hairs were significantly suppressed by co-colonization of a drought-enriched fungus ( M . alpina ) and a drought-depleted fungus ( E. nigrum ). Conclusions Collectively, our findings revealed host genotypes profoundly influence rhizosphere microbiome assembly and functional adaptation, as well as it provides evidence that drought-resistant plant recruitment of specific rhizosphere fungi enhances drought tolerance of drought-sensitive wheats. These findings significantly underpin our understanding of the complex feedbacks between plants and microbes during drought, and lay a foundation for steering “beneficial keystone biome” to develop more resilient and productive crops under climate change. \n Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40168-024-01770-8.",
"conclusion": "Conclusions Manipulation of rhizosphere microbiomes represents an effective and promising strategy for addressing the threat that changes in global climate such as drought pose to sustainable agriculture. We propose a conceptual paradigm that demonstrates that the genotype of the drought-resistant wheat cultivar restructures the rhizosphere microbiome and shapes functional adaptations and highlights the importance of beneficial fungal strains in enhancing the drought resistance of wheat (Fig. 7 ). More importantly, the rhizosphere microbiomes that were recruited by the drought-resistant genotype could serve as a “microorganism goldmine” for the design of reasonable synthetic communities for increasing agricultural productivity. Taken together, our results represent a significant advance in determining the molecular mechanisms of drought-enriched fungal strains in enhancing the drought resistance of plants and provide critical new knowledge on key ecological interactions between drought-resistant wheats and their rhizosphere microbiomes. These suggest that a functionally reliable “beneficial biome” will offer opportunities for sustainable agriculture and provide a new direction for breeding drought-resistant wheat and enhancing agricultural sustainability under global climate change scenarios. Fig. 7 Conceptual paradigm depicting the synergistic mechanisms between host plants and rhizosphere microorganisms that improve drought resistance in wheat. The drought-resistant wheat cultivar determines the direct response of plants to recruit genotype-specific microbial communities, such as M. alpina . M. alpina can feedback to plants and improve drought resistance by activating the CIPK-PP2C network to induce drought-responsive genes. However, M. alpina and E. nigrum together have a negative effect on the lateral roots and root hairs of wheat and lead to wheat becoming significantly more sensitive to drought stress",
"discussion": "Discussion In this study, we sought to verify whether the harnessed rhizosphere microbiome from the drought-resistant wheat cultivar have the ability to enhance the drought tolerance of drought-sensitive wheat under drought stresses using amplicons, metagenomics, and RNA-sequencing approaches. By profiling both bacterial and fungal communities in the drought-resistant and drought-sensitive cultivars at two different planting sites, we reveal that wheat genotypes exert significantly influences on the rhizosphere microbiomes, and drought-resistant wheat cultivar has more diverse bacterial and fungi taxa than drought-sensitive cultivar. Microbial interkingdom association network analysis indicates that the drought-resistant wheats possess higher network complexity of rhizosphere microbiomes, regardless of planting sites. Metagenomics sequencing data from these two wheat cultivars further suggest that the enriched rhizosphere microbiome affiliate to drought-resistant cultivar has greater investment in “Energy metabolism” and “Carbohydrate metabolism”. Moreover, our work provide evidence that colonization of the single strain M. alpine significantly enhances the plant growth under drought stresses. Through RNA-Seq, we further confirm that the strains M. alpine heighten plant adaptability of drought-sensitive cultivar to drought stresses through synergistic regulation of CIPK and PP2C genes. Below, we discuss how these findings have facilitated our understating of the single fungi strains confer drought tolerance to plant hosts. Host drought-resistant differentiation shapes distinct rhizosphere microbiome assembly and functional adaptation Host plants and their associated microbes interacted by multiple distinct mechanisms, including plant-to-microbiome, microbiome-to-microbiome, and microbiome-to-host, which sustains agricultural ecosystem services in response to ongoing environmental changes [ 28 , 52 , 64 ]. Our results demonstrated that the drought-resistant wheat cultivar possessed more diverse bacterial and fungal communities, particularly for specific fungal species. The consequences of ecological memory of recurrent drought to the soil microbial community [ 7 ] and alternations in the root exudates of plants under the non- continuous drought stresses [ 58 , 74 ] could at least partially explain this altered assembly of specific microbial taxa. Furthermore, multiple microbial attributes of the rhizosphere microbiome were jointly influenced by the plant genotype and planting site. These findings are consistent with previous studies [ 32 , 66 , 71 ], suggesting that host plant signature drives the differentiation in the ecological role of rhizosphere microbiomes in agricultural ecosystems. Increasing crop yields and adaptation of plants to drought stress are among the most important goals during the breeding process of drought-resistant wheats. Rhizosphere and root-associated microbiota play an essential role in plant growth and resilience [ 59 ]. The impact of plant breeding on the rhizosphere microbiome assemblage and functional profiles usually represents directional selection by root exudates that are released by plants with different genotypes [ 44 , 58 ]. Moreover, wheat varieties with different drought tolerances could recruit specific microbiota that formed complex microbial interkingdom association networks to reinforce the resistance of plants under harsh environmental conditions. We further found that the drought-resistant wheat cultivar possessed more complex bacterial–fungal interkingdom correlations with a higher proportion of negative edges in the microbial networks. Pioneer studies indicated that complex networks with greater connectivity are more resilient to environmental perturbations than networks with lower connectivity [ 56 , 76 ]. It has also been shown that the complexity of microbial networks contributed greatly to ecosystem multifunctionality [ 65 ]. In this sense, the higher complexity and proportion of negative edges between hub bacteria and fungi may indicate that rhizosphere microbiota derived from the drought-resistant wheat cultivar are more tolerant of drought stress, as different bacteria and fungi can complement each other. Interestingly, we found that plant pathogen fungi exhibited significantly higher abundance in the warmer climate area (e.g., SQ site) than semi-arid area (e.g., YL site), suggesting that warming climates may facilitate the outbreak of plant pathogenic fungi and further affects the plant growth via pathogenic-to-beneficial microbe interactions [ 17 , 54 ]. Apart from changing rhizosphere microbiome assemblages and interkingdom association networks, plant–microbe co-adaptation plays key roles in shaping functional adaptations of rhizosphere microbiomes in response to stresses [ 15 ]. Metagenomic analyses in our study revealed that the functional diversity and enriched functional profiles in rhizosphere microbiomes from the drought-resistant cultivar exhibited contrasting patterns in comparison with the drought-sensitive cultivar. Rhizosphere microbiomes belonging to the drought-resistant cultivar had greater relative abundances of genes involved in “Carbohydrate metabolism” and “Energy metabolism”, whereas functional genes encoding factors associated with cell integrity and involved in “Membrane transporter” and “Cell growth and death” were significantly enriched in the drought-sensitive cultivar. This is likely due to the rhizosphere microbiome from drought-resistant cultivar and drought-sensitive cultivar adopt distinct life history strategies to confront drought stress in the agroecosystem [ 45 ]. In the framework of trait-based life history theories in microbial ecology, drought-resistant microbes may adopt growth yield and stress tolerance strategies to cope with harsh conditions. On the other hand, drought-sensitive microbes tend to adopt resource acquisition strategy to maintain the microbial cellular osmolarity and integrity under adverse conditions. Tradeoffs in microbial life history can have consequences for turnover and transformation of soil availability nutrients [ 11 ], thereby directly mediating the gene expression of plants via microbiome-to-host interactions [ 20 ]. More importantly, rhizosphere microbiomes from drought-resistant cultivar with enriched “Energy metabolism”, in particular in “Carbon fixation” and “Nitrogen metabolism”, may reinforce the ability and flexibility of crops to deal with diverse environmental stresses by enhancing the acquisition of carbon and nitrogen for plants [ 16 ]. These results indicate that plant hosts exert a strong selective effect on the utilization of specific metabolic functions of rhizosphere microbiota. Complementary to the findings of pioneer studies that selection by the host plant via genetic features plays a key role in shaping rhizosphere microbiome communities [ 60 , 61 ], this work provides novel evidence that host plant signature profoundly influence not only ecological patterns of rhizosphere microbiomes but also their functional adaptations. Collectively, these findings offer a gateway to the manipulation of specific functions of bacteria and fungi to enhance the drought resistance of wheats during the plant breeding process. SynComs strongly inhibit the development of lateral roots and root hairs in wheat Plant root traits are central drivers of many ecosystem processes, plant roots can quickly respond to harsh environments in natural and agricultural ecosystems [ 69 ]. More importantly, root morphological features are vital indicators of plant hosts in responses to drought stress [ 43 ]. A pioneer study also highlighted the critical role of dominant microbiota in shaping the phenotypic traits of plant roots when confronting drought stress [ 42 ]. Importantly, long and thick L-type of lateral roots is better suited for water uptake during drought stress [ 19 ]. Plants exhibit substantial variation in root morphology in response to drought stresses and resource availability, but our understanding of how specific rhizosphere microbiomes covary with root morphology remains inconclusive. Here, we found that lateral roots and root hairs of the drought-susceptible cultivar were strongly inhibited by co-colonization by M. alpina and E. nigrum under drought stress. The underlying mechanisms currently include (i) competition for niche space, which directly determined whether the beneficial rhizosphere microbiome successful colonization on the root surface [ 9 , 14 , 25 ],(ii) inhibition from secondary metabolites, which determined metabolite exchange networks between rhizosphere microbiome and plant roots via potential allelopathy effect [ 41 , 58 ],(iii) regulation of gene expression in plants, which would affected plant hormone signal transduction and immune system activities (tolerance or avoidance) via the process of microbial inheritance in plants [ 1 , 38 ]. Furthermore, pioneer studies indicated that root hairs are able to influence the rhizosphere microbiome assemblage, and in turn rhizosphere microbes are able to interact with plant-modifying root hairs [ 34 , 55 ]. Root hairs have physical properties that play vital roles in the two-way interaction by specifically changing the plant–microbe interaction interface [ 44 ]. MYB36 is known to be a novel factor involved in the later stages of lateral root development and is expressed in cells surrounding lateral root primordia to regulate the proliferation–differentiation transition in the root meristem [ 18 ]. WOX11 is expressed in the founder cells of adventitious roots and activates LBD16, which has vital functions in the formation of lateral roots and adventitious roots [ 77 ]. SynCom-2 significantly inhibited the development of lateral roots and root hairs by significantly decreasing the transcription levels of MYB36 and WOX11 , which suggested that water uptake and nutrient acquisition for the plant had been impeded, which led to plant death. Although the fungal strain M. alpina positively contributed to the resistance of wheat to drought, it is worth noting that the fungal strain E. nigrum may have extraordinarily eliminated the positive role of M. alpina when these two fungi co-colonized the plant rhizosphere. It was suggested that a single fungal strain enriched from the drought-resistant variety maintained root growth. In contrast, the synthetic microbial community contained fungal strains from the drought-sensitive cultivar, which could have hampered plant growth by inhibiting lateral root development. Our study provides empirical evidence that fungal communities utilized by drought-resistant wheat may play an increasing ecological role in sustaining plant growth and root development. These findings provide new insights into the complex interactions between root morphological traits and rhizosphere microbiomes and pave the way to the deployment of “beneficial microorganisms” for sustainable agriculture. However, the dynamic interactions among root exudates, root morphology, and rhizosphere microbiomes are not fully understood and need further exploration. M. alpina enhances the drought tolerance of wheat by activating the CIPK-PP2C network There is mounting evidence that drought leads to dramatic shifts in rhizosphere microbiomes and in which rhizosphere microbiomes are selected by plant hosts [ 16 ]. However, the feedback effects of drought-resistant microbiomes on plant growth and fitness are still limited. Drought-resistant wheat recruits the genotype-specific fungus Mortierella , which is directly responsible for increasing the efficiency of the uptake of nutrients, including P and Fe, and the synthesis of triacylglycerols, phytohormones (e.g., indole-3-acetic acid), and 1-aminocyclopropane-1-carboxylate deaminase [ 49 ]. Apart from direct mechanism, Mortierella also strengthen indirect mechanism activities, including hydrogen cyanide, chitinase, protease, and antibiotic for plant growth promotion [ 27 , 53 ], which results in a positive effect on the protection of crops against harsh conditions [ 8 , 36 ]. In this study, M. alpina acted as a beneficial fungus with the ability to enhance the drought resistance of the drought-susceptible cultivar. When the host suffered from drought stress, M. alpina and the host generated synergistic responses by activating the CIPK-PP2C network. In accordance with a pioneer study, CIPK-PP2C complexes could act as molecular on–off switches to regulate the phosphorylation state of plant ion transporters involved in stress responses [ 35 ]. In particular, transcriptional regulation of CIPK9 and PP2C30 resulted in 100 times higher basal expression levels in wheat inoculated with M. alpina in comparison with the control group because of the strong synergistic effects of MAPKs. A previous study indicated that multiple MAPKs can be quickly activated in plants to cope with drought stress [ 78 ]. Our qRT-PCR data further confirmed that M. alpina regulated the gene expression of MAPKKK17 and MAPK3 in the drought-susceptible variety in response to abiotic stress. This result indicated that multiple stress signaling pathways, which have crosstalk features, were directly or indirectly activated to form complex signal transduction networks in host plants by colonization by M. alpine (Fig. S 18 ). On the basis of these findings, we propose a scheme where beneficial fungi play a key role in strengthening the drought resistance of plants by activating plant integrated responses to abiotic stimuli, although the molecular triggers of these synergistic responses remain to be characterized."
} | 4,714 |
27958317 | PMC5153641 | pmc | 2,366 | {
"abstract": "The arch-shaped single electrode based triboelectric nanogenerator (TENG) is fabricated using thin film of reduced graphene oxide nanoribbons (rGONRs) with polyvinylidene fluoride (PVDF) polymer used as binder to effectively convert mechanical energy into electrical energy. The incorporation of rGONRs in PVDF polymer enhances average surface roughness of rGONRs/PVDF thin film. With the combination of the enhancement of average roughness and production of functional groups, which indicate improve charge storage capacity of prepared film. Furthermore, the redox peaks obtained through cyclic voltammetry were identified more in rGONRs/PVDF composite in comparison to pristine rGONRs to confirm charge transfer capability of film. Herein, the output performance was discussed experimentally as well as theoretically, maximum voltage was obtained to be 0.35 V. The newly designed TENG to harvest mechanical energy and opens up many new avenues of research in the energy harvesting applications.",
"conclusion": "Conclusions In summary, we have synthesized the rGONRs through unzipping of CNTs. The charge storage and transfer capability of rGONRs/PVDF have been revealed by cyclic voltammetry. Therefore, we have designed single electrode TENG based on charge transfer between Al and rGONRs/PVDF thin film with finite size supported by kapton tape and modulating distance. In this approach, obtained AC voltage through finger touching convinced mechanical energy by TENG. The maximum output voltage would be achieved by TENG is 0.35 V. TENG could be quite stable upto 500 cycles and the energy generated by the TENG would be stored or can be directly used to operate portable electronics devices. This proposed work demonstrates the practicability of nanogenerator in which the mechanical energy is utilized and convert it into electrical energy.",
"discussion": "Results and Discussion rGONRs were synthesized as discussed in the methods and schematic diagram of the synthesis process was illustrated in Fig. 1 . Characterization of rGONRs X-ray diffraction (XRD) analysis was performed for the thin film prepared by rGONRs/PVDF composite as illustrated in Fig. 2 . Figure 2 reveals that the characteristic peaks of PVDF polymer are present around 17.2°, 18.3°, 19.6°, 38.5° and 44.69° which are in good agreement with the reported literature [JCPDS No. 42-1650]. The peak around 26.5° confirms the presence of highly reduced graphene oxide nanoribbons in composites. In addition, the existence of all individual peaks of both elements indicates that they exists in their original phases and there is no direct interaction which changes the phase of overall product. Therefore, we have considered the combination of these elements as a rGONRs/PVDF composite material. Raman spectroscopy of rGONRs/PVDF composite has been shown in Fig. 3 . The presence of D-band at around 1360 cm −1 proves the presence of certain defects in the structures of rGONRs. These defects must be due to the presence of PVDF or some vacancies in the hexagonal structure created during opening of CNTs. Further, the observed value of I 2D /I G from the Raman spectrum indicates the presence of 2–3 layers stacked together 42 . Atomic Force Microscopy (AFM) images of rGONRs/PVDF thin film was shown in Fig. 4 . In this analysis, we can observe the orientation and 3D topographical information about the sample surface. Figure S1 in the Supplementary Information shows the surface profilometry of the pristine rGONRs thin film as well as the rGONRs/PVDF thin film. From Figure S1 , it has been clearly observed that, average surface roughness of rGONRs/PVDF thin film higher than rGONRs film. The average surface roughness play vital role to trap charge and enhance charge storage capacity of materials. The degree of oxidation of graphene nanoribbons from MWCNTs were further illustrated by Fourier Transformation Infrared (FTIR) Spectroscopy. FTIR spectroscopy ( Figure S2 , Supplementary Information ) shows that the −OH stretching was present at 3441.72 and 2923.43 cm −1 in the high frequency range, −CH stretching and carboxylic group (−COOH) at 2358.12 cm −1 and 1634.23 cm −1 , respectively. Moreover, −CO stretching would be verified at 1384.05, 1270.20 and 1091.81 cm −1 . To extend our investigation, we had also examined the contact resistance by using two - probe resistivity measurement method. Figure S3 shows the current voltage (I-V) curve in which the contact resistance for pristine rGONRs have been found 6.11 Ω/μm 2 . While, composite was prepared using PVDF as binder, contact resistance increases upto 9.32 Ω/μm 2 because of the presence of functional groups in rGONRs/PVDF. The Cyclic Voltammetry (CV) analysis was performed to examine electrochemical performance of the electrodes. CV curve ( Figure S4 in Supplementary information ) of prepared sample measured at scan rate of 100 mV/s in 1M DMF with in potential range from −1.5 to +1.5 V. The redox peaks were identified more in rGONRs/PVDF composite in comparison to the pristine rGONRs, which confirm better charge transport of rGONRs/PVDF composite. Moreover, area under the CV curve of rGONRs/PVDF is larger than the pristine rGONRs, so that the charge storage capability of this material is good in comparison to pristine rGONRs, this results are in good agreement with the literature 43 . Field Emission-Scanning Electron Microscope (FE-SEM) and Transmission Electron Microscope (TEM) images of rGONRs are shown in Fig. 5e and f. These figures demonstrated that CNTs were unzipped completely in the form of oxidized graphene nanoribbons, having ~ 45 nm width. The elemental analysis is obtained using energy dispersive X-ray analysis (EDAX) of the prepared rGONRs, which confirms the composition by weight% of carbon and oxygen are found 66.20% and 33.80%, respectively ( Figure S5 ). Figure 5b shows the optical images of fabricated rGONRs/PVDF based TENG, where one electrode is made ground and from the other we get voltage. These optical images clearly show the arch-shaped structure of TENG. The surface of thin film with uniformly distributed nanoribbons which are clearly visible in the FE-SEM image of thin film as shown in Fig. 5d . The obtained rough surface of the thin film would enhance the surface area of the device and due to this the device would give effective output 44 . The rGONRs should be negatively charged because graphene oxide is negative charged as reported 45 . The charge on the rGONRs/PVDF film is fixed and confined by kapton tape. The total negative charge (Qr GONRs ) of rGONRs/PVDF film could induce positive charge in the Al foil. The fabricated device as flat panel capacitor, the capacitance (C) of capacitor could be written as C = εS/d, where ε, S, d and represents dielectric constant of air, surface area of rGONRs/PVDF film and separation distance between rGONRs/PVDF film and Al foil. The output current and voltage could be based on relation as: The triboelectric charge density could be increased by having the oxidized graphene nanoribbons on the surface, due to which the effective surface area has been enhanced. Figure 5d illustrates that the surface of the rGONRs/PVDF thin film was having the uniformly distributed nanoribbons structures. The working of the rGONRs/PVDF based TENG was performed by applying a mechanical force on the nanogenerator, therefore rGONRs/PVDF thin film and the Al foil can come in contact and released from each other. One of the electrode i.e. Al foil was connected to load resistance of 100 MΩ and was made ground. Once released, the rGONRs/PVDF thin film and Al would get apart from each other because of the elasticity stored in the kapton film, then come back in its original shape. The rGONRs/PVDF thin film surface was having negative surface electric potential can be upto 0.35 V by persistent contact-separation between the rGONRs/PVDF thin film and Al foil as illustrated in Fig. 6a . For measuring the voltage, TENG was connected to low noise voltage pre-amplifier. The output of the rGONRs/PVDF based single electrode TENG by applying the mechanical energy induced by finger touching was having highest voltage as 0.35 V. Moreover, we have removed rGONRs/PVDF thin film from Kapton. Now, Kapton is a contact surface in the fabricated TENG, corresponding electrical output voltage was found to be 0.16 V. When the compressive force was applied, the TENG would give positive voltage. While, compressive force was released it gave negative voltage, which is clearly depicted in Fig. 7 . The mechanism of the rGONRs/PVDF based TENG was schematically depicted as shown in Fig. 6 . At the original state, no contact between Al foil and rGONRs/PVDF thin film, there was no electric output and hence no charge transfer across them. Once the Al foil and rGONRs/PVDF thin film surface came in contact with each other, electrons were injected from Al foil into rGONRs/PVDF thin film due to the presence of negative charge on itself. According to the triboelectric series, Al is having more propensity to loss electrons and hence according to electrostatic phenomena more triboelectrically positive than rGONRs/PVDF thin film 46 . If the size of the rGONRs/PVDF thin film and Al foil was definite then the positive triboelectric charges, which were generated on the Al foil may drift to the ground through external load, giving electric field leakage. By releasing the mechanical force, TENG would instantly come back to its initial shape because of the flexibility present in the kapton tape. As the gap formed between the rGONRs/PVDF thin film and Al with definite size would be increasing, the triboelectric positive charge formed on the Al film was decreasing, which offers the flow of electrons from Al film to the ground. However, when the mechanical energy is applied again on the TENG to come in contact with each other i.e. Al and rGONRs/PVDF thin film, then induced triboelectric negative charges on the rGONRs/PVDF thin film would be increasing to balance the triboelectric positive charges on the Al foil and hence electrons would flow from ground into the Al foil. The full contact-separation cycle of electricity generation would be shown in Fig. 6b . In order to further understand the working principle of the as prepared rGONRs/PVDF based single electrode TENG, numerical analysis was investigated through numerical simulation based on COMSOL Multiphysics. The model was constructed same as prepared experimentally, which based on Al foil and rGONRs/PVDF thin film with same size of 3 × 4.5 × 0.05 cm 3 . Figure 6d illustrates the calculations, which were performed and results in the electric potential distribution with the distance of 0.2, 3, 6, 9 and 12 cm, respectively between rGONRs/PVDF and Al foil. However, when the gap distance is increasing, the potential difference between the Al foil and rGONRs/PVDF thin film can increase upto 2000 V. As the gap distance between rGONRs/PVDF film and Al increases the amount of charges present on Al foil decreases, which indicates electrons were transferred from ground to Al foil as the gap distance is increasing. The principle behind the working of the as prepared TENG could be explained on the basis of transfer of charge among ground and Al foil, by varying the gap distance in the rGONRs/PVDF film and the Al foil due to the leakage field present at the edges of the thin films having definite size. The distance which was made between the rGONRs/PVDF thin film and Al foil must not be small while comparing it with the dimensions of the rGONRs/PVDF film or Al foil. If the gap distance between the rGONRs/PVDF thin film and Al foil is very less than the sizes of Al foil or the size of both the materials are approximately much large, then the charge transfer among the ground and Al might be very less. If this could be done, then the as prepared device might not work properly."
} | 2,972 |
23555721 | PMC3608662 | pmc | 2,367 | {
"abstract": "In obligate symbioses, the host’s survival relies on the successful acquisition and maintenance of symbionts. Symbionts can either be transferred from parent to offspring via direct inheritance (vertical transmission) or acquired anew each generation from the environment (horizontal transmission). With vertical symbiont transmission, progeny benefit by not having to search for their obligate symbionts, and, with symbiont inheritance, a mechanism exists for perpetuating advantageous symbionts. But, if the progeny encounter an environment that differs from that of their parent, they may be disadvantaged if the inherited symbionts prove suboptimal. Conversely, while in horizontal symbiont acquisition host survival hinges on an unpredictable symbiont source, an individual host may acquire genetically diverse symbionts well suited to any given environment. In horizontal acquisition, however, a potentially advantageous symbiont will not be transmitted to subsequent generations. Adaptation in obligate symbioses may require mechanisms for both novel symbiont acquisition and symbiont inheritance. Using denaturing-gradient gel electrophoresis and real-time PCR, we identified the dinoflagellate symbionts (genus Symbiodinium ) hosted by the Red Sea coral Stylophora pistillata throughout its ontogenesis and over depth. We present evidence that S. pistillata juvenile colonies may utilize both vertical and horizontal symbiont acquisition strategies. By releasing progeny with maternally derived symbionts, that are also capable of subsequent horizontal symbiont acquisition, coral colonies may acquire physiologically advantageous novel symbionts that are then perpetuated via vertical transmission to subsequent generations. With symbiont inheritance, natural selection can act upon the symbiotic variability, providing a mechanism for coral adaptation.",
"conclusion": "Conclusions The ecological and evolutionary implications of employing both modes of symbiont transmission are substantial. Horizontal acquisition of novel symbionts may be a means by which coral species can adapt to environmental changes [56] . Although sexual reproduction in Symbiodinium occurs relatively infrequently [39] , [57] , [58] , new host-symbiont combinations can emerge that may lead to novel, advantageous, and potentially specific symbioses [5] . Horizontal symbiont acquisition in adult corals, however, may either not occur or may be transient [19] , [20] , [41] , [46] , [59] . In coral species with horizontal symbiont acquisition, the juvenile stage appears to be more flexible in acquiring symbionts not present in the adult population [25] – [28] , [42] , [60] . Thus, horizontal symbiont acquisition may allow an individual juvenile to obtain, and subsequently maintain, novel symbionts. In turn, this may increase the juvenile’s fitness and survival as it grows into an adult coral, potentially enabling short-term acclimation on an individual level. Horizontal symbiont acquisition, however, does not provide a mechanism for the perpetuation of the novel symbionts in subsequent generations. Since symbionts are not transferred to the progeny, the advantageous symbionts will be lost each generation when the progeny must acquire symbionts anew. In contrast, if juvenile corals with vertically transmitted symbionts are capable of acquiring novel, advantageous Symbiodinium that are maintained into adulthood, subsequent vertical symbiont inheritance to their progeny would facilitate the maintenance of novel Symbiodinium over generations. Vertical symbiont inheritance, punctuated with horizontal symbiont acquisition, provides an evolutionary mechanism for adaptation to environmental changes through the acquisition and maintenance of advantageous symbionts. S. pistillata , for example, may benefit from both modes of symbiont transmission. On the one hand, the progeny are equipped with inherited Symbiodinium , eliminating the risk of not obtaining their obligate symbionts. On the other hand, juveniles may acquire novel symbionts, potentially increasing their chances of survival in a new environment. If the horizontally acquired novel symbionts improve host fitness, and become abundant in the resulting adult coral, they will be transferred to the brooded progeny, thereby perpetuating the novel symbiosis. Natural selection can then act on the genetic variation in the symbiosis, potentially establishing novel host-symbiont combinations that may be advantageous during changing environmental conditions.",
"introduction": "Introduction Obligate mutualistic symbioses are ubiquitous on earth and play pivotal roles in many ecosystems [1] , [2] . By definition, in obligate mutualisms, the host must possess symbionts in order to survive. If a host secures the perpetuation of obligate symbionts by directly transferring symbionts to the offspring (vertical transmission) [3] , progeny encountering an environment that differs from that of their parent may be disadvantaged by hosting a suboptimal symbiont. On the other hand, if a host releases aposymbiotic progeny that must acquire symbionts from the environment (horizontal transmission), progeny may acquire symbionts that are beneficial in a new environment [4] . As partner fidelity is not absolute in horizontal transmission, strong partner choice and symbiont sexual recombination can allow mutualisms to persist and evolve in systems with horizontal transmission [2] , [5] , [6] . In horizontal transmission however, since subsequent offspring do not inherit the symbionts, advantageous symbionts may or may not be acquired again, leaving each generation to potentially gamble with the continuation of a beneficial symbiosis. Adaptation, that is maintained via natural selection and subsequent evolution, may be required for species to survive in a changing environment, but neither vertical nor horizontal symbiont acquisition strategies alone provide a mechanism for the adaptation of an obligate symbiosis via symbiont partner change. Vertical transmission provides a means for the perpetuation of symbionts, but offers no mechanism for the acquisition of novel symbionts. In contrast, horizontal transmission provides a mechanism for acquiring novel symbionts but limited means for perpetuating the novel symbionts. Combining both acquisition strategies may provide a mechanism for adaptation, but evidence of the same host species utilizing both transmission modes is rare. Phylogenetic analyses of specific obligate prokaryote-insect [7] , [8] and prokaryote-marine invertebrate [9] , [10] symbioses reveal predominant vertical symbiont transmission punctuated by infrequent horizontal symbiont acquisition. We investigated whether both symbiont transmission modes could occur in a eukaryote-invertebrate obligate symbiosis. Reef building corals have an obligatory mutualism with dinoflagellate algae (genus Symbiodinium ), which provides a nutritional foundation for host metabolism [11] and calcification [12] making them fundamental components of coral reef ecosystems. In the Symbiodinium genus, species remain largely unresolved, limiting classification to membership within nine Symbiodinium clades (named A–I) and multiple types within each clade [13] , [14] . Symbiodinium can exhibit different physiologies in response to variations in light and temperature [15] – [17] . Consequently, the same host can display different physiologies based on the Symbiodinium found within it [18] . Regardless of whether Symbiodinium are acquired horizontally or vertically, adult corals exhibit extremely stable and specific mutualisms with Symbiodinium \n [19] – [24] . In contrast, larvae (planulae) and/or juveniles of coral species with horizontal Symbiodinium acquisition can acquire non-parental symbionts [25] – [27] . While the juvenile stage may be key in establishing novel symbioses, no studies to date have demonstrated coral juveniles capable of successfully maintaining a novel symbiont type into adulthood [28] , nor have they provided a mechanism for the perpetuation of a novel symbiont to subsequent generations. Contrary to horizontal symbiont acquisition, vertical transmission is often regarded as a “closed” system that precludes symbiont diversity in all life stages [3] , [20] , [29] – [32] but see [33] . Regarding vertical symbiont transmission as a closed system may explain why the symbiont identity in planulae and juveniles of coral species with vertical symbiont transmission has not been determined. Only recently has the symbiont identities in eggs of one coral species with vertical symbiont transmission been documented [34] . Planulae and/or juvenile corals with maternally derived symbionts may be capable of subsequent horizontal acquisition, which would facilitate diversity. Importantly, the inheritance of symbionts via vertical transmission would perpetuate the novel symbiosis if it increased the fitness of current and subsequent host generations. Therefore, deciphering the symbiont acquisition strategies utilized by corals throughout ontogenesis is key to understanding corals’ ability, or lack thereof, to alter their symbionts based on the environmental conditions of the habitat in which they grow. We determined whether horizontal symbiont acquisition could occur in a coral host with vertical symbiont transmission. The coral Stylophora pistillata (Pocilloporidae) broods and releases planulae with vertically transmitted Symbiodinium \n [35] . S. pistillata is widely distributed throughout the Indo-Pacific and Red Sea [36] , and is among the most abundant reef building corals in the Gulf of Eilat, Red Sea [35] . In the Gulf of Eilat, S. pistillata adult colonies host two distinct Symbiodinium clades. Shallow water colonies (<17 m) associate with clade A Symbiodinium \n [37] , [38] (type A1 [39] ), while congeners sampled in deep-water harbor symbionts of clade C [38] (e.g. type C72 at 20–30 m [5] ). We examined the Symbiodinium genetic identity in S. pistillata adults, their released planulae, and juvenile colonies, in shallow and deep depths, using techniques capable of detecting both abundant and possible low-level symbiont populations. We determined whether shallow and deep-water adult S. pistillata colonies hosted previously undetected low-levels of the second Symbiodinium clade found in S. pistillata adults. We also identified the Symbiodinium inherited by the planulae. Due to physiological differences between Symbiodinium, which symbiont(s) the progeny inherit may affect their survivorship in different habitats [34] , [38] . Additionally, we looked at the Symbiodinium naturally occurring in juveniles at both shallow and deep depths. If S. pistillata juvenile colonies can acquire symbionts from the environment their dual mode of symbiont acquisition may enable rapid adaptation.",
"discussion": "Discussion In obligate symbioses, in which the symbiont is vertically transmitted from parent to offspring, the offspring are guaranteed to receive the obligate symbiont. Vertical symbiont transmission in itself, however, is not necessarily straightforward. If different parents within a species host different symbionts, or if the same parent hosts multiple genetically distinct symbionts, progeny may inherit all or any one of the symbionts. In the case of Stylophora pistillata , we detected low-level symbionts in some adult colonies, indicating the potential for diverse Symbiodinium combinations in planulae. Which symbiont(s) the offspring inherit may vary between and even within a single parent. In addition, if progeny that inherit symbionts can later acquire symbionts from the environment, the symbiont variation within the coral species may further increase. If the symbionts differ physiologically, then which of the various symbionts the progeny acquire (either from a parental or environmental source) may affect their fitness. The Symbiodinium genetic identity in coral species with vertical symbiont transmission has not been investigated extensively and we are aware of only one study that determined the Symbiodinium inherited in eggs of the coral Montipora capitata in Hawai’i [34] . While only approximately 35% of coral species vertically transmit Symbiodinium , these coral species belong to several widely distributed, dominant coral genera, e.g. Porites , Montipora , and Pocillopora \n [48] . Consequently, investigating the symbiont identity throughout ontogenesis in the numerous ecologically dominant coral species with vertical symbiont transmission is ecologically relevant to understanding coral–algal symbioses and coral reefs in general. It is equally important to determine if species with vertical Symbiodinium transmission are capable of symbiont acquisition from the environment. Our results corroborate previous Symbiodinium identification in adult S. pistillata in the Gulf of Eilat, whereby adult S. pistillata colonies host two different Symbiodinium clades as a function of depth [37] , [38] . Several coral species, with horizontal or vertical symbiont transmission, host different Symbiodinium , either at the same depth or over a depth gradient [15] , [49] , [50] . By employing molecular techniques with finer resolution, it has been demonstrated that, in some of these species, colonies can host one Symbiodinium type at abundant levels, in addition to a second symbiont type present at low-levels [51] . Using real-time PCR, we uncovered that the S. pistillata adult colonies sampled in shallow water only hosted clade A Symbiodinium, while colonies in deep-water could harbor low-levels of clade A in addition to the abundant levels of clade C Symbiodinium . The symbiont depth zonation observed in S. pistillata in the Gulf of Eilat, at both abundant and low-levels, may be due to symbiont niche partitioning [4] . Clades A and C Symbiodinium hosted by adult S. pistillata colonies display differential responses to both elevated temperature and irradiance [38] , [52] . Adult colonies hosting clade C are less resilient to thermal stress than colonies hosting clade A [38] . Additionally, cell size and chlorophyll content differ between Symbiodinium types A1 and C72 hosted by shallow and deep-water S. pistillata colonies, respectively [38] . These physiological differences may explain why in the present study, type C72 Symbiodinium appears mostly limited to S. pistillata in deeper water while type A1 is capable of surviving in colonies at both depths. The presence of clade A Symbiodinium in deep-water adults varied from individual to individual. Since we did not repeatedly sample the same colonies, we cannot ascertain whether low-levels of clade A are hosted permanently or transiently in some deep-water adult colonies. Hosting transient symbionts at abundant and low-levels, even those not known to associate with a given host, has been documented in temporal studies [20] , [46] , [53] . Regardless of whether S. pistillata maternal colonies hosted only one Symbiodinium clade, or had low-levels of the other Symbiodinium clade, all the planulae, from both depths, were released with only the abundant maternal Symbiodinium type. On the other hand, Padilla-Gomiño et al. \n [34] reported several instances in which Montipora capitata eggs harbored both parental and non-parental symbiont types. Although the techniques employed by Padilla-Gomiño et al. \n [34] may overestimate symbiont diversity [54] , clade level differences were noted, with some of the non-parental symbiont clades and types known to associate with M. capitata in Hawai’i. As M. capitata vertically transmits Symbiodinium , the authors raised three hypotheses to explain these results, including environmental contamination, sampling bias, and the potentiality of horizontal symbiont acquisition in M. capitata eggs [34] . Our study provides evidence for the possibility of horizontal symbiont acquisition in a species with vertical symbiont transmission since S. pistillata planulae contained only one Symbiodinium type while some of the juvenile colonies, at both shallow and deep depths, harbored mixed symbioses. The incongruity between the symbionts in adult S. pistillata, the planulae and juveniles is likely explained by events occurring during the juvenile phase, and we present four plausible scenarios that may lead to the symbiont depth zonation observed in adult S. pistillata colonies. For example, if planulae released from shallow water colonies, with their clade A Symbiodinium complement, settle and metamorphose in deep-water, the resulting juveniles will initially contain clade A Symbiodinium ( Figure 2 ). Two possible, not mutually exclusive, scenarios may then follow, both leading to the observed adult S. pistillata symbiosis with abundant clade C Symbiodinium in deep-water. First, juveniles that continue to maintain only clade A Symbiodinium may not survive to adulthood ( Figure 2 ). Alternatively, if clade A containing juveniles in deep-water horizontally acquire clade C Symbiodinium , clade C may outcompete clade A and become the abundant symbiont present in adulthood ( Figure 2 ). We may have witnessed a snapshot of this process in the nine juveniles that contained both Symbiodinium clades simultaneously. Similar scenarios may apply to shallow water S. pistillata ( Figure 2 ). 10.1371/journal.pone.0059596.g002 Figure 2 Schematic depicting potential scenarios (1–4) of symbiont inheritance and acquisition throughout Stylophora pistillata ontogenesis. (1) Shallow water adults and planulae only host clade A Symbiodinium (•). Planulae settling in shallow water will become adults hosting clade A. (2) Some planulae from shallow water adults may settle in deep-water. These juveniles may horizontally acquire clade C Symbiodinium (+) while juveniles hosting only clade A may perish (×). (3) Deep-water adults abundantly host clade C Symbiodinium (potentially low-levels of clade A), and planulae only inherit clade C. Upon settlement in deep-water, juveniles will maintain clade C or horizontally acquire clade A. (4) Planulae from deep-water adults may settle in shallow water. Juveniles only hosting clade C may perish; horizontally acquiring clade A may facilitate survival to adulthood. Horizontal symbiont acquisition in the juvenile phase offers the most parsimonious explanation for the presence of multiple Symbiodinium clades within several juvenile colonies and the lack of multiple symbiont clades within all planulae analyzed. Alternatively, very low-levels of background symbionts could be present in some planulae, which could explain the presence of multiple symbiont clades in some juveniles. The real-time PCR assay reliably detected a Symbiodinium clade comprising at least 1% of the total symbiont population. Given the large volume of planulae released from a single parental colony over the course of the spawning season, it is possible that a small number of planulae inherit multiple symbionts, although statistical analyses indicated that this is very unlikely (see results). Given the tools utilized, for this coral species, horizontal symbiont acquisition in addition to vertical symbiont transmission seems plausible, although it remains to be determined whether juveniles that acquire symbionts from the environment survive to adulthood. The occurrence of both vertical and horizontal symbiont transmission modes within a single host species, although not necessarily within a single individual, has previously been inferred in studies on vertically transmitted prokaryotic symbionts. These studies detected phylogenetic evidence of horizontal symbiont transmission [7] – [10] , [55] , although the frequency and life stage of acquisition were not determined. By sampling members of a coral species throughout its ontogeny, we were able to investigate a mutualism with inherited eukaryotic symbionts that may also engage in horizontal symbiont acquisition. Conclusions The ecological and evolutionary implications of employing both modes of symbiont transmission are substantial. Horizontal acquisition of novel symbionts may be a means by which coral species can adapt to environmental changes [56] . Although sexual reproduction in Symbiodinium occurs relatively infrequently [39] , [57] , [58] , new host-symbiont combinations can emerge that may lead to novel, advantageous, and potentially specific symbioses [5] . Horizontal symbiont acquisition in adult corals, however, may either not occur or may be transient [19] , [20] , [41] , [46] , [59] . In coral species with horizontal symbiont acquisition, the juvenile stage appears to be more flexible in acquiring symbionts not present in the adult population [25] – [28] , [42] , [60] . Thus, horizontal symbiont acquisition may allow an individual juvenile to obtain, and subsequently maintain, novel symbionts. In turn, this may increase the juvenile’s fitness and survival as it grows into an adult coral, potentially enabling short-term acclimation on an individual level. Horizontal symbiont acquisition, however, does not provide a mechanism for the perpetuation of the novel symbionts in subsequent generations. Since symbionts are not transferred to the progeny, the advantageous symbionts will be lost each generation when the progeny must acquire symbionts anew. In contrast, if juvenile corals with vertically transmitted symbionts are capable of acquiring novel, advantageous Symbiodinium that are maintained into adulthood, subsequent vertical symbiont inheritance to their progeny would facilitate the maintenance of novel Symbiodinium over generations. Vertical symbiont inheritance, punctuated with horizontal symbiont acquisition, provides an evolutionary mechanism for adaptation to environmental changes through the acquisition and maintenance of advantageous symbionts. S. pistillata , for example, may benefit from both modes of symbiont transmission. On the one hand, the progeny are equipped with inherited Symbiodinium , eliminating the risk of not obtaining their obligate symbionts. On the other hand, juveniles may acquire novel symbionts, potentially increasing their chances of survival in a new environment. If the horizontally acquired novel symbionts improve host fitness, and become abundant in the resulting adult coral, they will be transferred to the brooded progeny, thereby perpetuating the novel symbiosis. Natural selection can then act on the genetic variation in the symbiosis, potentially establishing novel host-symbiont combinations that may be advantageous during changing environmental conditions."
} | 5,688 |
32676585 | PMC7333181 | pmc | 2,372 | {
"abstract": "Stretchable electronics are of great significance for the development of the next-generation smart interactive systems. Here, we propose an intrinsically stretchable organic tribotronic transistor (SOTT) without a top gate electrode, which is composed of a stretchable substrate, silver nanowire electrodes, semiconductor blends, and a nonpolar elastomer dielectric. The drain-source current of the SOTT can be modulated by external contact electrification with the dielectric layer. Under 0-50% stretching both parallel and perpendicular to the channel directions, the SOTT retains great output performance. After being stretched to 50% for thousands of cycles, the SOTT can survive with excellent stability. Moreover, the SOTT can be conformably attached to the human hand, which can be used for tactile signal perception in human-machine interaction and for controlling smart home devices and robots. This work has realized a stretchable tribotronic transistor as the tactile sensor for smart interaction, which has extended the application of tribotronics in the human-machine interface, wearable electronics, and robotics.",
"conclusion": "3. Conclusions In summary, we have proposed an intrinsically stretchable organic tribotronic transistor without a top gate electrode, which consists of a stretchable substrate, silver nanowire electrodes, semiconductor blends, and a nonpolar elastomer dielectric. Using the contact electrification between the Al film and the PDMS dielectric layer, the drain-source current of the SOTT is increased (-4.15 μ A to -5.55 μ A) as the separation distance of the Al film goes up (0 to 250 μ m), with an excellent stability for more than 600 cycles. Composed of stretchable materials, the SOTT can be stretched both parallel and perpendicular to the channel directions, with excellent output performances at the strain range from 0 to 50% along two directions. The SOTT can be stretched for thousands of cycles with less than 10% decrease in output performances, showing an excellent robustness of the SOTT. Moreover, the SOTT can be conformably attached to the human hand. Through the tactile perception of the SOTT, the common smart home devices and the robot have been successfully controlled. This work has realized a stretchable tribotronic transistor as the tactile sensor for smart interaction, which has extended the application of tribotronics in human-machine interface, wearable electronics, and robotics.",
"introduction": "1. Introduction Stretchable electronics are grabbing more and more attention for a wide range of applications in wearable devices, soft mechanics, robotic skin, human-machine interfaces, and so on [ 1 – 6 ]. To date, a series of stretchable functional devices have been developed with prominent tactile-sensing properties based on various physical transduction mechanisms such as piezoresistivity [ 7 , 8 ], capacitance [ 9 ], magnetism [ 10 , 11 ], and optics [ 12 ]. However, most of the tactile-sensing mechanisms for stretchable electronics are passive, lacking direct interaction with human/environment [ 13 – 17 ]. This complicates the process of information acquisition and further influences the tactile perception ability of stretchable functional devices. Therefore, developing stretchable electronics with an active sensing mechanism is highly desired. Since 2012, the triboelectric nanogenerator (TENG) as a new energy technology derived from the Maxwell displacement current has been invented by Wang et al. [ 18 , 19 ], which can effectively convert mechanical energy into electricity [ 20 – 22 ]. In recent years, tribotronics as a new field has been proposed by using the triboelectric potential generated by TENG to control the carrier transport in semiconductors, which has established the direct interaction mechanism between human/environment and electronics [ 23 – 29 ]. Moreover, a variety of tribotronic functional devices have been demonstrated for tactile perception and control, including smart tactile switch [ 30 ], tactile-sensing arrays [ 31 ], active modulation of conventional electronics [ 32 ], and mechanosensation-active matrix [ 33 ]. In addition, tribotronic devices have demonstrated the diversity of material selection [ 24 – 36 ], which is very promising for the intrinsically stretchable electronics for active tactile sensing by further coupling with stretchable materials. Here, we propose an intrinsically stretchable organic tribotronic transistor (SOTT) without a top gate electrode, which is composed of a stretchable substrate, silver nanowire (Ag NW) electrodes, semiconductor blends, and a nonpolar elastomer dielectric. The drain-source current of the SOTT can be modulated by external contact electrification with the dielectric layer. Under 0-50% stretching both parallel and perpendicular to the channel directions, the SOTT exhibits good output performances. After being stretched to 50% for thousands of cycles, the SOTT can survive with excellent stability. Moreover, the SOTT can be conformably attached to the human hand, which can be used for tactile signal perception in human-machine interaction and for controlling smart home devices and robots. This work has realized a stretchable tribotronic transistor as the tactile sensor for smart interaction, which has extended the application of tribotronics in human-machine interface, wearable electronics, and robotics.",
"discussion": "2. Results and Discussion 2.1. Fabrication of the Stretchable Tribotronic Transistor Poly(3-hexylthiophene-2,5-diyl) (P3HT), as a polymer semiconductor, has a high hole mobility and a low band gap width [ 37 ]. The P3HT nanofibril (P3HT-NF) combined with the stretchable elastomer materials has high stretchability, which is good for developing stretchable semiconductor devices [ 38 , 39 ]. Among a lot of elastomers, the polydimethylsiloxane (PDMS) with a simple preparation process can sustain large strains. Moreover, the PDMS has a good triboelectric property, which is good for tribotronic devices [ 40 – 42 ]. Ag NWs have good conductivity, which have been widely used in the field of stretchable electrodes [ 43 , 44 ]. Therefore, in order to obtain highly stretchable SOTT, we exploit the P3HT nanofibril-percolated PDMS rubber composite as a stretchable semiconductor, the Ag NWs dispersed within the PDMS as a stretchable conductor, and the PDMS as a gate dielectric. Through the contact electrification between the external triboelectric layer and the PDMS gate dielectric layer, the drain-source current of the transistor can be modulated. The schematic illustration of the fabrication process for the SOTT is shown in Figure 1(a) . The detailed process is elaborated in Materials and Methods. The SOTT consists of a stretchable substrate, Ag NW electrodes, semiconductor blends, and a nonpolar elastomer dielectric, which is fabricated throughout a sequential lamination transfer process. To prepare the stretchable drain and source electrodes, Ag NWs were spray coated onto an octadecyltrimethoxysilane- (OTS-) pretreated silica wafer through a shadow mask (i), then followed by embedding into a nonpolar elastomeric PDMS substrate (ii). To build an ohmic contact between the Ag NW electrodes and the semiconductor, the PDMS substrate with Ag NW electrodes was immersed into HAuCl 4 ·H 2 O solution for the formation of gold nanoparticles on the Ag NWs by an in situ reduction method (iii). Scanning electron microscopy (SEM) images of the stretchable Ag NW electrodes before and after immersion into HAuCl 4 ·H 2 O solution indicate the successful formation of gold nanoparticles, as shown in Figure S1 . The stretchable semiconductor blends were prepared by a cooling and heating process. Briefly, P3HT was dissolved in m-xylene at 70°C and then cooled to room temperature for the formation of P3HT nanofibrils. After mixing with m-xylene-diluted PDMS, the P3HT NF solution was subsequently spin-coated onto the drain-source electrodes through a polyimide shadow mask to achieve a patterned semiconductor layer (iv). Contained in the transparent PDMS, the semiconductor layer has revealed great optical transparency, which is important for the application in wearable and bionic electronics (Figure S2 ). The stretchable dielectric layer, which is composed of PDMS, was spin-coated onto a polytetrafluoroethylene (PTFE) block and then transferred onto the P3HT NF/PDMS semiconductor layer to form the final SOTT (v and vii). The complete structure of the SOTT shown in Figure 1(b) was obtained by peeling off the whole device from the PTFE block (viii), which has demonstrated a simple structure without a top gate electrode. The channel length is about 500 μ m, as shown in Figure 1(c) . Since all components of the device are stretchable, the prepared SOTT can be stretched both parallel and perpendicular to the channel directions. Figure 1(d) shows the optical graphs of a stretched device in two directions. As clearly seen from the graphs, the device can be deformed without physical damage upon stretching. Moreover, optical microscopy and atomic force microscopy (AFM) images of the P3HT NF/PDMS blends have demonstrated that the semiconductor blends can be stretched without any obvious cracks under 50% strain, as shown in Figures 1(e) and 1(f) , which is very helpful for promoting the stretchability of the SOTT. 2.2. Mechanism and Performances of the Stretchable Tribotronic Transistor The working mechanism of the SOTT is presented in Figure 2(a) . The drain and source electrodes of the SOTT are connected with a voltage source. An aluminum (Al) film, as an external triboelectric layer, fully contacts with the dielectric layer in the initial state for electrification as shown in Figure 2(a) , i. The Al film is electrified with positive charges while the PDMS dielectric layer with negative charges for the difference in charge affinities. Owing to the electrostatic balance, electrical potential difference is not applied to the channel region, and no obvious changes take place in the drain current. When the Al film gradually separates from the PDMS dielectric layer by an external force as shown in Figure 2(a) , ii, negative charges on the dielectric layer surface will induce an inner charge polarization, which will build an inner electric field across the channel and the dielectric surface, leading the holes to accumulate at the interface of the channel and the dielectric layer. As a result, an enhancement zone is achieved in the p-type P3HT NF/PDMS channel, and the drain current is increased. The enhancement zone and the current will be enhanced until a maximum separation distance of the Al film is reached. When the Al film starts to come back to the original station, the inner electric field will be decreased for the depressed inner charge polarization, resulting in the previous accumulated holes diffusing away from the interface of the channel and the dielectric layer. Therefore, the enhancement zone is depressed, while the hole concentration in the interface and the drain current are decreased. Once the Al film reaches its initial state, the hole concentration in the interface and the drain current recover to the original value. It is worth noting that a depletion zone will be formed in the channel and the current will be decreased when a film with strong electronegativity contacts and then separates from the PDMS dielectric layer, such as the fluorinated ethylene propylene (FEP) film. This is the basic operation mechanism of the SOTT, which can be equivalent to a circuit in Figure 2(b) . The electrostatic potential generated from the contact electrification between the Al film and the dielectric layer is equivalent to an external gate voltage, which is illustrated in the energy diagram shown in Figure 2(c) . Figure S3 shows the simulation results of the electrostatic potential generated from the contact electrification between the Al film and the dielectric layer, which indicates that the triboelectric potential is dependent on the separation distance of the Al film. As shown in Figure S4 , the drain current changes with the reciprocating motion of the Al film, showing a consistency with the working mechanism that we discussed above. In addition, we have increased the velocity of the contact-separation movement, until the time of rise and fall for the drain current signal is not changed. The corresponding waveform of I d is shown in Figure S4 , which indicates that the response and recovery times are 80 ms and 90 ms, respectively, which exhibits that the device has a small hysteresis and shows a potential of SOTT to construct sensing electronics. To better evaluate the performance of the SOTT, the electrical characteristics of the SOTT at a different separation distance of the Al film without any mechanical stretching are systematically studied. The separation distance of the Al film is precisely controlled by a linear motor. Figure 2 (d) and Figure S5 show the relationship between the drain current and the separation distance. The inset image is the corresponding transfer characteristics of the SOTT, and the drain current changes with the traditional gate voltage are shown in Figure S6 . The drain current is obtained at a drain voltage of -30 V. With the separation distance increases from 0 to 250 μ m, the drain current is increased from -4.15 μ A to -5.55 μ A. Figure 2 (e) contains the output characteristic curves of the SOTT at a different separation distance from 0 to 250 μ m. The drain current increases with the increasing separation distance, with the drain voltage swept from 0 to -40 V, which is in good accordance with the working mechanism analyzed above. Moreover, as shown in Figure 2(f) , the current can be modulated by the periodic contact-separation motion of the external triboelectric layer for more than 600 cycles with little hysteresis, showing a high stability of the device. To examine the performances of the SOTT under mechanical strain, the electrical characteristics of the SOTT with increasing separation distance are collected by stretching the devices both parallel and perpendicular to the channel directions. Note that the SOTT was fabricated with stretchable components; the stretching strain distributed in the device is effectively suppressed and assumed to be distributed across the whole device, as shown in Figure S7 . Figure 3 (a) shows the transfer characteristics of the SOTT under 50% stretching strain parallel to the channel direction. More results are illustrated in Figure S8 , which has shown that the SOTT can survive after being stretched parallel to the channel direction. When the SOTT is stretched up to 50%, an increase in the drain current from -1.35 μ A to -1.75 μ A is obtained at the separation distance from 0 to 250 μ m. Besides, the output characteristic curve of the SOTT at 50% strain in the parallel stretching direction is shown in Figure 3(b) . More results are depicted in Figure S9 . With the increases of the separation distance, the drain current rises within a drain voltage of 0 to -40 V for the stretched SOTT. The insert image in Figure 3(b) is the optical microscope images of stretched SOTT parallel to the channel direction. It can be clearly observed that the channel was stretched to about 600 μ m without any rupture accrued. On the basis of these transfer curves, the variations of the drain current at different separation distance and stretching strain were calculated, as shown in Figure 3(c) . The device can maintain good performance when the device was stretched by up to 50% parallel to the channel direction. Compared to the parallel direction stretching, an increase from -2.42 μ A to -2.85 μ A in the drain current was obtained at the separation distance from 0 to 250 μ m when the SOTT was stretched to 50% perpendicular to the channel direction, as shown in Figure 3(d) . The device also has good output characteristics when the drain voltage swept from 0 to 40 V after being stretched, as shown in Figure 3(e) . The transfer and output characteristics of the SOTT at different stretching strains from 10% to 40% perpendicular to the channel direction are shown in Figure S10 and S11 , respectively. Also, the variations of the drain current at different separation distances and stretching strains perpendicular to the channel direction were calculated, as depicted in Figure 3(f) . An ideal modulation performance of the SOTT can be also maintained up to 50% stretching strain. Moreover, the distance resolution of the SOTT is illustrated in Figure S12 , showing that the SOTT has excellent distance resolution in the initial state and the stretched state. All these results suggest that the intrinsically stretchable organic tribotronic transistor can maintain good output performances after being stretched, which may promise a bright future of tribotronics in stretchable smart sensing electronics. Specifically, the SOTT shows unprecedented robustness when stretched repeatedly both parallel and perpendicular to the channel directions. As exhibited in Figure 4(a) , only a small shift can be observed in the transfer characteristics curves of a SOTT that was cycled to the stretched state. The normalized maximum current variations during cycling at 50% stretching strain parallel and perpendicular to the channel directions are exhibited in Figure 4(b) . After one thousand stretching cycles, the current variation at the maximum separation distance only decreased by about less than 10% both parallel and perpendicular to the channel directions, showing a high stretchability of the SOTT. Moreover, as shown in Figure 4(c) , the SOTT can be conformably attached to the human hand due to its stretchability, which is very beneficial for skin-inspired devices. Owing to the simple form of the structure, the merit of stretchability, and the retention of performance, the SOTT is very promising for active tactile sensing. Hence, in order to fulfill the potential of SOTT in active tactile-sensing applications, the SOTT was used for controlling smart home devices. As shown in Figure 4(d) , the SOTT is integrated on a finger as a tactile sensor. The finger can be divided into two states. In the initial state, the finger is so straight that the SOTT is not stretched. Positive charges will be induced on the dielectric layer of the SOTT, and a current increase can be observed when another finger touches the SOTT. When the finger is bent and the SOTT is stretched, a touch can still be perceived by the stretched SOTT, which has shown a high potential of the SOTT as a tactile sensor for a smart control system regardless of pristine or stretched states. Through the tactile perception of the SOTT, the common home devices, such as a table lamp, a bell, and an electric fan, can be wirelessly controlled by a finger touch, as shown in Figure 4(e) , which presents many potential applications for the SOTT in daily life, such as the self-care for the disabled. Moreover, as shown in Figure 4(f) , tactile sensing can promote the function of human-machine interaction, such as wirelessly controlling a robot. A SOTT is directly attached to the finger for robot control. A current change signal can be observed when the finger touches the SOTT, which is treated as a tactile-sensing signal. The signal is the original blinking signal from the SOTT, followed by the signal after being filtered, amplified, and relay-converted. Then, the output terminal of the latching relay is connected with a microcontroller which can send the robot control instructions through a wireless transmitting module. The robot posture will be controlled when a signal is received by the wireless receiving module on the robot. As a response to the tactile, the robot changes the posture from standing to crouching, as shown in Figure 4(g) . When the SOTT is in the stretched state, the posture of the robot can still be controlled by a finger touch, as depicted in Figure 4(h) , which has shown great application prospects of the SOTT in smart interaction. All these results have demonstrated the remarkable application potential of the SOTT in human-machine interface, wearable electronics, intelligent skin, and robotics."
} | 5,062 |
37194043 | PMC10186816 | pmc | 2,374 | {
"abstract": "Background The lithospheric microbiome plays a vital role in global biogeochemical cycling, yet their mutual modulation mechanisms remain largely uncharted. Petroleum reservoirs are important lithosphere ecosystems that provide desirable resources for understanding microbial roles in element cycling. However, the strategy and mechanism of modulating indigenous microbial communities for the optimization of community structures and functions are underexplored, despite its significance in energy recovery and environmental remediation. Results Here we proposed a novel selective stimulation of indigenous functional microbes by driving nitrogen and sulfur cycling in petroleum reservoirs using injections of an exogenous heterocycle-degrading strain of Pseudomonas . We defined such bacteria capable of removing and releasing organically bound sulfur and nitrogen from heterocycles as “bioredox triggers”. High-throughput 16S rRNA amplicon sequencing, metagenomic, and gene transcription-level analyses of extensive production water and sandstone core samples spanning the whole oil production process clarified the microbiome dynamics following the intervention. These efforts demonstrated the feasibility of in situ N/S element release and electron acceptor generation during heterocycle degradation, shifting microbiome structures and functions and increasing phylogenetic diversity and genera engaged in sulfur and nitrogen cycling, such as Desulfovibrio , Shewanella , and Sulfurospirillum . The metabolic potentials of sulfur- and nitrogen-cycling processes, particularly dissimilatory sulfate reduction and dissimilatory nitrate reduction, were elevated in reservoir microbiomes. The relative expression of genes involved in sulfate reduction ( dsrA , dsrB ) and nitrate reduction ( napA ) was upregulated by 85, 28, and 22 folds, respectively. Field trials showed significant improvements in oil properties, with a decline in asphaltenes and aromatics, hetero-element contents, and viscosity, hence facilitating the effective exploitation of heavy oil. Conclusions The interactions between microbiomes and element cycling elucidated in this study will contribute to a better understanding of microbial metabolic involvement in, and response to, biogeochemical processes in the lithosphere. The presented findings demonstrated the immense potential of our microbial modulation strategy for green and enhanced heavy oil recovery. \n Video Abstract Graphical Abstract \n Supplementary Information The online version contains supplementary material available at 10.1186/s40168-023-01553-7.",
"conclusion": "Conclusions This study disentangled the microbial community succession dynamics and revealed that the active biogeochemical sulfur and nitrogen cycles induced by exogenously introduced heterocycle-degrading bacteria could significantly shift the structure and function of the microbiome and shape new metabolic niches in deep subterranean environments. This approach breaches the obstacles of conventional direct nutrient injections, such as the rapid depletion of electron acceptors during their transportation and failure to achieve targeted activation of functional microorganisms. Meanwhile, the proposal of this engineered microbial intervention strategy and the revelation of the modulation mechanism will inform future studies on interconnections between microbial activities and lithospheric biogeochemical cycling. Our findings also offer prospects for expanding biotechnological applications for recovering the vast quantity of heavy oil resources that would otherwise remain unexploited. Further efforts are needed to comprehend the microbial community dynamics and their relationships with biogeochemistry in the lithosphere more deeply and to unravel mechanisms of interspecies interactions.",
"introduction": "Introduction Microbial metabolic activities are intricately intertwined with myriads of environmental interactions that collectively shape the biochemical dynamics of ecosystems. The most prevalent and ecologically significant elements, carbon, nitrogen, oxygen, and sulfur, exist in a variety of oxidation states and chemical forms, and their involvements in metabolically driven redox reactions completed by microorganisms substantially influence the biogeochemical cycling of ecosystems [ 1 ]. Mounting research on associations between microorganisms and elemental transformations in the habitats of oceans [ 2 – 4 ] and surface soils [ 5 – 7 ] is igniting new interests in the exploration of microbial resources, implications on Earth’s biogeochemistry and global climate, and novel approaches for energy development and environmental remediation [ 8 – 10 ]. However, knowledge about how microbiomes are involved in elemental cycling in lithospheric ecosystems remains limited. Although recent studies have characterized the geographic distribution of microbial life in shallow terrestrial subsurface ecosystems such as groundwater and surficial mangrove sediments [ 11 – 13 ], the modulation and response of the biogeochemical nitrogen and sulfur cycles involving microbial activities in the deep terrestrial lithosphere, particularly petroleum reservoirs that typically occur hundreds of meters below the Earth’s surface, are poorly understood due to difficulties in sample accessibility and processing. The lithosphere has a complex chemical environment, extreme physiochemical conditions, and phylogenetic and metabolically diverse microbial communities [ 14 ]. Despite the relatively low microbial biomass of the lithosphere (2–19% of Earth’s total biomass) [ 15 , 16 ], the richness of mineral constituents (e.g., Fe, Mn, N, S) and the vast biodiversity of crustal microbiota suggest significant microbial roles and impacts on Earth’s biogeochemical cycling. Petroleum reservoirs, as essential lithosphere components, encompass abundant carbon sources and heteroatomic chemicals in porous minerals. Meanwhile, the tremendous species diversity and metabolic versatility of the microbiome observed in petroleum reservoirs endow them with favorable attributes as natural laboratories for researching microbial participation in lithospheric biogeochemical cycling [ 17 , 18 ]. Multitudinous functional microorganisms capable of performing redox of carbon, nitrogen, and sulfur compounds have been identified in reservoirs [ 17 , 19 ]. Multiplex electron flows and elemental transformations will inevitably disrupt microbial homeostasis and induce shifts in the composition, diversity, and metabolic potential of the microbiome when subject to environmental variation. However, interactions between microbial metabolic activities and nutrient cycling remain elusive. Considering that microbial metabolism and growth are predominantly constrained in the nutrient-scarce lithosphere by the availability of electron acceptors [ 18 , 20 , 21 ], utilitarian indigenous microbial enhanced oil recovery (IMEOR) techniques have been proposed adopting the fundamental principle of replenishing reservoir electron acceptors. As promising alternatives to physiochemical approaches, they biostimulate indigenous microorganisms, strengthening microbial metabolism to alter crude oil properties by directly infusing air, nitrate, or sulfate into reservoirs [ 22 ]. Nevertheless, injected limited electron acceptors will be indiscriminately utilized and rapidly exhausted during their transportations in subsurface porous rocks; hence, the purpose of selectively activating functional microorganisms for long-term oil production enhancement has not yet been attained. An enormous challenge in heavy oil recovery is the high-viscosity property resulting from the massive quantities of heterocyclic compounds [ 23 , 24 ]. These polar compounds incorporate multiple heteroatoms, such as nitrogen, sulfur, oxygen, and metals (e.g., V, Fe, Ni) on their lattices of aromatics [ 23 – 25 ]. Therefore, the removal of heteroatoms is paramount for enhancing oil production (by reducing oil viscosity), alleviating downstream refining operations, and upgrading oil value. In particular, it is anticipated that the release of N/S elements due to carbon-sulfur and carbon-nitrogen bond cleavage could potentially yield electron acceptors (e.g., nitrate and sulfate) by microbial or redox-active matter-mediated oxidation processes [ 20 , 26 – 30 ], thereby altering the microbial ecology and further driving the cryptic nitrogen and sulfur cycles in the lithosphere. Microorganisms with superior heterocycle decomposition capabilities will serve as triggers for fueling element cycling and microbial community succession in the subsurface. However, the isolation of robust heterocycle-degrading bacteria that are adaptable to anaerobic environments has remained challenging, and the influence of introducing exogenous heterocycle-degrading bacteria on autochthonous microbial communities and interaction networks of microbial ecology in the lithosphere has yet to be resolved. Here, we presented an effective microbial regulatory strategy for driving nitrogen and sulfur cycles and stimulating functional microorganisms by introducing exogenous heterocycle-degrading bacteria into heavy oil reservoirs. A facultative anaerobic bacterial strain Pseudomonas sp., which had been demonstrated to have remarkable performance in S,N-heterocycle degradation [ 31 , 32 ], was isolated and charged into heavy oil reservoirs for field trails. The main objective was to explore the microbial succession and involvement in biogeochemical cycling following the modulation of the exogenous bacteria and its application prospect for recovering heavy oil resources. We characterized the consecutive taxonomic and functional profiles of production water (PW) and pristine sandstone core samples over the production lifetime of four oilwells. Through 16S rRNA, metagenomic, and quantitative gene transcription-level analyses, we were able to decipher the key role of the in situ release of N/S elements in structuring the indigenous microbial community of petroleum reservoirs. In addition, we examined the chemical constituents of heavy oil, PW, and gas samples to track the flow of N/S elements in varying valence states, with a particular focus on potential electron acceptors (e.g., nitrate and sulfate).",
"discussion": "Discussion Exploration of associations between element cycling and microbial metabolic activities provides a unique portal into the operation of microbial life in deep crustal environments. This study sheds light on the succession law of microbial communities in heavy oil reservoirs following our modulation and highlights the critical role of the release and transformation of N/S elements in affecting lithospheric biogeochemical cycling. A field microbial regulatory practice for enhanced heavy oil recovery was undertaken, revealing the mechanism of the manipulation of exogenous heterocycle-degrading bacteria to selectively stimulate indigenous functional groups related to sulfur and nitrogen metabolism. We propose naming this sort of bacteria with exceptional capabilities of dissociating heteroatoms from heterocycles or minerals the “bioredox triggers”. In detail, pristine subsurface environments are nutrient deficient, and thus microbial-driven heterocycle degradation and element cycles are stagnant. Upon the injection of exogenous Pseudomonas sp., N/S elements in heterocycles were dissociated and liberated as reduced inorganic nitrogen and sulfur compounds during bioaugmented heterocycle decomposition (Fig. 7 ). N/S chemicals might be oxidized by redox-active matter (e.g., ferrihydrite, manganese oxides, perchlorate) or microbe-mediated (biotic and abiotic) reactions to nitrate and sulfate (or their intermediate substances), when transported by pore water to zones with high ORP (oxidation/reduction potential) in highly heterogeneous oil reservoir environments ( Supplementary Results and Fig. S 15 ). These oxidative substances will serve as electron acceptors, thereby establishing multiplex metabolic niches in petroleum reservoirs. Changes in reservoir niches and the appearance of inorganic electron acceptors resulted in a rapid proliferation of SRB and NRB, which in turn accelerated the biodegradation of heterocycles [ 80 – 82 ]. Diversity analysis showed that the taxonomic diversity of microorganisms in PW was significantly enhanced. This phenomenon is possibly the consequence of the promoted functional redundancy, which means that the advent of new metabolic niches led to rising microbial clades with similar functions competing and cooperating within the niches, better buffering against taxonomic shifts caused by the intervention [ 64 , 83 ]. Furthermore, the phylogenetic and functional diversity of the lithospheric mineral microbiome was significantly higher than that of the PW microbiome. The view was supported by our delineation of co-occurrence subnetworks, which showed that more modules (dissimilar niches) were categorized in the network of lithospheric minerals (Supplementary Fig. S 7 ). Fig. 7 Schematic diagram of exogenous bacteria (“bioredox trigger”) modulating lithospheric indigenous microbial communities and element cycling. The early primary reservoirs (May, June) were primarily dominated by Proteobacteria and Firmicutes where electron acceptors were deficient. Our strategy of exogenous single-species manipulation was implemented in July by injecting heterocycle-degrader Pseudomonas strains into the subsurface. The N/S elements were in situ released and transformed to replenish extremely scarce electron acceptors in the lithosphere, activating the lithospheric N/S cycling and inducing microbial community succession (after Aug) We also discerned minor transitions in bacterial community compositions and functions of PW following the intervention. SRB and genes responsible for dissimilatory sulfate reduction were more prevalent in the former phase (Aug), whereas SRB subsequently diminished, and the reservoir was prone to the dominance of anaerobic respiration of nitrate and nitrite by NRB or NR-SOB (in Sep and Oct). The trend of SRB was in good concordance with the change in H 2 S concentration in reservoirs (Supplementary Fig. S 2 a), which increased dramatically during August and then declined after mid-September. Possible reasons for this microbial succession pattern could include the following: (1) Nitrogen content is generally much lower than sulfur content in petroleum asphaltenes and thus less nitrogen was released [ 84 ]; (2) Metagenomics and previous studies indicated that SRB, primarily affiliated with Desulfobacterota , possessed larger potentials for S-heterocycle biodegradation such as benzothiophene [ 80 , 81 ], whereby the release of sulfur was preferentially promoted; (3) The redox potentials ( \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{E}}^{0\\mathrm{^{\\prime}}}$$\\end{document} E 0 ′ ) of \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NO}}_{3}^{-}$$\\end{document} NO 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}$${\\mathrm{NO}}_{2}^{-}$$\\end{document} NO 2 - (+430 mV) and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NO}}_{2}^{-}$$\\end{document} NO 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}$${\\mathrm{NH}}_{4}^{+}$$\\end{document} NH 4 + (+440 mV) are much higher than that of \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{SO}}_{4}^{2-}$$\\end{document} SO 4 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}$${\\mathrm{HSO}}_{3}^{-}$$\\end{document} HSO 3 - (−516 mV) and \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{HSO}}_{3}^{-}$$\\end{document} HSO 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}$${\\mathrm{HS}}^{-}$$\\end{document} HS - (−110 mV), and therefore nitrate reduction and reduced sulfur oxidation are thermodynamically favored [ 21 ], resulting in increasing sulfate accumulation but little residual nitrate at inception; (4) The rapid proliferation of SRB caused fierce trophic competition among species, the depletion of sulfate, and elevated concentrations of hydrogen sulfide, inhibiting the growth of SRB; (5) SRB and dissimilatory sulfite reductase can be competitively inhibited with the accumulation of nitrite (or nitrate) in PW (9, 10). Co-occurrence network analysis revealed that the release of N/S markedly affected bacterial community structure and created new metabolic niches. Co-occurrence of ASVs in a module is indicative of similar niche adaptation and interspecies links signify their competitive or cooperative relationships [ 85 ]. Bacterial co-occurrence patterns indicated that the emergent bacteria, primarily affiliated with Desulfobacterota and Campilobacterota , could co-exist with Gammaproteobacteria (e.g., Pseudomonas ) and Firmicutes , which suggested that the directed stimulation of functional groups could rely on the synergistic modulation of electron acceptor provisioning and interspecies symbiosis (such as mutualism, commensalism, or competition). The analysis of topological properties indicated that the magnitude and connectedness of the subnetwork of the PW microbiota (8, 9, 10) showed an obvious increase after the intervention. The increase in connections could partially stem from community successions and the increased species diversity. Alternatively, the appearance of available nutrients and electron acceptors (e.g., \\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym} \n\t\t\t\t\\usepackage{amsfonts} \n\t\t\t\t\\usepackage{amssymb} \n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$${\\mathrm{NO}}_{3}^{-}$$\\end{document} NO 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}$${\\mathrm{SO}}_{4}^{2-}$$\\end{document} SO 4 2 - ) induced more convoluted and stronger interspecies cooperation or competence for nutrient acquisition, which is the canonical trait in N/S-transforming microbial systems. For example, nitrifiers Nitrosopumilus spp. facilitate the growth of anammox bacteria ( Ca Scalindua spp.) by generating nitrite from ammonia oxidation, but also compete for ammonia alongside them [ 1 ]. It is worth noting that higher community stability was speculated to be established for PW (8, 9, 10) due to the weaker interactions (smaller clustering coefficient) and higher connectedness, implying that the introduction of exogenous bacteria enabled lithospheric environments to remain comparably more resistant and resilient to environmental perturbations. Field experiments validated the extraordinary effectiveness of our microbial modulation in viscosity reduction, heteroatom removal, and incremental yield of heavy oil."
} | 5,231 |
24824794 | PMC4019531 | pmc | 2,375 | {
"abstract": "Coral reef ecosystems are based on coral–zooxanthellae symbiosis. During the initiation of symbiosis, majority of corals acquire their own zooxanthellae (specifically from the dinoflagellate genus Symbiodinium ) from surrounding environments. The mechanisms underlying the initial establishment of symbiosis have attracted much interest, and numerous field and laboratory experiments have been conducted to elucidate this establishment. However, it is still unclear whether the host corals selectively or randomly acquire their symbionts from surrounding environments. To address this issue, we initially compared genetic compositions of Symbiodinium within naturally settled about 2-week-old Acropora coral juveniles (recruits) and those in the adjacent seawater as the potential symbiont source. We then performed infection tests using several types of Symbiodinium culture strains and apo-symbiotic (does not have Symbiodinium cells yet) Acropora coral larvae. Our field observations indicated apparent preference toward specific Symbiodinium genotypes (A1 and D1-4) within the recruits, despite a rich abundance of other Symbiodinium in the environmental population pool. Laboratory experiments were in accordance with this field observation: Symbiodinium strains of type A1 and D1-4 showed higher infection rates for Acropora larvae than other genotype strains, even when supplied at lower cell densities. Subsequent attraction tests revealed that three Symbiodinium strains were attracted toward Acropora larvae, and within them, only A1 and D1-4 strains were acquired by the larvae. Another three strains did not intrinsically approach to the larvae. These findings suggest the initial establishment of corals– Symbiodinium symbiosis is not random, and the infection mechanism appeared to comprise two steps: initial attraction step and subsequent selective uptake by the coral.",
"introduction": "Introduction Reef-building corals engage zooxanthellae, as symbionts that supply them with photosynthetic products, which enable corals to effloresce in oligotrophic tropical seas. This coral–algal symbiosis is a fundamental pillar for biologically and economically important coral reef ecosystems. The symbiont algae, dinoflagellate genus Symbiodinium are divided into nine phylogenetically distinct genetic groups (clades A–I) [1] , and each clade consists of numerous genotypes (e.g. [2] ). Physiological responses to environmental stresses may differ among different clades and genotypes [3] , [4] . Sexual progeny of corals can acquire Symbiodinium by either of the two modes: vertical transmission (maternal inheritance) or horizontal transmission (acquisition from environment), and corals that acquire Symbiodinium from the environment are predominant [5] . The horizontal transmission is considered to be advantageous by enabling corals to acquire Symbiodinium adapted to their newly settled environments, however, there is still scant evidence supporting this idea [6] . Furthermore, it is still unclear how corals that acquire Symbiodinium by horizontal transmission recognize and acquire their symbionts from the environmental population. To answer this question, several infection experiments have been carried out using Symbiodinium culture strains or freshly isolated cells from adult corals or other zooxanthellate animals [7] – [12] . Although such infections were usually successful, the results often differed between experiments. For example, Cumbo et al. [11] reported that Acropora larvae can acquire a wide variety of Symbiodinium clades, whereas Yuyama et al. [12] demonstrated that the infectivity of Symbiodinium cells in Acropora tenuis juveniles can differ among the Symbiodinium clades. Thus, it is still unclear whether the new generations of host corals acquire their own Symbiodinium randomly or selectively. Previous field observations revealed that in the common reef-building coral Acropora , Symbiodinium genotype compositions often differ between recruitment/juvenile stages and adult populations [13] , [14] . Yamashita et al. [14] only detected clade A and/or D Symbiodinium in 55 naturally settled 2-week-old Acropora recruits, whereas clade C Symbiodinium , which are the dominant symbionts in adult Acropora corals, were never detected. The Acropora recruits tested by Yamashita et al. [14] were identified to the species level by Suzuki et al. [15] and they comprised at least 10 Acropora species, including the dominant Acropora species in this area, e.g., Acropora hyacinthus , Acropora digitifera , Acropora nasuta , Acropora intermedia , and Acropora selago . Thus, it is plausible to suggest that clades A and D Symbiodinium play important roles during the initial symbiosis. In the present study, to clarify whether Acropora corals acquire their symbionts selectively or randomly, we compared the Symbiodinium genotype compositions (at a finer scale compared with Yamashita et al. [14] ) in naturally settled Acropora recruits (approximately 2 weeks old) and in the adjacent seawater. Furthermore, we also conducted laboratory experiments in which the Acropora coral larvae were artificially infected with naturally occurring densities of various Symbiodinium genotypes. Our results demonstrated that in the initial stage of symbiosis, Acropora corals had an apparent preference for specific genotypes, even when these genotypes were present at much lower densities than the non-selected Symbiodinium . The mechanism of this preferential association was examined by an attraction experiment to determine whether the corals selected the symbionts or whether the Symbiodinium selected the corals. Our results from field observations and laboratory experiments suggest underlining two-step mechanisms of attraction and selection for the initial establishment of corals– Symbiodinium symbiosis.",
"discussion": "Discussion Our results clearly showed that Acropora corals do not randomly acquire Symbiodinium from surrounding environments, at the initial stage of their symbiosis. Environmental Symbiodinium (i.e., solitary Symbiodinium in the environment away from host animals) are highly diverse. Clades A, B, C, D, E, G, and H have been previously detected from the water column and/or sediment samples in the Pacific [14] , [17] – [19] . In the present study, we collected water samples during the coral spawning period of 2011 to clarify potential symbiont sources for the coral larvae, and detected clades A, C, D, and G. Therefore, it was plausible that apo-symbiotic corals acquired these diversified environmental Symbiodinium . In our study area, Urasoko Bay, several species of adult Acropora corals, Porites lutea , Cyphastrea serailia harbored only clade C Symbiodinium \n [14] , [20] , whereas Pocillopora eydouxi and Favites abdita harbored mainly clade C with background level of clade D Symbiodinium \n [20] . The environmental Symbiodinium populations in Urasoko Bay were previously reported to primarily consist of clade C, and clades A/D were relatively rare [14] . In the present study, however, almost equal number of clade C and D clones was recovered from water samples. Because ITS1 copy number of clade D Symbiodinium was estimated to be approximately three times higher that of clade C [21] , it is plausible to assume that clade C populations were approximately three times larger than clade D populations in the same period. Nevertheless, our present and previous field observations [14] demonstrated that naturally settled 2-week-old Acropora coral recruits usually harbored clades A and/or D Symbiodinium , which were rare in the environment. Littman et al. [22] reported that high cell densities (1000–4000 cells/mL) of Symbiodinium were present in the benthic communities of the Great Barrier Reef; furthermore, in Hawai‘i, Takabayashi et al. [18] demonstrated that the diversity of Symbiodinium was higher in sediments than that in the water column. Thus, sediments are considered to be an important source of symbionts for corals. In our study area of Urasoko Bay, Symbiodinium clade compositions in the sediments were previously reported [14] . However, Symbiodinium cell densities could not be determined due to the low Symbiodinium DNA concentrations in the sediment samples; the dominant Symbiodinium in the sediments belonged to clade C, whereas clade A and D were relatively rare, as found in the water column [14] . This result may indicate that the benthic Symbiodinium cell densities and diversity in our study area of Urasoko bay were lower than those in other coral reef environments. In previous field observations, clades A and/or D Symbiodinium were also detected from Acropora recruits, e.g., at the Great Barrier Reef [13] , [23] , [24] and Ishigaki Island [14] . In the present study, we only examined seven Acropora recruits from five Acropora species; Symbiodinium clade compositions within the recruits were in accordance with our previous observation in the same area [14] . Furthermore, ITS2 type analysis revealed that clade A type compositions were completely different between the environment and the recruits. Environmental A types consisted of type A1, A2 relative (free-living), and other A type group, whereas naturally settled recruits primarily harbored type A1 and a few A3 group members. Type A2 relative group was not detected in natural recruits. This result indicated that Acropora corals may engage in symbiosis with specific types in the specific clade. In contrast, clade D types within the recruits were not so different from environmental D types (but statistically different). Many clade D types have been determined by PCR–denaturing gradient gel electrophoresis (DGGE) fingerprinting, and several types are known to have two or three different ITS2 sequences within their genomes [25] . For example, type D1-4 (formerly type D1a) has both a D1 sequence and a D4 sequence in its genome [25] . PCR cloning separately detects D1 and D4 sequences. Therefore, D1 and D4 sequences recovered in our study were presumed to originate from a Symbiodinium cell with D1 and/or D4 sequence(s) in its genome. D1 and D4 were the most abundant sequences in both the water and recruit samples; thus, the clade D type compositions were not so different between the samples (p = 0.01009), compared to the case of A type compositions (p = 1.47×10 −6 ). However, type D3 relative Symbiodinium was only detected from water samples and never from recruit samples. This result indicates that Acropora corals do not randomly acquire environmental clade D Symbiodinium , as is the case with clade A Symbiodinium . We performed infection tests with apo-symbiotic planula larvae of Acropora tenuis using clade A–F Symbiodinium culture strains. Infection test results well reflected field observations, particularly type A1 (AJIS2-C2) and type D1-4 (CCMP2556) were readily acquired by A. tenuis larvae. Interestingly, type A1 strain UcM-C2, with an ITS region sequence identical to that of AJIS2-C2 [19] , was not often acquired by the larvae. This result may indicate that characteristics may differ even within the same ITS2 sequence. In the present study, although A. tenuis larvae were used in the laboratory infection tests, we could not find naturally settled A. tenuis recruits in our field samples. Two of 55 2-week-old Acropora recruits that were analyzed previously by Yamashita et al. [14] were identified as A. tenuis by Suzuki et al. [15] , and these corals harbored clade D Symbiodinium . In the present study, however, the laboratory infection test revealed that A. tenuis larvae can acquire clade A (type A1) and clade D (type D1-4) strains. Type A3 and clade B Symbiodinium strains also infected the larvae, although the infection rates were lower than that for types A1 and D1-4. Previous laboratory experiments reported that type A3 and clade B Symbiodinium were acquired by apo-symbiotic Acropora larvae/juveniles [12] , [26] . Although, we could not detect type A3 Symbiodinium in the water samples, one A3 clone was recovered from a natural recruit. Environmental type A3 Symbiodinium was detected by Yamashita and Koike [19] in water samples collected from Urasoko Bay. In contrast, clade B Symbiodinium was never detected from environmental waters or Acropora corals of Urasoko Bay [14] , [20] . Furthermore, type A2 relative and clade C Symbiodinium were readily detected in the water column of Urasoko Bay; however, these groups were never acquired by Acropora tenuis larvae. These results are in agreement with previous reports that type A2 relative groups were not acquired by cnidarian hosts [27] , [7] and that clade C Symbiodinium was relatively rare in Acropora recruits from the field [13] , [14] . In our trial, type A2 relative Symbiodinium strain ISS-C2-Sy could infect the larvae under a high cell density condition (14,000 cells/L), whereas at low and medium cell densities (140 or 1400 cells/L), the larvae did not acquire ISS-C2-Sy cells. This result may indicate that the larvae could acquire any Symbiodinium if the cells were abundant. However, in coral reef environments, Symbiodinium cell densities in the water column were usually only several hundred cells/L or less [14] , although higher cell densities have been reported from the Great Barrier Reef [22] . According to previous laboratory infection tests, Acropora larvae can acquire several Symbiodinium genotypes (e.g. [11] ), whereas Yuyama et al. [12] demonstrated that some Symbiodinium clades did not infect Acropora juveniles. Laboratory infection tests have revealed important information regarding the initial symbiotic stage of Symbiodinium and corals, e.g., the timing of symbiont acquisition by the coral larvae [9] , [10] , differential growth rates [23] , and gene expression [28] among juvenile polyps infected with different Symbiodinium . However, these previous experiments required high-density Symbiodinium cells, and thus might be not representative of natural infection behavior. Although the cell densities of symbiont source were low, corals established symbiosis with these sources in the field. Furthermore, types A1 and D1-4 cells were successfully acquired by the larvae at lower cell densities (140 cells/L). It is plausible to assume that corals and/or Symbiodinium attract each other to establish initial symbiosis. In our attraction tests, the A. tenuis larvae were not attracted by any of the Symbiodinium cultures, whereas five of eight tested Symbiodinium strains were found within the tubes with A. tenuis larvae. That is, type A1 (AJIS2-C2), type A2 relative (ISS-C2-Sy), clade B (CCMP1633), type D1-4 (CCMP2556), and clade E (MJa-B6-Sy) Symbiodinium cultures. Among these five strains, type A1, type D1-4, and clade E Symbiodinium cell densities were significantly higher in the tubes with larvae compared with that in the tubes without larvae. Thus, these three Symbiodinium strains are considered to have been attracted by the larvae. We used five string tubes in this attraction test, thus the possibility of a disproportionate distribution of Symbiodinium cells could not be discounted entirely. However, we found that 5–14 AJIS2-C2, CCMP2556, and MJa-B6-Sy cells were present in the tubes with larvae, which did not support this possibility, because we did not find any Symbiodinium cells in the tubes without larvae (control tubes). The attraction behavior of Symbiodinium cells has also been reported in previous studies. For example, in the laboratory experiments, motile Symbiodinium cells exhibited taxis for the coral mouth [29] , and some of the coral extracts can attract Symbiodinium cells (Takeuchi et al. personal communication). Indeed, motile-stage Symbiodinium have chemosensory responses [30] , and also they exhibit phototaxis toward green light [31] . Some of the corals, green fluorescence was concentrated around the mouth region of the larvae approximately 4–6 days after spawning [32] , which corresponded to the timing of symbiont acquisition by the coral larvae. In the environment, these abilities and/or characteristics of Symbiodinium and corals may be used for initial contact in the establishment of symbiosis. Hollingsworth et al. [31] , [32] generated the “beacon hypothesis,” suggesting that green fluorescent proteins may play a role in attracting motile Symbiodinium cells to apo-symbiotic larvae. In fact, motile-stage Symbiodinium cells have an eyespot that is considered to reflect green light [33] . In our preliminary experiment, dead Symbiodinium cells were not acquired by the larvae ( Text S1 ). Although dead cells and live non-motile cells may be different when the larvae acquire Symbiodinium cells, it is inferred also from this preliminary data that attraction behavior of Symbiodinium cells is one of the key functions for initiating the symbiosis. Our laboratory and field observations, however, indicated that Acropora corals do not acquire all types of attracted Symbiodinium cells. Type A2 relative, clade B, and E Symbiodinium cells were found within the tubes with larvae, but they did not infect the larvae in test tubes, and these genotypes were not found in the field coral recruits/juveniles. Furthermore, even with the readily acquired types A1 and D1-4, cell densities within the larvae were independent of the inoculated cell densities. In the present study, we inoculated type A1 and D1-4 Symbiodinium at different densities that ranged from 1 to 100 cells/larva. The A. tenuis larvae successfully acquired Symbiodinium cells even with a low Symbiodinium cell density (1 cell/larva). However, even when we inoculated high densities (100 cells/larva) of type A1 and D1-4 Symbiodinium , the larvae only harbored 4.3±0.8 and 10.9±3.5 cells of type A1 and D1-4, respectively. These results indicate that the initial symbiosis between corals and Symbiodinium is not determined by only infectivity of the Symbiodinium , and possibly involves selection by the corals. Several reports have indicated that corals recognize Symbiodinium cell-surface sugars (glycoproteins and glycolipids) by means of lectin at the onset of symbiosis [34] - [37] . Thus the initial establishment of symbiosis between apo-symbiotic corals and Symbiodinium cells involves a two-step selection; initial attraction step and subsequent selective uptake step. Both types A1 and D1-4 Symbiodinium groups have been detected from various hosts from various geographic ranges [38] , [25] . In addition, the combined results of our field observations and laboratory experiments suggested that type A1 and D1-4 Symbiodinium are the first symbiont partners of Acropora corals. Thus, a future challenge is to elucidate the benefits of symbiosis establishment between Acropora corals and type A1/D1-4 Symbiodinium ."
} | 4,772 |
36703435 | PMC9814380 | pmc | 2,376 | {
"abstract": "Direct electron transfer at microbial anodes offers high energy conversion efficiency but relies on low concentrations of redox centers on bacterium membranes resulting in low power density. Here a heat-treatment is used to delicately tune nitrogen-doping for atomic matching with Flavin (a diffusive mediator) reaction sites resulting in strong adsorption and conversion of diffusive mediators to anchored redox centers. This impregnates highly concentrated fixed redox centers in the microbes-loaded biofilm electrode. This atomic matching enables short electron transfer pathways resulting in fast, direct electrochemistry as shown in Shewanella putrefaciens ( S. putrefaciens ) based microbial fuel cells (MFCs), showing a maximum power output higher than the conventional non-matched nitrogen-doped anode based MFCs by 21 times. This work sheds a light on diffusion mediation for fast direct electrochemistry, while holding promise for efficient and high power MFCs.",
"introduction": "Introduction As a renewable energy source to harvest electricity by oxidizing substrates (electron donor) such as organic wastes, microbial fuel cells (MFCs) have received great attention in recent years 1 , 2 . The practical application of MFCs is still restrained by its relatively low power density in comparison to the conventional H 2 -O 2 fuel cells, due to the poor electron transfer between bacteria and electrode 3 – 8 . The electron transfer pathways of MFCs anodes are generally divided into two processes, namely, direct electron transfer and diffusive redox species-mediating transfer. It has also been noted that the electron transfer mediated by endogenously generated electron shuttles from bacteria is even more advantageous in MFCs than the membrane active centers-enabled direct electron transfer, since such highly concentrated mediators can produce high current density while eliminating adding additional electron mediators 9 – 12 . As an important redox moiety for electron transfer cycles in MFCs, Shewanella sp . as a model organism has been widely used and intensively investigated in MFCs. Flavin 13 , 14 as an electron mediator permits bacteria to utilize a remote electron acceptor that was is accessible to the cells. Flavin mononucleotide (FMN) has been utilized as a two-electron/proton redox mediator as FMN + 2e − + 2 H + → FMNH 2 9 , 15 , and the electron transfer mediated by Flavin is executed at 1, 5-nitrogens via a two-electron reaction between their Quinone and hydroquinone species 11 . FMN can be either in free-state or bonded state in MFCs anode. It has been proposed a possibility that the concentration of free Flavin in MFCs anode could greatly affect the power output while the bonded Flavins could allow direct electron transfer between the outer membrane cytochromes and the electrode 16 , 17 , but it has not been proved. The direct electrochemistry (DEC) is always more efficient and faster than a diffusive redox species-mediated indirect electrochemistry in the MFCs anode due to its short electron pathways but it is still very challenging to achieve high current density due to lack of high density of reaction active centers for the fast direct electron transfer. Currently, MFCs anodes mainly rely on mediation chemistry, an indirectly electron mediator-based electron transfer process 6 , 9 , 13 , in which the lowly diffusive flux of electron mediators as electron shuttles often suffers from diffusion limit presenting a diffusion control process. In particular, the MFC anodes are porous and the diffusion limiting even more seriously affect the electricity harvest in MFCs leading to high energy losses 16 . There is very high motivation to develop efficient approaches to accelerate the direct electrochemistry of mediators such as Flavin for high power MFCs while exploring fundamental insights of the direct electrochemistry. Functional nanomaterials have been used to modify MFCs anodes to improve the electron transfer process by either overcoming the steric effect of diffusive redox shuttles such as Flavin to access the electrode surface or offering proximity of electrode nanomaterial to cell membrane for direct electrochemistry 11 . However, these approaches can still not significantly improve the power density of MFCs. Among various functional materials used in MFCs, nitrogen-doped carbon materials have been investigated due to their good electrical conductivity, rich surface functional groups as well as unique electronic structure 18 – 20 . As an example, nitrogen-doped carbon nanoparticles have been used to increase anodic adsorption of flavins for facilitating shuttle-mediated extracellular electron transfer 21 . Previous work has applied plasma to implant N + ion for MFCs anode modification for enhanced biofilm growth and improved interaction between bacteria and electrode 22 . Another work has demonstrated that the use of amine-terminated ionic liquid (IL-NH 2 ) functionalized carbon nanotubes (CNTs) could improve the Flavin based interfacial electron transfer 23 . Although nitrogen-doped carbon materials as electrode materials have shown improved performance of MFCs, the doping process has never been rationally tailored to match the reaction sites of FMN for highly concentrated fixed reaction centers and further to convert a mediating electron transfer pathway to a direct electrochemistry process. In this work, a thermal treatment is used to delicately control nitrogen doping for atomic matching with Flavin (mediator) reaction sites, which results in strong adsorption to convert diffusive mediator molecules for anchored redox centers. The followed microbes loading on the electrode surface with completely fixed highly concentrated redox centers for a short electron pathway to eventually convert mediating electron transfer to a fast direct electrochemistry. The rational nitrogen-matching process is investigated in detail. Further, the detail atomic matching catalytic mechanism for a fast direct electrochemistry is discussed. Application of such a direct electrochemistry behavior of FMN enabled by a rationally doped nitrogen for matching atomic reaction sites is further demonstrated, and significant improvement has been achieved.",
"discussion": "Discussion Considering the concentration of self-excreted electron shuttles are often found at nM or μM concentrations 34 , 35 , while many other ions (e.g. potassium, sodium, phosphate, chloride, lactate and bicarbonate) are present in mM concentrations, the transport of electron shuttles by migration is negligible 16 . Thus, the total current density obtained by electroactive bacteria using electron shuttles is limited by the diffusion of electron mediator. More importantly, the relatively small diffusion coefficients of organic mediator molecules indicate that the diffusion is an inherently slow process especially in a porous anode due to the very low concentration gradient and extremely narrow mass transport channels. Previous works have demonstrated that electron transfer between carbon electrodes and flavins is also relatively sluggish and surfaces designed to better interact with these compounds would deliver a higher electron transfer rate. It has been reported that modification of an electrode surface through nitrogen doping enhances electron transfer between electrode and microbe and proposes that it may due to enhanced adsorption ability of electron shuttles adsorption 27 . However, fundamentally the mechanism of the interaction between the electron shuttles and electrode surface as well as its effect on direct electrochemistry has not been systematically investigated yet. The nitrogen doping effect on the interaction of nitrogen and FMN mediator was first studied by us. Figure 4c shows that the FMN-mediated electron transfer occurs at 1,5-nitrogens via a two-electron redox reaction between quinone (FMN) and hydroquinone (the oxidation product of FMN) species 11 , 36 . It is noted that the steric distance between the two electroactive sites (1,5-nitrogens) is 3.644 and 4.081 Å between the other two nitrogen (2,4-nitrogens) of FMN at the lowest energy state (Fig. 4a ). Interestingly, our measured XPS and corresponding simulation results clearly shows when the ratio of these N-Q and N-X for the nitrogen doping is three, the nitrogen distribution forms such a structure that the steric distance between N-X and N-Q is 3.660 and 4.216 Å between other two N-Q, respectively, which generates an atomic match with electroactive sites of FMN mediator. We argue that FMN can have strong interactions to preferentially adsorb on the surface with these atomically matching reaction sites and even anchor these FMN mediators on electrode surface for high adsorption of FMN. Further, it is known that FMN is a normally endogenously generated electron shuttle from bacteria, which can be easily pass through the membrane and thus should possess superior affinity to the cell membrane. We can reasonably argue that although the Flavin is just anchored on the surface of electrode, it could strongly attract the membrane of the microbe strains by the extremely high affinity to interact or adsorb or offer the most proximity to the cell membrane with very short electron transfer paths during the bacteria impregnation process for direct electrochemistry without the diffusion limit 11 . In addition, the atomic matched 2-electroactive centers of FMN with nitrogen on the electrode can allow simultaneous fast direct two-electron transfer process. This is why the highly concentrated, atomically matched FMN anode can enhance more than 21 folds high power density. This eventually convert a mediating electron transfer process to an anchored redox center (mediator)-based direct electrochemistry. Figure 4 schematically shows this atomic matching catalysis mechanism to convert a mediating electron transfer process (Fig. 4b ) to a fast direct electrochemistry electron pathway (Fig. 4a ), which gives vivid schematic explanation why the doping ratio of 3 to 1(N-Q and N-X) nitrogen can deliver the highest electrocatalytic activity and the largest catalytic current. Fig. 4 Schematic illustration of atomic matching catalysis mechanism. Schematic illustration of the enhancement mechanism of N-doped carbon structure for the direct electrochemistry of FMN. a Molecular structures and atomic distance of FMN and nitrogen sites of N-CNWs as well as the interaction between FMN and N-CNWs. b Schematic illustration of mediator-based indirect electrochemistry between FMN and non-matched nitrogen doped electrode. c Two-electron redox reaction equation of FMN and its chemical molecular structure. The two nitrogen sites highlighted by red circles is the active sites during the redox reaction. Notably, the carbonization temperature of PANI can be used to delicately tune the doping ratio of nitrogen on the electrode. There is optimal heating temperature (900 °C) to achieve a doping ratio of nitrogen (N-Q and N-X) for 3:1. The temperatures lower than 900 °C cannot produce enough N-Q to reach a ratio of 3:1, while a carbonization temperature higher than 900 °C may burn out carbon too much resulting in a less total nitrogen doping and also a lower doping ratio. The mechanism of the interaction between the electrode surface and electron shuttles is modeled via an appropriate theoretical approach. The calculations were based on spin-polarized density functional theory (DFT) using projector augmented wave (PAW) methods, as implemented in the Vienna ab initial simulation package (VASP). As shown in Supplementary Fig. 11, a plane-wave basis set with a kinetic-energy cut-off of 400 eV was used to expand the wave function of valence electrons. The adsorption energy of FMN adsorbed on different electrodes interface were shown in Supplementary Table 4 , the E ads of FMN adsorbed on N-CNWs/CC-900 (−0.260 eV) surface was smaller than the carbon surface (−0.007 eV), N-CNWs/CC-800 (−0.088 eV) and N-CNWs/CC-700 (−0.059 eV). Clearly, the results indicate that the surface of N-CNWs/CC-900 by the atomic match with the reaction sites of FMN was more favorable for Flavin adsorption, which is in good agreement with our experimental results discussed above. DFT calculation for N-CNWs/CC-1000 was also performed. Interestingly, the ratio of Quaternary N and Oxidized N of N-CNWs/CC-1000 is similar with N-CNWs/CC-900 according to the XPS configuration results of N-doping (Supplementary Table 2 ), the modeling and theoretical calculation results is also the same, indicating both can enable the atomic match. However, the matched atomic active sites after 1000 °C carbonization should be less than 900 °C carbonization due to its lower electrocalytic activity even with a same loss ratio of Quaternary to Oxidized N indicated by XPS results. The measured electrochemical active surface area of the former is smaller than the latter, which strongly supports the discussion above. To further confirm whether the atomic matching catalysis mechanism has a universal significance, the electrochemical behaviors of different electrode in phosphate buffer containing 2 μM riboflavin (RF) were investigated. RF is another Flavin mediator secreted from Shewanella spp , and its electroactive sites are similar with FMN (Fig. 4c and Supplementary Fig. 12a ). Supplementary Fig. 10b–d illustrate that the N-CNWs/CC-900 electrode also has the highest redox peak current and the lowest interfacial charge transfer resistance in RF solution. Obviously, the atomic matching catalysis enhancement mechanism of FMN is also applicable for RF. In conclusion, a thermal treatment is used to delicately tune nitrogen doping for atomic matching with the diffusive Flavin (mediator) reaction sites resulting to convert diffusive mediator molecules for anchored redox centers, and the followed microbes loading on the electrode surface with fully fixed atomic matched redox centers create an environment for a short electron pathway leading to fast direct electrochemistry. This eventually converts mediating electron transfer pathway to a fast direct electrochemistry by an atomic matching mechanism. The superior atomic match not only provides strong adsorption of flavins as fixed reaction centers but also enabling simultaneously 2-sites direct electrochemistry pathway for extremely higher power density, delivering a maximum power output of 2102 mW m −2 , which is more than 21 times higher than that of conventional non-matched nitrogen doped MFC anode. Considering the facile synthetic process, the long-term stability in nanostructures and surface properties and the superior electro catalytic performance of the N-CNWs/CC-900, this work demonstrates a feasible approach to turns a diffusive redox species-mediated electron transfer into a direct electrochemistry process, while exploring a new fundamental insights of direct electrochemistry, which can go through an atomic matching mechanism to realize fast and efficient direct electrochemistry."
} | 3,753 |
24375685 | null | s2 | 2,378 | {
"abstract": "Smooth, durable, ultrathin antifouling layers are deposited onto commercial reverse osmosis membranes without damaging them and they exhibit a fouling reduction. A new synergistic approach to antifouling, by coupling surface modification and drinking-water-level chlorination is enabled by the films' unique resistance against chlorine degradation. This approach substantially enhances longer-term fouling resistance compared with surface modification or chlorination alone, and can reduce freshwater production cost and its collateral toxicity to marine biota."
} | 140 |
36314978 | PMC9769858 | pmc | 2,379 | {
"abstract": "ABSTRACT Plant lignin is regarded as an important source for soil humic substances (HSs). Nonetheless, it remains unclear whether microbial metabolism on lignin is related to the genesis of unique HS biological activities (e.g., direct plant stimulation). Here, selected white-rot fungi (i.e., Ganoderma lucidum and Irpex lacteus ) and plant litter- or mountain soil-derived microbial consortia were exploited to structurally modify lignin, followed by assessing the plant-stimulatory activity of the lignin-derived products. Parts solubilized by microbial metabolism on lignin were proven to exhibit organic moieties of phenol, carboxylic acid, and aliphatic groups and the enhancement of chromogenic features (i.e., absorbance at 450 nm), total phenolic contents, and radical-scavenging capacities with the cultivation times. In addition, high-resolution mass spectrometry revealed the shift of lignin-like molecules toward those showing either more molar oxygen-to-carbon or more hydrogen-to-carbon ratios. These results support the findings that the microbes involved, solubilize lignin by fragmentation, oxygenation, and/or benzene ring opening. This notion was also substantiated by the detection of related exoenzymes (i.e., peroxidases, copper radical oxidases, and hydrolases) in the selected fungal cultures, while the consortia treated with antibacterial agents showed that the fungal community is a sufficient condition to induce the lignin biotransformation. Major families of fungi (e.g., Nectriaceae , Hypocreaceae , and Saccharomycodaceae ) and bacteria (e.g., Burkholderiaceae ) were identified in the lignin-enriched cultures. All the microbially solubilized lignin products were likely to stimulate plant root elongation in the order selected white-rot fungi > microbial consortia > antibacterial agent-treated microbial consortia. Overall, this study supports the idea that microbial transformation of lignin can contribute to the formation of biologically active organic matter. IMPORTANCE Structurally stable humic substances (HSs) in soils are tightly associated with soil fertility, and it is thus important to understand how soil HSs are naturally formed. It is believed that microbial metabolism on plant matter contributes to natural humification, but detailed microbial species and their metabolisms inducing humic functionality (e.g., direct plant stimulation) need to be further investigated. Our findings clearly support that microbial metabolites of lignin could contribute to the formation of biologically active humus. This research direction appears to be meaningful not only for figuring out the natural processes, but also for confirming natural microbial resources useful for artificial humification that can be linked to the development of high-quality soil amendments.",
"introduction": "INTRODUCTION Humic substances (HSs) are chromogenic polymers whose small components are supramolecularly assembled ( 1 , 2 ). Depending on molecular weights and water solubility, HSs are classified into fulvic acids, humic acids, and humins ( 2 ). Although humic structures are very complicated due to the compositional heterogeneity and the structural irregularity, it has been demonstrated that the presence of HSs leads to enhance soil fertility ( 3 – 5 ). The potential applicability of HSs as a soil conditioner was revealed by confirming their capacity to aggregate soil particles, thereby stabilizing soil structures ( 6 ). Another HS beneficial action on crops derives from their direct plant stimulation, wherein some humic components penetrate plant root interiors, followed by the induction of several gene and protein expressions of plants ( 7 – 9 ). These expression changes tend to coincide with different plant phenotypes. For instance, plant seed germination is accelerated in the presence of HSs, and this could be ascribed to the activation of auxin-like plant hormone pathways ( 10 ). Salt-induced abiotic stresses to plants are also mitigated with HSs, wherein the delayed degradation of plant high-affinity K+ transporter 1 is induced ( 9 ). Low-molecular-weight components of HSs appear to be involved in the direct plant stimulation owing to their high water solubility and plant root penetration capacity ( 2 ). The origin of HSs in the environments remains to be fully described, but structural evidence to underpin similarities between HSs and plant lignin has been continuously reported. Lignin-derived organic moieties (e.g., phenolic and carboxylic acid structures) are indeed widespread in HSs ( 2 , 11 ). The inherent recalcitrance of lignin combined with mineral association contributes to the supply of stabilized soil organic matter ( 12 ). Interestingly, Jiang et al. proved that lignin degradation could be accompanied by HS formation during composting ( 13 ). In addition, the positive correlation between HS production and populations of laccase-producing microbes is in accordance with the fact that microbial lignin metabolism results in the release of metabolic end-products that are recalcitrant and thus persistent in soils. It is also noticeable that new organic functional groups that are not found in lignin, but are found in HSs, can be formed by microbial metabolism of lignin ( 14 ). The contribution of plant lignin to soil humification is presumably significant, but it is still questionable whether microbial metabolism of lignin is essential for displaying plant-stimulatory activities of HSs. Microbial metabolites of lignin have been investigated, and key enzymatic reactions are suggested ( 15 – 17 ), but to our best knowledge, it appears that no clear connection between microbial metabolism-driven lignin structural modifications and plant-stimulatory activities has so far been reported. Since direct plant stimulation capacity is one of the main features of soil HSs, the connection must be proven for fully rationalizing the notion that lignin is the origin of HSs. Moreover, structural complexities of HSs make it hard to determine crucial structural changes of lignin resulting in humification. In this regard, it would be more acceptable to evaluate unique HS functionalities as a criterion of lignin humification. White-rot fungi are renowned for degrading lignin by employing ligninolytic exoenzymes ( 18 , 19 ), but bacteria incapable of direct lignin metabolism would likely be involved in the transformation processes by using fungal lignin metabolites as substrates. In other words, microbial consortia would play an important role in transforming lignin architecture and, in turn, humification in soils. In fact, the increase of bacterial population along with dead wood decomposition was experimentally identified ( 20 ). Depolymerized lignin products are also found to be metabolized with bacterial species ( 21 , 22 ). Thus, which types of environmental microbes are necessary to facilitate transformation of lignin toward biologically functional HSs needs to be evaluated. The aim of this study was to identify the essential microbial activities needed for biologically functional humification of plant lignin, thus elucidating the causal factors of HS plant-stimulatory properties. Selected white-rot fungi (i.e., Ganoderma lucidum and Irpex lacteus ) and microbial consortia derived from plant litters and mountain soils were recruited to metabolize plant lignin. A potato dextrose broth mixed with lignin was used to proliferate the selected fungi and to enrich the microbial consortia, and in the enrichment of environmental microbes, well-known antibiotics (i.e., erythromycin and chloramphenicol) were also treated to hamper bacterial proliferation and metabolism. Microbial community analyses before and after the enrichment were conducted based on next-generation sequences (NGSs) of DNAs extracted from the environmental samples (i.e., plant litters and mountain soils) and the enrichment cultures. After the cultivations, solubilized lignin metabolites were harvested in the supernatants, followed by analyses of their polymeric sizes, radical-scavenging capacity, chromogenic features, mass spectrometry, protein quantification, and total phenolic contents. Control lignin obtained from the same broths but not inoculated was also used for the comparison. In the case of the selected fungal cultures, additional analyses of exoenzyme profiling and solid-state 13 C nuclear magnetic resonance (NMR) were further conducted. Finally, lettuce growth was monitored in the presence of the lignin variants and compared with that of the control lignin, thereby assessing the enhancement of crop-stimulatory activity by microbial lignin metabolism. Leonardite humic acids used for agronomical purposes were also used for the comparison.",
"discussion": "RESULTS AND DISCUSSION Lignin-derived products by selected white-rot fungi. Brown colors in the supernatant parts of the liquid cultures of selected white-rot fungi (i.e., G. lucidum and I. lacteus ) were observed with a naked eye, while relatively pale brown colors were seen in those of the noninoculated control cultures (data not shown), suggesting that more lignin-derived components are dissolved with fungal metabolism. Chromogenic properties of the supernatant parts were quantitatively assessed by measuring the extent of visible light absorbance, and as shown in Fig. 1A , the absorbance at 450 nm per unit solid content of each supernatant significantly increased with increasing incubation times. This result could be ascribed to the higher solubilization of brown-colored lignin with fungal metabolisms. To confirm whether any chromogenic materials are synthesized by the fungi without lignin, the supernatants from fungi alone were also compared at 450 nm. As shown in Fig. S1 in the supplemental material, the absorbance intensity was much less than for fungi in combination with lignin, indicating that microbial lignin solubilization is the main factor to increase the absorbance intensity. FIG 1 (A) Visible light absorbance (450 nm); (B) Bradford assay-based protein quantification; (C) ABTS decolorization capacities; (D) total phenolic contents of solubilized products in noninoculated control and fungal cultures. Average and standard deviation ( n = 2) are shown. Another experimental set ( n = 3) showed the similar results. To further explore this possibility, 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) radical decolorization capacities and total phenolic contents of the supernatants were also evaluated, because the presence of products derived from lignin exhibiting polyphenolic moieties would likely be positively related to the two kinds of properties ( 23 ). As seen in Fig. 1 , similar patterns with the visible absorbance were recorded in both ABTS radical decolorization and total phenolic contents, suggesting that lignin-derived products are accumulated in the supernatants with the fungal cultivations. Then we also quantified exoenzymes of the fungi by using a Bradford assay to confirm the possible involvement of fungal metabolisms in the lignin solubilization. In general, the quantity of enzymes secreted was in line with the visible absorbance with incubation times ( Fig. 1 ). To profile the detailed exoenzyme functions, we then analyzed proteomics by separating proteins from the supernatants. As listed in Tables 1 and 2 , nonspecific oxidases such as peroxidase and Mn-dependent peroxidase were detected, suggesting that lignin is oxidized by these enzymes, thus enhancing their solubility by increasing oxygen-based functional groups and decreasing their molecular weights. In fact, it has been considered that these peroxidases are crucial for lignin biodegradation of white-rot fungi ( 15 ). Copper radical oxidases mainly capable of oxidizing alcohol groups were also detected in the both white-rot fungi employed. As Daou et al. suggested that this group of enzymes plays an important role in lignin structural changes by recognizing lignin phenolics as their substrates ( 24 ), they are likely to induce lignin solubilization by reacting with lignin phenolics. In addition, the catalytic action of copper radical oxidases results in the production of hydrogen peroxide, which is in turn used as an electron acceptor of lignin-degrading peroxidases ( 25 ). Other protein- and sugar-related enzymes identified would be due to the cosupply of potato containing proteins and carbohydrates to the fungal cultures ( 26 , 27 ). TABLE 1 Proteomic analysis of exoenzymes from I. lacteus fungal cultures a Accession no. Proteins identified Protein content (mol %) \n A0A0C9W4N5 \n Unplaced genomic scaffold scaffold_28, whole-genome shotgun sequence 0.23 ± 0.07 \n Q8LW55 \n Ribonuclease T2 3.2 ± 1.55 \n B0CNY1 \n Predicted protein 0.67 ± 0.27 \n P17576 \n Polyporopepsin 9.37 ± 4.88 \n P87212 \n Peroxidase 7.75 ± 5.23 \n A0A0P0I676 \n Mitochondrial choline dehydrogenase 5.03 ± 1.39 \n V5ND37 \n Metallopeptidase 1 0.59 ± 0.24 \n A0A0P0HVK7 \n Manganese peroxidase 3 (fragment) 3.97 ± 1.68 \n A0A0P0HPB0 \n Manganese peroxidase 3 6.24 ± 3.86 \n UPI000440B7AC \n Hypothetical protein STEHIDRAFT_156755 0.87 ± 0.38 \n UPI0004407FDA \n Hypothetical protein FOMMEDRAFT_141224 0.5 ± 0.37 \n A0A060SCW9 \n Glycoside hydrolase family 74/carbohydrate-binding module family 1 protein 0.23 ± 0.07 \n R7RWJ3 \n Glycoside hydrolase family 71 protein 0.47 ± 0.15 \n A0A0C3C4R3 \n Glycoside hydrolase family 3 protein 0.8 ± 0.06 \n S7S113 \n Glutaminase GtaA 0.28 ± 0.14 \n UPI00046216D2 \n Glucoamylase G2 0.63 ± 0.22 \n UPI0004415A45 \n Glucoamylase 0.93 ± 0.65 \n Q75NB5 \n Glucanase 4.15 ± 1.82 \n Q9Y724 \n Glucanase 0.98 ± 0.1 \n U6NLC6 \n Glucanase 0.44 ± 0.19 \n UPI000441823B \n Family S53 protease 0.9 ± 0.08 \n UPI0004416B61 \n Endopeptidase 1.67 ± 0.47 \n Q5W7K4 \n Endoglucanase 1.48 ± 0.21 \n UPI0004622D89 \n DUF1793-domain-containing protein 0.4 ± 0.03 \n Q0ZKA4 \n Copper radical oxidase 0.29 ± 0.01 \n A0A0P0HFU3 \n Choline dehydrogenase 1.45 ± 0.17 \n A0A0P0I834 \n Choline dehydrogenase 0.31 ± 0.1 \n A0A060SZQ9 \n Carboxypeptidase 0.44 ± 0.19 \n M2R432 \n Carboxypeptidase 0.4 ± 0.03 \n A0A0C3NIQ6 \n Carboxylic ester hydrolase 0.75 ± 0.29 \n R7SY09 \n Carboxylic ester hydrolase 0.36 ± 0.17 \n G3XKT3 \n Aspartic protease 1.79 ± 1.51 \n UPI000440F5C0 \n Aspartic peptidase A1 0.52 ± 0.22 \n W4KAT8 \n Aspartic peptidase 0.39 ± 0.12 \n R7SSF9 \n Alpha-1,2-mannosidase 0.75 ± 0.29 \n UPI000444A5EF \n Alpha/beta-hydrolase 0.36 ± 0.17 a The mol % (i.e., each emPAI/sum of emPAI × 100) and standard deviation ( n = 2) are shown. Proteins detected in both duplicates are listed based on UniRef100 database ( www.uniprot.org ). TABLE 2 Proteomic analysis of exoenzymes from G. lucidum fungal cultures a Accession no. Proteins identified Protein content (mol %) \n A0A0H2SDZ6 \n Tripeptidyl peptidase A 0.59 ± 0.33 \n Q8LW55 \n Ribonuclease T2 2.2 ± 0.27 \n P87212 \n Peroxidase 0.94 ± 0.08 \n A0A0P0I676 \n Mitochondrial choline dehydrogenase 4.73 ± 0.55 \n A0A0P0HVJ9 \n Manganese peroxidase 3 (fragment) 5.87 ± 2.8 \n UPI0004407FDA \n Hypothetical protein FOMMEDRAFT_141224 1.44 ± 0.16 \n A0A067QAQ4 \n Glycoside hydrolase family 5 protein 2.08 ± 0.68 \n UPI0004415A45 \n Glucoamylase 0.41 ± 0.07 \n UPI000441823B \n Family S53 protease 1.53 ± 0.22 \n R7SYU3 \n Endopeptidase 3.92 ± 0.53 \n Q0ZKA4 \n Copper radical oxidase 1.3 ± 0.42 \n A0A060SZQ9 \n Carboxypeptidase 0.5 ± 0.07 \n M2R432 \n Carboxypeptidase 1.03 ± 0.09 \n A0A0C9V748 \n Carboxypeptidase 2.21 ± 0.62 \n A0A0C3NIQ6 \n Carboxylic ester hydrolase 0.63 ± 0.24 \n G3XKT3 \n Aspartic protease 1.17 ± 0.15 \n UPI000440F5C0 \n Aspartic peptidase A1 1.17 ± 0.15 \n R7SSF9 \n Alpha-1,2-mannosidase 1.8 ± 0.17 a The mol % (i.e., each emPAI/sum of emPAI × 100) and standard deviation ( n = 2) are shown. Proteins detected in both duplicates are listed based on UniRef100 database ( www.uniprot.org ). Structural elucidation of lignin-derived products by selected fungi. Size exclusion chromatography (SEC) and ultrahigh resolution Fourier-transform ion cyclotron resonance (FT-ICR) mass spectrometry were used to evaluate structural changes of lignin in more detail as shown in Fig. 2 and 3 , respectively. Several sharp peaks showing fewer molecular weight (MWs) than the major peak corresponding to approximately a molecular weight of 8,000 were evident in SEC of the control (i.e., the lignin-containing broths that were not inoculated with the fungi), while the fungal cultures displayed one major peak near 8,000 MW that was dragged to lower MWs and a new peak showing a higher MW (i.e., 34,101 MW). By the comparison with SEC of the potato broth alone (data not shown), we concluded that peaks of the control ranging from 1,406 to 899 MW derived from lignin alone. Fungal metabolic pathways presumably oxidize them, thus allowing for their disappearance. The appearance of a new peak (i.e., 34,101 MW) in the fungal cultures would be attributable to solubilization of solid lignin by fungal metabolism. The overall absorbance at 280 nm in the SECs of the fungal cultures was higher than that of the control, indicating that more UV-absorbing products are solubilized with fungal metabolism. Given the multiple aromaticity of lignin capable of absorbing UV light ranges, the products would be from the solubilization of solid lignin. FIG 2 Size-exclusion chromatography of solubilized products in noninoculated control and fungal cultures. The green bracket indicates newly formed peaks with fungal metabolisms, and the black one indicates lignin-related peaks of the control. FIG 3 Van Krevelen plots showing the distribution of chemical classes based on the atomic H/C and O/C ratios of the assigned molecular formulas in supernatant components of noninoculated control and fungal cultures. The assignment of putative molecular formulas based on the mass spectrometric measurements visualized more significant changes in lignin before and after the metabolism. Molecules both composing carbon (C), hydrogen (H), and oxygen (O) and fittable to the lignin class of molar H/C and O/C ratios are highly likely to be lignin components ( Fig. 3 ). The molecules in the control presumably derive from the soluble part of lignin initially supplied. Interestingly, a significant portion of them was molecularly transformed with fungal metabolism, thus exhibiting higher H/C ratios (circled regions in Fig. 3 ). This would be caused by the benzene ring opening, thus transforming aromatic groups into aliphatic groups. It has been similarly suggested that fungi are able to open benzene rings of lignin ( 14 ). Quinone-involved redox cycling by white-rot fungi is linked to the production of hydroxyl radicals capable of oxidizing aromatic compounds, which could be further involved in the ring opening ( 28 ). The peroxidases detected ( Tables 1 and 2 ) presumably act as phenol oxidizers, thus forming quinone structure. It is also noticeable that molecules both belonging to the protein region and showing the composition of carbon (C), hydrogen (H), oxygen (O), and nitrogen (N) increased with fungal metabolism, which could be attributable to the presence of exoenzymes. Solid-state 13 C NMR revealed the structural features driven by fungal metabolism ( Fig. 4 ). Major peaks shown in the control are very similar to those of starch, which could be ascribed to components of the potato broth ( 29 ). However, the small peaks near 174 and 147 ppm that may correspond to carboxylic and phenol groups, respectively ( 30 ), suggest that some lignin-derived components are present in the supernatant of the control, which is consistent with the result of the SEC and mass spectrometry results ( Fig. 2 and 3 ). Interestingly, peaks near 32 ppm and corresponding to the alkyl group were newly formed with fungal metabolism ( Fig. 4 ), and given the shift of the compositional molecules from lignin toward higher H/C ratios shown in Fig. 3 , this shift would probably result from the benzene ring opening. Other interesting peaks enhanced with fungal metabolism are near 174 and 147 ppm (i.e., carboxylic and phenol groups, respectively). The intensity increases in these two peaks coincided with increasing peaks near 128 ppm, corresponding to aromatic rings, indicating that lignin-derived products accumulate in the supernatants of fungal cultures. Oxidative reactions by fungal exoenzymes may leave oxygen-based functional groups that are further involved in increasing hydrophilicity of lignin products. This interpretation is in line with the detection of carboxylic ester hydrolase in the supernatants of the fungal cultures ( Tables 1 and 2 ) and a previous report showing the abundance of ester bonds by the p -coumaric acid moiety in plant lignin ( 31 ). FIG 4 Solid-state 13 C NMR spectra of solubilized products in noninoculated control and fungal cultures. Red boxes indicate peaks associated with lignin-related functional groups. To further detail small lignin-derived phenolics in the supernatants, gas chromatography-mass spectrometry (GC-MS) was conducted (Fig. S2). Interestingly, one-benzene-ring-containing lignin-related aromatics such as vanillin and benzene acetic acid were not detected with the microbial cultivation. It would be related to phenol metabolism of white-rot fungi ( 32 ). Given that the GC column used is not able to detect oligomeric compounds, the chromatograms also indicate that the observed visible light colors with fungal growths could be attributable to the release of polyphenolic oligomers from insoluble lignin. Plant stimulation by lignin-derived products of selected fungi. We then tested whether the lignin solubilized by fungal metabolism is effective to stimulate plants. As shown in Fig. 5 , lettuce root morphologies were significantly modified, while the control samples which were not affected by fungal metabolism exhibited no difference from the lettuces treated with nothing. In particular, the axial roots tended to grow more rapidly with the solubilized lignin products by fungal metabolism. Root growth-stimulatory actions of HSs have been well demonstrated ( 2 , 10 ), and auxin-related plant hormone pathways appear to be involved in the stimulation ( 10 ). Structurally, phenolic and carboxylic groups in HSs play an important role in the stimulations ( 10 ) and seem to be present in the current fungal lignin-derived products. We then also compared the stimulatory activities with those of commercial humic acids that are used for agronomical purposes. There was a noticeable difference between the untreated and the humic-treated plant root growths. In addition, the magnitude of the root stimulation was almost comparable to that of lignin plus fungus inoculation (Fig. S3). Overall, these results indicate that fungal metabolism on plant lignin is involved in the production of biologically active organics. FIG 5 Root elongation of lettuce cultivated in MS agar mixed with solubilized products of noninoculated control and fungal cultures. Average and standard error ( n = 14) are shown, and the data were statistically analyzed using one-way analysis of variance (ANOVA) and Duncan’s test ( P < 0.05). Use of microbial consortia for lignin metabolism. Beyond the use of isolated white-rot fungi, we endeavored to study the role of environmental microbes for this phenomenon. The same potato broths containing lignin were used to enrich environmental microbes of plant litters and mountain soils. We also used antibacterial drugs (i.e., erythromycin and chloramphenicol) to confirm whether bacterial metabolism could aid in the functional humification. As shown in Fig. 6 , very similar properties of all the supernatants with those of the selected fungal cultures (i.e., enhancement of chromogenic, radical-scavenging capacities and total phenolic contents) were observed with the enrichment of environmental microbes. These results imply that some of the microbes enable lignin metabolism, wherein solid lignin components are solubilized. In addition, bacterial metabolism was found not to be obligatory, considering the inability of the antibacterial drugs to hamper the lignin solubilization. It thus appeared that the fungal community is a sufficient condition to induce such structural changes in lignin ( Fig. 6 ). FIG 6 (A) Size-exclusion chromatography wherein the green bracket indicates newly formed peaks with microbial metabolisms, and the black one indicates lignin-related peaks in the control. (B) ABTS decolorization capacities, (C) visible light absorbance (450 nm), and (D) total phenolic contents of solubilized products of noninoculated control and microbial consortia. AB, antibacterial agents. Average and standard deviation ( n = 2) are shown. Another experimental set ( n = 3) for panels B, C, and D showed similar results. Molecular composition assignments by the ultrahigh-resolution mass spectrometry clearly visualized the lignin transformation that is similar to that of the selected fungi. The molecules composing carbon (C), hydrogen (H), and oxygen (O) and also belonging to the lignin class shifted toward the range having higher H/C ratios (circled regions in Fig. 7 ). In addition, the movement toward higher O/C ratios was significantly observed, except for the litter treated with antibiotics (rectangular regions in Fig. 7 ). These changes could be caused by oxygenation of lignin and benzene ring opening. The presence of antibiotics hardly induces the changes in the distribution of molecules, and this raises the question of whether the enrichment of fungal species overwhelms bacterial metabolism in the given nutrient broth. FIG 7 Van Krevelen plots showing the distribution of chemical classes based on the atomic H/C and O/C ratios of the assigned molecular formulas in supernatant components of noninoculated control and microbial consortia. AB, antibacterial agents. The microbial enrichment patterns shown are based on next-generation sequencing (NGS)-based microbial community analyses whose rarefaction curves were saturated (Fig. S4), indicating that the communities profiled are representative of the cultures and the real environments. The significant decreases in the microbial diversity were recorded ( Fig. 8 , Fig. S5 and S6), meaning that carbon sources of the potato broth and lignin act as an enrichment pressure to help selected microbes flourish. Interestingly, different microbial families between plant litters and soils were enriched, which could be attributable to different microbial compositions of the inoculants. Nonetheless, Burkholderiaceae in bacteria and Nectriaceae , Hypocreaceae , and Saccharomycodaceae in fungi were commonly detected, suggesting that these could be important for lignin metabolisms. Indeed, Burkholderiaceae is renowned for aromatic degradation ( 33 ), and Nectriaceae and Hypocreaceae are also identified during natural wood decay ( 34 , 35 ). Their detailed roles in lignin metabolism need to be further investigated by using culture-dependent techniques. It is also remarkable that the treatment of the antibiotics induced very similar fungal communities, irrespective of the kinds of inoculants ( Fig. 8 ). FIG 8 Relative abundance of fungal communities of environmental samples used for the inoculation (i.e., plant litters and mountain soils) and the enriched cultures at the family level. Taxa with an abundance of <1% are included in “other.” AB, antibacterial agents; environmental, real environmental samples. Finally, it was confirmed that the root growth is enhanced with the enrichment of environmental microbes. As shown in Fig. 9 , the root lengths affected by lignin-derived products of the microbial consortia tended to increase compared with the control (i.e., lignin alone) and no treatment, although the extent of the increase appeared to be less than that by the selected fungal broths ( Fig. 5 ). These results suggest that plant litters and mountain soils contain microbial species enabling the production of lignin metabolites that could potentially lead to the biologically functional humification of lignin. On the other hand, the treatments of the antibiotics appeared to result in less root-stimulatory activity ( Fig. 9 ). This would be indicative of better lignin transformation efficiency by the cooperation between fungal and bacterial metabolism, which is supported by the fact that bacteria are coproliferated when wood is naturally decayed ( 20 ). HSs are known to stimulate plants in a various ways ( 2 ). Depending on the experimental situations, changes in the plant physiology of seed germination, abiotic stress responses, and aboveground biomass growth could be induced. Hence, further studies to assess such changes with the microbially transformed lignin products need to be performed. FIG 9 Root elongation of lettuce cultivated in MS agar mixed with solubilized products (0.1 g L −1 ) of noninoculated control and microbial consortia. Average and standard error ( n = 18) are shown, and the data were statistically analyzed using one-way ANOVA and Duncan’s test ( P < 0.05). AB, antibacterial agents. Here, we confirmed that selected white-rot fungi and microbial consortia derived from plant litters and mountain soils are able to not only solubilize lignin but also endow lignin with plant-stimulatory activity (i.e., root elongation). The biologically functional lignin humification by white-rot fungi coincided with the secretion of ligninolytic enzymes such as peroxidase and Mn-dependent peroxidase, copper radical oxidases, and carboxylic ester hydrolase. The litters and soils were found to contain microbial species enabling lignin transformation connected to plant-stimulatory capacity, wherein bacterial metabolisms appeared not to be obligatory, and Nectriaceae , Hypocreaceae , and Saccharomycodaceae in fungi and Burkholderiaceae in bacteria were mainly enriched. Given that most studies linking microbial lignin transformation to humification do not consider how biological activities of HSs originated ( 14 , 36 , 37 ), this study provides new knowledge that microbial metabolism endows lignin with plant-stimulatory activity."
} | 7,522 |
32874778 | PMC7439960 | pmc | 2,380 | {
"abstract": "Corals are associated with diverse microbial assemblages; however, the spatial-temporal dynamics of intra-species microbial interactions are poorly understood. The coral-associated microbial community varies substantially between tissue and mucus microhabitats; however, the factors controlling the occurrence, abundance, and distribution of microbial taxa over time have rarely been explored for different coral compartments simultaneously. Here, we test (1) differentiation in microbiome diversity and composition between coral compartments (surface mucus and tissue) of two Acropora hosts ( A. tenuis and A. millepora ) common along inshore reefs of the Great Barrier Reef, as well as (2) the potential linkage between shifts in individual coral microbiome families and underlying host and environmental parameters. Amplicon based 16S ribosomal RNA gene sequencing of 136 samples collected over 14 months, revealed significant differences in bacterial richness, diversity and community structure among mucus, tissue and the surrounding seawater. Seawater samples were dominated by members of the Synechococcaceae and Pelagibacteraceae bacterial families. The mucus microbiome of Acropora spp. was dominated by members of Flavobacteriaceae, Synechococcaceae and Rhodobacteraceae and the tissue was dominated by Endozoicimonaceae. Mucus microbiome in both Acropora species was primarily correlated with seawater parameters including levels of chlorophyll a, ammonium, particulate organic carbon and the sum of nitrate and nitrite. In contrast, the correlation of the tissue microbiome to the measured environmental (i.e., seawater parameters) and host health physiological factors differed between host species, suggesting host-specific modulation of the tissue-associated microbiome to intrinsic and extrinsic factors. Furthermore, the correlation between individual coral microbiome members and environmental factors provides novel insights into coral microbiome-by-environment dynamics and hence has potential implications for current reef restoration and management efforts (e.g. microbial monitoring and observatory programs).",
"conclusion": "Conclusions This study highlights that microbiomes inhabiting different physical microniches within the coral holobiont differ in their linkage between host and environmental factors. Microbiomes of Acropora spp. differed significantly among host compartments (surface mucus layer and tissue) and species ( A. tenuis and A. millepora ). Seawater parameters had the greatest influence on the mucus microbiome in both species whereas the tissue microbiomes showed differential patterns to environmental/host-physiological parameters, suggesting host-specific modulation of the tissue microbiome. While further research is needed to unequivocally define the drivers of coral microbiome variation, by investigating temporal variation in water quality and coral health measures and correlating these with microbial community dynamics across distinct host compartments in closely related species, this study has identified several intrinsic and extrinsic factors that contribute to microbiome composition in corals.",
"introduction": "Introduction Coral microbiomes include the well-characterized endosymbiotic dinoflagellates of the family Symbiodiniaceae , and a vast diversity of bacteria and archaea ( Bourne, Morrow & Webster, 2016 ; Frade et al., 2016a ; Rohwer et al., 2002 ). The microbiome has a fundamental role in the health and stability of the coral holobiont; it recycles nutrients, removes waste products and defends against pathogens ( Lema, Willis & Bourne, 2012 ; Morris et al., 2011 ; Rädecker et al., 2015 ; Rosado et al., 2019 ). The coral microbiome is influenced by a variety of intrinsic and extrinsic factors. Coral microbiomes are host species-specific and were thought to remain relatively stable over space and time ( Frias-Lopez et al., 2002 ; Rohwer et al., 2002 ). However, recent studies have proposed that spatial–temporal factors such as environmental parameters ( Chen et al., 2011 ), depth ( Glasl et al., 2017 ), geography ( Hong et al., 2009 ; Littman et al., 2009 ), seasonality ( Ceh, Van Keulen & Bourne, 2011 ; Chen et al., 2011 ; Hong et al., 2009 ; Koren & Rosenberg, 2006 ), coastal pollution ( Klaus et al., 2007 ), and the physiological status of the host ( Grottoli et al., 2018 ; Littman, Willis & Bourne, 2009 ) can also influence the occurrence and relative abundance of microbial taxa. For instance, Li et al. (2015) reported a dynamic relationship between the community structure of coral-associated bacteria and the seasonal variation in environmental parameters such as dissolved oxygen and rainfall. Glasl et al. (2019a) showed that although host-associated microbiomes were five-times less responsive to the environment compared to the seawater microbiome, they were still affected by environmental factors (e.g., temperature, turbidity, and nutrient concentration). The coral provides different microhabitats for its microbial associates, including the surface mucus layer, coral tissue, skeleton and gastrovascular cavity, each differing in microbial richness, diversity and community structure, often assessed through alpha- and beta-diversity metrics ( Agostini et al., 2012 ; Engelen et al., 2018 ; Pollock et al., 2018 ; Sweet, Croquer & Bythell, 2011 ). Each microhabitat has a unique set of biochemical features and harbors a specific microbial community ( Engelen et al., 2018 ; Pollock et al., 2018 ; Sweet, Croquer & Bythell, 2011 ). Hence, revealing microhabitat-specific host-microbiome associations and their specific sensitivities to environmental fluctuations is crucial to our understanding of coral holobionts. For example, the coral surface mucus layer is a polysaccharide-protein-lipid complex that provides an interface between the coral epithelium and the surrounding seawater ( Brown & Bythell, 2005 ). Here microbes take advantage of a nutrient-rich medium and particular microbiome members found in the coral mucus overlap with both the tissue and the seawater microbial communities ( Bourne & Munn, 2005 ; Brown & Bythell, 2005 ; Glasl, Herndl & Frade, 2016 ; Sweet, Croquer & Bythell, 2011 ). In contrast to the extracellular polymeric nature of the surface mucus layer, the coral tissue consists of two distinct layers (epidermis and gastrodermis) and a connective-tissue layer, the mesoglea ( Muller-Parker, D’Elia & Cook, 2015 ). The coral tissue harbors photosymbiotic dinoflagellates (family Symbiodiniaceae), that can provide up to 100% of energy required by their coral host ( Muller-Parker, D’Elia & Cook, 2015 ). The Symbiodiniaceae community has been shown to vary in tandem with the bacterial community in early life stages of corals ( Quigley et al., 2019 ) and this may be caused by the release of complex organic molecules such as the organosulfur compound dimethylsulfoniopropionate (DMSP; Bourne et al., 2013 ; Frade et al., 2016b ). The coral tissue microbiome is mostly represented by bacteria belonging to the phyla Proteobacteria and Actinobacteria. For example, the gammaproteobacterial Endozoicomonas are abundant in the coral’s endodermal tissue and are often considered ‘true’ coral symbionts ( Bayer et al., 2013 ; Glasl et al., 2019b ; Neave et al., 2016 ; Neave et al., 2017 ). When compared to the surface mucus layer, the microbial community in the tissue is significantly less dense and diverse ( Bourne & Munn, 2005 ; Koren & Rosenberg, 2006 ), likely attributed to the more spatially stable and host controlled environment ( Bourne & Munn, 2005 ), although divergent evidence suggests the mucus is less diverse than the tissue ( Pollock et al., 2018 ). Furthermore, tissue-associated bacterial communities form aggregations within the coral cell layers, also referred to as coral-associated microbial aggregates (CAMAs), and are often co-localized near algal symbiont cells, highlighting potential metabolic interactions between symbionts ( Wada et al., 2019 ). In this study, we test the hypotheses that different coral compartments (surface mucus layer and tissue) of Acropora spp. harbor distinct microbial communities and that different intrinsic and extrinsic factors explain microbiome dynamics within these compartments. Furthermore, we aim to identify significant correlations of individual bacterial families associated with coral tissue and mucus with host-physiological and seawater parameters.",
"discussion": "Discussion Microbial communities associated with corals are continually exposed to fluctuations in the surrounding environment and the physiology of their host. Previous studies have demonstrated changes in the coral microbiome in response to thermal stress ( Ainsworth & Hoegh-Guldberg, 2009 ; Grottoli et al., 2018 ; Lee et al., 2015 ; Thurber et al., 2009 ), ocean acidification ( Grottoli et al., 2018 ; Thurber et al., 2009 ), organic matter enrichment ( Garren & Azam, 2012 ), bleaching events ( Bourne et al., 2008 ) and other environmental and physiological factors ( Glasl et al., 2019a ; Guppy & Bythell, 2006 ; Kelly et al., 2014 ; Li et al., 2015 ; Pollock et al., 2018 ). However, the coral microbiome is not homogenous across the animal and an improved understanding of the sensitivity of the microorganisms inhabiting each coral compartment is needed. This study highlights compositional differences in the bacterial communities associated with coral mucus and coral tissue, as well as with the surrounding seawater, findings that are largely consistent with previous studies ( Apprill, Weber & Santoro, 2016 ; Bourne & Munn, 2005 ; Engelen et al., 2018 ; Pollock et al., 2018 ; Sweet, Croquer & Bythell, 2011 ). Furthermore, the high similarity between mucus and seawater microbiomes (see Tables S1 and S2 , Figs. 2 and 3 ) and the high dissimilarity between tissue and seawater microbiomes suggests that the mucus microbial community is more strongly influenced by the external environment than the tissue community. Similar results have been reported for other coral species in the Caribbean ( Orbicella faveolata , Diploria strigosa , Montastraea cavernosa , Porites porites and Porites astreoides ), where mucus and seawater shared significantly more microbial taxa than those shared by tissue and seawater microbiomes ( Apprill, Weber & Santoro, 2016 ). Our results also support that mucus microbiomes are richer and more diverse than tissue microbiomes, which is a pattern corroborated by many previous studies ( Bourne & Munn, 2005 ; Koren & Rosenberg, 2006 ). Despite the host species-specificity of the coral microbiomes, some bacterial taxa were ubiquitously associated with a particular coral compartment. For example, Flavobacteriaceae and Synechococcaceae dominated the mucus of both species, while Endozoicimonaceae dominated the tissue microbiome of both Acropora species. However, overall microbiome composition also showed some overlap between host compartments, consistent with previous reports of overlap between the mucus and tissue microbiomes of other coral species ( Engelen et al., 2018 ; Sweet, Croquer & Bythell, 2011 ). This intersection is a natural feature of the coral holobiont as both compartments are within the same host and because the constituents of the surface mucus layer are originally produced inside the tissue ( Bythell & Wild, 2011 ). The sharing of some microbial taxa between compartments may also arise due to methodological challenges associated with retrieving samples that are exclusively mucus or coral tissue ( Sweet, Croquer & Bythell, 2011 ), and hence these methodological limitations can obscure differences between the mucus and seawater microbiomes ( Brown & Bythell, 2005 ). Explanatory factors of mucus microbiome variation We hypothesized that the coral mucus microbiome, which is in direct contact with seawater, would be primarily correlated with seawater parameters, whereas the tissue microbiome would be most affected by the physiological state of the coral host. Mucus is highly hydrated: mucocyte cells release their secretions in a condensed form which then undergo a massive swelling upon hydration, forming a visco-elastic gel ( Brown & Bythell, 2005 ). Surface mucus can therefore be influenced by the presence of nutrients dissolved in the surrounding seawater ( Tanaka, Ogawa & Miyajima, 2010 ). As expected, environmental factors (i.e., seawater parameters) were influential in shaping the mucus microbiome of both species ( A. millepora and A. tenuis ), consistent with recent studies relating changes in the mucus microbiome with environmental perturbations ( Li et al., 2015 ; Pollock et al., 2018 ). However, the extent of influence from environmental parameters (10% of variation) on the mucus microbiome was much lower than the influence of environment on the seawater microbiome (32% of variation), suggesting that other factors also play a role in modulating the mucus microbiome. For instance, the surrounding environment may interact with host physiology and together they alter the bacterial community structure of the mucus. Mucus is a nutrient-rich medium fueled by the photosynthetic activity of the Symbiodiniaceae ( Brown & Bythell, 2005 ) and therefore it is expected that some degree of variation in its chemical composition is explained by host-Symbiodiniaceae factors. For example, A. millepora and A. tenuis at the sampling site (Geoffrey Bay at Magnetic Island) associate with distinct Symbiodiniaceae ( LaJeunesse et al., 2018 ; Ulstrup & Van Oppen, 2003 ; Van Oppen et al., 2001 ). A. millepora colonies were associated with Durusdinium ( Van Oppen et al., 2001 ) whereas A. tenuis harbored Cladocopium spp. ( Ulstrup & Van Oppen, 2003 ). Links between mucus chemical composition and microbiome community structure have been proposed ( Tremblay et al., 2011 ). Physiological factors regulating the dynamics of production and release of the surface mucus layer could also contribute to regulating mucus microbial composition ( Glasl, Herndl & Frade, 2016 ). Fluctuations of NH \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{4}^{+}$\\end{document} 4 + , NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − , Chl a and POC in the surrounding seawater significantly correlated with the mucus microbiome variation in Acropora species. Li et al. (2015) and Chen et al. (2011) suggested that rainfall had a crucial effect on bacterial community variation in the coral microbiome, being mostly associated with an increase in the relative abundance of the Bacilli group ( Chen et al., 2011 ; Li et al., 2015 ). In the present study, NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − (and its collinear variables daylight, particulate nitrogen and grainsize of sediments; Table S3 ) had the greatest influence on microbiome structure, being a significant factor for both studied species. The link between rainfall and increasing nutrients (such as NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − ) is well established for inshore reefs ( Fabricius, 2005 ). In the current study, higher amounts of particulate and dissolved nutrients (but a decrease in TSS), corresponded to an increase in mucus-associated Synechococcaceae, Pirellulaceae, OCS155 and Rhodobacteraceae and a decrease in Halomonadaceae. For instance, Synechococcaceae in the mucus was highly positively correlated with NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − and negatively correlated with TSS. These findings corroborate previous work in which the abundance of free-living Synechococcus in shallow coastal waters decreased significantly under lower nutrient (especially nitrate) and higher TSS concentrations ( Uysal & Köksalan, 2006 ). Dissolved nutrients, such as nitrogen and phosphorus, can affect coral physiology and drive changes in the associated microbial community ( Shaver et al., 2017 ; Thompson et al., 2015 ). For example, organic-rich nutrients from terrestrial run-off negatively affect the health of corals and promote rapid growth of opportunistic heterotrophic bacteria (e.g., Vibrionales, Flavobacteriales and Rhodobacterales), thus affecting the overall composition of the coral microbiome ( McDevitt-Irwin et al., 2017 ; Weber et al., 2012 ). In our study, the abundance of Flavobacteriaceae and Rhodobacteraceae in the mucus of A. tenuis correlated with TSS and NH \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{4}^{+}$\\end{document} 4 + , respectively. The coral holobiont, including cyanobacteria related to Synechococcus spp. ( Lesser et al., 2004 ), can also efficiently take up inorganic nitrogen, for example, as nitrogen is required by the photosynthesis production of its Symbiodiniaceae symbionts ( Yellowlees, Rees & Leggat, 2008 ). In fact, NH \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{4}^{+}$\\end{document} 4 + can be assimilated by both coral and its Symbiodiniaceae ( Pernice et al., 2012 ), and recent work has implicated bacteria such as Vibrio and Alteromonas in the incorporation and translocation of NH \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{4}^{+}$\\end{document} 4 + into coral tissues and associated Symbiodiniaceae ( Ceh et al., 2013 ). Nitrifying members of the mucus microbiome, such as ammonium oxidizing bacteria (e.g., Pirelullaceae) and archaea, are fueled by NH \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{4}^{+}$\\end{document} 4 + ( Beman et al., 2007 ; Siboni et al., 2008 ; Yang et al., 2013 ), and NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − can be respired by nitrate reducers putatively active in coral microbiomes ( Siboni et al., 2008 ; Yang et al., 2013 ). Interestingly, Pirellulaceae abundances in the mucus of A. millepora positively correlated with concentrations of environmental NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − , the products of ammonium oxidation. These nitrogen-cycling processes mediated by microbes are highly dependent on oxygen availability, but because oxygen concentration in the mucus shows strong diel fluctuations ( Shashar, Cohen & Loya, 1993 ), it is possible that both aerobic (e.g., nitrification) and anaerobic (e.g., denitrification) processes happen within the mucus layer at different times of the day. Temporal dynamics in the coral mucus microbiome are thus likely influenced by the individual and collective metabolic capabilities of the diverse assemblage of microbes and by nutrient availability in the surrounding waters. Explanatory factors of tissue microbiome variation The statistical relation between the coral tissue microbiome and the environmental and physiological parameters differed between coral species. Whereas the tissue microbiome of A. tenuis corresponded to both environment and host physiology, A. millepora correlated only with environmental parameters. This difference may be associated to specific features of each species, through which A. millepora could modulate the internal environment and create more stable intra-tissue conditions than A. tenuis (e.g., via skeletal light modulation, host morphology and tissue thickness, sensu \n Enriquez, Mendez & Iglesias-Prieto, 2005 ). A non-mutually exclusive alternative explanation is the influence of the algal symbiont (Symbiodiniaceae) genotype associated to the host. Little (2004) investigated Symbiodiniaceae communities associated with A. millepora and A. tenuis on Magnetic Island demonstrating that the coral-algal endosymbiotic relationship in Acropora spp. is distinct between species, dynamic and flexible (corals associate with different Symbiodiniaceae types at different life stages, for example), and contributes significantly to physiological attributes of the coral holobiont. For example, different algal genotypes can affect the nutrient availability (e.g., carbon and nitrogen) in the coral holobiont ( Pernice et al., 2015 ; Bayliss et al., 2019 ). Environmental factors such as seawater temperature can also lead to temporal changes in the symbiont community ( Cooper et al., 2011 ; Howells et al., 2012 ; Rocker, Willis & Bay, 2012 ). As the microbiome is strongly associated to the coral holobiont, any disturbance in the host-Symbiodiniaceae relationship may have indirect effects on the microbial composition and its response to environmental and physiological factors. Other studies demonstrate the influence of Symbiodiniaceae on the host microbial community and also support the idea that these two components of the coral holobiont are finely tuned ( Glasl et al., 2017 ; Grottoli et al., 2018 ; Littman, Bourne & Willis, 2010 ; Littman, Willis & Bourne, 2009 ; Quigley et al., 2019 ). In the present study, Endozoicimonaceae were strongly positively correlated with the Symbiodiniaceae density in the tissue of A. tenuis and negatively correlated with NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − /NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − in A. millepora (see Fig. 5 ). These results are to some extent at odds with experimental results showing a stable dominance of Endozoicimonaceae in tissues of Pocillopora verrucosa irrespective of excess dissolved organic nitrogen and despite a bleaching phenomenon concomitant with structural changes in its Symbiodiniaceae community ( Pogoreutz et al., 2018 ). Besides the diversity of Symbiodiniaceae associated to each coral species, other factors can affect the coral and its response to environmental parameters, such as photochemical efficiency (Fv/Fm) and symbiont density ( Cunning & Baker, 2014 ; Da-Anoy, Cabaitan & Conaco, 2019 ). For instance, Da-Anoy, Cabaitan & Conaco (2019) demonstrated a greater reduction of Fv/Fm in A. tenuis in response to elevated temperatures compared to A. millepora and the temperature responses of the corals did not directly correlate with their associated Symbiodiniaceae. This suggests that other species-specific physiological factors could modulate the responses of the coral to the environment and, indirectly, influence the tissue-associated microbiome. One such factor is the way coral-associated microbial aggregates (CAMAs) are distributed throughout the tissue, which varies within populations and can vary among coral species ( Work & Aeby, 2014 ; Wada et al., 2019 ). Total suspended solids (TSS) was the only environmental parameter measured in the present study that significantly related to the tissue microbiome of both coral species. TSS can impact corals by limiting light availability for photosynthesis and decreasing Symbiodiniaceae densities, which can indirectly affect microbial communities ( Fabricius, 2005 ; Pollock et al., 2014 ). High levels of suspended solids characterize the environment of inshore reefs such as those found around Magnetic Island. The decrease in TSS is strongly associated with an increase in the abundance of tissue-associated Synechococcaceae and Cryomorphaceae. Cryomorphaceae are typical copiotrophs in the phylum Bacteroidetes and their increase in the tissue of A. millepora could relate to declines in coral holobiont health."
} | 6,855 |
29194914 | null | s2 | 2,384 | {
"abstract": "The ability to recognize self and to recognize partnering cells allows microorganisms to build social networks that perform functions beyond the capabilities of the individual. In bacteria, recognition typically involves genetic determinants that provide cell surface receptors or diffusible signalling chemicals to identify proximal cells at the molecular level that can participate in cooperative processes. Social networks also rely on discriminating mechanisms to exclude competing cells from joining and exploiting their groups. In addition to their appropriate genotypes, cell-cell recognition also requires compatible phenotypes, which vary according to environmental cues or exposures as well as stochastic processes that lead to heterogeneity and potential disharmony in the population. Understanding how bacteria identify their social partners and how they synchronize their behaviours to conduct multicellular functions is an expanding field of research. Here, we review recent progress in the field and contrast the various strategies used in recognition and behavioural networking."
} | 273 |
35960256 | PMC9805206 | pmc | 2,386 | {
"abstract": "Abstract The congruence between phylogenies of tightly associated groups of organisms (cophylogeny) reflects evolutionary links between ecologically important interactions. However, despite being a classic example of an obligate symbiosis, tests of cophylogeny between scleractinian corals and their photosynthetic algal symbionts have been hampered in the past because both corals and algae contain genetically unresolved and morphologically cryptic species. Here, we studied co‐occurring, cryptic Pocillopora species from Mo′orea, French Polynesia, that differ in their relative abundance across depth. We constructed new phylogenies of the host Pocillopora (using complete mitochondrial genomes, genomic loci, and thousands of single nucleotide polymorphisms) and their Symbiodiniaceae symbionts (using ITS2 and psbA ncr markers) and tested for cophylogeny. The analysis supported the presence of five Pocillopora species on the fore reef at Mo′orea that mostly hosted either Cladocopium latusorum or C. pacificum. Only Pocillopora species hosting C. latusorum also hosted taxa from Symbiodinium and Durusdinium . In general, the Cladocopium phylogeny mirrored the Pocillopora phylogeny. Within Cladocopium species, lineages also differed in their associations with Pocillopora haplotypes, except those showing evidence of nuclear introgression, and with depth in the two most common Pocillopora species. We also found evidence for a new Pocillopora species (haplotype 10), that has so far only been sampled from French Polynesia, that warrants formal identification. The linked phylogenies of these Pocillopora and Cladocopium species and lineages suggest that symbiont speciation is driven by niche diversification in the host, but there is still evidence for symbiont flexibility in some cases.",
"introduction": "1 INTRODUCTION For organisms that form symbiotic partnerships with microbes, the identity and composition of symbionts is often critical to the performance and function of the holobiont (the host and its symbionts) (Rohwer et al., 2002 ; Roughgarden et al., 2018 ; van Oppen & Medina, 2020 ). Especially when host‐symbiont partnerships are obligate, variation among host individuals and species in their symbiont identity may explain some of the variation in their response to environmental gradients or disturbances (Abbott et al., 2021 ; del Campo et al., 2020 ; Innis et al., 2018 ). However, the extent to which the interaction between host and symbiont itself evolves remains unclear in many cases. Therefore, identifying host‐symbiont specificity, and the extent to which host and symbiont lineages codiversify through shared evolutionary histories, is necessary for understanding the ecological and evolutionary dynamics of the holobiont, and the degree to which symbioses promote or restrict adaptation to climate change (Compant et al., 2010 ; Kaltenpoth et al., 2014 ; Seah et al., 2017 ; Takiya et al., 2006 ). Most scleractinian corals form obligate symbiotic relationships with photosynthetic dinoflagellate algae in the family Symbiodiniaceae. In some corals, symbiont flexibility and the environment often play a substantial role in the composition of Symbiodiniaceae hosted (Boulotte et al., 2016 ; Cunning et al., 2015 ; Putnam et al., 2012 ; Quigley et al., 2019 ). However, there is also evidence for some level of host‐symbiont specificity (Forsman et al., 2020 ; LaJeunesse & Thornhill, 2011 ; Parkinson & Baums, 2014 ), though direct evidence for concordant evolutionary histories is often considered lacking (Rowan & Powers, 1991 ; van Oppen & Medina, 2020 ). The degree of host–symbiont specificity, and the potential for coevolution, should relate to the mode of symbiont transmission (Baird et al., 2009 , 2021 ; Goodnight, 2000 ; Hartmann et al., 2017 ; Zeng et al., 2017 ). Most (~71%) scleractinian coral species obtain their symbionts horizontally (i.e., acquired from the environment), which is common in species that spawn gametes with external fertilization (broadcast spawning) (Baird et al., 2009 ). These coral species are expected to host greater symbiont diversity, and necessarily host algal species that are able to live outside the coral (Fujise et al., 2021 ; Quigley et al., 2017 ), reducing specificity and decoupling their evolutionary histories. In contrast, other coral species obtain their symbionts vertically (i.e., from parent to offspring; Hirose et al., 2000 ; Hirose & Hidaka, 2006 ). Vertical transmission is more common in coral species with internal fertilization, where the embryo develops within the polyp before release as a motile planula larva (brooders) (Baird et al., 2009 ). Coral species with vertical transmission are expected to host lower algal diversity (Bongaerts et al., 2015 ), and algal species that are highly host specialized and often unable to live and propagate outside the coral host (Fujise et al., 2021 ). As a result, partnerships with hosts that vertically transmit algal symbionts are expected to be more stable over time, facilitating coadaptation of partners as a result of their shared reproductive fate (Fisher et al., 2017 ). Therefore, especially in species with vertical transmission, the main cause of symbiont speciation is thought to be niche diversification provided by speciation of the host, since the host provides the habitat that modulates natural selection on the symbiont (Thornhill et al., 2014 ). Nevertheless, there are few direct tests of this hypothesis (Lewis et al., 2019 ; Turnham et al., 2021 ). Estimates of cophylogeny, or the concordance between the phylogenies of hosts and symbionts, provide a means to identify host—symbiont specificity and shared evolutionary histories. Congruence between the phylogenies of coral hosts and algal symbionts may be the product of coevolution or cospeciation. Strictly speaking, coevolution involves reciprocity; that is, evolutionary change in one species in response to the traits of a second species, followed by evolutionary response by the second species to the change in the first species (Janzen, 1980 ; Thompson, 1994 ). Cospeciation is concordant patterns of speciation that do not necessarily involve reciprocity, such as speciation of the coral host and subsequent tracking and diversification of the algal symbionts (Lewis et al., 2019 ), or shared biogeography and similar responses to common environments (Althoff et al., 2012 ; Nuismer et al., 2010 ). However, cophylogeny between corals and their photosynthetic algal symbionts has not yet been formally tested or quantified (LaJeunesse et al., 2010 ; Pinzon & LaJeunesse, 2011 ), though there is recent evidence that the age at which algal symbiont species diverged from each other corresponds to when their hosts diversified (Turnham et al., 2021 ), evolutionary associations between modes of symbiont transmission and reproduction (Hartmann et al., 2017 ), evolutionary stability and maintenance of photosynthetic algae in corals (Gault et al., 2021 ), and cophylogeny between corals and bacterial symbionts (O'Brien et al., 2019 ; Pollock et al., 2018 ). Estimates of cophyologeny are important because several factors could limit the concordance between host and symbiont phylogeny when it is expected to be high (i.e., when there is vertical transmission). Some coral species with vertical transmission appear to exhibit a mixed mode of symbiont acquisition, allowing larvae to associate with symbionts that are not detected in the maternal colony (Quigley et al., 2018 ). A mixed mode of symbiont transmission may allow juveniles to have greater flexibility in different environments than strictly vertical transmission. Even if symbionts are transmitted vertically, an individual can host a diverse community of ‘background’ symbionts, even when there is often only one taxa detected, that are transitory and may depend on environmental conditions (Kriefall et al., 2022 ; Lee et al., 2016 ; Rouzé et al., 2019 ; Strader et al., 2022 ). As a result, even in species with vertical transmission, the relationships between scleractinian host and symbiont can be dynamic over ecological timescales (Quigley et al., 2019 ; Reich et al., 2017 ). Certain species of algal symbionts may share evolutionary histories with the coral host, while other species within the same host clade may not. Identifying associations between host and symbiont that are maintained in different habitats and in response to environmental stressors is therefore necessary for confidently detecting coevolutionary relationships, as well as generating new hypotheses for what drives these associations. Tests of cophylogeny between corals and their algal symbionts have been hampered in the past by unresolved taxonomy and phylogenetic relationships within both the coral host and the algal symbiont. Both the coral host and the symbiont contain morphologically‐cryptic species, making it difficult to identify associations without rigorous genomic information. In particular, the coral genus Pocillopora is notorious for containing multiple cryptic species because of morphological plasticity and overlapping morphological phenotypes (Marti‐Puig et al., 2014 ; Paz‐García et al., 2015 ; Pinzón et al., 2013 ). Despite being a well‐studied coral genus (Flot et al., 2008 ; Gélin et al., 2017 ; Johnston et al., 2017 ; Oury et al., 2021 ; Schmidt‐Roach et al., 2014 ), with a monophyletic radiation approximately 3 million years old (Johnston et al., 2017 ), the identification and placement of some mitochondrial lineages is not yet fully resolved. Similarly, Symbiodiniaceae, the dinoflagellates that form the symbiotic partnership with scleractinians, have recently received taxonomic revision at the family and genus level (LaJeunesse et al., 2018 ), and the extent to which variation in genetic sequences reflects species is only just beginning to be clarified (Lewis et al., 2019 ; Thornhill et al., 2014 ; Turnham et al., 2021 ). Here, we investigate Pocillopora –Symbiodiniaceae associations in multiple co‐occurring cryptic Pocillopora species that differ in their responses to thermal stress and in their niche space across depths at Mo′orea, French Polynesia (Burgess et al., 2021 ; Johnston et al., 2022 ). Pocilloporid corals are a particularly interesting group because colonies that broadcast spawn actually transmit their symbionts vertically in their eggs (Hirose et al., 2000 ; Johnston et al., 2020 ; Massé et al., 2013 ; Schmidt‐Roach et al., 2012 ), unlike many other broadcast spawning corals that do not transmit their symbionts to offspring (Baird et al., 2009 ). Eggs are relatively small (~100 μm diameter) and negatively buoyant upon release (Schmidt‐Roach et al., 2012 ), in contrast to other broadcast spawning species that release eggs without algal symbionts, which possibly reflects selection to facilitate fertilization while avoiding symbiont stress from radiation at the sea surface that causes harm to hosts (Hartmann et al., 2017 ). Previous studies on Pocillopora have found both symbiont specificity and flexibility (Cunning et al., 2013 ; Noreen et al., 2013 ; Pinzon & LaJeunesse, 2011 ; Rouzé et al., 2019 ; Schmidt‐Roach et al., 2013 ). We used multiple genomic data sets to reconstruct phylogenetic relationships and determine the placement of mitochondrial haplotypes that were not included in previous Pocillopora phylogenomic analyses (Johnston et al., 2017 ). We also assessed Symbiodiniaceae relationships using sequence variation in both ITS2 and psbA ncr markers in order to determine whether there is support for their cophylogeny, or multiple colonization events of distantly related Symbiodiniaceae into Pocillopora . Given the recent evidence of codiversification described for some Pocillopora lineages with Cladocopium (Turnham et al., 2021 ), we hypothesize that there will be specific Pocillopora — Cladocopium associations that are not likely to differ across depths, reflecting cospeciation.",
"discussion": "4 DISCUSSION Identifying the dependence of symbiont lineages on their host lineages is important for understanding the evolutionary history and ecological responses of the holobiont (Corbin et al., 2017 ). Despite the obligate symbiotic relationships between photosynthetic algae (Symbiodiniaceae) and scleractinian coral hosts being a classic example of a holobiont, explicit tests of phylogenetic congruence have been absent. Furthermore, formal, quantitative tests of cophylogeny between corals and their algal symbionts have been hampered in the past because both corals and algae contain genetically unresolved and morphologically‐cryptic species. Here, we used multiple genomic data sets from samples collected across multiple depths and sites at Mo′orea, French Polynesia, to resolve previously well‐studied mitochondrial lineages into species and found strong evidence of cophylogeny between Pocillopora species and their algal symbionts. Our analyses supported the presence of five co‐occurring Pocillopora species on the fore reef of Mo′orea that largely hosted two species of host‐specialized algae ( Cladocopium latusorum and C. pacificum). Pocillopora acuta can also be found at Mo′orea but is only found on the fringing reef, not on the fore reef, and largely hosts Durusdinium in this environment (Rouzé et al., 2019 ; Strader et al., 2022 ) and was therefore not included in our analyses. In general, the Cladocopium phylogeny, including multiple clades within species, mirrored the Pocillopora phylogeny (Figure 7 ). Certain Cladocopium species and lineages within species showed stronger dependence with specific Pocillopora species than others. Such cophylogeny is probably a consequence of the life history of broadcast‐spawning Pocillopora , which, unlike many broadcast spawning corals from other genera, transmit algal symbionts to the next generation in the eggs (Hirose et al., 2000 ; Schmidt‐Roach et al., 2012 ). However, we also found that some Pocillopora colonies were dominated by symbiont taxa from Symbiodinium and Durusdinium . Taxa from Symbiodinium and Durusdinium also occur in other host taxa ranging from jellyfish (Mammone et al., 2021 ; Sachs & Wilcox, 2006 ) to clams (DeBoer et al., 2012 ; Poo et al., 2021 ), and may have a free‐living stage (Fujise et al., 2021 ; Pochon et al., 2014 ). The presence of symbiont taxa from Symbiodinium and Durusdinium suggests that there may also be some degree of horizontal transmission in broadcast‐spawning Pocillopora , as has been observed in the brooding P. damicornis (Epstein et al., 2019 ). While speciation of C. latusorum and C. pacificum appears to be driven by niche diversification provided by speciation in Pocillopora (Turnham et al., 2021 ), we are now in a position to understand more about certain associations of symbiont clades within more Pocillopora species. Furthermore, Pocillopora still harbours generalist symbionts that may also be transmitted vertically but could be acquired horizontally as well. Recently, Turnham et al. ( 2021 ) described two new species of Symbiodiniaceae in the genus Cladocopium and found that they codiversified with different Pocillopora lineages, probably as a result of maternal vertical transmission of symbionts. They found that C. latusorum codiversified with the lineage that contains P. meandrina , haplotype 8a, and P. grandis , and that C. pacificum codiversified with P. verrucosa , and that this speciation event in Cladocopium coincided with the diversification of these extant Pocillopora lineages approximately 3MYA. Here, we provide additional symbiont associations for several Pocillopora species and provide a formal test of cophylogeny. Specifically, we found that, in addition to P. meandrina , haplotype 8a, and P. grandis , haplotypes 11/2 also host C. latusorum . In addition to P. verrucosa , haplotype 10 also hosts C. pacificum . Furthermore, we show that there are unique associations between Pocillopora species and different lineages within C. latusorum and C. pacificum , providing further evidence for the cospeciation hypothesis. In the past, cophylogeny between corals and their algal symbionts was expected to be rare or absent (van Oppen & Medina, 2020 ) because many coral species exhibit horizontal transmission (Baird et al., 2009 ), many algal symbiont species exhibit flexibility to associate with different hosts (LaJeunesse et al., 2018 ), and even coral species with vertical transmission can have some degree of horizontal acquisition of symbionts (Quigley et al., 2018 , 2019 ). Other coral species, including members from Porites and Montipora , also exhibit vertical transmission and harbour host‐specialist species of Cladocopium (Forsman et al., 2020 ; Hoadley et al., 2021 ; LaJeunesse et al., 2004 ; LaJeunesse & Thornhill, 2011 ), but the extent of cophylogeny within these groups remains to be tested. Members of the genus Cladocopium associate with a broad diversity of hosts in addition to corals, including other cnidarians, clams, ciliates, flatworms, foraminifera, and sponges, but can be highly host specialized (LaJeunesse et al., 2018 ). Therefore, an important next step to test the links between vertical transmission and cospeciation is to compare estimates of cophylogeny among more groups of corals and compare groups with vertical versus horizontal transmission. The functional differences, if any, between C. latusorum and C. pacificum in their effects on the coral host are not clear. However, the high specificity and evidence for shared evolutionary history still generates hypotheses for the degree to which symbioses in this group promote or restrict adaptation to climate change (Compant et al., 2010 ). For example, the ability to switch or shuffle symbionts in response to environmental change may be lower in coral species with high symbiont specificity. To test this, it will be informative to know the extent to which Pocillopora species hosting C. latusorum or C. pacificum differ in their ability to switch or shuffle symbionts. In our data set, only Pocillopora species that typically host C. latusorum showed the ability to host taxa from Symbiodinium and Durusdinium . In fact, only those with similar relative abundances between 5 and 20 m ( P. grandis , and haplotype 11 and 2: Johnston et al., 2022 ) hosted symbionts from genera other than Cladocopium . For example, Durusdinium glynnii , a heat‐tolerant host generalist (LaJeunesse et al., 2010 ), was detected in five haplotype 11/2 colonies. D. glynni forms stable, long‐term associations with hosts that are often found in highly variable environments (Innis et al., 2018 ; LaJeunesse et al., 2010 ; McGinley et al., 2012 ; Rouzé et al., 2019 ; Sawall et al., 2014 ). Interestingly, haplotypes 11/2 experienced the greatest bleaching mortality following the bleaching event at Mo′orea in 2019 (Burgess et al., 2021 ), thus the association with D. glynnii in the survivors sampled here could reflect the outcome of selective mortality or symbiont shuffling following this event. Furthermore, the decreased thermal tolerance of haplotypes 11/2 may relate to their unique association with C. latusorum Clade III (Figure 5 ), reduced fitness from introgression among hosts (Kim et al., 2018 ; Muhlfeld et al., 2009 ), or the interplay between both (Miller et al., 2010 ). Similarly, four P. grandis colonies contained 12%–91% Symbiodinium spp. This genus contains both free‐living and obligate symbionts, and generalist species found in many tropical environments and in a diversity of hosts such as jellyfish, giant clams, and other coral genera (LaJeunesse, 2017 ). Some taxa are highly infectious and when attracted to certain hosts, can swarm in high densities next to these available hosts (Yamashita et al., 2014 ). In the sea anemone, Exaiptasia pallida , colonization by S. microadriaticum was found to be akin to a form of parasitism in which there is no apparent benefit to the host (Gabay et al., 2018 ). It is interesting that Symbiodinium was only detected in P. grandis . In the Red Sea, P. verrucosa , increasingly hosts Symbiodinium with decreasing latitude, indicating that there may be some benefit to this symbiosis with increasing sea water temperature (Sawall et al., 2014 ). But whether the Symbiodinium taxa hosted in P. grandis at Mo′orea is the same as that hosted in the Red Sea is not known. Additionally, whether this symbiosis with P. grandis is long‐lasting or brief is not known, nor whether this Symbiodinium spp. is more attracted to P. grandis than other Pocillopora species at Mo′orea. Alternatively, the Cladocopium lineages hosted by P. grandis may be poorer competitors with this Symbiodinium spp. than those hosted by other Pocillopora , or the immune system of P. grandis may be less able to detect this Symbiodinium spp. A higher rate of sexual reproduction within a Cladocopium species or clade could lead to greater potential for adaptation to different environments. Recently, cytological evidence of reproduction of C. latusorum in P. meandrina was found to be of mixed mode; mainly asexual with sexual reproduction occurring in <1%–5% of cells in hospite (Figueroa et al., 2021 ). While there is some evidence that sexual reproduction in free‐living Symbiodiniaceae may be triggered by stress, such as nutrient deficiency (Pfiester, 1989 ), the triggers of sexual reproduction in symbiotic Symbiodiniaceae, such as C. latusorum (Fujise et al., 2021 ), are not known. These triggers may be both biotic and abiotic, differ by Cladocopium species or lineage, and/or be controlled by the host. The depth differences we detected in the symbiont ITS2 sequence diversity and type profiles within Pocillopora haplotype 10 may indicate the potential for adaptive divergence in C. pacificum , as hypothesized in other coral species (Bongaerts et al., 2011 , 2013 ). By genetically identifying species of both Pocillopora and Symbiodiniaceae, our analysis overcomes a major hurdle that has prevented tests of coevolution and cospeciation in corals and their symbiotic algae. For example, we can also confirm that haplotype 10 (Forsman et al., 2013 ) is a distinct Pocillopora species that warrants formal identification. Although haplotype 10 hosted the same algal species ( C. pacificum ) as its sister species P. verrucosa (haplotypes 3a, 3b, 3f, and 3h), it tended to host a different psbA ncr clade and ITS2 type profile. Haplotype 10 and P. verrucosa are both more common at 20 m than 5 m, though haplotype 10 is far more abundant than P. verrucosa at all depths at Mo′orea (Johnston et al., 2022 ). Furthermore, geographic sampling of Pocillopora to date indicates that haplotype 10 may be endemic to French Polynesia while P. verrucosa is widely distributed from the Tropical Eastern Pacific to the Red Sea and Arabian Gulf (Forsman et al., 2013 ; Gélin et al., 2017 ; Mayfield et al., 2015 ). As a result, we hypothesize that haplotype 10 diverged in sympatry from P. verrucosa , and that this is reflected in some of the genetic divergence in C. pacificum lineages among these two Pocillopora species. \n Pocillopora haplotypes showing evidence for introgression and similar ecology also exhibited similar Cladocopium composition. In contrast to Johnston et al. ( 2017 ), in which colonies were sampled from across the Pacific and where evidence of hybridization was found only between the youngest Pocillopora species (the brooders P. acuta and P. damicornis ), here we found evidence of introgression in multiple branches of the phylogeny in co‐occurring Pocillopora corals from Mo′orea, French Polynesia. Genomic data indicated fewer species than mitochondrial data. Because mitochondrial evolution proceeds more slowly than nuclear evolution in anthozoans (Shearer et al., 2002 ), our genomic analysis suggests relatively recent introgression between P. meandrina (mitochondrial haplotype 1a) and haplotype 8a, and between haplotypes 11 and 2, even though these haplotypes resolved as reciprocally monophyletic in our genomic analysis. Our finding of introgression between P. meandrina (mitochondrial haplotype 1a) and haplotype 8a also contrasts with Gélin et al. ( 2017 ) who designated haplotype 1a and haplotype 8a as distinct primary species hypotheses (PSH) using mostly mitochondrial markers (Figure 3 ). Similarly, we found no evidence that haplotype 3f (PSH 16) is a distinct species from haplotypes 3a, 3b, 3h (PSH 13) using nuclear and mitochondrial genomes (compare with Gélin et al., 2017 ). Furthermore, despite a lack of divergence in the mitochondrial genome in P. meandrina and P. grandis , these species are distinct, confirming previous findings (Johnston et al., 2017 ). Similarly, mitochondrial data suggests that haplotype 2 is ancestral to haplotype 11 and P. ligulata . Haplotype 2 is geographically widespread, found from Clipperton Atoll in the Tropical Eastern Pacific to Madagascar in the western Indian Ocean, but is relatively rare throughout this range (Gélin et al., 2017 ; Marti‐Puig et al., 2014 ; Pinzón et al., 2013 ) and at Mo′orea (Burgess et al., 2021 ). In contrast, haplotype 11 has, to date, only been sampled from French Polynesia (Burgess et al., 2021 ; Forsman et al., 2013 ; Johnston et al., 2022 ). Given the basal placement of haplotype 2 in the mitochondrial phylogeny, and their respective geographic distributions, we hypothesize that haplotype 11 and P. ligulata diverged from haplotype 2 in isolation. However, the nuclear introgression observed between haplotypes 11 and 2, and the lack of differentiation in symbionts hosted, suggest that there is some gene flow between these haplotypes at Mo′orea. At Clipperton Atoll, haplotype 2 hosts a yet to be described species that is different from that hosted in Mo′orea, which is neither C. latusorum nor C. pacificum , that is closely related to C. goreaui (Figure 5a ; taxa beginning with BB07) (Pinzon & LaJeunesse, 2011 ). Whether there are fitness impacts when interspecific crosses occur between Pocillopora that host different algal species, such as the failure to transmit the algal symbiont or reduced symbiont growth if transmitted, is not known. Given the emerging evidence, the mosaic of Pocillopora species distributions, and thus respective interactions, is probably evolutionarily complex across time and space, giving rise to lineages and species following a pattern of reticulate evolution, as has been hypothesized for many other coral species (Arrigoni et al., 2016 ; Diekmann et al., 2001 ; Veron & Stafford‐Smith, 2000 )."
} | 6,726 |
25225493 | PMC4150442 | pmc | 2,388 | {
"abstract": "The ability of microorganisms to thrive under oxygen-free conditions in subsurface environments relies on the enzymatic reduction of oxidized elements, such as sulfate, ferric iron, or CO 2 , coupled to the oxidation of inorganic or organic compounds. A broad phylogenetic and functional diversity of microorganisms from subsurface environments has been described using isolation-based and advanced molecular ecological techniques. The physiological groups reviewed here comprise iron-, manganese-, and nitrate-reducing microorganisms. In the context of recent findings also the potential of chlorate and perchlorate [jointly termed (per)chlorate] reduction in oil reservoirs will be discussed. Special attention is given to elevated temperatures that are predominant in the deep subsurface. Microbial reduction of (per)chlorate is a thermodynamically favorable redox process, also at high temperature. However, knowledge about (per)chlorate reduction at elevated temperatures is still scarce and restricted to members of the Firmicutes and the archaeon Archaeoglobus fulgidus . By analyzing the diversity and phylogenetic distribution of functional genes in (meta)genome databases and combining this knowledge with extrapolations to earlier-made physiological observations we speculate on the potential of (per)chlorate reduction in the subsurface and more precisely oil fields. In addition, the application of (per)chlorate for bioremediation, souring control, and microbial enhanced oil recovery are addressed.",
"introduction": "INTRODUCTION Microorganisms inhabit subsurface environments 100s of meters below Earth’s surface where oxygen is most often lacking. The development of the first modern oil wells in the 19th century opened the “gate to the deep biosphere” and not long after that scientists discovered the first microbes thriving in these environments ( Bastin, 1926 ; Ginzburg-Karagicheva, 1926 ; Gahl and Anderson, 1928 ). Particularly the studies of Tausson (1925) and Zobell (1945) gained detailed insight into the microbial oxidation of hydrocarbons by indigenous subsurface microbes. A large number of studies in the following decades tightened the concept of an active and diverse microbial subsurface community. The development of improved anaerobic culturing techniques during the second half of the 20th century resulted in another step forward in the identification of anaerobes and their physiology ( Magot et al., 2000 ; Youssef et al., 2009 ). These indigenous subsurface microbes were isolated and often deposited in publicly accessible strain collections. A major driver for investigating the microbiology of oil reservoirs has been the biogenic in situ formation of hydrogen sulfide from sulfate, causing souring. The detrimental effects associated with the formation of hydrogen sulfide (high toxicity, sulfide stress cracking, corrosion, precipitation of metal sulfides) increase the production and refinery costs of petroleum ( Tang et al., 2009 ) and have created a generally negative image of microorganisms in oil fields from the beginning of modern oil recovery ( Bastin et al., 1926 ). However, particular microorganisms indigenous (or introduced) to the subsurface may have characteristics that are desirable during oil recovery, and it might be beneficial to stimulate these further in situ . The most prominent example is the mitigation of souring by nitrate-reducing communities in oil fields ( Hubert and Voordouw, 2007 ; Gieg et al., 2011 ). Additionally, growing effort is spent on the development of new strategies for microbial enhanced oil recovery (MEOR), or other processes (e.g., conversion of coal to methane) that use the “help of microorganisms” for increasing hydrocarbon recovery."
} | 932 |
38930834 | PMC11206051 | pmc | 2,389 | {
"abstract": "Electrospinning is a cost-effective and flexible technology for producing nanofibers with large specific surface areas, functionalized surfaces, and stable structures. In recent years, electrospun nanofibers have attracted more and more attention in electrochemical biosensors due to their excellent morphological and structural properties. This review outlines the principle of electrospinning technology. The strategies of producing nanofibers with different diameters, morphologies, and structures are discussed to understand the regulation rules of nanofiber morphology and structure. The application of electrospun nanofibers in electrochemical biosensors is reviewed in detail. In addition, we look towards the future prospects of electrospinning technology and the challenge of scale production.",
"conclusion": "6. Conclusions Electrospinning is a simple and efficient method for producing nanofibers. Electrospun nanofibers have large specific surface area, high porosity, and diversified structure. In addition, it is easy to functionalize the surface of electrospun nanofibers. These advantages make electrospinning technology have great potential in electrochemical sensors. In the past decades, a large number of studies have laid the foundation for the control of the structure and morphology of electrospun nanofibers, and many interesting composite nanofibers structures have been witnessed in practical applications of electrochemical sensing. In this review, an overview was provided on the morphological and structural regulation of nanofibers by electrospinning technology as well as its application in electrochemical sensors. The factors affecting the morphology of the nanofibers and the strategies for producing nanofibers with special morphology were discussed. Generally, the performance of nanofibers can be improved and their application can be expanded through the following strategies: (i) nanofiber diameter; (ii) special morphology (hollow, core-sheath, porous and bead-like); (iii) functionalization of surfaces; (iv) metallization of nanofibers; and (v) carbonization to form CNFs. The working principles and advantages of various nanofibers obtained by these strategies in the detection of different substances were also discussed in this review. In summary, many achievements have been made in both the theory and application of electrospinning technology in recent decades. Electrospinning has been involved in many fields, such as energy storage, tissue engineering, environmental protection, smart textiles, electronic devices, and physical and chemical sensors. In particular, the current research on electrospinning has a tendency of miniaturization, simplification, and economization, making it possible for the technique to play an important role in intelligent medical treatment. However, there are still many issues. For example, process and safety problems still exist in the large-scale production of electrospinning. In addition, it is necessary to establish a complete evaluation system for the safety of electrospun nanofibers.",
"introduction": "1. Introduction Nanofibers, as one-dimensional nanomaterials, have attracted much attention due to their unique advantages of large specific surface areas, functionalized surfaces, and stable structures. The large specific surface areas contribute to the excellent adsorption performance of nanofibers [ 1 , 2 ]. In particular, nanofibers with suitable pore size distribution can provide a large number of sites to accommodate the high loading of active materials [ 3 ]. The electrospinning technique is a process of spinning polymer solutions or melts under a strong electric field. Under the action of the electric field, the spinning fluid expands into a tiny jet that solidifies into a fiber. Electrospinning devices are simple and have low cost. Electrospinning can produce a wide variety of materials, and the process is controllable. These advantages make it one of the most popular techniques for producing polymer nanofibers. In addition, carbon nanofibers can be obtained by carbonized polymer precursors [ 4 ]. Electrospinning is also a common technique for producing composite nanofibers. In contrast with the traditional spinning technique, a high-voltage electrostatic field can stretch polymer solution (or melt) into nanofiber, making the diameter of the fiber produced by electrospinning to be as small as one nanometer [ 5 ]. Moreover, the technique is simple, cost-effective, and versatile, making it suitable for industrial production, as will be discussed later in the review. Nowadays, with the development of industry, more and more pollutants of inorganic and organic contaminants have been produced. Therefore, qualitative and quantitative analyses of these pollutants play an important role in environmental protection and food safety [ 6 , 7 , 8 , 9 ]. Many techniques have been developed such as fluorescence, UV-Vis spectroscopy, mass spectrometry, and electroanalysis [ 10 , 11 , 12 , 13 , 14 ]. Among them, electrochemical sensors have received more and more attention due to their excellent sensitivity, accuracy, wide detection range, easy operation, and low price [ 15 ]. In biosensors, electrospun nanofibers can be used as substrate materials or functional components of sensors. As the base material, electrospun nanofibers can provide a large surface area to enhance the adsorption of biomolecules, thus improving the sensitivity and detection limit of the sensors. At the same time, the pore structure of electrospun nanofibers is also conducive to the diffusion and transfer of biomolecules, enhancing the response speed and stability of the sensors. In addition, electrospun nanofibers can also be used as functional components, such as fixed carriers of biomolecules and fixed substrates of biometric molecules. By immobilizing biomolecules or biometric molecules onto electrospun nanofibers, a highly sensitive and selective detection of different biomolecules can be achieved. In this review, we outline the principle of electrospinning technology, the regulation of nanofiber morphology, and the application of electrospun nanofibers in electrochemical sensors. The challenge and prospect of electrospinning technology are also prospected."
} | 1,560 |
36245414 | PMC10108327 | pmc | 2,390 | {
"abstract": "The growing consumption of electrical and electronic equipment leads to high\namounts of electronic waste (e-waste), which is now considered the\nfastest-growing waste stream at the national and international levels. As well\nas being a potential secondary resource due to its precious metals content,\ne-waste also contains strategic metals and plastics. For instance, mobile phones\nhave about 25–55% plastic substances. A few studies have been performed to\ninvestigate the potential of indigenous bacteria in metals’ bioleaching from the\npolluted environment. Heterotrophic bioleaching potential in acidic conditions\nhad been preliminarily investigated. Two soil types of iron ore were considered\nthe source of indigenous bacteria. Despite the acidophilic nature of the\nbacterial consortium, they continued their leaching activity regardless of\nalkaline conditions. Maximum biorecovery rate related to copper (4%) responding\nto the main soil, owing to the higher copper content of mobile phone waste.\nChromium had the least recovery rate (⩽0.002%). Overall, the maximum metal\nrecovery rate was 4.7%, achieved by tailing heterotrophs at an e-waste loading\nof 10 g l −1 . Statistical analysis had shown that there was no\nsignificant difference between the metal recovery rates and soil type or even\nthe solid-liquid ratio ( p > 0.05). Although acidophilic\nindigenous heterotrophs could not be an appropriate alternative for a large\namount of metal recovery process, they might have considerable potential in the\nbioremediation of e-waste plastic fractions and metals in low concentrations\nsimultaneously.",
"conclusion": "Conclusions The main objective of this study was to obtain preliminary data on the feasibility of\nthe heterotrophic consortia in solubilizing hazardous metals and degrading plastics\ncontained in WPCBs in acidic conditions. The iron mine soil had shown rich total\nheterotrophic bacterial counts of up to 130 × 104 and 149 × 104 in the main soil and\nthe tailing soil, respectively. In bioleaching set-ups, pH tended to neutral/slightly alkaline values, probably due\nto the alkaline nature of the WPCBs. Despite the acidophilic nature of the bacterial\nconsortium, they continued their leaching activity regardless of alkaline\nconditions. The highest biorecovery rate was achieved by heterotrophs of the main\nsoil corresponding to copper (4%) at S/L = 10 g l −1 , owing to the higher\namount of this metal in WPCBs. Other metals had a nominal bioleaching rate; among\nthem, chromium showed the least recovery rate (⩽0.002%). It might be concluded that\nchromium has a higher degree of toxicity to these bacteria. Hence, it is suggested\nto conduct quantification tests to measure the toxicity level of the heavy metals\nand determine the bacterial tolerance towards these toxic metals. Considering the\nentire metals involved, the highest recovery rate was 4.7%, achieved by heterotrophs\nof tailing at S/L = 10 g l −1 . Although acidophilic indigenous\nheterotrophs could not be an appropriate alternative for the extensive metal\nrecovery process, they could have considerable potential in the bioremediation of\nWPCBs plastic fractions and metals in low concentrations simultaneously. This\nreveals that central iron ore soil of Iran country can be an appropriate source of\nmicroorganisms with potential for plastic degradation.",
"introduction": "Introduction Environmental protection as a constitute of ecological health is preventing undesired\nchanges to ecosystems and their components ( Hamilton et al., 2018 ). Therefore, it\nfocuses on identifying, evaluating and eliminating environmental hazards ( Nriagu, 2019 ). Tracking\nthese environmental challenges is a beneficial step to identifying harmful exposures\nand implementing strategies towards reducing these risks, and creating\nhealth-supportive environments ( WHO, 2018 , 2020 ). The era of accelerated technological progress has brought comfort to lives but also\nintroduced harmful contaminants and pollution to the environment ( Basu et al., 2022 ). The\ngrowing consumption of electrical and electronic equipment leads to high amounts of\nelectronic waste (e-waste), which is now considered the fastest-growing waste stream\nboth at the national and international levels ( Abalansa et al., 2021 ; Forti et al., 2020 ; UN, 2017 ). E-waste is a\nheterogeneous combination of various materials ( Purchase et al., 2020 ). As well as being a\npotential secondary resource due to its precious metals content, e-waste also\ncontains strategic metals and plastics ( Ankit et al., 2021 ). Critical metals include aluminium, arsenic, cadmium, chromium, cobalt, gallium, lead,\nmercury, nickel, platinum, tin, titanium, zinc and other rare earth elements ( FederalRegister, 2021 ;\n Rai et al., 2021 ).\nThese metals exist on earth naturally and are also produced by humankind’s\nactivities. Some of these metals are essential for the growth of organisms. Still,\ntheir higher concentrations significantly damage humans’ health since they are\nnon-degradable and persist long-term in the ecosystem ( Ayangbenro and Babalola, 2018 ). The adverse\nhealth effects on humans include severe toxic effects on the brain, heart, liver,\nkidney and skeletal system damage ( Ali et al., 2021 ). On the other hand, plastics represent more than about 20% of the mass content of\nwaste electrical and electronic equipment ( Basu et al., 2022 ). For instance, mobile\nphones have about 25–55% plastic substances. It was estimated that the consumption\nof plastics in electrical and electronic equipment would reach more than 6 billion\npounds in 2022. By 2040, about 400 million tons of CO 2 emissions will be\nexpected due to global total e-waste plastic production and incineration ( Singh et al., 2020 ).\nE-wastes contain various kinds of polymers. High-impact polystyrene (HIPS) and\nacrylonitrile butadiene styrene (ABS), polystyrene (PS), polyethylene, polypropylene\n(PP), styrene-acrylonitrile, acrylonitrile butadiene styrene/polycarbonate (ABS/PC),\npolyamide, polyurethane, polyphenylene oxide (HIPS/PPO), polybutadiene\nterephthalate, polyamide 66 (Nylon 66) and polyphenylene sulphide are known as the\nmain polymers in printed circuit boards (PCBs). Also, PS, ABS and PP are the main\npolymers in e-waste constituting about 31%, 16% and 13%, respectively ( Arshadi et al., 2019 ). That\nis why minimization of the harmful effects of the contained contaminants should come\nunder serious consideration ( Basu et al., 2022 ). Although advanced incineration and modern landfills can effectively manage the\ne-waste issue, these technologies are not sustainable and appropriate for long-term\napplication ( Awasthi and Li,\n2019 ). Therefore, it falls to the recycling industry to help sustainable\nwaste management, environmental protection and resource conservation through\nremoving hazardous materials and economic recovery of metal resources from e-waste\n( Awasthi and Li,\n2019 ; Işıldar et al.,\n2018 ). However, recycling e-waste is very difficult due to its complexity\nand toxicity ( Singh et al.,\n2020 ). Yet, for the widespread application of e-waste recycling, it is essential to consider\nprocesses with a minor release of toxic substances and affordability ( Rai et al., 2021 ). To this\naim, several approaches have been recommended for effective and eco-friendly\nrecycling of e-waste from which biological methods and bioremediation are widely\naccepted ( Basu et al.,\n2022 ). Microorganisms are appropriate candidates for solving complex problems in different\nenvironmental aspects of life due to their exclusivity and unpredictable nature, and\nbiosynthetic capacity ( Kumar\nand Gopal, 2015 ). A variety of microorganisms pose excellent capability\nfor sustainable biotechnology, especially the heterotrophic communities with\npromising biosynthetic activity ( Kumar and Gopal, 2015 ; Zuñiga et al., 2020 ). Bioleaching is a promising technology for bioremediation that uses capable\nmicroorganisms to recover various metals from different minerals and waste\nmaterials, including e-waste, through microbial solubilization of metals ( Gopikrishnan et al.,\n2020 ). Although bioleaching has been performed since prehistoric times, its\napplication in metal recovery from e-waste is still emerging ( Valix, 2017 ). Many microorganisms are applied for metal recovery, especially those isolated from\nmining sites, including autotrophic and heterotrophic organisms ( Gopikrishnan et al.,\n2020 ). Among autotrophs, iron- and sulphur-oxidizing bacteria, particularly\nbacteria from the genera Acidithiobacillus and\n Leptospirillum , are widely studied and reported to play a\nsignificant role in bioleaching, especially in acidic conditions ( Gopikrishnan et al., 2020 ;\n Valix, 2017 ). Regarding heterotrophic bacteria, there are few available reports; of those, the\ngenera Bacillus and Pseudomonas are introduced as\nthe most heterotrophic strains used in bioleaching of various metals. These bacteria\nconsume organic carbon and secrete organic acids that help metal solubilization\n( Gopikrishnan et al.,\n2020 ; Valix,\n2017 ). Since these are neutrophilic bacteria, their optimal growth pH is\nbetween 5 and 9 ( Jin and Kirk,\n2018 ). Alongside the metal solubilization, some reports regarding the\nbiodegradation of high molecular weight organic polymers using potent microorganisms\nsuch as Aspergillus glaucus , A. niger ,\n Pseudomonas sp., Bacillus siamensis ,\n Klebsiella pneumonia , Micrococcus sp.,\n Staphylococcus sp. and Moraxella sp. However,\nthe degradation mechanism is not well-defined. Researchers have proposed different\npaths, including aerobic biodegradation through aerobic respiration resulting in\nH 2 O and CO 2 release and molecular structure break of\nplastics due to the secreted enzymes and other extracellular compounds ( Kathiresan, 2004 ; Zeenat et al., 2021 ). There have been several studies on the bioleaching process using microorganisms\nisolated from different environments; meanwhile, a few studies have been performed\nto investigate the potential of indigenous microorganisms in the environment. Due to\nthe abundance of heterotrophic bacteria, in this study, bioleaching potential in\nacidic conditions has preliminarily been investigated. They consider that the best\nbioleaching efficiencies typically occur at low pH values ( Díaz et al., 2018 ). The plastic\ndegradation capability of the isolated bacteria was also studied to develop\nefficient microorganisms and their products.",
"discussion": "Results and discussion Mobile phone contents and soil bacteria enumeration So far, several pieces of research have been performed using autotrophs, and less\nis available about heterotrophs in metal recovery. In contrast, heterotrophic\nbacteria have great potential for remediation of organic pollutants. The main\npurpose of this study was to investigate the ability of the heterotrophic\nconsortium of mineral areas to recover heavy metals from polluted environments.\n Figure 2 shows the\nconcentration of total metals in the mobile phone waste in this survey. Figure 2. The concentrations of metals in WPCBs as ppm (mobile phone). According to Figure 2 ,\nnickel, copper, tin, chromium, aluminium and zinc were measured as heavy metals\nwith the highest concentrations in the mobile phone waste. It was reported that,\non average, PCBs typically are made up of about 7.0% Fe, 27.0% Cu, 2.0% Al, 1.5%\nZn, 0.5% Ni, 2000 ppm ( Ajiboye et al., 2019 ). Ajiboye et al. (2019) analysed the\nWPCBs metal composition and observed that most metals included base metals such\nas copper (41.64%), iron (11.11%), aluminium (5.05%), nickel (2.28%), zinc\n(1.79%), tin (1.63%) and lead (1.09%). Yamane et al. (2011) also declared\nthat their studied PCBs from discarded computers mainly consist of copper, lead,\naluminium, iron, tin, cadmium and nickel. The present study did not investigate other metals, including sodium, potassium,\nmagnesium, iron, etc. Due to the abundance of tailing in the Bafq iron mine\narea, two soil samples were investigated, including the main soil and tailing.\nAlso, the community and morphology of bacterial isolates are indicated in Figure 3 . Figure 3. Initial bacterial community and morphological characteristics as the soil\ntype. As shown in Figure 3 ,\ntailing had a bit more bacterial community than the main soil. However, both\nsoil samples had enough bacterial content to perform bioleaching experiments.\nAlso, the formed bacterial colonies of both samples were green with a dry\nsurface, a dark core and lighter edges. Bacterial adaptation Bacterial adaptation was carried out through the gradual addition of the WPCBs.\nAlongside, main environmental parameters, including OD 595 and DO,\nwere monitored, and the results are presented in Figure 4 . Figure 4. OD 595 and DO variations during bacterial adaptation (e-waste\nwas gradually added in three stages with concentrations of 30, 60 and\n100 mg). On the first days of adaptation, high optical densities were observed. However,\nweekly WPCBs level-up decreased the OD trend ( Figure 3 ). As opposed to\nOD 595 , DO charts had an increasing trend during the adaptation\nprocess, which could probably result from continuous agitation inside the\nincubator shaker providing the required oxygen for the aerobic bacteria ( Bates et al., 2016 ).\nThe decreasing trend of OD can be attributed to two reasons; Adding higher\nloadings of WPCBs increases the toxicity of the medium containing the bacteria,\nwhich can affect bacterial growth ( Wu et al., 2018 ). Also, according to some papers, high concentrations of DO lead to an inhibitory\neffect on bacterial growth, despite oxygen being an electron acceptor through\naerobic bacterial metabolism ( Baez and Shiloach, 2014 ; Guezennec et al.,\n2016 , 2017 ;\n Li et al.,\n2015 ). This deleterious effect could be due to the oxidative stress\ninduced by the activation of reactive oxygen species due to the change in\noxidation states of metal ions in the medium ( Baez and Shiloach, 2014 ; Guezennec et al.,\n2017 ). Bioleaching experiments Figure 5 shows the\nenvironmental parameters measured during bioleaching experiments. Figure 5. pH, OD 595 and DO variations during bioleaching. As shown in Figure 5 , in\nall bioleaching set-ups, the bacterial consortia of the main soil showed\nincreasing OD. It seems that the preadapted heterotrophs of tailing have\nmaintained their growth trait, being influenced negatively by increasing DO\nduring the bioleaching period. Although microorganisms need more time to adapt\nto higher levels of e-waste ( Shaikh et al., 2018 ), it can be\nobserved that the main soil heterotrophic bacteria have adapted well to the\ntoxic e-waste. However, at S/L = 15 g l −1 , the isolates of the main\nsoil and tailing had a similar OD trend, showing decreasing growth on the ending\ndays, which can be due to the high toxicity of e-waste loading ( Wu et al., 2018 ). Although the isolates were meant to be acidophiles acidifying the medium, an\nopposite phenomenon was observed. In all bioleaching set-ups, pH tended to have\nneutral/slightly alkaline values. The fact is that WPCBs have an alkaline nature resulting in the release of\nalkaline compounds ( Rasoulnia et al., 2021 ). Since heterotrophic bacteria produce\norganic acid due to carbon consumption, this leads to a decrease in the medium\npH ( Feliatra et al.,\n2019 ; Mei et\nal., 2021 ). Yet, the amount produced organic acids in this study was\nnot enough to neutralize the alkalinity of e-waste. Also, the bacterial communities were monitored periodically. Heterotrophic\nenumeration changes during the bioleaching process are displayed in Figure 6 . Figure 6. Variation of heterotrophic bacterial communities during metals\nsolubilizations. Although it is believed that higher concentrations of e-waste lead to inhibition\nof bacterial activity ( Wu\net al., 2018 ), the main soil bacterial community showed an increasing\ngrowth trend during bioleaching of all three WPCBs loadings in this study ( Figure 6 ), suggesting\nthat the studied conditions are providing the favoured substances, especially\ncarbon source for these bacteria. Debbarma et al. isolated some indigenous\nmicrobial strains possessing the potential to degrade e-waste plastics. As they\nobserved, the microbial consortium showed accelerated growth in the presence of\ne-waste ( Debbarma et al.,\n2018 ). Meanwhile, the bacterial community of tailing soil decreased\nduring the process, except for the density of 15 g l −1 , in which the\nbacterial population of tailing soil showed an increasing trend within the first\nweek of bioleaching and then started to decrease ( Figure 6 ). The calculated rate of metal recovery is presented in Table 1 . Table 1. Metal recovery rate achieved by acidophilic heterotrophs * . Metals Main soil Tailing soil 5 g l −1 (%) 10 g l −1 (%) 15 g l −1 (%) 5 g l −1 (%) 10 g l −1 (%) 15 g l −1 (%) Aluminium (Al) 0.003 0.004 0.04 0.02 0.05 0.02 Chromium (Cr) 0.001 0.002 0.002 0.002 0.003 0.001 Copper (Cu) 3 4 3 2 3 3 Nickel (Ni) 0.3 0.1 0.2 0.03 0.5 0.1 Tin (Sn) 0.1 0.2 0.2 0.1 0.4 0.2 Zinc (Zn) 0.4 0.2 0.1 0.1 0.6 0.04 Total 3.6 4 3.3 1.8 4.7 3.7 * \n E % = C 2 / C 1 × 100 \n . - E %: metals recovery rate. - C 1 : metal content in the e-waste\nsample. - C 2 : metal content in the solution. As presented in Table\n1 , the hiighest biorecovery rate corresponded to copper (4%) by\nheterotrophs of the main soil at S/L = 10 g l −1 . The higher leaching\nvalue of copper could be owing to the large proportion of this metal in WPCBs.\nThe biorecovery rates for other metals were negligible; among them, chromium was\nthe metal with the least recovery rate (⩽0.002%). It might mean that the\nisolates were more susceptible to chromium than other metals. Rodríguez et al.\ninvestigated the tolerance behaviour of some heterotrophic bacteria isolated\nfrom the natural waters of a mining area. They reported an overall toxicity\npattern for all tested bacteria (Hg > Co > Cr > Zn > Ag = Cu),\nshowing that chromium was more toxic than copper ( Escamilla-Rodríguez et al., 2021 ). Figure 7 compares the\nweight changes of different e-waste loadings (5, 10 and 15 g l −1 )\nduring bioleaching experiments. Figure 7. Changes in WPCBs weight (mg) after bioleaching process based on S/L\nratio. Considering total studied metals, the highest recovery rate was 4.7%, achieved by\nheterotrophs of tailing at S/L = 10 g l −1 . Regarding the WPCBs weight\ndecrease due to bioleaching, the most weight change was done by the heterotrophs\nof tailing at S/L = 15 g l −1 , leading to a 2840 mg decrease ( Figure 7 ). Although the\nmechanism of the considerable WPCBs weight decrease is not surgically known, the\ncapability of these acidophilic heterotrophs in this matter suggests their\nprobable ability to consume plastics as a source of carbon or to penetrate\neasily into the polymers and biodegrade them into monomers ( Kathiresan, 2004 ;\n Zeenat et al.,\n2021 ). Although these heterotrophic isolates did not show significant\nmetal leaching performance, their application in the remediation of plastic\npollution could be beneficial. Analysed surface morphology by SEM microscopy is illustrated in Figure 8 . Figure 8. WPCBs appearance (a) before and (b) after bioleaching process. The results from SEM analysis ( Figure 8 ) indicated that the bioleached sample had a more crushed\nsurface than the raw one, which confirms the bacterial activity. Statistical analysis Correlations of metals’ biorecovery rate with solid-liquid ratio and soil type\nwere evaluated using Two-way ANOVA tests ( Table 2 ). Table 2. Correlation of biorecovered metals with soil type and solid-liquid\nratio. Metals p -Value Soil type Solid-liquid ratio Aluminium (Al) 0.414 0.422 Chromium (Cr) 0.643 0.385 Copper (Cu) 0.230 0.354 Nickel (Ni) 0.952 0.738 Tin (Sn) 0.519 0.192 Zinc (Zn) 0.950 0.394 Total 1.000 0.422 As shown, there was no significant difference between the metal recovery rates\nand soil type or even the solid-liquid ratio ( p > 0.05)."
} | 4,975 |
28168102 | PMC5289447 | pmc | 2,391 | {
"abstract": "Background Fungal endophytes are highly diverse ubiquitous asymptomatic microorganisms, some of which appear to be symbiotic. Depending on abiotic conditions and genotype of the plant, the diversity of endophytes may confer fitness benefits to plant communities. Methods We studied a crop wild relative (CWR) of strawberry, along environmental gradients with a view to understand the cultivable root-derived endophytic fungi that can be evaluated for promoting growth and tolerating stress in selected plant groups. The main objectives were to understand whether: (a) suboptimal soil types are drivers for fungal distribution and diversity; (b) high pH and poor nutrient availability lead to fungal-plant associations that help deliver fitness benefits; and (c) novel fungi can be identified for their use in improving plant growth, and alleviate stress in diverse crops. Results The study revealed that habitats with high pH and low nutrient availability have higher fungal diversity, with more rare fungi isolated from locations with chalky soil. Plants from location G were the healthiest even though soil from this location was the poorest in nutrients. Study of environmental gradients, especially extreme habitat types, may help understand the root zone fungal diversity of different functional classes. Two small in vitro pilot studies conducted with two isolates showed that endophytic fungi from suboptimal habitats can promote plant growth and fitness benefits in selected plant groups. Discussion Targeting native plants and crop wild relatives for research offers opportunities to unearth diverse functional groups of root-derived endophytic fungi that are beneficial for crops.",
"conclusion": "Conclusions This study targeted a small geographical area and a CWR from a single plant taxon to understand the fungal endophytic diversity along environmental gradients. A greater number of ‘rare’ fungi was found in the location with high pH and poor nutrient availability. It has already been recognized that the plant microbiome should be considered in crop breeding strategies ( Berg et al., 2016 ). Studying endophytes of CWRs will help understand the plant-fungus relationship and their impact on growth and stress tolerance in crop species, and further research will shed light on habitat adaptation and plant succession in marginal and challenging habitats from where many CWRs originated. As a result of domestication and migration, it is likely that present crop species have become devoid of beneficial microbes which are part of the rhizosphere of CWRs ( Hale, Broders & Iriarte, 2014 ; Pérez-Jaramillo, Mendes & Raaijmakers, 2016 ). Using less resources for better crop production is a sustainable formula for future crop growing systems, and endophytes may play a significant role ( Jones, 2013 ). Based on the preliminary results obtained from our findings, detailed studies are in progress to understand the potential beneficial effects of these fungi to plants. Endophytes from challenging environments could offer opportunities to develop smart solutions to combat both abiotic and biotic stressors in the age of climate change and continued human population growth.",
"introduction": "Introduction Endophytes are ubiquitous asymptomatic microorganisms that live within plants in their natural habitats and are often symbiotic, which can have significant impacts on plant communities. The distribution of endophytes across environmental gradients reveals unique associations between fungi and plants ( Redman & Rodriguez, 2007 ). Lau & Lennon (2012) propose that, when plants are faced with environmental change, they may benefit from association with diverse soil microbes that respond rapidly to such changes. These co-habiting belowground microbial communities could help maintain plant fitness by adapting to stressors associated with global climate change. Steps have been taken in understanding the relationships between endophytes and their hosts in a range of circumstances to study both biotic and abiotic stresses ( Malinowski & Belesky, 2000 ; Bultman & Bell, 2003 ; Rodriguez & Redman, 2008 ; Oelmuller et al., 2009 ; Larriba et al., 2015 ) and these results underpin a framework for research to understand unique plant–fungal interactions and crop development strategies for the future in an age of climate change. Crop wild relatives (CWRs) have provided breeders with genes for pest and disease resistance, abiotic stress tolerance, and quality traits in an ever-increasing number of food crops ( Xiao et al., 1996 ; Hajjar & Hodgkin, 2007 ; McCouch et al., 2007 ; Khoury, Laliberté & Guarino, 2010 ). CWRs and other plants from suboptimal environments are also a source of genetic traits that can help productivity in challenging conditions ( Mickelbart, Hasegawa & Bailey-Serres, 2015 ). Breeding of field crops for drought tolerance can be realized by using CWRs as discussed in tomato ( Schauer, Zamir & Fernie, 2005 ), wheat ( Nevo, 2007 ; Witcombe et al., 2008 ) and barley ( Nevo, 2007 ; Lakew et al., 2011 ). Selection of superior plant genotypes and optimizing growing conditions to achieve the desired performance of crops is the conventional system currently used ( Hale, Broders & Iriarte, 2014 ). They suggest that modifying the plant-associated microbiome may be an intriguing complementary strategy for crop improvement. Applying this to develop better crops using crop wild relatives (CWRs) and their microbiomes offers new possibilities. On the other hand, the number of studies reported on asymptomatic fungi that are associated with the root zone of CWRs is still very few. Murphy, Doohan & Hodkinson (2015) found that novel, horizontally transmitted endophytes from a wild relative of barley may help the crop to grow successfully in nutrient-poor soil, which will help reduce fertilizer inputs while maintaining acceptable yields. The findings of this line of research may help to grow field crops in a more sustainable, cost-effective and environmentally friendly way ( Murphy, Doohan & Hodkinson, 2015 ). As Hale, Broders & Iriarte (2014) suggested, in the same way that diversity loss was caused in crop species by domestication due to genetic bottlenecks, the process of migration may have resulted in loss of associated microbial diversity due to the physical dislocation of host plants from their co-evolved microorganisms. Comby et al. (2016) found that, for both bacterial and fungal endophytes, there were strong spatial and temporal variations, and emphasized the importance of the establishment of a collection of cultivable endophytes that can be evaluated for their beneficial effects upon crops. Non-pathogenic endophytes might be of interest in the search for plant growth promoters or biological control agents ( Alabouvette, Olivain & Steinberg, 2006 ; Berg et al., 2016 ). Two of the most important abiotic stressors are drought and salinity. In a world of changing climate, the importance of increasing the tolerance of crops to drought and salinity is critical ( Tester & Langridge, 2010 ). Our current study was carried out to see whether suboptimal soil characteristics lead to environment-specific plant–fungal associations that can confer fitness benefits and improve plant performance. We studied the distribution of fungal endophytes in one of the CWRs of strawberry, Fragaria vesca , across an environmental gradient within 1.6 km 2 of protected land in Chafford Hundred, south-east England, managed by the Essex Wildlife Trust. The main objectives of this study included: • Identification by ITS sequencing of cultivable endophytic fungi from roots of Fragaria vesca , on a spatial scale within a small geographical area of a protected woodland; • Characterization of soil from seven collection locations for their key mineral element composition, pH and humus level; • Diversity analysis of fungi in different soil types with regards to ubiquity and rareness; • Assessment of wild Fragaria vesca plant characteristics to determine the influence of endophytes; • Small pilot studies to understand the effect of two endophytic fungi on functional traits on fast-growing model species (selected cereal, legume and brassica).",
"discussion": "Discussion Ubiquity against rarity along environmental gradients Fragaria vesca from seven collection locations yielded at least 61 different endophytes, demonstrating the ability of this plant species to associate with a wide range of fungal groups. The differences in fungi recovered from different locations reinforce the idea that the interactions in plant-endophyte associations vary with environmental conditions ( Wali et al., 2009 ), even between habitats that are very proximal geographically (few hundred meters). The locations F and G were notable in their greater root fungus diversity. The NMDS plot ( Fig. 3 ) showed that location G was well separated, particularly from locations C and D, demonstrating their difference to each other in terms of associated endophytes and underlying environmental gradients. The NMDS with overlaid environmental vectors correlates endophytes at location G with higher pH and calcium levels, and low levels of phosphate and humus. This suggests that the 16 rare fungi observed only in samples from location G ( Table 2 ) may be key beneficial endophytes that help the host plant to tolerate environmental stresses. Trichoderma isolates were found only in location F which may substantiate the notion that root colonization by Trichoderma spp. can confer resistance to abiotic stresses, as well as improve the uptake and use of nutrients ( Harman et al., 2004 ). These two locations also had relatively little small vegetation, both in quantity and diversity, which may also have an influence upon fungal diversity. The more common fungi may be associated with, and dependent on, specific non-strawberry species, which may be uncommon or absent at location G. With reduced dominance of common fungi, other endophytes, which are adapted to different environments, may then be able to colonize the F. vesca plants. Plant morphology, P-limiting soil and fungal diversity Phomopsis columnaris , Ilyonectria robusta and Dactylonectria sp. were more abundant in locations that had relatively nutrient-rich soil. In suboptimal environments as location G, it would be beneficial for plants to be able to have a mutualistic relationship with microbes that can release the precipitated P to make it available to the plant. Fungal endophytes have been shown to increase plant P uptake ( Behie & Bidochka, 2014 ), and Hartmann et al. (2009) showed that roots exude a variety of compounds giving the plants a level of selectivity over potentially endophytic microbes. The potential of microorganisms for synthesis and release of pathogen-suppressing metabolites have been shown in previous in vitro studies ( Vassilev, Vassileva & Nikolaeva, 2006 ), strongly supporting manipulation of soil microbiota for agricultural applications. This could explain the improved fitness in terms of increased size of plants at location G despite its poor-nutrient soil ( Table 3 and Fig. 4 ). 10.7717/peerj.2860/table-3 Table 3 Soil analysis data from each location. Location p H a NH \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{4}^{+}$\\end{document} 4 + (ppm) b NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{2}^{-}$\\end{document} 2 − (ppm) b NO \\documentclass[12pt]{minimal}\n\\usepackage{amsmath}\n\\usepackage{wasysym} \n\\usepackage{amsfonts} \n\\usepackage{amssymb} \n\\usepackage{amsbsy}\n\\usepackage{upgreek}\n\\usepackage{mathrsfs}\n\\setlength{\\oddsidemargin}{-69pt}\n\\begin{document}\n}{}${}_{3}^{-}$\\end{document} 3 − (lb acre −1 ) b P (lb acre −1 ) b Ca 2+ (ppm) a Humus (%) a A 7.9 ± 0.1 0 <1 <10 <10 3,500 9.1 ± 1.4 B 8.4 ± 0.1 <5 <1 15 25 17,500 11.8 ± 0.5 C 7.9 ± 0.0 <5 <1 15 75 17,500 11.5 ± 2.2 D 8.3 ± 0.1 <5 <1 15 75 17,500 11.7 ± 2.0 E 8.6 ± 0.0 <5 <1 ≤10 75 25,000 3.9 ± 1.6 F 8.6 ± 0.1 <5 <1 <10 10 43,750 8.4 ± 2.8 G 8.9 ± 0.2 <5 <1 <10 <10 43,750 2.7 ± 1.2 Notes. a Average shows mean ± SE of three samples. b Average shows mode of three samples. The significance of soil pH was found to be critical for the presence of a specialised sebacinalean endophyte ( López-García, 2016 ) while Bååth & Anderson (2003) observed fungal biomass increased with increasing soil pH. Cultivated strawberry often suffer from chlorosis when grown in high pH soil ( Yao & Guldan, 2015 ), but wild strawberry is often found on chalky land ( Grubb, Green & Merrifield, 1969 ). It is possible that Fragaria vesca has an established symbiosis with endophytes that help with nutrient availability in high pH conditions. The pH gradient observed from locations A to G suggests a longer period of topsoil accumulation at A. At location G the more recent exposure of chalk bedrock is still evident, including visible chalk in the substrate. LaMotte Chemical Products (1967) states that 75 lb acre −1 of readily available phosphorus is the minimum required level for agricultural use, and 200 to 300 lb acre −1 or more is desirable. Location G had a very low P content (<10 lb acre −1 ) compared to other locations ( Table 3 ). It is therefore noteworthy that the F. vesca plants at G were apparently as healthy as at any of the other collection locations, with the highest amount of roots and the longest root. Vigorous growth of long and dense roots is an important prerequisite for efficient acquisition of macro and micro nutrients ( Wang et al., 2016 ) and root hair length is critical for the acquisition of limited nutrients such as P ( Brown et al., 2012 ). Valentinuzzi et al. (2015) reported that P and iron deficiency had a positive effect on the nutritional content of strawberry fruits without negative effects on fruit yield and quality. The above finding suggest that it could be possible to grow strawberry in nutrient-limited soils enriched with endophytes to improve the resilience and qualitative aspects of the crop. Potentially beneficial fungal endophytes among ‘rare’ fungi Majority of ‘rare’ fungi identified from both locations F and G are of interest for further detailed studies. The isolate Humicola sp.-like F2A(13), recovered from location F, showed to be most closely related to Humicola spp. (Sordariales, Chaetomiaceae), a fungal genus found to confer benefits to plants ( Lang et al., 2012 ; Radhakrishnan et al., 2015 ). Some Humicola species are asymptomatic root fungi such as H. fuscoatra in tomato roots ( Menzies et al., 1998 ), while in other cases they showed promise to be useful in the biological control of plant pathogens ( Ko et al., 2011 ; Piper & Millson, 2012 ; Wicklow et al., 1998 ; Wicklow, Jordan & Gloer, 2009 ). Yang et al. (2014) also discussed its role as a biocontrol agent against Phytophthora spp. Isolates, including Volutella rosea -like G3B(10), with 100% ITS sequence match to Volutella rosea were found only in location G. Volutella is able to solubilize and mineralize P from inorganic and organic pools of total soil P to make them available to plant roots ( Vassilev, Vassileva & Nikolaeva, 2006 ; Richardson, 2001 ). Carnivorous plants that live in nutrient-poor and stressful environments were found to benefit from the presence of Volutella ( Quilliam & Jones, 2012 ). Another endophyte, most closely related to Paraphoma sp. (Pleosporales, Pleosporaceae) was found in both F and G locations. Zhang et al. (2012) found that a specific strain of Paraphoma improved the biomass of its host plant. Endophytes related to other fungal genera of interest include Trichoderma , Knufia and Exophiala . Trichoderma can stimulate plant growth and defence responses have already been widely used for biological control of plant diseases (e.g., Harman et al., 2004 ; Druzhinina et al., 2011 ); Knufia and Exophiala species have been found to be highly stress-tolerant fungi in hot and arid environments ( Zakharova et al., 2014 ). Improvement of plant growth in culture by Humicola sp.-like F2A(13) and Volutella rosea -like G3B(10) Biomass of rye and common vetch was influenced when plants were grown with both Humicola sp.-like F2A(13) and V. rosea -like G3B(10) isolates. As a direct impact of endophyte inoculation, the fresh weight root/shoot ratio of both rye and common vetch was increased ( Fig. 5 ). Wali et al. (2009) and Ding, Kupper & McNear Jr (2015) also found higher root/ shoot ratios in different species of Festuca with endophytes compared to endophyte-free plants. On the other hand, the root/shoot ratio of dry weight in both plant species were less in the presence of endophytes. Although the root dry weight/fresh weight percentage was significantly lower in endophytic treatments, this was probably so due to longer and thinner roots in rye and greater branching in common vetch plants. These conditions, on the other hand, may have effectively improved the surface area when compared to controls, thereby allowing a relative increase in nutrient uptake. Association with beneficial fungi stimulates increased growth of plants, presumably by increasing nutrient acquisition of the plant. In our study, both endophytic isolates tested in vitro promoted increase in root fresh weight. A variety of factors may be involved in the stimulation of plant growth when P availability is low, one of the critical variables being root architecture ( Burridge & Jochua, 2016 ). Endophytic fungi may effectively improve P absorption by altering root architecture ( Ding, Kupper & McNear Jr, 2015 ). The significant increase in shoot dry weight and in root/shoot fresh weight percentage in both rye and common vetch have apparently highlighted this aspect ( Fig. 5 ). The increase in the ratio of shoot dry weight/fresh weight percentage also suggests the improved quality of the shoot system in both plants. Interaction with fungi may be one of the reasons why plants from location G had longer roots in the wild. It seems reasonable to suggest that plants with long and dense root systems can compete more effectively for nutrients both in situ and in vitro as found in this study. Effect of Humicola sp.-like F2A(13) endophytic isolate on salt tolerance Plants from the Brassicaceae family are known to have no mycorrhizal association, so it was envisaged that other root endophytes, including those derived from other species, might have particular significance for this family of plants ( Card et al., 2015 ). In their review of Brassicaceae endophytes, Card et al. (2015) noted that endophytes recovered from crop wild relatives (CWRs) of brassica are beneficial for different characteristics related to growth and tolerance to stresses in plants. In this work, the brassica plant radish was studied for salt tolerance and water use efficiency under in vitro conditions using inoculation with the Humicola sp.-like F2A(13) and V. rosea -like G3B(10) endophytic isolates. Our results indicated that both endophytes helped radish to use water more efficiently. Since evaporation from the medium surface and transpiration through plant tissue should be the only routes of water loss from the medium (escaping through the Magenta ® vessel B caps which allow passive air exchange), the rate of transpiration was presumably influenced by the presence of fungi. Plants co-cultured with Humicola sp.-like F2A(13) had greater shoot length and might therefore be expected to lose more water through a greater plant surface area. Neverthless, these plants depleted the least amount of medium volume ( Fig. 5 ). Radhakrishnan et al. (2015) found that a Humicola sp. isolate significantly increased shoot length and protein content in salt-stressed plants of soybean. In agreement with these data, our results suggest that both Humicola sp.-like F2A(13) and V. rosea -like G3B(10) isolates are promising fungi to be further explored in their potential ability to confer salt-stress tolerance. Research in using co-inoculation with multiple microorganisms rather than with a single inoculant can improve plant yield ( O’Callaghan, 2016 ). We suggest that the use of other endophytes identified from F and G locations could also be explored further in terms of potential bioinoculants for different crops and model plant species, using in vitro and ex vitro systems, either individually or in combinations. Endophytes of CWRs along environmental gradients Suboptimal environments offer opportunities to explore and understand the fungal diversity that could be useful for agricultural applications. For example, novel endophytes of plants growing in geothermal soils have been shown to confer high temperature tolerance to the host plants ( Rodriguez & Redman, 2008 ; Zhou et al., 2015 ). The majority of the CWRs originate from diverse habitat types, including hotspots in the subtropical and tropical regions of the world ( Castañeda-Álvarez et al., 2016 ). Studying CWRs and their root-derived and foliar endophytes along environmental gradients could help uncover a whole new array of beneficial fungi. Detailed studies on the benefit of fungi sourced from P-limited soil types (e.g., from F and G locations) are ongoing to understand whether selected endophytes can support better growth of plants on P-limited growing media."
} | 5,425 |
39511591 | PMC11545793 | pmc | 2,392 | {
"abstract": "Background Microalgae have emerged as sustainable alternatives to fossil fuels and high-value petrochemicals. Despite the commercial potential of microalgae, their low biomass productivity is a significant limiting factor for large-scale production. In the photoautotrophic cultivation of microalgae, achievable cell density levels depend on the light transmittance of the production system, which can significantly decrease the photosynthetic rate and biomass production. In contrast, the mixotrophic cultivation of microalgae using heterotrophic carbon sources enables high-density cultivation, which significantly enhances biomass productivity. The identification of optimal production conditions is crucial for improving biomass productivity; however, it is typically time- and resource-consuming. To overcome this problem, high-throughput screening (HTS) system presents a practical approach to maximize biomass and lipid production and enhance the industrial applicability of microalgae. Results In this study, we proposed a two-step HTS assay that allows effective screening of heterotrophic conditions compatible with new microalgal isolates. To confirm the effectiveness of the HTS assay, three microalgal isolates with distinctive morphological and genetic traits were selected. Suitable cultivation conditions, including various heterotrophic carbon sources, substrate concentrations, and temperatures, were investigated using a two-step HTS assay. The optimized conditions were validated at the flask scale, which confirmed a significant enhancement in the biomass and lipid productivity of each isolate. Moreover, the two-step HTS assay notably enhanced economic and temporal efficiency compared to conventional flask-based optimization. Conclusions These results suggest that our two-step HTS assay is an efficient strategy for investigating and optimizing microalgal culture conditions to maximize biomass and lipid productivity. This approach has the potential to enhance the industrial applicability of microalgae and facilitate the seamless transition from laboratory to field applications. Supplementary Information The online version contains supplementary material available at 10.1186/s12934-024-02550-7.",
"conclusion": "Conclusions The two-step HTS assay described herein was shown to be an effective strategy for optimizing microalgal mixotrophic conditions to achieve optimal production of biomass and lipids. Using this assay, we optimized the mixotrophic culture conditions for novel microalgal strains, confirming a significant enhancement in biomass and lipid productivity. While this approach is partly dependent on the microalgal species, the variety of organic substrate, and the instruments employed, it is undoubtedly a time- and cost-effective strategy compared to conventional optimization methods. Furthermore, it can simplify the transition from laboratory-scale experiments to field applications, suggesting its potential to contribute significantly to the industrial applicability of microalgae.",
"discussion": "Results and discussion Isolation of three novel microalgae strains Large-scale microalgae cultivation facilities for biodiesel production are typically exposed to the external environment to facilitate solar energy utilization, resulting in irregular cultivation conditions that are strongly influenced by abiotic factors, such as the temperature and amount of sunshine [ 24 ]. Therefore, isolating superior species with a high tolerance to various abiotic conditions is crucial for large-scale industrial applications of microalgae [ 25 ]. Previously, it has been reported that microalgae isolated from regions with environmental fluctuations have outstanding adaptability to various conditions. It is anticipated that microalgae species are well adapted to various conditions and have a robust capacity for valuable material production in environmentally dynamic region [ 26 , 27 ]. Thus, we aimed to isolate a novel microalgal species with excellent adaptability and biodiesel production capacity from harsh and environmentally variable areas, such as polluted urban streams. To isolate novel microalgal strains, sequentially diluted environmental samples were spread onto BG11 agar plates. Separated colonies were clearly obtained in samples diluted 100- to 1,000-fold. To isolate uncontaminated strains, we selected colonies based on several criteria, including robust growth, green color, large size, and clear boundaries. The 18 S rRNA gene was amplified and sequenced to identify each strain. The results revealed an overall predominance of the genera Chlorella and Desmodesmus . However, we focused on the other strains, because these two genera are well known to be fertile and have distinct heterotrophic properties [ 25 , 28 ]. Only three strains did not belong to the genera Chlorella and Desmodesmus during the initial screening. For further experiments, we finalized three strains that have not been studied and exhibit have distinct morphological and taxonomic characteristics (Fig. 1 ; Table 1 ). KGG-7, identified as Chlamydomonas sp., showed morphological characteristics similar to those of the well-known laboratory strain Chlamydomonas reinhardtii (Fig. 1 a). The KGG-9 strain showed an atypical sickle-shaped morphology and was taxonomically identified as Monoraphidium sp. by 18 S rRNA sequencing (Fig. 1 b). Monoraphidium sp. has emerged as an important strain for biodiesel production because of its high lipid content [ 29 ]. KGG-18 cells exhibited multiple unicellular characteristics, with a wrinkled cell surface and cell-to-cell clumping (Fig. 1 c). It was identified as Hariotina sp., which has not been studied until recently [ 30 ]. It is difficult to establish initial culture conditions for newly isolated microalgae that have not been studied extensively. Thus, we considered it suitable for validation of our two-step HTS assay and utilized it in subsequent experiments. \n Fig. 1 Light micrograph (400X) of microalgal isolates. (a) \n Chlamydomonas sp. KGG-7, (b) \n Monoraphidium sp. KGG-9, and (c) \n Hariotina sp. KGG-18 \n \n Table 1 Identification of isolated microalgal strains Strain Closest species (% Identity of 18 S rRNA) Assembled 18 S rRNA length (bp) GenBank Accession number KGG-7 Chlamydomonas sp. CCAP 11/132 (99.82%) 1741 OM218994 KGG-9 Monoraphidium convolutum strain AS7-3 (99.94%) 1741 OM218993 KGG-18 Hariotina sp. QW-2010a strain FACHB-2320 (99.96%) 2792 OM218995 \n Heterotrophic carbon source screening To identify the optimal heterotrophic carbon source for microalgae, the preference for carbon sources must be examined for each strain. However, this process is labor-intensive and requires considerable effort. To simplify and perform the screening process rapidly, we used a Biolog GENIII microplate, which is a commercially developed microbial profiling system. This microplate has 71 organic substrates in each well, enabling us to analyze the availability of these substrates with a single incubation. To explore its availability, inocula of Chlamydomonas sp. KGG-7 and Monoraphidium sp. KGG-9, and Hariotina sp. KGG-18 were prepared in a flask culture, diluted, and dispensed into each well. After cultivation, different cell growth were observed in each well, and the cell concentrations were determined by absorbance measurements. To normalize the raw values, the absorbance values measured in the wells containing the organic carbon substrate were divided by the absorbance values measured in the wells without the organic substrate and these values are plotted in Fig. 2 . The mean and standard deviation of the values are represented in a box plot. We conducted four replicate experiments and measured absorbance twice in each experiment. To identify the organic substrates favorable for heterotrophy, we focused on samples in which the final cell concentration was more than doubled. In Chlamydomonas sp. KGG-7, the final cell concentration increased by 2.12-fold only in the presence of α-ketoglutaric acid, whereas none of the other organic substrates caused an increase exceeding two-fold (Fig. 2 a). α-Ketoglutaric acid is known as a key intermediate in the tricarboxylic acid cycle and is involved in cellular energy supply and the metabolism of carbon and nitrogen [ 31 ]. Generally, α-ketoglutaric acid has been widely studied for dietary supplementation in human health due to its clinical effects [ 32 ]. It is used as an agent that increases antioxidant activity for protection against oxidative stress in humans and animals [ 33 ]. Recently, it has also been reported that the supplementation of α-ketoglutaric acid in microbial fermentation significantly enhances the production of ɛ-polylysine, which is a high-value product used as a food preservative, emulsifying agent, and enhancer of anticancer agent, suggesting that α-ketoglutaric acid can be used as a supplement to promote microbial cell growth [ 34 ]. However, α-ketoglutaric acid utilization as a carbon source as well as supplements in Chlamydomonas has not been reported previously. Therefore, α-ketoglutaric acid was considered unsuitable as a heterotrophic organic carbon source in microalgal culture. The results also showed that similar levels of absorbance for most heterotrophic substrates, suggesting the need for further modification of the conditions for Chlamydomonas sp. when utilizing the Biolog GENIII microplate. To further optimize, we excluded Chlamydomonas sp. KGG-7 in the subsequent screening experiments. \n Fig. 2 Heterotrophic carbon source screening for novel microalgae strains. (a) \n Chlamydomonas sp. KGG-7, (b) \n Monoraphidium sp. KGG-9, and (c) \n Hariotina sp. KGG-18. The normalization of raw values was conducted by scaling to each negative control using following formula, OD S /OD N . OD S is the optical density of each well with different organic substrate, and OD N is the optical density without the organic substrate (negative control). Experiments were conducted in four-repeated and error bars indicate standard deviation of mean \n In Monoraphidium sp. KGG-9, numerous organic substrates resulted in a more than two-fold increase in cell concentration. Therefore, we focused on organic substrates that increased the cell concentration by more than four times. When cellobiose, sucrose, and glucose were used, the cell concentration increased by 4.84-fold, 4.83-fold, and 4.40-fold, respectively (Fig. 2 b). Monoraphidium sp. is considered a promising candidate for biofuel production because of its notable lipid-producing ability, indicating that the heterotrophic properties of Monoraphidium sp. have been relatively well studied [ 29 , 35 ]. In previous studies, glucose has been frequently utilized as an organic substrate for the heterotrophic cultivation of Monoraphidium ; however, the optimal concentration of glucose needs to be further investigated. Compared to sucrose and cellobiose, glucose has several advantages as a substrate for mixotrophic cultivation, such as broad usability, monosaccharide characteristics, and industrial applicability. Thus, we selected glucose as a screening substrate for Monoraphidium sp. KGG-9, and optimized its concentration in subsequent experiments. For Hariotina sp. KGG-18, we filtered and selected the substrates that exhibited a double or greater increase in cell concentration. The cell concentrations showed a 2.73-fold increase in the wells supplied with lactose and a 2.53-fold increase in the wells supplied with maltose (Fig. 2 c). These results indicate that lactose and maltose can be candidates for subsequent experiments. Hariotina was initially classified as a genus distinct from Coelastrum based on its morphological characteristics. However, it was later transferred to Coelastrum in 1899 and was reinstated as Hariotina in 2002 [ 36 ]. Hariotina sp. has not been well studied, either taxonomically or physiologically, suggesting that it needs to be examined experimentally. To accumulate an experimental dataset and precisely investigate the substrate preferences of mixotrophic cultures, we conducted additional experiments using both lactose and maltose. Overall, we successfully selected the organic substrate candidates preferred by the newly isolated microalgal strain through microplate-based screening with minimal repeated experiments, suggesting that this is a promising and time-efficient approach for the characterization of novel strains. HTS of temperature and organic substrate concentration Specific optimization of the cultivation conditions is necessary to maximize the productivity of microalgal biomass and valuable materials. For photosynthetic microalgal growth, abiotic parameters such as temperature, light intensity, and CO 2 supply are generally considered the most important [ 37 ]. Temperature is also an important parameter in heterotrophic and mixotrophic cultures. Determining the optimal heterotrophic carbon substrate concentration is crucial because overfeeding of organic substrates can inhibit cell growth [ 38 ]. Similar to screening for organic substrate preferences, optimizing culture conditions using a flask-by-flask approach is labor-intensive. To reduce the input labor and conduct a time- and cost-efficient optimization process, we utilized the previously developed PhotoBiobox system [ 22 ]. This system comprises a high-throughput photobioreactor that enables the precise control of temperature, light intensity, and gas supply at the microplate scale. To determine the optimal organic substrate concentration and temperature, each row of a 96-well microplate was configured with an organic substrate concentration ranging from 0 to 30 g L − 1 , and each column of a 96-well microplate was adjusted within the range of 15 to 40 ℃ using the PhotoBiobox. After allowing sufficient time for cell growth, the absorbance of each well was measured to assess microalgal growth in response to the temperature and organic substrate concentration. When glucose was used as the carbon source by Monoraphidium sp. KGG-9, relatively high absorbance was observed in the temperature range of 25 to 27 °C and the glucose concentration range of 10 to 20 g L − 1 (Fig. 3 a). The maximum and minimum values over the entire plate were 1.67 and 0.47, respectively, indicating at least a three-fold increase in biomass production as a result of optimal temperature adjustment and organic substrate supplementation. These results represented that the optimal conditions for temperature and glucose concentration may be within ranges showing relatively high absorbance, suggesting that further investigations at the flask scale are necessary for a more detailed optimization. \n Fig. 3 High-throughput screening of optimal temperature and carbon substrate concentration using PhotoBiobox. The columns were set to range from 15.0 ℃ to 40.0 ℃ and the rows were set to range from 0 to 30 g L − 1 of substrate concentration. The values represent optical density measured by absorbance at 700 nm. (a) Optimization of glucose concentration and incubation temperature for Monoraphidium sp. KGG-9. (b) Optimization of maltose concentration and incubation temperature for Hariotina sp. KGG-18 \n Maltose and lactose were selected as promising heterotrophic substrates for Hariotina sp. KGG-18 through GENIII-based organic substrate screening. Subsequently, we examined to simultaneously determine the optimal substrate concentration and temperature. The absorbance values were comparatively higher in the maltose concentration range of 15 to 30 g L − 1 and the temperature range of 27 to 31 °C (Fig. 3 b). The maximum and minimum absorbance values were 0.61 and 0.19, respectively, suggesting a 3.2-fold increase in biomass production due to temperature adjustment and organic substrate supplementation. These results indicated that these maltose concentrations and temperatures may be optimal for mixotrophic culture of Hariotina sp. KGG-18. When lactose was used as the organic carbon source, no significant differences in microalgal cell growth were observed (Additional File 1: Figure S1 ). This result contradicts the result shown in Fig. 2 c, suggesting the potential of lactose as a carbon source. The difference between the results may be attributed to the unclear lactose concentration of the GENIII microplate and the insufficient culture conditions provided by the PhotoBiobox for lactose metabolism. For example, efficient cultivation with lactose requires adequate aeration and agitation [ 39 ], but the inner part of the PhotoBiobox is too enclosed, resulting in insufficient aeration. Due to these limitations, lactose was eliminated as a candidate and instead maltose was selected as the optimal heterotrophic substrate for mixotrophic cultivation of Hariotina sp. KGG-18. Despite the exceptional case, these results confirmed the optimal heterotrophic carbon source and temperature range through an HTS experiment without labor-intensive and time-consuming processes. Validation of the two-step HTS assay To validate the two-step HTS assay, flask cultivation was performed under the optimal conditions obtained through the two-step HTS assay. The optimal conditions were determined within a specific range; however, flask cultivations were performed at the middle point of the optimal temperature range since precise temperature control down to 1 °C is unlikely to be meaningful in typical incubators. The concentration of organic carbon substrates can be easily adjusted and is very important for cost effectiveness, so we evaluated the maximum and minimum values within the optimal range identified by the screening experiments. In addition, mixotrophic conditions were compared with photoautotrophic conditions to explore the sole effects of organic substrate addition. To investigate the optimal glucose concentration for Monoraphidium sp. KGG-9, glucose was supplemented into flasks at concentrations of 0, 10, or 20 g L − 1 and the flasks were incubated at 26 °C. Microalgal cell growth in the flask culture was determined by measuring the optical density. The sample without glucose supplementation was used as a control. When glucose was provided at 10 g L − 1 and 20 g L − 1 , cell growth reached 4.84 ± 0.081 and 4.7 ± 0.177, respectively (Fig. 4 a). These values were approximately six-fold higher than the value of 0.72 ± 0.004 observed in the absence of glucose supplementation. The final biomass production was calculated by measuring the weight of the dried cells. When glucose was provided at 10 g L − 1 and 20 g L − 1 , the maximum dried cell weights reached 4.25 ± 0.212 g L − 1 and 4.35 ± 0.212 g L − 1 , respectively (Fig. 4 b). These values were approximately three times higher than the concentration of 1.5 ± 0.141 g L − 1 measured without glucose supplementation. When glucose is added, the average time required for Monoraphidium sp. KGG-9 to double the biomass is 2 days. However, there was no significant difference in cell growth between the 10 g L − 1 and glucose 20 g L − 1 conditions. When glucose consumption was measured, the concentration of 5.09 ± 0.03 g L − 1 and 5.01 ± 0.85 g L − 1 decreased in the medium, respectively (Fig. 4 c). The growth of Monoraphidium sp. KGG-9 dramatically increased with glucose supplementation. However, only approximately 5 g L − 1 of glucose was consumed during cultivation, regardless of the initial glucose concentration. This result was consistent with the results shown in Fig. 3 a, which showed sufficiently high growth at 25 °C and 5 g L − 1 glucose with no significant differences when compared to glucose concentrations of 10 g L − 1 and 20 g L − 1 . \n Fig. 4 Comparison of microalgal growth, biomass production, and substrate consumption on flask cultivations. (a) Microalgal cell growth by measuring optical density, (b) biomass production, and (c) glucose consumption in Monoraphidium sp. KGG-9. (d) Microalgal cell growth by measuring optical density, (e) biomass production, and (f) maltose consumption in Hariotina sp. KGG-18. Experiments were conducted in triplicate and error bars indicate standard deviation of mean \n To optimize the organic substrate concentration for Hariotina sp. KGG-18, maltose was used in accordance with our preliminary results. Based on the results shown in Fig. 3 b, Hariotina sp. KGG-18 was cultivated at 29 °C with supplementation of 0, 15, or 30 g L − 1 of maltose. Microalgal cell growth in the flask culture was determined by measuring the optical density. The sample without maltose supplementation was used as a control. When maltose was supplied at 15 g L − 1 and 30 g L − 1 , the cell growth reached 0.71 ± 0.034 and 1.04 ± 0.062, respectively. These values were approximately 2.29-fold and 3.35-fold higher than the value of 0.31 ± 0.012 observed in the absence of maltose supplementation (Fig. 4 d). The final biomass concentrations after supplementation with 15 g L − 1 and 30 g L − 1 maltose reached 2.15 ± 0.071 g L − 1 and 3.2 ± 0.141 g L − 1 , respectively. These concentrations were also 1.39-fold and 2.06-fold higher than the concentration of 1.55 ± 0.212 g L − 1 measured in the absence of maltose supplementation (Fig. 4 e). When maltose is added at concentration of 15 g L − 1 and 30 g L − 1 , the time required for Hariotina sp. KGG-18 to double its biomass is 1.47 days and 2.23 days, respectively. When supplied with 15 g L − 1 of maltose, only 1.65 ± 0.25 g L − 1 of maltose was consumed, whereas 8.01 ± 0.86 g L − 1 of maltose was consumed when maltose was provided at a concentration of 30 g L − 1 (Fig. 4 f). Maltose consumption was approximately five times higher when 30 g L − 1 maltose was supplied. The performance of flask culture in the presence of 30 g L − 1 maltose was enhanced, in contrast to the comparable growth observed for maltose concentrations of 15 g L − 1 and 30 g L − 1 in the PhotoBiobox screening. This was attributed to greater substrate consumption in 30 g L − 1 maltose condition. However, the increase in biomass was relatively limited compared to the total amount of maltose provided. In this study, we focused on the optimization of organic carbon source utilization based on the general abiotic conditions only varying temperature, so it will remain as further challenges that how to improve the overall conversion yield from substrate to biomass. Several strategies, such as the alteration of microalgal physiology by modulating light, nutrients, and environmental conditions and the engineering by controlling carbon partitioning and energy route, can contributed to improve the efficient carbon conversion [ 40 ]. Overall, these findings showed a meaningful correlation between the screening results and flask validation, suggesting that our two-step HTS assay was highly effective at maximizing microalgae biomass production. To assess the effects of organic substrate supplementation on lipid production, the lipid content and productivity of each microalgal strain were measured (Table 2 ). The sample without organic carbon substrate supplementation was used as a control. For Monoraphidium sp. KGG-9, the lipid content was measured at 47.49 ± 2.11% in the absence of glucose, but 54.03 ± 0.25% and 56.59 ± 1.16% in the presence of 10 g L − 1 and 20 g L − 1 glucose, respectively. Compared to photoautotrophic cultivation, lipid content under mixotrophic cultivation slightly increased in Monoraphidium sp. KGG-9. In contrast, the lipid content under mixotrophic conditions was similar to or slightly lower than that under photoautotrophic conditions for Hariotina sp. KGG-18. When maltose was provided at 15 g L − 1 and 30 g L − 1 , the lipid content was measured at 35.31 ± 0.98% and 35.86 ± 0.96%, respectively. In the absence of maltose supplementation, the lipid content was measured at 36.09 ± 1.13%. According to previous reports, supplementation with heterotrophic carbon sources has been shown to reduce lipid accumulation in microalgae [ 29 , 41 ]. This appears to be consistent with our findings of Hariotina sp. KGG-18. In contrast, glucose supplementation promoted lipid accumulation and cell growth in Monoraphidium sp. KGG-9, resulting in a more than three-fold increase in final lipid productivity (Fig. 5 a). The dramatic increase in lipid production resulting from glucose supplementation is likely due to the redistribution of photosynthetic energy and improved carbon flux. Typically, microalgae generate acetyl-CoA, ATP, and NADPH by photosynthesis in the absence of organic substrates. In contrast, when organic carbon substrates are introduced, the energy consumption responsible for photosynthesis can be redirected to lipid biosynthesis, resulting in enhanced lipid production [ 42 ]. In addition, the increased carbon flux associated with additional glucose supplementation can lead to increased levels of glycerol-3-phosphate (G3P), which further promotes lipid biosynthesis [ 43 , 44 ]. Taken together, organic substrate supplementation would be either positive or negative for lipid content, but this cannot be easily determined without experimental evidence because it involves complex physiological changes in each microalgal species. Nevertheless, heterotrophic supplementation usually enhances cell growth greatly, so that the final lipid productivity would be quite improved. \n Table 2 Comparison of biomass and lipid productivity in novel microalgal strains Strain Culture condition Final biomass yield (g L − 1 ) Lipid content (%) Biomass productivity (g L − 1 d − 1 ) Lipid productivity (mg L − 1 d − 1 ) Monoraphidium sp. KGG-9 Without glucose 1.50 ± 0.10 47.49 ± 2.11 0.17 ± 0.01 78.98 ± 3.954 Glucose 10 g L − 1 4.25 ± 0.15 54.03 ± 0.25 0.47 ± 0.02 255.09 ± 11.53 Glucose 20 g L − 1 4.35 ± 0.15 56.59 ± 1.16 0.48 ± 0.02 273.37 ± 7.74 Hariotina sp. KGG-18 Without maltose 1.45 ± 0.15 36.09 ± 1.13 0.13 ± 0.01 47.69 ± 8.45 Maltose 15 g L − 1 2.15 ± 0.05 35.31 ± 0.98 0.20 ± 0.00 69.04 ± 4.18 Maltose 30 g L − 1 3.20 ± 0.10 35.86 ± 0.96 0.29 ± 0.01 104.26 ± 1.83 \n \n Fig. 5 Comparison of lipid productivity between photoautotrophic and mixotrophic cultivation. (a) Lipid productivity in Monoraphidium sp. KGG-9. (b) Lipid productivity in Hariotina sp. KGG-18. Experiments were conducted in triplicate and error bars indicate standard deviation of mean. Asterisks represent statistically significant difference, as determined by a Student t -test (* P < 0.05, ** P < 0.01, and *** P < 0.001) \n Indeed, in the case of Hariotina sp. KGG-18, the final lipid productivity in mixotrophic cultivation using maltose was increased by up to two-fold, which was attributed to increased cell growth (Fig. 5 b). Not only in the case of Hariotina sp. KGG-18, but also in the case of Monoraphidium sp. KGG-9, the lipid productivity was greatly enhanced by the synergistic effect of increased lipid content and cell growth resulting from optimized organic substrate supplementation. In previous studies, it was reported that lipid production was regulated through the combination of temperature and light intensity in Monoraphidium dybowskii Y2, achieving a biomass yield of 1.79 g L − 1 and a lipid productivity of 66.17 mg L − 1 d − 1 under autotrophic conditions [ 45 ]. Monoraphidium sp. FXY-10, which was newly isolated in Lake Fuxian, also exhibited biomass yield at 3.96 g L − 1 and lipid productivity at 148.74 mg L − 1 d − 1 in heterotrophic culture conditions using glucose [ 29 ]. Compared to Chlorella strains, which are well-known for biofuel production strains, Monoraphidium sp. KGG-9 show the higher biomass yield and lipid productivity. Chlorella zofingiensis was cultivated in in 60 L flat panel photobioreactors and exhibited a biomass yield of 1.587 ± 0.016 g L − 1 and lipid productivity of 22.30 ± 0.90 mg L − 1 [ 46 ]. Chlorella protothecoides showed a lipid productivity of 177.3 mg L − 1 through two-stage fed-batch culture using optimized major nutrient conditions involving carbon, nitrogen, and phosphorus sources [ 47 ]. In this study, we achieved the biomass yield of 4.35 g L − 1 and the lipid productivity of 273.37 mg L − 1 d − 1 in Monoraphidium sp. KGG-9 through the comprehensive optimization of organic carbon substrate supplementation, considering the type and concentration of substrate and temperature at the same time. In addition, fortunately, the novel strain KGG-9 also seems to have its own inherently superior performance, which allows achieving the highest lipid product record in microalgal species ever reported. These results indicate that our two-step HTS assay is practical for investigation of optimal culture conditions which can maximize the productivity of biomass and target products. Hariotina sp. has not been extensively studied so far; however, our results provide valuable experimental data for further research. This suggests that our two-step HTS assay can be utilized to rapidly investigate the cultivation characteristics of unexplored microalgae. In summary, the two-step HTS assay was effective at enhancing lipid productivity by optimizing mixotrophic culture conditions, suggesting the enhancement of the industrial applicability of novel microalgae. Typically, the capacity of microalgae to utilize heterotrophic carbon sources is highly dependent on the microalgal strain [ 48 ]. In industrial-scale production, light independence through the supplementation of organic sources can considerably reduce production costs and space requirements [ 49 ]. This suggests that the novel microalgal strains isolated in this study have the potential to be utilized as industrial microalgal strains due to they showed excellent growth and lipid production using heterotrophic sources. To further minimize the costs of industrial bioprocesses using microalgae, it is essential to substitute the current heterotrophic carbon substrate with low-cost carbon sources, such as waste feedstocks. Food wastes, including wastewater from food processing, anaerobic digestion wastewater, and food residues, are considered sustainable growth media for microalgal biorefinery production [ 50 ]. Monoraphidium littorale exhibited significantly higher biomass production and lipid content when cultured with vegetable waste, such as digested rotten potato supernatant, than with the control medium [ 51 ]. The newly isolated Monoraphidium sp. SVMIICT6 has also been used to efficiently treat dairy wastewater, resulting in substantial biomass accumulation and elevated proportions of lipids and carbohydrates [ 52 ]. The nutrient-rich microalgae biomass produced through wastewater treatment can be utilized in various applications, such as biofertilizers and value-added products. Based on these findings and our study, Monoraphidium sp. KGG-9 is a suitable novel microalgal strain due to its high biomass and lipid productivity, as well as its potential for utilizing low-cost and sustainable substrates. It is also expected that further research utilizing sustainable low-cost substrates will additionally enhance the industrial applicability of newly microalgae strains. To confirm the economic and temporal efficiencies of the HTS assay, we compared the two-step HTS assay with the conventional flask-based optimization method in terms of consumable materials and costs (Table 3 ). In conventional methods, a substantial number of materials, including experimental equipment, media, and incubators, are required to optimize the different carbon substrate types, concentrations, and temperatures. However, the two-step HTS assay can handle diverse conditions simultaneously, thereby significantly reducing the amount of material and equipment required. The total duration of the optimization process using the two-step HTS assay was 8 days, whereas the flask-based method took 150 days. The estimated consumable cost for the entire optimization of the HTS assay was USD 21.30, whereas the flask-based strategy incurred a cost of USD 599.60. Additionally, the workforce cost was calculated based on time considerations for the overall optimization process. The two-step HTS required USD 464, indicating a significant reduction compared to the workforce cost of USD 8,700 for flask-based optimization. The HTS assay can effectively address several issues associated with conventional flask-based methods, including the management of a large number of flasks, demands for labor and resources, limited cultivation space, and the risk of contamination due to prolonged cultivation periods [ 53 , 54 ]. It also reduces the time required for sample preparation and enables the rapid detection of cell growth. Furthermore, these advantages also suggest that HTS assays are highly beneficial for strain selection. The HTS method allows for the rapid identification of optimal nutrient conditions across a large number of strains, reducing the number of experiments by more than 2,000 times [ 19 ]. Through this, the method allows for rapid and comprehensive exploration of different experimental conditions, which can minimize the overall experimental time and lead to significant cost savings [ 55 ]. Several challenges, such as variations between 96-well conditions and difference arising from scale-up processes, remain to extend the HTS assays and optimized conditions to industrial applications. To address these, integration of HTS assay and continuous automated experimental operations can be minimize error and increase consistency between experiment [ 56 ]. Machine learning-integrated hybrid optimization method can also be used to compensate for differences between experimental conditions and perform more fine-grained condition exploration [ 57 ]. Based on these results, the two-step HTS assay has spatial, temporal, and cost advantages over the conventional method, indicating that it is expected to be effectively utilized in the pre-optimization of industrial processes. Furthermore, this practical strategy can be applied to other industrial microorganisms and microalgae, indicating the potential to expand its usage across various industrial sectors. \n Table 3 Comparative analysis of two-step HTS assay and flask-based conventional method Parameters Two-step HTS Conventional* Number of materials for 71 different substrate types screening (in triplicate) 3 213 Number of materials for substrate concentration screening (in triplicate) 3 24 Number of incubators for 12 different temperature screening 1 12 Volume of media required for overall process (mL) 28.8 8350 Estimated time of total screening for overall process (day) 8 150 Estimated cost of consumables for overall process $ (USD) 21.30 599.60 Estimated cost of workforce for overall process # (USD) 464 8700 * Calculated based on a volume of 50 mL in a 125 mL flask, which could accommodate 25 flasks in one incubator, JSSI-200CL (JSR, Gongju, South Korea) $ For consumables for the overall process, BG-11 medium (Sigma-Aldrich Co., Saint Louis, MO, USA), a 125 mL Erlenmeyer flask with a vent cap (Corning, Corning, NY, USA), a GENIII plate (Biolog, Hayward, CA, USA), and 96 well plate (Thermo Fisher Scientific, Waltham, MA, USA) were used # Calculated based on US federal minimum hourly wage (US$ 7.25)"
} | 8,844 |
36652522 | PMC9848640 | pmc | 2,394 | {
"abstract": "Mycorrhizae are symbiotic associations between terrestrial plants and fungi in which fungi obtain nutrients in exchange for plant photosynthates. However, it remains unclear how different types of mycorrhizae affect their host interactions and productivity. Using a long-term experiment with a diversity gradient of arbuscular (AM) and ectomycorrhizal (EcM) tree species, we show that the type of mycorrhizae critically controls the effect of diversity on productivity. With increasing diversity, the net primary production of AM trees increased, but EcM trees decreased, largely because AM trees are more effective in acquiring nitrogen and phosphorus. Specifically, with diversity increase, AM trees enhance both nutrient resorption and litter decomposition, while there was a trade-off between litter decomposability and nutrient resorption in EcM trees. These results provide a mechanistic understanding of why AM trees using a different nutrient acquisition strategy from EcM trees can dominate in subtropical forests and at the same time their diversity enhances productivity.",
"introduction": "INTRODUCTION High plant diversity often promotes primary productivity ( 1 , 2 ). To maintain their growth and productivity and thus coexist, different species in plant communities often have complementary resource acquisition strategies ( 3 , 4 ). One of the most critical acquisition strategies for soil resources is mycorrhizal symbiosis, which can expand plant root surface area and thus plant access to nutrients ( 5 , 6 ). Nutrient resorption and litter decomposition provide ~90% of the annual nitrogen (N) and phosphate (P) needs for tree growth ( 7 , 8 ). However, how mycorrhizal plants balance or coordinate these two pathways to optimize their nutrient acquisition in high-diversity ecosystems is still unknown. There are two dominant mycorrhizal types associated with trees, arbuscular mycorrhizae (AM) and ectomycorrhizae (EcM) ( 9 ), which may regulate plant-soil feedbacks and drive plant diversity ( 10 , 11 ). While AM tree species generally experience negative feedback and dominate in high-diversity ecosystems, EcM tree species often exhibit positive feedback and promote monodominance or familial dominance ( 12 ). This difference has largely been attributed to different types of mycorrhizae. While AM fungi offer low root protection against soilborne pathogens and AM trees exhibit conspecific inhibition ( 10 – 12 ), EcM fungi form a mantle around tree roots that better protects against pathogens and facilitates conspecific species growth ( 10 – 12 ). At the same time, AM fungi have low host specificity and complement each other spatially during nutrient foraging in soils, thereby promoting plant coexistence and productivity ( 13 , 14 ). In contrast, EcM fungi have high host specificity and could establish a common fungal network to transfer nutrients and signals between conspecific trees, which usually promotes their dominance ( 15 , 16 ). Thus, plant-soil feedback theory may explain why AM trees are more diverse than EcM trees, but not why AM trees with lower resistance to pathogens should be more competitive than EcM trees in mixed subtropical forest ( 10 , 12 ). Our understanding of the drivers of AM trees’ competitive advantage is still inadequate. The contrasting nutrient acquisition strategies of AM and EcM plants may play important roles in regulating the relationship between plant diversity and productivity ( 4 , 9 , 17 – 19 ). Compared to AM plants, EcM plants supposedly favor a more conservative nutrient strategy and are often associated with higher resorption efficiency but lower litter decomposition rate ( 8 , 20 , 21 ). This difference may stem from the coevolutionary history of plants with fungi ( 22 ). The evolution from AM to EcM symbioses was accompanied by an increase in plant lignification and of symbiosis with lignin-degrading fungi such as Agaricomycetes ( 22 – 24 ). Therefore, EcM trees often have higher lignin concentrations and produce more recalcitrant litter than AM trees, resulting in slower litter decomposition and nutrient mineralization but higher soil organic matter accumulation ( 20 , 25 ). In a high-diverse ecosystem where nutrient competition intensifies, AM trees may rely more on rapid litter decomposition to gain inorganic nutrients ( 9 , 25 ). In contrast, EcM trees may be dependent more on EcM fungi to mine and absorb nutrients directly from the organic matter because EcM fungi retain degradative enzymes from their saprotrophic ancestors ( 9 , 26 ). Also, as the decomposition capacity of EcM fungi is lower than that of saprotrophic fungi ( 27 ), EcM plants may have to rely on other nutrient conservation strategies, such as nutrient resorption, to ensure their nutrient needs ( 7 ). When productivity increases with diversity, plant communities likely adjust their nutrient acquisition strategies via either physiological plasticity or species reordering ( 28 , 29 ). However, there is a lack of experimental evidence regarding how trees with different mycorrhizal symbioses regulate their nutrient resorption and litter decomposition for optimizing nutrient acquisition in high-diversity ecosystems, as AM and EcM trees rarely are planted together at the same site in subtropical areas. We used a controlled biodiversity-ecosystem functioning (BEF) experiment in a subtropical forest in Southeast China to study these issues. We hypothesized that as plant nutrient demand increases with diversity, AM and EcM trees use different strategies to increase their nutrient supply. More specifically, the acquisitive AM trees would mainly rely on inorganic nutrient supply to meet their growing nutrient demands via litter decomposition, while the conservative EcM trees would enhance aboveground nutrient resorption to alleviate their nutrient limitation.",
"discussion": "DISCUSSION Results from our long-term field experiment showed that as tree species richness increased, EcM tree biomass rapidly decreased, but AM tree biomass progressively increased, leading to the dominance (>90%) of AM trees in the highest diversity treatment ( Fig. 1 ). This finding is consistent with high dominance of AM trees in diverse tropical and subtropical forests ( 25 , 30 ) where EcM trees often account for less than 10% of tree stems ( 10 , 12 ). The high species diversity of AM trees is often attributed to conspecific negative density dependence (CNDD) driven by host-specific pests and pathogens ( 10 , 31 ). However, CNDD is unable to explain the dominance of AM over EcM trees because EcM fungi often provide better protection against pathogens and have high mycorrhizal abundance ( 10 , 12 ). Alternative mechanisms may underlie the dominance of AM trees over their EcM counterparts in high-diverse subtropical and tropical forests. The first and foremost role of mycorrhizal symbioses is to facilitate plant nutrient acquisition ( 32 , 33 ), which may be key to maintaining high productivity in highly diverse ecosystems ( 3 , 33 ). AM and EcM fungi have distinct nutrient acquisition strategies: While AM fungi are highly efficient in uptake of inorganic nutrients, EcM fungi evolved a unique capacity to acquire nutrients directly from organic matter ( 25 , 34 ). This difference may grant competitive advantages of AM trees over EcM trees in mixed forests through several mechanisms. First, competition for nutrients generally intensifies as the diversity increases, and AM fungi offer a competitive edge for their hosts over EcM fungi in nutrient acquisition. Soils in warm and humid subtropical and tropical regions are highly weathered and thus P limited as most P is bound with iron (Fe) and aluminum (Al) ( 32 , 35 ). Also, soils in these areas are less N limited because warm and humid climates favor litter decomposition and nutrient mineralization by saprotrophic microbes ( 36 ) and are conducive to biological N fixation ( 37 ). Although AM fungi have little saprophytic ability, they are uniquely efficient in obtaining less mobile nutrients such as P and Fe ( 32 ). For example, AM fungi can promote P-solubilizing bacteria, which release carboxylates and phosphatases to mobilize soil inorganic and organic P as well as micronutrients such as Zinc (Zn), Fe, and manganese (Mn) ( 38 ). Increased AM fungi diversity in highly diverse ecosystems could enhance nutrient acquisition such as decomposition process, thereby promoting productivity ( Fig. 4 ). In contrast, EcM fungi are able to directly mine nutrients (mainly for N) trapped in organic matter ( 39 ) but have a low ability to acquire mineral nutrients in soil ( 25 ). Consequently, P limitation and fast microbial decomposition of litter in high-diverse subtropical and tropical forests favor AM trees over EcM ones ( Fig. 2 ) ( 40 ). Second, the positive feedback between plant traits and decomposition rate would benefit AM trees more than EcM trees in the highly diverse subtropical forests ( 25 , 41 ). Compared with EcM trees, AM trees produce high-quality litter with low concentrations of lignin and secondary metabolites ( 30 ). With the increase in diversity, lignin concentration of AM litter further decreased, leading to an increase in AM litter decomposition ( Fig. 5 ). The increased litter decomposition promotes a more inorganic nutrient economy, which would benefit AM trees and their diversity more than EcM trees ( 9 , 25 ). A recent global-scale synthesis identified decomposition rate as the primary driver of the distribution of mycorrhizal symbioses in different climate zones ( 40 ). Our results provide direct evidence illustrating how microbial decomposition rate is related to mycorrhizal types in (sub)tropical forests. Another major but often overlooked mechanism affecting plant competition is the ability of competitive species to reuse nutrients via nutrient resorption. Nutrient resorption allows for the conservation of nutrients that may be otherwise returned to the soil via litterfall ( 42 ). Our results showed that compared with EcM trees, AM trees have higher nutrient resorption efficiency ( Fig. 2 and fig. S5). The predicted negative trade-off between litter decomposition and nutrient resorption occurred in EcM trees but not in AM trees. This likely is because AM plant litter decomposition is more affected by litter lignin concentration than by litter N and P concentrations ( Fig. 5 and fig. S13). Because leaf cell degradation is a prerequisite for nutrient resorption during leaf senescence ( 43 , 44 ), the decreased lignin concentration for AM trees not only increases litter decomposition but also weakens the lignin barriers and results in more “complete nutrient resorption” in high-diversity community ( 45 ). Consequently, high resorption further enhances nutrient reservoirs in AM trees, which allows fast growth and better competition against their EcM counterparts in a diverse community. An unexpected finding was that EcM trees enhanced litter decomposition but not nutrient resorption efficiency with increasing tree species richness. One possible explanation is that, in humid and warm subtropical forests, the decomposition pathway is always more beneficial to plants than the resorption pathway, because resorption requires plants to synthesize enzymes to degrade and remobilize leaf nutrients ( 45 ). Therefore, as competition increased with tree richness, EcM trees tended to reduce C investment in nutrient resorption, resulting in a lower N and P resorption. A global meta-analysis also showed that, in contrast to boreal and temperate ecosystems, EcM plants resorb less N than AM plants in subtropical and tropical habitats ( 46 ). In addition, we observed that EcM tree monoculture communities has a high abundance of symbiotic fungi but low pathogens, whereas AM tree monoculture communities were dominated by saprotrophic and pathogenic fungi ( Fig. 4 ). As tree richness increased, shifts in microbial communities, particularly a decrease in the relative abundances of Agaricomycetes, most of which are EcM fungi, corresponded with a decrease in NPP of EcM subcommunity but have a relatively small impact on NPP of AM subcommunity ( Fig. 5B ). These results imply that AM trees were less dependent on their fungal partners than EcM trees for nutrient acquisition and that the increasing dominance of AM trees hampered the formation of ectomycorrhizal networks in the mixed forests ( 15 , 47 ). In addition, the SEM also revealed a direct and positive effect of tree species richness on NPP of AM subcommunities. Compared to EcM subcommunities with the same species diversity, AM subcommunities in this study had a higher phylogenic diversity (table S1). Thus, our results further confirmed the previous finding at this site that tree species richness effects were strongly associated with phylogenic diversity ( 1 ). Together, our findings suggest that more efficient nutrient-acquiring strategies ( Fig. 6 ), rather than the microbial-mediated negative plant-soil feedback, drive the dominance of AM trees in high-diversity ecosystems. However, as a large proportion of operational taxonomic units (OTUs) were unclassified, better characterization of how mycorrhizal fungi mediate plant-soil feedback still relies on advances in sequencing technologies. Fig. 6. Conceptual model of nutrient strategies of AM and EcM trees. AM trees with high leaf quality exhibit nutrient acquisition strategies and often dominate in high-diversity communities, whereas EcM trees with low leaf quality show nutrient conservative strategies and generally dominate in monocultures. The differences in nutrient acquiring strategy and diversity pattern between AM and EcM trees could be associated with the coevolution of trees and fungi. The onset of AM-tree symbiotic association occurred before the emergence of lignin-degrading Agaricomycetes, making AM tree nutrient acquisition dependent on saprophytic microbes to degrade lignin. Post-emergence of Agaricomycetes, EcM fungi evolved from humus and wood saprotrophic ancestors and retained part of their enzymatic ability. Thereby, compared to AM trees, EcM trees tend to scavenge N trapped in the organic matter directly. Ma, million years. Our study provides direct evidence demonstrating that types of mycorrhizae critically mediate plant interactions and the biodiversity-productivity relationship in diverse forests. Our analyses further reveal that differences in mycorrhizal nutrient acquisition strategies, both nutrient acquisition from soil and nutrient resorption within the plant, contribute to the competitive edge of AM trees over EcM ones. These findings provide an alternative mechanism for explaining why and how AM trees usually dominate in high-diversity subtropical forests. Our results also have some implications for predicting the population dynamics and carbon balance of forest ecosystems in subtropical and tropical regions. Deforestation, urbanization, and agricultural practice have resulted in large-scale loss of EcM forests ( 48 ). Global changes such as N deposition and climate warming also favor AM trees over EcM trees ( 40 , 49 ) . The spread of AM trees in the future would accelerate the nutrient cycle via promoting decomposition, which potentially increases vegetation C sequestration at the cost of soil C stability ( 50 , 51 ). Together, our findings suggest that mycorrhizal responses to climate change factors need to be considered in forest management, particularly in the selection of tree species for plantation for both timber production and climate change mitigation."
} | 3,898 |
39497183 | PMC11533408 | pmc | 2,397 | {
"abstract": "Background Corals are the foundational species of coral reefs and coralligenous ecosystems. Their success has been linked to symbioses with microorganisms, and a coral host and its symbionts are therefore considered a single entity, called the holobiont. This suggests that there may be evolutionary links between corals and their microbiomes. While there is evidence of phylosymbiosis in scleractinian hexacorals, little is known about the holobionts of Alcyonacean octocorals. Results 16S rRNA gene amplicon sequencing revealed differences in the diversity and composition of bacterial communities associated with octocorals collected from the mesophotic zones of the Mediterranean and Red Seas. The low diversity and consistent dominance of Endozoicomonadaceae and/or Spirochaetaceae in the bacterial communities of Mediterranean octocorals suggest that these corals may have a shared evolutionary history with their microbiota. Phylosymbiotic signals were indeed detected and cophylogeny in associations between several bacterial strains, particularly those belonging to Endozoicomonadaceae or Spirochaetaceae , and coral species were identified. Conversely, phylosymbiotic patterns were not evident in Red Sea octocorals, likely due to the high bacterial taxonomic diversity in their microbiota, but cophylogeny in associations between certain coral and bacterial species was observed. Noteworthy were the associations with Endozoicomonadaceae , suggesting a plausible evolutionary link that warrants further investigations to uncover potential underlying patterns. Conclusions Overall, our findings emphasize the importance of Endozoicomonadaceae and Spirochaetaceae in coral symbiosis and the significance of exploring host-microbiome interactions in mesophotic ecosystems for a comprehensive understanding of coral-microbiome evolutionary history. Supplementary Information The online version contains supplementary material available at 10.1186/s42523-024-00351-2.",
"conclusion": "Conclusion While many bacteria associated with corals are host-specific, only a minority of bacterial phylotypes associated with corals display cophylogenetic patterns indicative of long-term host-microbe relationship, especially with Endozoicomonas . This result emphasizes the idea that, although host-microbe cophylogeny likely plays a role in phylosymbiosis, other factors, such as biogeographic influences, also play an important role in shaping this pattern. Furthermore, the different degrees of cophylogeny between coral microbes and their hosts underscore the fact that the microbiome is not a singular entity subject to uniform selection and that the abundance of the symbionts within the holobiont does not correlate with their common evolutionary history. Instead, it comprises a multitude of distinct participants with varying degrees of historical association with both the host and each other. This study reveals important patterns between coral hosts and their microbiota, offering valuable insight into their potential evolutionary history. However, to better understand whether these patterns indicate coevolution or are driven by other factors, such as vicariance, further research is needed. Investigating whether these patterns remain consistent throughout the coral’s life cycle and how environmental factors influence host-microbe interactions would add more clarity.",
"discussion": "Discussion We investigated the diversity and composition of bacterial communities associated with octocorals from the mesophotic zone of the Mediterranean Sea and Red Sea and assessed whether we can detect evolutionary patterns in these host-microbe associations. Our study revealed that Mediterranean and Red Sea octocoral holobionts harbor specific bacterial communities, but signals of phylosymbiosis were found only in octocorals from the Mediterranean Sea. However, indications of cophylogeny were detected in 13 of the 14 octocoral species investigated, suggesting that octocorals share a common evolutionary history with a few specific bacterial symbionts, primarily belonging to the Endozoicomonadaceae and Spirochaetaceae ."
} | 1,026 |
28381068 | PMC5380558 | pmc | 2,398 | {
"abstract": "A binary spike-time-dependent plasticity (STDP) protocol based on one resistive-switching random access memory (RRAM) device was proposed and experimentally demonstrated in the fabricated RRAM array. Based on the STDP protocol, a novel unsupervised online pattern recognition system including RRAM synapses and CMOS neurons is developed. Our simulations show that the system can efficiently compete the handwritten digits recognition task, which indicates the feasibility of using the RRAM-based binary STDP protocol in neuromorphic computing systems to obtain good performance.",
"conclusion": "Conclusions A RRAM-based binary STDP protocol was proposed and experimentally demonstrated in a RRAM-based crossbar array. An unsupervised online pattern recognition system is designed to demonstrate the protocol. The simulations indicate that the system can efficiently learn and classify the handwritten digit patterns from MNIST database, which suggests that the RRAM-based binary STDP protocol is a potential learning approach that can be used for brain-inspired computing systems.",
"discussion": "Results and discussion After forming operation, the SET/RESET bias voltage (swept from 0 to +2.5/−2.5 V, then back to 0 V) was applied to the TE, with the BE grounded. The typical DC current–voltage (I–V) characteristics are shown in Fig. 1d . The devices show bipolar resistive switching behavior with abrupt SET process from high-resistance state (HRS) to low-resistance state (LRS) and gradual RESET process from LRS to HRS. As shown in Fig. 2 , multilevel resistance states can be achieved by controlling the current compliance value during SET and modulating stop voltage during RESET. The achieved multilevel resistance states show the devices’ robustness to the disturb pulses during SET and RESET, as shown in Fig. 3 . These multilevel resistance states reached by SET and RESET will enable the realization of binary STDP protocol. Fig. 2 Measured multilevel resistance states characteristics of the cells in the crossbar RRAM arrays. a Using current sweeping mode for SET. b Using voltage sweeping mode for RESET \n Fig. 3 Robust multilevel resistance states behaviors to the disturb voltage pulses with the amplitudes of 0.7 V/−1.0 V and the width ranging from 10 ns to 1 ms \n A binary STDP protocol is proposed and demonstrated as shown in Fig. 4 . The time overlap between the pre-pulse at the TE and the post-pulse at the BE leads to a change of the device conductance. The parameter delta t (∆ t ) is defined as ∆ t = t \n post − t \n pre , where t \n post is the time when post-pulse arrived BE and t \n pre is the time when pre-pulse arrived TE. Figure 4a is the waveforms used in the protocol. It includes two pre-pulses (pre I and pre II) and one post-pulse (post). For the convenience of displaying, we set the time span of post as 2 μs. With ∆ t varying from −1000 ns to 3000 ns, only when the pre I meets the post between t \n 1 and t \n 2 (Fig. 4b ) could the RRAM device be switched to LRS, which corresponds to long-term potentiation (LTP). If the pre II meets the post between t \n 2 and t \n 3 (Fig. 4c ), the RRAM device will be switched to HRS, which corresponds to long-term depression (LTD). In other situations, the super imposed waveforms will not switch resistance states. In the protocol, LTP and LTD are determined by the type of pre and delta t . For different initial resistance states (from about 4 × 10 2 Ω to 4 × 10 4 Ω), the device performs similar switching behavior. Fig. 4 A binary STDP protocol. a Waveforms used in this protocol. Including pre I, pre II, and post. b , c Measured binary STDP characteristics. LTP and LTD are determined by pre type and delta t \n \n An unsupervised online learning system (Fig. 5 ) consists of 28 × 28 pre-neurons, and 15 post-neurons are designed based on the binary STDP protocol above, with RRAM cells working as synapses. Leaky integrate-and-fire (LIF) circuits and pulse generators were adopted to construct the post-neurons. Pre-neurons are fully connected with post-neurons by the crossbar structure. Besides, post-neurons connect to each other through inhibitory synapses. In post-neuron, the LIF circuit collects currents from synapses which are connected to it. The LIF circuit integrates the currents and intrigues the pulse generators when the internal potential exceeds a fixed threshold. The fired pulse generators will generate three signals: the feedback signal, the inhibiting signal, and the signal for the next layer. The feedback signal is used to update synapse weights. The inhibiting signal can inhibit other neurons by inhibitory synapses. The signal for the next layer shows the recognition results. Fig. 5 Schematic of an unsupervised online learning system. RRAM cells work as the synapses. Leaky integrate-and-fire circuits and pulse generators construct the post-neurons. Crossbar structure enables pre- and post-neurons fully connected. Fired post-neuron generates three signals: feedback signal, inhibit signal, and signal to next layer neuron \n Training session and test session are two main sessions for a pattern recognition system. At training session, learning events occur at discrete time periods (learning epochs). In this system, the implementation of training session is as follows. Before training session, all synaptic weights were initially set to random values, with an expected value G \n E = 0.5 × ( G \n LRS + G \n HRS ), where G \n LRS is the conductance of LRS and G \n HRS is the conductance of HRS. At the beginning of one learning epoch, handwritten 2D digit patterns from MNIST database are converted into one dimension binary input information. Corresponding to input 1/0, pre-neurons input pre I/pre II into the system (Fig. 6a ). During communication stage (from 0 to 1), the pre I/pre II voltage was set to −0.2 V/0 V. Post-neuron which has maximum sum of current will fire first. The fired neuron reduces sum currents of other neurons by the inhibiting signal and thus becomes the only fired neuron. At the same time, the fired neuron sends a feedback signal (post) to the connected synapses (Fig. 6a ). The post consists of a negative pulse (−0.7 V, t s) and a positive pulse (1.0 V, t s). The post encounters the pre I and the pre II at the synapses which are connected to the fired neuron. According to binary STDP protocol, LTP/LTD can only be achieved at synapses connecting the fired neuron, corresponding to the pre I/pre II at the time t /2 t , and thus, other synapses are not affected. The evolution of the synaptic weight maps is shown in Fig. 6b . Through one learning epoch, the input information is stored into the synapses which are connected to the fired neuron. Afterwards, another learning epoch follows. The training session is over when all learning epochs are completed. Fig. 6 One learning epoch. a Waveforms used for the learning epoch, including pre I, pre II and post, which correspond to input information 0, 1, and feedback signal. Parameter t \n f means integrating time required to fire post-neuron. b The evolution of synaptic weights map. Black pixels and white pixels learned at time t and 2 t , respectively \n In the system, the time t \n f which is needed to fire a post-neuron is determined by the input current I , the RC parameter of the integrating circuit, and the threshold voltage \\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}$$ {V}_{\\mathrm{th}}\\left({t}_{\\mathrm{f}}=-\\mathrm{RC} \\ln \\left(\\frac{V_{t h}}{IR}+1\\right)\\right) $$\\end{document} V t h t f = − R C ln V t h I R + 1 . The input current is related to the input information and synaptic weights. For a constant threshold voltage, a higher input current leads to a shorter t \n f . As a result, communication stage should be wide enough in order that all input information can fire a post-neuron. In our simulation, we presume that our circuit works at an ideal state; hence, parameter t can be considered an ideal value. In order to update weights at various fire time, the time span between the input waveform and the feedback waveform should be 2 t . A learning and classification task of ten patterns is used to verify the function of this system. The weight maps after training session are shown in Fig. 7 . Ten-digit patterns fired ten different post-neurons randomly, which shows that patterns are learned and classified successfully. Fig. 7 Simulated learning results. Ten handwritten digits from MNIST database have been learned and classified by the system"
} | 2,184 |
40213002 | PMC11935182 | pmc | 2,399 | {
"abstract": "Compute‐in‐memory (CIM) is a pioneering approach using parallel data processing to eliminate traditional data transmission bottlenecks for faster, energy‐efficient data handling. Crossbar arrays with two‐terminal devices such as memristors and phase‐change memory are commonly employed in CIM, but they encounter challenges such as leakage current and increased power usage. Three‐terminal transistor arrays have potential solutions, yet large‐scale electrolyte‐gated transistors (EGTs) demonstrations are uncommon due to compatibility issues with existing photolithography processes. Herein, a 20 × 20 EGTs array is designed using indium‐gallium‐zinc‐oxide as the semiconductor channel and polyacrylonitrile (PAN) doped with C 2 F 6 LiNO 4 S 2 as the electrolyte. Each transistor unit in the array can serve as a synapse, exhibiting a large conductance range, low energy consumption (6.984 fJ) for read–write operations, excellent repeatability, and quasilinear update characteristics. It has been confirmed that the EGTs array not only enables precise device programming but also virtually eliminates signal interference between neighboring devices during the programming process. Using 54 transistors in the EGTs array, unsupervised learning with a winner‐takes‐all neural network is successfully demonstrated. After 50 training iterations, the neural network achieves perfect 100% accuracy in classifying test‐set letters. The work demonstrates the potential of EGTs for constructing large‐scale integration synaptic array toward efficient computing architectures.",
"conclusion": "3 Conclusions In summary, we fabricated a 20 × 20 synaptic transistor array utilizing photolithography, with IGZO as the semiconductor channel and PAN doped with C 2 F 6 LiNO 4 S 2 as the electrolyte. The experiments have demonstrated that individual synaptic transistors within the EGTs hold significant potential in neuromorphic computing. In addition, we verified the minimal interference between adjacent devices in the EGTs array during the programming process, affirming that our proposed EGTs array not only facilitates precise device programming but also virtually eliminates interference between neighboring devices. Based on 54 transistors within the EGTs array, unsupervised learning was successfully demonstrated through experiments using a WTA neural network. After 50 training iterations, the neural network successfully achieved perfect classification of test‐set letters, with an accuracy of 100%. This implies that our designed EGTs array can be used to construct a larger unsupervised learning system. The work demonstrates the potential of EGTs for constructing large‐scale integration synaptic array toward efficient computing architectures.",
"introduction": "1 Introduction Compute‐in‐memory (CIM), as an innovative computing paradigm, has garnered significant attention and research interest. [ \n 1 , 2 , 3 \n ] It integrates computing and storage cell in parallel, effectively eliminating the need for frequent data transformation between memory and processing units in the von Neumann architecture, holding immense potential for a wide range of applications. [ \n 4 , 5 , 6 , 7 , 8 \n ] In CIM, parallel data processing can be achieved through vector–matrix multiplication using Ohm's law (for multiplication) and Kirchhoff's law (for accumulation) in a crossbar array structure. [ \n 9 , 10 , 11 \n ] Synaptic devices are employed to accurately copy the synaptic plasticity in the biological brain, underlying the precise vector–matrix multiplication. Two‐\nterminal devices, such as memristors [ \n 12 , 13 , 14 , 15 \n ] and phase‐change memory, [ \n 15 , 16 , 17 \n ] have been frequently used as synaptic devices. However, hardware arrays consisting of these two‐terminal devices to implement CIM encounter challenges in cross‐talk induced, sneak path current, and nonlinear current–voltage characteristics. [ \n 18 , 19 , 20 \n ] \n Three‐terminal devices possess the potential to precisely modulate the conductivity of individual devices in an array, owing to their extra terminals for programming and reading. [ \n \n 21 \n \n ] Among these three‐terminal devices, electrolyte‐gated transistors (EGTs), with their large range of conductivity, low read–write energy consumption, and fast switching speed, have emerged as ideal candidates for emulating biological synapses. [ \n 2 , 22 , 23 \n ] However, current research on EGTs primarily focuses on enhancing the performance of individual transistors, such as reducing energy consumption, [ \n \n 24 \n \n ] increasing tensile strength, [ \n \n 25 \n \n ] and improving proton mobility. [ \n \n 26 \n \n ] There is less emphasis on the fabrication of large‐scale, high‐density arrays of EGTs, which are essential for the implementation of integrated applications. [ \n 27 , 28 \n ] Many of the reported EGTs currently use organic or liquid electrolytes, posing significant challenges for manufacturing methods predominantly based on photolithography. [ \n 29 , 30 , 31 \n ] For instance, photolithography can damage organic solid‐state electrolytes. Although EGTs employing inorganic electrolyte materials demonstrate potential for large‐area fabrication, [ \n 32 , 33 \n ] they often require higher energy consumption. Therefore, exploring alternative manufacturing methods, using organic solid‐state electrolytes as insulating layers to produce large‐scale EGTs arrays, becomes crucial. In practical arrays of three‐terminal devices, while individual devices possess the capability for precise control, a common gate electrode is often shared by a row or column of devices. In such scenarios, programming one device can often lead to interference with adjacent devices. [ \n \n 34 \n \n ] To achieve precise control of devices within the array without causing programming interference to neighboring devices, a programming methodology tailored for EGTs arrays is required. Unsupervised learning is a learning paradigm within the field of machine learning that, in contrast to supervised learning, does not require labeled target outputs, making it more adaptable to large‐scale datasets. Furthermore, it aids in the discovery of latent structures, patterns, and features within data without prior knowledge of these details, thus holding significant value in data exploration, feature selection, and data understanding. [ \n 35 , 36 , 37 , 38 , 39 \n ] In the context of unsupervised learning applications based on synaptic transistors, current research predominantly employs single device for simulating neural network training. [ \n \n 40 \n \n ] Evidently, this approach lacks persuasiveness because it does not account for the mutual interference between multiple devices in an array and fails to demonstrate the practical effects of array‐based implementations. [ \n 41 , 42 , 43 , 44 , 45 \n ] Therefore, it is crucial to utilize real EGTs arrays to showcase unsupervised learning. In this study, a 20 × 20 EGTs array with a coplanar gate structure was fabricated using photolithography, with indium‐gallium‐zinc‐oxide (IGZO) as the semiconductor channel and polyacrylonitrile (PAN) doped with C 2 F 6 LiNO 4 S 2 as the electrolyte. We employed EGTs as synaptic units in a hardware neural network and found that EGTs exhibit a large range of conductivity, low read–write energy consumption (6.984 fJ), excellent conductivity reproducibility, and quasilinear update characteristics, indicating their significant potential in neuromorphic computing. In addition, we verified the minimal interference between adjacent devices in the EGTs array during the programming process, confirming that our proposed EGTs array not only enables precise device programming but also virtually eliminates interference between neighboring devices. Furthermore, based on 54 transistors within the EGTs array, unsupervised learning was successfully demonstrated through experiments using a winner‐takes‐all (WTA) neural network. After 50 training iterations, the neural network successfully achieved perfect classification of test‐set letters, with an accuracy of 100%, highlighting the foundation laid by our designed EGTs array for constructing a larger‐scale unsupervised learning system.",
"discussion": "2 Results and Discussion 2.1 The Processing of EGTs Arrays and Characterization of Electrolyte Thin Films The fabrication process of the EGTs array, as shown in Figure \n \n 1 a , is detailed in Experimental Section . Figure 1b shows a photograph of the array. To more clearly display the morphology of single device in the array, we also captured optical images of individual devices in the array without electrolyte coating and annotated the corresponding dimensions, as shown in Figure 1c . As a matter of fact, it is commonly acknowledged that top gate or bottom gate structure, which is used in the fabrication of conventional MOSFET, is helpful for the miniaturization of the device. However, the incompatibility between the electrolyte and the conventional Si‐based technology prevent the integration of EGTs. Therefore, we propose a coplanar structure of EGTs to accomplish this integration, as illustrated in Figure 1d . This specific structure allows the electrolyte film to be drop cast on top of the array, effectively avoiding damage to the film from other fabrication processes. [ \n 46 , 47 , 48 , 49 \n ] The preparation of this EGTs array based on the coplanar gate structure provides valuable experience and reference for manufacturing large‐scale arrays using similar electrolyte materials. In terms of the device's integration density, considering the stability and repeatability of the array, so far, we have only been able to achieve this level of integration density due to technical limitations. Although the density does not seem very high, we believe that higher integration densities can be achieved with improvements in process technology. Please refer to our previous work for more details. [ \n \n 50 \n \n ] Here, we propose the following potential challenges for further integration. In this integration scheme, with the reduction of device size and the increase in integration density, the difficulties of interconnection and signal acquisition will intensify. [ \n \n 51 \n \n ] The planar structure of a single transistor with a coplanar gate used in the EGTs array is depicted in Figure S1a, Supporting Information. This structure leverages the double‐layer effect to directly couple the gate voltage to the semiconductor channel. When an external voltage is applied at the gate, ions accumulated at the interface of PAN/C 2 F 6 LiNO 4 S 2 electrolyte and the electrode result in a significant double‐layer capacitance and strong lateral field coupling. [ \n \n 52 \n \n ] This is crucial for constructing artificial synaptic networks. In the search for the PAN/C 2 F 6 LiNO 4 S 2 mixing ratio with the maximum areal capacitance, we characterized the film capacitance under different mixing ratios, as shown in Figure S1b, Supporting Information. From the graph, it is observed that a substantial double‐layer capacitance exists in the low‐frequency range, but as the frequency increases from 20 Hz to 2 MHz, the capacitance significantly decreases. The areal capacitance of the film varies with the doping concentration of C 2 F 6 LiNO 4 S 2 . When PAN is mixed with C 2 F 6 LiNO 4 S 2 at a 1:1 mass ratio, the resulting electrolyte film exhibits the highest specific capacitance at a frequency of 20 Hz, reaching 0.55 μF cm −2 . In contrast, when the mass ratio of PAN to C 2 F 6 LiNO 4 S 2 is 2:1, the specific capacitance decreases to 0.48 μF cm −2 ; when the mass ratio is 3:1 ratio, it further decreases to 0.46 μF cm −2 . The specific capacitance of the pure PAN electrolyte film, without C 2 F 6 LiNO 4 S 2 doping, is only 0.33 μF cm −2 . Furthermore, a detailed analysis was conducted on the surface of the PAN/C 2 F 6 LiNO 4 S 2 mixed electrolyte film with a 1:1 mass ratio, and its scanning electron microscope (SEM) image is displayed in Figure 1e . The left image represents the overall morphology of the film, whereas the right SEM image is an enlarged view of the film's surface. The SEM images indicate a relatively uniform electrolyte surface with limited roughness. This result was confirmed by energy‐dispersive X‐ray spectroscopy (EDS) analysis for element F, representing C 2 F 6 LiNO 4 S 2 , as shown in Figure 1f . Figure 1 Fabrication process and physical characteristics of EGTs array and electrolyte film. a) Process diagram of the EGTs array based on photolithography. b) Photo of the prepared EGTs array. c) Picture of a single transistor in the array. d) 3D structure diagram of a single electrolyte transistor. e) SEM image of the surface of the electrolyte film (left) and a local magnified view of the film's surface (right). f) EDS analysis result for the element F (representing C 2 F 6 LiNO 4 S 2 ). © 2024 WILEY‐VCH GmbH 2.2 The Electrical and Synaptic Performance of EGTs As shown in Figure \n \n 2 a , the transfer characteristics curve of EGTs ( V \n D = 0.4 V) is obtained by performing a forward sweep of V \n G from −2.5 to 3 V and then returning to −2.5 V. This curve exhibits a distinct hysteresis window, which is advantageous for simulating the plasticity of synaptic performance in subsequent steps. The threshold voltage ( V \n th ) for individual transistors in the EGTs array is −0.002 V, with a subthreshold swing of 0.26 V decade −1 . Under the conditions of V \n DS = 0.0006 V, pulse signals with pulse widths and amplitudes of 3 ms and 1 V, respectively, were applied to the gate terminal, triggering the excitatory postsynaptic current (EPSC), as depicted in Figure 2b . This phenomenon occurs because when a positive voltage is applied to the gate, a forward electric field forms between the gate and the channel. This drives positively charged ions to gradually migrate to the interface between the gate dielectric and the channel, inducing channel current. As the pulse voltage is removed from the gate, the ions gradually diffuse back to their initial state, causing the EPSC to return to its initial state as well. Figure 2 The electrical and synaptic performance of individual transistors in the EGTs array. a) Hysteresis transfer curve of individual transistors in the EGTs array within the voltage range of −2.5 to 3 V, showing a large hysteresis window indicative of significant potential for achieving synaptic characteristics. b) EPSC current triggered by a pulse voltage ( V \n G = 1 V, V \n DS = 0.0006 V, t \n pulse = 3 ms), with a calculated energy consumption of 6.984 fJ. c) EPSC triggered by a pair of presynaptic pulses (1 V, 60 ms), where the EPSC triggered by the second pulse is 45.5 nA, significantly larger than the 35.7 nA triggered by the first pulse. d) EPSC triggered by different pulse frequencies (1–10 Hz, V \n G = 2 V, V \n DS = 0.05 V, t \n pulse = 50 ms), showing an increase in triggered EPSC with increasing frequency. e) Current gain at different frequencies, G = (A5 – A1)/A1, indicating an increase in current gain from 0.06 to 1.04 as the frequency increases from 1 to 10 Hz, suggesting significant potential for high‐pass filtering applications. f) EPSC responses to 50 consecutive pulses ( V \n G = 1 V, t \n pulse = 50 ms). g) Changes in EPSC as the voltage amplitude gradually increases from 1 to 9 V ( V \n DS = 0.05 V, t \n pulse = 50 ms). h) Changes in EPSC as the pulse voltage width increases from 50 to 950 ms ( V \n G = 2 V, V \n DS = 0.05 V). © 2024 WILEY‐VCH GmbH The maximum value of the EPSC is 3.88 nA. The individual energy consumption of a single pulse peak for a transistor in the array was calculated based on the formula ( 1 ): \n (1) \n E pre pulse = I peak ⋅ V DS ⋅ t \n where I \n peak and t represent the peak and width of the pulse, respectively, and the calculated minimum energy consumption for a single pulse of the transistor is 6.984 fJ, which is comparable with the energy consumption of a biological synapse. [ \n 30 , 53 , 54 , 55 , 56 , 57 \n ] To further investigate the synaptic transistor's paired pulse facilitation (PPF) behavior, we applied two consecutive presynaptic pulses to the gate terminal of the synaptic transistor. Each pulse had an amplitude of 1 V, a width of 60 ms, and an interval of 40 ms. As shown in Figure 2c , the EPSC triggered by the second pulse is 72.1 nA, significantly higher than the 43.6 nA triggered by the first pulse. The reason for this double‐pulse facilitation is the relatively slow ion motion rate in the solid‐state electrolyte film, which cannot respond promptly to the pulse voltage at the gate. When the subsequent pulse stimulus arrives, the ions that did not fully respond during the first pulse further contribute, resulting in an increase in the EPSC. [ \n 58 , 59 , 60 \n ] Figure 2d demonstrates the simulation of the synaptic transistor's high‐pass filtering characteristics using six sets of pulse signals with different frequencies (pulse width of 50 ms, pulse amplitude of 2 V, V \n DS = 0.05 V). We set the amplitude of the EPSC triggered by the first pulse as A1 and the amplitude of the EPSC triggered by the fifth set of pulses as A5. When 1 Hz frequency pulses are applied to the gate, the EPSC remains relatively unchanged. However, as the pulse frequency increases from 1 to 10 Hz, the response of the synaptic transistor gradually intensifies, as shown in Figure 2e . The gain of the EPSC ((A5–A1)/A1) increases from 0.06 to 1.04, indicating that the device can serve as a high‐pass filter in the information transmission process. This phenomenon occurs as with the increase in pulse frequency, the time intervals between presynaptic pulses shorten, leaving insufficient time for ions to return to their equilibrium positions. This results in more ions accumulating at the gate dielectric and channel surfaces, leading to larger EPSC. [ \n 57 , 61 , 62 , 63 \n ] Figure 2f displays the EPSC response after applying 50 presynaptic pulses (pulse width of 50 ms, V \n G = 1 V, V \n DS = 0.4 V) to the synaptic transistor. It is observed that the amplitude of the subsequent EPSC is higher than that of the previous one. This occurs because if a second stimulus is triggered before the residual ions fully recover, these residual ions combine with the ions triggered by the second stimulus, leading to a higher amplitude of the subsequent EPSC. This mechanism is similar to PPF. Figure 2g illustrates the effect of different voltage stimuli on the postsynaptic current (pulse width of 60 ms, V \n DS = 0.05 V). As the applied voltage gradually increases from 1 to 9 V, the generated EPSC increases from 35.76 to 178 nA. This indicates that the peak of the EPSC increases with the increment of the stimulus voltage. As higher voltages are applied, a greater number of ions are induced in the electrolyte layer, resulting in an increase in the EPSC. Figure S2a, Supporting Information, displays the relationship between pulse amplitude and the corresponding EPSC size. It is evident from the graph that as the stimulus voltage increases, the generated EPSC exhibits an approximate linear relationship. This also suggests that it performs well in a programming approach involving a gradual increase in voltage amplitude. [ \n 64 , 65 \n ] By applying electrical signals of varying pulse widths ( V \n G = 2 V, V \n DS = 0.05 V) to the gate terminal of the synaptic transistor, ranging from 50 to 950 ms, we obtained EPSC responses under different pulses width stimuli. As shown in Figure 2h , with an increase in pulse duration, the EPSC generated by the synaptic transistor also increases, and the time required for it to return to its initial state becomes longer. This is because, as the pulse duration extends, more ions accumulate at the interface between the gate dielectric layer and the active layer, resulting in the accumulation of more electrons at the channel interface and, consequently, an increase in EPSC. Figure S2b, Supporting Information, depicts the relationship between pulse width and the corresponding EPSC size. It is apparent from the graph that the increase in pulse width results in EPSC responses that approximate a linear trend, indicating its notable effectiveness in a programming approach involving a gradual increase in voltage pulse width. [ \n 64 , 65 \n ] \n 2.3 Array Programming Interference Analysis and Uniformity Characterization In the EGTs array, transistors within the same column share a common gate electrode port. Therefore, when programming a specific device within the array, other devices in the same column may also experience programming interference ( Figure \n \n 3 a ). [ \n 66 , 67 \n ] This is disadvantageous for subsequent differential computing tasks. To address this issue, we devised an independent programming method based on the EGTs array design. During the programming, a selected device receives programming pulses on its GL while setting SL to 0 V and DL to 0.1 V. For nontargeted devices, we implemented an inhibition procedure by setting their DL to a low voltage level (0 V) (Figure 3b ). This reduced interference by keeping these nontargeted devices deactivated. To demonstrate the practical effect, we selected four adjacent devices within the same column, sharing a common gate electrode port. We applied a continuous sequence of 50 trigger pulses to the common gate electrode port (pulse voltage of 1 V with a duration of 60 ms). We configured devices for excitation and inhibition, as previously described. The results showed a significant increase in channel conductance for the two selected devices, whereas the other two nonselected devices maintained nearly constant channel conductance (Figure 3c ). To provide a clearer representation of the change in channel conductance for these devices, we calculated the relative change in channel conductance for the four adjacent devices sharing the common gate electrode port. The relative change in channel conductance for B01, B02, B03, and B04 was 0%, 70.5%, 0%, and 70.4%, respectively (Figure 3d ). This indicates that this programming method is highly effective for achieving independent programming. Furthermore, this pulse programming method allows for individual programming of devices within the array that have different conductance states. To evaluate the durability of EGTs, more than 2200 P–D pulses ( V \n P = 2 V, V \n D = ‐2 V, t \n pulse = 50 ms) were applied to the EGTs. As shown in Figure 3e , even after the application of over 2200 pulses, EGTs devices remained stable and exhibited excellent repeatability compared with the initial measurements. Figure 3 Crosstalk in the array and device durability. a) An example of a 4 × 4 array, with one gate controlling a column of devices. b) The configuration for reducing device programming interference in the EGTs array (left) and the specific arrangement of individual devices within the array (right). c) Changes in the conductance of four adjacent devices sharing a common gate under the trigger of 50 consecutive excitatory pulses (pulse voltage: 1 V, width: 60 ms). The selected device's conductance shows a significant increase, whereas the conductance of the unselected devices remains relatively unchanged. d) The electrical conductance changes for the two selected devices are 70.5% and 70.4%, whereas the unselected two devices show no electrical conductance change, registering at 0%. e) Durability characteristics of EGTs subjected to over 2200 pulses. © 2024 WILEY‐VCH GmbH To achieve CIM in EGTs arrays, small differences between devices are necessary. To assess the mutual differences between devices, we characterized the electrical characteristics of 100 devices in the array. By performing transfer characteristics curve measurements on 100 transistors selected from a 20 × 20 array of synaptic transistors, V th and SS were extracted. Figure S3a, Supporting Information, displays the distribution of V th for these 100 devices, ranging from −0.8 to 0.6 V, the average of the V th for the selected 100 devices is 0.19 V, with a standard deviation of 0.22 V. Figure S3b, Supporting Information, shows the distribution of SS for these 100 devices, which generally falls within the range of 0.1–0.9 V decade −1 , the average of the SS for the selected 100 devices is 0.51 V decade −1 , with a standard deviation of 0.14 V decade −1 . Using the number of pulses to represent the learning iterations, 50 pulses (with a pulse width of 60 ms, pulse amplitude of 1 V, and V \n DS = 0.4 V) were administered at various locations within the array. The corresponding EPSC of each device was measured at 1, 5, 10, 20, 30, and 50 pulses after stimulation. The varying color intensity observed in the grid cells of the figure indicates the strength of the EPSC generated by the corresponding unit in the array. The grid cells in the figure exhibit varying color intensity, reflecting the magnitude of EPSC generated by each unit in the array. With the increase in the number of learning iterations, the memory level showed improvement, as illustrated in Figure S3c, Supporting Information. The synaptic transistor array successfully demonstrated dynamic memory, suggesting that it has promising potential for practical applications in image processing. Figure S3d, Supporting Information, shows the channel conductance variation of a single transistor in the array under a single‐period pulse stimulation, with linearity values of 1.16 for the potentiation phase and 3.29 for the depression phase, demonstrating acceptable linearity. [ \n 68 , 69 \n ] A two‐layer perceptron is a basic neural network model consisting of 400 input neurons, 100 hidden neurons, and 10 output neurons, as shown in Figure S3e, Supporting Information. Through repeated forward propagation and backpropagation processes, the two‐layer perceptron can gradually learn the weights that adapt to the input data, enabling classification or prediction of the input data. Figure S3f, Supporting Information, displays the accuracy of the transistor over 125 training cycles, with a significant improvement in perceptron accuracy in the first 10 training cycles. Throughout the entire training process, the highest accuracy achieved is 92.02%. 2.4 Implementing Unsupervised Learning Functionality Based on EGTs Array The EGTs array was configured as a WTA neural network, using EGTs as analog synapses, for the classification of standard letters z, v, n, and their noisy versions ( Figure \n \n 4 a ). This network was trained based on the actual EGTs array, meaning that no external computer simulation models were used during both the training and inference processes. The network's input layer consists of nine neurons, each corresponding to one of the nine pixels in the pattern, whereas the output layer is composed of three neurons, each corresponding to one of the three output categories (Figure 4b ). The network's output results Y \n \n j \n are primarily determined by formula ( 2 ): \n (2) \n Y i = ∑ i = 1 9 x i W i , j ∑ i = 1 9 x i ( G ( i , j ) + − G i , j − ) \n where x i is the input voltage for the ith pixel, where dark pixels represent high‐level input voltage and light pixels represent low‐level input voltage. Here, W i , j represents the synaptic weight, which is equal to the difference in channel conductance between two adjacent devices ( G ( i , j ) + − G i , j − ). Using this differential subtraction method to represent synaptic weights is effective in achieving both positive and negative weight values and can effectively increase the dynamic range of synaptic weights. Furthermore, representing synaptic weights using the differential subtraction of channel conductance between adjacent devices assists in reducing device‐to‐device variability, as adjacent devices exhibit highly similar characteristics. [ \n 32 , 70 \n ] \n Figure 4 The architecture diagram for unsupervised learning. a) Standard images of letters “z”, “v”, and “n” and their noisy versions with only one pixel flipping. b) Architecture of the neural network. c) Flowchart for WTA network training based on EGTs array. d) Training process of the letter“v” as an example within the neural network system. © 2024 WILEY‐VCH GmbH The training process for unsupervised learning is illustrated in Figure 4c . Initially, the EGTs array needs to undergo an initialization step to ensure that the devices within the array are in their initial state. Next, we randomly select one of the standard letters “z”, “v”, or “n” as an input image, with each image representing one training epoch. The training phase involves using the standard patterns of the letters z, v, and n, as well as their noisy versions. Train until the synaptic weights of each neuron are sufficiently updated. In the process of conducting unsupervised learning, a pulse width of 50 ms and a pulse amplitude of 1 V were applied to the gate, whereas a drain voltage of 0.4 V was used. Taking the standard letter “v” as an example, the training process is as follows (see Figure 4d ): First, the image of the standard letter “v” is decomposed into nine pixels, and these pixels are fed into the neural network. Subsequently, following the WTA rule, synaptic weights within the neural network undergo updates. The specific weight update process is as follows: after inputting the standard “v” letter's pattern into the neural network model, the network generates three output results. These three output results are then compared, and the output neuron corresponding to the maximum output value is declared as the winner. Only the synaptic weights connected to the winning neuron are updated, whereas other synaptic weights remain unchanged. Synaptic weights connected to the winning neuron and associated with high‐level input are strengthened. Conversely, synaptic weights connected to the winning neuron and associated with low‐level input are suppressed. For example, if, after inputting the standard letter “v,” neuron Y 1 produces the highest output value, Y 1 is declared as the winner. In this case, synaptic weights connected to Y 1 , such as W 1 , 1 , W 3 , 1 , W 4 , 1 , W 6 , 1 , and W 8 , 1 , would be increased ( G 1 , 1 + , G 3 , 1 + , G 4 , 1 + , G 6 , 1 + , and G 8 , 1 + increased and G 1 , 1 − , G 3 , 1 − , G 4 , 1 − , G 6 , 1 − , and G 8 , 1 − unchanged), whereas synaptic weights connected to Y \n 1 , such as W 2 , 1 , W 5 , 1 , W 7 , 1 , and W 9 , 1 , would be suppressed ( G 2 , 1 + , G 5 , 1 + , G 7 , 1 + , and G 9 , 1 + unchanged, G 2 , 1 − , G 5 , 1 − , G 7 , 1 − , and G 9 , 1 − increased). After a sufficient number of training iterations, a test set consisting of both the standard and noisy versions of these letters is introduced into the trained neural network model to evaluate its clustering performance with respect to these letters. \n Figure \n \n 5 a illustrates the clustering results for letters in the EGTs array before and after training. It can be observed that before training, many patterns are misclassified, as shown in the upper part of Figure 5a . At this point, the network does not yet possess the capability to effectively cluster letters. However, after multiple rounds of training with the standard and noisy patterns of the letters z, v, and n, the EGTs array gains the ability to successfully cluster the letters, with an accuracy of 100%, as depicted in the lower part of Figure 5a . To better understand the convergence behavior of the algorithm, we define a specialization function Si for each neuron i , representing the pattern x (z, v, n) for which the neuron produces the maximum output yi. During the training of the neural network, when each neuron uniquely corresponds to different letter patterns and remains unchanged, we can consider the network's classification a success. Figure 5b shows the changes in the output neurons representing different letters during the training process. From the figure, it can be seen that after 24 training cycles, the neural network achieves successful classification. Subsequently, neurons 1, 2, and 3 correspond to patterns z, v, and n, respectively. Between points A and B, neuron 3 points to both letters v and n . This is due to the initial weight distribution of the neural network system not creating a significant distinction between the patterns v and n , leading to an illusion of misclassification. However, with the increasing number of training cycles, the synaptic weights of the neural network receive updates. When confronted with the input of v and n once again, the system successfully classifies them. Figure 5c demonstrates the evolution of 27 synaptic weights during the neural network training process. Before any training, the weight distribution chart shows no apparent patterns. With an increasing number of training cycles, the distribution chart representing the letter patterns in the synaptic weights gradually becomes clearer. After ≈30 training cycles, neurons 1, 2, and 3 have exhibited relatively clear letter patterns, corresponding to the letters z, v, and n. To better illustrate this process, we use the fourth transistor of neuron 1 and the second transistor of neuron 3 as examples, detailing the evolution of their weights, as shown in Figure 5d . They start from similar initial values and then evolve according to the WTA rule. After multiple training iterations, they eventually form two distinct states. The WTA algorithm demonstrated in this work can also be extended to multilayer WTA networks to address more complex input features. Figure 5 Presentation of unsupervised learning results. a) Classification results of the test set by the neural network system before and after training. Many patterns were misclassified before training, but after training, it effectively classifies standard letters and noisy letters in the test set. b) Changes in the different letters represented by the output neurons during the training process. c) Evolution of synaptic weights of the three output neurons during 50 rounds of training; as training progresses, the shapes represented by each output neuron become gradually clearer. d) Example of the synaptic weight update process for the fourth synapse of the first neuron and the second synapse of the third neuron. © 2024 WILEY‐VCH GmbH"
} | 8,590 |
34632243 | PMC8490943 | pmc | 2,400 | {
"abstract": "Abstract Mine wastes pollute the environment with metals and metalloids in toxic concentrations, causing problems for humans and wildlife. Microorganisms colonize and inhabit mine wastes, and can influence the environmental mobility of metals through metabolic activity, biogeochemical cycling and detoxification mechanisms. In this article we review the microbiology of the metals and metalloids most commonly associated with mine wastes: arsenic, cadmium, chromium, copper, lead, mercury, nickel and zinc. We discuss the molecular mechanisms by which bacteria, archaea, and fungi interact with contaminant metals and the consequences for metal fate in the environment, focusing on long‐term field studies of metal‐impacted mine wastes where possible. Metal contamination can decrease the efficiency of soil functioning and essential element cycling due to the need for microbes to expend energy to maintain and repair cells. However, microbial communities are able to tolerate and adapt to metal contamination, particularly when the contaminant metals are essential elements that are subject to homeostasis or have a close biochemical analog. Stimulating the development of microbially reducing conditions, for example in constructed wetlands, is beneficial for remediating many metals associated with mine wastes. It has been shown to be effective at low pH, circumneutral and high pH conditions in the laboratory and at pilot field‐scale. Further demonstration of this technology at full field‐scale is required, as is more research to optimize bioremediation and to investigate combined remediation strategies. Microbial activity has the potential to mitigate the impacts of metal mine wastes, and therefore lessen the impact of this pollution on planetary health.",
"conclusion": "5 Conclusions Metal contamination from mining causes serious environmental impacts and risks to human health on a global scale. Microorganisms inhabit mine wastes and microbial activity can both mobilize metals from mine wastes, and sequester metals from contaminated waters. Microbes have a range of resistance mechanisms to deal with metal toxicity, and can also indirectly change metal speciation. These microbial processes influence the environmental fate of metals, transferring them between the terrestrial and aquatic environments, and consequently their mobility and likelihood to cause adverse impacts on planetary health. Metal contamination can cause decreased efficiency in soil functioning as microbes need to expend energy to maintain and repair cells when challenged with toxic metals. But perhaps the effect of this is less than would be expected, particularly observed in analyses of long‐term field studies rather than laboratory experiments where soils were spiked with metal salts. This may be due to microbial metal tolerance, resistance and community adaption. It appears that the impact of metals that are essential elements and subject to homeostasis (e.g., Zn and Cu), or metals that have a close biochemical analog (e.g., As), have less of an impact on soil function compared to those which are neither (e.g., Cd, Cr, and Pb). This is rather simplistic though, as microbial metal and mineral transformations will also have an effect, as will mixtures of metals that are commonly encountered in mine wastes, plus geochemical factors like pH and organic carbon availability. While there are many laboratory studies that investigate the addition of metal salts to soils, more studies using metal‐impacted mine wastes are required to fully understand the combination of the waste mineralogy, geochemistry, and multiple metal contaminants. Ideally these should investigate long‐term metal exposure to allow for the adaption of microbial communities. As well as this, more long‐term field studies are required, especially that consider the geomicrobiological processes that influence metal speciation and microbial community dynamics. It is important that fungi as well as prokaryotes are considered in these systems. It would be exciting to see more molecular biology approaches applied to understand how microbial communities respond and adapt to metal contamination in the field, to show which resistance mechanisms are most significant and if this response that can explain the observed (bio)geochemistry. Stimulating beneficial microbial activity to remediate metal‐impacted mine wastes has had some successes, particularly the application of constructed wetlands or passive bioreactors to generate (microbially mediated) anoxic and sulfidic conditions. The combination of different remediation strategies (oxic and anoxic) may confer advantages against the use of a single system. Optimisation of remediation strategies needs to be addressed in order to provide low‐cost approaches for governments and companies to address metal contamination in water and soils, which affects almost every country in the world. Further research is required to allow bioremediation to be applied in future at large scale, particularly around engineering, costs, challenges with consistency, acceptable rate of reaction, and disposal of wastes. Alternative approaches could be also be assessed such as the potential for microbial formation of metal phosphate minerals and metal sequestration in association with manganese oxides; it would be interesting to see whether these processes can be stimulated at large‐scale to remediate mine wastes. Fungi in particular grow quickly, make large amounts of biomass under aerobic conditions and can form phosphate and manganese oxides; more study of their potential is required. Finally, we have described microbial metal resistance mechanisms at the genetic level for single species, but can these be applied to develop new and novel treatment strategies, or even to optimise field based bioremediation? Understanding microbe‐metal interactions may hold the key to limiting metal toxicity in mine wastes, by informing remediation strategies to prevent the formation of acid mine waters, to recover metals from mine wastes and to remediate metal‐impacted mine waters, therefore improving environmental outcomes and planetary health.",
"introduction": "1 Introduction Mine wastes are the unwanted by‐products left behind after the ores of economic interest have been extracted, and can be solid, liquid or gaseous (Lottermoser, 2010 ). Mining has occurred for thousands of years, and almost every country has a legacy of mine waste that contains toxic materials including various metals and metalloids (Hudson‐Edwards et al., 2011 ). It has been estimated that 20–25 Gt of solid mine wastes are produced globally every year (Lottermoser, 2010 ). These contain waste rock, poorly extracted ore minerals, gangue minerals, tailings, processing chemicals and residues that are stored at or near mine sites, or in the past were discharged to rivers or wetlands (Hudson‐Edwards et al., 2011 ). Pollution of the environment by metals and metalloids from mine wastes is a global environmental issue due to their widespread distribution and potential toxicity to humans, plants and wildlife. To improve planetary health we need interdisciplinary solutions to limit the impact of soil and water pollution; for mine wastes this has to involve geochemical, mineralogical and microbiological considerations. The fate and transport of metals and metalloids in mine wastes are controlled by physical, chemical and biological processes. Microorganisms colonize and inhabit mine wastes (Figure 1 ), they can tolerate high concentrations of metals and metalloids, and transform/detoxify them through metabolism or resistance mechanisms. This review focusses on the impact of microbiological activity on the behavior of metals in the environment, and how microorganisms can be used to remediate metal and metalloid contaminated mine wastes. We review microbial interactions with the most common potentially toxic metals and metalloids found in mine wastes, together with how microbial processes influence their environmental mobility. We also describe the effect of these contaminants on the essential microbial activity that contributes to soil functioning, and the impact of biogeochemical cycling on the fate of metals and metalloids. Throughout this review we have referred to metals and metalloids found in mine wastes as “metals.” We discuss specific field sites and how microbial processes can be enhanced to mitigate the impacts of metal pollution to improve planetary health (Hudson‐Edwards, 2016 ). Phytoremediation, that is growing plants to stabilize mine tailings, or the detoxification of mine wastes by hyperaccumulators (e.g., Wang et al., 2017 ) is not covered. Figure 1 Diverse range of microorganisms found in a former metal mine (Cornwall, UK). (a) Secondary mineral coatings forming within the mine. (b) Range of organisms isolated from the secondary mineral coatings. Scale bars 2 mm. Image source: T. Sbaffi. (c) Composition of the prokaryotic community in the secondary mineral coatings. Image source: Bakes, 2020 . Section 2 of this review describes microbe‐metal interactions including microbial metabolisms, how microbial oxidation and reduction impact metal mobility, metal complexation by microbially generated ligands, microbial metal resistance mechanisms and biosorption. Section 3 then covers how microbes interact with arsenic, cadmium, chromium, copper, lead, mercury, nickel, and zinc. Section 4 of this review describes case studies where microbial metabolism has been stimulated to remediate metal contamination at mine sites with acidic, neutral and alkaline wastes."
} | 2,398 |
32641733 | PMC7343832 | pmc | 2,402 | {
"abstract": "Photosynthetic microorganisms such as cyanobacteria, purple bacteria and microalgae have attracted great interest as promising platforms for economical and sustainable production of bioenergy, biochemicals, and biopolymers. Here, we demonstrate heterotrophic production of spider dragline silk proteins, major ampullate spidroins (MaSp), in a marine photosynthetic purple bacterium, Rhodovulum sulfidophilum , under both photoheterotrophic and photoautotrophic growth conditions. Spider silk is a biodegradable and biocompatible material with remarkable mechanical properties. R. sulfidophilum grow by utilizing abundant and renewable nonfood bioresources such as seawater, sunlight, and gaseous CO 2 and N 2 , thus making this photosynthetic microbial cell factory a promising green and sustainable production platform for proteins and biopolymers, including spider silks.",
"introduction": "Introduction Growing awareness of climate change, depletion of nonrenewable fossil resources, and global food and water crises have recently spurred efforts to develop “sustainable cell factory” platforms for the production of valuable biocompounds/chemicals. Ideally, these next-generation cell factories should employ eco-friendly and sustainable bioprocesses and solely depend on renewable nonfood bioresources as feedstocks. We have been developing a purple nonsulfur bacterium, Rhodovulum sulfidophilum , that confers advantages from both photosynthetic 1 , 2 , and halophilic 3 , 4 abilities as a potential alternative workhorse to replace current heterotrophic microbial cell factories 5 . R. sulfidophilum is a marine anoxygenic photosynthetic bacterium with versatile metabolic capabilities that produces biohydrogen 6 , bioplastic 7 , and extracellular nucleic acids 8 . The most important points are its ability to grow under photoautotrophic conditions by utilizing low-cost and abundant renewable resources such as light (energy), CO 2 (carbon source), and N 2 (nitrogen source) via photosynthesis and nitrogen fixation processes 9 – 11 and its ability to grow in seawater, which could lower the risk of biological contamination during cultivation 5 . Nature provides extremely strong and tough biomaterials, such as spider silk 12 , limpet teeth 13 , and bagworm silk 14 . Spider dragline silk, in particular, has been extensively studied due to its outstanding features, including high tensile strength, high extensibility, and low weight 15 , 16 . In addition, the biodegradable and biocompatible features of spider dragline silk have made it suitable for biomedical and eco-friendly applications 17 . Major ampullate spidroin (MaSp) is produced in the major ampullate gland of spiders, and spun silk fibers are mainly composed of multiple types of MaSp, such as MaSp1 and MaSp2 18 , 19 . MaSp has a conserved primary structure comprising three domains: a repetitive central domain and nonrepetitive N -terminal and C -terminal domains. The MaSp repetitive domains are arranged in alternating blocks of polyalanine (crystalline) and glycine-rich (amorphous) sequences, which are responsible for the high tensile strength and high elasticity, respectively, of spider silk fibers 20 , 21 . Current mass production of spidroins has been achieved using recombinant host organisms because of low yields from spider silk glands and the cannibalistic and territorial nature of spiders 22 , 23 . Spidroins have been successfully expressed in recombinant bacteria ( Escherichia coli ) 24 , 25 , yeasts ( Pichia pastoris ) 26 , insects (silkworm Bombyx mori ) 27 , plants (tobacco and potato) 28 , and animals (mice and mammalian cell cultures) 29 , 30 . Using bacterial or yeast fermentation technologies, a few venture companies have launched various prototypes made of artificial spider silk fibers 31 . However, it is still a great challenge to produce spidroins on a large scale with a sustainable production process, even though spider silk is an eco-friendly and sustainable material. Moreover, the hydrophobic tandem sequences of MaSp1 could reduce productivity by microbial fermentation. Besides, high price of spider silk due to high production cost also remains a challenge to be resolved. Raw materials that used in heterotrophic microbial fermentation systems could contribute up to 70% of production cost 32 . Here, we develop an economical and sustainable marine photosynthetic microbial cell factory using R. sulfidophilum , which is a marine purple nonsulfur bacterium that is capable of producing the hydrophobic repetitive sequence of MaSp1 using small amount of organic substance under photoheterotrophic or photoautotrophic growth conditions. Although very little information is available for recombinant protein expression in R. sulfidophilum except for studies related to its photosynthetic apparatus 33 , 34 . To the best of our knowledge, this is the first report of heterologous spidroin production using photosynthetic and halophilic bacteria with abundant carbon and nitrogen sources under seawater conditions.",
"discussion": "Results and discussion Construction of MaSp1-expressing R. sulfidophilum The introduction of exogenous plasmid DNA into R. sulfidophilum via bacterial conjugation using pCF1010-derived plasmids and E. coli S17-1 as a donor strain was reported 34 . This transformation was achieved based on the RP4/RK2 mating system. In this study, we used another broad-host-range vector, pBBR1MCS-2, harboring a kanamycin resistance gene, mob (mobility) gene and transfer origin ( oriT ), which have been widely used in Gram-negative bacterial conjugation 35 , 36 . In the chromosome of R. sulfidophilum (accession no. NZ_CP015418), two tellurite resistance genes encoding the TerB-family tellurite resistance protein were present at the loci ‘A6W98_RS06280’ and ‘A6W98_RS17070’. Both kanamycin and tellurite resistance features were used as selection markers to distinguish positive conjugants of R. sulfidophilum . The newly constructed pBBR1-P trc -MaSp1 plasmid contained (i) a trc promoter (P trc ), which is a hybrid ( trp and lacUV5 promoters, differs from tac promoter by 1 bp) constitutive strong promoter in E. coli 37 , (ii) the ribosome-binding site (RBS) sequence “AGGAGA”, which is derived from the upstream region of the puf operon (encoding a light-harvesting protein and a reaction center complex) in R. sulfidophilum 38 , and (iii) a repetitive domain sequence of the MaSp1 gene from Nephila clavipes , which had been codon-optimized for E. coli 24 (Fig. 1a, b , Supplementary Table 2 ). This gene cassette was located in the multiple cloning site of pBBR1MCS-2 but in the opposite direction of the lac promoter (P lac ) to avoid the influence of the lac promoter on our target protein expression. Fig. 1 Heterologous expression of spider dragline silk proteins in the recombinant marine photosynthetic bacterium Rhodovulum sulfidophilum under photoheterotrophic conditions. a A recombinant R. sulfidophilum harboring the broad-host-range vector pBBR1MCS-2 with a MaSp1 repetitive domain from Nephila clavipes was developed to express spider dragline silk protein. b A gene cassette containing the trc promoter (P trc ) and MaSp1 -(1-mer, 2-mer, 3-mer, and 6-mer) was inserted into pBBR1MCS-2, and a histidine tag was present at the N -terminus of MaSp1 (pink-color box). c Tris-Tricine SDS-PAGE (16.5%) of soluble proteins from four days of recombinant R. sulfidophilum cultures. d Western blot using monoclonal anti-His•Tag antibody, which targets histidine-tagged MaSp1-(1-mer, 2-mer, 3-mer, or 6-mer) proteins. Photoheterotrophic production of different sizes MaSp1 Approximately 0.4 g of cell wet mass (CWM) was obtained from 50 mL of a recombinant R. sulfidophilum culture grown to the stationary growth phase under photoheterotrophic conditions, namely, marine broth (MB) with LED illumination at 730 nm and irradiation at 20–30 W m −2 , for 4 days. Although the overexpression of the recombinant MaSp1 proteins was not detected clearly in all the recombinant R. sulfidophilum cultures by SDS-PAGE (Fig. 1c ), we confirmed the positive expression of the MaSp1 proteins for all the newly constructed recombinant R. sulfidophilum cells harboring pBBR1-P trc -MaSp1-(1-mer, 2-mer, 3-mer, or 6-mer) by western blotting (Fig. 1d ) and liquid chromatography–tandem mass spectrometry (LC–MS/MS) analyses (Supplementary Data 1 39 ). The single repetitive domain in our constructs contains 33 amino acid residues as follows: NH 2 -SGRGGLGGQGAGAAAAAGGAGQGGYGGLGSQGT-COOH. The theoretical molecular weights for the target proteins, including nonspidroin sequences (His-Tag, S-Tag, enterokinase, and thrombin cleavage sites) at the N -terminus, are 7.9 kDa for the 1-mer (81 aa), 10.5 kDa for the 2-mer (114 aa), 13.1 kDa for the 3-mer (147 aa), and 20.9 kDa for the 6-mer (246 aa). Indeed, all the target protein bands in western blots appeared at slightly higher positions than their corresponding theoretical molecular weights. This gel shifting is due to the hydrophobicity of silk proteins in general, which affects protein–SDS interactions to reduce gel mobility 40 , 41 . In addition to the confirmation of MaSp1 proteins expression, we also performed a brief estimation of the amount of MaSp1 proteins obtained from the recombinant R. sulfidophilum cultures, which was ~3–10 mg L −1 (1-mer = 3.4 mg L −1 , 2-mer = 3.9 mg L − 1 , 3-mer = 10.2 mg L −1 , and 6-mer = 6.8 mg L −1 ) or 3.5–6.9% of total proteins based on western blotting semiquantification (Supplementary Fig. 1 ). For comparison, heterologous expression of spidroins in a well-established and widely used recombinant E. coli system was able to produce ~0.3–1.2 g L −1 purified spidroin 23 , 42 . Nevertheless, to our knowledge, this was the first report of successful biosynthesis of artificial spider silk proteins in a marine photosynthetic bacterium under photoheterotrophic conditions. Further attempts in expressing artificial spider silk proteins with sizes close to the native spider dragline silk (~100-mer or ~300 kDa), which had been achieved in metabolically engineered E. coli would be applicable in R. sulfidophilum as well. However, there are many challenges need to be resolved for the host in advance such as metabolic capability (high demand for glycine and alanine tRNAs) and stability of the genetic constructs (long and highly repetitive DNA sequences) 42 , 43 . Photoautotrophic growth and heterotrophic MaSp1 production The most remarkable result of this study is the demonstration of next-generation microbial cell factories based on marine photosynthetic organisms in which we can apply an photoautotrophic growth mode by using renewable nonfood feedstocks and seawater as the cultivation medium. R. sulfidophilum harboring pBBR1-P trc -MaSp1-(6-mer) was cultured in Daigo’s artificial seawater (ASW) medium with light from LEDs (730 nm, 20–30 W m −2 ) with a bicarbonate salt (1 g L −1 ) as an inorganic carbon source and nitrogen gas (0.5 L d −1 ) as a nitrogen source for 7 days (Fig. 2a ). The largest repeat, MaSp1-(6-mer), was chosen for subsequent experiments because higher molecular weight of MaSp1 would contribute more tensile strength to the spider silk fiber. Sodium bicarbonate was used to supply inorganic carbon because bicarbonate salts have greater solubility and lower logistic and transportation costs than gaseous CO 2 44 . Fig. 2 Photoautotrophic growth and heterotrophic production of artificial spider silk protein in the recombinant marine photosynthetic bacterium Rhodovulum sulfidophilum . a Recombinant R. sulfidophilum harboring pBBR1-P trc -MaSp1-(6-mer) was cultivated using 20 mL of Daigo’s artificial seawater (ASW) SP for marine microalgae medium in a 20 mL glass vial with a rubber stopper at 30 °C with continuous far-red LED light (730 nm, 20–30 W m −2 ) for 7 days. Inorganic carbon was supplied as 1 g L −1 sodium bicarbonate, while nitrogen was supplied via daily nitrogen gas bubbling at 0.5 L d −1 . Both marine broth (MB) and ASW media contained 100 mg L −1 kanamycin. b Biomass accumulation of recombinant R. sulfidophilum in various medium compositions based on cell dry mass (CDM). c Semiquantitative quantification of MaSp1-(6-mer) expression in crude cell lysate of R. sulfidophilum by western blot using a monoclonal anti-His•Tag antibody. d MaSp1-(6-mer) yield by recombinant R. sulfidophilum in various medium compositions. Mean data (±SD) accompanied by different letters are significantly different with p values < 0.05 ( n = 3 independent biological replicates). (C = NaHCO 3 , YE = 0.4 g L −1 yeast extract, N 2 = nitrogen gas, and P = 0.5 g L −1 KH 2 PO 4 ). In our previous study, we had examined the cell growth of R. sulfidophilum under different light conditions, such as intensity (8 and 50 W m −2 ) and wavelength (730, 800, and 850 nm) 45 . In this study, we evaluated the effect of a few additional nutrients (yeast extract, vitamin, iron, and phosphorus) that are deficient in ASW medium on the growth of recombinant R. sulfidophilum . The cell dry mass (CDM) decreased from 0.90 g L −1 (with all nutrients) to 0.66 g L −1 and 0.39 g L −1 in the absence of yeast extract and phosphorus, respectively (Supplementary Fig. 2 ). In subsequent experiments, we also observed that the recombinant R. sulfidophilum was unable to grow in ASW medium without the supply of any of NaHCO 3 , N 2 gas, or phosphorus (Fig. 2b , ASW + N 2 , ASW + C + N 2 , ASW + C + P, and ASW + P + N 2 ). The CDM (~0.4 g L −1 ) in these ASW cultures was most likely from the inoculums or seed cultures (MB) even after the samples were washed with 2% sodium chloride. Thus, carbon, nitrogen, and phosphorus sources are all necessary for the growth of recombinant R. sulfidophilum in ASW medium. As expected, the cell growth increased significantly from 0.34 ± 0.02 g L −1 (ASW + C + N 2 ) to 0.58 ± 0.08 g L −1 (1.7-fold increase) and 0.81 ± 0.02 g L −1 (2.4-fold increase) in the presence of yeast extract (ASW + C + N 2 + YE) and phosphorus (ASW + C + N 2 + P), respectively. The highest CDM was achieved by adding together yeast extract and phosphorus, which yielded 1.04 ± 0.06 g L −1 (3.1-fold increase) or almost 70% of the CDM in nutrient-rich MB medium (1.48 ± 0.01 g L −1 ). An ~0.2 mg L −1 recombinant MaSp1 protein yield and an MaSp1 content accounting for 2% of total proteins were observed in ASW + N 2 , ASW + C + N 2 , ASW + C + P, and ASW + P + N 2 (Fig. 2c, d and Supplementary Fig. 3 ), which might be carry-over from inoculum as explained in the previous section. MaSp1 protein production was promoted by the addition of yeast extract, which significantly increased the yield of MaSp1 protein from 0.12 ± 0.10 mg L −1 (ASW + C + N 2 ) to 3.93 ± 2.76 mg L −1 (ASW + C + YE + N 2 ). Yeast addition also increased the percentage of MaSp1 in the total protein from 1.2 ± 1.0 to 6.9 ± 5.3%. Interestingly, the addition of phosphorus had an adverse effect on MaSp1 protein production even though it could significantly promote CDM increments. Compared to growth in ASW + C + YE + N 2 , growth in ASW + C + YE + P + N 2 decreased the yield of MaSp1 protein to 2.71 ± 1.09 mg L −1 and the percentage of MaSp1 in total proteins to 3.9 ± 1.6%. These results could be explained by the function of each component, where the yeast extract (autolyzed yeast cells) is mainly a nitrogen source, which promotes protein biosynthesis 46 – 48 . Meanwhile, phosphorus is an essential macronutrient and heteroelement in many important cellular compounds that promotes the growth of primary producers 49 , 50 . Although further optimization on the ASW medium is necessary to achieve cell growth and MaSp1 yield comparable to those in MB medium (CDM = 1.48 ± 0.01 g L −1 ; MaSp1 yield = 52.28 ± 11.20 mg L −1 ), we demonstrated photoautotrophic growth and heterotrophic synthesis of silk proteins by using mainly renewable nonfood feedstocks, small amount of organic substance, and seawater as the cultivation medium. Purification of MaSp1 for spider silk fiber formation To obtain a sufficient amount of MaSp1 protein for fiber extrusion, we performed nine-liter-scale jar fermentation for the production of MaSp1-(6-mer), the largest repeat available in this study (Fig. 3a ). In general, the size of the spidroins have positive correlation to tensile strength until a certain size of molecular weight. Large proteins possess more interchain and intrachain interactions, more entanglements, and less chain-end defects 42 , 43 , 51 . Purifications of MaSp1-(6-mer) were carried out using affinity chromatography via histidine tag, which was present at the N -terminus of the MaSp1 gene cassette (Supplementary Fig. 4a ), and gel filtration chromatography (Supplementary Fig. 4b ). The purified MaSp1-(6-mer) appeared in eluent fractions 1 and 2 after His-Tag purification. Purified MaSp1-(6-mer) in the eluent fractions 10–12 was combined after gel filtration and then subjected to desalting and lyophilization. In the end, we obtained ~10 mg of purified MaSp1-(6-mer) (Fig. 3b ) from ~40 g of CWM. Silk fibers were produced by pipetting 10 wt% purified MaSp1-(6-mer) dissolved in hexafluoroisopropanol (HFIP) into a coagulation bath, followed by hand-drawing using forceps (Fig. 3c ). The best results were obtained using 90% (v/v) 2-propanol as the coagulation bath, which induced relatively mild dehydration that allowed efficient drawing 52 . Analysis using scanning electron microscopy showed that the fibers exhibit approximately constant diameters of 10–20 μm and a surface marked by striations parallel to the fiber axis. Fracture surface analysis revealed an internal structure consisting of microfibrils (Fig. 3d, e ). Fig. 3 Large-scale production and fiber extrusion of MaSp1-(6-mer) artificial spidroin. a Nine-liter-scale production of MaSp1-(6-mer) using marine broth containing 100 mg L −1 kanamycin under photoheterotrophic conditions and continuous far-red LED light (850 nm, 15 W m −2 ) at 30 °C for 7 days. b Lyophilization of pure MaSp1-(6-mer) after His-Tag affinity and gel filtration chromatographic purifications. c Fiber extrusion was performed via hand-drawing using forceps with 10% (w/v) purified MaSp1-(6-mer) dissolved in HFIP, while 2-propanol was used as a coagulation bath. d Scanning electron micrographs of the hand-drawn spider silk fibers at the surface. e Scanning electron micrographs of the break point of the spider silk fiber. In conclusion, we have successfully established a promising marine photosynthetic microbial cell factory using the purple nonsulfur bacterium R. sulfidophilum and demonstrated photoheterotrophic expression of artificial spider silk protein and silk fiber formation in this system and, more importantly, under photoautotrophic growth condition. Future work is needed to improve cell growth and protein expression under photoautotrophic growth conditions through methods such as supplementing seafood processing wastewater 53 into ASW medium and modifying the recombinant protein expression system. In principle, this marine photosynthetic microbial cell factory should also be suitable for the production of other biocompounds, which will contribute greatly to research communities and society in efforts to promote green, sustainable, and cost-effective bioprocesses."
} | 4,860 |
39419997 | PMC11487081 | pmc | 2,403 | {
"abstract": "Global functional adaptation after local mechanical stimulation, as in mechanobiology and the mimosa plant, is fascinating and ubiquitous in nature. This is achieved by locally sensing mechanical deformation with precise thresholds, processing this information via biochemical circuits, followed by downstream actuation. The integration of such embodied intelligence allowing for mechano-to-chemo-to-function information-processing remains elusive in man-made systems. By merging the fields of chemical circuits and metamaterials, we introduce adaptive metamaterial hydrogels (meta-gels) that can accurately sense mechanical stimuli (local touch and global strain), transmit this information over long distances via reaction-diffusion signaling, and induce downstream mechanical strengthening by growing nanofibril networks, or soft robotic actuation through competitive swelling. All elements of the sensor-processor-actuator system are embedded in the device, functioning autonomously without external feeding reservoirs. Our concept enables designing advanced life-like materials systems that synergistically combine two worlds – chemical circuits for chemical information-processing and metamaterial unit cells for physical information-processing.",
"introduction": "Introduction Living entities interact with and adapt to their environment by applying the principles of decentralized embodied intelligence using a sensor-processor-actuator paradigm 1 , 2 . This paradigm inspires the next generation of life-like systems with unprecedented capacity for adaptation and autonomous operation. For example, sensing of mechanical forces allows to detect obstacles for guiding movement or to change physical properties 3 , 4 . The underlying information-processing, i.e., mechanotransduction, is a multi-step mechanism in which force events are converted into (bio)chemical signals and processed in nonlinear biochemical signaling circuits to finally achieve mechanical adaptation as an actuator function 5 – 7 . The mimosa plant is an excellent example of local sensing, long-range information-transmission, and global adaptation through actuation 8 . In stark contrast, classical responsive materials do not allow for equally complex behavior because they lack processor functions. The progress in materials relevant to this study has focused so far mainly only on the mechano-sensing and actuation parts, such as mechanoresponsive polymers that exhibit continuous activation of bonds without a critical threshold, and soft robotic devices that can bend under a direct external stimulus 9 – 16 . The challenge of incorporating information-processing elements into mechanical materials can be tackled from different fields. Complex autonomous decision-making requires nonlinearities in chemical or physical dynamics, as in chemical reaction networks (CRNs) or mechanical metamaterials 1 , 2 , 17 – 20 . CRNs use chemical feedback loops to generate exotic behavior, e.g., self-acceleration through positive feedback or transient states and oscillations through negative feedback 21 – 23 . CRNs combined with transport processes result in reaction-diffusion (RD) phenomena such as propagating chemical signals or self-organized patterns 17 , 24 . CRNs have been used to program autonomous lifetimes and periodic behavior in self-assemblies and materials 25 – 29 . However, in these cases, typically, a homogeneous chemical signal dictates a spatially uniform bulk material behavior. The development of materials with a fully integrated mechanotransduction system that enables precise sensing of a macroscopic deformation state and local-to-global adaptation remains an unresolved challenge, even though it is highly relevant technologically. Mechanical metamaterials offer a complementary perspective on the topic. Metamaterials use porous unit cell structures with specific geometries to produce nonlinear mechanical response impossible for bulk materials 18 , 30 – 33 . These nonlinearities can be used for logic gates, signal propagation, and processing. This area is largely driven by solid mechanics 34 . Combining it with the world of CRNs and responsive materials can open unprecedented possibilities for synergistic chemical and material intelligence in next-generation life-like material systems. Herein, we introduce a platform concept to realize the sensor-processor-actuator paradigm in soft robots and mechanical materials, integrating the RD dynamics of a pH-autocatalytic CRN 23 as the information-processing element with metamaterial unit cells as sensory elements. Autocatalysis serves as a signal enhancement, and the coupling to diffusive transport allows for a sharp, self-sustaining chemical front, which ensures local-to-global signal transmission throughout the material 35 , 36 . The sensing step is the local detection of a mechanical force. We design a metamaterial unit for force-threshold-dependent activation. The actuation, that is, mechanical strengthening and shape-morphing, is achieved by using pH-triggered materials that react downstream of the CRN processor. This concept allows us to incorporate nontrivial spatiotemporal information-processing into materials and build autonomous, freestanding soft robots with mimosa-like system-level adaptivity induced by local forces.",
"discussion": "Discussion The implementation of the sensor-processor-actuator paradigm with autonomous information-processing and operation inspired by the embodied intelligence of living systems is one of the greatest challenges in engineering new functional soft materials. New generic platform concepts are needed, where the complex material-level decision making (thresholds, response amplitude, pathways, etc.) arise from relatively simple nonlinear elements. Such elements of embodied intelligence can be provided with different approaches, e.g., CRNs, metamaterials, or also neuromorphic semiconductor devices, where each domain has different strengths 1 , 2 . In this work, we merged elements of two otherwise separated fields of embodied intelligence— autocatalytic CRNs and metamaterial strain gates—to realize a synergistic combination of nonlinear modules for the design of adaptive materials systems. Unlike previous studies focusing on bulk programmability 25 – 29 , our approach realizes spatiotemporal sensing, information propagation, and structural actuation in spatially mesostructured meta-gels. This is crucially important for advanced local-to-global information-transmission and adaptation scenarios. The information-transmission system in this work is empowered by the urea-urease OH – -autocatalytic front reaction. Autocatalysis multiplies the signal chemically, and the coupling with diffusive transport results in sharp, long-distant RD fronts. The RD mechanism ensures fast and non-exhausting signal transmission from a local sensory event to all over the object, regardless of its size. In addition, the enhanced chemical signal is a strong enough effector to trigger a downstream process without killing the front. This strategy may be expanded to other autocatalytic processes as long as the initial dormant state is sufficiently stable. However, the advantage of the urea-urease system is its biocompatibility and the availability of many pH-responsive systems to engineer downstream functional processes. Here, we demonstrated applications in soft robotics using pH-responsive polymers and in self-strengthening materials exploiting pH-triggered nanofibrillation. Nontrivial life-like behaviors always arise from non-equilibrium operation. The actuation speed is limited by the front propagation that is engineered to be slow enough to provide a sufficiently long-lasting dormant state, and the water transport responsible for swelling. Therefore, CRN-empowered soft robots have a slower response and smaller generated forces than hard robots. However, they are especially suitable for creating complex autonomous (computer-chip-free) dynamics that mimic living systems, and the generated forces are potentially enough to interface them with biosystems on a small size scale. Our gel devices presented here respond to one-time events. Resettability and cyclic operation would be extremely challenging in such a closed, compartmentalized system. The current state of pH CRNs does not provide suitable activator-inhibitor-type chemical dynamics for closed systems; a potential solution could be external refueling using vasculature or further hydrogel elements. From a wider perspective, mechanical stimulation is one of the most ubiquitous sensory inputs in fields from biomaterials to macroscopic structural materials, but building the interface from mechanics to CRNs and back to engineer life-like functions remains largely elusive. The hydrogel framework is excellent for confined, freestanding 3D devices that can naturally process force and touching events and robustly accommodate spatiotemporal chemistry. In this work, we first created hydrogel actuators sensing and processing local non-thresholded compressive forces, and secondly, we transformed stretching into thresholded compression with the strain gate. We introduced how comparably simple metamaterial unit cells can generate fundamentally new self-controlled behavior allowing for (1) distinct thresholding of activation strain and (2) also for delocalizing sensor units away from the actor unit. Distant off-robot activation opens fundamentally new design opportunities in soft robotics, where sensing, processing, and acting are spatially separated and thus can be coupled to different input and output signals (e.g., mechanical and chemical). Depending on the strength of the mechanical signal (strain and stress), our system reacts in a strictly binary way: In the case of low strain, it remains unchanged, but in the case of high strain, it adapts everywhere. We leveraged this behavior for self-strengthening materials that become stiffer and tougher once they are stretched to a supercritical level. We believe that the great progress in mechanical metamaterials and 3D printing opens possibilities for more complex material designs. Looking out to the future: The presented concept, integrating RD signaling, metamaterial sensing, and adaptive downstream processes, offers a new perspective for future multi-sensory soft material constructions, enabling quasi-intelligent fate-selection and complex autonomous decision-making and self-controlled operation in space and time."
} | 2,613 |
35028420 | PMC8691124 | pmc | 2,404 | {
"abstract": "Biologists and engineers are making tremendous efforts in contributing to a sustainable and green society. To that end, there is growing interest in waste management and valorisation. Lignocellulosic biomass (LCB) is the most abundant material on the earth and an inevitable waste predominantly originating from agricultural residues, forest biomass and municipal solid waste streams. LCB serves as the renewable feedstock for clean and sustainable processes and products with low carbon emission. Cellulose and hemicellulose constitute the polymeric structure of LCB, which on depolymerisation liberates oligomeric or monomeric glucose and xylose, respectively. The preferential utilization of glucose and/or absence of the xylose metabolic pathway in microbial systems cause xylose valorization to be alienated and abandoned, a major bottleneck in the commercial viability of LCB-based biorefineries. Xylose is the second most abundant sugar in LCB, but a non-conventional industrial substrate unlike glucose. The current review seeks to summarize the recent developments in the biological conversion of xylose into a myriad of sustainable products and associated challenges. The review discusses the microbiology, genetics, and biochemistry of xylose metabolism with hurdles requiring debottlenecking for efficient xylose assimilation. It further describes the product formation by microbial cell factories which can assimilate xylose naturally and rewiring of metabolic networks to ameliorate xylose-based bioproduction in native as well as non-native strains. The review also includes a case study that provides an argument on a suitable pathway for optimal cell growth and succinic acid (SA) production from xylose through elementary flux mode analysis. Finally, a product portfolio from xylose bioconversion has been evaluated along with significant developments made through enzyme, metabolic and process engineering approaches, to maximize the product titers and yield, eventually empowering LCB-based biorefineries. Towards the end, the review is wrapped up with current challenges, concluding remarks, and prospects with an argument for intense future research into xylose-based biorefineries.",
"conclusion": "8. Conclusion and future perspectives Xylose is a readily available sugar with potential to serve as feedstock for biorefineries. For the economic viability of lignocellulose biorefineries, the efficient conversion of hemicellulosic sugars into value-added products is mandatory. Glucose-based commercially developed bioprocesses are prevalent while xylose-based ones are evolving at an industrial scale. Recent developments in biomass pretreatment technologies have led the way to extract xylose from the hemicellulosic fraction of plant cell walls with desired yields with a small amount of plant cell wall inhibitors. In nature, the xylose metabolising microorganisms are scanty compared to those metabolising glucose. Therefore, bioprospecting of novel microorganisms that could assimilate xylose separately or in combination with glucose with faster conversion rates will significantly promote efficiency of LCB-based biorefineries. However, the xylose uptake rates of the well-known xylose assimilating microorganisms are significantly lower than those assimilating glucose. Despite the exemplary developments in xylose bioconversion, there are still several challenges which need to be fixed for developing efficient microbial cell factories for high level manufacturing of biochemicals and biofuels. These challenges include efficient xylose transportation into microbial cells, faster uptake & metabolism of xylose similar to glucose, continuous availability of redox cofactors for maintaining homeostasis, glucose repression during co-fermentation, and feedback, substrate, and product mediated inhibition. Recent advancements in enzyme/metabolic/pathway engineering along with system/synthetic biology approaches have been employed to overcome these challenges but have been met with limited success. Though, xylose-based bioproduction has shown significant progress in the last few decades, many obstacles still need to be addressed to realize xylose as a feedstock at the industrial level.",
"introduction": "1. Introduction Biomass is a potential alternative to non-renewable and non-sustainable fossil fuels causing massive harm to the atmosphere through colossal carbon emission and generation of pollutants. 1 Analogous to a petroleum refinery, a biorefinery processes biomass into multiple products with a green and sustainable approach leading to low carbon biomanufacturing technologies. 1,2 First generation biorefineries making use of edible feedstocks such as sugar, starch, and vegetable oils for generating biofuels are well established, but pose a significant concern and are a regular subject of the food vs. fuel debate. 3 On the other hand, second generation biorefineries based on non-edible feedstocks such as lignocellulosic biomass (LCB) do not interfere in any food chain and offer a clear value proposition for the production of bulk and speciality chemicals. LCB is the most abundant feedstock on the planet (∼200 billion tonnes) with a significant contribution stemming from post-harvest agricultural residues. It is composed of lignin (15–20%), the outermost protective layer, cellulose (40–50%), the inner amorphous and crystalline component of the secondary wall, and hemicellulose (25–30%) microfibrils that connect the outermost and inner cellulose layers ( Fig. 1A ). 4 Cellulose is a linear homo-polymer of d -glucose units connected by β-1,4-glycosidic bonds, and hemicellulose is a complex hetero-polymer containing d -xylose, l -arabinose, d -glucose, l -galactose, d -mannose, d -glucuronic acid and d -galacturonic acid ( Fig. 1B ). Hemicelluloses constitute 26% dry weight in hard woods, 22% in soft woods, and up to 25% in agro-residues with various polymeric forms such as xylan, arabinoxylan, xyloglucan, and glucuronoxylan. 5,6 To utilize this three-dimensional polymeric structure as the feedstock for fermentative production of value-added chemicals, the polymer is converted into simple fermentable sugars. However, the major limitation is that most of the microorganisms are incapable of metabolizing all the fermentable sugars present in LCB, especially pentoses. The pentose sugars are present in the hemicellulosic fraction with xylan as the major polysaccharide which is composed of β-1,4-linked xylose residues. The depolymerization of the hemicellulosic fraction generates a mixture of sugars containing ∼90% xylose. In fact, xylose is the second most abundant sugar available after glucose in LCB ( Fig. 1A ). 7 Despite this, the application of xylose as a potential feedstock is overlooked for biorefineries and it is discarded as waste or incinerated for energy purposes. This is due to a lack of efficient fermentation systems, as many of the microorganisms do not have a native pathway for metabolizing xylose. In addition, uptake of xylose is suppressed in the presence of glucose due to carbon catabolite repression. 8 That is why the number of literature reports using glucose as a substrate for bioproduction is much larger in comparison to that using xylose. However, while exploiting biochemical platforms, the techno-commercial success of an LCB-based biorefinery largely thrives on the revival of the carbohydrate economy, which in turn is dependent on efficient depolymerization of both the structural polysaccharides to simple sugars and their subsequent valorisation to various commercially important products either through chemical or biotechnological routes. 9,10 Therefore, efficient conversion of xylose is necessary and it is imperative to find robust microbial systems for metabolizing xylose for simultaneous assimilation of glucose and xylose for the pragmatic development of profitable LCB-based biorefineries. Fig. 1 (A) Structural components of lignocellulosic biomass (LCB). (B)Illustration of the composition of individual subunits of LCB and compositions of sugar and sugar acids in the hemicellulosic fraction. Considering the challenges associated with xylose utilization, the current review (i) covers the efficient pretreatment processes assisting in xylan extraction from different LCB residues, (ii) discusses the bottlenecks impeding xylose assimilation and strategies to overcome them, (iii) describes the major native and engineered microbial cell factories available for efficient bioconversion of xylose to chemical building blocks, (iv) includes implementation of elementary flux mode analysis to understand the optimal pathway for xylose utilization to produce biomass and end metabolites with a case study of succinic acid, and (v) briefly covers alternative chemical catalysis of xylose for manufacturing value-added products. Finally, the limitations and future perspectives for constructing microbial cell factories to effectively utilize xylose and produce a wide array of products are included. 1.1 Pretreatment strategies for the extraction of fermentable sugars from LCB Recalcitrance is a natural and intrinsic feature of any LCB, originating from its three principal constituents, cellulose, hemicellulose, and lignin, that chemically interact to form a complex network popularly known as a lignin–carbohydrate complex (LCC). 11,12 During biorefining via a biochemical route, pretreatment is an imperative module that disrupts the lignocellulosic matrix by breaking LCC linkages leading to delignification and partial or complete hydrolysis of xylan, thereby improving the surface characteristics of biomass and enhancing the accessibility of cellulose for enzymatic hydrolysis. Invariably, most of the traditional pretreatment strategies primarily result in lignin removal, releasing fermentable sugars from the thermolabile hemicellulosic/xylan fraction, or are focused on selective delignification enriching the biomass in glucan and xylan fractions. 13 1.1.1 Pretreatment method targeting xylan hydrolysis Conventional techniques like steam explosion (SE), liquid hot water (LHW), dilute acid (DA), and hydrothermal (HT) pretreatments result in the solubilization of the hemicellulose fraction and partial lignin removal. 14 However, the extent of xylan hydrolysis and release of inhibitors during pretreatment significantly depends on the process severity. Process variables such as solid loading during pretreatment, temperature, pressure, residence time and concentration of acid in case of DA pretreatment, biomass composition and pretreatment reactor configuration directly or indirectly govern the successful xylan extraction as monomers, oligomers or its degradation products like furfural, the release of lignin-derived inhibitory derivatives and loss of cellulose as glucose or its dehydrated product namely 5-hydroxymethylfurfural (HMF) in the hydrolysed fraction. 15–17 Generally, SE, LHW and HT pretreatments favour deacetylation of thermolabile acetyl groups attached to the hemicellulose backbone and cause release of acetic acid in a temperature range of 180–250 °C. 16 Since acetic acid is weak compared to inorganic acids, partial xylan hydrolysis occurs, and the resulting pre-hydrolysates are predominant in xylooligosaccharides (XOS) with fewer xylose monomers. 18 Yao et al. have recently confirmed that the pH of the medium plays a decisive role in the breaking of LCC linkages. 15 Thus, HT pretreatment likely induces deacetylation and catalyses the cleavage of glycosidic linkages within the xylan backbone, but the addition of strong acid even at low concentration reduces pH that preferentially breaks the ester linkages between lignin and xylan. 15 Therefore, during DA pretreatment, lower temperatures are recommended (120–180 °C) as the addition of acid demands lower operating conditions favouring xylan hydrolysis. Further, combinatorial pretreatment involving a low concentration of inorganic acid and water facilitates the release of xylose monomers from the hemicellulose backbone. It enhances the efficiency of the process owing to milder operating conditions and less inhibitor generation, while preserving the cellulosic fraction in the biomass. Table 1 exclusively showcases a few examples of previously published literature where SE, LHW and DA pretreatments and their combinations selectively hydrolysed the xylan fraction (>85%) and gave <25% delignification. Since HT and DA pretreatments are among the most popular, efficient, and economically attractive pretreatment strategies that lead to selective xylan hydrolysis keeping the glucan fraction in the biomass intact, these technologies have been scaled up to semi-pilot and pilot plant levels as well, as shown in Table 2 . The following section describes conventional pretreatment methods, which lead to enrichment of xylan and glucan fractions in the biomass, targeting selective delignification. State of the art showcasing pretreatment strategies leading to selective xylan hydrolysis a Type of LCB Type of pretreatment Pretreatment conditions Biomass composition (%) Removal (%) References Before pretreatment After pretreatment Xylan Lignin Poplar DA Temp: 170 °C; time: 8.5 min; H 2 SO 4 : 0.5% (w/w) Gln-57.9; HC-17.5; KL-24.6 Gln-74.2; HC-<2.0; KL-25 99 — \n 198 \n MS HT pretreatment at low acid Temp: 180 °C; 10 min; H 2 SO 4 : 0.3% (w/w) Gln-41.9; #XMG-22.1; KL-22.0 Gln-64.4; #XMG-5.0; KL-29.3 86.4 20.2 \n 199 \n CS DA in steam gun Temp: 160 °C; time: 5 min; H 2 SO 4 : 2% (w/w) Gln-34.0; Xln-22.0; KL-12.3 Gln-57.4; Xln-3.2; KL-24.8 ∼92.8 ∼<1.5 \n 200 \n SCB DA Temp: 140 °C; time: 8 min; H 3 PO 4 : 0.2% (w/v) Gln- 40.1; HC-27.5; TL-18.5 Gln-58.5; HC-1.8; TL-29.05 96.5 14.8 \n 201 \n SG DA Temp: 160 °C; time: 30 min; H 2 SO 4 : 1% (w/w) Gln-33.5; Xln-22.7; KL-16.3 Gln-53.2; Xln-0.8; KL-33.3 98.6 18.3 \n 202 \n WS DA Temp: 140 °C; time: 90 min; H 2 SO 4 : 0.5% (w/w) Gln-43.2; Xln-24.4; KL-20.8 Gln-59.1; Xln-2.4; KL-30.7 91.5 — \n 203 \n CS HT Temp: 180 °C Gln-36.1; Xln-21.4; TL-13.6 Gln-33.0; Xln: 5.4; TL-13.5 74.9 — \n 200 \n CC HT Temp: 207 °C Gln-28.8; Xln-29.6; KL-18.6 Gln-54.5; Xln-10.2; KL-21.8 80.4 33.1 \n 204 \n SS LHW Temp: 220 °C; time: 5 min Gln-33.13; HC-26.2; KL-18.2 Gln-56.7; Xln-2.0; KL-37.0 96.5 6.6 \n 205 \n SCB H 3 PO 4 catalysed SE Temp: 195 °C; time: 7.5 min; H 3 PO 4 : 0.95% (w/w) Gln-31.8; Xln-12.2; KL-24.3 Gln-49.7; Xln-2.3; KL-31.9 90.6 14.4 \n 206 \n H 2 SO 4 catalysed SE Temp: 195 °C; time: 7.5 min; H 2 SO 4 : 0.2% (w/w) Gln-49.4; Xln-3.3; KL-31.5 86.6 12.1 a MW: maple wood; SCB: sugarcane bagasse; SG: switchgrass; WS: wheat straw; CC: corn cob; SS: sugarcane straw; DA: dilute acid; SE: steam explosion; LHW: liquid hot water; HT: hydrothermal; Gln: glucan; Xln: xylan; HC: hemicellulose; KL: Klason lignin: TL: acid soluble and insoluble lignin; #XMG: xylan, mannan and galactan. Acid catalysed SE and DA pretreatment carried out at semi-pilot and pilot scales with different types of lignocellulosic feedstock a LCB type Reactor type Reaction conditions Biomass composition (%) Composition of pre-hydrolysate References Untreated Pretreated Sugars Non-sugar component CC Screw steam explosive extruder Pressure: 15.5 bar Gln-42.23 Xylose: 27.5 wt% Acetic acid: 1.1 wt% \n 207 \n Time: 5.5 min HC-39.01 XOS: 2.4 wt% TP: 1.7 wt% H 2 SO 4 : 2.4% (w/w) + steam explosion KL-14.42 Glucose: 3.9 wt% Furfural: 0.5 wt% Arabinose: 3.7 wt% 5 HMF: 0.2 wt% SCB 350-L SS reactor with stirrer & thermal oil heating Temp: 120 °C Gln-45.1 Gln-54.6 C5: 17.4 g L −1 Acetic acid: 2.3 g L −1 \n 208 \n Time: 10 min HC-26.9 HC-10 TA: 7.5 g L −1 H 2 SO 4 : 1% (w/v) KL-22.2 KL-32 C6: 1.6 g L −1 Furfural: 0.8 g L −1 5 HMF: 0.2 g L −1 WS Continuous pretreatment reactor (250 kg day −1 ) Temp: 160 °C Gln-47.1 Gln-63.1 Xylose: 29.2 g L −1 Acetic acid: 1.9 g L −1 \n 209 \n Pressure: 5.2 bar Time: 10 min HC-24.3 HC-1.0 Glucose: 8.4 g L −1 Furfural: 0.9 g L −1 H 2 SO 4 : 0.5% (v/v) KL-28.5 KL-35.8 Arabinose: 2.6 g L −1 5 HMF: 0.6 g L −1 EG 150 L horizontal Andritz reactor Temp: 180 °C; time: 15 min; H 2 SO 4 : 2.4% (w/w) + steam explosion Gln-38.5 Gln-55.5 92% xylan recoverable and 74% as xylose Acetic acid: 2.9 wt% \n 210 \n Xln-11.0 Xln: 0.8 Furfural: 0.9 wt% KL-25.2 KL-37.1 5 HMF: 0.2 wt% RS Continuous pretreatment reactor (250 kg day −1 ) Temp: 162 °C; time: 10 min Gln-37 Gln-51.8 100 g xylose in hydrolysate/kg initial dry substrate Acetic acid: 2 g L −1 \n 211 \n Final H 2 SO 4 : 0.35% (w/w) Xln-20 Xln-3.6 Furfural: 1.2 g L −1 Preasoaking in acid: 0.5 h acid TL-13.4 TL-28.8 5 HMF: 1.1 g L −1 WS Steam explosion in a 30L rig Pressure: 12 bar; time: 12 min; final H 3 PO 4 : 1.2% (w/v); acid pre-soaked biomass introduced Gln-41.6 Xylose: 17.7 wt% — \n 212 \n Xln-30.3 TL-19.3 CS Gln-38.5 Xylose: 13.9 wt% — Xln-24.3 TL-18.3 MS Gln-47.0 Xylose: 14.7 wt% — Xln-25.1 TL-26.15 a CC: corn cobs; SCB: sugarcane bagasse; RS: rice straw; WS: wheat straw; EG: Eucalyptus grandis ; CS: corn stover; MA: Miscanthus; Gln: glucan; Xln: xylan; HC: hemicellulose; KL: Klason lignin; TL: acid soluble and insoluble lignin; C5: pentose sugars; C6: hexose sugars; XOS: xylooligosaccharides; GOS: glucooligosaccharides; DA: dilute acid; TPL: total phenolics; TA: total aromatics; wt%: Wt in g/100 g biomass. 1.1.2 Pretreatment strategies favouring glucan and xylan enrichment The use of sodium hydroxide (NaOH) during pretreatment is one of the most popular and industrially scalable delignification strategies. It cleaves LCC linkages (phenolic α-aryl, phenolic α-alkyl, and phenolic and non-phenolic β-aryl ether linkages) between lignin and hemicellulosic fractions, and improves the surface properties and digestibility of cellulose. 19 Unfortunately, hemicellulose, being amorphous, acetylated and thermolabile, is easily extracted when a high NaOH concentration is used above 70 °C, resulting in significant losses (≥35%). Hence, there are isolated reports of alkali pretreatment wherein xylan enrichment in the solid fraction has been successfully demonstrated. For example, Zhang and associates reported <20% xylan removal from wheat straw and sugarcane bagasse when they pretreated the biomasses with 0.5 M NaOH at 80 °C for 6 h. The resulting pretreated wheat straw and sugarcane bagasse contained 89.9% and 92.9% carbohydrate fraction, respectively. 20 Earlier, while evaluating various pretreatment methods for anaerobic digestion of Miscanthus floridulus , alkaline peroxide (2% H 2 O 2 at 35 °C for 24 h at pH 11.8) pretreatment removed >70% lignin, enriching pretreated biomass with 99.82 and 83.03% glucan and xylan fractions, respectively. 21 In yet another variation, Gong et al. (2020) achieved >70% delignification of corn stover by treating it with 5% alkaline methanol at 80 °C for an hour and retaining ∼89.5 and 88.5% glucan and xylan fractions in the solid biomass. 22 1.1.3 Pretreatment strategies favouring biomass fractionation & holistic utilization of biomass components The two-stage fractionation process has been another lucrative alternative for xylan removal in the first stage, followed by delignification in the later stage. Recently, beechwood was subjected to a two-step fractionation process in which pre-hydrolysis at 150 °C for 90 min was performed with 20 mM H 2 SO 4 . As a result, ∼85.8 wt% xylan was recovered in stage I. When in the second step, organosolv treatment was performed with a 1 : 1 ethanol–water mix and 80 mM H 2 SO 4 at 150 °C for 70 min, ∼82.7 wt% lignin yield was obtained in the liquid fraction leading to the generation of a highly digestible cellulose-rich pulp. 23 Earlier, Smit and Huijen evaluated seven different feedstocks: wheat straw, corn stover, beechwood, poplar, birchwood, spruce, and pine for mild organosolv pretreatment with 50% acetone and <50 mM H 2 SO 4 at 140 °C for 2 h. Irrespective of biomass type, 87–97% xylan hydrolysis was observed. Poor delignification yields were obtained only in spruce and pine, while glucan recoveries ranged between 68 and 94%. Later the group precipitated the dissolved lignin by diluting with water, leading to effective fractionation of all three components of different LCBs. 24 Recently, Xu et al. devised a mild technique for hemicellulose extraction from poplar wood with a binary solvent system containing formic acid and water. Pretreatment at 90 °C for 4 h resulted in 73.1% xylose yield while the solvent was recovered by fractional distillation and recycled back for a second round of pretreatment. 25 The following section emphasizes the use of novel solvents for complete LCB fractionation. Chen et al. used 1% H 2 SO 4 with 75% choline chloride to fractionate cellulose of switchgrass from lignin and xylan fractions. Treatment with this acidified deep eutectic solvent (DES) at 120 °C for 25 min removed 76% of the xylan fraction along with 51.1% delignification. Five cycles of recycling and reuse of this acidified liquor enriched the hydrolysed xylan and lignin fraction. Later, the group used xylose-rich liquor for furfural production at 160 °C for 15 min with 2% w/v AlCl 3 and recovered lignin. 26 Very recently, a biphasic acidic water/phenol system was used for the fractionation of Populus wood chips. 27 This unique biphasic system enriched the water-soluble phase with 77% xylose and negligible by-products when the chips were subjected to 120 °C for an hour. In contrast, the phenolic phase contained 90% dissolved lignin (90%), leaving solids retaining 96% of the original cellulosic fraction. 27 Likewise, a novel biphasic system comprising 2-phenoxyethanol and acidified water (70 : 30) was used to fractionate rice straw. 28 Pretreatment at 130 °C for 2 h led to cellulose-rich (86.48% retained) biomass, facilitated by 92.1 and 63.16% removal of hemicellulose and lignin fractions, respectively. Later, 92.6% pure lignin was recovered by simple precipitation and 81.83% of xylan/xylose enriched in the aqueous phase. 28 Yang et al. evaluated the effect of p -toluenesulfonic acid ( p -TsOH) on the fractionation of three feedstocks: corncobs, wheat straw, and miscanthus. Pretreatment at 80 °C for 10 min resulted in significant removal of lignin and xylan, leaving a cellulose-rich pulp. Later, spent liquor was diluted to precipitate lignin, and the reusability of p -TsOH was shown to be ∼5 times higher. 29 A similar attempt was made by yet another green hydrotrope, maleic acid (MA), for the effective fractionation of birchwood. 30 At 100 °C and 50 wt% MA, 94.5% of the cellulosic fraction was obtained as a solid after 30 min. Lignin was precipitated by dilution, and the solubilised xylan was converted to furfural with ∼70% yield. Furthermore, MA displayed ∼3 times recyclability with comparable performance. 30 Earlier, the cosolvent enhanced lignocellulose fractionation (CELF) method was developed for pre-treating corn stover using 0.5% H 2 SO 4 and tetrahydrofuran (THF) in the ratio of 1 : 1. The dilute acid hydrolysed xylan to xylose which later dehydrated to furfural, while THF led to lignin dissolution enriching the cellulosic biomass. Later, the group separated furfural from THF. The latter was recovered by vacuum distillation and recycled, leaving lignin as a powder. 31"
} | 5,803 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.